Potential and innovation of the selective laser melting technique in the manufacture of platinum jewelery

One of the revolutionary characteristics of selective laser melting is the possibility of processing materials that are notoriously difficult to use with other production techniques due to various melting mechanisms and modelling the alloys used in direct 3D metal printing. In the field of precious alloys,
this aspect is particularly interesting in the manufacture of platinum-based jewellery, a notoriously problematic element in the production phase due to its properties which include a high melting temperature of the alloys available on the market today and high reactivity with the materials it comes
into contact with. Consequently, production costs are higher, specific furnaces need to be used and, on average, the items produced often have defects. An assessment of the real revolutionary scope of using the SLM technique in making platinum jewellery was carried out by comparing SLM and traditional
production techniques, not only considering the technical aspects but also, and above all, the impacting economic/financial implications on the production line, in order to understand whether, and to what extent, introducing the SLM technique would lead to improvements in platinum jewellery manufacturing.

INTRODUCTION

In the article presented at the Santa Fè Symposium 2017 (1), we dealt with a general comparison between precious metal micro-casting and direct 3D printing processes (SLM) to understand whether this latter technique was effectively more advantageous than both classic and direct casting.

Among the production cases in which selective laser melting proved to be better, we identified the production of small series, the creation of hollow jewellery or items with complex geometries and no welding, or, when using difficult materials or those that are impossible to use in micro-casting.

Platinum jewellery production could be included in these cases since the casting of this material is famously more difficult than jewellery in gold or silver alloy (3). Moreover, the platinum jewellery market, despite renewed interest in this precious metal in the last 20 years, still has considerably
lower demand than gold or silver jewellery, therefore the machinery is not used to its full productive capacity.

In order to analyse whether, and to what extent, the SLM technique is competitive in relation to micro-casting in platinum jewellery production, we conducted a true-to-life production comparison between the two techniques in collaboration with Progol3D®️, a top reference for selective laser melting,
and Stilnovo S.r.l., a company in San Salvatore Monferrato (Valenza Po’ jewellery district), the producer of OEM jewellery and a reference for platinum casting.

The market segment chosen for comparing the two techniques was wedding and engagement rings since those are the most representative for platinum in the USA and Europe at the moment. The idea of eternal that has always been associated to platinum due to its resistance over time, means that this metal
is particularly in demand for wedding bands. This can be seen, for example, from data regarding the USA market in 2016 where, compared to a 10% drop in the production of platinum jewellery the year before, American spending on platinum engagement rings grew by 5% (2), making this segment even more predominant
than in the past.

WHAT HAS CHANGED COMPARED TO GOLD JEWELLERY PRODUCTION

Producing platinum jewellery with traditional methods notoriously involves more problems than producing items in gold, especially in regard to micro-casting. On the other hand, producing platinum jewellery using the Selective Laser Melting (SLM™) technique is not particularly any more difficult than
making items in gold alloy, which makes this technique interesting for platinum jewellery manufacture. In general, it can be seen that, for metals used in jewellery-making, the more difficult it is to produce with micro-casting, the easier it is to 3D print.

The greatest production difficulty found in micro-casting comes from differences in the thermo-physical properties of platinum alloys compared to gold alloys. First of all, the considerably higher temperature of liquid platinum alloys leads to the use of different heat-resistant materials for making
the moulds to withstand the greater temperatures. Instead of traditional materials based on calcium sulphate and cristobalite, other materials more resistant to high temperatures are required in which silica is blended with phosphate-based binding agents, which requires longer and more strenuous preparation
(4). Coating properties generally vary more drastically compared to traditional heat-resistant materials in the case of imperfect mixing, both in relation to components and processing times, making these materials more sensitive to storing and aging conditions, causing oscillations that are hard to
control in the surface quality and in the mechanical resistance of the cylinders (5).

Even if specific materials are used, cylinder resistance is still critical if they are heated to more than 900°C (6), a limit that leads to a much greater difference in temperature between molten metal and the cylinder during the casting phase compared to gold alloys with the metal consequently losing
heat faster on entry. This effect, together with viscosity and the surface tension of platinum alloys being greater than that of gold alloys, make it more difficult to fill the moulds completely, especially in the narrower areas, and it is therefore necessary to use centrifugal casting machines to partially
alleviate the problem (7). Increasing the centrifugal force helps the metal to fill the mould, but it also increases the chances of refractory fragment detachments which could then be incorporated into the metal when cooling. All these problems limit the quantity of metal that can be used on each tree
to a much lesser amount than can be used with gold or silver alloys, with the consequent reduction in production capacity. Filling problems and greater shrinkage in going from a liquid to a solid state (7) also mean that a more robust feeding system is needed, which, in turn, leads to a more unfavourable
ratio between scrap and pieces produced. A greater quantity of scrap implies higher production costs, which are further increased by the greater cost of refining platinum alloys compared to gold alloys, due to the more complex procedure and verification process. All these additional difficulties make
casting platinum jewellery more susceptible to variable results not to mention the need for more specifically skilled technicians.

The SLM™ process, on the other hand, has no particular problems compared to gold alloy production. In fact, the fundamental properties for metal-laser interaction, first and foremost, reflectivity and thermal conductibility, are more favourable for platinum alloys than for gold or silver alloys. This
means that less energy is required for laser melting and there is no need to add elements to the alloy to favour laser radiation absorption.

QUALITY COMPARISON

The quality comparison between platinum jewellery created through SLM™ and micro-casting was carried out by producing several ring models in Stilnovo’s BRIDAL series, a collection that best incorporates the concept of eternal associated to platinum jewellery since it comprises rings with the MULTISIZE
solution, covered by a patent (Application number 102017000104245, filed on 18 th September 2017).

The multisize patent is a system that leads to a new conception of the ring as an item that can easily change its diameter and therefore always be a perfect fit.

Changing the size of a ring has always been quite a problem for the jeweller as well as for the final wearer. Because a ring is long-lasting, and is sometimes even handed down from mother to daughter, it is quite possible that the need to change the size will arise sooner or later.

The operation is easy enough if the ring only has one mounted ring shank and a centre, but it becomes more and more complicated as shapes develop and is absolutely difficult when the whole shank is mounted: size modification, i.e. making the diameter smaller or larger is, in fact, obtained traditionally
by cutting the shank at the opposite side to the centre and adding or removing some of the metal. When the ring is mounted along the entire surface, it is dangerous to enlarge or tighten it by even just one size because the mounting may become insecure: in fact, changing the curve of the ring inevitably
modifies the diamond or precious stone setting, something with can compromise the reliability of the ring’s “hold” on them.

With the MULTISIZE RING solution, the internal part of the ring shank has a slot for an interchangeable B structure in various thicknesses (Figures 1, 2 and 3).

In our study, the A frames were made in platinum, while it was decided to use titanium for the sheet metal.

A simple KEY, a titanium hook, made into the shape of a treble clef, was used to pull out the interchangeable part from its slot in the fixed part when the size needed to be modified. Once the slot track in the fixed part A is empty, it is easy to position a new interchangeable part by hand and change
the size.

In order to compare micro-casting and direct metal printing, 10 models from the BRIDAL collection were chosen, comprising wedding bands, solitaires and trilogies, whose frames are shown in Figures 4 to 13. Production and the characteristics of the internal interchangeable parts were not taken into consideration
in this study since they were not made in platinum alloy but were preferentially made in gold or titanium due to the mechanical properties needed for the piece of the frame to be repeatedly inserted and removed without becoming deformed.

A wedding band model, called ETERNAL, which features 360° pavè (Figure 14), was initially chosen for the comparison but was later discarded due to the difficulty of removing the support required in SLM™ production.

6 rings of each model were made for both production technique in question of which 2 were to be sacrificed for destruction analysis, with the exception of the two wedding band models, of which three men’s size and three women’s size samples were made. The overall total of rings made for the study was
120 pieces of which 40 were to be sacrificed for destruction analysis. The list of pieces produced is summarized in Table 1.

In order to make the comparison more like a real production test, jewellery creation was divided between two producers: Stilnovo for micro-casting and Progol3D® for selective laser melting.  Each of the two producers is specialized in one of the two techniques being tested and is able to optimize the
process to obtain the best possible quality.

To assess quality differences given exclusively from the type of productive course and not ascribable to the different composition of the alloys used, the alloy 95PtGaInCu was used in both SLM™ and in micro-casting. Using the same composition for micro-casting and SLM™ made it possible not to give one
technique an advantage over the other thanks to the relative ease with which platinum can be melted by laser interaction so that no adjustments to the composition were needed for the SLM™ process, which would have been required had gold alloys been used. In fact, this composition is in the Progold range
as a micro-casting alloy and is also in use for SLM production at Progol3D®.

In regard to micro-casting production, the waxes were created with a Projet MJP 2500W 3D printing system using VisiJet M2 Cast wax. The cylinders were prepared with PRO HT Platinum Gold Star® keeping a water/plaster ratio of 33:100. The refractory firing cycle is outlined in Figure 15. Cylinder temperature
during casting was 850°C.

The cylinder plastering and firing phases were grouped as much as possible averaging between minimizing production times and the need to retrieve scrap.

For the melting and cylinder firing process, a Yasui VCC centrifugal casting machine was used with a casting temperature of 250°C above the alloy’s liquid state. After the cylinders were cooled, the refractory residues were removed from the metal by immersing them in hydrofluoric acid at room temperature.

A final sanding was carried out to complete refractory elimination.

As for selective laser melting, the jewellery was produced using a ReaLizer SLM50 laser printer equipped with a 100W fibre laser, collimated to a ray of 10 μm. The circular construction plate was 70 mm in diameter.

The layer thickness used for printing was 20 μm, favouring printing resolution over production speed, in consideration of the market segment involved in the study.

The printer was fed with 95PtGaInCu in powder form, obtained by gas atomization of the alloy and sieving to remove any coarse particles.

The shape of the powder particles was observed under a scanning electron microscope (SEM) and the particle size distribution was determined using a laser granulometer (Malvern, Hydro 2000S).

After the printing phase, the jewellery underwent shot peening to eliminate any partially melted powder on the surfaces which would cause the unrefined pieces to be rougher.

Both in micro-casting and direct metal printing, all the rings were re-fired to solubilize the alloy and reduce the internal tensions by furnace treatment at 1150°C for one hour, followed by rapid cooling in water.  In the case of wedding bands, the pieces were later hardened by furnace treatment at
650°C for an hour with slow cooling.

Whatever the production technique, every ring made was assessed using the following quality standards:

–          Surface appearance “as cast” or “as print”, impact of feed residues and supports.

–          Identification of any macroscopic non-conformity defects.

– Measurement of the internal diameter of the rings, variations to the nominal and deviations in measurements between rings of the same model.

On the two sacrificial samples for each model, the following was also carried out:

–          Measurement of surface roughness in both “as cast” or “as print” and after sanding or shot peening

–          Assessment of the internal quality by trimming and lapping the rings.

All the produced items that were not used for destruction analysis (altogether 40 micro-cast rings and 40 printed rings, subdivided into 10 models) were then polished and eventually mounted at Stilnovo for a final evaluation of the quality. The final quality assessment on the completed item was made
by Stilnovo’s internal quality control department which was not aware of the type of production technique used for each ring to be assessed. The standards normally adopted for high jewellery article control were applied.

At the same time, fundamental data were registered to compare the technological and economical aspects of micro-casting and direct metal printing, such as:

–          Production times

–          Production scrap

–          Technician impressions in the polishing phases

–          Technician impressions on the mounting

In order to correctly collect the data on finishing operations, an evaluation sheet, subdivided by phase, was attached to each ring and each technician was asked to complete it.

EVALUATION OF THE PHYSICAL, MECHANICAL AND TECHNOLOGICAL  CHARACTERISTICS

Surface appearance

The first comparison made between rings produced by micro-casting and by SLM™️ involved the appearance of the surfaces both when raw and after sanding or shot peening. This included assessing the impact on the surfaces of additional elements needed to create the item, in other words, feeders in
the case of micro-casting and supports in the case of SLM™️. The invasiveness and weight of these elements had direct repercussions on the quality of the rings due, for example, to the need to reconstruct the surfaces involved, and economically, because of being directly proportional to the percentage
of production scrap and process times.

This paragraph will evaluate the presence and invasiveness of feeders and supports in terms of the surface extension involved and the residue aspect, while the paragraph on the economic and financial repercussions will report the findings regarding scrap and production times.

Figures 18 to 25 compare the feeding and support systems of the 10 models selected for production.

From the comparison of the additional elements needed in production with the two techniques being examined, it was immediately obvious how the effect on the surfaces was completely different in the two cases. In micro-casting, where additional elements are considerable, the geometry of the directly fed
area of the jewellery item was totally lost, while in SLM™, the geometries below the residues of the supports, built as a grid, were generally visible.

Examples of support and feeding residues on the rings can be seen in Figures 26 and 27.

Support in SLM™ generally involves a greater surface area of the item, but, if the effective area of contact with the supports is taken into consideration, that is, the areas where the grid teeth actually touch the item and spoil the surface, the values are lower compared to the areas affected by feeding
in micro-casting.

There are cases, however, as in the example of the ETERNAL wedding band, in which, although the maximum geometry of the ring is maintained, the loss of detail due to the massive presence of support residue, makes production by selective laser melting, unsuitable.

For the solitaire 4 and trilogy 1 models, a good compromise was obtained in SLM™ by using a growth orientation that minimized the corner surfaces so that support was required, but with some supports in areas more difficult to reach than for other models when it came to removing them (Figures 28 and 29).
In these cases, a favourable use of support parameters leads to creating elements that are easier to detach thus partially compensating for the greater dexterity required for their removal.

In regard to the overall appearance of the surfaces, micro-cast rings were generally less rough in the raw state (example in Figures 30 and 31) and after surface treatment (Figures 32 and 33). However, surface irregularities were often observed, mainly between all the excess material burrs, which did
not appear with SLM™. These defects will be analysed in more detail in later paragraphs.

Roughness

To provide a quantitative evaluation of the differences between the surfaces, roughness measurements were carried out using a Taylor Hobson FTS INTRA 02 profilometer. The total roughness (Rt) of the profile was chosen as a comparison parameter corresponding to the difference between the highest and lowest
surface points. This value, in fact, represents the thickness of the precious material that must be removed in polishing to obtain an aesthetically satisfactory surface. The values were registered both for the items when raw (“as cast” in the case of micro-casting and “as print” in the case of SLM™)
and after surface sanding or shot peening.

Indeed, raw surface treatment is a production practice at both Progol3D®, by shot peening to reduce roughness and homogenize the surface appearance, and Stilnovo by sanding, mainly to eliminate refractory residues. The actual roughness that the jewellery-maker will come across in the roughing stages
is, in both cases, that of the treated item, and it will be from these values that the quantity of the material to be removed in order to obtain a smooth surface will depend.

Measurements were taken on several areas of the jewellery items corresponding to surfaces with various orientations in respect of the growth direction of SLM pieces and wax growth in micro-casting. Points with no obvious surface defects were measured in order to give an average Rt value net of macroscopic
surface irregularities.

In regard to wedding bands, the growth directions selected for 3D printing, for waxes in micro-casting and metal in SLM™ were the same, shown in Figure 34. Measurements were taken in direction 1 (surface parallel to the growth, direction perpendicular to z), in direction 2 (surface parallel to the growth,
direction parallel to z) and in direction 3 (surface perpendicular to the growth, direction perpendicular to z).

On the other hand, the solitaires and trilogies were printed with a different “standing” in SLM™ than the waxes in micro-casting due to the different type of support used. To be precise, growth was carried out with the pieces directed vertically in SLM™ and horizontally for waxes. In this case, measurement
directions were decided according to the growth direction, as shown in Figure 35 for SLM™ rings and in Figure 36 for micro-cast rings. The direction indicated with 4 therefore corresponds to a parallel surface in the growth direction, with the measurement taken perpendicularly to z, while 5 refers to
a variable surface due to inclination with measurement along z.

Table 2 shows the average values recorded on raw pieces divided by direction with the respective standard deviations, while Table 3 shows the values for items after sanding or shot peening.

The results are summarized in the graph in Figure 37.

As already noted in observing the raw surfaces, the roughness values are clearly greater in SLM™ compared to micro-casting. This is not a surprising result since surface roughness is one of the weak points of the SLM™ technique. In SLM™, the roughness registered is also greater on average than can usually
be found for gold alloys, a finding in line with the values reported in a study conducted by Progold® in 2015 (8), which noted how, compared to gold alloys, the greater presence of partially melted powder particles on the surfaces, leads to higher roughness on the raw item (Figure 38).

The higher degree of roughness registered in SLM™ in direction 3 compared to the other measurements is attributable to the surface progress caused by compounding the melting lines, which give a meniscus effect with greater height in the centre of the line and less height at the edges (Figure 39). In
the wax mould, the meniscus effect is much less pronounced (Figure 40), so much so that the roughness resulting from this effect, measured in direction 3, is much less than that caused by the subdivision into layers on the long z piece, which is the main cause of roughness in other directions.

The standard deviation recorded in SLM™ compared to micro-casting derives from an already higher deviation between different points of the same jewellery item belonging to equivalent areas.  These differences are mainly due to the different surface direction measured in respect of the movement that the
“wiper” makes during platform “recoating” (8), which results in the powder particles adhering differently to the surfaces.

The roughness on the items produced with micro-casting was, however, more constant both on individual items and in consideration of the various models.

Lastly, the effect of surface treatment, whether sanding or shot peening, on the roughness of the pieces in both production methods reduced the roughness values by about half compared to the “as cast” or “as print” condition.

The overall less surface roughness found in micro-casting generally implies that the jewellery-maker will have to remove less material in the roughing stage in order to achieve a compact surface. This is only true, however, if the piece has no areas with excess material, such as burrs, or spaces, like
surface dents. In these cases, the material lost and the processing time can increase considerably.

Defectology

Micro-casting

As already mentioned above, the jewellery items produced by micro-casting had a clearly higher incidence of macroscopic defects than those produced by SLM™, even after casting parameters were optimized.

The most commonly found defects were surface irregularities, such as burrs due to excess or lack of material.

In the first case (Figure 41), the cause was the partial rupture of the refractory so that cracks formed where they had filled with metal. This type of defect is generally very simple to correct since the excess material can easily be removed in a short time.

In some models, however, such as trilogy 1, the presence of details separated by tiny spaces, made this type of defect more critical, with cases like the one in Figure 42 where refractory rupture had caused different areas of the item to join up.

Phosphate-based refractory resistance variability, resulting from greater susceptibility to variations in storage conditions and the high temperatures of the casting metal, was the most probable source of other types of defect found.

The detachment of tiny portions of refractory led to the appearance of irregularities in some pieces in the form of cavities in cases where these micro-detachments became trapped in the metal (Figures 43 and 44), or of various-sized hollows whenever the micro-detachments were external to the metal and
created tiny round craters on the edges (Figure 45).

The high temperature of the metal, which causes reactions in the refractory, was probably responsible for the irregular surfaces and porosity found in some areas of the micro-cast jewellery items, like those in Figure 46 and in Figure 47, where the roughness was considerably greater than the average
of the surrounding areas.

In other items, surface defects seemed to have been caused by a combination of micro-detachments and refractory reaction (Figures 48 and 49).

The defects shown in Figures 43 to 49 were more damaging for the item compared to the previous since the problem was the lack of material rather than material in excess. In fact, this would have forced the technician to remove more material in order to obtain an even surface or to carry out repairs if
the cavities were deep, with a consequent greater loss of material and longer processing times.

Besides defects ascribable to metal-refractory interaction, problems were also found that derived from other production phases.

For example, the ovalling found in one of the micro-cast model 8 solitaires (Figure 50) was due to probable tension in the waxes or to problems in the plaster casting stage. Although deformed, in these cases, the jewellery-maker can quickly intervene to put the ring back into its original shape, practically
without altering the size so that this defect is of no particular consequence.

Another defect found was bent grips in the models where the grips were particularly long, especially in the model 4 solitaire. This problem (Figure 51), due in all probability to bending the waxes during plastering, can be resolved by adding a terminal ring to stop the grips from moving (Figure 52).

The rupture shown in Figure 53, on the other hand, was attributable to the mechanical stress that occurred in the cylinder cooling stage. In this case the ring was obviously non-compliant.

In order to further investigate the causes of rupture, the wedding band was sectioned horizontally and analysed under an electronic microscope.

In the internal part of the ring, where the fracture occurred, a cavity was found which was most likely due to refractory inclusion, given the results of the EDX analysis of the internal residues which highlighted the presence of silicone.

The cavity, which extended to both halves of the sectioned wedding band (Figures 54 and 55), had reduced the effective section of the ring thus drastically lowering the mechanical resistance, therefore the stress caused by the shrinkage in cooling exceeded the ultimate tensile strength causing the ring
to fracture.

SLM™

The macroscopic defects observed in the jewellery items produced with SLM™ were clearly fewer than those found with micro-casting. In fact, while the surfaces had a higher degree of roughness, only in the case of one ring produced was a real irregularity found in the form of swelling in one area of the
piece (Figures 56 and 57).

This type of defect occurs in SLM™ when the powder does not melt perfectly and so some partially non-melted particles remain and disrupt powder distribution in subsequent printing layers.

In this case in particular, since the defect only involved a small part of the item’s upper area, incomplete melting was probably caused by a variation in the average granulometry of the powder in the growth zone, due, for example, to the accumulation of agglomerates of partially molten particles within
the powder distributed by the “wipers” as the printing process continued.

Since the problem was material in excess and not a lack of it, correcting this type of defect was of no particular importance. However, it may happen that the swelling can be associated to widespread porosity in the area concerned, again caused by imperfect melting.

Dimensional coherence

An analysis of the correct nominal size and the deviations that could be found between the various same model rings was carried out on all the items produced by measuring the internal diameter, which can be directly correlated to the actual size of the ring.

For greater precision, the diameters were measured using a calibre (Mitutoyo), averaging three values in different positions, and also by means of photographic analysis using a Keyence digital microscope, suitably calibrated for maximum measurement accuracy.

Table 4 shows the data relating to the internal diameter of the rings. In the averages calculated for micro-casting, the ovalized ring in Figure 50 was not considered due to the difficulty of establishing the real diameter.

From the data obtained, it can be deduced that the internal diameter measurement in relation to the nominal value was always less in SLM™ compared to micro-casting for each of the ring models produced.

The origin of this reduction in internal diameter is obviously different for the two techniques: in SLM, it is caused by an imperfect correction of the width of the single laser trace while in micro-casting, it is caused by refractory shrinkage during the firing stage, from metal shrinkage as it goes
from liquid to solid and from the item’s contraction during cooling at room temperature.

In the case of SLM™, using platinum instead of gold was not a variable that could have affected dimensional variations, while in micro-casting, the higher temperatures and more notable shrinkage during phase change could have been cause for greater discrepancy in the nominal size for platinum rings rather
than gold rings. Repeatability on rings of the same model was generally greater in SLM™, with maximum standard deviations of ± 0.03 mm compared to ± 0.04 mm and more found in some micro-cast models. Given the greater oscillation found in micro-casting, any correction upstream of the internal dimensions,
by altering the design, for example, would be less effective.

Internal porosity

To analyse the porosity inside the items, the first technique considered was computerized tomography, a technique that has the advantage of not being destructive and able to investigate the entire volume of the jewellery item. However, the results obtained were not deemed satisfactory in terms of image
resolution, a problem caused by the high density of platinum which caused such elevated absorption of the beam, that analysing the thickness of the rings was extremely imprecise.

As an alternative to tomography, it was decided to make a direct analysis of the ring sections by cutting two out of the six rings produced for each model. In order to acquire a more complete evaluation of the internal volumes of the rings, sections from different areas of the items were analysed. To
be precise, one ring of the sacrificial pair was sectioned in the A plane represented in Figure 60, while the other was sectioned in the B planes (Figure 61), perpendicular to the first, in four different areas of the ring. After resin incorporation and lapping, the sections were photographed at 50X
to digitally analyse porosity using the software inside the Keyence microscope that had been used to take the images.

Figure 81, while Table 5 shows the percentage porosity values found in the various models considering porosity on the A and B planes of each ring, weighed along the entire surface of each analysed section.

The level of porosity found in the pieces can be quantified as medium-low in both production techniques with lower values in SLM™ compared to micro-casting, which, on average, had a twice as high porosity. For both cases, there was a notable variability between different pieces and between different
areas of the same sample, with sections that had practically total density and others that suffered from higher porosity.

In micro-casting, areas with singular high-volume porosity were observed, like, for example, the cavities in Figure 82, as well as a high number of small porosity clouds, as in the case of shrinkage porosity shown in Figure 83.

The porosity found in SLM™ was not in the form of cavities but single spherical pores (Figure 84), probably caused by gas, or areas with tiny, regularly placed spaces (Figure 85) due to imperfect melting between adjacent laser tracks.

Besides percentage porosity on all the items, locating any pores is also extremely important in jewellery-making: pieces with a dense interior but surface porosity are more difficult to finish than those that are more porous overall but have a more compact surface.

From this point of view, it can be seen how the porosity found in some areas of the SLM™ items was mainly inside the pieces and more rarely on the surface areas. This effect derives directly from the melting modality and item growth. Inside one single “layer”, the external surface is, in face, melted
as one single outer layer and the laser parameters are optimized in order to ensure the almost total absence of porosity in each laser track. The inside is then melted with parallel laser scans. Porosity tends to gather at the joints between the internal scans or between the outer and inner layers,
which are generally at least 150-200 μm from the surfaces, in an area that is unlikely to be removed in the polishing phase. In micro-casting, porosity distribution is more varied: there are surface cavities, visible also macroscopically on the external surfaces and mainly due to tiny refractory
detachments, and shrinkage porosity, which appears to concentrate more inside the items.

Lastly, the case of fracture found in one of the micro-cast wedding bands is to be considered: in this case, porosity, although concentrated inside the ring, was so extensive that it compromised the item’s mechanical endurance.

Metallographic appearance

To evaluate the dimension of the crystalline granules in the micro-cast rings and in those printed with SLM™, acid attacks were carried out on model 1 “as cast” and “as print” wedding bands.

Comparison confirmed what had already been seen in the past for samples in gold and platinum alloys: the average size of crystalline granules was drastically greater for micro-cast items (Figures 86 and 87) compared to SLM™ items (Figures 88 and 89). Therefore, in reality, signs of melting traces could
be made out but not of individual granules, even at high magnification.

The SLM™ sample showed the presence of micro-cracks, made visible by the acid (Figure 90). Mechanical tests, reported in the paragraph below, were carried out also to evaluate the effective impact of this defect on the properties of the SLM™ pieces.

Mechanical characteristics

The mechanical characteristics of jewellery, such as hardness, elongation and ultimate tensile strength, have direct repercussions not only on the item’s mechanical resistance but also on technological parameters, such as mounting and polishing.

For this reason, the mechanical performances of the items produced by micro-casting and SLM™ were compared, the alloy used being equal. Micro-hardness tests were carried out on “as print” or “as cast” model 1 wedding bands, both after re-firing (1 hour at 1150°C) and after hardening (1 hour at 650°C),
using a Vickers FUTURE-TECH hardness tester. The ultimate tensile strength (UTS) and elongation (A %) values were, on the other hand, obtained from traction tests carried out with an INSTRON dynamometer on specifically created specimens, shown in Figure… In this case the values were measures on “as
cast” or “as print” specimens and on specimens subjected to re-firing treatment, to evaluate possible mechanical differences that may affect the mounting phase.

The greater hardness found for “as print” items compared to “as cast” ones was more than likely due to the smaller dimensions of the crystalline granule in SLM™ and to higher internal tensions in printed items. The re-firing treatment, which, in the case of the alloy used, had the double effect of lowering
the samples’ internal tensions and of solubilizing, meant that, in both cases, hardness could be lowered to below 190 HV, therefore making mounting possible. After aging, in both cases, hardness increased considerably, although it was greater for micro-cast pieces, for which resistance to wear and tear
could therefore be greater than for SLM™ pieces. The difference observed could be caused by the presence of the micro-cracks seen in the SLM™ wedding bands when tested with acid, which favours indenter penetration into the sample.

In regard to traction tests, the samples printed by SLM™ had a greater ultimate tensile strength in the “as print” state than the “as cast” samples, to the expense, however, of lower ductility.

After the re-firing treatment, the ultimate tensile strength lowered for both types of sample but was still higher in the case of SLM™. The results of elongation at fracture, on the other hand, show an inversion between SLM™ and micro-casting. In fact, although ductility increased with thermal treatment
in both cases, the increase in SLM™ was considerably greater.

After re-firing, the samples produced by SLM™ therefore had higher UTS and elongation at fracture values, a fact that indicates how the micro-cracks observed after metallographic attack on printed samples can probably be attributed to mechanical properties rather than granule size and to possible internal
defects in the micro-cast samples.  Better performances after re-firing suggest a better behaviour of items during mounting.

Finishing: technician impressions

The impressions of sector technicians play a fundamental role in the possible success of a new production technique. Jewellery production is not exempt from this rule: even though the quality of a product may seem to be excellent according to technical analysis, if, during the processing phases, it does
not ‘convince’ the workers, with all probability, the production technique will not be adopted in the future. For this reason, we considered it essential to collect the opinions of the “jewellery makers” involved in item finishing in order to be able to add more subjective, but equally important to
the overall assessment of a new production method, evaluations to the quantitative ones, like times and losses due to finishing. The 80 rings created and not subject to destructive tests, were therefore sent for finishing and evaluation. Each of the working phases was carried out by the same technician
both for items created by micro-casting and by SLM™ in order to have the same judgment gauge for both techniques.

