'Shop-floor' process controls for lost foam.

With more opportunity for variability than other methods, process controls are critical. Following is one foundry's approach to controlling the process.

Effective process control is key to the success of any foundry. Nowhere is this truer than in the lost foam foundry.

The lost foam casting process is more susceptible to process changes than many similar processes because of the numerous critical variables and the complex interactions between them - which often are not fully understood. As a result, the success or failure of any lost foam facility is largely determined by the process controls that are developed and practiced.

Over the last 10-15 years, lost foam has evolved rapidly to the point of being a viable means of producing complex castings that would be difficult and expensive to produce using conventional methods. Advances in process knowledge and technology have greatly removed the challenge of producing a casting that closely resembles the initial pattern. Instead, the most daunting challenges today are related to the production of large numbers of castings that consistently meet and exceed customers' quality requirements. Implementation and use of proper process controls is the only means by which lost foam casting producers can consistently and economically cast high-quality parts.

As in any stable casting process, defective castings are created by undesirable process changes, many of which may go unnoticed until the scrap is measured. Because lost foam is a new, rapidly evolving technology, it lacks the long history of research and development that has standardized process controls to a large degree in conventional processes. As a result, lost foam foundries lacking sufficient process controls sometimes utilize the "don't touch anything" approach when casting results are good and are often left in confusion when scrap castings are produced because of unnoticed process changes.

Determining Controls to Use

The ultimate goal of process control is to minimize undesirable process variations. As variations are controlled and the process is stabilized, castings that consistently meet customer quality requirements may be produced. In order to determine which controls are necessary, customer critical-to-quality (CTQ) items must first be identified. These CTQs, which also are known as process outputs (or Ys), are the casting properties deemed important to the customer. They may be as specific as a particular critical dimension or as general as casting appearance. They are the focus of the customer, and production of castings that meet CTQ requirements must be the goal of any foundry.

Once CTQs are identified, the variables that affect them must be identified. These inputs (or Xs) are the specific settings at each process step that can mean the difference between success and failure at any given time. In order to implement and use any process control, the inputs and outputs must be measurable, since it is impossible to control what cannot be measured and it is difficult to meet requirements that cannot be defined.

Once the critical variables are identified, and the means of measuring them and their effect on the CTQs are established, the means of capably controlling process variability can be investigated. This is the last, and most important, stage of process control implementation.

In order to maximize the benefits of process controls, items such as time requirements, associated costs and "ease of use" must be considered. For example, can the data be collected, processed and utilized in a manner timely enough to be effective in maintaining the process and the end product? Do the costs of the testing equipment, supplies and time required to perform the activities justify the amount of control provided? Are special operator skills required and are the control methods easy enough to be used by the people who must deal with them? These items must be addressed before full potential can be attained.

As in any industry, not all controls are used at every manufacturing facility, and opinions on necessity or effectiveness of particular controls vary. This article looks at the individual steps of the lost foam process from one foundry's perspective, identifying typical CTQs, inputs, outputs and specific controls utilized in each one. Additionally, many of the process controls required in the lost foam process also are required in conventional casting processes, particularly in areas of melting and finishing. While they are as important in lost foam as they are elsewhere, they are not covered in this article.

* Foam Pattern Inspection

Though numerous process controls are utilized for foam pattern production, the initial step with respect to the casting process is usually incoming foam pattern inspection. In order to produce "good" castings, it is necessary to start with "good" patterns. Since pattern quality dictates casting quality, CTQs are to strongly affected by the incoming patterns. The primary areas of concern are dimensions, cosmetics and castability.

Pattern dimensions obviously play a major role in determining final casting dimensions and must therefore be carefully controlled. For any new product launch, foam patterns from a new tool are evaluated for required features and measured using a coordinate measuring machine (CMM) to establish dimensional capability. Then, these results are compared to customer prints. Foam expansion and shrinkage also can cause dimensional issues and must be accounted for through shrinkage studies to determine the required aging time. Because of aging issues, dimensional control for multiple-piece foams can be improved through the use of precision gluing fixtures, which often won't accept any distorted or improperly aged foams. Additional dimensional process controls at pattern inspection include age verification and dimensional verification of patterns at specified intervals. Recent developments in air matrix-based measuring equipment promise to alleviate some of the current foam-measuring difficulties, making more precise measurements possible.

Pattern cosmetic concerns include bead fill and fusion, flashing at parting lines, open glue seams, and dents, cracks, tears, and other damage. The primary control is a visual inspection audit of a certain number of pieces per run.

Variations in the foam molding or gluing operations can create castability issues. To detect such variation, precise pattern weights are obtained from a random sampling of a predetermined quantity that represents the run. Significant changes in pattern weight can indicate changes in bead density, composition, bead blend ratios or glue amounts and can negatively impact the casting process. Obtained pattern weights may be plotted on statistical process control (SPC) charts for comparison with historical data and to determine trends or stability. Such charts allow "out-of-control" conditions to be differentiated from normal process variation and can be used to trigger appropriate operator reactions.

* Rigging 

The next process step is rigging, in which patterns are assembled onto gating systems either individually or as clusters. In the rigging process, CTQs are addressed through proper pattern-to-pattern spacing, formation of sound glue joints at the pattern-to-runner contact areas, and elimination of glue drips and foam damage due to improper handling. For low-volume applications, process control often is heavily dependent on operator training and skill, visual inspections and part-specific instructions and fixtures. As lost foam expands into high volume applications, automatic pattern-handling and gluing equipment may be incorporated to eliminate operator variability. In all cases, control of glue temperature and application is required in order to eliminate poor glue joints that can lead to poor casting cosmetics, coating and sand inclusions and metal fill defects.

