Printer Friendly
The Free Library
23,403,340 articles and books

Using Computer Modeling to Optimize casting Processes.

Through examples, four engineers detail how computer modeling has resulted in structure and manufacturability improvements, while reducing lead times and production costs.

In today's global manufacturing environment, time is money. Casting buyers Want foundries to deliver quality cast components with short lead times, regardless of the number of steps in the component's production cycle. To this end, foundries must take a greater role in the initial design of its cast components as well as gating/ riser design instead of taking the cue from its customer. The one tool that has become indispensable to foundries working with their customers on optimized casting design with minimal lead times is computer modeling. Using solid models, finite element and fatigue analysis, and casting process modeling, foundrymen can make numerous no-cost design revisions on the computer screen.

In an AFS Engineering Div. Process Design & Modeling Committee panel presentation at the 2000 AFS Casting Congress, four casting engineers discussed casting examples that have benefited from analysis performed on computerized casting models. This article highlights some of those examples and provides foundries with some "food for thought" as to how the computerized casting modeling tool can aid them further in the design and casting of their components.

Optimized Designs

Computer modeling for most foundries focuses on casting process modeling. This analysis ensures the proper mold filling and solidification of the component. However, as foundries have begun to embrace casting process modeling, its capabilities have extended beyond this basic use.

Following are four examples in which casting process modeling combined with other computer modeling to optimize cast component manufacturability.

Component: A 54-lb ductile iron bracket.

Problem: The casting was experiencing three areas of porosity. Poured at 2600F (1426C) with a fill time of 9 sec into a green sand mold, the casting uses two 2.5 x 3.75-in. side risers and a 2 x 3 in. top riser. Although the casting fills with little turbulence during pouring, solidification modeling shows a temperature profile that isn't uniform. The hottest temperatures are at the bottom of the casting where the metal enters the mold, and the coldest metal is near the top of the casting and in the risers. Figure 1 illustrates the predicted porosity by the simulation and the porosity as shown in production.

Solution: In an attempt to solve the solidification problem, various stages of the casting modeling were analyzed. The solidification software time charts showed that the feed paths from the side risers were frozen off before the casting solidified. As a resuit, the shrinkage that produced the porosity couldn't be fed.

Three possible solutions were considered for the porosity problem:

* gating molten metal into the risers;

* increasing the riser contact areas;

* lowering of the side risers and angle contact.

Gating metal into the risers would ensure that the risers receive hot metal late in the filling cycle, however, that metal will "waterfall" into the casting, causing turbulence and reoxidation. Because the two other solutions didn't appear to pose any significant problems (except the extra cleaning cost for increased riser contacts), they were tried.

Increasing the contact area of the risers from 1.2 sq in. to 2.6 sq in. did alleviate some of the porosity problems, however, simulation predicted that porosity still would exist. Any further increase in contact area wasn't feasible due to the further increase in cleaning cost that accompanied the change. The solution was to lower the riser contacts and angle them downward. As a result, the area with porosity was fed more directly, eliminating the defect.

Component: A 92-lb steel digger tooth.

Problem: The foundry was importing its sand, resulting in high shipping costs. In an effort to reduce this expense, the foundry wanted to increase its casting to sand ratio by reducing the size of its molds. In the case of the digger teeth, the foundry needed to determine how close three of the castings could be placed within a single mold without adversely affecting casting quality.

Solution: Using simulation software, the foundry modeled castings with 1-4 in. of sand space between them (Fig. 2). The castings modeled with 1 in. separations showed a significant increase in solidification time (especially the center casting) due to the superheat generated in the sand. Castings modeled with 2 in. separations still showed the effects of superheat in the solidification times, but were lessened. The castings with 3 and 4 in. separations didn't appear to be affected by each other's heat.

The foundry also had to check the effect of the superheat on the feeding effectiveness of the top riser. As the solidification time of the casting increases, the effectiveness of the feeder decreases. Modeling was used to determine the "feed safety margin" for the risers in the same increments as the casting spacing was tested. As the modeling showed, 3 in. was the optimal distance between risers for casting feeding and solidification.

Component: Two 164-lb steel pump casings poured through one single-filtered riser.

Problem: The castings exhibited a visible porosity defect in the volute waterway. It was thought to be gas from an improperly vented shell core. However, the defect did not disappear after several changes in core practice.

Solution: Although the defect did not appear as shrinkage, it was investigated as a solidification problem via modeling. Three-dimensional fraction solid plots (Fig. 3) showed an isolated hot spot at the problem location on both castings. This also was clearly identified by the macroporosity plot.

In solidification modeling, a chill was placed on one of the two casting's problem areas. In further analysis of the 3-D fraction solid and macroporosity plots (Fig. 3), the shrinkage had been corrected on the casting with the chill, but not in the other one. As a result, a chill was added to both castings, solving the porosity problem.

Component: A 1138-lb stainless steel head casting for a double suction energy pump.

Problem: The configuration of the cooling jacket core posed serious doubts about the integrity of the wall around the stuffing box bore. The question was whether the narrow channel would solidify properly or exhibit heavy shrinkage.

Solution: In this case, the solution was to keep the same design as casting process modeling became an affirmation of the stability of the original design. The simulation was performed to determine if the side risers could feed the area in question. The fraction solid section plot (Fig. 4) indicated directional solidification from the area in question to the side riser, and the macroporosity plot indicated soundness in the area. In fact, the isothermal plot of the casting and mold section showed that the narrow channel over the jacket core was kept hot and open throughout solidification.

