Hardware
“Simulation is always too slow“. This statement, first voiced twenty years ago, is, in principle, still valid. If a foundry or their customer/designer has a question regarding the casting process or the properties of a casting, an answer is needed immediately. At the same time, questions are getting more detailed and more comprehensive. Hence, the physics to an- swer these questions gets more and more complicated. This leads to an increase in calculation time, which literally elim- inates the progress that has been made in computer speed over the last twenty years. The hardware development has made a significant contribution to the distribution and utili- zation of casting process simulation. Computers are able to calculate at least one thousand times faster than fifteen years ago, but at the same time, they cost less than one twentieth of what they did in 1990. In spite this positive development, the challenges for future hardware are still equally daunting: ten times higher resolution, using ten times more physics, has to be calculated ten times faster than today.
This only relates to requirements for simulation itself. The tool of the future is autonomous casting process op- timization, meaning the sequential or parallel calculation of many designs to automatically find the optimum solu- tion. The pathway out of this “poly-lemma” is the paral- lelization of the software. Future speed increases will no longer be defined by the frequency of a processor, but by the number of integrated CPUs (cores) (Fig. 27). There- fore, the challenge lies with the software developers. The hardware provides the platform, but the utilization of the architecture requires specific new ways of programming. A 4-core CPU is already the standard configuration of mainstream PCs. In a few years, we will find cluster com- puters under the desk of every engineer, perhaps without him even knowing it. The vision of casting process sim- ulation is to interactively modify geometries or process parameters on-screen and to receive an almost immediate answer from the optimization tool.
Figure 26. Different microstructures have an impact on fatigue life. In this case the prediction of the fatigue life was initiated for a suspension part, benchmarking sand (prototype) versus permanent mold (series production) casting processes. The microstructure of the castings was hereby transferred into the fatigue life prediction tool. It was shown that the fatigue life of the permanent mold casting exceeded the one of the sand casting. (A. Egner-Walter et al. 2004)
Figure 27. Development of casting simulation performance since 1995. The most powerful workstations in 1995 were approximately 120 times slower than today. The utilization of parallelized software algorithms and corresponding multi-processor hardware (clusters) leads to calculation times of one one-onethousands of the old calculation times.
20 International Journal of Metalcasting/Spring 10
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88 |
Page 89