The first phase of the finishing process is eliminating any added element residues from the rings that were used in their production but which are not part of the item, i.e. feeders in micro-casting and supports in selective laser melting. As can be seen by the opinions expressed, summarized in Figure
87, from this point of view, the SLM™ technique suffered from some models having supports that were more complicated to remove, like those used for the model 4 solitaire in Figure 28 and in the model 1 trilogy in Figure 29. Removing supports from the internal areas of the item required greater dexterity
on the part of the technician and increased the probability of the item being spoiled in this phase. Substantially similar results were observed in evaluating the difficulty of roughing (Figure 90). In fact, opinions generally regarding the roughness and compactness of the item’s surface revealed that
rings produced by SLM were 80% less difficult to work than those made by micro-casting. The only significant difference was the presence of a non-compliant micro-cast ring identified in this phase, which had caused particular roughing difficulties before it was definitively put aside.  The results obtained
with this evaluation are particularly interesting if one considers that one of the weak points of selective laser melting is high surface roughness.

The impressions reported by technicians on this point revealed that, at least in the case of platinum, all that was needed was slightly greater pressure or thicker paper during this preliminary phase to eliminate the additional surface roughness with little additional effort compared to micro-cast pieces.

This extra effort was, however, amply compensated by the quality of the SLM metal that the technician found (Figure 94): the percentage of surfaces evaluated as excellent in SLM™ in terms of compactness was, in fact, close to 100%, while in micro-casting, evaluations were more varied with only about
60% of surfaces considered as excellent, 25% as average, 10% as low quality due to evident porosity, and two rings were judged as non-compliant.

No particular difference was observed in the cleaning phases (Figure 95), while in mounting, the overall evaluations in both cases went from low to none (Figure 96). The mechanical properties of the metal therefore resulted as more than good both for the micro-cast alloy and the printed alloy.

Quality Control: evaluation

The Quality Control judgment is fundamental for understanding if the jewellery items produced conform to residual porosity criteria and the aesthetics defined by high jewellery.    We therefore subdivided the rings into those that directly passed verification, those that needed quick laser repairs in
order to be compliant and those that were judged as non-repairable.

There was considerable diversity in the results obtained in SLM™ and in micro-casting: while three quarters of the printed rings immediately passed the checks, only half of those produced by micro-casting achieved the same result (Figure 97).

The judgments expressed by Quality Control confirmed the data obtained from analysing the macroscopic defects and internal porosity of the sacrificed samples: the items produced in platinum by SLM™ were immediately found to be less faulty than those produced by micro-casting.

In regard to non-compliances, no SLM™ item was found to be such compared to two items made through micro-casting and submitted for finishing. Furthermore, one sacrificial wedding band was found to be non-compliant due to breakage.

The final appearance of the ten ring models after mounting and polishing can be seen in Figures 98, 99 and 100 for micro-casting and in Figures 101, 102 and 103 for 3D printing.

ECONOMIC AND FINANCIAL ASSESSMENT

Semi-processed production times

Both micro-casting and selective laser melting production processes were organized to reflect the timing and subdivision into typical phases of real-life production. Micro-casting production was subdivided onto 11 trees, listed in Table 8. Cylinder firing cycles, which are the longest production phase
for micro-casting, were grouped in order to obtain the best compromise between production times and scrap recovery. In order to imitate what happens in real production, it was, in fact, decided to re-use the casting scrap, adding it to new alloy to compensate the percentage of material used for producing
authentic jewellery. This procedure is normally carried out to limit the quantity of precious metal needed, both because of the cost of the raw material itself and for the cost of refining the scrap. To be precise, new alloy was used for the first group of four trees while the scrap added to new alloy
to reach the weight of the tree to be cast, was used for the next three trees and the last four.

In SLM™, production was distributed over 7 printing plates (Table 9), created in descending order according to the height of the items to be produced. In fact, this optimizes the use of the powder by producing the highest items first for which the quantity of powder needed to fill the printing space
is greater.

The average times for each cylinder and the total times that the machinery was in use, subdivided by phase, are shown in Table 10 for micro-casting. Table 11, on the other hand, shows the average times for each printed plate and the total times that the machinery was in use, subdivided by phase, for
SLM™. The time it took the technicians to carry out production was also recorded. In fact, a higher total of man hours not only increases production costs, it also implies less possibility of having an automated production process.

From the data shown, it can be noted how machinery usage times are lower (-20%) in micro-casting compared to direct metal printing. In both cases, one production phase required a longer usage of machinery compared to the others: in micro-casting, firing the plaster casts took 55% of the total production
time, while in SLM™, the printing phase actually took 85% of the overall time. Both phases, however, do not require any intervention from technicians and only weigh on production costs in terms of machinery usage and electricity.

Looking at man hours however, the situation is the opposite: despite the longer machinery times, the SLM™ technique required less technician time than micro-casting (-20%), and is therefore more inclined towards automation.

Another important fact for evaluating a production technique is undoubtedly the total production time, considered as the actual time it took to create a jewellery lot. This timing takes into account the effective daily hours (8 hours subdivided into two groups of 4 with a 1-hour break in between), the
weekly working days (5) and which processes can continue during the night because no human supervision is required. Also taken into consideration are processes that can be done at the same time as well as any waiting times, such as the time needed for pickling in acid or for partially drying the cylinders,
which do not involve machinery or technicians, but are part of the production process all the same.

The time divisions for micro-casting and SLM production phases are shown in Tables 12 and 13 respectively, considering the production sequence actually used for making rings. This includes subdividing micro-casting into 3 groups of cylinders in order to be able to use less precious material by re-casting
scrap, and the creation of 3 SLM™ plates with the lowest last instead of inserting them with the others to save time, in order to be able to use less powder in the machine at the beginning.

The total production time was equal to 5 working days for micro-casting and 5.5 for SLM™, which was therefore slightly slower. It should, however, be considered that the operations done on the sixth day in SLM™ do not prevent starting the production of a second lot because they can be done at the same
time. This means that, in the case of consecutive lots, the production capacity of 60 rings, equal to those made for this study, with the subdivision between phases, can be considered the same in the two cases.

Finishing times

The overall finishing times are shown separately in terms of semi-processed production times because the finishing needed the same phases for the items made with micro-casting as those made with SLM™. The discriminating factor in this phase was therefore the ease with which residues from feeders and
supports could be removed and the quality of the items in terms of surface roughness and compactness and of residue porosity. In fact, generally speaking, external porous surfaces or particularly irregular surfaces force the technician to remove more material in order to reach more compact areas of
the item with the consequent increase in working times and finishing loss.

After analysing the times needed to remove feeder and support residues, it can be noted that the former, on average, was done faster given also the relative simplicity of the ring geometries in the fed areas. The average time required in this phase was also more homogenous in micro-casting, while in
SLM™, variability grew depending on the positioning of the supports with longer times required for models where technicians indicated residue removal as more complex on their work difficulty evaluation sheet.

However, observing roughing times, it appears that, except for a few cases, the rings produced with SLM™ required similar or even shorter working times than those produced by micro-casting. This fact agrees with the impressions on the complexity of this phase shown in Figure 93 which, on average, saw
SLM™ rings as easy to work as the micro-cast ones but with a better surface quality. Polishing phases did not reveal any substantial differences between the techniques even in terms of working times and the same can be said for mounting, with the exception of the model 4 solitaire, which recorded a
slightly longer time with micro-casting.

Repairs, needed to a greater extent on micro-cast items but which were of short duration, led to a slighter longer overall time for micro-cast jewellery.

Finishing loss

The material removed from the rings in the finishing phases has an immediate effect on production costs since it cannot be totally recovered by refining materials that come from processing. To calculate these costs, an average loss of 5% of precious metal from finishing was estimated. Table 16 shows
the average values for losses for each model and for each technique during all the finishing operations.

The overall losses were greater in SLM™ or in micro-casting, depending on the model being processed. Analysing the individual phases, it can however be noted that, in the feeder elimination phases, micro-casting had greater losses than SLM™, while in the roughing phases, selective laser melting always
lost more material. These results can be easily explained by the greater quantity of surface roughness in SLM™ in the latter case. The impact of the losses recorded in terms of production costs, assuming a loss of 5% in the polishing recovering phase, is summarized in Table 17.

Raw material production costs

For a correct evaluation of the final cost of the items, we also considered the differences in the raw material costs. In fact, the two production methods differ in terms of the price of the raw material and the number of refining actions required to make the same quantity of jewellery. In regard to
raw material costs, we estimated, by evaluating the market prices, a higher amount for buying ‘new’ raw materials of 0.3 €/g for the powder needed in the SLM™ process compared to the micro-casting alloy, due to the cost of atomization. The same cost difference was taken into consideration between the
breaking down and atomization of new raw material retrieved from refined platinum. In order to assess the impact of refining, the first calculation regarded the various yields of the two production processes in terms of the ratio between pieces produced and casting scrap. The weights recorded and percentage
production yields are listed in Table 18 for micro-casting and Table 19 for selective laser melting.

The different yields found in the two production processes had a direct repercussion on the refining needed in the two cases with the consequent effect on production costs.  The calculation of production costs due to refining was carried out assuming that:

the 60 rings produced for this study represented a typical production lot, equal to about 500g of raw jewellery. To make the 60 micro-cast rings for this study, production scrap was re-cast twice, starting from 1 Kg of the initial alloy. We considered that we would have to refine everything after one
production lot.

To consider a situation in SLM™ similar to that of micro-casting, we assumed that, also in this case, everything would need to be refined after having re-melted the scrap twice. Moreover, for this study, the printer was initially loaded with 2.8 kg of powder, a standard production condition.

Refining costs, both fixed and those that depend on the quantity of material, were calculated using the average of the prices applied by 6 different Italian market suppliers (Table 20).

Focusing on SLM™, the quantity of powder that was initially put into the printer to produce jewellery lots was 2.8 kg. When creating one single lot, no scrap needed to be re-melted and, at the end of the print, the quantity of powder still in the machine was about 2 Kg, the rest having been used to produce
the items (500g) and supports (300g approx.). The second lot was also created without re-melting while, in order to make a third production lot, the scrap (mainly from supports) needed to be re-melted and 1000g of new powder had to be added in order to fill the printing platform to cover the height
of the items to be printed. Re-melting the material twice was only required to produce a fifth lot and, only after the sixth lot did all the powder need to be refined. In order to start producing a seventh lot, 1000g of new powder had to be added to that made with refined material.

The data relating to the powder needed for production by SLM™ and the material to be refined are shown in Table 21 while Table 22 reports the relative costs.

As a comparison, the refining costs for 3 Kg of jewellery produced with micro-casting were calculated, bearing in mind that, after every 500g lot, about 0.5 Kg of scrap would need to be refined (Tables 23 and 24).

Despite the lower raw material costs, the cost per gram of jewellery produced was 7% higher in the case of micro-casting mainly due to the set amounts for each refining, which mainly corresponded to the costs for the concentration sample. These expenses were, of course, to be added to the hours of machinery
usage, man hours and energy consumptions in order to have a complete picture of the effective cost per gram of jewellery produced with the two techniques in question.

Total production costs of the rings

With the data presented in the previous paragraphs, including production times, production lots and yields, the final industrial costs for producing each individual model could be calculated. In order to do so and to make the comparison as close to reality as possible, the following assumptions were
made:

Production capacity was calculated for the two techniques based on the effective use of both systems, considering the ring lot made for this study as a quantity produced in a week.

We considered a total machinery amortization time of five years, taking into account the current average fiscal amortization period in Italy. We did not consider the potential duration of the systems in work hours because all the machinery will most probably become obsolete before the end of its lifecycle.

The costs linked to consumer materials were divided uniformly between the items made, calculating an average cost and not the specific cost of each item produced.

In the calculation hypothesis, we intentionally left out the physical space needed to carry out industrial activities which, in regard to 3D printing, are decidedly smaller than that taken up by a lost wax casting system. The same goes for the capital invested into the electrical and hydraulic systems
needed for micro-casting.

We also did not take into consideration any waste disposal costs (crucibles, plaster, acids, etc.)  deriving from the lost wax casting process.

We also lowered the benchmark by considering two hypothetical companies that exclusively produce items in platinum. This would mean lower exploitation of resources, which could be common to platinum, gold and silver processes.

Dividing machinery usage and human resource costs for each model was carried out based on the percentage weight of each ring produced in respect of the total cylinder cast or the print plate.

The hourly cost of the technicians was taken as the same for both SLM™ and micro-casting, and similar for each person involved in the production and finishing processes.

The consumer materials needed for SLM™ and micro-casting production are listed in Table 25.

The results of production cost calculations for each model were subdivided into semi-processed production costs, finishing costs (including losses of material during processing) and refining costs, shown respectively in Tables 26, 27 and 28.

What emerged from semi-processed production costs was the enormous impact of micro-casting system underuse which led to unfavourable amortization costs compared to the SLM™ technique. This resulted in a higher production cost for almost every ring model, with the exception of the wedding bands for men

System underuse is due to the widespread practice in many companies of internalizing the platinum casting phase mainly for strategic rather than economic reasons, preferring not to entrust the process to third parties. Moreover, the platinum jewellery segment is a precious metal jewellery niche with
production demand standing at about 60 times less than demand for gold, another element that contributes to not optimising the use of the machinery.

The overall finishing costs, however, showed a more varied trend with a general advantage for the SLM™ technique except for those items that involved greater difficulty in removing supports and in roughing.

In regard to refining costs, all the models resulted more favourable with SLM™ due to the higher cost per gram for micro-cast jewellery.

Lastly, looking at the overall costs (Table 29), SLM™ production was clearly less expensive compared to micro-casting in 5 solitaire models and for the ladies’ wedding band sizes, while trilogy 1 and wedding band 4 for men were less expensive in micro-casting. Finally, in three cases, the model 8 solitaire,
model 2 trilogy and wedding band 1 for men, the costs were practically identical, given that the differences found could easily be cancelled out by tiny variations in the production phases. It should also be underlined that, the additional cost linked to re-firing two non-compliant, micro-cast wedding
bands was not taken into consideration in the calculations, therefore only 57 out of the 60 micro-cast rings were truly suitable for sale compared to all 60 in SLM™.

Investment capital

The amount of investment capital required to start producing the semi-processed goods involved in this comparison, net of the necessary resources for the finishing phases, which are the same for micro-casting and SLM™, are slightly higher for SLM™. In fact, the greater quantity of machinery needed to
start micro-casting activities is only partially compensated by the high cost of a selective laser melting printer and by the need for more precious material, estimated as 2.8 Kg of powder against 1 Kg of alloy in micro-casting, to be able to print the jewellery items to their full height.

It is also true that, in the case of micro-casting, there is a better offer of machinery which could lead to a lower amount of investment capital, while in 3D printing, the investment capital calculated here is the minimum required for using this technique. Another difference is that the investment capital
for micro-casting is all instrumental while part of the capital in 3D printing is financial. This favours an SLM™ company in the case of liquidation since selling precious metal is easier than selling second-hand machinery.

It should be noted, however, that, as previously mentioned, we did not take into consideration the higher cost of the systems needed for the good working order of micro-casting machinery, which include a more complex electricity system, a hydraulic system that must be installed onto each machine with
chilled water and a vacuum system to convey the air and dispose of fumes during cylinder firing phases.

It should also be added that a lost wax casting system requires a working space of at least 50 m² which, at market price in Italy, would cost about 100,000 €, while the space need for 3D printing is potentially less than 1 m².

Environmental impact

The environmental impact is a standard that is becoming increasingly important in the overall assessment of a production process. For this comparative study, environmental impact was quantified for each production technique by calculating the Carbon Footprint (CF), which refers to the amount of greenhouse
gas (GHG) released during production, expressed in terms of CO2 equivalent mass.

The released GHG comparison was carried out considering all the phases and materials needed to complete jewellery production. Calculation of emissions caused by production and by disposing of the materials used was carried out using the data taken from the EcoInvent 2.2 database, while the data on greenhouse
gases deriving from the production of electricity to power the machinery were taken from the National System for Environmental Protection ((ISPRA), based on the production of electricity for the Italian network (9). Calculations did not include greenhouse gases caused by raw material extraction and
system and machinery construction.

It can be noted from the results how the total of greenhouse gases released into the environment during the production of 60 rings with SLM™ was half the amount generated with micro-casting. Greater electricity consumption, the gases released in plaster firing phases and the general use of materials
with a high environmental impact are the main cause of this result.

CONCLUSIONS

We can conclude from this study that, from a qualitative point of view, the jewellery produced with SLM™ is better both in terms of macro surface defects and internal porosity. This fact is confirmed by technician evaluation and the number of items that needed to be corrected by laser, as well as by
the lack of non-conformities compared to the three non-compliant wedding bands produced by micro-casting. The potential re-firing of a non-compliant piece is also more unfavourable in terms of time and costs compared to a hypothetical re-print.

Production times were slightly slower in SLM™, although the technique is more effectively exploited with small platinum production lots compared to a micro-casting system.  The greater production yield with selective laser melting also limits resorting to refining, with advantages on costs.

The overall costs were significantly in favour of the SLM™ technique for many of the models created, with only two evident advantageous cases in micro-casting. All this in the face of a slightly higher initial activity start-up investment and half the environmental impact.

In consideration of the data collected, we can conclude that, for companies that deal only in platinum productions and with weekly lots of about 500g of raw jewellery, the SLM™ technique is decidedly more advantageous than micro-casting, since it is more suitable for the small quantities of platinum
jewellery produced and has, on average, a better quality compared to the same items produced by micro-casting.

We can therefore confirm that, as our work presented at the Santa Fè Symposium 2017 hypothesized, platinum jewellery production can be included in cases in which the SLM™ technique is truly an added value compared to traditional casting.

References

Damiano Zito et al., “Why Should We Direct 3D Print Jewelry? A Comparison between Two Thoughts: Today and Tomorrow,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2017, ed. E. Bell et al. (Albuquerque: Met-Chem Research, Inc., 2017).

Teresa Fryé and Joerg Fischer-Buehner, “Platinum Alloys in the 21st Century: A Comparative Study,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2011, ed. E. Bell (Albuquerque: Met-Chem Research, Inc., 2011).

GFMS Platinum Group Metals Survey 2017, Thomson Reuters Eikon™.

G. Ainsley et al., “Platinum Investment Casting Alloys,”
Platinum Metals Review
22, no. 3 (London: Johnson Matthey & Co. Limited, July 1978): 78.

P. Lester et al., “The Effect of Different Investment Powders and Flask Temperatures on the Casting of Pt Alloys,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2002, ed. E. Bell (Albuquerque: Met-Chem Research, Inc., 2002): 321-334.

U.E. Klotz and T. Drago, “The Role of Process Parameters in Platinum Casting,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2010, ed. E. Bell (Albuquerque: Met-Chem Research, Inc., 2002): 287-326.

Damiano Zito et al., “Definition and Solidity of Gold and Platinum Jewelry Produced Using Selective Laser Melting (SLM™ ) Technology,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2015, ed. E. Bell et al. (Albuquerque: Met-Chem Research, Inc., 2015): 455-492.

Istituto Superiore per la Protezione Ambientale, “Fattori di Emissione Atmosferica di CO
₂ e Sviluppo delle Fonti rinnovabili nel settore elettrico”, (2017).

Figure


1


. Eternal model frame and interchangeable part

Figure


2


. “KEY” for replacing the spring

Figure


3


. Sequence for changing the interchangeable part on Trilogy

Figure


4


. Model 1 wedding band frame

Figure


5


. Model 4 wedding band frame

Figure


6


. Model 4 solitaire frame

Figure


7


. Model 5 solitaire frame

Figure


8


. Model 7 solitaire frame

Figure


9


. Model 8 solitaire frame

Figure


10


. Model 15 solitaire frame

Figure


11


. Model 16 solitaire frame

Figure


12


. Model 1 trilogy frame

Figure

13


. Model 2 trilogy frame

Figure

14


. ETERNAL model wedding band


frame

Figure


15


. plaster firing cycles

Figure


16


. Feeders and supports used to produce the model 1 wedding band

Figure


17


. Feeders and supports used to produce the model 4 wedding band

Figure

18


. Feeders and supports used to produce the model 4 solitaire ring

Figure


19


. Feeders and supports used to produce the model 5 solitaire ring

Figure 20 . Feeders and supports used to produce the model 7 solitaire ring

Figure 21 . Feeders and supports used to produce the model 8 solitaire ring

Figure 22 . Feeders and supports used to produce the model 15 solitaire ring

Figure 23 . Feeders and supports used to produce the model 16 solitaire ring

Figure 24 . Feeders and supports used to produce the model 1 trilogy ring

Figure 25 . Feeders and supports to produce the model 2 trilogy ring

Figure 26 . SLM™ support residue on the surface of a ring

Figure 27 . Feeder residue on the surface of a ring

Figure

28


. Internal support in the model 4 solitaire

Figure

29


. internal support in the model 2 trilogy

Figure 30 . Raw model 4 wedding band produced by micro-casting

Figure 31 . Raw model 4 wedding band produced by SLM™

Figure 32 . Model 4 wedding band produced by micro-casting, after sanding

Figure 33 . Model 4 wedding band produced by SLM ™, after shot peening

Figure 34 . Roughness measurement directions on wedding bands

Figure 35 . Roughness measurement directions on SLM™ solitaires and trilogies

Figure 36 . Roughness measurement directions on micro-cast solitaires and trilogies

Figure 37 . Average roughness found in the various directions in micro-casting and SLM ™

Figure 38 . Surface roughness on the vertical wall of a raw SLM™ wedding band, 300X

Figure 39 . Surface roughness on the horizontal wall of a raw SLM™ wedding band, 300X.

The parallel tracks left by the laser scans can be seen.

Figure 40 . Surface roughness on the horizontal wall of a raw micro-cast wedding band, 300X.

Figure 41 . Excess material on the side of a micro-cast ring

Figure 42 . Example of refractory breakage

Figure 43 . Cavity on the surface of a model 4 solitaire, probably caused by a refractory micro-detachment encased in the molten metal

Figure 44 . enlargment of the defect in Figure 43

Figure 45 . Subsidence probably caused by fragments of detached refractory stuck on the surface of the molten metal

Figure 46 .

Figure 47 . Details of the surface in Figure 46

Figure 48 . Surface porosity in a micro-cast solitaire

Figure 49 . Details of the area in Figure 48

Figure 50 . Micro-cast model 8 solitaire with evident ovalling

Figure 51 . Deformation of the grips of a micro-cast model 4 solitaire

Figure 52 . Added ring to stabilize the position of the grips on model 4 micro-cast solitaires

Figure 53 . Breakage in one of the micro-cast wedding bands

Figure 54 . Internal cavity in a micro-cast wedding band.

Note the correspondence with the item’s fracture zone.

Figure 55 . Extension of the cavity in the other half of the sectioned wedding band

Figure 56 . Surface swelling in a model 2 trilogy produced by SLM™

Figure 57 . Surface swelling in the model 2 trilogy in Figure 56 compared with the standard profile

Figure 58 . Example of digital measurement

Figure 59 . Offset compared to the nominal internal diameter measurement with standard deviation

Figure 60 . Section A plane

Figure 61 . Section B planes

Figur e 62. Micro-cast wedding band 1

Figure 62 . SLM™ wedding band 1

Figure 63 . Micro-cast wedding band 4

Figure 64 . SLM™ wedding band 4

Figure 65 . Micro-cast solitaire 4

Figure 66 .  SLM™ solitaire 4

Figure 67 . Micro-cast solitaire 5

Figure 68 . SLM™ solitaire 5

Figure 69 . Micro-cast solitaire 7

Figure 70 . SLM™ solitaire 7

Figure 71 . Micro-cast solitaire 8

Figure 72 . SLM™ solitaire 8

Figure 73 . Micro-cast solitaire 15

Figure 74 . SLM™ solitaire 15

Figure 75 . Micro-cast solitaire 16

Figure 76 . SLM™ solitaire 16

Figure 77 . Micro-cast trilogy 1

Figure 78 . SLM™ trilogy 1

Figure 80. Micro-cast trilogy 2

Figure 81. SLM™ trilogy 2

Figure 82. cavity in micro-cast ring section

Figure 83. porosity from shrinkage in micro-cast ring section

Figur e 84. Porosity from gas in SLM™ ring section

Figure 85. Porosity inter-hatches in SLM™ ring section

Figure 86. "as cast" micro-cast wedding band after metallographic attack, 50x

Figur e 87. “as cast” micro-cast wedding band after attack, 200x

Figure 88. "as print" SLM™ wedding band after metallographic attack, 50x

Figure 89.  "as print" SLM™ wedding band after attack, 200x

Figure e 90. Micro-cracks visible in the SLM wedding band after metallographic attack

Figure 91. Traction tester

Figure 92. Evaluation of the difficulty in removing supports/feeders

Figur e 93. Evaluation of roughing difficulties

Figure 94. Evaluation of the surface quality after roughing

Figure 95. Evaluation of polishing difficulties

Figure 96. Evaluation of mounting difficulties

Figure 97. Stilnovo s.r.l.’s internal Quality Control evaluation of the jewellery produced by SLM™ and by micro-casting

Figure 98. Micro-cast solitaires

Figure 99. Micro-cast trilogies

Figure 100. Micro-cast wedding bands

Figure 101. SLM™ solitaires

Figure 102. SLM™ trilogies

Figure 103. SLM™ wedding bands

Figure 104. Overall machine times for technique used and total man hours

Figure 105. Kg of CO2 equivalent produced by each of the two techniques

Table

1


. List of items produced for each model and production technique

Table 2 . as cast / as print roughness

Table 3 .Roughness after sanding or shot peening

Table 4 . Internal diameters measured for each model compared to nominal diameters

Table 5 . Average percentage porosity found for the two production techniques in question

Table 6. Vickers micro-hardnesses on micro-cast and printed model 1 wedding bands

Table 7. Mechanical characteristics

Table 8. Casting subdivision

Table 9. Print subdivision and relative production time

Table 10. Average and total machine and technician times for producing the rings with micro-casting

Table 11. Average and total machine and technician times for producing the rings with SLM™

Table 12. Subdivision of micro-casting production phase times

Table 13 . Subdivision of SLM production phase times

Table 14. Finishing operation times (in minutes) for rings produced by micro-casting

Table 15. Finishing operation times (in minutes) for rings produced by SLM™

Table 16. Overall finishing losses

Table 17. Costs relating to losses registered in the finishing phase,

assuming a 5% loss in the recovering phases

Table 18. Percentage production yield with micro-casting

Tab le19. Percentage production yield with SLM™

Tab le 20. Average refining costs on the Italian market

Table 21. Material to be refined for producing with SLM™

Table 22. Refining costs for producing with SLM™

Table 23. Material to be refined for producing with micro-casting

Table 24. Costs of refining for producing with micro-casting

Table 25 . Consumer materials needed for production

Table 27. Finishing costs, including losses, for each ring subdivided by model

Table 28. Refining costs for each ring subdivided by model

Table 29. Total production costs, including costs for producing semi-processed items,

finishing and refining

Table 30. Investment capital needed to start production with micro-casting and SLM™

Table 31. Kg of CO
₂ equivalent produced in micro-casting

Table 32. Kg of CO
₂ equivalent produced in SLM™

Introduzione

Le finiture superficiali decorative nere possono essere ottenute tramite diverse tecnologie accessibili e vengono realizzate quotidianamente ricorrendo a un’ampia gamma di metodi di applicazione. Anche se tali metodi vengono utilizzati da anni su scala industriale, ognuno di essi è caratterizzato da limiti. Essendo elevata la probabilità che i substrati di base trattati siano di colore bianco – come nel caso dell’alluminio, acciaio e argento – o pre-placcati con uno strato bianco utilizzato per proteggere il substrato – come nel caso della lega di zinco e dell’ottone -, la resistenza all’abrasione è un fattore qualitativo chiave utilizzato come indice di riferimento per tutte le categorie di trattamento superficiale decorativo nero. Poiché lo strato nero si usura, quello bianco resta esposto, mostrando un contrasto di colore che è ben percepito dall’occhio umano, cosa che semplifica all’acquirente finale la valutazione della qualità del prodotto via via che invecchia. 

Specifico della galvanoplastica, il processo di placcatura nera ha subito diverse modifiche negli ultimi anni. Con l’entrata in vigore del regolamento europeo REACH, alcune sostanze chimiche precedentemente utilizzate come additivi annerenti o agenti ossidanti non possono più essere utilizzate. Metalli storicamente utilizzati per la galvanoplastica nera come il nichel, il cobalto e il cromo sono costantemente sotto esame, incrementando la consapevolezza di marchi, acquirenti ed enti governativi, cosa che comporta, anno dopo anno, l’aumento di richieste legate a un uso limitato o all’eliminazione di questi metalli, soprattutto nei casi in cui lo scopo finale del prodotto è quello di essere indossato o essere a contatto con la pelle umana.

Alcuni processi di finitura nera conformi alla direttiva REACH comunemente utilizzati nel settore della galvanoplastica decorativa si basano sul rutenio nero e il rodio nero. Questi trattamenti vengono generalmente detti “flash”, con uno spessore massimo ottenibile di 0,2 micron, e si usurano piuttosto facilmente nel tempo malgrado la durezza elevata di entrambi i metalli.

Questo documento seguirà lo sviluppo di un processo di galvanoplastica basato su oro nero e la realizzazione di prove comparative relative alle applicazioni decorative nere più comunemente utilizzate al giorno d’oggi nel nostro settore e analizzerà i possibili vantaggi di tale processo in termini sia di strato finale che di strato tecnico.   

Tecnologie di finitura decorativa nera

I numerosi tipi di metodi di finitura nera disponibili variano a seconda del substrato originale dal trattare e dal settore di applicazione. Anche se questi processi sono piuttosto numerosi, al momento di esaminarli presentano limiti in termini di applicazione, design e costi.

