Eddie Bell Neutec USA, Albuquerque, NM, USA

Eddie Bell is the President of Neutec/USA and one of the founders of the Santa Fe Symposium on Jewelry Manufacturing Technology. He is an expert in the goldsmith industry, and is considered one of the most highly qualified experts in the process of lost wax casting. He participates at the most important international gold conferences, consistently presenting original papers with a very practical approach.

The feed sprue is an important link in the process of making a good casting. Too often little thought goes into feed sprue design and placement or it is treated as part of the design exercise where esthetics rule over function. Since science and mathematics are difficult to apply to determining the feed sprue design, jewelers must depend on experience. In an effort to explore what works and what doesn’t work, this paper will survey feed sprue problems and solutions.

Feeding Success

Introduction Some cynics have described the criteria used by most jewelers for feed sprue design and placement as an esthetic exercise rather than a technical exercise. While that may be true in some instances, I don’t take such a dim view. Actually, I appreciate an esthetically pleasing feed sprue, (Figure 1) as long as it works well. I think our industry is learning and, in contrast to ten years ago, factories today are using reasonably good feed sprues. However, an area that may not be well understood is the thermal aspect of casting and how that relates to the sprue system as a whole and to the alloy being cast. Perhaps we can apply the theory used in industrial casting to improve our practice and the quality of jewelry castings.

Figure 1 A feed sprue that combines function and beauty; the pattern casts well and the sprue witness is easy to remove.

Feed sprues What do we know about feed sprue design and performance? We know that the feed sprue should be as friction-free as possible. The feed sprue must be capable of flowing liquid metal into the pattern to replace the volume lost due to the shrinkage of solidification (1). Therefore, the feed sprue must freeze later than the pattern it is attached to. We also know that turbulence must be kept at a minimum and that smooth-sided, tapered sprues help reduce turbulence. We know that as the metal moves from the source to the pattern we must maintain a pressure system. June 2005


Reducing the area of the sprue system as it goes downstream does this. The jewelry casting literature is full of casting advice that is backed up by complicated equations and mathematical rules. But these rules and equations are not easy to apply to jewelry models, so they are not used. However, they are based on theory and we can use the theory as a guide, even if we don’t do the math. One of the most mysterious aspects of the feed sprue is the influence that the thermal conductivity of the alloy cast has on how they work. How well heat is transferred from the main sprue to the feed sprue, and on to the pattern, should influence shrinkage porosity and hot tears or cracks. Naturally, each alloy has unique thermal properties but, in practice, jewelry casters use a one-size-fits-all policy. It is common for a pattern to be cast in various alloys with the same feed sprue design. For example, if the same main sprue and feed sprue are used on a specific pattern and cast in yellow gold, white gold and sterling silver, the yellow gold might produce sound castings, the white gold might have no-fills or porosity and the silver might have cracks. This could be explained by the difference in thermal properties these three metals have. Usually we succeed to solve such a problem by adjusting flask and/or metal temperature for the obvious reason that it is a simple solution compared to making a new model and designing a new feed sprue. However, when we push out to the limits of what is possible to cast, a simple change in temperature may not work without causing other problems, and the sprue system must be changed. An example of cracked ring shanks due to the thermal behavior in a sprue system can be seen in Figure 2. The diameter of the main sprue in relationship to the length and diameter of the feed sprue determine the temperature at the feed sprueto-pattern junction. The large main sprue (left) conducted too much heat to the feed sprue-to-pattern junction and caused cracked shanks in the silver casting. Since the castings were made with low metal and flask temperatures, reducing the diameter of the main sprue (right) was necessary to solve the problem.

