The switch part has also influce on the dynamic results ?

Hello Everyone,

I built two suspension templates. The only one difference is with and without Switch Part.

Meanwhile the topologieplan between them is the same. That means, force or/and joint elemnt connect always between A part and B part.

What I can not understand:

1. There come out the difference of resutls on the dynamic simulation (the results in quasi-static simulation are the same) . On my opionion, the results should be totly the same. Because switch part = part (massloss) + fixed joint.

The simulation model can be found in attachment.

2. The difference appears not immediately in the beginning of simulation, howeverappare for a while.

The simulation models are not in attachment.

Attached Files (2)

_front_sus_without_switch.tpl.zip

_front_sus_with_switch.tpl.zip

0 greg.wheeler over 3 years ago

Which version are you using? Recently we introduced a feature that automatically removes switch parts when you do a rigid->flexbody swap. Switch parts can overconstrain your flexible body and ruin your dynamic results.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 zhenhui xu over 3 years ago

Thanks for your answer.

I use version 2019.02. And there is none of flex part in the model.

I dont know the reason about the difference. And sometimes we should build complex template that may be contain many switch parts. In this case could or shoud I use switch part?
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 greg.wheeler over 3 years ago

Sorry, I assumed there were flexible parts involved.

You should open a case with your local MSC support so they can investigate.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 zhenhui xu over 3 years ago

ok, thanks for your answer again.

I will create a case.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 Jesper Slattengren over 3 years ago

If the model is relatively small, the introduction of switch parts can change the results. Each part that you add to the model contributes with 15 equations, and the fixed joint required for the switch part adds another 6 equations.

The bigger and more complex the model is, the less addition of a few extra massless parts does not matter, but if it is a small model it can be some difference.

This tells me that you have not gone through the exercise to set the error and hmax to get robust results.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel