Solving Casting Problems 
with New Sleeve Technology

Ronald C. Aufderheide Ralph E. Showman Foundry Products Division Ashland Specialty Chemical Company Division of Ashland Inc.

Abstract

Foundrymen are constantly being confronted with challenges to improve their operations and lower costs while at the same time producing higher quality castings. One of the ways to achieve lower costs and improve the soundness of a casting is to incorporate the use of exothermic riser sleeves. This can lead to improved yield while solving shrinkage problems. However, at the same time, the use of exothermic riser sleeves can create other problems. This paper will discuss two defects that, under certain conditions, can be created by the use of exothermic riser sleeves in ductile iron.

The first defect is a surface "fish-eye" defect that is caused by the buildup of exothermic sleeve material in the molding sand. This defect doesn't occur on a casting just because it is made using exothermic sleeves; rather, it occurs on the ductile iron castings that are made with sand that has been contaminated with exothermic sleeve materials. Tests showed that the presence of fluorine in the exothermic sleeve formulations contributed to the formation of fish-eye defects. A new exothermic sleeve was developed that did not contain fluorine and that eliminated the fish-eye defect.

The second defect is a degradation of the graphite nodules in ductile iron castings. Testing showed that the amount of flake graphite is related to the type and composition of the exothermic sleeve. The degradation was highest when sand-based exothermic sleeves were used. Fiber-based exothermic sleeves produced slightly less degradation, and the new-technology cold box-based fiber-free and fluorine-free exothermic sleeves produced the least amount of degradation when exothermic sleeves were used. Insulating riser sleeves did not show any degradation tendencies. Depending upon the type of exothermic sleeves used, special considerations need to be made with respect to the placement, size, and quantity of sleeves used so that no contaminated metal gets into the casting itself.

Introduction

Ductile iron castings have unique riser requirements compared to the feeding of other metals. The volume changes in the casting are not a simple contraction as the metal cools. For example, when the graphite nodules are formed, the metal actually expands, which can push metal back into the riser and gating system if these are not properly designed. This, along with the subsequent contraction of the metal as it cools, creates a strong demand on the feeding capabilities of the riser.

Now, more than ever, foundries are trying to find ways to reduce their overall cost to produce a casting. One way to reduce costs is to incorporate the use of exothermic sleeves around the risers. This allows the use of smaller risers that improve yield and reduce the contact surface area of the riser to the casting, which costs money to grind off.

There are a variety of sizes, shapes, and formulations available in the exothermic category of riser sleeves. Traditionally the sleeves were made of fibrous refractory combined with a blend of materials that produce an exothermic reaction more commonly known as a thermite reaction. The most common fuel material is aluminum. When mixed with an oxidizer and an initiator/fluxing material and exposed to extreme heat, the aluminum is oxidized, giving off heat as the reaction proceeds.

2Al + Fe₂0₃ ----> Al₂0₃ + 2Fe + Heat 
(Fluoride initiator / Flux + Heat)

In addition to fiber-based exothermic sleeves, sand-based exothermic sleeves had been gaining favor with many ductile iron foundries. Sand-based, high-density sleeves are formulated to contain more aluminum fuel and to generate a greater amount of heat. This heat is first required to raise the temperature of the sand-based sleeve, before favorably influencing the temperature of the metal in the riser.

Finally, in 1997 Fiber-free New Technology Sleeves were introduced, providing another exothermic sleeve alternative. The refractory material is a round alumina silicate material bonded by cold box resin technology. During the development of this technology, it became apparent that the requirements for an exothermic sleeve for ductile iron applications are different from those for an exothermic sleeve used to make steel castings. This is especially true for cold risers in ductile iron.