Figure 2. Cracked ring shanks were the result of using a 15.8mm in diameter main sprue (left). The incidence of cracks was also influenced by ring size since larger rings had shorter feed sprues. Using a 10.5mm diameter main sprue (right) solved the problem. The flask and metal temperature were not changed. 70

Jewelry Technology Forum

I have seen patterns that cast very well in yellow gold and nickel white gold, but when palladium white gold was cast the results were not acceptable. The feed sprue had to be changed in order to get sound casting. The investment should not be much above 600ºC when the metal is cast. The casting temperature of Pd white gold is usually 250-300ºC higher than other gold alloys, so the greater temperature difference between the metal and the mold causes the Pd white gold alloy to lose heat more quickly than the lower temperature alloys as it flows though the sprue system. If the casting temperature of the flask is increased to compensate for metal freezing too early, the risk is high that the investment will be reacted. The result will be contaminated sprue metal that cannot be recast successfully. In Figure 3, you can see reacted investment that could not be cleaned off of the casting using a highpressure water stream. The solution is to design a sprue system that allows the metal to reach the pattern with less temperature drop. Short conically tapered sprues work best with Pd white gold.

Figure 3. High metal and flask temperature combine to react the investment material and bond it tightly to the metal surface

Risers If we study industrial sand casting where much money has been spent for research, we can see that they take the thermal dynamics of the mold/metal relationship very carefully into account. Generally, their castings are large and they use generous amounts of metal in the sprue system to assure that they get a sound casting. Because of high metal value, jewelers are motivated to design their sprue systems to use as little metal as possible, and sometimes the casting suffers as a result. One could argue that the industrial caster has an advantage because the metals they cast are low value, but so are the castings, and the percentage of yield is relative to the investment for both jewelers and industrial casters. Industrial casters have problems we don’t normally have in that they cast in sand molds and have to deal with dirt and dross. This influences the design of their sprue system, which they call rigging (Figure 4). Industrial casters have a pouring basin where the metal is poured that is attached to a vertical sprue (sometimes called a down sprue). At the bottom of the sprue is a pour-box, or sprue well, where the metal has time to settle the turbulence of pouring before it proceeds on. June 2005


The sprue is situated as far from the casting as possible for the same reason. Attached to the pour box are one or more runners. The runners usually run parallel to the casting and allow the metal flow to become laminar. The pattern cavity is attached at the bottom by one or more in-gates. The runners usually extend past the last in-gate, so any dirt or slag can be caught and not carried into the casting. Depending on surface to volume calculations for the casting, risers are attached to the in-gate and sometimes to the casting, depending on the geometry of the casting. Risers serve as reservoirs of liquid metal that feed the casting to replace the shrinkage (volume lost) during solidification of the metal and provide a pressure head to help fill thin sections. Risers are designed to supply liquid metal to the area of the casting where shrinkage voids occur. Risers can function to remove gas and dross from the casting as well.

Figure 4. Generic rigging system on industrial casting Lets compare the sprue system in a typical jewelry tree. Our sprue button is equivalent to their pour basin, and we have a sprue attached as they do. If our patterns are attached some distance from the top end of the sprue, we have the equivalent of their sprue well (Figure 5). We use runners sometimes, but most of the time, we don’t. We shouldn’t have to deal with dirt and dross, so runners are optional. Our feed-sprue is equivalent to their in-gate. What we are missing is the riser. How do we get away with not including the thermal control that the riser provides to the industrial caster? Should we consider using risers and, if we did, what would they look like?

Figure 5. If the main sprue extends beyond the top row of patterns, it serves as a sprue well to allow the turbulence of the first metal entering the mold to settle 72

Jewelry Technology Forum

IMost of the time, we have luck on our side. Generally, our castings have a very high surface area relative to the volume and that helps solidify the metal in the casting quickly. The quick solidification makes it possible for the feed sprue and main sprue to act as the riser. Of course, in order for the feed sprue and main sprue to act as the riser, the metal in the pattern must solidify toward the feed sprue, and the sprue system must have enough thermal capacity to keep liquid metal feeding into the pattern to prevent shrinkage voids. Other times, luck abandons us and we must do something to get rid of the porosity. The most common solution is a secondary feed sprue. If the problem is incomplete filling, adding another feed sprue is a good idea, but if the problem were porosity, maybe a riser would be a better option. Sometimes a second feed sprue makes the pattern hard to mold and it would be difficult to make a clean wax pattern. A good riser design is a sphere on a short rod attached to the section of the casting that needs feed. The sphere should have 1.2 times the volume of the section it is intended to feed and can be added to the master model after evaluation casting have been done to determine the best size and location. If the geometry of the pattern requires the feed sprue to be longer than usual, we couldn’t expect that the thermal capacity of the main sprue to prevent the feed sprue from freezing before the pattern. An old dental casting practice of putting a sphere in inline with the feed sprue proves to be an effective way of managing porosity.

