2Concept To Reality Spring 2002

By applyingoptimization

technology to casting

process design,manufacturers canproduce less-expensive,better-quality parts inrecord time.

This model of a process design for a cast ironflywheel was optimized with OPTICastsimulation software.

c2r Smiley feat 5/9/02 9:28 AM Page 2

O

3Concept To Reality Spring 2002

P R O C E S S A U T O M A T I O N

By Lawrence E. SmileyFinite Solutions, Inc.

One of the most fundamental processes for manufacturing metalparts is casting. This technology can be traced back severalthousand years to a time when ancient civilizations cast weaponsand cooking utensils. Today, casting thrives and is widely used tomake automotive, aeronautic and machine parts.

The science of process design for casting has developedrelatively slowly until recently. Rule-of-thumb and empiricalprocedures, along with much trial and error, characterized thisdiscipline over the past century. Within the last decade,computer simulation of the process has come into general use,providing the casting engineer with a more scientific basis upon

which to design. However, evenwith simulation, the designprocess itself has tended to bebasically a trial-and-error searchfor an “adequate” design.

The state of the art insimulation is now changingeven more rapidly. In August2001, the first commercialsystem for optimization of thecasting process was released.Based on a marriage of AltairEngineering’s optimizationsoftware and Finite Solutions’SOLIDCast simulator, the newsystem – dubbed OPTICast –offers casting engineers arational and automated designmethodology that was notpreviously available.

Riser RelevanceCasting engineers struggle with physics to produce a good,

high-quality part. One of the basic complications is that whenliquid metal cools and freezes, it contracts and becomes dense. Ifthe casting process is not designed correctly, then thiscontraction will result in holes, or “shrinkage porosity,” withinthe metal part. Such a defect can cause a cast part to fail understress or even interfere with the function of the part if machiningoperations expose the defect.

The usual solution is to provide a shape attached to thecasting which can serve as a delivery source of liquid metalduring solidification, “feeding” metal into the casting tocompensate for the contraction. In the jargon of the trade, these

Get theMost

from YourCastingProcess

c2r Smiley feat 5/9/02 9:29 AM Page 3

which features of the model would be made available as“design variables” and which outputs could be used for“objective functions” and “constraints.”

For example, developers designated the size of the riser,initial material temperatures and pouring time as keydesign variables. They determined that one of the mostuseful combinations of potential constraints and objectivefunctions involved specifying the level of predictedshrinkage porosity within the casting as a constraint andasking the system to maximize the process yield – definedas the ratio of the net weight of the casting to the weight oftotal metal poured – as the objective function. This resultsin the system finding the least amount of metal which willproduce a cast part with a desired quality level.

OPTICast in ActionThe optimization software speeds up the casting design

process, as illustrated in the following example, in whichthe casting designer’s goal is to decide if the current designhas resulted in a “good part” at the lowest possible cost.

The first step is for the user to identify the modelfeatures which are to be considered design variables, asshown in Figure 1. Here, one of the large side risers,displayed in red, has been selected. The minimum andmaximum scale factors are specified, defining the designspace for this riser. The user also needs to indicate a “pinpoint,” which is a point on the selected shape that will stayat a constant location while the shape expands and

contracts. Once these items are specified, the next feature

(one of the other risers) is selected as anadditional design variable, and the sameinformation is entered. This continues until allrisers have been selected.

The user then optionally selects one or moreconstraints and the objective function from alist of system outputs. Once these selectionshave been made, the optimization project islaunched. This just requires the user to clicka menu selection. The process after thatpoint is totally automatic. The software runs a series of process

simulations with changes in the design variables. The systemautomatically modifies the model each time, and results arecompared to the constraint(s) and objective function.

Once the optimization is complete, the final model isavailable for display and postprocessing of results withinSOLIDCast. In addition, the progress of the optimization

shapes are known as “risers” or “feeders” and are removedfrom the casting net shape by cutting or grindingoperations.

This situation creates something of a dilemma for thecasting engineer. Larger risers tend to ensure the absenceof shrinkage defects in the casting. However, risers add tothe cost of producing a part: the metal in the risers is notpart of the finished casting, but it must be melted, poured,removed and handled as part of the process. Therefore, thesmaller the risers, the lower the cost of producing a casting.

