CAE opens new frontiers in casting design: applying cutting-edge, integrated computer-aided engineering (CAE) technologies to optimize casting designs can open doors for foundries, placing them at the head table with their customers.

The pace of change in today's marketplace is so rapid that time-to-market has to be the overriding priority. All too often casting designers feel forced to compromise on innovative casting ideas or approaches because of a perception that there simply isn't enough time. With rapidly evolving CAE technologies including comprehensive casting process simulation, automatic casting process optimization and new computer-based component design tools, metalcasters and casting designers can work together, concurrently, to optimize component design and casting process parameters up-front.

Many of today's metalcasters still don't see casting design as a function of the foundry. However, as end-users continue to push value-added responsibilities (machining, painting, assembly and, yes, casting design) down the supply-chain, it is advantageous for foundries to latch on to these responsibilities. By becoming a full-service supplier, a foundry can dictate a component's make-up to benefit themselves (higher profit margins) and their customer's product (an optimized component design that takes advantage of what the casting process offers). This second point is where CAE sets the stage.

The ever-increasing demands of casting customers for reduced prices with increased quality and performance requires cast components that are designed to their highest potential. The casting's optimal shape (taking advantage of part consolidation when possible) with the thinnest walls and necessary mechanical properties to meet expected loads and stresses must be designed in the early stages of production. Only when casting design is involved at the early stages of product design can the true potential of the casting process be unleashed. This usually only can be achieved, however, when metalcasters are present at the design table.

But metalcasters' presence at the design table only can be achieved when they have the necessary resources and technology (CAE) to assist in the optimization of a casting design. While CAE technology encompasses casting process modeling, it also reaches beyond it to incorporate casting process optimization and component design optimization.

For foundries to compete in 10 years on component price, quality and performance, they will have to rely on CAE technology as much as they rely on casting process modeling today.

This article will demonstrate the advanced implementation of CAE simulation technology with cast components using three metalcasting examples. Metalcasters working with designers can optimize component design in less time and with improved casting quality using these engineering tools.

Automatic Optimization

Although metalcasting design approaches and casting processes are flexible, few casting designers really take advantage of this flexibility. They design most castings using standard design rules, often limiting the use of technology tools to finite element analysis (FEA).

Within the last few years, the potential of automatic shape optimization has been explored, and today we can see the first casting designs created largely with automatic computational shape optimization. In several casting markets, this technology seems quite promising for optimizing cast component designs.

To show the potential of computational component design optimization, an aluminum rear suspension arm was designed. Only the mounting dimensions and the expected mechanical loads were given as initial conditions for this relatively simple casting design. Part design optimization tools were employed to obtain the initial component design. It was then translated to 3D CAD for cleanup, stress analysis and casting process. Simulation evaluated the feasibility of the design and the design finally was manufactured in a rapid prototyping casting process.

The required data for shape optimization was minimal; only the mounting locations/dimensions and the mechanical loads on the component were necessary. Other objectives included minimizing casting weight and meeting distortion specifications under load. The part was meshed and the optimization software began its analysis (Fig. 1).

[FIGURE 1 OMITTED]

A basic optimization rule is to remove material sections from the part that are below a specified stress level when applying the test load. The optimization resulted in geometry for the arm component that met the criteria. The optimized structure, available as a finite element mesh, then had to be reengineered using a 3D-CAD program so that the data could be used for further design changes and casting production layout.

A casting process simulation was performed to check this first casting design approach for possible casting defects (Fig. 2). The microstructure, resulting mechanical properties and anticipated residual stress also then could be assessed. Transferring the CAD data into the simulation program to review the results took only 30 min. The simulation delivered the distribution of local thermal modulus (the volume/surface ratio of a casting section). This allowed the casting designer to rig appropriately sized risers on the arm for these critical areas (Fig. 3).

[FIGURE 2-3 OMITTED]

The rigging layout then was applied to the first prototype casting. The first casting produced did show one 0.8-mm shrinkage defect when checked with high-resolution computer tomography. The defect found had been predicted previously at this position by simulation. This was considered a non-critical area and no change was made to the casting.

The contribution of various CAE technologies used for this project (automatic shape optimization, CAD, casting simulation and rapid prototyping) can best be measured by the short time spent to complete the project. Casting designers were able to go from receipt of specifications to a "zero" defect aluminum prototype in seven working days.

Process Optimization

Steadily increasing computer performance is a driving force for the application of CAE tools in casting development. Looking into the future, the potential of computational process optimization can be seen.

With a high-pressure diecast aluminum housing as an example, automatic casting process optimization was applied using a multi-processor computer for fast, closed loop improvement iterations. Various scenarios were considered by the software to determine not just an acceptable solution, but an optimum solution.

The aluminum housing produced in the high-pressure diecasting process was exposed to dramatic temperature changes from cycle to cycle. This led to substantial thermal stress demonstrated through casting process simulation (Fig. 4).

[FIGURE 4 OMITTED]

Optimum die life depends on smooth temperature gradients and moderate quenching or heating rates. Until now, when using simulation alone, engineers relied on computer-based trial and error--running simulations, reviewing the results manually and attempting further simulations to find the best solution.

