Simulating the casting process for jewelry

a speech by Vera Benincasa

Simulating processes has been mandatory in many companies from the most diverse manufacturing sectors for years. One only need think of the aviation industry where companies must be absolutely certain that the components produced have no micro-structural defects whatsoever and where a rejected part cannot be re-processed. 
Casting process simulation identifies areas subject to defects and helps to design the casting system in the most efficient manner. It also allows the causes of any inefficiency to be analysed and to understand how to increase productivity.
This system has been used for over twenty years in smelting processes in the automotive and aerospace industries and for some years now, it has also been adopted by our sector. 
In the jewellery world, lost wax micro-casting has always been linked to operator experience or trial and error. Nowadays, simulation software can optimize the entire process starting from the very first prototype design to the mass production of jewellery items. 


The micro-casting process is one of the oldest methods for producing many types of article.
Different peoples and cultures used this process to produce tools, objects and statues in bronze. One very famous example is the Bronze Riace Warriors that were salvaged from the sea in 1907, 2500 years after they were made in Greece in the sixth century BC.
The process developed over the centuries, evolving from a simple artistic method and demonstrating exceptional versatility.
Micro-casting, or lost wax casting, has been used for many years in the automotive and aerospace technology sectors, however, this process, although highly reliable, still presents defects.  

In the jewellery world, the most common and problematic defects are:
- incomplete filling of the cast
- porosity by retraction.
While cases of the first type of defect are limited to specific kinds of production (filigree, particular alloys, complex geometries...), the second can be found in all micro-cast products because it is intrinsically linked to the solidification process of the metal alloy.

It is on this latter category of defect that we will be focusing our attention.
Up until a few decades ago, “Trial and Error” was the obligatory method of trying to solve the problem of porosity by retraction.
With practice and experience, the aim was to minimize and hide the defect.
In order to obtain a commendable result, modest amounts of time and metal needed to be invested.
Simulating the process in a virtual environment minimizes this investment and achieves a better result in less time.

Nowadays, simulation software has reached a high level of precision so that excellent results can be obtained in terms of product development times and the production process can be improved.
The use of casting process simulation software in the jewellery sector is relatively recent although it is constantly on the increase due to the growing request for quality on the part of the big jewellery companies.

Simulation software works when there is a deep knowledge of the entire process. For this reason, the more it is used, the more information will be available to configure the process and the more the results provided by the software will correspond to industrial needs.

The software used for this study is produced by the French ESI Group and is called ProCAST.
It is an advanced and complete tool and has been on the market for over 20 years. It is widely used in various industrial fields. The software is based on finite element technology and is able to simulate a long list of real processes. This study focuses on the module for simulating the lost wax casting process.

 

Figure 1 – processes that can be simulated with Procast

A knowledge of metallurgy and the production process is necessary to be able to use the software to its best advantage. With the help of the SW, the process technologist can work out the best conditions for a reliable and robust micro-casting process.

The software consists of the following environments:
- MESH
- CAST
- VISUAL

MESH is the environment in which our object, starting from the CAD model, is broken down into minor elements (mesh) that the then software uses to understand the points on which to calculate the thermal exchange and solidification equations. The operator decides on the size of the mesh in accordance with various factors. Besides the object of our study, the mould into which we will be casting the metal in order to accurately simulate our process must also be designed and “meshed”.

CAST is the environment where all the parameters to be taken into account in our process are inserted: type of alloy, process temperature, pressure of the metal as it goes into the mould, entry section, thermal exchange, radiation phenomena, etc...

VISUAL is the ideal environment for observing and measuring the simulation results and, depending on what is being analysed and measured, different physical measurements can be viewed (temperature range, solid fraction, voids, porosity, flow speed, etc...)
With the help of Procast, it is extremely easy to foresee where the defects of a particular micro-cast will be and how large these defects effectively are.

Thanks to the software, we can quickly and fully investigate the porosity by retraction defect.

The problem of porosity by retraction is strictly correlated to the solidification concept. In fact, porosity by retraction occurs when the metal status turns from liquid to solid: the metal undergoes a volumetric contraction and, where the volume retracts, dendritic structures, which can become larger as the alloy hardens, can emerge on the surfaces.

Figure 2 – Diagram showing volume according to temperature

Dendrites are tree-like structures that form during metal alloy solidification. The metal forms crystals that grow and solidify in the most energetically favourable crystallographic directions. If cooling is fast, dendrite growth is limited. On the other hand, if cooling is slow, the dendrites will be larger and, in the worst scenarios, visible to the naked eye in the volumetric retraction zone.

Figure 3 - Diagram of the structure of an alloy’s dendritic growth

Volumetric contraction is intrinsic to the solidification process and therefore, porosity by retraction is an inevitable defect in the casting process.
Porosity by retraction cannot be eliminated but it can be conveyed to strategic points by encouraging directional solidification.

When a metal alloy solidifies, the last volume to do so, that is, the one that stays “hot” the longest, will be the part that contains porosity by retraction.
In designing a casting system, it is therefore essential to study the thermo-dynamic factors that lead to controlled solidification: the casting channels, feeders and risers are to be designed and sized so as to be able to correctly feed the piece being produced and, at the same time, “keep” retraction out of the areas of interest.

To study the solidification of micro-cast objects, it is important to consider the cooling module.
The cooling module, or thermal module, is given by the ratio between mass and surface of an object M=V/S.
Starting with volume, if the surface of the object is greater, solidification time will drop drastically. Solidification time depends on M and also on the type of material and the object’s geometry.
Studying solidification time is essential for shifting the direction of solidification. 

Let’s take a very simple object, such as a wedding band, for example.
Being circular and symmetrical, the point in which we put the feeder is of no importance. What is important, however, is the feeder’s section and geometry.
Below is an example of the solidification simulations of the same ring with three different feeders in ascending section size.

Figure 4 – solidification analysis of the wedding band in the three study cases

As can be seen in the picture above, the feeder with the largest section tapered towards the metal entry section is the one that fills the cast and directs ring solidification correctly. 

As further evidence of the correctness of feeder 3’s design, we can see, again by simulation, a reduction in porosity (in purple) in the diagrams below.

Figure 5 – porosity in the rings using feeders with sections of ascending size

Now let’s look at another simple ring geometry, but this time with a variable section.

Figure 6 – ring with variable section

In this case, since the geometry is not symmetrical, the point at which we feed the ring is of fundamental importance. Figure 7 below shows the progress of solidification based on the point chosen for positioning the cast feed.

Figure 7 – ring solidification with feeder in point A or point B

Solidification observed in the previous figure leads to porosity in the areas highlighted in figure 8.

Figure 8 – evidence of porosity by retraction found after ring simulation with feeding in point A or point B

The results can be verified by observing the actual components cast. When using simulation software, it is extremely important to calibrate the reliability of the software with one’s own casting process.  
The photographs below show the surfaces of the ring analysed and cast with the two different feeder positionings:

Figure 9 – porosity visible on the raw cast ring.
On the left, ring fed from position A, on the right, ring fed from position B

Figure 10 – evidence of a macroscopic pore on the surface of the ring fed from position B

Figure 11 – two polished wedding bands.
On the right, porosity by retraction on the ring fed from position B

Similarly, taking a larger object, we can see that the same directional solidification rules can be applied in this case too.
The figure being examined is a “C” which could be used to make half a bracelet.

Figure 12 - "C-shape” for producing bracelets

In the first example, let’s consider the item with the same type of feed but cast with different parameters. The response variation to the change in temperature, both in the mould and in casting, is particularly notable.