 L’anodizzazione è un processo per la finitura di leghe in alluminio che utilizza l’ossidazione elettrolitica della superficie in alluminio per produrre un rivestimento protettivo di ossido. Questi rivestimenti possono essere colorati in una gamma di tonalità limitata, tra cui il nero. Anche se l’anodizzazione ha un vasto impiego funzionale a livello di industria, il processo viene utilizzato massicciamente come metodo decorativo in numerosi settori tra cui quello degli elettrodomestici, dell’elettronica e soprattutto quello dei telefoni cellulari in virtù del suo rapporto costo/qualità. Dal momento che il processo consiste essenzialmente nella colorazione dello strato di ossido naturale dei metalli, l’anodizzazione fa parte dell’alluminio in sé fornendo un legame completo e proprietà di adesione eccezionali. Gli svantaggi principali del processo in generale è che tende a esercitare un impatto negativo sui dettagli; il suo uso è pertanto parziale, ovvero limitato a forme semplici; l’applicazione stessa è limitata all’alluminio e a pochissimi altri metalli non ferrosi, il che restringe l’uso complessivo del processo dal punto di vista decorativo. 

I processi basati su vernici e rivestimenti sono procedimenti a base di resine diluite in acqua o solventi che offrono la più ampia gamma di colori ed effetti di tutti i metodi discussi. Anche se le sequenze di lavorazione possono variare, vernici e rivestimenti possono essere applicati a qualsiasi substrato metallico o addirittura plastico. In virtù di ciò vengono utilizzati praticamente in tutti i settori esaminati; fanno tuttavia la parte del leone – nel mercato degli ornamenti decorativi di qualità superiore – nel settore delle montature per occhiali. Con questo tipo di applicazione è possibile ottenere numerose varianti di nero.  I rivestimenti cataforetici monostrato (e-coating) sono piuttosto convenienti; tuttavia presentano problemi di durata o resistenza all’ossidazione a seconda del tipo di resina utilizzato. Questo inconveniente può essere eluso con l’uso di sistemi a 2 e 3 strati di vernice applicati a spruzzo utilizzando più resine; tuttavia alla fine il costo risulta più elevato a causa del tempo di applicazione e della perdita di materia prima fino al 70%. L’aspetto negativo dell’uso di vernice è dato dal fatto che ostacola il movimento meccanico di parti funzionali quali fermagli, catene e ganci. Nel caso specifico dell’applicazione a spruzzo, si lega con qualsiasi substrato con cui viene a contatto; pertanto l’uso in gioielleria è limitato a causa della presenza costante di pietre preziose. La vernice viene inoltre tacciata di “finitura a buon mercato” a causa dell’aspetto simile alla plastica che tende a conferire al metallo; pertanto molti settori di beni di lusso, a seconda del posizionamento del marchio, non ne prendono in considerazione l’uso.

La deposizione fisica da vapore (PVD) è un processo che utilizza una camera da vuoto per pressurizzare un target in vapore che viene quindi trasformato in un film condensato sulla superficie degli oggetti. I substrati applicabili sono limitati alla temperatura necessaria per realizzare una determinata deposizione, il che può costituire una grande limitazione per il suo utilizzo. Questo processo è in grado di produrre diversi colori, tra cui il nero. Uno dei processi più recenti utilizzati è il Diamond Like Carbon (DLC) che produce uno strato nero lucente estremamente duro e resistente all’abrasione. L’investimento iniziale in materia di attrezzature è piuttosto elevato, il che rende la PVD di gran lunga la tecnologia più costosa tra quelle discusse, essendo difficilmente fattibile dal punto di vista economico al di fuori della Cina. Anche se molto resistente ad abrasione e usura, la resistenza chimica della PVD tende a essere piuttosto debole a causa della formazione di fori microscopici durante la lavorazione. Questo problema può in ultima analisi essere superato con più trattamenti; tuttavia il costo del trattamento esercita un notevole impatto. Dal momento che il settore degli orologi utilizza l’acciaio inossidabile come materiale di base e richiede requisiti elevati in materia di resistenza all’abrasione e che le geometrie che tratta sono standardizzate, costituisce il candidato ideale per la PVD, che difatti viene ampiamente utilizzata in tutto il settore.

I trattamenti di galvanoplastica nera sono numerosi e utilizzano un’ampia varietà di metalli per ottenere il risultato target. Due dei mercati decorativi più esigenti per questo tipo di placcatura sono il settore della gioielleria e quello degli accessori moda. Entrambi i settori possono essere suddivisi in segmenti di fascia alta e bassa. I segmenti di fascia bassa utilizzano metalli quali il cromo, il nichel, lo stagno, il cobalto o leghe costituite da tali metalli principalmente per una questione di costi. Anche se questi metalli sono convenienti, la finitura ottenuta non è sufficiente per superare le prove climatiche normative richieste da alcuni marchi. Questo fatto, abbinato alla crescente richiesta di finiture prive di nichel e alle normative più severe come nel caso del regolamento REACH dell’Unione Europea, fa sì che i segmenti di fascia più alta ricorrano a metalli quali il palladio, il rodio e il rutenio per ottenere un colore nero. La gioielleria di fascia alta tende a utilizzare rodio nero mentre gli accessori moda di fascia alta si servono di rutenio nero. Anche se il rutenio nero e il rodio nero sono in grado di superare la maggior parte delle prove climatiche se utilizzati con le sequenze di pre-placcatura corrette, entrambi hanno uno spessore massimo ottenibile di 0,2-0,3 micron, per cui tendono a presentare problemi di resistenza all’abrasione. Questo inconveniente viene al momento eluso in entrambi i settori di fascia alta coprendo il metallo con una vernice per migliorare l’abrasione o semplicemente accettando il fatto che il colore si usurerà nel tempo.

Il che porta a chiedersi se sia possibile ottenere uno strato nero resistente all’abrasione senza l’uso di vernice, evitando al contempo l’uso di nichel nella lavorazione e mantenendo la conformità al regolamento REACH. Se è così, il mercato sarà in grado di sostenerne il costo?         

Oro nero senza nichel

Sia il rodio che il rutenio hanno uno spessore massimo ottenibile di 0,2-0,3, principalmente dovuto al fatto che si tratta di sistemi monometallici che si infragiliscono con l’aumentare dello spessore. Nell’ottica di migliorare la resistenza all’abrasione, il primo obiettivo sarà quello di mantenere una durezza soddisfacente rendendo al contempo lo strato più malleabile e ampliando la gamma dello spessore raggiungibile. Omettendo l’uso di nichel e cobalto in modo da preservare l’ipoallergenicità e l’uso di rame, che resta la fonte principale di ossidazione, è stato studiato un elettrolita bimetallico che includeva oro, palladio e ferro. Il sistema studiato presentava un pH alcalino per ricevere bene i metalli selezionati, con l’oro nella concentrazione più alta e il ferro in quella più bassa dei metalli utilizzati per sviluppare il composto chimico.

Tabella 1 – Caratteristiche elettrolitiche dell’oro nero

Valutazione della superficie della lega in oro nero

Durante la lavorazione, l’elettrolita produce una lega di colore nero costituita dal 49% di palladio, il 39% di oro e il 12% di ferro in peso, con un titolo in oro depositato di circa 12,5 kt. Per valutare il colore della lega viene utilizzato il sistema di coordinate colore CIELab, concentrandosi sulla coordinata L durante la valutazione di finiture nere.

La coordinata L è il valore della luminosità che in questo caso determina la gradazione scura complessiva del deposito nero, con il valore L più basso che equivale a una tonalità più scura di nero. Con la lega in oro nero è stato misurato un valore L pari a 58.

Tabella 2 – Confronto delle coordinate colore di oro nero, rutenio nero e rodio nero

Questo valore rientra nel range più elevato dei colori neri standard del settore con coordinate L che vanno da 50 a 60 nel caso del rodio nero e una disponibilità molto più ampia di tonalità di nero con il rutenio che presenta coordinate L che spaziano tra 32 e 60. Le formule più comunemente utilizzate sia del rodio che del rutenio hanno coordinate L pari a 58 o 59.

È stato realizzato un test basato su celle di Hull per identificare l’intervallo di densità corrente, il cui risultato è stato un pannello nero con finitura a specchio prodotto a 1,0 A/dm² dopo 10 minuti di tempo di deposizione. A questo si aggiunge il fatto che l’elettrolita ha dimostrato un buon potere di penetrazione ed estensione, con un valore di densità corrente ottimale pari a 1,0 A/dm².

Sono state eseguite prove di spessore con l’uso di un microscopio SEM/EDX. Prove di laboratorio hanno dimostrato la possibilità di ottenere un deposito fino a 2 micron mantenendo la finitura a specchio prevista per il settore della placcatura decorativa di fascia alta.  Con misurazioni accurate dello spessore, sono state calcolate le velocità di placcatura, ottenendo 1 micron in 12-15 minuti a 1,0 A/dm².

Osservazioni chimiche

La chimica si è dimostrata stabile in un arco di tempo di 12 mesi, dando prova dell’assenza di elementi chimici di partecipazione in caso di mantenimento corretto del pH. Nell’elettrolita è stato rilevato un pH fluttuante, soggetto a un calo giornaliero pari allo 0,2.

L’abbassamento del pH può essere corretto con l’aggiunta di sali di regolazione del pH; tuttavia ciò comporterà il monitoraggio giornaliero della soluzione del pH da parte degli operatori della soluzione di oro nero, conferendo all’elettrolita caratteristiche di manutenzione simili a quelle di un bagno al solfato per oro.

Figura 1 – Variazione del pH in un periodo di 18 giorni

Sequenze di placcatura nei segmenti di mercato

Sono state valutate due della sequenza di placcatura nera più consolidate e standard in ciascun segmento di mercato di fascia alta. Nel caso della gioielleria è stata valutata la finitura nera sopra l’argento come uno dei casi più standardizzati. Partendo da un materiale di base come l’argento 925, viene applicato palladio come strato di placcatura iniziale seguito da rodio nero nei cicli di finitura più comuni. In questo esempio il palladio viene applicato come strato tecnico con un duplice scopo. Il primo è il miglioramento della resistenza complessiva all’ossidazione fungendo da barriera alla migrazione del rame e impedendo la corrosione dovuta a fattori ambientali. Più spesso è lo strato di palladio, più forte è la resistenza alle prove climatiche. Il secondo motivo per l’uso del palladio è la protezione dell’elettrolita. Il palladio in quanto metallo non è soggetto a corrosione in un ambiente acido; lo sono invece l’argento e il rame con cui viene legato. Se viene saltato il passaggio del palladio, l’argento e il rame finiranno per deteriorare la qualità finale depositata in quanto la loro presenza aumenta con il tempo nel bagno di rodio sotto forma di contaminazione metallica.

Il settore degli accessori di alta moda utilizza un ciclo di finitura completamente diverso per ottenere un colore simile. Il motivo principale risiede nel fatto che i materiali di base iniziali sono tipicamente l’ottone o leghe di zinco, che richiedono una sequenza di placcatura diversa per il conseguimento di una finitura di qualità elevata.  I cicli più comunemente utilizzato sono costituiti da 5-7 processi di placcatura che comportano numerosi strati per ottenere vantaggi tecnici. La sequenza scelta inizia con il rame alcalino per l’adesione, seguito da rame acido per la brillantezza della superficie, passando al bronzo bianco per una migliorata durezza e quindi al palladio per la resistenza all’ossidazione; vengono infine applicati 1-2 strati di rutenio. In questo caso vengono utilizzati 2 strati di rutenio; il primo strato è costituito rutenio grigio chiaro che viene quindi seguito da rutenio nero, in modo da migliorare, in ultima analisi, l’indossabilità complessiva dei prodotti.  

L’integrazione della lega di oro nero nei cicli di placcatura selezionati consente di misurare l’aggiunta del nuovo strato utilizzando i due processi più standard del settore come benchmark qualitativo. 

Sequenza di placcatura della gioielleria in oro nero

Nel caso delle modifiche proposte al ciclo di lavorazione della gioielleria, l’oro nero era stato scelto per sostituire il palladio come strato tecnico, visto che la lega di oro nero in sé contiene una percentuale elevata di palladio, oltre al fatto che entrambi gli elettroliti hanno un pH alcalino simile, cosa che comporta deviazioni di lavorazione minori.

Parti di argento sterling 925 sono state trattate direttamente con 0,3 micron di lega di oro nero seguiti da 0,2 micron di rodio nero. Parti aggiuntive sono state trattate con 0,3 micron di palladio e 0,2 micron di rodio nero da utilizzare come riferimento nelle prove di abrasione e corrosione. A seguito dei trattamenti i campioni trattati utilizzando l’oro nero come substrato presentavano un colore visibilmente più scuro rispetto a quelli trattati con palladio, anche se la finitura finale era lo stesso rodio nero applicato utilizzando gli stessi parametri. Questo processo viene definito in questo documento come sequenza di processo 1.

Figura 2 – Fila superiore: sequenza di placcatura scelta come ciclo di riferimento, fila inferiore: ciclo di prova con l’implementazione di oro nero.

Sequenza di placcatura degli accessori moda in oro nero

Le modifiche selezionate per il ciclo del processo degli accessori moda erano diverse vista la complessità del processo di placcatura in sé. In questo caso ci siamo concentrati sul ciclo di finitura che include 1 strato di rutenio nella sequenza. Il deposito in oro nero è stato testato come strato tecnico utilizzandolo per sostituire lo strato in palladio per mantenere un costo simile, creando spazio quantitativo per l’oro nero. Questo processo viene definito in questo documento come sequenza di processo 2.

Figura 3 – Fila superiore: sequenza di placcatura scelta come ciclo di riferimento, fila inferiore: ciclo di prova con l’implementazione di oro nero.

Inoltre l’oro nero era stato implementato nel processo come finitura finale dato che condivide un colore simile al rutenio grigio canna di fucile. Poiché entrambi i depositi hanno una coordinata L di 58, questa verrà utilizzata come riferimento per la resistenza degli ori neri all’ossidazione con uno standard del segmento di mercato.

Figura 4 – Ciclo di test con l’implementazione dell’oro nero come strato finale da testare in contrasto al rutenio.

Le parti in ottone sono state lavorate con le sequenze di placcatura del segmento di mercato standard e le deviazioni di lavorazione dell’oro nero relative per l’analisi qualitativa.  Questo processo viene definito in questo documento come sequenza di processo 3.

Analisi comparativa

Sulla base delle suddette informazioni le parti trattate con i cicli di placcatura più comuni dai due segmenti di mercato selezionati sono state sottoposte ad analisi comparativa a fronte di quelle trattate con lo strato di oro nero inserito nella sequenza. Sono stati utilizzati metodi di analisi normativa standardizzata per simulare l’ossidazione e l’abrasione.

Resistenza all’abrasione

Il metodo di test utilizzato per misurare la resistenza all’abrasione nel mercato degli accessori di alta moda è il test Turbula, progettato per simulare usura o abrasione. Viene utilizzata una macchina specifica in cui la rotazione è basata su un perno, creando un ambiente più aggressivo rispetto a un miscelatore standard.

Il mezzo utilizzato è una ceramica a forma di piramide allungata con una forma diversa che garantisce punti di contatto sia acuminati che piatti, simulando due forme aggressive di abrasione. L’analisi viene eseguita con un mezzo che pesa circa 41 grammi per 100 pezzi. Per ogni analisi sono state trattate cinque parti con una fase di valutazione fissa seguita da un ciclo di Turbula a un tempo fisso e una velocità di 72 giri/min.

Risultati della sequenza di gioielleria

A seguito di un ciclo di abrasione di 3 minuti, i campioni di argento 925 che hanno utilizzato il deposito di oro nero come strato intermedio si sono dimostrati più resistenti all’abrasione rispetto alle parti trattate con il processo in palladio tradizionale. Un’ulteriore valutazione ha determinato un miglior legame tra lo strato in oro nero e rodio nero rispetto ai metodi di placcatura tradizionali.

Figura 8 – Fila superiore: palladio come strato intermedio; fila inferiore: oro nero come strato intermedio entrambi dopo il ciclo di abrasione

Risultati della sequenza degli accessori moda

A seguito di un ciclo di abrasione di 30 minuti, i campioni di ottone che hanno utilizzato il deposito di oro nero come strato intermedio si sono dimostrati più resistenti all’abrasione rispetto alle parti trattate con il processo in palladio tradizionale, che hanno mostrato la perdita completa dello strato di rutenio non superando in tal modo il test.

Figura 9 – A sinistra: oro nero come strato intermedio; a destra – palladio come strato intermedio entrambi dopo un ciclo di abrasione di 30 minuti 

Resistenza al sudore artificiale

Il test del sudore artificiale a cui sono sottoposti gli accessori di fascia alta è descritto in NFS 80-772:2010-10; si tratta di una prova basata sul contatto diretto, più aggressiva rispetto ai test atmosferici. In base a questo metodo di test, i campioni vengono messi a contatto diretto con un feltro assorbente che è stato immerso in una soluzione di sudore artificiale. Il campione viene sigillato ermeticamente e mantenuto a una temperatura costante di 55 °C. I campioni vengono quindi tenuti per una durata predeterminata di tempo a incrementi di 24 ore.

A seguito di un ciclo di sudore artificiale di 24 ore i campioni di argento 925 che hanno utilizzato il deposito di oro nero come strato centrale hanno dimostrato una resistenza superiore rispetto alle parti trattate con la procedura basata su palladio tradizionale. La superficie dei pezzi che utilizzano il processo in palladio ha mostrato segni notevoli di aggressione chimica.

Figura 9 – Fila superiore: palladio come strato intermedio; fila inferiore: oro nero come strato intermedio entrambi dopo un ciclo di sudore artificiale di 24 ore

Nebbia salina

A seguito della normativa ISO 9227, gli oggetti da testare vengono sospesi in una camera sigillata e sono sottoposti a una nebbia salina costante per una durata di tempo predeterminata. Il test è progettato per simulare un ambiente corrosivo e verifica la resistenza dei substrati in un tempo di ciclo di 96 ore.

Per testare la resistenza alla nebbia salina della lega di oro nero, è stata utilizzata la sequenza di processo 3 in quanto in questo caso la lega di oro nero viene esposta come strato finale.

A seguito del ciclo di nebbia salina di 96 ore, i campioni in ottone finiti sia in rutenio che in lega di oro nero non hanno mostrato segni di corrosione.

Figura 9 – A sinistra: oro nero come strato superiore; a destra: rutenio come strato superiore a seguito di un ciclo di nebbia salina di 96 ore

Conclusioni

L’analisi di laboratorio iniziale ha dimostrato che la lega in oro nero ha migliorato i risultati qualitativi se utilizzata come strato tecnico/intermedio, in particolare la resistenza al sudore artificiale, se applicata sull’argento come sostituto del palladio e finita con rodio nero. Quando testata come strato finale, la resistenza alla corrosione era simile a quella del rutenio, ma date le notevoli differenze di costo tra palladio/oro e rutenio e unitamente al fatto che i due elettroliti producono lo stesso colore, l’uso in questo senso è improbabile.

Il pH fluttuante dell’elettrolita di oro nero può essere visto come una limitazione in quanto restringe l’uso del prodotto a stabilimenti di placcatura qualificati non potendo essere utilizzato nei processi in beaker, che sono comuni nel settore della gioielleria.

 Le prove proseguiranno in futuro in quanto l’elettrolita di oro nero consente di ottenere un’ampia gamma di spessori e – considerando le numerose e diverse sequenze di processo utilizzate in tutto il settore delle finiture decorative – apre numerose opportunità per questo processo, anche attraverso la combinazione della galvanoplastica con altre tecnologie quali la PVD.

In ultima analisi, la migliorata resistenza all’abrasione e alla corrosione resterà un obiettivo centrale per molte aziende del settore della finitura delle superfici. Miglioramenti qualitativi agli articoli commercializzati migliorano la soddisfazione del cliente, rafforzano la reputazione del marchio e soprattutto riducono l’impronta complessiva sull’ecosistema grazie a prodotti che durano più a lungo.

Stato dell’arte

Le leghe di palladio sono state al centro dell’attenzione al simposio di Santa Fe, in particolare nel 2006 e nel 2009 anni in cui la richiesta di Pd nel settore della gioielleria ha registrato una rapida crescita. Sono stati pubblicati diversi documenti sul mercato [1], sul processo di microfusione [2-4]
e sulla fabbricazione e la lavorazione [5, 6] del palladio. Il primo studio sulla microfusione relativo alle leghe 950Pd è stato presentato da Fryé [4]. A causa delle limitate informazioni allora disponibili, lo scopo essenziale del documento era quello di ottenere una migliore comprensione delle caratteristiche
di colata delle leghe 950Pd utilizzate nel settore orafo. Il lavoro di Battaini [2] mirava a presentare le principali proprietà fisiche e chimiche delle leghe dentali basate su palladio e a trasferire l’esperienza acquisita nel campo odontoiatrico a quello dell’oreficeria.

Nel 2007 uno studio sulla microfusione delle leghe 950Pd è stato realizzato dal fem per conto della Palladium Alliance International (PAI). I risultati sono stati presentati in occasione del simposio di Santa Fe del 2008 e pubblicati nel 2009 [3]. Nella sezione che segue verranno sintetizzate le conclusioni
principali di questi studi.

2.1 Sfide di colata tipiche associate alle leghe di palladio 950

Il palladio 950 commerciale contiene Ru, Ga o Co come componenti principali. Nella Tabella 1 viene fornito un elenco delle leghe con le loro principali proprietà e caratteristiche.

Tabella 1:            Panoramica delle leghe di palladio commerciali tratta dalle schede tecniche dei loro produttori. AC = as-cast, CW = lavorazione a freddo, AN = ricottura.

Il rutenio (Ru) è un metallo del gruppo del platino dal colore bianco che lega bene con il palladio. Le leghe contenenti Ru mostrano una temperatura di fusione più elevata, in quanto il rutenio incrementa la temperatura del solidus e del liquidus in base al diagramma di fase binario. Il Ru ha una solubilità
limitata nel palladio; pertanto vengono prodotte solo leghe 950Pd-Ru. L’incremento dell’intervallo di fusione tramite l’aggiunta di Ru richiede temperature di colata più elevate rispetto al palladio puro. Ciò comporta una sollecitazione termica più elevata sui crogioli e i materiali di microfusione
durante la colata. L’intervallo di fusione di 950PdRu è molto ristretto, solo pochi gradi Celsius. Alcuni valori forniti nelle schede tecniche dei produttori per 950PdRu sono in contraddizione con le informazioni sul diagramma di fase. Ciò viene attribuito a ulteriori leganti non specificati o alla
difficoltà di determinare l’intervallo di fusione effettivo.

Durante la solidificazione si formano dendriti e il Ru viene segregato nel cuore di queste ultime e la fusione restante si arricchisce di Pd. Solitamente l’intervallo di fusione cresce a causa della segregazione, ma nel caso di 950PdRu ciò avviene in modo molto limitato. Di conseguenza la lega mostra
un congelamento quasi isotermico e quindi un riempimento dello stampo molto limitato durante la colata. Tale processo è stato investigato in dettaglio per 950PtRu [15].

Il Ru si dissolve nel palladio in modo da formare una soluzione solida. Le leghe Pd-Ru sono relativamente morbide a causa della piccola differenza nella dimensione dell’atomo di Pd e Ru. La durezza tipica è di circa 100-120 HV1 nella condizione di ricottura o as-cast. Per migliorare le proprietà meccaniche
vengono frequentemente utilizzate aggiunte di Gallio (Ga).

Il diagramma di fase binario di Pd con Ga viene mostrato nella Figura 1. Ga ha un punto di fusione molto basso (29 °C) e la sua aggiunta abbassa in modo significativo la temperatura del solidus e del liquidus di Pd. La solubilità massima in Pd è una percentuale in massa di Gallio pari a 8 ca.. A una
concentrazione di Ga più elevata si formano numerosi composti intermetallici in reazioni di fase complesse. Non sono noti studi sistematici sull’indurimento per precipitazione delle leghe a contenuto elevato di Pd nella letteratura aperta. Tali studi, tuttavia, sono disponibili per le leghe di Pt [16]
e i risultati possono essere trasferiti a quelle di Pd.

L’indurimento per precipitazione è ben noto e applicato per le leghe 950Pt [16, 17]. Tuttavia la superiorità del gallio nel palladio è più elevata che nel platino. Pertanto sono necessarie qualità più elevate di Ga per ottenere lo stesso livello di durezza – ovvero per un dato contenuto di Pd, ad esempio
950Pd, la durezza conseguibile è inferiore. La risposta all’indurimento di Pt legato con Ga è segnalata come instabile e pertanto classificata come non praticabile per un indurimento affidabile da parte di alcuni autori [16].

Figura 1:              Approccio tradizionale delle leghe di palladio dure con un contenuto di Ga più elevato. Sezione del sistema Pd-Ga (a sinistra) confrontata con il sistema Pt-Ga (a destra) calcolata utilizzando il database TCNOBL1 e ThermoCalc.

Prove sperimentali e indagini corrispondenti in un precedente studio del fem [3] si sono concentrate su due leghe, una con Ru/Ga e l’altra con Ag/Ga/Cu. Di conseguenza non è possibile trarre alcuna conclusione generale in ordine all’adeguatezza delle leghe per la colata di palladio in base alla composizione
della lega. Sulla base dell’analisi dei difetti nelle colate industriali, sembra che le leghe con un contenuto di gallio relativamente alto tendano ad avere una suscettibilità più elevata alla formazione di cricche nelle parti as-cast. La formazione di cricche si è rivelata un problema complesso. Una
failure analysis approfondita ha rivelato che il meccanismo sottostante è correlato alle proprietà e alle condizioni di colata particolari del materiale di microfusione. Si deve notare che le colate prive di cricche della lega Ru/Ga sono state ottenute n modo riproducibile durante prove di colata presso
il fem e sono inoltre ottenute in qualità elevata e in modo riproducibile da molti fonditori industriali che hanno cooperato al progetto.

Il silicio è un’impurità tipica che si verifica nei processi di microfusione. Se viene utilizzato materiale di scarto per la rifusione la rimozione di eventuali residui della microfusione è della massima importanza [2]. Tali residui di ossidi possono decomporsi durante la fusione, in particolare in condizioni
riducenti (gas di formatura: Ar/H
₂ or N
₂/H
₂) che devono essere evitate. L’ossigeno rilasciato passa in soluzione solida all’interno della fusione ed evapora durante la solidificazione formando una porosità del gas significativa. Il silicio forma un eutettico a punto di fusione basso (Pd + Pd
₃Si) a una temperatura di 782 °C. Tale eutettico sui bordi dei grani è responsabile delle crepe di ritiro. Un esempio del risultato catastrofico delle impurità di silicio viene mostrato nella Figura 2. L’albero di colata è diventato completamente fragile. Numerose cricche nelle parti provocano
più fratture che si verificano lungo i bordi dei grani interdendritici. Il numero 4 nella parte inferiore destra dell’immagine mostra l’incremento della concentrazione di Si determinata dall’analisi EDX.

Figura 2:              Frattura a caldo dovuta a contaminazione con residui della microfusione.

3 Processo di sviluppo

3.1 Identificazione di componenti della lega adeguati

Sono stati selezionati potenziali leganti nella tavola periodica degli elementi. Alcuni elementi hanno dovuto essere esclusi in quanto volatili, tossici, allergenici o radioattivi, troppo reattivi nelle condizioni tipiche della microfusione o insolubili. I principali requisiti per la nuova lega erano:

Intervallo di fusione sufficiente di almeno 25 K

Durezza media (130-160 HV1)

Struttura a grano fine

La lega 950PtRu è stata definita come riferimento per lo sviluppo delle nuove leghe 950Pd. 950PdRu è caratterizzata da un promettente colore grigio-argento rispetto al colore grigio della maggior parte delle leghe 950Pd. Contiene metalli del gruppo del platino al 100% e pertanto non richiede gas protettivo
durante la lavorazione. Tuttavia la fluidità della lega è molto bassa e alcuni produttori non la consigliano per la colata.

Tra i candidati restano solo pochi elementi. Per poter superare le scarse proprietà di colata di 950PdRu sono necessari i seguenti miglioramenti:

Ampliamento dell’intervallo di fusione  Aggiunta di Co, Fe o Cu

Miglioramento delle proprietà di colata, in particolare del riempimento dello stampo    Aggiunta di Co

Ottimizzazione della segregazione, riduzione delle reazioni alla microfusione     Aggiunta di Sn

Miglioramento del colore e della durezza            Aggiunta di Cr, Fe, B

Affinamento del grano  Aggiunta di Fe, W, Zr

Nella Figura 3 viene mostrato l’intervallo di fusione variabile di varie leghe 950Pd in cui il Ru viene sostituito da un terzo elemento (Me). Sul lato sinistro della figura troviamo la lega binaria 950PdRu; sul lato destro le leghe binarie 950PdMe. Alcuni elementi come Au hanno a malapena effetto sull’intervallo
di fusione e sulla temperatura del liquidus. Altri elementi (Ag, Cu,Cr) hanno un effetto medio sull’intervallo di fusione e sulla temperatura del liquidus. Nel caso del rame sono necessarie quantità relativamente alte per ottenere un effetto. Gli effetti maggiori si hanno con l’aggiunta di Co e Fe.
Tuttavia, a causa della loro tendenza all’ossidazione, la quantità deve essere limitata al 2% max.

Figura 3:              Effetto dell’aggiunta di leganti a 950PdRu. Il Ru è sostituito da un terzo elemento metallico (Me). L’asse X fornisce la quantità del terzo elemento in percentuale di massa. Calcolo effettuato utilizzando il database SNOB3 e ThermoCalc.

