GEOMETRY DEPENDENT PATTERN ALLOWANCE PREDICTION FOR CASTINGS
Open Access
- Author:
- Kochar, Vishal Madanlal
- Graduate Program:
- Industrial Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- April 24, 2006
- Committee Members:
- Robert Carl Voigt, Committee Chair/Co-Chair
Richard Allen Wysk, Committee Member
Timothy William Simpson, Committee Member
Md Amanul Haque, Committee Member - Keywords:
- Pattern allowance
casting
geometry
modulus
steel
restraint
solidification
parting line
heat treatment - Abstract:
- Castings have been used for thousands of years and are an important part of the manufacturing industry today. Among the various casting alloys, steel castings are able to maintain their market share because they are inexpensive, readily cast to near-net-shape, are easy to weld and have good mechanical properties. Sand casting is the dominant and most economical method for casting steel. Complex, dimensionally accurate steel castings depend on using proper shrinkage (pattern) allowances during the production of the initial pattern. The correction factor applied to a pattern's dimensions in order for it to produce acceptable castings is called the "Pattern Allowance." One area in which the steel casting industry can reduce costs and increase its market share is by preventing the formation and repair of dimensionally non-conforming castings. It is not uncommon for steel foundries to take several pattern adjustment and casting trial cycles between when a customer places an order and when the first acceptable casting is ready. A key way to ensure that the dimensional conformance of castings without pattern dimensional rework is to better predict initial pattern allowances. This in turn would significantly reduce the lead-time to produce dimensionally acceptable castings. This research investigates the nature and causes of dimensional errors in steel castings produced in green sand, no-bake and shell molds. The significant process and design variables that affect dimensional variability are identified. Also errors introduced into casting dimensions due to incorrect pattern allowances (shrinkage allowances) are quantified. The major process and geometry variables that affect steel casting shrinkage allowances are identified. The influences of casting geometry and mold restraint and their interaction on the shrinkage behavior of feature dimensions are quantified, and predictive models for improved pattern allowances have been developed. Higher pattern allowance variation is observed in features of a casting made in green sand, followed by no-bake and shell, respectively, for similar size features. Pattern allowances measured in this study ranged from -8.0 % to 13.0 %. Features crossing the mold parting line show similar within-feature pattern allowance variations but have much lower nominal pattern allowance values that are usually negative. The amount of variation due to geometry increases as the slenderness ratio of a feature increases. There is a significant effect of heat treatment on the casting dimensions and the effective pattern allowance values. There is an average increase [0.61%] in the pattern allowance indicating that most features shrink further after heat treatment. Pattern allowances can vary somewhat from foundry-to-foundry for same casting; however, the geometry dependent trends in pattern allowance were consistant from foundry to foundry. Pattern allowance values are found to be dependent on molding method, restraint type, degree of restraint, feature slenderness, feature modulus and parting line location and a multiple linear regression model incorporating these factors is developed. The geometry dependant pattern allowance [GDPA] model showed 36% improvement in pattern allowance error when compared to the standard pattern allowance value of 2.08%. About 85% of the features of a validation casting showed improvement in predicting pattern allowance estimates using the GDPA model. Even though this research focused primarily on steel castings, the methodology used in this research can be easily applied to different materials that undergo shrinkage upon solidification.