The function of the sprue button The traditional teaching in jewelry casting is that the sprue button provides the function of the riser and, as we have already mentioned, the sprue button is equivalent to the pour basin. I think we all can accept that having a funnel to pour the metal in is a benefit. But does it help manage shrinkage porosity? When the funnel is filled with metal, depending on the design of the cast tree, the sprue button may or may not function as a useful heat and liquid metal reservoir. We must consider all aspects of a particular cast tree whether it is a single pattern of a one-off custom design or a production tree with a main sprue central to a number of patterns. For example, the sprue button may be effective at managing shrinkage porosity in the one pattern cast and have little or no benefit when casting full trees. The sprue button will benefit the quality of a cast pattern if it influences progressive solidification. If you are casting two or three patterns, you generally don’t have a central sprue and you need some mass of metal that will remain liquid longer than the feed sprue and pattern, in that order. The classic way to do this is to attach the feed sprue to the button and, therefore, the button serves as the heat reservoir. When casting a tree with a central sprue and many patterns with a high surface area and low volume, the sprue button may be of little use as a heat reservoir. The height of the metal above the closest pattern influences the pressure used to fill the pattern, but we will address that later. First, I want to concentrate on the thermal advantage of the sprue button, if it exists. In theory, we should be able to get a rough estimate of how beneficial the sprue button is by how much sink or piping there is after solidification. June 2005


Figure 6A is a section view of a sprue button before solidification. Figure 6B is the same button after solidification. When light-weight (low volume), high-surface area patterns are cast, we would expect to see relatively small sprue button shrinkage. This indicates it is not feeding the patterns and may not be needed. When thick section patterns are cast, the sprue button may show considerable shrinkage, as Figure 6C shows. This indicates it is doing a useful job and is needed. In practice, as we all know, the sprue button can take many mysterious shapes. Depending on the alloy cast and the equipment used, the shrinkage might be a pipe as shown in Figure 6C or a skin can form on the top surface preserving the shape of the liquid volume on the outside, but have an internal void as shown in Figure 6D.

Figure 6. Shrinkage may be an indicator of how much metal is being fed to the cast patterns from the sprue button A few years ago, I started to think about the volume of the button as it relates to pressure. Casters know that incomplete filling defects are more prevalent at the bottom (sprue button end) of the tree than at the top. An explanation for this is that the head pressure the liquid metal provides becomes lower as the patterns gets closer to the bottom of the tree. In other words, the height of the liquid metal in the sprue above the feed sprue inlet controls the pattern filling pressure provided by the metal. When casting one of a kind jewelry or master models, the usual method is to attach feed sprues directly to the sprue button as shown in Figure 7A. I cut the button off of a NeuSprue (which is mostly hollow) and weighed it; the volume is the same as about 100 mm of main sprue. Since the height of a liquid column is more important than the diameter, I surmised that instead of a 25 mm tall button, casting with a 100 mm buttonless sprue would use the same amount of metal and provide four times the pressure. If a taller flask is used, the same patterns can be attached to the top of a sprue as seen in Figure 7B and the weight of the sprue button subtracted from the total metal cast. As pictured, the height of the sprue in Figure 7B is twice that of the sprue button (Figure 7A) providing double the filling pressure while using half as much metal. The expected difficulty for pattern filling can control how high the sprue is. A longer sprue can be used for more difficult-to-fill patterns and a shorter sprue for less difficult patterns. Also, the small diameter sprue is easier to recycle than the large sprue button because it can be cut up to mix with fresh alloy in the ratio needed. The arrangement at the right (Figure 7C) is not recommended because it would likely cause shrinkage porosity in the cast patterns because of lack of heat reservoir to keep metal feeding the pattern as it shrinks. 74

Jewelry Technology Forum

Figure 7. Casting with a long sprue and no sprue button (B) uses less metal than the traditional sprue button (A) while providing twice the filling pressure. The configuration on the far right is not recommended I showed this to Tyler Teague and he started experimenting with it. Some of the first trees he cast are shown in Figure 8. They are cast in .925 silver and 18K gold.