L’effetto della segregazione durante il processo di solidificazione può essere simulato tramite le cosiddette simulazioni di Scheil-Gulliver. L’intervallo di fusione effettivo di una lega solitamente aumenta, in quanto l’equilibrio termico completo che viene presunto nei diagrammi di fase dell’equilibrio
non viene conseguito durante un processo di raffreddamento relativamente rapido. Ciò comporta una variazione continua della composizione chimica della fase liquida mentre la solidificazione procedere e questo effetto può essere studiato tramite le simulazioni di Scheil-Gulliver. L’effetto di tali variazioni
della composizione non equilibrata della fusione sulla temperatura del solidus viene mostrato nella Figura 4 per una serie di leghe 950Pd-30Ru-Co,Fe. La 950Pd-Ru binaria mostra un intervallo di solidificazione molto stretto. L’aggiunta del 20 ‰ di Fe+Co riduce la temperatura del solidus e il processo
di segregazione si fa più pronunciato. La segregazione di Fe e Co nella fase liquida comporta una riduzione della temperatura del solidus effettiva e consente un intervallo di fusione di circa 30-100 K. Ciò sembra promettente in termini di miglioramento del riempimento dello stampo, migliore alimentazione
(riduzione del microritiro) e riduzione delle reazioni di microfusione.

Figura 4:              Calcolo di Scheil-Gulliver. Segregazione di leghe 950Pd-30Ru con contenuto di Fe e Co variabile.

Prove di microfusione

Dalle considerazioni di cui sopra è stata derivata una serie di composizioni della lega come mostrato nella Tabella 2. Come lega di riferimento è stata utilizzata una lega 950PdRu acquistata presso C. Hafner, Pforzheim, Germania. Le leghe sono state preparate tramite fusione ad arco da elementi puri
con una purezza del 99,9% o superiore (acquistati presso HMW Hauner Metallische Werkstoffe, Germania). Il campione a forma di bottone è stato laminato a freddo nella lamina che è stata utilizzata per la microfusione centrifuga di alberi tipici conformemente alla Figura 5 con una massa di circa 100 g.
Su questi alberi è stata realizzata una caratterizzazione di base che comprendeva la determinazione del colore, il rilascio di metalli, la durezza, la risposta al termoindurimento e la microstruttura. L’albero conteneva una serie di parti di gioielleria tipiche inclini ai difetti tipici della colata.
La griglia è stata utilizzata per la prova del riempimento dello stampo mentre il campione a forma di lamina per le prove del rilascio di metalli e delle misure del colore. Sono state preparate e analizzate circa 35 composizioni di leghe. Nella Tabella 2 viene riportata una scelta di queste composizioni.
Sono state selezionate le leghe più promettenti e modificate nel seguente passaggio.

Tabella 2:            Composizioni delle leghe testate in prove su piccola scala (selezione)

Figura 5:              Configurazione dell’albero di colata e parti di colata

La colata ha richiesto un controllo di processo sofisticato allo scopo di garantire condizioni di colata riproducibili e affidabili (Figura 7). Come fonditrice è stato utilizzato il modello TCE10 della Topcast, Italia, che ha consentito la fusione e la colata entro 40-60 s dall’inizio del processo di
riscaldamento. Per tutte le prove di colata è stato utilizzato un crogiolo al quarzo di qualità elevata del tipo "KGZ"di Porzellanfabrik Hermsdorf, Germania. Questo tipo di crogiolo si era rivelato adatto per le leghe di platino in un studio precedente. La temperatura del metallo è stata controllata
durante la fusione e la colata con una termocamera. Ciò ha consentito una valutazione dettagliata della temperatura del metallo superiore a quella del pirometro integrato nella fonditrice. Persino la temperatura del cilindro ha potuto essere controllata tramite termocoppie montate sull’albero o vicino
alla superficie del cilindro interna per documentare il surriscaldamento della microfusione. Tuttavia tali misure richiedono uno sforzo molto elevato e pertanto sono state utilizzate solo in una quantità molto limitata di prove di colata.

La temperatura del cilindro è stata selezionata in base alle dimensioni e alla forma delle parti ed è stata di 650°C per la maggior parte delle prove di colata. Questa temperatura si è dimostrata il miglior compromesso tra riempimento dello stampo elevato e bassa porosità da ritiro.  Per ridurre il più
possibile le reazioni di microfusione è stato utilizzato gesso bicomponente a base fosfatica (platino Ransom&Randolph). Dopo la fusione le parti sono state sottoposte a prova non distruttiva tramite tomografia computerizzata e metallografia tradizionale.

Figura 6:              Fonditrice e controllo di processo (per la descrizione vedere il testo)

Un riempimento dello stampo ottimizzato richiede una configurazione dell’albero adeguata. In base all’esperienza dei progetti di fusione precedenti con platino, le parti sono state montate sul lato principale relativo al senso di rotazione della fonditrice. Nella Figura 7 viene illustrata la configurazione
della colata, le forze di azione e un esempio della simulazione del processo di riempimento dello stampo. A causa de montaggio delle parti sul lato principale il metallo è obbligato a scorrere verso la punta dell’albero. Le parti vengono quindi riempite gradualmente dalla punta verso l’attacco di colata
dell’albero. Dettagli sul processo di microfusione e sulla simulazione di colate sono disponibili in [18, 19].

Figura 7:              Condizione di colata nella colata centrifuga L’orientamento delle parti sul lato principale dell’albero, relativo alle forze in gioco. Le frecce blu indicano il senso di rotazione. Le frecce indicano le forze in gioco: arancio (inerzia), rosso (gravità) verde (forza risultante).
Simulazione del processo di riempimento dello stampo.

Dopo la fusione l’albero è stato tagliato e documentato come mostrato nella Figura 8. La qualità della superficie è stata valutata utilizzando l’aspetto del campione a lamina. Il campione griglia ha fornito informazioni sul riempimento dello stampo, che è stato dato come percentuale dei punti di intersezione
riempiti della griglia. È stata effettuata un’ispezione metallografica sugli anelli con canale di alimentazione doppio e singolo che sono inclini alla porosità da ritiro. La sezione metallografica del campione a lamina è stata utilizzata per la misura del colore prima e dopo una prova di rilascio di
metallo in saliva artificiale. I risultati sono stati suddivisi in tre categorie che vengono riportate nella Figura 9 per il riempimento dello stampo, la reazione di microfusione e la porosità. La microstruttura e le dimensioni del grano sono state determinate utilizzando la microscopia elettronica
a scansione. Sono stati indagati possibili difetti quali cricche, mancata omogeneità chimica o inclusioni (Figura 10).

Figura 8:              Risultati della colata e routine di valutazione (per la descrizione vedere il testo)

Figura 9:              Criteri di valutazione per riempimento dello stampo, qualità della superficie e porosità

Figura 10:            Microstruttura nello stato as-cast di leghe selezionate

Tabella 3:            Principali risultati di leghe selezionate in prove di colata full-size

Nella Tabella 3 vengono forniti i risultati di alcune composizioni delle leghe selezionate. 950PdRu mostra buone proprietà di base con dimensione di grano media e solo pochissime cricche intergranulari. L’aggiunta di Co riduce le dimensioni del grano e contribuisce a evitare completamente le cricche.
Le aggiunte di Fe sono promettenti in termini di riduzione delle dimensioni del grano, ma provocano problemi con una porosità del gas massiccia. Malgrado ciò il Fe viene mantenuto come potenziale legante. Il Cu non ha apportato alcun vantaggio reale e ha ridotto la durezza comportando la distorsioni
degli anelli già durante la sgessatura. Ciò ha portato all’esclusione del Cu. L’aggiunta di Cr ha comportato reazioni eccessivamente forti con la microfusione e cricche marcate, per cui è stato escluso dallo studio. L’aggiunta di Sn ha significativamente modificato la morfologia del grano, il che è
stato valutato come uno svantaggio. Tuttavia il riempimento dello stampo è stato ottimo e le prestazioni complessive buone il che ha consentito di mantenere Sn nell’elenco di leganti promettenti.

Infine le leghe contrassegnate in verde sono state studiate ulteriormente mentre quelle contrassegnate in rosso sono state escluse dallo studio. In base a questa valutazione le leghe più promettenti sono state selezionate e ottimizzate in ulteriori passaggi tramite l’aggiunta di leganti extra. Tutte
le leghe hanno mostrato una bassa durezza in questa fase dello sviluppo; pertanto l’obiettivo di ulteriore miglioramento era l’incremento della durezza.

Per migliorare la durezza la letteratura ha fornito il boro (B) e l’alluminio (Al) come componenti promettenti [20, 21]. Tuttavia entrambi gli elementi non sono di facile aggiunta a causa della loro elevata reattività. Ciò ha richiesto la preparazione di pre-leghe con quantità accuratamente regolate
di Al e B. L’uso di tali pre-leghe ha consentito al fonditore di preparare leghe contenenti Al e B senza i rischi di ossidazione durante la fusione iniziale della lega. Ha inoltre permesso al produttore di controllare la quantità di legante in modo molto preciso. Nella Figura 11 vengono forniti risultati
a livelli diversi di B e Al. Piccole aggiunte di B nell’ordine dell’1‰ (lega PD1502) incrementano in modo significativo la resistenza e la durezza preservando al contempo la duttilità. Le aggiunte di Al hanno un effetto simile, ma sono necessarie quantità molto più elevate di questo elemento. È stato
rilevato un incremento lineare approssimativo della resistenza e della durezza.

Questi risultati sono stati determinati tramite la prova di trazione delle barre che sono state colate in stampi di rame. Tale colata comporta un raffreddamento rapido e potrebbe non essere rappresentativa per la microfusione. Pertanto sono state eseguite ulteriori prove tramite microfusione. I livelli
di resistenza e la durezza vengono mantenuti dalla microfusione, ma la duttilità è significativamente più bassa. Questo è tipico ed è solitamente un effetto del struttura del grano colonnare e più grezza dopo la microfusione.

Figura 11:            Proprietà meccaniche di leghe selezionate basate su 950PdRu con aggiunte di B e Al

Ulteriori ottimizzazioni nelle prove di colata aggiuntive hanno avuto come risultato le composizioni fornite nella Tabella 4. Il contenuto di boro è stato ridotto per motivi di sicurezza, in quanto un più elevato livello di B può provocare frattura a caldo in condizioni di raffreddamento inadeguate.
Una combinazione di Al e B ha fornito le proprietà migliori e affidabili in prove di colata ripetute. La combinazione di metalli è stata più efficace che singole aggiunte di quantità persino più elevate. Variazioni di leghe che utilizzano Fe e Sn offrono il vantaggio di un affinamento del grano e un
migliorato riempimento dello stampo rispettivamente.

Tabella 4:            Composizioni di leghe ottimizzate che soddisfano i requisiti di durezza.

Un confronto con altre leghe bianche a caratura elevata (Figura 12) mostra alcuni specifici vantaggi delle leghe 950Pd di recente sviluppo. Rispetto a 950PdRu la durezza viene significativamente incrementata a livelli di 140-160 HV1, il che è considerato ottimale. Una durezza più elevata può essere vantaggiosa
per una migliorata resistenza ai graffi, ma compromette la formabilità del materiale durante l’incastonatura delle pietre. Le leghe mostrano una risposta al termoindurimento che potrebbe essere utilizzata, nel caso in cui sia necessaria una durezza più elevata. Il confronto con le leghe di platino all’avanguardia
950 (colonne verdi) mostra proprietà superiori rispetto a 950PtRu, ma una durezza inferiore a 950PtRuGa, che è a volte considerato troppo duro. La durezza è anche paragonabile a quella delle leghe in oro bianco Pd a 18k contenenti zinco per una migliorata durezza (colonna gialla).

Ulteriori proprietà da prendere in considerazione sono il colore e la densità delle leghe. Per le leghe bianche l’indice di giallo (YI D1925) è lo standard accettato per la valutazione del colore [22]. Valori YI inferiori a 18 sono considerati "bianco premium", il che significa che le leghe
non richiedono rodiatura. Le leghe basate su 950PdRu mostrano valori YI inferiori a 10, paragonabili a quelli delle leghe 950Pt. La differenza di colore tra 950PdRu e 950PtRu è a malapena visibile all’occhio umano. In contrasto le leghe in oro bianco Pd a 18k con un valore YI intorno a 18 appaiono molto
più gialle. La densità delle leghe basate su 950PdRu è vicina a 12 g/cm³, che è il 60% delle leghe 950Pt e il 75% delle leghe in oro bianco a 18k. La densità più bassa consente al produttore di produrre gioielleria più massiccia allo stesso peso o gioielleria leggera, ad esempio orecchini o pendenti.
La combinazione di proprietà etichetta le leghe 950PdRu di recente sviluppo come "leggere, brillanti e resistenti".

Figura 12:            Confronto con leghe commerciali (proprietà tipiche conformemente a [23])

4 Sintesi e conclusioni

Nel presente documento viene descritto lo sviluppo di leghe 950 Pd con migliorate proprietà per applicazioni di gioielleria. Sulla base di calcoli termodinamici sono state selezionate composizioni di leghe promettenti, che sono poi state fuse e colate tramite microfusione centrifuga. Sono stati utilizzati
materiali per microfusione e crogioli tipici dimostratisi adatti per le leghe in platino. Questi materiali sono stati trovati adatti anche per le leghe 950Pd. La fusione è stata surriscaldata di circa 80 °C prima della colata. La temperatura del cilindro era di 650 °C nella maggior parte delle prove
di colata. Le leghe contenenti boro possono essere sensibili al raffreddamento del cilindro. I cilindri devono pertanto essere lasciati raffreddare lentamente a temperatura ambiente prima della sgessatura.

Le nuove leghe si basano su 950PdRu e contengono componenti aggiuntivi per ampliare l’intervallo di fusione (Co, Sn, B), per ridurre la granulometria (Fe) e per incrementare la durezza (Al, B). La temperatura tipica del liquidus di tali leghe era di 1560-1570 °C, leggermente inferiore a 950Pd50Ru. L’aggiunta
dei summenzionati componenti incrementa la durezza dai 100 HV1 ca. della lega binaria morbida 950PdRu a 140-160 HV1. Si presume che tale intervallo di durezza costituisca la durezza ideale per l’incastonatura delle pietre e la finitura, in grado di fornire sufficiente resistenza ai graffi quando si
indossano i gioielli. Tale durezza viene anche raggiunta nel caso di leghe 950Pt da medio a dure di leghe di oro bianco senza nichel a 18k. Il colore delle leghe 950 Pd è paragonabile a quello delle leghe 950 Pt. Entrambi i gruppi di leghe mostrano un indice di giallo di circa 1, che è significativamente
più bianco delle leghe in oro bianco premium (YI <18). La densità di 950Pd è di circa il 40% e il 25% inferiore rispetto a 950Pt e all’oro bianco Pd a 18k rispettivamente. La bassa densità è un vantaggio per gioielli leggeri o massicci.

La lega binaria 950PdRu si contraddistingue per la bassa durezza. Le nuove leghe 950Pd danno prova di durezza superiore mantenendo al contempo una buona capacità di riempimento dello stampo, reazioni di microfusione e crogiolo basse e sufficiente resistenza alla frattura a caldo con condizioni di processo
adeguate. Possono pertanto diventare un’opzione per le applicazioni di gioielleria.

5 Ringraziamenti

Il presente lavoro è stato sostenuto finanziariamente da Norilsk Nickel, Russia. Si ringrazia Linus Drogs (AuEnterprises, USA) per il supporto al progetto e la consulenza durante la sua realizzazione. Un particolare ringraziamento va ai colleghi del fem per il loro contributo all’indagine SEM, alla metallografia
e all’analisi chimica.

6

7 Riferimenti

1.            Swan, N. and B.J. Williams.
Palladium – light, bright and precious – a world view. in
The Santa Fe Sympoisum. 2006. Albuquerque, USA.

2.            Battaini, P.
Investment casting behavior of palladium-based alloys. in
The Santa Fe Sympoisum. 2008. Albuquerque, USA.

3.            Fischer-Bühner, J., A. Basso, and M. Poliero.
Challenges for Palladium Casting Alloys. in
The Santa Fe Sympoisum. 2009. Albuquerque, USA.

4.            Fryé, T.
Palladium casting: an overview of essential considerations. in
The Santa Fe Sympoisum
. 2006. Albuquerque, USA.

5.            Battaini, P.
The working properties for jewelry fabrication using new hard 950 palladium alloys. in
The Santa Fe Sympoisum. 2006. Albuquerque, USA.

6.            Mann, M.B.
Palladium: manufacturing basics for servicing, assembly and finishing. in
The Santa Fe Sympoisum. 2007. Albuquerque, USA.

7.            Johnson-Matthey.
JM jewellery alloys. Available from:

http://www.noble.matthey.com/jewelry

.

8.            Hafner, C.
Palladium alloys
. Available from:

http://www.c-hafner.de/de/leistungen-und-produkte/edelmetall-produkte/werkstoffe/werkstoffe-fuer-schmuck-und-dekorative-anwendungen/
.

9.            Heimerle+Meule.
WKD_189-950. Available from:

http://www.heimerle-meule.com/fileadmin/dateien-global/02_Products/Schmuckhalbzeuge/Produktuebersicht/Werkstoffdatenblaetter/Legierungen/Pd_500_585_950/englisch/WKD_189-950eng.pdf
.

10.          Wieland. 2016; Available from:

http://www.wieland-edelmetalle.de/produkte/schmuckhalbzeuge/palladium/page.html?L=0
.

11.          Agosi.
Agosi palladium alloys. Available from:

http://www.agosi.de/wp-content/uploads/2015/09/AG_AgosiManufaktur.pdf
.

12.          Legor.
Legor palladium alloys. Available from:

http://products.legor.com/EN/download
.

13.          Hoover&Strong. 2016; Available from:

https://www.hooverandstrong.com/media/pdfs/MaterialSpecs-950Palladium-01-2015.pdf
.

14.          UnitedPMR.
950 Palladium Grain (PD950). 2016; Available from:

http://www.unitedpmr.com/palladium_950_grain.php
.

15.          Klotz, U.E., et al.
Platinum investment casting: material properties, casting simulation and optimum process parameters. in
The Santa Fe Sympoisum. 2015. Albuquerque, USA.

16.          Biggs, T., S.S. Taylor, and E. van der Lingen,
The Hardening of Platinum Alloys for Potential Jewellery Application. Platinum Metals Review, 2005. 49(1): p. 2-15.

17.          Kretchmer, S.,
Heat-treatable platinum-gallium-palladium alloy for jewelry. 2003, Palenville, NY, US.

18.          Heiss, T., U.E. Klotz, and D. Tiberto,
Platinum Investment Casting, Part I: Simulation and Experimental Study of the Casting Process.
Johnson Matthey Technology Review, 2015. 59(2): p. 95-108.

19.          Klotz, U.E., T. Heiss, and D. Tiberto,
Platinum Investment Casting, Part II: Alloy Optimisation by Thermodynamic Simulation and Experimental Verification. Johnson Matthey Technology Review, 2015. 59(2): p. 132-141.

20.          Böhm, W.,
Legierung auf der Basis von Platin, Palladium oder Gold. 2011.

21.          Blatter, A., J. Brelle, and R. Ziegenhagen,
Allaige à base de palladium. 2006, PX Holding S.A., 2304 La-Chaux-de-Fonds, CH.

22.          Henderson, S. and D. Manchanda,
White gold alloys.
Gold Bulletin, 2005. 38(2): p. 55-67.

23.          Fryé, T. and U.E. Klotz.
Mechanical properties and wear resistance of platinum alloys: a comparative study. in
The Santa Fe Symposium on Jewerly Manufacturing Technology. 2018. Albuquerque, NM, USA: Met-Chem Research.

Introduzione: il processo galvanico

Per processo galvanico si intende la deposizione di un metallo o di una lega metallica mediante un fenomeno di elettrolisi in cui l’energia elettrica sviluppata all’interno del sistema è convertita in energia chimica dando luogo ad una serie di reazioni di ossido–riduzione. Il risultato di questo fenomeno
prevede che la corrente elettrica che attraversa un sistema elettrolitico consente la riduzione di ioni metallici disciolti nella soluzione elettrolitica per formare un deposito metallico su un elettrodo. Questa tecnica, di fatto, consente di modificare le proprietà della superficie di un oggetto
ed ha dunque un utilizzo in ambito industriale per la protezione di strutture e oggetti metallici dagli effetti della corrosione. Non meno importante è anche lo scopo decorativo: nel settore orafo e fashion, si è soliti realizzare gioielli ed accessori moda depositando strati di diverso spessore
di metalli più nobili su metalli di partenza meno preziosi.

Il sistema più semplice possibile per effettuare un processo galvanico prevede (Figura 1):

Figura 1: Rappresentazione schematica di un sistema elettrolitico.

Generatore di tensione continua

: Costituisce il vero e proprio motore del bagno galvanico in quanto fornisce l’energia e la corrente necessaria per i processi di ossido-riduzione. Nello specifico, il generatore di tensione continua sfrutta un circuito raddrizzatore consentendo la trasformazione della corrente alternata di rete in
corrente continua con residui di corrente alternata inferiori al 5%.

Catodo

: Elettrodo negativo, sede dei processi di riduzione. Di fatto consiste nell’oggetto da galvanizzare su cui verranno depositati i metalli disciolti nella soluzione elettrolitica. Questi, infatti, sono ridotti all’interfaccia fra l’elettrodo e la soluzione. Il potenziale che consente la riduzione del
metallo al catodo è detto potenziale di deposizione. Se si conosce la distribuzione della corrente attorno al catodo, è possibile anche avere un’idea dello spessore di metallo depositato in ogni parte dell’oggetto da galvanizzare.

Anodo

: Elettrodo positivo, sede dei processi di ossidazione. Gli anodi possono essere attivi (o solubili) oppure inerti (o insolubili). Nel primo caso, il processo di ossidazione prevede il discioglimento del metallo costituente l’elettrodo che da stato di ossidazione nullo si trasforma in specie ionica disciolta
all’interno della soluzione. Nel caso degli anodi inerti, essi non partecipano alla reazione anodica ma svolgono soltanto ruolo di supporto all’ossidazione garantendo lo scambio elettronico alla loro superficie e di fatto la chiusura del circuito.

Bagno galvanico

: Soluzione elettrolitica all’interno della quale sono disciolti i sali dei metalli da depositare al catodo. Costituisce il mezzo che consente il trasporto della corrente mediante gli ioni presenti al suo interno. Il sistema elettrolitico sarà, dunque, costituito da un solvente (acqua nella quasi totalità
dei casi) che ha la capacità di ionizzare le specie disciolte al suo interno. Nello specifico, sono presenti i sali dei metalli da depositare e i sali conduttori, ovvero delle specie facilmente ionizzabili che sono in grado di consentire il trasporto di corrente nella soluzione mediante conduzione
ionica. La corrente dunque attraversa il sistema elettrolitico attraverso le specie ioniche disciolte all’interno e consente la riduzione al catodo dei metalli disciolti all’interno. Solitamente a completamento di un bagno galvanico, sono presenti anche ulteriori additivi inorganici o organici
che consentono di ottenere depositi di maggiore compattezza, lisci o brillanti, influenzando la struttura dei depositi.

•Parametri caratteristici di un processo galvanico

Ogni tipologia di bagno galvanico esprime al massimo le proprie performance se vengono rispettati una serie di parametri. Essi dipendono dal tipo di metallo o lega da depositare e dalla chimica che costituisce il sistema elettrolitico. Di seguito sono riportati quelli caratteristici:

Differenza di potenziale

: È il parametro mediante il quale viene fornita l’energia necessaria per il processo di elettrodeposizione. Ogni ione metallico ha un suo specifico valore di differenza di potenziale in seguito al quale avviene la sua riduzione e la sua conseguente deposizione al catodo. In linea di principio, i metalli
con il valore più negativo di potenziale standard di riduzione (Tabella 1) sono anche quelli più facilmente elettrodepositabili. Questi potenziali, però, sono valori di equilibrio mentre i processi galvanici sono intrinsecamente dei processi dinamici, oltre al fatto che spesso ci si trova con
parametri di temperatura e concentrazione diversi da quelli standard. Il potenziale al quale avviene la deposizione è detto potenziale di deposizione. Questo potenziale, varia con la concentrazione del metallo nel bagno e dipende anche dalla densità di corrente.

Tabella 1: Potenziali standard di riduzione delle specie chimiche più comuni.

Densità di corrente

: Molto più della tensione è questo il parametro più importante del processo galvanico. Essendo, quello galvanico, un processo dinamico, è la corrente generata dalla differenza di potenziale il parametro maggiormente connesso alla formazione e crescita del deposito metallico. Il vero parametro che determina
la quantità di elettrodeposito formato al catodo è la quantità di carica che fluisce durante il processo elettrolitico. Sicuramente un parametro migliore da controllare per gestire al meglio la quantità di carica che sopraggiunge all’oggetto da galvanizzare è la densità di corrente, ovvero la
quantità di carica che fluisce attraverso una unità di superficie in un’unità di tempo misurata in A/dm2. Si chiamano zone ad alta densità di corrente le parti dei pezzi trattati che ricevono più corrente rispetto alle altre. In genere, sono le parti più appuntite, quelle più esposte agli anodi,
le parti iniziali o terminali dell’oggetto immerso nel bagno galvanico. Le zone a bassa densità di corrente, invece, sono l’esatto opposto, dunque le zone centrali degli oggetti e le parti più nascoste.

Temperatura

: Sebbene in maniera minore, anche questo parametro contribuisce a fornire l’energia necessaria affinché avvenga il processo di elettrodeposizione. È un parametro legato alla cinetica del processo elettrolitico determinandone efficacia e velocità. La temperatura contribuisce a regolare conducibilità
e potere penetrante del bagno galvanico.

Tempo di trattamento:

Corrisponde al tempo necessario per depositare il metallo o la lega ed ottenere un deposito di buona qualità e dello spessore desiderato. Come facilmente intuibile, maggiore è il tempo di trattamento, maggiore sarà la quantità di metallo depositata. Per ogni processo galvanico è definito un tempo ottimale
di trattamento derivante da un compromesso fra qualità del deposito e quantità di metallo da depositare;

Efficienza catodica:

Espressa in milligrammi di deposito per Ampere-minuto (mg/ Amin), indica la quantità di metallo o lega metallica depositata in un minuto lavorando con una corrente di un ampere. Consente di capire quanto efficacemente è possibile depositare un metallo definendo una stima di quanta corrente è effettivamente
responsabile della formazione del deposito. L’efficienza catodica di un bagno dipende da molti fattori ed è variabile a seconda di temperatura, tensione, concentrazione di metalli ed additivi nel bagno.

È importante sottolineare che i valori dei parametri caratteristici di un bagno galvanico non sono stringenti ma generalmente è possibile definire un intervallo più o meno ampio di buona operatività per ognuno dei parametri precedentemente descritti.

Come effettuare un buon deposito galvanico

Prima di entrare nel dettaglio di come ottenere un buon deposito galvanico, è opportuno definire cosa si intende per deposito di buona qualità. È piuttosto intuitivo il fatto che la qualità di un deposito galvanico dipenda dalla particolare applicazione cui tale deposito è destinato. In alcuni casi,
ad esempio, può essere sufficiente che il metallo ricopra omogeneamente gli oggetti da trattare e che quindi sia sufficiente che il deposito abbia una buona adesione al substrato. Nel caso di depositi destinati al settore orafo-decorativo, alla condizione precedentemente descritta si deve aggiungere
anche l’assenza di porosità che conferisce al deposito un aspetto lucido e brillante ed è anche richiesta una buona resistenza alla corrosione. In altri casi si valuterà anche lo spessore e la durezza del deposito.

Per ottenere un deposito galvanico di buona qualità è senza dubbio necessario disporre di attrezzatura adeguata oltre che di prodotti di qualità ma spesso potrebbe non essere sufficiente. Nella maggior parte dei casi, le ragioni per cui non è stato ottenuto un buon deposito sono da ricercare o in un
non corretto rispetto dei parametri caratteristici dello specifico bagno galvanico o in una non corretta esecuzione della procedura di preparazione dei pezzi prima di effettuare il deposito galvanico desiderato.

Rispetto dei parametri caratteristici del processo galvanico

Per quanto concerne il primo aspetto, è infatti importante rimanere all’interno dei range di lavoro ottimali di ogni singolo parametro caratteristico del bagno galvanico al fine di ottenere un deposito di buona qualità. Non è certo che se non viene rispettato qualcuno di questi valori si ha sicuramente
un problema nel deposito ma sicuramente si è al di fuori della regione di massime performance della soluzione galvanica e ciò potrebbe portare all’incorrere di uno o più difetti nel deposito o, nella peggiore delle situazioni, potrebbe addirittura compromettere definitivamente il bagno galvanico
costringendo l’utilizzatore a dismetterlo definitivamente.

Sono di seguito riportati, parametro per parametro, gli accorgimenti più comuni finalizzati ad ottenere un deposito galvanico di buona qualità:

Differenza di potenziale

: È sicuramente questo il parametro su cui prestare maggiore attenzione assieme alla densità di corrente. Solitamente è definito per ogni processo un range di tensioni all’interno del quale è possibile ottenere un deposito di buona qualità.

