Following Cadence Design Systems’ acquisition of the Design & Engineering (D&E) business, the D&E Community Forum on the Nexus Community Center has been migrated in the Cadence Physical System Simulation Community. For more details and if you have any questions, please refer to the FAQ.

Stats

  • State Not Answered
  • Replies 3
  • Subscribers 1
  • Views 39
  • Members are here 0
More Content

Performance when running complex analysis with Ftire and flex bodies

Vojtech Brychta

Hello, I need to run a long analysis of a vehicle running on a rough road. I need to use Ftire + relatively large flex body (20GB). The obvious issue is that the computational time gets out of reasonable timeframe. However, the CPU load remains very low. I have then following questions:

What steps should I take to improve the performance when using FTire and Flex bodies?
Is there a way to fully utilize the performance my computer provides?

Thank you.

SIncerely,

Vojtech Brychta

Parents
  • 0

    Hello Vojtech

    Could you please let us know what are the solver settings and integrator that you are using? Here are some general suggestions-

    1. Turn off FTire’s separate animation window if it is on (e.g., SEPARATE_ANIMATION…0 ) and run in batch/no-GUI.

    2. Reduce flex modes / use better reduction (CMS) to shrink the 20 GB model, and lower rough-road resolution to only what’s needed for your target frequency range.

    3. Set a sensible thread count (NTHREAD ≈ number of physical cores) and, if using CPU affinity, ensure the thread count does not exceed the number of cores in MSC_ADAMS_THREAD_AFFINITY_SET0.

    4. Make sure you have enough RAM so the run doesn’t page to disk, and keep model/temp files on a fast local SSD.

    Regards,
    Rishi

Reply
  • 0

    Hello Vojtech

    Could you please let us know what are the solver settings and integrator that you are using? Here are some general suggestions-

    1. Turn off FTire’s separate animation window if it is on (e.g., SEPARATE_ANIMATION…0 ) and run in batch/no-GUI.

    2. Reduce flex modes / use better reduction (CMS) to shrink the 20 GB model, and lower rough-road resolution to only what’s needed for your target frequency range.

    3. Set a sensible thread count (NTHREAD ≈ number of physical cores) and, if using CPU affinity, ensure the thread count does not exceed the number of cores in MSC_ADAMS_THREAD_AFFINITY_SET0.

    4. Make sure you have enough RAM so the run doesn’t page to disk, and keep model/temp files on a fast local SSD.

    Regards,
    Rishi

Children
  • 0 in reply to Rishikesh Mokashi

    Hello Rishi!

    thanks for the information!

    So far I have done following:

    - reduced the visualization of the MNF to lower the size of MNF (from 20GB to ~8GB) - doesn't really help with the calculation speed

    - run a small sample at reduced sample rate and then disabled modes by energy tolerance

    - run the analysis multithreaded in batch mode, keeping everything in the same place.

    The run is sped up considerably while maintaining similar results. The dynamic solver setting is following:

    Integrator GSTIFF (tried HHT, the driver wasn't able to follow a trajectory so the sim failed - not sure what's the reason)

    Formulation I3 (tried SI2 - didn't see any impact on performance)

    Error - 0.4

    Hmax 1e-3

    Kmax 6

    Maxit 10

    Any suggestions on the solver settings ?

    Thanks again!

    Have a nice day,

    Vojtech

  • 0 in reply to Vojtech Brychta

    Hi Vojtech,

    I would expect the HHT to be faster, however since you mentioned the sim failed it would be good to investigate the model to check what the reason behind this is. Please feel free to contact support by logging a request.

    Regarding solver settings, you may want to test with the following settings

    GSTIFF + Original corrector

    Error seems very loose, you may want to tighten this say1e-02 to start with

    Hmax: 0.005

    Regards,
    Rishi

Cadence Guidelines

Community Guidelines

The Cadence Design Communities support Cadence users and technologists interacting to exchange ideas, news, technical information, and best practices to solve problems and get the most from Cadence technology. The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information. By accessing, contributing, using or downloading any materials from the site, you agree to be bound by the full Community Guidelines.