Cold risers are those risers that are filled with metal after the metal has moved through the casting, as opposed to hot risers that are filled by the gating system before the metal goes into the casting. This makes the metal in the cold riser colder and closer to solidifying. In order to get an exothermic reaction to start, the metal needs to give up some of its energy to the riser sleeve. However, if too much energy is given up, the surface of the riser begins to skin. Once the skin has formed, the exothermic reaction of the riser sleeve is not enough to remelt it, and the riser becomes less efficient in its feeding capabilities. To solve this, a special, fast-igniting exothermic sleeve is needed so that the energy taken out of the metal in the cold riser is minimized. It has been found that cold ductile iron risers exhibit improved performance when their formulation has been optimized so that they ignite at lower temperature and energy levels, have a faster ignition time, and burn at higher temperatures with more energy. The result is a flatter feed pattern in the riser, as shown in Figure #1.

Figure #1: Piping of standard exothermic-sleeved riser (Left) versus the flatter feed of a New Technology fast-igniting exothermic sleeve (Right).

Increasing the amount of aluminum can increase the heat generated by the riser, but there are some limitations. First of all, after the exothermic reaction is completed, the riser must then rely on its insulating properties to keep the metal in the riser hot. Unfortunately, as more exothermic material is added to the riser formulation, the amount of insulating material is reduced, thereby also reducing the insulating properties after the exothermic reaction is completed. The key is to balance the amount of each to produce the optimum sleeve performance. Sand-based exothermic sleeves are not good insulators.

The sand has the same thermal characteristics as the sand mold, so the sand-based exothermic sleeve must rely on the heat of its exotherm. However, the increased use of exothermic sleeves has brought with it a unique set of problems, which is the subject of this article.

We will be covering two different defects, both of which are directly affected by the use of exothermic riser sleeves. The fish-eye defect appears on the surface of the casting as a round depression with a raised center, as shown in Figure #2. The second defect that will be discussed is a degradation of the graphite nodule from spherical to flake form. This flake structure can potentially extend into the casting, resulting in severe reductions in the physical properties of the casting and its subsequent performance. This effect was first noted during the microstructure analysis of the fish-eye defect and the further development work on new exothermic sleeve formulations and their effect on ductile iron.

Figure #2: Fish-eye Defect

The Fish-eye Defect

Fish-eye defects are unique to ductile iron foundries. Very little is understood about their cause, and even less has been written about them. Before the wide use of exothermic sleeves, fish-eye defects were only seen occasionally in green-sand foundries when an excessive amount of clay balls was present in the sand. With the growing use of exothermic riser sleeves, there has been an increase in the presence of fish-eye defects. It has been found that these defects are linked directly to the use of exothermic riser sleeves. Initially, it was theorized that the fish-eyes were caused by high levels of fluorine in the molding sand, residual unburnt pieces of sleeves in the molding sand, gas reactions from the sleeve materials, and/or sleeve materials coming in contact with the casting surface.

In order to determine the cause, tests were run using five different contamination materials: cryolite, crushed unburnt exothermic riser sleeves, crushed burnt exothermic riser sleeves, crushed 0-Fluorine exothermic sleeves, and crushed insulating sleeves. These contaminants were placed on the pattern surface for one set of tests. For the second set of tests, they were mixed in with green sand and used as facing sand for the test castings. In order to determine the amount of sleeve materials needed to produce a fisheye defect, tests were run on a foundry's sand that actually did produce the defect. With this information, the amount of contamination used in the testing was deliberately set at twice the calculated amount.

The first set of tests, where the contamination materials were placed directly on the casting surface, did not create any defects. This was extremely interesting in light of the initial theories of high fluorine levels (from the cryolite initiator material in exothermic sleeves), gas from the sleeve, and/or contact with sleeve material causing the defect.

The second set of tests, where the contaminants were mixed in with green sand and used as facing sand to make the casting, were much more informative. Figure #3 shows the contaminated facing sand on the pattern. Table 1 lists the results of these tests. Although fluorine initially was thought to be a major contributor to producing fish-eye defects, in both cases cryolite contamination did not produce a defect. The only time that a defect occurred was when material from a traditional exothermic sleeve was mixed in with the facing sand. However, when the fluorine initiator in the exothermic sleeve was replaced with a nonfluorine-containing initiator, no defects were produced. The defect was also produced regardless of whether the exothermic sleeve was new (as would be the case if a mold wasn't poured and the sleeve was shaken out with the mold in a closed system) or if it was burnt.