Densità di corrente

: Per essere sicuri di non incorrere in alcun problema il parametro su cui fare totale affidamento per controllare il giusto apporto di energia necessaria a formare correttamente il deposito è rappresentato dalla densità di corrente. Lavorare con valori di densità di corrente all’interno dei range prestabiliti
garantisce sicuramente di fornire il corretto apporto di carica al catodo e dunque di formare un deposito dotato delle giuste caratteristiche chimico-fisiche. È possibile valutare qualitativamente e quantitativamente il range di densità di corrente ideale mediante test in cella di Hull o test
a catodo-piegato (Figura 2). Se si conosce la distribuzione della corrente intorno al catodo, si può avere una buona stima di come il metallo rivestirà l’intero oggetto: in quali parti lo spessore sarà maggiore o minore. Come mostrato in Figura 3, il deposito galvanico tenderà a formarsi e ad
accrescere maggiormente agli angoli e ai bordi perché zone di alta densità di corrente e molto poco nelle zone nascoste o in generale più lontane dall’anodo perché zone di bassa densità di corrente. Se si ottiene una distribuzione del metallo non desiderata, è possibile seguire degli accorgimenti
per migliorarla come modificare il modo in cui gli oggetti sono legati ai telai(Figura 4), modificare distanza anodo-catodo o sfruttare effetti di schermatura (Figura 5)

Figura 2: Esempio di un test cella di Hull ( sx) e di un catodo piegato (dx).

Figura 3: Rappresentazione schematica del modo in cui tende ad accrescersi un deposito galvanico. Nelle zone di alta densità di corrente (L) il deposito è in misura maggiore rispetto alle zone di bassa densità di corrente (D).

If the metal distribution obtained is not the one required, some precautions can be taken to improve it:

Modify the way in which the items are linked to the frames

: knowing that most of the current will accumulate in the corners of the items being plated and knowing that the current always tends to follow the shortest route between two conductors, in order to reduce the load accumulation at the corners of the objects, it is a good idea to place them as far away
from the anodes as possible. For this reason, if working in beakers, the recommendation is to place the objects parallel to the anode instead of perpendicular (Figure 4).

Figura 4: Per ottimizzare l’omogeneità del deposito galvanico è opportuno disporre gli oggetti da galvanizzare esattamente di fronte all’anodo.

Modify the anode-cathode distance:

the anode should be shaped so that the anode-cathode distance is as similar as possible for all surface points and slightly further away from borders thus favouring current passage towards lower current density areas.

Take advantage of shielding effects:

when working with several objects at the same time, the pieces can be linked so that the corners shield one another thus creating effective obstacles against load accumulation in these areas (Figure 5).

Figura 5: Rappresentazione schematica di un possibile effetto di scheramatura reciproca degli oggetti da galvanizzare disposti parallelamente all’anodo.

 

Temperatura

: Ogni bagno galvanico ha una sua precisa temperatura di lavoro. Essa è strettamente correlata alla conducibilità della soluzione e al potere penetrante del bagno, ossia la capacità di depositare omogeneamente metallo anche in zone a densità di corrente molto bassa. L’ eventuale diversa conducibilità
è un aspetto critico quando si co-depositano più metalli per formare una lega: una variazione di temperatura introduce una variazione della percentuale dei metalli in lega depositati. Di contro, lavorare con temperature eccessive potrebbe danneggiare i componenti chimici del sistema oppure accrescere
eccessivamente l’efficienza del processo con conseguente deposito di scarsa qualità. In definitiva, la temperatura è assolutamente un parametro da non trascurare e di conseguenza è richiesto un suo controllo efficace mediante termostati e termocoppie evitando eccessive fluttuazioni del suo valore.

Tempo di trattamento

: Per ogni tipo di bagno galvanico esiste un tempo minimo di trattamento al di sotto del quale il deposito metallico non è uniforme. Come facilmente intuibile, maggiore è il tempo di trattamento, maggiore sarà lo spessore del deposito ottenuto. È bene ricordare, però, che non tutti i bagni galvanici
hanno la capacità di produrre spessori elevati di deposito. Per questo motivo, soprattutto nel caso di bagni ideati per realizzare spessori di dimensioni inferiori a 0,5 micron, è definito anche un tempo massimo di deposizione oltre il quale il deposito potrebbe non essere di buona qualità. Nel
caso di bagni a spessore, viene solitamente indicato il tempo necessario per depositare un micron di deposito ed anche in questo caso può essere definito un tempo massimo di deposizione corrispondente allo spessore massimo di deposito che il bagno galvanico garantisce di ottenere con buona qualità.

Efficienza catodica

: Come già anticipato, non si tratta di un vero e proprio parametro da impostare quanto piuttosto una caratteristica intrinseca della soluzione che dipende da altri parametri. Tuttavia è un aspetto da non trascurare per ottenere un buon deposito galvanico in quanto fornisce un’idea sugli spessori di
metallo che la soluzione elettrolitica è effettivamente in grado di garantire con buona qualità. Un valore di efficienza basso, infatti, indica che il bagno galvanico è adatto per una messa a colore, un flash, e di conseguenza difficilmente sarà performante per effettuare spessori dell’ordine
del micron (Figura 6).

Figura 6: Immagine SEM della sezione di un campione in cui si osservano i diversi strati galvanici dei quali è stato misurato lo spessore.

2.2. Rispetto della chimica del bagn

In alcuni casi la distribuzione del metallo può essere migliorata agendo sulla chimica del bagno andando ad esempio a modificarne l’efficienza o la sua conducibilità mediante l’utilizzo di additivi.

Un bagno galvanico dovrebbe essere sempre mantenuto all’interno dei valori di riferimento di concentrazione dei suoi vari componenti. I motivi per i quali un bagno può modificare la sua composizione possono essere:

– Decomposizione delle sostanze chimiche

– Fenomeni di drag-in e drag-out

È raro che un bagno non richieda aggiunte. Dal momento che sono necessarie, il consiglio è quello di effettuarle frequentemente e in piccole quantità in modo che le sostanze chimiche non eccedano mai al di fuori dei range di lavoro. Aggiunte di grande quantità sono spesso sconsigliate per via di sotto-reazioni
accessorie non desiderate che possono incorrere oppure per eventuali eccessi di impurità che possono essere inclusi contestualmente alle specie chimiche da aggiungere nel bagno.

Non è un aspetto secondario anche la presenza di inquinanti di tipo metallico o organico. La prima tipologia è dovuta generalmente a cross-contaminazioni fra bagni oppure al disgregamento di parti di anodo e catodo o di eventuali altri oggetti metallici che possono venire in contatto con il bagno galvanico
o infine all’ utilizzo di acqua non propriamente demineralizzata. La contaminazione organica può essere dovuta ancora a cross-contaminazione con sgrassature e neutralizzazioni oppure alla semplice sporcizia che può accidentalmente presentarsi all’interno della soluzione galvanica o infine alla
presenza di residui di additivi degradati e non più funzionanti e all’utilizzo di acque contaminate. C’è il rischio, infatti, che simili contaminazioni possano essere incluse all’interno del deposito galvanico diminuendone di conseguenza la sua qualità. Per ovviare a questi inconvenienti, si
possono effettuare periodicamente delle filtrazioni della soluzione o trattamenti purificatori con carbone attivo o processi di dummy-plating. Nel caso di soluzioni elettrolitiche di elevato volume è sempre suggerito lavorare utilizzando un sistema di filtrazione in continuo mentre per piccoli
utilizzatori si possono effettuare filtrazioni su carta (Figura 7) fermo restando che un accorgimento molto utile è quello di coprire la soluzione o rintanicarla quando non la si usa per lunghi periodi.

Figura 7: Esempi di un filtro di carta usato per filtrare del precipitato ferroso (a sx) e di cartucce filtro utilizzate nei sistemi di filtrazione di impianti galvanici (a dx).

2.3. Rispetto degli step di preparazione

La qualità del deposito è anche dipendente dalla condizione dell’oggetto da galvanizzare e dalla fase preparatoria. Gli oggetti da trattare devono essere lucidati in modo da eliminare porosità e qualsiasi altra imperfezione superficiale prima di procedere alla deposizione galvanica. È perciò necessario
che gli oggetti da galvanizzare siano di buona qualità e accuratamente preparati prima di effettuare il trattamento (Figura 8).

Figura 8: Confronto fra due anelli di ottone non lucidato (a sx) e lucidato (a dx).

Durante la fase di preparazione dei pezzi da galvanizzare, si effettua la pulizia della loro superficie da ogni contaminante e l’attivazione della stessa in modo tale da ottimizzare l’adesione del metallo successivamente elettrodepositato. Gli step da seguire, teoricamente, dipendono dalla superficie
e dal tipo di lega di partenza su cui si vanno ad elettrodepositare i metalli. Rimanendo circoscritti all’ambito della gioielleria e del fashion, è possibile seguire il seguente schema standard per le fasi di preparazione degli oggetti (Figura 9):

Figura 9: Rappresentazione schematica degli step preliminari di pulizia e attivazione delle superfici degli oggetti da galvanizzare.

– Lavaggio ad ultrasuoni

– Sgrassatura elettrolitica

– Neutralizzazione

Ognuna delle precedenti fasi è seguita da un lavaggio e risciacquo degli oggetti.

–
Ultrasuoni: La pulizia ad ultrasuoni consente di eliminare i residui di grassi, oli e paste di pulitura dai pezzi da galvanizzare introdotte dalla procedura di lucidatura. Il principio di funzionamento è quello della cavitazione generata dagli ultrasuoni: la vibrazione degli elementi piezoelettrici
presenti nella lavatrice ad ultrasuoni produce onde ad alta frequenza che a loro volta generano all’interno della soluzione delle bolle che colpiscono la superficie degli oggetti ad alta energia provvedendo alla rimozione di contaminanti eventualmente presenti. Normalmente la soluzione contenente
il sapone per gli ultrasuoni lavora ad una specifica temperatura che favorisce lo scioglimento delle paste di pulitura in stretta combinazione con l’azione pulente esercitata dai relativi saponi e con l’azione meccanica esercitata invece dagli ultrasuoni stessi. Di conseguenza, affinché la procedura
di lavaggio ad ultrasuoni sia efficace, è necessario che la soluzione contenga dei saponi opportuni ed operi ad una specifica temperatura altrimenti l’azione sgrassante non è abbastanza efficace. ( Figura 10)

Figura 10: Fasi caratteristiche del processo di cavitazione tipico del bagno ad ultrasuoni.

– Sgrassatura elettrolitica:

Questa seconda fase preparatoria richiede l’uso della corrente elettrica. Oltre ad effettuare una ulteriore pulizia degli oggetti provenienti dal lavaggio ad ultrasuoni, questo processo consiste in una attivazione chimica della superficie da elettrodepositare. In seguito al processo di elettrolisi si
sviluppano bolle di idrogeno sui pezzi che garantiscono la contestuale pulizia ed attivazione delle superfici metalliche così da ottimizzare al massimo la successiva elettrodeposizione. La soluzione di sgrassatura è solitamente alcalina.

– Neutralizzazione:

La neutralizzazione è un semplice processo chimico tramite il quale vengono neutralizzate tutte le sostanze inquinanti ed incompatibili con i successivi processi galvanici. La soluzione deve essere chimicamente opposta a quella di sgrassatura. Poiché la sgrassatura è quasi sempre alcalina, la neutralizzazione
richiede una soluzione per lo più acida. Con la neutralizzazione, gli oggetti da trattare sono perfettamente puliti e la superficie è neutra e pronta per l’elettrodeposizione.

Lavorando quasi esclusivamente con soluzioni acquose, risulta chiaro che per ottenere un deposito galvanico di buona qualità sia fondamentale adoperare l’acqua corretta. La qualità dell’acqua in galvanica influenza fortemente il risultato finale del processo galvanico. Per questi scopi, l’acqua deve
essere esente da contaminazione organica e con un basso contenuto salino (inferiore ai 5 microsiemens). Per questo motivo solitamente gli impianti galvanici industriali sono dotati di colonne con carbone attivo e resine a scambio ionico. Per le soluzioni galvaniche, dunque, la scelta migliore
è quella di usare acqua deionizzata.

Cause di difetti su un deposito galvanico

Quando non è stato ottenuto un deposito di buona qualità si dice che esso manifesta dei difetti. Esistono una vasta gamma di imperfezioni che emergono sulla superficie dell’oggetto su cui si effettua una deposizione galvanica che ne inficiano aspetto estetico e proprietà chimico-fisiche.

3.1. Tipologie di difetti su un deposito galvanico

Provando a schematizzare, è possibile definire dei difetti puntuali, ossia poco estesi che si manifestano localizzati in maniera più o meno regolare sulla superficie del deposito galvanico; dei difetti estesi o di superficie, vale a dire quei difetti che interessano omogeneamente tutta la superficie
dell’oggetto trattato o vaste zone continue di esso; ed infine esistono dei difetti di adesione e di coesione relativi alla capacità del deposito galvanico di aderire al metallo sottostante e di rimanere integro superando le forze di tensione che necessariamente insorgono durante i processi di
nucleazione e crescita di uno strato galvanico su di una superficie.

Fra i difetti puntuali, quelli più comuni sono (Figura 12):

Figura 12: Esempi di diverse tipologie di difetto localizzati. In alto da sx: macchie scure sul deposito (cerchiate in rosso), macchie bianche sul deposito (cerchiate in rosso), macchie scure post-trattamento (cerchiate in rosso). In basso da sx: velature, bolle (cerchiate in giallo) e puntinature
(cerchiate in rosso).

Macchie scure sul deposito ( Burning spots)

: si tratta di macchie irregolari sulla superficie del deposito. Possono presentarsi al centro del deposito ma di solito sono più frequenti alle estremità degli oggetti, nelle zone di alta densità di corrente

Macchie bianche sul deposito ( Stains)

: Si tratta di macchie molto ravvicinate fra loro non necessariamente di piccole dimensioni.

Puntinatura ( pitting)

: Si tratta di micro-porosità generalmente di forma sferica concava presenti in maniera irregolare sul deposito.

Bolle e vescicolature ( bubbles/ vesicles)

: Si tratta di vere e proprie bolle, di accrescimenti di forma sferica che si generano sul deposito. Solitamente tendono a formarsi nelle zone di alta densità di corrente ma possono anche trovarsi in altre zone ed essere inglobate all’interno del deposito galvanico.

Striature ( streaking)

: Possono manifestarsi o come anelli concentrici che procedono dalle zone di alta a quelle di bassa densità di corrente oppure come strisce generate sempre a partire dai bordi dell’oggetto.

Velature ( hazes/ cloudiness):

Si tratta di zone casuali della superficie galvanizzata in cui il deposito è traslucido, nebuloso, come appunto ricoperto da un velo biancastro.

Macchie scure post-trattamento (post- oxidation):

Consiste nella comparsa di macchie immediatamente dopo aver effettuato la deposizione oppure nelle immediate fasi successive all’asciugatura.

Di seguito sono, invece, riportati i più comuni difetti di superficie: (Figura 13)

Figura 13: Esempi di diverse tipologie di difetto che interessano aree estese. Da sx: bruciature, deposito opaco, colore non uniforme.

Bruciature ( Burning)

: Si ha quando l’intero deposito o porzioni di esso presentano una finitura di grana grossa con aspetto spento e poco lucido, un deposito ruvido, rugoso e spesso molto poco tenace.

Deposito opaco ( Dull deposit):

Si tratta di un deposito non lucido e brillante in zone estese dell’oggetto e ben definite.

Colore differente o non uniforme ( discoloration)

: In alcuni casi si possono osservare zone dello stesso oggetto avente colori differenti oppure l’intero deposito con un colore più chiaro o più scuro rispetto a quello desiderato soprattutto nel caso di deposizione di una lega. Sono contemplati in questa tipologia di difetto anche i fenomeni di iridescenza
del deposito dovuti a spessori inferiori rispetto ai parametri suggeriti.

Basso livellamento (low levelling):

Caratteristico dei depositi a spessore, si ha quando il deposito non è disteso omogeneamente e si possono identificare delle discontinuità simili ad una serie di piani sovrapposti oppure simili a delle porosità non rivestite.

Basso potere di penetrazione (low throwing power):

Difetto che si manifesta con assenza di deposito o colore diverso in specifiche zone dell’oggetto da trattare, soprattutto zone di bassa densità di corrente (Figura 14).

Figura 14: Esempio di una catena ruteniata a bassa temperatura e tensione rispetto ai valori di riferimento. I problemi di penetrazione sono evidenti se si osservano le diverse zone della catena con assenza di deposizione e con deposito non omogeneo.

Infine si devono considerare i difetti a causa dei quali avviene il distacco dello strato elettrodepositato dal substrato. Si parla di difetti di adesione quando avviene immediatamente dopo la deposizione galvanica o addirittura contestualmente allo stesso processo galvanico. Solitamente esistono due
modalità secondo cui il deposito può distaccarsi: (Figura 15)

Figura 15: Esempi di difetti di adesione. A sx in alto sfarinatura di un deposito di rutenio, a dx in basso sfogliatura di un deposito di nichel.

Lack of adhesion is often due the objects being improperly prepared or to the absence of pre-strata underneath the final one, or the deposition parameters (specifically, temperature and tension) not being respected. However, when deposit loss occurs at a later moment, often following the application
of variable degrees of stress on the plated object, the defects are referred to as
cohesion defects
. At the time of their electro-plating, the metals are subject to tension forces which can be so intense as to spoil the deposit in two ways (Figure 16):

– Sfogliatura (peeling): quando il deposito si sfalda secondo delle lamine

– Sfarinatura ( blistering): quando il deposito si frantuma del tutto a formare una polvere fina, un deposito, di fatto, farinoso

Figura 16: Esempi di difetti di coesione in un deposito di nichel. In alto, presenza di cricche nel deposito in seguito a piegamento del campione; in basso, sfaldamento del deposito in seguito a piegamento del campione.

Quando, invece, la perdita del deposito avviene in un momento successivo alla deposizione galvanica, spesso in seguito all’applicazione sull’oggetto galvanizzato di stress più o meno intensi, si parla di difetti di coesione. Quando soggetti a stress, i metalli elettrodepositati sono soggetti a delle
forze di tensione le quali possono essere così intense da deteriorare il deposito secondo due modalità (Figura 16)

3.2. Cause più comuni di difetti

Le ragioni per cui un deposito possa presentare dei difetti sono molteplici e spesso uno stesso difetto può essere dovuto a più di un aspetto. Viceversa, una causa di difetto può manifestarsi in più di una tipologia di difetto. Fare un elenco dettagliato di tutti i possibili difetti e delle loro cause
senza contestualizzarli ad uno specifico processo galvanico può essere molto complicato e poco esaustivo, tuttavia è possibile, almeno in linea generale, raggruppare le cause di difetto in tre categorie:

Difetti dovuti al non corretto rispetto dei parametri:

In questa categoria sono considerati sia i difetti dovuti al non aver rispettato i parametri caratteristici suggeriti dalla scheda tecnica dello specifico sistema elettrolitico, sia i difetti dovuti all’utilizzo di strumentazione non adeguata come anodi danneggiati o diversi da quelli consigliati, cavi
parzialmente ossidati, strumentazione elettrica non adeguata ecc…

Difetti dovuti ad una procedura di preparazione non corretta:

In questa categoria sono considerati i difetti che insorgono in seguito all’assenza o alla non corretta esecuzione di uno o più step preparatori alla elettrodeposizione

Difetti dovuti all’utilizzo di prodotti non adeguati:

In questo caso si intendono bagni formulati con sostanze chimiche di bassa qualità o bagni galvanici che non hanno tutti i valori all’interno dei parametri di funzionamento (pH, densità, titoli di metalli ecc…).

3.2.1 Difetti dovuti al mancato rispetto dei parametri caratteristici del processo galvanico

I difetti dovuti al non coretto rispetto dei parametri di uno specifico bagno galvanico sono fra le cause di difetto più immediate e di più semplice risoluzione: basta, infatti, correggere il parametro per poter tornare ad ottenere depositi di buona qualità. Ripercorrendo i parametri tipici di un
bagno, dunque, sono indicate le cause più probabili di difetto:

Differenza di potenziale non corretta: Solitamente lavorare con valori di tensione troppo elevati o troppo bassi rispetto a quelli consigliati può comportare problemi di adesione nel deposito elettroformato, o possibile variazione del colore del deposito oppure bruciature. (Figura 17)

Figura 17: Doratura rosa eseguita a differenze di potenziale diverse. Lavorando con tensioni al di sotto del range (a dx) la lega si arricchisce in oro e il deposito assume tonalità più gialle rispetto alla corretta lega depositata lavorando con il giusto valore di tensione (a sx).

Densità di corrente non corretta

: Si verificano sostanzialmente gli stessi difetti osservati per valori non corretti della tensione, essendo i due parametri correlati.

Temperatura non corretta:

Lavorare a temperature non corrette produce differenti condizioni che possono generare effetti rilevanti come bruciature del deposito, problemi di adesione, diverso colore o diversa composizione della lega depositata (Figura 18)

Figura 18: Esempio di campione ruteniato a temperatura inferiore rispetto al range di lavoro. Nella zona centrale (bassa densità di corrente) non è presente deposizione alcuna, nelle zone periferiche (alta densità di corrente) si osserva deposito non uniforme.

Tempo di deposizione non corretto:

Sicuramente aumentare il tempo di deposito consente di avere spessori maggiori ma eccedere nei tempi può causare deposito opaco o problemi di coesione del deposito. Un tempo troppo breve può generare difetti di tipo puntuale o di superficie.

Efficienza catodica non corretta

: Non è un parametro di lavoro ma, come detto in precedenza, dipende dai precedenti parametri e influenza la qualità del deposito. Se il suo valore non è corretto, è possibile osservare problemi di adesione, velature o bruciature oppure uno spessore di deposito diverso dalle aspettative.

3.2.2. Difetti dovuti ad una procedura di preparazione non corretta

I difetti dovuti ad una non corretta procedura di preparazione delle superfici da galvanizzare oppure quelli dovuti ad un utilizzo di strumentazione non adeguata sono fra le cause di difetto più comune ed allo stesso tempo quella più trascurate: spesso, ritenendo non importante la procedura di preparazione,
si cerca di migliorare la qualità del deposito agendo sui parametri di deposizione o, in misura peggiore, intervenendo sulla chimica del bagno galvanico con il rischio di compromettere definitivamente il processo stesso.

Sono di seguito indicati i difetti più comuni associati ad una non corretta esecuzione delle fasi di preparazione degli oggetti da galvanizzare:

Lucidatura:

Se l’articolo da trattare è eccessivamente poroso, sicuramente il deposito galvanico non sarà in grado di eliminare tale porosità e non si otterrà un deposito omogeneo con probabilità di avere problemi di adesione ed eventuali difetti di tipo puntuale sul deposito (Figura 19)

Figura 19: Confronto fra un campione rodiato senza precedente lucidatura e un campione lucidato e rodiato.

Lavaggio ad ultrasuoni:

Sono i responsabili della rimozione di paste di lucidatura e contaminazione organica. Se presenti in superficie possono dare luogo a macchie sul deposito o problemi di adesione.

Sgrassatura

: È sicuramente lo step più importante nella fase di attivazione della superficie da elettrodepositare. Una sua non corretta esecuzione può comportare deposito disomogeneo o problemi di adesione e la maggior parte dei difetti puntuali (Figura 20)

Figura 20: Esempi di difetti osservati su campioni rodiati precedentemente sgrassati con sgrassatura non più funzionante.

Neutralizzazione e lavaggi:

Statisticamente sono le fasi maggiormente trascurate durante l’attivazione delle superfici da trattare. Una loro non corretta esecuzione comporta solitamente difetti di punto ed anche la possibilità di contaminare i bagni galvanici introducendo conseguentemente ulteriori cause di difetto generate da
contaminanti.

3.2.3. – Difetti dovuti all’utilizzo di prodotti non adeguati

Rappresentano le cause di difetto meno probabili specialmente nel caso in cui si utilizzano bagni galvanici pronti all’uso di comprovata qualità. Tuttavia è possibile che con l’utilizzo della soluzione possano introdurvisi contaminazioni di tipo organico o inorganico che a loro volta possono dar luogo
a difetti localizzati sul deposito o problemi di adesione.

Nel caso di bagni galvanici di elevati volumi, l’utilizzo del bagno e fenomeni di drag-out comportano il consumo dei costituenti della soluzione elettrolitica i quali dovranno essere ripristinati. Una loro mancanza produce una serie di difetti come bruciature o velature. È possibile anche che il consumo
dei costituenti del bagno possa provocare una variazione di pH della soluzione il che può causare oltre che difetti di tipo puntuale anche problemi di adesione o addirittura può determinare anche l’instaurarsi di fenomeni di precipitazione del metallo.

Le operazioni di ripristino dei costituenti del bagno galvanico devono essere effettuate seguendo attentamente le informazioni tecniche al fine di impedire aggiunte eccessive che possono generare ulteriori difetti.

Lo scenario macroeconomico per il settore orafo

Il terzo trimestre 2018 ha visto la domanda mondiale di gioielleria in oro registrare un andamento positivo, con un vero e proprio rimbalzo delle quantità (+6%) dopo i dati fiacchi o negativi dei trimestri precedenti. Secondo il World Gold Council, la debolezza dei prezzi dell’oro ha stimolato gli
acquisti, in particolare in India (in recupero dopo gli scorsi trimestri in calo), in Cina (favoriti anche dalla festa di Qixi, equivalente cinese di San Valentino) ed in molti altri mercati del Sud Est Asiatico. Ancora in difficoltà, invece, sempre secondo le statistiche diffuse dal World Gold
Council, i paesi del Medio Oriente.

Fig. 1 – Domanda mondiale

Fonte: elaborazioni Intesa Sanpaolo su dati World Gold Council – Gold Demand Trend

Il rimbalzo della domanda mondiale di gioielleria in oro, a fronte di prezzi in calo, si è riflesso anche nella dinamica delle esportazioni italiane. Nel terzo trimestre è, infatti, proseguito l’andamento negativo delle esportazioni di gioielleria in oro (-8,8% la variazione tendenziale dei valori
in euro), conseguente anche al confronto con un 2017 particolarmente brillante. L’evoluzione delle quantità ha, invece, registrato un significativo balzo in avanti (+35,7%), che sottintende bruschi movimenti nei valori medi unitari (VMU, valori divisi per le quantità).

Fig. 2 – Evoluzione delle esportazioni di gioielli in oro* (var. % tendenziali)

Nota: (*) Codice 711319. Fonte: elaborazioni Intesa Sanpaolo su dati Istat

Nel complesso dei primi nove mesi, le vendite all’estero sono diminuite del 4,1% in valore, mentre sono aumentate del 20,8% in quantità (per un ammontare pari a 28 tonnellate), implicando una diminuzione dei valori medi unitari pari al 20% circa nella media dei paesi.

Il dettaglio geografico evidenzia come il calo dei valori in euro sia diffuso a quasi tutte le principali destinazioni, con l’eccezione degli Stati Uniti (+3,2%), del Regno Unito (+27,2%) e del Sud Africa (+12%). Allo stesso modo l’andamento espansivo delle quantità ha riguardato tutti i paesi, con
solamente gli Emirati Arabi Uniti a registrare un calo del 15,2%, confermandosi come uno dei mercati dove le vendite di gioielleria in oro Made in Italy soffrono maggiormente.

Tabella 1 – Evoluzione delle esportazioni di gioielli in oro* nei primi nove mesi 2018

(var. % tendenziali)

 

 

Spicca il dato francese: gli invii di gioielli in oro verso la Francia nei primi nove mesi del 2018 sono cresciuti addirittura dell’86,6%, pari a 7 tonnellate in più rispetto allo stesso periodo del 2017, con un contestuale crollo dei valori medi unitari di oltre il 50%, che potrebbe anche riflettere
modifiche nella fissazione dei transfer price all’interno dei gruppi multinazionali francesi con basi produttive nel settore in Italia. Va infatti sottolineato come lo scorso anno i valori medi unitari impliciti nei flussi di export dall’Italia alla Francia avessero raggiunto livelli particolarmente
elevati rispetto al dato medio nazionale (con un balzo del 32%). Si tratterebbe, pertanto, di una sorta di “normalizzazione” sui livelli medi registrati negli anni precedenti. I dati relativi al terzo trimestre, poi, evidenziano un recupero di ritmi positivi dell’export verso la Francia anche
in termini di valori (+16%).

Anche nei confronti degli Stati Uniti le quantità di gioielli in oro esportate hanno registrato un notevole incremento (+51,4%, pari a 5 tonnellate aggiuntive), sottolineando il forte interesse di questo mercato per il gioiello italiano, in questo caso espresso anche dai dati in valore.

L’andamento positivo delle quantità vendute all’estero è coerente con l’evoluzione ancora fortemente in crescita dell’indice della produzione industriale del settore (che include anche la bigiotteria e l’argenteria). La produzione ha registrato nei primi 10 mesi una crescita dell’8,3%, con una
nuova significativa accelerazione nel mese di ottobre. Si tratta di un ritmo di sviluppo molto elevato (sebbene in rallentamento rispetto al +17,4% medio del 2017), nettamente superiore al +2,6% tendenziale registrato in media dalla produzione manifatturiera italiana nello stesso periodo.

Fig.  3 – Evoluzione dell’indice di produzione industriale (var. % tendenziale, dati grezzi)

 

Fig.  4 – Evoluzione dell’indice di fatturato

(var. % tendenziale, dati grezzi)

 

 Fonte: elaborazioni su dati ISTAT

 

 Fonte: elaborazioni su dati ISTAT

Anche l’indice di fatturato (che è in valore) è rimasto in crescita, sebbene in questo caso il rallentamento rispetto ai dati del 2017 sia stato tale da riportare il settore dell’oreficeria e bigiotteria sugli stessi ritmi del manifatturiero.