Figure 8. Sound 18K and .925 Ag cast on Neutec/ J-zP casting machine. Picture by Tyler Teague, Jett Research Tyler expanded on the idea and introduced it to production casting. He reported on his work with buttonless trees in production at the Santa Fe Symposium in 2002. Buttonless casting saves money because 25% to 33% less metal is in the sprue. It appears that the old jewelry casters myth about the need for a sprue button is just a myth. Understanding the theory of the riser in industrial casting gives us an idea why we don’t always need the sprue button. When the pattern cast is thin, it solidifies really fast. In such casting, the main sprue is large enough and the feed sprue is short enough that the main sprue acts as a riser and keeps the feed sprue open for the time necessary to feed the pattern properly. Figure 9. June 2005


Figure 9. The colors represent metal temperature, blue is solidified, orange is solidifying and shrinking, red is near liquidus and feeding the shrinkage and yellow has enough superheat to act as the riser, providing liquid metal, heat and pressure to the pattern.

Buddha on a riser A few years ago I was helping a caster who cast Buddhas in pure silver and gold. Pure metal is difficult because liquidus and solidus are the same temperature; so special steps have to be taken to manage shrinkage. A Buddha is a special kind of sphere and we all know that a sphere is the most difficult geometry to cast because the volume is very high relative to the surface area. The shrinkage shows up in such casting as huge voids under a skin of metal in the belly and leg area (Figure 10). As the metal cools, the metal skin wrinkles and pulls away from the cavity wall. This happens even when a very heavy sprue system was used. The solution was to take advantage of the geometry and add a riser to the casting. The head being smaller than the body, which is smaller than the leg portion, causes the metal to solidify progressively from the head to the feet. By adding a riser with a greater volume than the leg portion, a rather modest feed sprue (Figure 11) can be used and the shrinkage is driven to the riser, which will be removed. So you see, this is a special kind of riser necessary to cast such a shape in a pure metal. If the Buddha was to be cast in an alloy, another feed system could be used that would be easier to remove.

Figure 10

Figure 11

Summary While the design and placement of feed sprues and overall sprue practice in jewelry casting has improved in the last ten years, we still have some problems such as non76

Jewelry Technology Forum

filling, cracks and porosity. The thermal dynamics in the sprue system is an area of major importance that has been neglected. The influence specific alloys have on how well the sprue system works is also not taken into consideration as demonstrated by our one-size-fits-all approach to spruing patterns. In particular, the rising popularity of palladium white gold has brought on new temperature related casting problems, many of which are best solved by sprue design changes that decreases the temperature loss of metal as it travels through the sprue system to the pattern. Industrial casting practice can be applied to jewelry casting if we appreciate why they design their metal feed system the way they do and drawing parallels between industrial casting methods and jewelry casting methods. In doing so we can understand the use of a riser and see how the main sprue serves that purpose in most jewelry casting. That understanding brings the function of the sprue button into question. It appears that several years of production casting without sprue buttons on the tree prove that they are not needed on typical jewelry product casting. Buttonless casting can also aid one-of-a-kind casters improve quality and reduce the amount of metal in process as well as make recycling sprues easier. Finally, it is shown that using a riser makes it possible to produce sound heavy section casting in pure silver and pure gold.

References (1) Paul Finelt, “Basic Metallurgy”, Proceedings, Santa Fe Symposium on Jewelry Manufacturing Technology 1987, p.29 • Klaus Wiesner, Metal Flow Optimizing - An Important Step to successful Casting, Proceedings, Santa Fe Symposium on Jewelry Manufacturing Technology 1999, p 1 • Lee A. Plutshack et al, Riser Design, ASM International Handbook, Casting Volume 15, 4th edition, 1998, p 577 • J. S. Campbell Jr., Casting and Forming Processes In Manufacturing, McGraw Hill, 1950, p 283 • Taylor Lyman, Editor, Casting Design Handbook, American Society for Metals, 1962

June 2005