Figure # 3: Contaminated Facing Sand

The results of these tests showed that in order to avoid the formation of fish-eye defects, several things could be done:

• Don't pour ductile iron when the sand is contaminated with fluorine containing exothermic sleeves.
• Dilute contaminated sand before using it to pour ductile iron castings. 
• Avoid using large quantities of fluorine-containing exothermic sleeves.
• Use insulating sleeves.
• Use 0-Fluorine exothermic sleeves.

Microstructure Degradation

During the initial investigation into what caused fish-eye defects, the microstructure of the metal was examined in the areas adjacent to the defect and areas in contact with the exothermic sleeve in the riser. These examinations revealed another problem that wasn't expected. A large percentage of the graphite nodules had degraded into flake graphite. Closer examination showed that the degradation was not limited just to the surface of the metal but extended all the way across the upper portion of the riser itself, as can be seen in Figure #4.

An analysis of the metal at the top, center, and base of the riser revealed varying levels of aluminum. Table #2 below lists the carbon, silicon, magnesium, and aluminum levels at three locations within the riser. The aluminum level is extremely high in the top portion of the riser. Typically aluminum levels in excess of about 0.10% will create flake graphite in ductile iron.  As the aluminum level goes down, the graphite degradation is reduced. The concern is how much aluminum will the riser pick up and how far down will the graphite degradation extend. In one case, 33% of the riser contained flake graphite.

Figure # 4: Ductile Iron Riser Cross-Sectioned

The total amount of aluminum in the metal could continue to build up if not monitored and controlled. High aluminum levels in the metal will not only prevent the formation of graphite nodules, but also can cause pinhole defects in green-sand molding. 

	C	Si	Mg	Al
Top	3.48	2.40	0.03	0.30
Center	3.50	2.12	0.045	0.047
Base	3.56	2.14	0.05	<0.01
Table # 2: Analysis of Riser

There is good reason for concern since the risers on ductile iron castings represent a high percentage of the base charge for the next metal melted. In addition to the aluminum pickup from exothermic riser sleeves, aluminum also can come from other sources such as inoculants and alloys, as well as from other charge materials. With this in mind, it was decided to investigate the different types of sleeves on the market to see how they compared in terms of aluminum pickup. Table #3 lists the results of how the different types of sleeves affect the amount of flake graphite that is formed in the riser.

Table #3: Comparison of Aluminum and Flake Graphite with Different Sleeves

Exothermic sand risers showed the highest amount of flake graphite formation, followed by the traditional fiber-based exothermic riser sleeve. Two different fiber-free New Technology exothermic riser sleeves were also run. The "Hi-EX" New Technology sleeve contained a slightly higher level of aluminum than the current production grade that is being sold in the market to try to match the amount of aluminum contamination caused by the exothermic fiber-based riser. Although the amount of aluminum contamination was similar, the graphite degradation was only 15%. A second "Lo-EX" New Technology sleeve, which contained about half the amount of aluminum fuel, was tested, as well as an insulating sleeve which does not contain any aluminum fuel. These showed less than 2% and no flake graphite respectively.

There are several ramifications if the flake graphite contamination extends into the ductile iron casting itself. For example, the mechanical properties of the casting, such as tensile strength and ductility, will be reduced, which can cause the part to fail in service. Unfortunately, these problems are not always detected by normal testing. Many ductile iron foundries check the aluminum level in their metal on a regular basis, which is a practice that all foundries should implement.