In use, this family of parts has not experienced rejections or failures in 4 yr of production.

Modeling Optimization at Design Conception

As opposed to the previous examples, some foundries become involved in the design of its cast components at conception. At this point, the foundry is able to optimize a cast component's structure and manufacturability from the beginning of the process instead of reengineering later in the process.

Two of the computer-aided engineering analysis methods used to optimize the structure of cast components and the process of metalcasting from conception are topology optimization and multidisciplinary response surface optimization. Both methods utilize casting modeling as the foundation for design iterations.

Following is a look at the two methods and examples of optimized cast parts.

Topology Optimization--Topology optimization is a method of layout optimization that is most effectively employed at the early stages of the design process, ideally before the design is conceived. In topology, the only input required is a layout of the package space, a definition of the loads and constraints on the structure, and a mass target. The optimization process starts with a uniform topology (uniform mass distribution) and then modifies the topology to minimize the compliance of the structure. In other words, the mass is moved around in the package structure to achieve the stiffest design for a given mass target. Multiple loading conditions can be input as well as weighing factors.

Topology optimization can quickly provide information to the design engineer as to the optimal layout of the mass before concept designs are developed. This can reduce the number of iterations significantly during design since the optimal part configuration will be known in the beginning. If employed on existing designs, the process often can be used to reduce structure weight while maintaining performance or improve performance while maintaining mass.

Figure 5 is a topology optimization case study of a Volkswagen engine bracket casting. At left is the original component design with a package space definition performed by filling in the cutout areas. These cutout areas on the model have mass since these areas are used to connect to other structures in the automobile. Seven different load cases also were defined for different operating conditions.

At right, in Fig. 5, is an illustration of the topology optimization results. The density isosurface plot shows the distribution of the mass for the object. This information was used to develop the final design as shown, which reduced the mass by 23% while still meeting all of the performance targets.

Multidisciplinary Response Surface Optimization--This is an optimization method utilized in the conceptual design stage to ensure manufacturability of design. Utilizing solidification modeling, this optimization process can increase the quality of casting and the productivity of foundries. The process is as follows:

1. Set-up the optimization problem, including the definition of the baseline problem and a description of the component geometry (finite element, finite difference mesh, material properties, etc.).

The process is as follows:

2. Define the process variables and constraints, including fill rates, mold temperatures, etc.

3. Define the shape variables and constraints.

4. Define the objective function (such as minimizing porosity or solidification time)--the key to optimization.

5. Set the input parameters, which include a solidification modeling input deck along with definitions of all variables, constraints and objective functions.

6. Submit the analysis to the solver system (software) and perform an initial run.

7. Compare the results of this analysis with the objective function. If the convergence criteria for the objective function is met, then the result is an optimized process. If it is not met, then the process repeats itself. The optimization software then will reconfigure the input to the solver and launch a new analysis. The optimization software controls the input of the analysis each time after checking the results of the previous run. The greater the number of variables, the greater number of iterations required.

In an example of response surface optimization with solidification modeling is a 4700-lb high manganese steel idler casting (for a roller mill) in Fig. 6 with two risers and a feeding channel. The variables are the riser height and diameter. The constraint is to maintain directional solidification. This was monitored by determining when the riser was frozen from the casting and calculating the percent liquid metal still in the casting. The constraint ensured that this percentage was zero, thus minimizing the solidification time.

Figure 6 shows the initial casting riser design and the results of this riser design after an optimization study. In the initial design, an insulating sleeve wasn't used on the riser. In the first design iteration, a 2-in, sleeve was added to the riser, which aided directional solidification. In further iterations, riser diameter and height were increased and decreased. The optimization analysis converged with six iterations at 12 min/iteration. The result was that the riser radius was increased by 0.1 in., the riser height was reduced by 3.6 in., riser volume was reduced by 13%, and there was a 20-mm reduction in solidification time. Although both the initial design and optimized design produce sound castings, the new geometry will save the foundry money by reducing the melt required and the cycle time.

Make It Right the First Time

The goal of computer modeling is to make the casting right the first time. By performing design changes on the computer screen, foundries save time and production costs. As the demand for shorter lead times continues, foundries taking the "cast and see" approach without computer modeling analysis will find it difficult to compete. This article was adapted from a panel presentation at the 2000 AFS Casting Congress.
COPYRIGHT 2000 American Foundry Society, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

 Reader Opinion




Article Details
Printer friendly Cite/link Email Feedback
Author:Faivre, Tony
Publication:Modern Casting
Date:May 1, 2000
Previous Article:Use Simulation to Analyze Macrosegregation, Hot Tears, Heat Treatment in Steel Castings.
Next Article:Southeastern Foundrymen Discuss the State of the Region.

Related Articles
Linking foundry operations with computer-aided engineering.
QUESTION: "What has been your greatest success story with solidification modeling and rapid prototyping? What other opportunities do you think this...
Modeling Saves $270,000 on One Aluminum Diecast Part.
Eliminating Modeling 'Trial & Error' with Casting Process Optimization.
Improving competitiveness through technological development: six AFS development projects aimed at improving quality, reducing cost and shortening...
CAE opens new frontiers in casting design: applying cutting-edge, integrated computer-aided engineering (CAE) technologies to optimize casting...
Sessions: current as of 2/20/05.
Charting a course for magnesium: magnesium casting is in an enviable position to enter new and growing markets, but a course of action must be...
Improving productivity through casting process modeling.

Terms of use | Copyright © 2014 Farlex, Inc. | Feedback | For webmasters