A livello territoriale, dove i dati sono disponibili solamente in valore e per l’aggregato che include anche la bigiotteria, i primi nove mesi dell’anno hanno confermato l’andamento in calo visto a livello nazionale. In particolare, nel complesso dei primi nove mesi dell’anno è stata Vicenza
a registrare la diminuzione più pronunciata dei valori (-4,8%), con un netto peggioramento nei mesi estivi (-8,1% la variazione tendenziale registrata tra luglio e settembre 2018 ed il corrispondente periodo del 2017). In forte frenata nel terzo trimestre anche l’export di Arezzo (-7,1%),
che porta il dato dei primi nove mesi a -2,3%. Rimane stabile sui ritmi (sempre negativi) della prima parte dell’anno l’export di Valenza Po (misurato come per gli altri poli produttivi con il dato relativo all’intero territorio provinciale).

Fig. 5 – Evoluzione delle esportazioni di gioielleria e bigiotteria (var. % tendenziali in valore su dati provinciali)

Fonte: elaborazioni su dati ISTAT

Nel dettaglio, per Vicenza, i dati negativi hanno riguardato molti mercati, in particolare quelli emergenti: Hong Kong (-16,4% nei primi nove mesi), Emirati Arabi Uniti (-19,2%), Giordania (‑32,4%), Romania (-22,3%) e Turchia (-12,8%). Spiccano, in particolare, i veri e propri crolli
di cui sono stati protagonisti durante i mesi estivi i valori esportati verso la Giordania (-56,4%) e la Turchia (-42%).

Da sottolineare invece come, all’opposto, gli Stati Uniti (primo mercato di sbocco nel 2017) abbiano registrato un andamento stabile nella media dei primi nove mesi, ma con un miglioramento durante il terzo trimestre. Segnali di recupero anche nelle esportazioni verso il Regno Unito (dopo
il risultato negativo del 2017), che con i dati del terzo trimestre sembrano aver arrestato la caduta; bene anche l’export diretto in Sudafrica, in netta accelerazione nel terzo trimestre.

Tab. 2 – Esportazioni del distretto orafo di Vicenza

Fonte: elaborazioni su dati ISTAT

Anche le esportazioni di Arezzo stanno continuando a risentire di cali importanti e continui negli Emirati Arabi Uniti (-17,9% tra gennaio e settembre, con un -27% nel terzo trimestre). Come Vicenza, poi, anche Arezzo ha visto un brusco peggioramento delle vendite in Turchia, con un ‑26,7%
maturato nei mesi estivi. Migliorano, invece, i risultati verso gli USA, tornati in territorio positivo tra luglio e settembre, ma non sufficientemente da riportare in crescita il dato complessivo dei primi nove mesi del 2018 (-8,2%). Rispetto a Vicenza, invece, Arezzo riesce a mantenere
su buoni livelli le vendite ad Hong Kong: dopo l’eccezionale +21,6% del 2017, l’export di Arezzo è cresciuto dell’1,8% nei primi nove mesi. Bene anche il ritmo di sviluppo del mercato francese (rimasto intorno al +20%); da sottolineare infine gli ottimi risultati di vendita, sebbene su
livelli limitati, che hanno interessato Panama ed il Libano.

Tab. 3 – Esportazioni del distretto orafo di Arezzo

Fonte: elaborazioni su dati ISTAT

L’export di Valenza Po è invece nettamente più concentrato dal punto di vista geografico, con quasi i tre quarti delle vendite focalizzate su Svizzera e Francia. Come evidenziato anche a livello nazionale, è in particolare il risultato verso la Francia a condizionare i dati complessivi. Nei
primi nove mesi del 2018 gli invii di oreficeria da Valenza alla Francia hanno sperimentato un calo del 20,9%, per poi migliorare in modo importante durante l’estate. L’export verso la Svizzera ha invece evidenziato nei dati più recenti un brusco peggioramento (-10,3%) dopo una prima
parte dell’anno più tonica, così come è avvenuto anche per l’export diretto verso gli USA (-9,8% nel terzo trimestre che lascia comunque il complesso dei primi nove mesi in territorio positivo, +3%). Male, dopo lo straordinario +55,2% del 2017, l’export diretto ad Hong Kong (-14,6%, in
peggioramento a -25,6% nel periodo luglio-settembre).

Tab. 4 – Esportazioni del distretto orafo di Valenza Po

Fonte: elaborazioni su dati ISTAT

Le prospettive per i mesi conclusivi dell’anno sembrano improntate ad un cauto ottimismo. Il risveglio della domanda mondiale, seppure in un contesto di elevata incertezza, porta a ritenere probabile il proseguimento, anche negli ultimi mesi dell’anno, delle tendenze riscontrate nella prima
parte. Ci aspettiamo che la domanda proveniente dai mercati asiatici e dagli Stati Uniti continui a mantenersi tonica, mentre dovrebbero proseguire le difficoltà riscontrate nei paesi del Medio Oriente.

Le prospettive per il 2019 appaiono più incerte, condizionate dal possibile recupero dei prezzi dei preziosi (si veda il prossimo paragrafo) che potrebbero avere un impatto negativo sui segnali di risveglio della domanda di gioielli registrati nella seconda metà del 2018. Nel nostro scenario,
comunque, il prezzo dell’oro dovrebbe rimanere – in media d’anno – intorno ai 1,250 USD all’oncia, un livello di poco inferiore rispetto a quello medio del 2018.

Responsibly Sourced Artisanal Gold: A Tour of Peruvian and Colombian Mining Communities.

What is ASM or Artisanal and small-scale mining?

Artisanal and small-scale mining can best be described as ‘low intensity’ mining, encompassing all types of mining from alluvial deposits to hard rock extraction1. Estimates vary, but in general it is accepted that ASM’s provide about 20% of the gold mined each year but employ about 80% of the mining
population. Operations tend to be informal, very low technology based and labor intensive. In many cases, miners use hand tools such as hammers and chisels, with the more advanced operations using nothing more complex than small scale excavation equipment such as back-hoes and dump trucks. ASM communities
also tend to be very marginalized and have little access to resources, particularly state or industry support. In many cases the need to mine is poverty driven and mining operations are often excluded from the normal banking sector, many having no bank accounts or access to credit. A sad consequence
of this is that many are subject to criminal activity in many forms.

The aim of both the Fairmined and Fairtrade organizations is to improve the above situation. By introducing standards which ensure that all gold and silver produced at artisanal and small-scale mining operations certified by the respective organization are mined in a responsible manner, they improve
the lives and working conditions of miners and mining communities. Once the mine becomes certified, they are not only guaranteed a fair price for the metals they sell, but also a premium to be paid directly to the mine by the purchaser for every gram of gold and silver purchased.

It must be noted that there is not enough ASM gold mined to fulfill all of the requirements for gold, whether for jewelry, electronics or medical uses. The point of responsible sourcing is not to eradicate industrial scale mining, it is to improve the lives and working conditions of miners and mining
communities in the developing world. Many customers, particularly those described as Generation X and Millennials, have different, less traditional buying habits. They want a story, something they can relate to, they want their purchases to make a difference socially and environmentally. This makes
responsible sourcing of gold and silver very relevant in today’s jewelry market.

Figure 1. Miners at the Sotrami mine, Peru.

Figure 1 shows a husband and wife who work at the Sotrami mine at Santa Filomena in Peru. This is a small scale mine employing 700 workers of which 160 are miners working at the face. Some mines are privately owned, and some, like Sotrami, operate as cooperatives while others are lone operators or single-family
operations.

Mercury use.

Both Fairmined and Fairtrade are concerned with mining communities in the developing world, and with this comes many problems with unregulated artisanal and small-scale mining. One of the biggest problems is the use of mercury. Extracting gold using mercury is a relatively inexpensive, simple and quick
process. Gold bearing ore is crushed into a sand like consistency and mercury is mixed in, usually with your hands and in many cases your feet as well. The resulting amalgam is then separated from the waste products and the mercury burnt off to leave gold typically of around 75% to 80% pure.

Figure 2. Using mercury to extract gold from ore
².

Figure 3. Mercury-gold amalgam
².

Figure 4. Burning off mercury
³.

Unfortunately, mercury is also highly toxic, and not just the liquid mercury. The fumes from burning off mercury are also highly toxic. Heavy and prolonged exposure causes irreversible damage to the human body. Mercury poisoning damages the brain, heart, lungs, kidneys and immune system. It results in
birth defects with many children born to mothers who have mercury poisoning being educationally subnormal with decreased intelligence. Mercury contamination also results in a toxicized ecosystem ‘ water, plant life, fish and animals are all affected. Because of bioaccumulation, mercury works its way
up the food chain and can bio-magnify and increase concentration in certain plant and animal species4. This is a major reason to use certified ASM gold and silver ‘ to guarantee that mercury has not been used in the gold extraction process.

Responsibilities for Mine Certification.

For a mine to become Fairmined or Fairtrade certified, certain criteria must be fulfilled (1,5,6):

Mines are required to participate in the social development of their communities.

Mines must eliminate child labor from their organization. No one under the age of 15 can be contracted to work in the mining organization and under 18’s must work in non-hazardous conditions.

There must be health and safety training for all employees and minimum health and safety standards must be met. Mandatory use of personal protective gear must be observed and enforced at all times, and working conditions continuously improved.

The mines must recognize and respect the rights of employees to form or join trade unions and collectively negotiate their working conditions – freedom of association and collective bargaining.

The responsible use of chemicals is mandatory. If mercury is used in the extraction process it must be controlled and plans/timelines agreed to eliminate its use altogether. If cyanide processing is used it must be handled using responsible practices. Chemicals have to be reduced to a minimum and wherever
possible eliminated over an agreed time period.

Responsible use of any premiums paid for under the Fairmined and Fairtrade certification schemes.

Why do Fairmined and Fairtrade metals cost so much?

The previous section describes the basic framework that mining organizations have to agree to and implement to comply to the standards and qualify for the premiums. From the jeweler’s point of view, a few of the questions most ask when looking to purchase responsibly sourced artisanal gold and silver
are why does it cost so much, where does this additional money go to and what is it used for? These are all very fair questions because certified ASM gold and silver is expensive and the jeweler pays considerably above market rate for the materials they purchase.

Normally, the gold and silver purchased by jewelry material suppliers is already in-country, they just call up their bullion supplier or refiner and buy it. However, with Fairmined and Fairtrade metals, unless your business is in Peru, Colombia, Bolivia, Mongolia or a few African countries, it isn’t
in-country and it has to be exported from the mine and then imported into your country of business.

Both Fairmined and Fairtrade use the same basic system. A fixed percentage of the gold or silver market price is agreed between the mine and the purchasing company, guaranteeing a fair price to the mine for their product. The Fairmined and Fairtrade purchasing procedures regulate this to prevent mines
being taken advantage of. As well as this agreed percentage of the gold or silver market price, the purchaser also pays a price per gram of fine metal purchased, and this is paid directly to the mine. All fees are paid prior to the metals being shipped from the mine. Additionally, there are transport,
insurance and logistics costs to get the metals from the mine to an airport, through customs export, onto a plane to your destination country, through customs import then from arrival airport to your plant.

This can be somewhat complex, however by doing this, the miners get a fair market value for the metals sold, and the payment of the premium directly to the mine which is then spent on both mine and community improvements. This premium typically, but not exclusively, gets banked and when it reaches a
certain amount, or after a certain time, the managers who run the mine or owners who run the cooperative decide what they should spend it on. They’re typically responsible in their choices, which are monitored, and listen to the community before deciding what to do but it is always spent for the good
of the mine and community.

Examples of how the premiums are spent range from safety equipment to medical needs, bringing mains electricity or running water to the community, or building schools and places of worship. In the case of one of the mines visited by the author, they decided the money would be best spent on mains electricity,
so they wired up the town. The next year they were looking into supplying mains water ‘all water has to be delivered by truck to the mine daily. But, after consultation, they decided to purchase a cell phone tower. Many of the mine workers do not permanently live on site but work there for three week
shifts, so communication with home and family is important to them, hence the cell phone tower.

Another item they purchased was a football pitch. They built a plateau out of waste rock from the mining operation and then had the pitch built on it ‘ they actually have two at this mine, one at about 1,900 meters and another at 2,750 meters. For the mine workers and community in general, this was an
extremely good spend. The workers now have a sporting activity they can do for both fitness and entertainment, and they have their own league that involves other mines, improving inter-community relations.

Figure 5. The cell phone tower at the Sotrami mine, Peru.

Figure 6. A football pitch at the Sotrami mine, Peru.

These are just two examples of what Fairmined and Fairtrade money has contributed to for a mining community. This money does good, helping to improve the work, the social environment and the community. This is the case with all Fairmined and Fairtrade mines.

In both Peru and Colombia there are two different types of ASM mining activity, these being regulated and unregulated mining activity.

Regulated Mining.

Mining accounts for only a small percentage of the GDP in Colombia however for Peru, about 15% of the GDP is from ‘on the books’ mining. There are no reliable figures regarding how much ‘off the books’ or ‘unregulated’ mining there is. The Government wants all mining activities to be legal and regulated
for a number of reasons. Primarily it will increase the tax base and help grow the economy. This tax base is used for many infrastructure projects and services such as building roads and construction, medical care and education to name a few. They also want to implement health and safety regulations
to safeguard the miners, and eradicate crime and exploitation. To encourage and incentivize miners to become legal and regulated, the government offers a 5% tax refund to a regulated mine for every kilo of gold exported.

Unregulated mining.

Unregulated mines are exactly as described and considered illegal. There are no rules and miners often work for little money under poor working conditions. Unregulated miners can die due to their working conditions, from poor health and safety, and the use of mercury and other chemicals. They are also
taken advantage of in many ways by criminals who exploit them, and unfortunately for many there is no way out of this cycle.

Unregulated mines are typically remote and difficult to reach. Those observed by the author in Peru were often high up in mountainous areas with no roads or tracks leading to them, and tens of kilometers away from the nearest paved roads. Everything needed for the mining operation has to be carried up
the mountainside either by donkey or by hand. Supplies such as food and water, tools and equipment, any wood for shoring up tunnels to make them safe, mercury to process the ore, all of these need to be transported to the remote areas where these mines exist. The mortality rate of the miners is relatively
high. Tunnels are unsupported ‘ they have to get wood to the general area first, then up the mountain, so often they don’t bother and the inevitable happens and tunnels can cave in. Problems with mercury poisoning are also high. In recent years at the request of the Peruvian government, the doctor at
Sotrami’s medical center tested blood samples from unregulated miners and found all to be over the safe legal limit for mercury. As a result of this the owners of the unregulated mines did not allow further tests to be carried out for fear of being closed down. The unregulated miners need the work and
so, as previously stated, get caught in the cycle.

Macdesa & Sotrami Mines, Peru.

The Macdesa and Sotrami mines lie within the Chaparra and Lucanas provinces respectively in Peru, about 600km south of Lima and roughly in between Nazca and Arequipa. Both of these provinces are in the Atacama Desert, which is the driest places on Earth. This is predominantly a mountainous and dusty
region, with the dust being the fineness of cement. It is quite a challenging environment in which to mine and to live.’

The Macdesa mine started life as a hole in the ground like all the other mines in the region. Many of these mines were left idle after Peruvian independence from Spain in 1821. When people started resuming mining, a relatively recent thing, the area was very much the ‘wild west’ with rival mining groups
fighting each other. Miners were barely making a living, safety was non-existent and life was very hard. As a consequence of this, at Macdesa, about 350 miners agreed that the only way to improve was to work together and form a community.

Macdesa is about 1,500 meters in altitude and at the beginning, the miners used hand tools and wheelbarrows to move ore from the mine. They would load approximately 60kg of ore from the mine into sacks and carry these to their donkeys. Once these animals were loaded up, they were walked 50 kilometers
down the mountain to the processing plant near the coast and main highway. However, the miners were taken advantage of and were only given about 80% of the market value. To compound this, the scales were rigged, the assays were rigged, and the ore was processed incorrectly. As a result, the miners rarely
saw any more than 50% of the value of the gold they mined.

So, in order to improve their yields and payouts, the miners started processing ore themselves by hand with mercury, but didn’t really know what they were doing. They determined that they needed to invest in their own technology and hire people who did know what they were doing. The story was very similar
at Sotrami. The original 350 ‘Socio’s’ ‘ founders – took out just enough wages to live by. The remainder of the money they received for their gold was reinvested to solve their problems and improve their processes and yields. Other things they did were:

Purchased proper mining drills.

Made rock and dirt roads to get trucks up and down from the processing plant to the mine.

Improved housing for mine workers and their families.

Invested in safety equipment and implemented a training program for all workers.

Improved conditions in the mine. Air was pumped into the mining shafts to help prevent pulmonary problems. The walls & ceilings were shored up to make the tunnels safe. Rail tracks were installed in each tunnel to enable them to easily move the ore out of the mine.

They began to assay on-site. This way, they knew exactly what purity their dore bars were and so what to expect as a return.

They developed processes to become mercury-free.

They purchased a boring machine, enabling them to test drill and see where the vein goes, greatly reducing the amount of mining needed and increasing their yield of gold recovered to ore mined.

The story for the Sotrami mine is similar. The founders formed the mine thirty one years ago and the village of Santa Filomena has grown around the mine at about 2,750 meters elevation. The population of Santa Filomena is ≈ 2,000, of which 700 work directly for the mine, with the remaining population
working support jobs. Similar to Macdesa, all the profits are invested back into the mine and the community. Workers work 20 days on, 10 days off which are unpaid, and usually eleven hour shifts. Many miners live away from the mine and travel home for their 10 days off to see family, but some have their
families with them, especially if both parents work at the mine.

Very often, local and national governments see mining as a kickstart industry. Thanks to the Mayor in the local town of Challa, who also involved the central government, 40km of asphalt road has been laid from the main Pacific Coast Highway, cutting the drive time to the mine in half. People in the lower
areas have taken advantage of the road and started crop farming, because now they can grow their produce and easily transport it to market without damaging it en route. In many cases where mining is a success story, its presence indirectly gives opportunities to others. The road is also important for
other reasons, one being that for Sotrami there is no running water onsite as such. All the water they need is transported up the mountain three times a day using three trucks.

Typically, nothing goes to waste. Rather than just discarding the rock, they use it for improvements. Waste rock is used to create flat areas which are then built on to expand the mine buildings and the town.

Figure 7. Waste rock used to create level building space at Sotrami, Peru.

Macdesa has three tunnels working their mine, whereas Sotrami only has one and all miners have to enter and exit the mine through this entrance and down this shaft. The mine is 630 meters deep, has 13 levels, and it takes the miners 35 minutes to climb down and 45 minutes to climb back up after completing
their shift.

Figure 8. The mine entrance at Sotrami, Peru.

Figure 9. The entrance shaft at Sotrami, Peru.

Figure 10. One of the mine tunnels at Sotrami, Peru.

Inside the mine, timbers are used to shore up the roof but generally there’s only lighting at the workface and for safety at the ore extraction rails. Although these timbers all have to be brought up from the coast because there are no trees growing in the vicinity of the mines, the good thing is that
the climate is so dry that there are no bugs to eat away at the wood and rot is generally not an issue.

Safety equipment must be worn by all when down the mines, particularly hard hats and especially respirators because the one thing that is a constant is the dust. The particles can cause serious lung problems if miners are unprotected. This is very different from the mines in Colombia, which are in an
entirely different environment.

Ore Processing: Hard Rock.

As far as ore processing is concerned both mines use similar processes. They both originally used mercury but understood the toxicity of this to themselves and the environment and so switched to a much more environmentally friendly cyanide process.

Figure 11. Large rotation barrel.

The ore is first crushed and then processed to the consistency of sand in a very large rotation barrel. Water is introduced to help the grinding process and also to form a slurry, which exits the barrel through a filter, making sure that the particles are small enough for downstream processing. Once
filtered, the slurry is pumped into cyanide treatment tanks. The gold present becomes suspended in the cyanide solution and is then pumped into reaction tanks containing activated carbon particles. The carbon particles are about the size of a grain of rice and they attract the gold, taking it out of
suspension. Typically, three successive tanks are used to maximize the gold yield; the process overall yields 96% of the gold entering the process as crushed ore.

At this point, the used cyanide solution is disposed of into cyanide waste pools. This may sound drastic, but the process is environmentally safe and not to be confused with the extremely environmentally damaging process of cyanide leaching, where the chemical is exposed to the ore in an untreated way
and comes into direct contact with the ecosystem with few or no constraints or containment. These cyanide waste pools have very durable membranes and the waste is chemically treated so that when it is exposed to ultra violet light it eventually converts to carbonates, making the waste pools non-toxic.
Once full, these pools are filled in then either planted or built on.

Figure 12. The slurry filter.

Figure 13. Activated carbon tanks.

Figure 14. A waste cyanide pool. The photograph is taken from a previous pool that has been filled, covered and planted.

Figure 15. Steel mesh electrodes in the plating tank.

The next part of the process is to filter out the gold laden carbon particles from the tanks, transfer these to an electrolytic cell and then plate the gold onto steel mesh. Once this part of the process is complete, the final part is to load the gold laden mesh into an oven and the gold is melted off
and cast into dore bars. The gold purity is now of the order of 80% and the product can be further processed or sold into the market as dore.

Spending the premiums.

As discussed earlier, premiums have been spent on various projects: mains electricity, running water, cell phone towers and football pitches. If the miners are asked why they do what they do and why they are happy to get certified by Fairmined and Fairtrade, there are numerous reasons. These include
environmental responsibility, better working conditions, prosperity, but they all agree that a major reason is for their children. These are relatively young enterprises and the mine workers want their children to have better lives and opportunities than they had. These miners came from virtually nothing,
created businesses and founded communities, and are flourishing.

A great example of this is their schools. At Macdesa, they have spent premium money on their kindergarten and elementary school, which are permanent block buildings, fenced in, clean, safe, and filled with computers. The children are taught how to use both the computers and the internet at a very early
age, and after school, the adults are also taught how to use computers, so they are put to very good use.

Figure 16. The elementary school at Macdesa.

Figure 17. Computers purchased with Fairmined and Fairtrade premiums.

The miners at Sotrami have used their premiums in similar ways. Sotrami has an excellent medical centre and this is very important not only for the mine, but also for the entire surrounding community. The mine funds this with their premiums, but they let anyone who needs it use it, regardless of whether
they are employed by Sotrami or not. They see it as their community duty to do so. It’s well equipped and they have a doctor assigned to them permanently. The next item on their wish list is an X-ray machine. Imagine having to travel 70km down a mountain on unpaved roads for X-rays on a broken bone.

The Peruvian mines discussed are success stories, and the same can be said for the Colombian mines.

Iquira, Coodmilla and Gualconda mines, Colombia.

The gold mines in Colombia discussed here are much smaller in scale than the Peruvian mines. The Iquira mine is in the Huila district, and the Coodmilla and Gualconda mines are in the Narino district, both areas being very lush and at altitude. Similar to the Peruvian mines, the mines in Narino are remote
and about a four-hour drive on unpaved roads from the nearest major highway.

Many of the same problems regarding regulated and unregulated mining in Colombia exist as noted for Peru, and until recently many areas of Colombia were dangerous and off limits to outsiders. There is still a great deal of illegal mining that the Colombian government wants to regulate, but implementation
has been a problem. Just telling miners they can’t mine anymore without providing any alternatives does not work. As a consequence, most communities don’t have an alternative and so go back to illegal mining just to survive. Colombia now has a formal ban on mercury but this is difficult to enforce for
various reasons. The Colombian government has also recently changed the banking laws as an anti-corruption and anti-terrorism countermeasure. A consequence of this is that several mining organizations have lost their ability to export their gold, however Fairmined is working with the mines and the government
to change this.

The Iquira Cooperative is southwest of Bogota in the Huila region of Colombia. This region is famous for coffee and a number of the miners at Iquira are also coffee farmers. The mines were on their land, they knew the gold was there, but they did not begin mining until 2004 when they organized as a cooperative
of 11 shareholders. By 2010 there were 35 cooperative shareholder members of which 8 are women and the co-op has 11 legally registered mines. Initially they would sell their gold informally to the local market, however becoming Fairmined certified has given them the ability to export.

Figure 18. Narino, Colombia
⁷.

Each mine is a separate business, but they all work as a cooperative for banking, selling and exporting purposes, resulting in better deals and reduced costs. These are relatively small mines and typically go about 500 meters into the mountain on one, two or three levels. About 20% of employment in Iquira
is by the cooperative and all of the workers come from the region.

Figure 19. A mine entrance at Iquira, Colombia.

The Colombian mines are also very different from those in Peru as far as environmental conditions are concerned. While the Peruvian mines are very dry and dusty due to the arid region, the Colombian mines are very damp. Water runs down through porous rock and hits a gold-containing quartz layer, which
is impervious, and so runs down the layer into the mine. The miners say they look for this because where there’s water, there’s gold.

The mines all have safety systems in place as part of their certification and qualification for the premiums. Each mine has a gas detection and alarm system to protect the workers, along with safe rooms in case of gas leaks or collapses. All mines and processing plants have first aid kits and all workers
have the required personal protective equipment. A number of mines in the cooperative employ women to handle the explosives and do the blasting, as well as work at the mine face.

Figure 20. A safe room in the mine at XXXX

Figure 21. Miners in Colombia
¹.

Figure 22. Sacks of ore waiting to be transported to the processing facility.

Figure 23. Sacks of ore in the crusher, Iquira, Colombia.

Once the ore is mined and bagged it is transported to the processing plant. Each bag contains approximately 50kg of ore which is processed in a similar manner to that used at the Peruvian mines, except on a smaller scale. Due to the nature of the deposits, after crushing, a flotation table is used to
separate out the heavier elemental gold and gold-containing particles. They do not use carbon rice but settlement tanks and let the gold rich solutions separate out by gravity before further processing into dore bars. There is a very similar process used at the Coodmilla and Gualconda mines in the Narino
district ‘ crushing followed by flotation table then cyanide treatment.

Figure 24. The flotation table at Gualconda, Colombia.

Figure 25. Gold separated from waste from the settlement tanks.

The Coodmilla cooperative, which is run as a non-profit organization, has been in operation for forty years and has four mine titles of which two have been certified by Fairmined. They have one hundred hectares of land covered by their titles and they are currently only working about three hectares.
Unfortunately, both Coodmilla and Gualconda are currently out of certification, but not due to mining practices. They are having difficulty meeting the new banking requirements outlined earlier. Fairmined is working with them and the Colombian government to resolve this.

Figure 26. A gold bearing quartz vein.

Figure 27. The Gualconda mine, Narino, Colombia.

Both of these mines are remote. Gualconda has been built into the jungle using the natural slope of the hillside to help with their processes. For Gualconda at least it has been a difficult journey up until this stage. Back in 1974 the mine was mined completely using hand tools and mercury, requiring
466 grams of mercury to process one ton of ore. Once used, this waste mercury was discarded directly into the river and therefore into the ecosystem. The mine has made great progress in cleaning up the mercury contaminated sites however there is still evidence of the mercury processing used that requires
cleaning up.

Figure 28. The mercury contaminated site at Gualconda.

Between 2001 and 2006 the mine was shut down due to armed conflict involving paramilitary forces and coca growing. Because of this, about one hundred families were displaced, leaving the area for the local city, but the miners state that life in the city did not suit them. In 2006 they formed their cooperative
but their ASM policy was in its infancy and very poor with regard to processing and environmental care. Between 2009 and 2013 they still did not have power for equipment, so they built a water mill using parts sourced from a junkyard. At this point they were still using mercury but had determined that
to be a responsible ASM source, they needed to phase out the use of mercury altogether. Initially, their policy was to reuse the mercury rather than just discharge it into the river, preventing environmental damage and reducing the quantity required to process one ton of ore from the previous 466 grams
to 25 grams. In 2015 they finally got electricity at the mine and this allowed them to redesign their processing plant, fully eliminating the need for mercury and adopting the more environmentally friendly and safe cyanide-based process that they use today. This gave them an increase of 20% in efficiency,
but also increased costs ‘ cyanide is more expensive than mercury when used to process gold bearing ore. However, the increase in efficiency and the knowledge that with Fairmined certification they would get better prices for their gold, plus a premium, made this process the chosen route to help them
fulfill their desire to be a responsible ASM source. After three years this is just coming to fruition.

The aim of the miners is not to stand still but to improve continuously, which includes decontaminating the remaining area at the mine where mercury was used for processing. This mine is now considered a model for responsible ASM mining and they have tours through every two weeks to see what they have
done and how they have done it. This is all within the backdrop of Colombia still having many social conflicts and corruption ‘ violence and drug trafficking is still prevalent in these areas. It is still the case that many mining titles are given to large mines but the smaller independent mines often
go ignored.

Why should you use responsibly sourced ASM gold?

There are many reasons to buy responsibly sourced and certified ASM gold. It may be a good way to get more customers into your store and so can be good for your business. If your customer base is made up of older generations, introducing a product line using ASM gold can be very attractive to both the
younger jewelry buying consumer and also customers who may not have considered gold jewelry, but making a purchase that makes a difference is attractive to them. There are jewelers who have decided to convert completely to Fairtrade and Fairmined gold, or at least as much as possible, but you don’t
have to do this. Every little bit helps and there’s absolutely no obligation, total conversion of every gold product you make is not necessary. Many jewelers who use ASM gold try one line or one collection to begin with, and if they have success, or the idea shows promise with their customer demographic,
they expand their offerings. The story is there: where it comes from, what the extra money is used for. It’s giving back to developing world c ommunities to improve their quality of life. This fact alone often gives customers a reason to buy jewelry made using ASM gold.

How can you purchase ASM gold?