Figure 13- solidification – on the left, Tcil:Tc1 Tfus:Tf1; on the right, Tcil:Tc2 Tfus:Tf2
(with Tc2>Tc1 and Tf2>Tf1)

Figure 14 - porosity – on the left, Tcil:Tc1 Tfus:Tf1; on the right, Tcil:Tc2 Tfus:Tf2
(with Tc2>Tc1 and Tf2>Tf1)

 

As can be seen from the diagrams, as the temperature rises, the size of the porosities decreases. This happens because the metal is given more time to solidify in a directional manner. In this case, however, only modifying the process parameters does not solve the problem at the root.
It is therefore necessary to modify the feed. Let’s look at two different feeds.

Figure 15 - feed A – feed B

The figure below shows the simulation of the solidification process in both cases.

Figure 16 – Bracelet solidification with feed position A (left) and feed position B (right)

Analysing the figure on the left, it can be noted that the six feed spokes are solidifying before the bracelet itself has solidified (as in the previous example), thus “blocking” the way for the metal to continue to feed the object correctly. In the figure on the right, however, we can see how the four spokes are feeding the piece well, resulting in a directional solidification towards the heart of the column.

Figure 17 – porosity analysis in the two feeding positions, A and B

Evidence of the efficiency of type B feeding position can be given by analysing the porosity.
In figure 16, it can be noted how, in case B, the object has no porosity, while in case A, there are six porosity nuclei exactly where the metal took longer to cool.

The accurateness of these simulations is shown in the photographs below.

 

Figure 18 – raw cast bracelets: on the left, feeding position A, on the right, feeding position B

Figure 19 – detail that already shows porosity by retraction on the raw bracelet using feeding position A

Figure 19 – on the left, bracelet fed from position A, on the right, from position B

The analysis of these simple geometries demonstrates the validity of simulation. The software can precisely predict which areas will be affected by defects and their size.
The micro-casting simulation process is a useful tool for the technologist who cannot totally eliminate the “Trial and Error” process but can limit it in the virtual simulation environment, thus reducing product industrialization times and costs.

Table 1 – advantages of simulation calculated on actual study cases

The fundamental tool for using casting simulation software is CAD 3D modelling.
In fact, as already mentioned, in order to be able to simulate the casting process, it is absolutely necessary to start from a 3D model of both the casting system we want to simulate and the mould in which we will be casting the metal.
The more accurate the initial model is, the more accurate the simulation results will be.
Moreover, CAD modelling offers the advantage of being able to rapidly design and simulate various types of feeding points and casting systems.
By simulating different feeding points, we can find the best one for our item.
By carrying out casting simulation at the beginning of the design process, it would immediately be possible to identify errors in the design and intervene by modifying the model’s geometry.   
If modifying the item is not possible, then acting on the other parameters (feeding, process parameters, etc..) will become necessary.

Once the importance of simulating each individual detail has been understood, new possibilities for more complex casting systems can be explored. By simulating an entire casting tree, for example, it is possible to analyse the entire process and optimize it.

Figure 20 – cast simulation of a tree

In conclusion, introducing this technology into the jewellery supply chain is undoubtedly helpful for moving towards better production performances and benefits companies that want to use additional forces and means in their production processes.

Resources and study are needed to take best advantage of this technology. Nevertheless, the benefits resulting from its usage (savings in time and means as well as the effectiveness of the results obtained) eliminate every uncertainty. In time, this will be the only way to proceed in jewellery industrialization, as has already happened in other production sectors.


Responsible of process engineering and maintenance at Mattioli SpA in Turin

Vera Benincasa is responsible of process engineering and maintenance at Mattioli SpA in Turin, a jewelry company working for the main brands in the jewelry industry and with its own line of jewels. Vera Benincasa has collaborated with this company since 2013.

© 2019JTF - LEGOR GROUP S.p.A. - CF e P.Iva IT 00844230284 - Reg. Imprese Vi - Nr. REA 147954 - Capitale sociale i.v. 3.000.000€ - Privacy policy