ANSYS Benchmark #7: Procedure
PADT has created this online benchmark database for the ANSYS user community. The success of this effort will rely upon the user community running and submitting results as follows.
Other benchmarks can be seen on the Benchmark Index.
Large Model Test
This benchmark is designed to test the capacity of your machine. Large memory 64bit machines will do well…others may not finish this before you need them back in production.
Simulation Details: A bolt holds 2 plates together with contact [friction] used between all parts. After preload is established, a fixed displacement is applied to one plate forcing it to slide relative to the other plate until the bolt hole clearance is closed.
The benchmark should be run in batch mode as the only significant process on you system. Results that are substantially faster or slower than other systems of similar configuration will be flagged for verification.
Procedure for Benchmarking
- Download the input macro file, bench07.mac, and run it in batch mode.
- A plot of the final usum will be created in the working directory, verify that the max displacement is close to
0.0783 (it will vary slightly depending upon solver, OS and version)
- An ouput file called bench7.dat will be created in the working directory.
- Feel free to vary the solver (eqslv) and any launch settings –m –db –lm –dis –pp –np etc… but don’t alter other solution settings like nlgeom,on or nsubst which would change the results
- Using the solution’s wall time and CPU time listed at the end of the output file to fill in the form below.
- Add relevant comments at the bottom of the form.
- Click the Submit button at the bottom of this page, and you will then be redirected to the Results List where your new entry should now appear.
Note:
- Linux hardware information can be found using cat /proc/cpuinfo and cat /proc/meminfo.
- Windows hardware information can be found using CPUZ [http://www.cpuid.com/cpuz.php
Benchmark Inputs
|