Figure #5 shows a ductile iron yoke production casting that had a dark area on the machined surface on the cope side of the casting. The location was just below an exothermic riser. The initial thoughts were that the casting had not "cleaned-up" and some of the oxidized "as-cast" surface remained. However, further examination of the microstructure below the defect revealed flake and vermicular graphite in the dark areas rather than nodules. Contaminated metal from the riser had apparently fed into the casting, causing the degraded microstructure. Changing the location of the risers solved the problem. Rather than placing an individual riser on top of each casting, a single riser was placed in the runner system between the ingates of two separate castings. This changed the feeding pattern so that no contaminated metal entered into the casting. It also improved the casting yield.

Unmachined and machined castings	Dark area on machined surface
Microstructure in dark area, 100x	Microstructure in adjacent area, 100x

Figure #5: Production casting showing graphite degradation that extended into the casting

Identifying the Causes

It has been shown that exothermic sleeves can cause flake graphite within ductile iron risers and that this flake structure can potentially extend into the casting. The next question is how do the different materials in an exothermic sleeve formulation contribute towards the graphite degradation. To answer this question, a design of experiment was performed looking at the percent of total aluminum, type of aluminum, percent of iron oxide, percent of fluorine-containing compounds and the percent of other types of initiators.

The results of the design of experiment showed that the amount of flake graphite was affected by the percent of total aluminum and any other ingredient that either reduced the ignition temperature of the exothermic reaction or increased its burn temperature. As the percent of total aluminum increased, the percentage of flake graphite also increased. In addition, when the ignition temperature decreased or the burn temperature increased, the amount of flake graphite increased. However, it was discovered that when fluorine is taken out of the exothermic sleeve formulation, no flake graphite was formed, even when the ignition temperature of the sleeve was decreased and the burn temperature was increased.

Discussion

Feeding ductile iron castings requires a higher-performance exothermic riser sleeve than has traditionally been used for steel applications. This is especially true in cold-riser applications where the metal has a chance to cool before it goes into the riser. In coldriser applications, exothermic sleeves that have a lower ignition temperature and burn hotter perform better. This is because of the lower temperature of the metal in the riser and the lower amount of superheat of the metal.

This need for faster ignition and hotter-burning exothermic riser sleeves also increased the degradation of the graphite nodules into flake graphite. However, with the development of the 0-Fluorine New Technology riser sleeve, this problem has been solved. In addition, the 0-Fluorine New Technology sleeve solves fish-eye defects created by the contamination of the molding sand with traditional exothermic riser sleeves. The high-performance, 0-Fluorine, New Technology exothermic riser sleeves also provide extremely efficient feeding for ductile iron cold-riser applications.

Conclusions

Although different in their appearance and formation, Fish-eye defects and ductile iron graphite degradation have one thing in common: they are both influenced by the use of exothermic riser sleeves. Ductile iron graphite degradation is caused by the use of exothermic riser sleeves, and Fish-eye defects are caused by the contamination of the molding sand from the same exothermic riser sleeves.

Using 0-fluorine New Technology riser sleeves can prevent both of these defects.

Good foundry practices should also be used to choose the best riser location to be able to feed the casting with the smallest riser possible. This will improve the foundry's casting yield and lower its overall cost to produce castings.

Acknowledgements

Thanks to Ben Carr and Mark Hysell in Ashland's Metals Application Laboratory for their assistance on this project.

References

"New Developments in Riser Sleeve Technology", R. Aufderheide, R. Showman, H. Twardowska, AFS Transactions, 1998, pp. 395-400.

"Graphitization of Pure Fe-C alloy During Annealing", D.N. Khudokormov, V.M. Kordev, Russian Casting Production, Dec. 1967, pp. 589-590.

"Observation of Reactions During the Combustion of Exothermic Materials by Differential Thermal Analysis", C. Pelhan, N. Majcen, paper to 1970 International Foundry Congress.

"Exothermic Riser Sleeves Can Cause Flake Graphite in Ductile Iron", R. Showman, R. Aufderheide, C. Lute, paper 01-086, AFS Casting Congress, Dallas, 2001.

Foundrymen's Guide to Ductile Iron Microstructures, AFS, Des Plaines, Ill., 1984.

Metals Handbook Ninth Edition, Volume 15, Casting, ASM International, 1988.