The best way to find out where to purchase ASM gold is to contact Fairtrade or Fairmined (ARM), which can be done via their respective websites, and they will provide you with a list of suppliers and jewelers who are licensees and registered with them. Both organizations have similar systems: if you
want to use the name and mark, you have to be licensed. Depending on volume of business, you may have to pay to be licensed, undergo an audit, and pay a fee to the organization for every gram sold. If you do not want to use the name or mark, you can still purchase the gold ‘ there are no restrictions
on buying it. In this case you can still call it responsibly sourced gold from artisanal mines, you just can’t call it Fairmined or Fairtrade.

The aim of all businesses is to make money. However, this gold is relatively expensive and so to keep it as attractive as possible to the consumer, when pricing jewelry made from Fairmined and Fairtrade gold, it is advised not to mark-up the premium that has been paid to the mine. To keep the cost of
the piece as low as possible, ask the metal supplier to give you the cost of the gold purchased if it was not Fairtrade or Fairmined. This price can be used to calculate mark-ups and then the cost of the premium added in at the end. The more Fairmined and Fairtrade gold sold, the more money goes to
help improve the lives of miners and the mining communities.

Conclusions.

Where gold is sourced from to make jewelry is a personal choice for the jeweler. Suppliers of gold decide which direction they want to go with ‘ responsible or not – and so jewelers can also decide where and who to buy from. You have a choice. Jewelers who want to take part in the responsible sourcing
initiative can do so. There are three main sources to choose from:

Gold that has been mined responsibly but on an industrial scale. Where there is an audit trail, you know where the gold is coming from and you can inform your customers of this.

Gold that is produced from 100% recycled sources. This source of gold has already paid its environmental mining price, and by choosing this option you are getting, arguably, the most environmentally friendly gold supply by using what is already above ground.

Gold that has been mined responsibly from ASM communities. This choice directly helps developing world mining communities improve both their working conditions, their environment, and their lives in general.

Or, you can choose not to take part – it’s your choice. But consider the following:

• Being responsible can be good for business.
• All gold has been mined at some point.
• The overall aim should be to remove irresponsible mining from our industry.
• There must be transparency and traceability from mine and/or recycler to retailer.
• There is still a need for both industry and consumer education.

Finally, is ASM gold any different from any other gold? The answer, of course, has to be no, it isn’t. Gold is gold, wherever it comes from. However, it makes where it came from different, it makes how you get it different, and it makes the environmental and social price it has paid to get to you different.

References:

www.fairmined.org

Photo credit:
www.artisanalgold.org

Photo credit: Marieke Heemskerk

en.wikipedia.org/Mercury_poisoning

www.fairtrade.org.uk

www.responsiblemines.org

Photo credit: Fairmined Family

Potential and innovation of the selective laser melting technique in the manufacture of platinum jewelery

One of the revolutionary characteristics of selective laser melting is the possibility of processing materials that are notoriously difficult to use with other production techniques due to various melting mechanisms and modelling the alloys used in direct 3D metal printing. In the field of precious alloys,
this aspect is particularly interesting in the manufacture of platinum-based jewellery, a notoriously problematic element in the production phase due to its properties which include a high melting temperature of the alloys available on the market today and high reactivity with the materials it comes
into contact with. Consequently, production costs are higher, specific furnaces need to be used and, on average, the items produced often have defects. An assessment of the real revolutionary scope of using the SLM technique in making platinum jewellery was carried out by comparing SLM and traditional
production techniques, not only considering the technical aspects but also, and above all, the impacting economic/financial implications on the production line, in order to understand whether, and to what extent, introducing the SLM technique would lead to improvements in platinum jewellery manufacturing.

INTRODUCTION

In the article presented at the Santa Fè Symposium 2017 (1), we dealt with a general comparison between precious metal micro-casting and direct 3D printing processes (SLM) to understand whether this latter technique was effectively more advantageous than both classic and direct casting.

Among the production cases in which selective laser melting proved to be better, we identified the production of small series, the creation of hollow jewellery or items with complex geometries and no welding, or, when using difficult materials or those that are impossible to use in micro-casting.

Platinum jewellery production could be included in these cases since the casting of this material is famously more difficult than jewellery in gold or silver alloy (3). Moreover, the platinum jewellery market, despite renewed interest in this precious metal in the last 20 years, still has considerably
lower demand than gold or silver jewellery, therefore the machinery is not used to its full productive capacity.

In order to analyse whether, and to what extent, the SLM technique is competitive in relation to micro-casting in platinum jewellery production, we conducted a true-to-life production comparison between the two techniques in collaboration with Progol3D®️, a top reference for selective laser melting,
and Stilnovo S.r.l., a company in San Salvatore Monferrato (Valenza Po’ jewellery district), the producer of OEM jewellery and a reference for platinum casting.

The market segment chosen for comparing the two techniques was wedding and engagement rings since those are the most representative for platinum in the USA and Europe at the moment. The idea of eternal that has always been associated to platinum due to its resistance over time, means that this metal
is particularly in demand for wedding bands. This can be seen, for example, from data regarding the USA market in 2016 where, compared to a 10% drop in the production of platinum jewellery the year before, American spending on platinum engagement rings grew by 5% (2), making this segment even more predominant
than in the past.

WHAT HAS CHANGED COMPARED TO GOLD JEWELLERY PRODUCTION

Producing platinum jewellery with traditional methods notoriously involves more problems than producing items in gold, especially in regard to micro-casting. On the other hand, producing platinum jewellery using the Selective Laser Melting (SLM™) technique is not particularly any more difficult than
making items in gold alloy, which makes this technique interesting for platinum jewellery manufacture. In general, it can be seen that, for metals used in jewellery-making, the more difficult it is to produce with micro-casting, the easier it is to 3D print.

The greatest production difficulty found in micro-casting comes from differences in the thermo-physical properties of platinum alloys compared to gold alloys. First of all, the considerably higher temperature of liquid platinum alloys leads to the use of different heat-resistant materials for making
the moulds to withstand the greater temperatures. Instead of traditional materials based on calcium sulphate and cristobalite, other materials more resistant to high temperatures are required in which silica is blended with phosphate-based binding agents, which requires longer and more strenuous preparation
(4). Coating properties generally vary more drastically compared to traditional heat-resistant materials in the case of imperfect mixing, both in relation to components and processing times, making these materials more sensitive to storing and aging conditions, causing oscillations that are hard to
control in the surface quality and in the mechanical resistance of the cylinders (5).

Even if specific materials are used, cylinder resistance is still critical if they are heated to more than 900°C (6), a limit that leads to a much greater difference in temperature between molten metal and the cylinder during the casting phase compared to gold alloys with the metal consequently losing
heat faster on entry. This effect, together with viscosity and the surface tension of platinum alloys being greater than that of gold alloys, make it more difficult to fill the moulds completely, especially in the narrower areas, and it is therefore necessary to use centrifugal casting machines to partially
alleviate the problem (7). Increasing the centrifugal force helps the metal to fill the mould, but it also increases the chances of refractory fragment detachments which could then be incorporated into the metal when cooling. All these problems limit the quantity of metal that can be used on each tree
to a much lesser amount than can be used with gold or silver alloys, with the consequent reduction in production capacity. Filling problems and greater shrinkage in going from a liquid to a solid state (7) also mean that a more robust feeding system is needed, which, in turn, leads to a more unfavourable
ratio between scrap and pieces produced. A greater quantity of scrap implies higher production costs, which are further increased by the greater cost of refining platinum alloys compared to gold alloys, due to the more complex procedure and verification process. All these additional difficulties make
casting platinum jewellery more susceptible to variable results not to mention the need for more specifically skilled technicians.

The SLM™ process, on the other hand, has no particular problems compared to gold alloy production. In fact, the fundamental properties for metal-laser interaction, first and foremost, reflectivity and thermal conductibility, are more favourable for platinum alloys than for gold or silver alloys. This
means that less energy is required for laser melting and there is no need to add elements to the alloy to favour laser radiation absorption.

QUALITY COMPARISON

The quality comparison between platinum jewellery created through SLM™ and micro-casting was carried out by producing several ring models in Stilnovo’s BRIDAL series, a collection that best incorporates the concept of eternal associated to platinum jewellery since it comprises rings with the MULTISIZE
solution, covered by a patent (Application number 102017000104245, filed on 18 th September 2017).

The multisize patent is a system that leads to a new conception of the ring as an item that can easily change its diameter and therefore always be a perfect fit.

Changing the size of a ring has always been quite a problem for the jeweller as well as for the final wearer. Because a ring is long-lasting, and is sometimes even handed down from mother to daughter, it is quite possible that the need to change the size will arise sooner or later.

The operation is easy enough if the ring only has one mounted ring shank and a centre, but it becomes more and more complicated as shapes develop and is absolutely difficult when the whole shank is mounted: size modification, i.e. making the diameter smaller or larger is, in fact, obtained traditionally
by cutting the shank at the opposite side to the centre and adding or removing some of the metal. When the ring is mounted along the entire surface, it is dangerous to enlarge or tighten it by even just one size because the mounting may become insecure: in fact, changing the curve of the ring inevitably
modifies the diamond or precious stone setting, something with can compromise the reliability of the ring’s “hold” on them.

With the MULTISIZE RING solution, the internal part of the ring shank has a slot for an interchangeable B structure in various thicknesses (Figures 1, 2 and 3).

In our study, the A frames were made in platinum, while it was decided to use titanium for the sheet metal.

A simple KEY, a titanium hook, made into the shape of a treble clef, was used to pull out the interchangeable part from its slot in the fixed part when the size needed to be modified. Once the slot track in the fixed part A is empty, it is easy to position a new interchangeable part by hand and change
the size.

In order to compare micro-casting and direct metal printing, 10 models from the BRIDAL collection were chosen, comprising wedding bands, solitaires and trilogies, whose frames are shown in Figures 4 to 13. Production and the characteristics of the internal interchangeable parts were not taken into consideration
in this study since they were not made in platinum alloy but were preferentially made in gold or titanium due to the mechanical properties needed for the piece of the frame to be repeatedly inserted and removed without becoming deformed.

A wedding band model, called ETERNAL, which features 360° pavè (Figure 14), was initially chosen for the comparison but was later discarded due to the difficulty of removing the support required in SLM™ production.

6 rings of each model were made for both production technique in question of which 2 were to be sacrificed for destruction analysis, with the exception of the two wedding band models, of which three men’s size and three women’s size samples were made. The overall total of rings made for the study was
120 pieces of which 40 were to be sacrificed for destruction analysis. The list of pieces produced is summarized in Table 1.

In order to make the comparison more like a real production test, jewellery creation was divided between two producers: Stilnovo for micro-casting and Progol3D® for selective laser melting.  Each of the two producers is specialized in one of the two techniques being tested and is able to optimize the
process to obtain the best possible quality.

To assess quality differences given exclusively from the type of productive course and not ascribable to the different composition of the alloys used, the alloy 95PtGaInCu was used in both SLM™ and in micro-casting. Using the same composition for micro-casting and SLM™ made it possible not to give one
technique an advantage over the other thanks to the relative ease with which platinum can be melted by laser interaction so that no adjustments to the composition were needed for the SLM™ process, which would have been required had gold alloys been used. In fact, this composition is in the Progold range
as a micro-casting alloy and is also in use for SLM production at Progol3D®.

In regard to micro-casting production, the waxes were created with a Projet MJP 2500W 3D printing system using VisiJet M2 Cast wax. The cylinders were prepared with PRO HT Platinum Gold Star® keeping a water/plaster ratio of 33:100. The refractory firing cycle is outlined in Figure 15. Cylinder temperature
during casting was 850°C.

The cylinder plastering and firing phases were grouped as much as possible averaging between minimizing production times and the need to retrieve scrap.

For the melting and cylinder firing process, a Yasui VCC centrifugal casting machine was used with a casting temperature of 250°C above the alloy’s liquid state. After the cylinders were cooled, the refractory residues were removed from the metal by immersing them in hydrofluoric acid at room temperature.

A final sanding was carried out to complete refractory elimination.

As for selective laser melting, the jewellery was produced using a ReaLizer SLM50 laser printer equipped with a 100W fibre laser, collimated to a ray of 10 μm. The circular construction plate was 70 mm in diameter.

The layer thickness used for printing was 20 μm, favouring printing resolution over production speed, in consideration of the market segment involved in the study.

The printer was fed with 95PtGaInCu in powder form, obtained by gas atomization of the alloy and sieving to remove any coarse particles.

The shape of the powder particles was observed under a scanning electron microscope (SEM) and the particle size distribution was determined using a laser granulometer (Malvern, Hydro 2000S).

After the printing phase, the jewellery underwent shot peening to eliminate any partially melted powder on the surfaces which would cause the unrefined pieces to be rougher.

Both in micro-casting and direct metal printing, all the rings were re-fired to solubilize the alloy and reduce the internal tensions by furnace treatment at 1150°C for one hour, followed by rapid cooling in water.  In the case of wedding bands, the pieces were later hardened by furnace treatment at
650°C for an hour with slow cooling.

Whatever the production technique, every ring made was assessed using the following quality standards:

–          Surface appearance “as cast” or “as print”, impact of feed residues and supports.

–          Identification of any macroscopic non-conformity defects.

– Measurement of the internal diameter of the rings, variations to the nominal and deviations in measurements between rings of the same model.

On the two sacrificial samples for each model, the following was also carried out:

–          Measurement of surface roughness in both “as cast” or “as print” and after sanding or shot peening

–          Assessment of the internal quality by trimming and lapping the rings.

All the produced items that were not used for destruction analysis (altogether 40 micro-cast rings and 40 printed rings, subdivided into 10 models) were then polished and eventually mounted at Stilnovo for a final evaluation of the quality. The final quality assessment on the completed item was made
by Stilnovo’s internal quality control department which was not aware of the type of production technique used for each ring to be assessed. The standards normally adopted for high jewellery article control were applied.

At the same time, fundamental data were registered to compare the technological and economical aspects of micro-casting and direct metal printing, such as:

–          Production times

–          Production scrap

–          Technician impressions in the polishing phases

–          Technician impressions on the mounting

In order to correctly collect the data on finishing operations, an evaluation sheet, subdivided by phase, was attached to each ring and each technician was asked to complete it.

EVALUATION OF THE PHYSICAL, MECHANICAL AND TECHNOLOGICAL  CHARACTERISTICS

Surface appearance

The first comparison made between rings produced by micro-casting and by SLM™️ involved the appearance of the surfaces both when raw and after sanding or shot peening. This included assessing the impact on the surfaces of additional elements needed to create the item, in other words, feeders in
the case of micro-casting and supports in the case of SLM™️. The invasiveness and weight of these elements had direct repercussions on the quality of the rings due, for example, to the need to reconstruct the surfaces involved, and economically, because of being directly proportional to the percentage
of production scrap and process times.

This paragraph will evaluate the presence and invasiveness of feeders and supports in terms of the surface extension involved and the residue aspect, while the paragraph on the economic and financial repercussions will report the findings regarding scrap and production times.

Figures 18 to 25 compare the feeding and support systems of the 10 models selected for production.

From the comparison of the additional elements needed in production with the two techniques being examined, it was immediately obvious how the effect on the surfaces was completely different in the two cases. In micro-casting, where additional elements are considerable, the geometry of the directly fed
area of the jewellery item was totally lost, while in SLM™, the geometries below the residues of the supports, built as a grid, were generally visible.

Examples of support and feeding residues on the rings can be seen in Figures 26 and 27.

Support in SLM™ generally involves a greater surface area of the item, but, if the effective area of contact with the supports is taken into consideration, that is, the areas where the grid teeth actually touch the item and spoil the surface, the values are lower compared to the areas affected by feeding
in micro-casting.

There are cases, however, as in the example of the ETERNAL wedding band, in which, although the maximum geometry of the ring is maintained, the loss of detail due to the massive presence of support residue, makes production by selective laser melting, unsuitable.

For the solitaire 4 and trilogy 1 models, a good compromise was obtained in SLM™ by using a growth orientation that minimized the corner surfaces so that support was required, but with some supports in areas more difficult to reach than for other models when it came to removing them (Figures 28 and 29).
In these cases, a favourable use of support parameters leads to creating elements that are easier to detach thus partially compensating for the greater dexterity required for their removal.

In regard to the overall appearance of the surfaces, micro-cast rings were generally less rough in the raw state (example in Figures 30 and 31) and after surface treatment (Figures 32 and 33). However, surface irregularities were often observed, mainly between all the excess material burrs, which did
not appear with SLM™. These defects will be analysed in more detail in later paragraphs.

Roughness

To provide a quantitative evaluation of the differences between the surfaces, roughness measurements were carried out using a Taylor Hobson FTS INTRA 02 profilometer. The total roughness (Rt) of the profile was chosen as a comparison parameter corresponding to the difference between the highest and lowest
surface points. This value, in fact, represents the thickness of the precious material that must be removed in polishing to obtain an aesthetically satisfactory surface. The values were registered both for the items when raw (“as cast” in the case of micro-casting and “as print” in the case of SLM™)
and after surface sanding or shot peening.

Indeed, raw surface treatment is a production practice at both Progol3D®, by shot peening to reduce roughness and homogenize the surface appearance, and Stilnovo by sanding, mainly to eliminate refractory residues. The actual roughness that the jewellery-maker will come across in the roughing stages
is, in both cases, that of the treated item, and it will be from these values that the quantity of the material to be removed in order to obtain a smooth surface will depend.

Measurements were taken on several areas of the jewellery items corresponding to surfaces with various orientations in respect of the growth direction of SLM pieces and wax growth in micro-casting. Points with no obvious surface defects were measured in order to give an average Rt value net of macroscopic
surface irregularities.

In regard to wedding bands, the growth directions selected for 3D printing, for waxes in micro-casting and metal in SLM™ were the same, shown in Figure 34. Measurements were taken in direction 1 (surface parallel to the growth, direction perpendicular to z), in direction 2 (surface parallel to the growth,
direction parallel to z) and in direction 3 (surface perpendicular to the growth, direction perpendicular to z).

On the other hand, the solitaires and trilogies were printed with a different “standing” in SLM™ than the waxes in micro-casting due to the different type of support used. To be precise, growth was carried out with the pieces directed vertically in SLM™ and horizontally for waxes. In this case, measurement
directions were decided according to the growth direction, as shown in Figure 35 for SLM™ rings and in Figure 36 for micro-cast rings. The direction indicated with 4 therefore corresponds to a parallel surface in the growth direction, with the measurement taken perpendicularly to z, while 5 refers to
a variable surface due to inclination with measurement along z.

Table 2 shows the average values recorded on raw pieces divided by direction with the respective standard deviations, while Table 3 shows the values for items after sanding or shot peening.

The results are summarized in the graph in Figure 37.

As already noted in observing the raw surfaces, the roughness values are clearly greater in SLM™ compared to micro-casting. This is not a surprising result since surface roughness is one of the weak points of the SLM™ technique. In SLM™, the roughness registered is also greater on average than can usually
be found for gold alloys, a finding in line with the values reported in a study conducted by Progold® in 2015 (8), which noted how, compared to gold alloys, the greater presence of partially melted powder particles on the surfaces, leads to higher roughness on the raw item (Figure 38).

The higher degree of roughness registered in SLM™ in direction 3 compared to the other measurements is attributable to the surface progress caused by compounding the melting lines, which give a meniscus effect with greater height in the centre of the line and less height at the edges (Figure 39). In
the wax mould, the meniscus effect is much less pronounced (Figure 40), so much so that the roughness resulting from this effect, measured in direction 3, is much less than that caused by the subdivision into layers on the long z piece, which is the main cause of roughness in other directions.

The standard deviation recorded in SLM™ compared to micro-casting derives from an already higher deviation between different points of the same jewellery item belonging to equivalent areas.  These differences are mainly due to the different surface direction measured in respect of the movement that the
“wiper” makes during platform “recoating” (8), which results in the powder particles adhering differently to the surfaces.

The roughness on the items produced with micro-casting was, however, more constant both on individual items and in consideration of the various models.

Lastly, the effect of surface treatment, whether sanding or shot peening, on the roughness of the pieces in both production methods reduced the roughness values by about half compared to the “as cast” or “as print” condition.

The overall less surface roughness found in micro-casting generally implies that the jewellery-maker will have to remove less material in the roughing stage in order to achieve a compact surface. This is only true, however, if the piece has no areas with excess material, such as burrs, or spaces, like
surface dents. In these cases, the material lost and the processing time can increase considerably.

Defectology

Micro-casting

As already mentioned above, the jewellery items produced by micro-casting had a clearly higher incidence of macroscopic defects than those produced by SLM™, even after casting parameters were optimized.

The most commonly found defects were surface irregularities, such as burrs due to excess or lack of material.

In the first case (Figure 41), the cause was the partial rupture of the refractory so that cracks formed where they had filled with metal. This type of defect is generally very simple to correct since the excess material can easily be removed in a short time.

In some models, however, such as trilogy 1, the presence of details separated by tiny spaces, made this type of defect more critical, with cases like the one in Figure 42 where refractory rupture had caused different areas of the item to join up.

Phosphate-based refractory resistance variability, resulting from greater susceptibility to variations in storage conditions and the high temperatures of the casting metal, was the most probable source of other types of defect found.

The detachment of tiny portions of refractory led to the appearance of irregularities in some pieces in the form of cavities in cases where these micro-detachments became trapped in the metal (Figures 43 and 44), or of various-sized hollows whenever the micro-detachments were external to the metal and
created tiny round craters on the edges (Figure 45).

The high temperature of the metal, which causes reactions in the refractory, was probably responsible for the irregular surfaces and porosity found in some areas of the micro-cast jewellery items, like those in Figure 46 and in Figure 47, where the roughness was considerably greater than the average
of the surrounding areas.

In other items, surface defects seemed to have been caused by a combination of micro-detachments and refractory reaction (Figures 48 and 49).

The defects shown in Figures 43 to 49 were more damaging for the item compared to the previous since the problem was the lack of material rather than material in excess. In fact, this would have forced the technician to remove more material in order to obtain an even surface or to carry out repairs if
the cavities were deep, with a consequent greater loss of material and longer processing times.

Besides defects ascribable to metal-refractory interaction, problems were also found that derived from other production phases.

For example, the ovalling found in one of the micro-cast model 8 solitaires (Figure 50) was due to probable tension in the waxes or to problems in the plaster casting stage. Although deformed, in these cases, the jewellery-maker can quickly intervene to put the ring back into its original shape, practically
without altering the size so that this defect is of no particular consequence.

Another defect found was bent grips in the models where the grips were particularly long, especially in the model 4 solitaire. This problem (Figure 51), due in all probability to bending the waxes during plastering, can be resolved by adding a terminal ring to stop the grips from moving (Figure 52).

The rupture shown in Figure 53, on the other hand, was attributable to the mechanical stress that occurred in the cylinder cooling stage. In this case the ring was obviously non-compliant.

In order to further investigate the causes of rupture, the wedding band was sectioned horizontally and analysed under an electronic microscope.

In the internal part of the ring, where the fracture occurred, a cavity was found which was most likely due to refractory inclusion, given the results of the EDX analysis of the internal residues which highlighted the presence of silicone.

The cavity, which extended to both halves of the sectioned wedding band (Figures 54 and 55), had reduced the effective section of the ring thus drastically lowering the mechanical resistance, therefore the stress caused by the shrinkage in cooling exceeded the ultimate tensile strength causing the ring
to fracture.

SLM™

The macroscopic defects observed in the jewellery items produced with SLM™ were clearly fewer than those found with micro-casting. In fact, while the surfaces had a higher degree of roughness, only in the case of one ring produced was a real irregularity found in the form of swelling in one area of the
piece (Figures 56 and 57).

This type of defect occurs in SLM™ when the powder does not melt perfectly and so some partially non-melted particles remain and disrupt powder distribution in subsequent printing layers.

In this case in particular, since the defect only involved a small part of the item’s upper area, incomplete melting was probably caused by a variation in the average granulometry of the powder in the growth zone, due, for example, to the accumulation of agglomerates of partially molten particles within
the powder distributed by the “wipers” as the printing process continued.

Since the problem was material in excess and not a lack of it, correcting this type of defect was of no particular importance. However, it may happen that the swelling can be associated to widespread porosity in the area concerned, again caused by imperfect melting.

Dimensional coherence

An analysis of the correct nominal size and the deviations that could be found between the various same model rings was carried out on all the items produced by measuring the internal diameter, which can be directly correlated to the actual size of the ring.

For greater precision, the diameters were measured using a calibre (Mitutoyo), averaging three values in different positions, and also by means of photographic analysis using a Keyence digital microscope, suitably calibrated for maximum measurement accuracy.

Table 4 shows the data relating to the internal diameter of the rings. In the averages calculated for micro-casting, the ovalized ring in Figure 50 was not considered due to the difficulty of establishing the real diameter.

From the data obtained, it can be deduced that the internal diameter measurement in relation to the nominal value was always less in SLM™ compared to micro-casting for each of the ring models produced.

The origin of this reduction in internal diameter is obviously different for the two techniques: in SLM, it is caused by an imperfect correction of the width of the single laser trace while in micro-casting, it is caused by refractory shrinkage during the firing stage, from metal shrinkage as it goes
from liquid to solid and from the item’s contraction during cooling at room temperature.

In the case of SLM™, using platinum instead of gold was not a variable that could have affected dimensional variations, while in micro-casting, the higher temperatures and more notable shrinkage during phase change could have been cause for greater discrepancy in the nominal size for platinum rings rather
than gold rings. Repeatability on rings of the same model was generally greater in SLM™, with maximum standard deviations of ± 0.03 mm compared to ± 0.04 mm and more found in some micro-cast models. Given the greater oscillation found in micro-casting, any correction upstream of the internal dimensions,
by altering the design, for example, would be less effective.

Internal porosity

To analyse the porosity inside the items, the first technique considered was computerized tomography, a technique that has the advantage of not being destructive and able to investigate the entire volume of the jewellery item. However, the results obtained were not deemed satisfactory in terms of image
resolution, a problem caused by the high density of platinum which caused such elevated absorption of the beam, that analysing the thickness of the rings was extremely imprecise.

As an alternative to tomography, it was decided to make a direct analysis of the ring sections by cutting two out of the six rings produced for each model. In order to acquire a more complete evaluation of the internal volumes of the rings, sections from different areas of the items were analysed. To
be precise, one ring of the sacrificial pair was sectioned in the A plane represented in Figure 60, while the other was sectioned in the B planes (Figure 61), perpendicular to the first, in four different areas of the ring. After resin incorporation and lapping, the sections were photographed at 50X
to digitally analyse porosity using the software inside the Keyence microscope that had been used to take the images.

Figure 81, while Table 5 shows the percentage porosity values found in the various models considering porosity on the A and B planes of each ring, weighed along the entire surface of each analysed section.

The level of porosity found in the pieces can be quantified as medium-low in both production techniques with lower values in SLM™ compared to micro-casting, which, on average, had a twice as high porosity. For both cases, there was a notable variability between different pieces and between different
areas of the same sample, with sections that had practically total density and others that suffered from higher porosity.

In micro-casting, areas with singular high-volume porosity were observed, like, for example, the cavities in Figure 82, as well as a high number of small porosity clouds, as in the case of shrinkage porosity shown in Figure 83.

The porosity found in SLM™ was not in the form of cavities but single spherical pores (Figure 84), probably caused by gas, or areas with tiny, regularly placed spaces (Figure 85) due to imperfect melting between adjacent laser tracks.

Besides percentage porosity on all the items, locating any pores is also extremely important in jewellery-making: pieces with a dense interior but surface porosity are more difficult to finish than those that are more porous overall but have a more compact surface.

From this point of view, it can be seen how the porosity found in some areas of the SLM™ items was mainly inside the pieces and more rarely on the surface areas. This effect derives directly from the melting modality and item growth. Inside one single “layer”, the external surface is, in face, melted
as one single outer layer and the laser parameters are optimized in order to ensure the almost total absence of porosity in each laser track. The inside is then melted with parallel laser scans. Porosity tends to gather at the joints between the internal scans or between the outer and inner layers,
which are generally at least 150-200 μm from the surfaces, in an area that is unlikely to be removed in the polishing phase. In micro-casting, porosity distribution is more varied: there are surface cavities, visible also macroscopically on the external surfaces and mainly due to tiny refractory
detachments, and shrinkage porosity, which appears to concentrate more inside the items.

Lastly, the case of fracture found in one of the micro-cast wedding bands is to be considered: in this case, porosity, although concentrated inside the ring, was so extensive that it compromised the item’s mechanical endurance.

Metallographic appearance

To evaluate the dimension of the crystalline granules in the micro-cast rings and in those printed with SLM™, acid attacks were carried out on model 1 “as cast” and “as print” wedding bands.

Comparison confirmed what had already been seen in the past for samples in gold and platinum alloys: the average size of crystalline granules was drastically greater for micro-cast items (Figures 86 and 87) compared to SLM™ items (Figures 88 and 89). Therefore, in reality, signs of melting traces could
be made out but not of individual granules, even at high magnification.

The SLM™ sample showed the presence of micro-cracks, made visible by the acid (Figure 90). Mechanical tests, reported in the paragraph below, were carried out also to evaluate the effective impact of this defect on the properties of the SLM™ pieces.

Mechanical characteristics

The mechanical characteristics of jewellery, such as hardness, elongation and ultimate tensile strength, have direct repercussions not only on the item’s mechanical resistance but also on technological parameters, such as mounting and polishing.

For this reason, the mechanical performances of the items produced by micro-casting and SLM™ were compared, the alloy used being equal. Micro-hardness tests were carried out on “as print” or “as cast” model 1 wedding bands, both after re-firing (1 hour at 1150°C) and after hardening (1 hour at 650°C),
using a Vickers FUTURE-TECH hardness tester. The ultimate tensile strength (UTS) and elongation (A %) values were, on the other hand, obtained from traction tests carried out with an INSTRON dynamometer on specifically created specimens, shown in Figure… In this case the values were measures on “as
cast” or “as print” specimens and on specimens subjected to re-firing treatment, to evaluate possible mechanical differences that may affect the mounting phase.