With new computational optimization technology, the software can find the best thermal balance for the situation at hand. Automatic optimization allowed designers to establish target goals, consider process variables and take constraints into account. The results of each simulation were evaluated automatically, and the software proposed the appropriate process variables for the next simulation loop, which again was run automatically.

The closed optimization loop for an automatic casting process optimization consisted of the simulation engine, an automatic simulation result evaluation and an optimizer making decisions for parameter variations. In this case, modified cooling power and media were considered.

As a result of using automatic optimization, cycle times were cut and the overall solidification time of the casting reduced. Also, more homogeneous die temperatures were obtained, while maximum temperatures were reduced and variations minimized (Fig. 5).

[FIGURE 5 OMITTED]

The maximum die temperatures were reduced by 188F (87C). Variation during the casting cycle was reduced from 298.4F (148C) to a variation of 136.4F (58C). This provided reduced temperature gradients and minimized thermal shock during the die coating procedure. In addition, porosity was reduced in the process, thereby improving overall casting quality.

Design Optimization

Another example of "advanced state of the art" was demonstrated using a wheel design. The simulation result was an optimized casting design with respect to the casting practice--the wheel's behavior under load and the layout of a reliable production process.

At the same time, the simulation documented melt flow and cooling during die filling and solidification, as well as the residual stress development. Aspects of construction, tool manufacturing and the foundry process also were taken into consideration simultaneously.

Many casting designers outside the metalcasting industry use FEA to evaluate mechanical loads on a casting design. In this endeavor, few consider that the thermal stresses created in the casting processes often have an impact on the casting's quality. Neglecting such residual stresses might result in an incorrect analysis and substandard casting production.

The combined use of casting process simulation and stress analysis was documented using a magnesium automotive wheel casting. The wheel, gating system, casting die and the cooling channels were constructed directly in 3D CAD. The CAD model was used for comprehensive casting process simulation.

During simulation of the casting processes, mold filling, solidification and microstructure formation were examined as well as the development of residual stresses caused during the casting process. These calculated residual stresses were taken as a preliminary load on the wheel and analyzed further using FEA.

During the optimization of the wheel geometry and the casting process, aspects concerning the casting process and stresses were taken into account. The wheel was cast using a low-pressure diecasting process. The elements of a low pressure diecasting machine that are most important for the simulation are the movable upper die half (top core), the fixed bottom die half (bottom core) and the slides (side cores).

To calculate the residual stresses, the temperatures calculated during the casting process simulation were transferred directly for FEA. The temperatures were gradually deposited onto the wheel as external loads. The calculation results show maximum residual stresses of 45 MPa on the backside of the spokes. The residual stresses were caused by the hot spot in the wheel hub. In the FEA, a rotating bending test was carried out on the wheel with a test force of 3500 N applied at an axial distance of 600 mm from the hub.

The test force introduced a bending moment onto the wheel. This led to tensile stresses within the upper spokes and to compressive stresses within the lower spokes. Therefore, the tensile residual stresses were increased in the upper spokes, and, on the opposite side, the compressive stresses caused by load are reduced. The resulting stress at the most loaded spoke is approximately 150 MPa, which was above the fatigue strength under reversed bending stresses (80 MPa).

Taking into consideration the simulation results, the spokes of the wheel were reinforced in their cross section, and the base of the wheel was thickened continually toward the spokes. In the transition area between bottom core and slides, the tool was cooled additionally with air from the outside. Additional cooling also was placed in the area of the wheel horn, and cooling increased in the area of the hub at the end of filling.

On the one hand, the objective of these measures improved the directional solidification and thus the feeding possibilities in critical areas. On the other hand, the reinforcement of the spokes supported the reduction of stresses under load. Based on the simulation results, increased cooling of the hub minimized the temperature differences between the base and the hub (as shown during casting ejection). In this way, the residual stresses within the spokes were reduced significantly.

The result almost completely eliminated porosity (Fig. 6). At the same time, the total load on the wheel under the combination of residual stresses and the bending load was reduced to less than 75 MPa (Fig. 7). This met the acceptable criteria for both soundness and stresses.

[FIGURE 6-7 OMITTED]

Simultaneous Integration

The world of casting design is faced with some of the most revolutionary changes since the introduction of CAD in the 1980s. The newly available computer-based technologies such as casting process simulation, computational optimization of casting design and optimization of casting processes will alter the designing environment.

These examples show why CAE technologies including casting optimization, casting process simulation and stress analysis have to be used simultaneously and integrated into the design and manufacturing chain. For the future, metalcasting engineers will need to expand their knowledge and use of CAE technologies to communicate effectively throughout the entire process. This will be the only way to ensure that casting process conditions will be integrated early into the casting design process to take advantage of casting's true potential.

For More Information

"Optimized Development of Castings and Casting Processes," G. Hartmann and A. Egner-Walter, International Congress on FEM Technology 2001, Berlin, Germany (October 2001).

About the Authors

Magma Giessereitechnologie GmbH, Aachen, Germany, and its North American arm Magma Foundry Technologies, Inc., Arlington Heights, Illinois, develop casting simulation tools and answers for a variety of metalcasting processes and materials.

Tim McMillin Magma Foundry Technologies, Inc., Arlington Heights, Illinois Gotz Hartmann and Achim Egner-Walter Magma Giessereitechnologie GmbH, Aachen, Germany