AnsweredAssumed Answered

contact with friction at 2000 Hz in motion analysis

Question asked by Dan Shi on Nov 29, 2013
Latest reply on Dec 2, 2013 by Jared Conway

Like • Show 0 Likes0
Comment • 6

Hello,

I have one model, a ball bouncing on a vibrating plate in the motion analysis.

When the plate vibrates at 50 Hz, Motion analysis - 1000 frames per second, the model works well for the contact between ball and plate.

It works well both for material and friction consideration.

However, when the frequency is 2000 Hz, Motion analysis setup is 10,000 frames per second, the model works well when only considering the material.

Considering friction, it stops work when the ball impacts with the plate for the first time.

Could someone give me some advice for what setup should I change?

Thank you very much for any advice.

Bill McEachern Nov 29, 2013 10:19 AM

Correct Answer

what happens when you keep the frames per cycle the same? Does the 50 Hz anaysis work at 5 frames/cycle as opposed to the 20 frames/cycle that does work?

See the reply in context

No one else had this question

Outcomes

Bill McEachern Nov 29, 2013 10:19 AM

what happens when you keep the frames per cycle the same? Does the 50 Hz anaysis work at 5 frames/cycle as opposed to the 20 frames/cycle that does work?

Actions

Dan Shi @ Bill McEachern on Nov 29, 2013 11:26 AM

Thanks very much. For 50 Hz, 5 frames/cycle does not work either.

So I change to 20 frames/cycle for 2000 Hz. And I found the vibrating amplitude also influence whether the setting frames/cycle is enough. For example, the plate vibrating amplitude increase, the frames/cycle should increase. From my view, it seems that if the contact force is bigger, the frames/cycle should be bigger. Could you give me some advice how to choose the frames/cycle?

Actions

Bill McEachern @ Dan Shi on Nov 29, 2013 12:10 PM

Its a time resolution thing. The faster the events the more frames you need to supply enough time resolution on the computations form them to capture what you want. The response magnitudes would be a second order effect in my estimation. In one cycle how much information is necessary to capture the fidelity you want and to keep the solution from accumulating errors that can leads to divergence.

Actions

Jared Conway Nov 30, 2013 2:19 PM

I also suspect the number of frames is the most important here but also whether you are using built in materials or your own and also the contact and accuracy settings and solver could have an effect.

Actions

Dan Shi @ Jared Conway on Dec 2, 2013 4:01 AM

I use the built in material.

And now the choice for material is Acrylic, Aluminum, Nylon, Rubber, Steel. Is it possible that I can choose the specific material or other material, such as Aluminum 3003-H18, or there are not so many differences among different Aluminum for the contact?

As for the contact and accuracy setting solver, is it the accuracy in the motion study properties?

I have already chosen Use Precise Contact.

For the Accuracy, what is the basic principle to choose? What I use now is 0.0001.

For the Cycle rate is 1cps.

Could you give me some advice on how to set up this?

Thanks a lot.

Actions

Jared Conway @ Dan Shi on Dec 2, 2013 1:12 PM

"it possible that I can choose the specific material or other material," > yes you can if you have the properties

"or there are not so many differences among different Aluminum for the contact? " > this is probably also true

your other settings are fine, at this point i'd recommend posting a model if you're still having trouble. there is no right/wrong answer to this, it is a matter of trying different values and reviewing the documentation on what is recommended and how it works

Actions

6 Replies

Bill McEachern Nov 29, 2013 10:19 AM
Unmark Correct
Correct Answer
what happens when you keep the frames per cycle the same? Does the 50 Hz anaysis work at 5 frames/cycle as opposed to the 20 frames/cycle that does work?
Like • Show 1 Like1
Actions
- Dan Shi @ Bill McEachern on Nov 29, 2013 11:26 AM
  Mark Correct
  Correct Answer
  Thanks very much. For 50 Hz, 5 frames/cycle does not work either.
  So I change to 20 frames/cycle for 2000 Hz. And I found the vibrating amplitude also influence whether the setting frames/cycle is enough. For example, the plate vibrating amplitude increase, the frames/cycle should increase. From my view, it seems that if the contact force is bigger, the frames/cycle should be bigger. Could you give me some advice how to choose the frames/cycle?
  Like • Show 0 Likes0
  Actions
  - Bill McEachern @ Dan Shi on Nov 29, 2013 12:10 PM
    Mark Correct
    Correct Answer
    Its a time resolution thing. The faster the events the more frames you need to supply enough time resolution on the computations form them to capture what you want. The response magnitudes would be a second order effect in my estimation. In one cycle how much information is necessary to capture the fidelity you want and to keep the solution from accumulating errors that can leads to divergence.
    Like • Show 1 Like1
    Actions
Jared Conway Nov 30, 2013 2:19 PM
Mark Correct
Correct Answer
I also suspect the number of frames is the most important here but also whether you are using built in materials or your own and also the contact and accuracy settings and solver could have an effect.
Like • Show 0 Likes0
Actions
- Dan Shi @ Jared Conway on Dec 2, 2013 4:01 AM
  Mark Correct
  Correct Answer
  I use the built in material.
  And now the choice for material is Acrylic, Aluminum, Nylon, Rubber, Steel. Is it possible that I can choose the specific material or other material, such as Aluminum 3003-H18, or there are not so many differences among different Aluminum for the contact?
  
  As for the contact and accuracy setting solver, is it the accuracy in the motion study properties?
  I have already chosen Use Precise Contact.
  For the Accuracy, what is the basic principle to choose? What I use now is 0.0001.
  For the Cycle rate is 1cps.
  Could you give me some advice on how to set up this?
  
  Thanks a lot.
  Like • Show 0 Likes0
  Actions
  - Jared Conway @ Dan Shi on Dec 2, 2013 1:12 PM
    Mark Correct
    Correct Answer
    "it possible that I can choose the specific material or other material," > yes you can if you have the properties
    
    "or there are not so many differences among different Aluminum for the contact? " > this is probably also true
    
    your other settings are fine, at this point i'd recommend posting a model if you're still having trouble. there is no right/wrong answer to this, it is a matter of trying different values and reviewing the documentation on what is recommended and how it works
    Like • Show 0 Likes0
    Actions

contact with friction at 2000 Hz in motion analysis

Outcomes

Related Content

Recommended Content