The greater hardness found for “as print” items compared to “as cast” ones was more than likely due to the smaller dimensions of the crystalline granule in SLM™ and to higher internal tensions in printed items. The re-firing treatment, which, in the case of the alloy used, had the double effect of lowering
the samples’ internal tensions and of solubilizing, meant that, in both cases, hardness could be lowered to below 190 HV, therefore making mounting possible. After aging, in both cases, hardness increased considerably, although it was greater for micro-cast pieces, for which resistance to wear and tear
could therefore be greater than for SLM™ pieces. The difference observed could be caused by the presence of the micro-cracks seen in the SLM™ wedding bands when tested with acid, which favours indenter penetration into the sample.

In regard to traction tests, the samples printed by SLM™ had a greater ultimate tensile strength in the “as print” state than the “as cast” samples, to the expense, however, of lower ductility.

After the re-firing treatment, the ultimate tensile strength lowered for both types of sample but was still higher in the case of SLM™. The results of elongation at fracture, on the other hand, show an inversion between SLM™ and micro-casting. In fact, although ductility increased with thermal treatment
in both cases, the increase in SLM™ was considerably greater.

After re-firing, the samples produced by SLM™ therefore had higher UTS and elongation at fracture values, a fact that indicates how the micro-cracks observed after metallographic attack on printed samples can probably be attributed to mechanical properties rather than granule size and to possible internal
defects in the micro-cast samples.  Better performances after re-firing suggest a better behaviour of items during mounting.

Finishing: technician impressions

The impressions of sector technicians play a fundamental role in the possible success of a new production technique. Jewellery production is not exempt from this rule: even though the quality of a product may seem to be excellent according to technical analysis, if, during the processing phases, it does
not ‘convince’ the workers, with all probability, the production technique will not be adopted in the future. For this reason, we considered it essential to collect the opinions of the “jewellery makers” involved in item finishing in order to be able to add more subjective, but equally important to
the overall assessment of a new production method, evaluations to the quantitative ones, like times and losses due to finishing. The 80 rings created and not subject to destructive tests, were therefore sent for finishing and evaluation. Each of the working phases was carried out by the same technician
both for items created by micro-casting and by SLM™ in order to have the same judgment gauge for both techniques.

The first phase of the finishing process is eliminating any added element residues from the rings that were used in their production but which are not part of the item, i.e. feeders in micro-casting and supports in selective laser melting. As can be seen by the opinions expressed, summarized in Figure
87, from this point of view, the SLM™ technique suffered from some models having supports that were more complicated to remove, like those used for the model 4 solitaire in Figure 28 and in the model 1 trilogy in Figure 29. Removing supports from the internal areas of the item required greater dexterity
on the part of the technician and increased the probability of the item being spoiled in this phase. Substantially similar results were observed in evaluating the difficulty of roughing (Figure 90). In fact, opinions generally regarding the roughness and compactness of the item’s surface revealed that
rings produced by SLM were 80% less difficult to work than those made by micro-casting. The only significant difference was the presence of a non-compliant micro-cast ring identified in this phase, which had caused particular roughing difficulties before it was definitively put aside.  The results obtained
with this evaluation are particularly interesting if one considers that one of the weak points of selective laser melting is high surface roughness.

The impressions reported by technicians on this point revealed that, at least in the case of platinum, all that was needed was slightly greater pressure or thicker paper during this preliminary phase to eliminate the additional surface roughness with little additional effort compared to micro-cast pieces.

This extra effort was, however, amply compensated by the quality of the SLM metal that the technician found (Figure 94): the percentage of surfaces evaluated as excellent in SLM™ in terms of compactness was, in fact, close to 100%, while in micro-casting, evaluations were more varied with only about
60% of surfaces considered as excellent, 25% as average, 10% as low quality due to evident porosity, and two rings were judged as non-compliant.

No particular difference was observed in the cleaning phases (Figure 95), while in mounting, the overall evaluations in both cases went from low to none (Figure 96). The mechanical properties of the metal therefore resulted as more than good both for the micro-cast alloy and the printed alloy.

Quality Control: evaluation

The Quality Control judgment is fundamental for understanding if the jewellery items produced conform to residual porosity criteria and the aesthetics defined by high jewellery.    We therefore subdivided the rings into those that directly passed verification, those that needed quick laser repairs in
order to be compliant and those that were judged as non-repairable.

There was considerable diversity in the results obtained in SLM™ and in micro-casting: while three quarters of the printed rings immediately passed the checks, only half of those produced by micro-casting achieved the same result (Figure 97).

The judgments expressed by Quality Control confirmed the data obtained from analysing the macroscopic defects and internal porosity of the sacrificed samples: the items produced in platinum by SLM™ were immediately found to be less faulty than those produced by micro-casting.

In regard to non-compliances, no SLM™ item was found to be such compared to two items made through micro-casting and submitted for finishing. Furthermore, one sacrificial wedding band was found to be non-compliant due to breakage.

The final appearance of the ten ring models after mounting and polishing can be seen in Figures 98, 99 and 100 for micro-casting and in Figures 101, 102 and 103 for 3D printing.

ECONOMIC AND FINANCIAL ASSESSMENT

Semi-processed production times

Both micro-casting and selective laser melting production processes were organized to reflect the timing and subdivision into typical phases of real-life production. Micro-casting production was subdivided onto 11 trees, listed in Table 8. Cylinder firing cycles, which are the longest production phase
for micro-casting, were grouped in order to obtain the best compromise between production times and scrap recovery. In order to imitate what happens in real production, it was, in fact, decided to re-use the casting scrap, adding it to new alloy to compensate the percentage of material used for producing
authentic jewellery. This procedure is normally carried out to limit the quantity of precious metal needed, both because of the cost of the raw material itself and for the cost of refining the scrap. To be precise, new alloy was used for the first group of four trees while the scrap added to new alloy
to reach the weight of the tree to be cast, was used for the next three trees and the last four.

In SLM™, production was distributed over 7 printing plates (Table 9), created in descending order according to the height of the items to be produced. In fact, this optimizes the use of the powder by producing the highest items first for which the quantity of powder needed to fill the printing space
is greater.

The average times for each cylinder and the total times that the machinery was in use, subdivided by phase, are shown in Table 10 for micro-casting. Table 11, on the other hand, shows the average times for each printed plate and the total times that the machinery was in use, subdivided by phase, for
SLM™. The time it took the technicians to carry out production was also recorded. In fact, a higher total of man hours not only increases production costs, it also implies less possibility of having an automated production process.

From the data shown, it can be noted how machinery usage times are lower (-20%) in micro-casting compared to direct metal printing. In both cases, one production phase required a longer usage of machinery compared to the others: in micro-casting, firing the plaster casts took 55% of the total production
time, while in SLM™, the printing phase actually took 85% of the overall time. Both phases, however, do not require any intervention from technicians and only weigh on production costs in terms of machinery usage and electricity.

Looking at man hours however, the situation is the opposite: despite the longer machinery times, the SLM™ technique required less technician time than micro-casting (-20%), and is therefore more inclined towards automation.

Another important fact for evaluating a production technique is undoubtedly the total production time, considered as the actual time it took to create a jewellery lot. This timing takes into account the effective daily hours (8 hours subdivided into two groups of 4 with a 1-hour break in between), the
weekly working days (5) and which processes can continue during the night because no human supervision is required. Also taken into consideration are processes that can be done at the same time as well as any waiting times, such as the time needed for pickling in acid or for partially drying the cylinders,
which do not involve machinery or technicians, but are part of the production process all the same.

The time divisions for micro-casting and SLM production phases are shown in Tables 12 and 13 respectively, considering the production sequence actually used for making rings. This includes subdividing micro-casting into 3 groups of cylinders in order to be able to use less precious material by re-casting
scrap, and the creation of 3 SLM™ plates with the lowest last instead of inserting them with the others to save time, in order to be able to use less powder in the machine at the beginning.

The total production time was equal to 5 working days for micro-casting and 5.5 for SLM™, which was therefore slightly slower. It should, however, be considered that the operations done on the sixth day in SLM™ do not prevent starting the production of a second lot because they can be done at the same
time. This means that, in the case of consecutive lots, the production capacity of 60 rings, equal to those made for this study, with the subdivision between phases, can be considered the same in the two cases.

Finishing times

The overall finishing times are shown separately in terms of semi-processed production times because the finishing needed the same phases for the items made with micro-casting as those made with SLM™. The discriminating factor in this phase was therefore the ease with which residues from feeders and
supports could be removed and the quality of the items in terms of surface roughness and compactness and of residue porosity. In fact, generally speaking, external porous surfaces or particularly irregular surfaces force the technician to remove more material in order to reach more compact areas of
the item with the consequent increase in working times and finishing loss.

After analysing the times needed to remove feeder and support residues, it can be noted that the former, on average, was done faster given also the relative simplicity of the ring geometries in the fed areas. The average time required in this phase was also more homogenous in micro-casting, while in
SLM™, variability grew depending on the positioning of the supports with longer times required for models where technicians indicated residue removal as more complex on their work difficulty evaluation sheet.

However, observing roughing times, it appears that, except for a few cases, the rings produced with SLM™ required similar or even shorter working times than those produced by micro-casting. This fact agrees with the impressions on the complexity of this phase shown in Figure 93 which, on average, saw
SLM™ rings as easy to work as the micro-cast ones but with a better surface quality. Polishing phases did not reveal any substantial differences between the techniques even in terms of working times and the same can be said for mounting, with the exception of the model 4 solitaire, which recorded a
slightly longer time with micro-casting.

Repairs, needed to a greater extent on micro-cast items but which were of short duration, led to a slighter longer overall time for micro-cast jewellery.

Finishing loss

The material removed from the rings in the finishing phases has an immediate effect on production costs since it cannot be totally recovered by refining materials that come from processing. To calculate these costs, an average loss of 5% of precious metal from finishing was estimated. Table 16 shows
the average values for losses for each model and for each technique during all the finishing operations.

The overall losses were greater in SLM™ or in micro-casting, depending on the model being processed. Analysing the individual phases, it can however be noted that, in the feeder elimination phases, micro-casting had greater losses than SLM™, while in the roughing phases, selective laser melting always
lost more material. These results can be easily explained by the greater quantity of surface roughness in SLM™ in the latter case. The impact of the losses recorded in terms of production costs, assuming a loss of 5% in the polishing recovering phase, is summarized in Table 17.

Raw material production costs

For a correct evaluation of the final cost of the items, we also considered the differences in the raw material costs. In fact, the two production methods differ in terms of the price of the raw material and the number of refining actions required to make the same quantity of jewellery. In regard to
raw material costs, we estimated, by evaluating the market prices, a higher amount for buying ‘new’ raw materials of 0.3 €/g for the powder needed in the SLM™ process compared to the micro-casting alloy, due to the cost of atomization. The same cost difference was taken into consideration between the
breaking down and atomization of new raw material retrieved from refined platinum. In order to assess the impact of refining, the first calculation regarded the various yields of the two production processes in terms of the ratio between pieces produced and casting scrap. The weights recorded and percentage
production yields are listed in Table 18 for micro-casting and Table 19 for selective laser melting.

The different yields found in the two production processes had a direct repercussion on the refining needed in the two cases with the consequent effect on production costs.  The calculation of production costs due to refining was carried out assuming that:

the 60 rings produced for this study represented a typical production lot, equal to about 500g of raw jewellery. To make the 60 micro-cast rings for this study, production scrap was re-cast twice, starting from 1 Kg of the initial alloy. We considered that we would have to refine everything after one
production lot.

To consider a situation in SLM™ similar to that of micro-casting, we assumed that, also in this case, everything would need to be refined after having re-melted the scrap twice. Moreover, for this study, the printer was initially loaded with 2.8 kg of powder, a standard production condition.

Refining costs, both fixed and those that depend on the quantity of material, were calculated using the average of the prices applied by 6 different Italian market suppliers (Table 20).

Focusing on SLM™, the quantity of powder that was initially put into the printer to produce jewellery lots was 2.8 kg. When creating one single lot, no scrap needed to be re-melted and, at the end of the print, the quantity of powder still in the machine was about 2 Kg, the rest having been used to produce
the items (500g) and supports (300g approx.). The second lot was also created without re-melting while, in order to make a third production lot, the scrap (mainly from supports) needed to be re-melted and 1000g of new powder had to be added in order to fill the printing platform to cover the height
of the items to be printed. Re-melting the material twice was only required to produce a fifth lot and, only after the sixth lot did all the powder need to be refined. In order to start producing a seventh lot, 1000g of new powder had to be added to that made with refined material.

The data relating to the powder needed for production by SLM™ and the material to be refined are shown in Table 21 while Table 22 reports the relative costs.

As a comparison, the refining costs for 3 Kg of jewellery produced with micro-casting were calculated, bearing in mind that, after every 500g lot, about 0.5 Kg of scrap would need to be refined (Tables 23 and 24).

Despite the lower raw material costs, the cost per gram of jewellery produced was 7% higher in the case of micro-casting mainly due to the set amounts for each refining, which mainly corresponded to the costs for the concentration sample. These expenses were, of course, to be added to the hours of machinery
usage, man hours and energy consumptions in order to have a complete picture of the effective cost per gram of jewellery produced with the two techniques in question.

Total production costs of the rings

With the data presented in the previous paragraphs, including production times, production lots and yields, the final industrial costs for producing each individual model could be calculated. In order to do so and to make the comparison as close to reality as possible, the following assumptions were
made:

Production capacity was calculated for the two techniques based on the effective use of both systems, considering the ring lot made for this study as a quantity produced in a week.

We considered a total machinery amortization time of five years, taking into account the current average fiscal amortization period in Italy. We did not consider the potential duration of the systems in work hours because all the machinery will most probably become obsolete before the end of its lifecycle.

The costs linked to consumer materials were divided uniformly between the items made, calculating an average cost and not the specific cost of each item produced.

In the calculation hypothesis, we intentionally left out the physical space needed to carry out industrial activities which, in regard to 3D printing, are decidedly smaller than that taken up by a lost wax casting system. The same goes for the capital invested into the electrical and hydraulic systems
needed for micro-casting.

We also did not take into consideration any waste disposal costs (crucibles, plaster, acids, etc.)  deriving from the lost wax casting process.

We also lowered the benchmark by considering two hypothetical companies that exclusively produce items in platinum. This would mean lower exploitation of resources, which could be common to platinum, gold and silver processes.

Dividing machinery usage and human resource costs for each model was carried out based on the percentage weight of each ring produced in respect of the total cylinder cast or the print plate.

The hourly cost of the technicians was taken as the same for both SLM™ and micro-casting, and similar for each person involved in the production and finishing processes.

The consumer materials needed for SLM™ and micro-casting production are listed in Table 25.

The results of production cost calculations for each model were subdivided into semi-processed production costs, finishing costs (including losses of material during processing) and refining costs, shown respectively in Tables 26, 27 and 28.

What emerged from semi-processed production costs was the enormous impact of micro-casting system underuse which led to unfavourable amortization costs compared to the SLM™ technique. This resulted in a higher production cost for almost every ring model, with the exception of the wedding bands for men

System underuse is due to the widespread practice in many companies of internalizing the platinum casting phase mainly for strategic rather than economic reasons, preferring not to entrust the process to third parties. Moreover, the platinum jewellery segment is a precious metal jewellery niche with
production demand standing at about 60 times less than demand for gold, another element that contributes to not optimising the use of the machinery.

The overall finishing costs, however, showed a more varied trend with a general advantage for the SLM™ technique except for those items that involved greater difficulty in removing supports and in roughing.

In regard to refining costs, all the models resulted more favourable with SLM™ due to the higher cost per gram for micro-cast jewellery.

Lastly, looking at the overall costs (Table 29), SLM™ production was clearly less expensive compared to micro-casting in 5 solitaire models and for the ladies’ wedding band sizes, while trilogy 1 and wedding band 4 for men were less expensive in micro-casting. Finally, in three cases, the model 8 solitaire,
model 2 trilogy and wedding band 1 for men, the costs were practically identical, given that the differences found could easily be cancelled out by tiny variations in the production phases. It should also be underlined that, the additional cost linked to re-firing two non-compliant, micro-cast wedding
bands was not taken into consideration in the calculations, therefore only 57 out of the 60 micro-cast rings were truly suitable for sale compared to all 60 in SLM™.

Investment capital

The amount of investment capital required to start producing the semi-processed goods involved in this comparison, net of the necessary resources for the finishing phases, which are the same for micro-casting and SLM™, are slightly higher for SLM™. In fact, the greater quantity of machinery needed to
start micro-casting activities is only partially compensated by the high cost of a selective laser melting printer and by the need for more precious material, estimated as 2.8 Kg of powder against 1 Kg of alloy in micro-casting, to be able to print the jewellery items to their full height.

It is also true that, in the case of micro-casting, there is a better offer of machinery which could lead to a lower amount of investment capital, while in 3D printing, the investment capital calculated here is the minimum required for using this technique. Another difference is that the investment capital
for micro-casting is all instrumental while part of the capital in 3D printing is financial. This favours an SLM™ company in the case of liquidation since selling precious metal is easier than selling second-hand machinery.

It should be noted, however, that, as previously mentioned, we did not take into consideration the higher cost of the systems needed for the good working order of micro-casting machinery, which include a more complex electricity system, a hydraulic system that must be installed onto each machine with
chilled water and a vacuum system to convey the air and dispose of fumes during cylinder firing phases.

It should also be added that a lost wax casting system requires a working space of at least 50 m² which, at market price in Italy, would cost about 100,000 €, while the space need for 3D printing is potentially less than 1 m².

Environmental impact

The environmental impact is a standard that is becoming increasingly important in the overall assessment of a production process. For this comparative study, environmental impact was quantified for each production technique by calculating the Carbon Footprint (CF), which refers to the amount of greenhouse
gas (GHG) released during production, expressed in terms of CO2 equivalent mass.

The released GHG comparison was carried out considering all the phases and materials needed to complete jewellery production. Calculation of emissions caused by production and by disposing of the materials used was carried out using the data taken from the EcoInvent 2.2 database, while the data on greenhouse
gases deriving from the production of electricity to power the machinery were taken from the National System for Environmental Protection ((ISPRA), based on the production of electricity for the Italian network (9). Calculations did not include greenhouse gases caused by raw material extraction and
system and machinery construction.

It can be noted from the results how the total of greenhouse gases released into the environment during the production of 60 rings with SLM™ was half the amount generated with micro-casting. Greater electricity consumption, the gases released in plaster firing phases and the general use of materials
with a high environmental impact are the main cause of this result.

CONCLUSIONS

We can conclude from this study that, from a qualitative point of view, the jewellery produced with SLM™ is better both in terms of macro surface defects and internal porosity. This fact is confirmed by technician evaluation and the number of items that needed to be corrected by laser, as well as by
the lack of non-conformities compared to the three non-compliant wedding bands produced by micro-casting. The potential re-firing of a non-compliant piece is also more unfavourable in terms of time and costs compared to a hypothetical re-print.

Production times were slightly slower in SLM™, although the technique is more effectively exploited with small platinum production lots compared to a micro-casting system.  The greater production yield with selective laser melting also limits resorting to refining, with advantages on costs.

The overall costs were significantly in favour of the SLM™ technique for many of the models created, with only two evident advantageous cases in micro-casting. All this in the face of a slightly higher initial activity start-up investment and half the environmental impact.

In consideration of the data collected, we can conclude that, for companies that deal only in platinum productions and with weekly lots of about 500g of raw jewellery, the SLM™ technique is decidedly more advantageous than micro-casting, since it is more suitable for the small quantities of platinum
jewellery produced and has, on average, a better quality compared to the same items produced by micro-casting.

We can therefore confirm that, as our work presented at the Santa Fè Symposium 2017 hypothesized, platinum jewellery production can be included in cases in which the SLM™ technique is truly an added value compared to traditional casting.

References

Damiano Zito et al., “Why Should We Direct 3D Print Jewelry? A Comparison between Two Thoughts: Today and Tomorrow,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2017, ed. E. Bell et al. (Albuquerque: Met-Chem Research, Inc., 2017).

Teresa Fryé and Joerg Fischer-Buehner, “Platinum Alloys in the 21st Century: A Comparative Study,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2011, ed. E. Bell (Albuquerque: Met-Chem Research, Inc., 2011).

GFMS Platinum Group Metals Survey 2017, Thomson Reuters Eikon™.

G. Ainsley et al., “Platinum Investment Casting Alloys,”
Platinum Metals Review
22, no. 3 (London: Johnson Matthey & Co. Limited, July 1978): 78.

P. Lester et al., “The Effect of Different Investment Powders and Flask Temperatures on the Casting of Pt Alloys,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2002, ed. E. Bell (Albuquerque: Met-Chem Research, Inc., 2002): 321-334.

U.E. Klotz and T. Drago, “The Role of Process Parameters in Platinum Casting,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2010, ed. E. Bell (Albuquerque: Met-Chem Research, Inc., 2002): 287-326.

Damiano Zito et al., “Definition and Solidity of Gold and Platinum Jewelry Produced Using Selective Laser Melting (SLM™ ) Technology,”
The Santa Fe Symposium on Jewelry Manufacturing Technology 2015, ed. E. Bell et al. (Albuquerque: Met-Chem Research, Inc., 2015): 455-492.

Istituto Superiore per la Protezione Ambientale, “Fattori di Emissione Atmosferica di CO
₂ e Sviluppo delle Fonti rinnovabili nel settore elettrico”, (2017).

Figure


1


. Eternal model frame and interchangeable part

Figure


2


. “KEY” for replacing the spring

Figure


3


. Sequence for changing the interchangeable part on Trilogy

Figure


4


. Model 1 wedding band frame

Figure


5


. Model 4 wedding band frame

Figure


6


. Model 4 solitaire frame

Figure


7


. Model 5 solitaire frame

Figure


8


. Model 7 solitaire frame

Figure


9


. Model 8 solitaire frame

Figure


10


. Model 15 solitaire frame

Figure


11


. Model 16 solitaire frame

Figure


12


. Model 1 trilogy frame

Figure

13


. Model 2 trilogy frame

Figure

14


. ETERNAL model wedding band


frame

Figure


15


. plaster firing cycles

Figure


16


. Feeders and supports used to produce the model 1 wedding band

Figure


17


. Feeders and supports used to produce the model 4 wedding band

Figure

18


. Feeders and supports used to produce the model 4 solitaire ring

Figure


19


. Feeders and supports used to produce the model 5 solitaire ring

Figure 20 . Feeders and supports used to produce the model 7 solitaire ring

Figure 21 . Feeders and supports used to produce the model 8 solitaire ring

Figure 22 . Feeders and supports used to produce the model 15 solitaire ring

Figure 23 . Feeders and supports used to produce the model 16 solitaire ring

Figure 24 . Feeders and supports used to produce the model 1 trilogy ring

Figure 25 . Feeders and supports to produce the model 2 trilogy ring

Figure 26 . SLM™ support residue on the surface of a ring

Figure 27 . Feeder residue on the surface of a ring

Figure

28


. Internal support in the model 4 solitaire

Figure

29


. internal support in the model 2 trilogy

Figure 30 . Raw model 4 wedding band produced by micro-casting

Figure 31 . Raw model 4 wedding band produced by SLM™

Figure 32 . Model 4 wedding band produced by micro-casting, after sanding

Figure 33 . Model 4 wedding band produced by SLM ™, after shot peening

Figure 34 . Roughness measurement directions on wedding bands

Figure 35 . Roughness measurement directions on SLM™ solitaires and trilogies

Figure 36 . Roughness measurement directions on micro-cast solitaires and trilogies

Figure 37 . Average roughness found in the various directions in micro-casting and SLM ™

Figure 38 . Surface roughness on the vertical wall of a raw SLM™ wedding band, 300X

Figure 39 . Surface roughness on the horizontal wall of a raw SLM™ wedding band, 300X.

The parallel tracks left by the laser scans can be seen.

Figure 40 . Surface roughness on the horizontal wall of a raw micro-cast wedding band, 300X.

Figure 41 . Excess material on the side of a micro-cast ring

Figure 42 . Example of refractory breakage

Figure 43 . Cavity on the surface of a model 4 solitaire, probably caused by a refractory micro-detachment encased in the molten metal

Figure 44 . enlargment of the defect in Figure 43

Figure 45 . Subsidence probably caused by fragments of detached refractory stuck on the surface of the molten metal

Figure 46 .

Figure 47 . Details of the surface in Figure 46

Figure 48 . Surface porosity in a micro-cast solitaire

Figure 49 . Details of the area in Figure 48

Figure 50 . Micro-cast model 8 solitaire with evident ovalling

Figure 51 . Deformation of the grips of a micro-cast model 4 solitaire

Figure 52 . Added ring to stabilize the position of the grips on model 4 micro-cast solitaires

Figure 53 . Breakage in one of the micro-cast wedding bands

Figure 54 . Internal cavity in a micro-cast wedding band.

Note the correspondence with the item’s fracture zone.

Figure 55 . Extension of the cavity in the other half of the sectioned wedding band

Figure 56 . Surface swelling in a model 2 trilogy produced by SLM™

Figure 57 . Surface swelling in the model 2 trilogy in Figure 56 compared with the standard profile

Figure 58 . Example of digital measurement

Figure 59 . Offset compared to the nominal internal diameter measurement with standard deviation

Figure 60 . Section A plane

Figure 61 . Section B planes

Figur e 62. Micro-cast wedding band 1

Figure 62 . SLM™ wedding band 1

Figure 63 . Micro-cast wedding band 4

Figure 64 . SLM™ wedding band 4

Figure 65 . Micro-cast solitaire 4

Figure 66 .  SLM™ solitaire 4

Figure 67 . Micro-cast solitaire 5

Figure 68 . SLM™ solitaire 5

Figure 69 . Micro-cast solitaire 7

Figure 70 . SLM™ solitaire 7

Figure 71 . Micro-cast solitaire 8

Figure 72 . SLM™ solitaire 8

Figure 73 . Micro-cast solitaire 15

Figure 74 . SLM™ solitaire 15

Figure 75 . Micro-cast solitaire 16

Figure 76 . SLM™ solitaire 16

Figure 77 . Micro-cast trilogy 1

Figure 78 . SLM™ trilogy 1

Figure 80. Micro-cast trilogy 2

Figure 81. SLM™ trilogy 2

Figure 82. cavity in micro-cast ring section

Figure 83. porosity from shrinkage in micro-cast ring section

Figur e 84. Porosity from gas in SLM™ ring section

Figure 85. Porosity inter-hatches in SLM™ ring section

Figure 86. "as cast" micro-cast wedding band after metallographic attack, 50x

Figur e 87. “as cast” micro-cast wedding band after attack, 200x

Figure 88. "as print" SLM™ wedding band after metallographic attack, 50x

Figure 89.  "as print" SLM™ wedding band after attack, 200x

Figure e 90. Micro-cracks visible in the SLM wedding band after metallographic attack

Figure 91. Traction tester

Figure 92. Evaluation of the difficulty in removing supports/feeders

Figur e 93. Evaluation of roughing difficulties

Figure 94. Evaluation of the surface quality after roughing

Figure 95. Evaluation of polishing difficulties

Figure 96. Evaluation of mounting difficulties

Figure 97. Stilnovo s.r.l.’s internal Quality Control evaluation of the jewellery produced by SLM™ and by micro-casting

Figure 98. Micro-cast solitaires

Figure 99. Micro-cast trilogies

Figure 100. Micro-cast wedding bands

Figure 101. SLM™ solitaires

Figure 102. SLM™ trilogies

Figure 103. SLM™ wedding bands

Figure 104. Overall machine times for technique used and total man hours

Figure 105. Kg of CO2 equivalent produced by each of the two techniques

Table

1


. List of items produced for each model and production technique

Table 2 . as cast / as print roughness

Table 3 .Roughness after sanding or shot peening

Table 4 . Internal diameters measured for each model compared to nominal diameters

Table 5 . Average percentage porosity found for the two production techniques in question

Table 6. Vickers micro-hardnesses on micro-cast and printed model 1 wedding bands

Table 7. Mechanical characteristics

Table 8. Casting subdivision

Table 9. Print subdivision and relative production time

Table 10. Average and total machine and technician times for producing the rings with micro-casting

Table 11. Average and total machine and technician times for producing the rings with SLM™

Table 12. Subdivision of micro-casting production phase times

Table 13 . Subdivision of SLM production phase times

Table 14. Finishing operation times (in minutes) for rings produced by micro-casting

Table 15. Finishing operation times (in minutes) for rings produced by SLM™

Table 16. Overall finishing losses

Table 17. Costs relating to losses registered in the finishing phase,

assuming a 5% loss in the recovering phases

Table 18. Percentage production yield with micro-casting

Tab le19. Percentage production yield with SLM™

Tab le 20. Average refining costs on the Italian market

Table 21. Material to be refined for producing with SLM™

Table 22. Refining costs for producing with SLM™

Table 23. Material to be refined for producing with micro-casting

Table 24. Costs of refining for producing with micro-casting

Table 25 . Consumer materials needed for production

Table 27. Finishing costs, including losses, for each ring subdivided by model

Table 28. Refining costs for each ring subdivided by model

Table 29. Total production costs, including costs for producing semi-processed items,

finishing and refining

Table 30. Investment capital needed to start production with micro-casting and SLM™

Table 31. Kg of CO
₂ equivalent produced in micro-casting

Table 32. Kg of CO
₂ equivalent produced in SLM™