Given today’s extremely competitive environment, it isessential to find the “perfect” riser size, not only to keepdefects from forming but also to hold extraneous metal costto a minimum. The OPTICast system does just that – andmore.

On the Optimization PathThe initial design of the OPTICast system began in early

2000 as a project sponsored by the Edison MaterialsTechnology Center (EMTEC), an organization in the stateof Ohio whose mission, in part, is to encourage technologydevelopment and application in materials processingindustries, including foundries. EMTEC provided thefunding for the project and brought together theparticipants, whose mission was to link casting simulationand optimization technology to further improve the castingdesign process.

To accomplish that goal, Altair and Finite Solutions’team members developed an interface to the Altairoptimization software and created a method toautomatically adjust model geometry and initialconditions based on changes suggested by theoptimization system. Developers considered otherissues, such as identifying

4Concept To Reality Spring 2002

To optimize thedesign of one ofthe flywheel’srisers, the usersimply selectsthe riser (shownin red) and entersthe horizontaland verticallimits on its size.The softwaredoes the rest.

Fig. 1

c2r Smiley feat 5/9/02 9:29 AM Page 4

can be examined eithergraphically or by loadinga spreadsheet showingprogressive values ofdesign variables andoutput data.

Seeing is BelievingOne of the reports

OPTICast generates is a plotof progressive values of theobjective function. Theprocess yield in the examplewas increased from an initialvalue of 63% to more than80% by running 15 processsimulations. This yieldincrease was achieved bydownsizing the risers to about73% of their initial size.

In this case, the amount of metal saved per casting pouredwas about 100 lbs (45 kg); total material savings were 75,000lbs/yr. The annual cost savings for production of 750 partsamounted to about $9,000 (U.S.) dollars. What’s more,annual energy savings due to reduction in the amount ofmetal required to be melted were about 27,000 kwhr, with aprojected cost savings in kwhr of $2,700/yr.

For verification of the optimization results, the user canexamine the final optimized model. For example, a time plotshows progressively solidified areas within the casting andrisers (Figure 2). Another plot demonstrates calculatedmetal feeding patterns due to contraction of the metal(Figure 3).

From one relatively simple example, you can see thatoptimization has the potential to significantly reduce costsand energy usage while maintaining or improving quality inthe casting process. The total amount of time required todefine this project, run the optimization and analyze resultswas less than half a day, yet the annual payback was morethan the initial cost of the software. Given the number andvariety of parts cast in the world industry, the application ofoptimization to casting process design has tremendouspotential.

In addition to yield improvement, it is possible toapproach quality maximization for critical parts throughproper selection of constraints and the objective function.For example, the strength of the material in a casting isrelated to the cooling rate of the metal while the casting is

solidifying: higher cooling rates generally translate intobetter material properties (higher yield and ultimatestrength). Optimizing a casting for maximum cooling rate,while constraining the formation of internal defects, canresult in a part that can withstand higher loads – and thatperforms better over its lifetime.

In the past several months, there have been a number ofsuccessful applications of OPTICast in a variety of castingprocesses. EMTEC Program Manager Percy Gros says, “Wehave member companies that have expressed their desire toimmediately make great use of this new OPTICast product.”

Application of optimization to casting process designessentially represents a new paradigm in the industry. It is nolonger necessary for the design engineer to examine eachprocess simulation and make a decision as to whether theprocess is complete, and if not, what should be changednext. Using optimization, the engineer can spend his/hertime more productively on problem definition. This nudgesthe design engineer, perhaps unconsciously, into taking amore rational approach to the overall design process and willultimately result in more successes. Those companies thataggressively adopt this new technology will find themselvesat a competitive advantage in today’s global market.

5Concept To Reality Spring 2002

P R O C E S S A U T O M A T I O N

Fig. 2

Fig. 3

Postprocessing results fromOPTICast include a time plot ofprogressive solidificationwithin the cast part (Fig. 2) anda plot of final results showingfeed material as supplied by therisers (Fig. 3). Noindication ofshrinkage porosity isseen in the castingitself.

Lawrence E. Smiley is President of Finite Solutions, Inc.,Cincinnati, OH. Before launching the company in 1993, heworked at Reliable Castings Corp., Bishopric Products Co.and Republic Steel Corp.

c2r Smiley feat 5/9/02 9:29 AM Page 5