* Coating

Perhaps the most critical group of inputs in the lost foam process is found in the coating step. Coating provides a physical barrier between the compacted sand and the pattern or metal and can affect heat transfer from the casting to the mold. It also plays a major role during mold filling at pouring, as it controls the removal of the liquid and gaseous products generated by foam decomposition, which in turn controls the metal front as it fills the mold. Critical variables may be categorized under either coating properties or coating application technique, with both able to dramatically affect customer CTQs in areas of surface appearance, casting defects and metallurgy.

Multiple standardized coating tests are performed on incoming materials and on coatings prior to and during use, including viscosity, temperature, permeability, density, percent solids and bacteria tests. Usually, a single test is used to set up a tank for usage, and the remaining tests are used to verify that the coating meets established criteria. Test results are used to make setup changes, and results may be plotted using control charts to verify stability, identify potentially damaging trends and indicate out-of-control conditions. Established handling and testing procedures are used along with operator training to ensure that incoming material is acceptable and that coating properties remain stable.

For low- and medium-volume jobs, part-specific coating instructions and operator training are important for control of coating techniques. For higher volume jobs, automation may be implemented to reduce operator variability. Regardless of the coating application method, wet weights of a specified number of clusters are taken immediately after dipping and may be plotted using SPC charts. Results reflect the actual amount of coating per cluster and can be used to verify dipping or draining techniques and coating properties.

* Drying

After coating, the wet clusters are placed in humidity- and temperature-controlled rooms for drying prior to molding. Primary CTQ considerations include dimensional distortion due to excessive temperature and cosmetic issues that can be aggravated when clusters that have not been thoroughly dried are cast. Cluster dry weights are used to verify dryness and can give feedback on coating properties. Control charts may be utilized to clearly indicate variations in coating properties or drying times. Other process controls include automatic dry-house temperature and humidity controls, use of tags to verify drying time, visual inspection and part-specific rack-loading instructions.

* Molding

In the molding process, clusters are placed on a bed of sand, and loose, unbonded sand is rained around the cluster. As the sand is introduced, vibrational energy is applied to move and compact the sand around the patterns. One CTQ issue is dimensional distortion of the foam patterns caused by moving sand during the molding cycle. Other CTQ areas are related to sand compaction, including metal penetration, actual "noncompaction" defects and "mold lift" defects caused by insufficient compaction or burden above the pattern. Sand composition and properties such as temperature, grain fineness, angularity, and loss on ignition also can affect the molding process and casting results.

Molding process controls include the use of automatic molding "recipes" that set sand fill rates, times and energy inputs at various stages in the molding cycle. Part-specific instructions combined with visual inspection for proper sand level at various stages are utilized to verify proper selection of a part-specific "recipe" and ensure proper equipment operation. Sand system temperature control, sand tests and part-specific fixtures and molding aides also are used.

* Pouring

Primary pouring CTQ concerns are related to casting cosmetics and the occurrence of casting defects such as porosity, inclusions and collapse. Coating control becomes most critical during the pouring process, as it plays a major role in controlling the fill of the mold "cavity." Part- specific settings such as metal temperature range, pouring times and vacuum levels are specified and operator instructions and training are used in conjunction with manual pouring processes.

* Cleaning and Finishing

Process controls in the cleaning and finishing processes are similar to those used in conventional processes, with emphasis placed on dimensional verification through hard-gauging, fixturing and layouts using CMM and calipers. Part-specific instructions normally include visual inspection for casting defects, and process audits are used to monitor the casting process in general. As it is difficult to "inspect in quality," process controls in this last process step primarily serve as verification that the process controls upstream are operating properly.

Total System Control

When the process is properly controlled, no conventional casting process can match the "deliverables" of lost foam. Seeming subtle process shifts, however, can lead to tremendous casting difficulty ties and high scrap rates, particularly when such variations go unnoticed.

While lost foam process controls have evolved over recent years through improved tests, equipment and expanded process knowledge, customers continue to demand more from the process in the form of tighter tolerances, reduced piece weights and machining stock, and general quality improvements. As a result, only those foundries that implement and use effective controls at each process step can maintain processes and consistently produce castings that meet CTQ requirements and exceed customer expectations.

This article was adapted from an oral presentation at the 1998 AFS International Conference on Lost Foam. Conference Proceedings are available from AFS Publications at 800/537-4237.

RELATED ARTICLE: Citation Foam Casting Co.

Having first shipped production castings only 11 years ago, today's Citation Foam Casting Co. offers the largest independent iron foam casting capacity in the U.S. at 26,000 tons (two shifts). The foundry pours all standard grades of gray and ductile iron, and produces more than 80 different part numbers in weights of 2-700 lb for customers such as GM Powertrain, Mercedes Benz, General Electric, Dana Corp., MerCruiser, Volvo Truck, Freightliner, Caterpillar, Detroit Diesel, Lincoln Electric, U.S. Motors and others.

The 200-employee firm has undergone much expansion in recent years, including the addition of a second and third molding line in 1996 and 1997. The foundry reportedly maintains in-house scrap and customer returns that consistently rank among the lowest of any division in the Citation family. Further, it has reported that its mature, high-volume parts typically run at 2% or less internal scrap with less than 0.3% customer returns.

Michael J. Lessiter, editor