Which tool should I use?

Hi all.

I have to simulate an impact of a pendulum on a test piece. Normally, it wouldn't have any problem, but the impact has to be repeated. That is, the pendulum strikes the specimen, returns to the initial position, and strikes again. This process is repeated for X times. Which tool can I use to perform the test? Explicit dynamics? Static structural? I really don't know...

Thank you very much.


  • mjimji Forum Coordinator

    Generally impact analysis are handled by explicit dynamics, you can use either Ansys Explicit Dynamics or Ansys Ls-Dyna, both can accomplish this goal.

  • @Catacaldos

    I have seen this question asked several times in the last couple of years, but I have not yet seen a detailed answer on the workflow to achieve repeated impacts.

    One idea, that is very constrained in its implementation details, is ball impact where X balls are lined up in a row and given an initial velocity toward the test piece. The test piece is angled so that each ball bounces off in a direction that clears the way for the next ball to impact. The balls have to be spaced in the setup so they don't collide with each other. Requiring an angle of impact probably is a significant change from the physical test.


    I don't know how to have a single ball make the first impact with the test piece, then have the simulation restarted with the single ball returned to its starting position for the next impact while the test piece remains deformed and retains the stress of the last impact for the next impact, and to repeat that process X times. Please provide details on how this might be done in Explicit Dynamics.

  • mjimji Forum Coordinator

    One can define multiple steps in Ansys explicit Dynamics,

    step 1, with a very short duration, prescribe a velocity.

    step 2, deactivate the velocity BC, let impact happen, solver calculate the part velocity

    step 3, prescribe the velocity BC to the second impact velocity

    Step 4, deactivate the velocity BC and second impact

    With Ls-Dyna, you can define the velocity with keyword.

  • @mji

    Thanks for this idea.

    Step 1 can use a Velocity Initial Condition and have a step end time long enough to allow the deformation of the part and some rebound of the hammer.

    You need a velocity BC to activate and deactivate to get the "Initial Velocity" effect on subsequent steps for a very brief time so it is not part of the impact, because you can't set an Initial Condition for the second and subsequent steps.

    This would only work well for the case when there are constraints keeping the hammer from having only 1 DOF, the impact direction.

    The End Time of each Step would need to be carefully selected so the hammer has either stopped moving forward or is rebounding.

  • Thanks for your answers.

    But, if I make this, the computer is going to be "thinking" for days, don't it?

  • @Catacaldos Yes it will.

    @mji There is a problem using this approach of assigning a velocity BC over a very short time during a simulation. It is not the same as an Initial Condition on the body, which assigns that velocity to every node in the body.

    When I have tried to assign a velocity BC in liu of an initial condition, I could not pick a Solid, only a Face. Even if I pick every bounding face of a body, the nodes on the inside of the body did not get assigned the velocity BC. I tried using a Nodal Selection that included all the nodes in the body, but I could not use a Nodal Selection as a Velocity Boundary condition.

Sign In or Register to comment.