AnsweredAssumed Answered

How to make a simple flex simulation of a go kart rigid axle?

Question asked by Adam Fink on Oct 1, 2013
Latest reply on Oct 3, 2013 by Jared Conway

Like • Show 0 Likes0
Comment • 5

Hi,

As a starter project (SW2010), I am trying to make a crude simulation of a Go Kart rear axle flex when using a combination of 3 bearing supports (see pic below).

For that matter I use the "bearing" fixtures and apply a force at the extremities of my axle

Should be quite simple, yet I keep getting the " model is unstable..." error when using "direct sparse" solver or "status code 9 : matrix singular.." when using the "FFEplus"solver

I must be missing something very obvious. Could someone shade a light on this or suggest a relevent reference document/video tutorial for axle/bearing simulationst?

Thanks in advance for your help,

No one else has this question

Outcomes

Billy Wight Oct 1, 2013 2:53 AM

Your model is underconstrained. Particularly in axial translation and rotation. Think about the real life problem - you have at least one constraint in the axial direction with a shaft clamp, circlip, or something and you have rotational constraint with either the brake applied or the torque to the drive sprocket. Add in the additional constraints and it should solve fine. Be sure to check your solution with some simple beam calculations.

Actions

Jared Conway @ Billy Wight on Oct 1, 2013 4:05 AM

Slam dunk billy. The other thing is that you're making the assumption that the loading is perfectly balanced and hat the shaft won't rotate. Because the mesh isn't perfect, numerically it won't be stable. So you will also need to constrain rotation in at least one place. There are a bunch of threads about bearings that you can learn more about this. No specific tutorials unfortunately.

Actions

Adam Fink Oct 2, 2013 8:38 AM

BiIly and Jared, thanks for the explanation.

Btw, I came across another forum thread suggesting to use "inertial relief" to cancel instabilities due to mesh imperfection. Would this make sense for me to use this option or should I definitely try to add some explicit constraints to prevent shaft rotation/translation as per Billy's suggestion.

Actions

Billy Wight @ Adam Fink on Oct 2, 2013 11:05 AM

If you were to constrain one node to stop rotation and axial translation, then the results should be the same as with inertia relief enabled. If you were to constrain the part as it is with circlips/shaft clamps, and the sprocket/rotor interface the results would likely be a bit different, but also more applicable to the actual situation. I only use inertia relifef when there is no other reasonable option to stabilize the model.

Actions

Jared Conway @ Billy Wight on Oct 3, 2013 3:15 AM

Adam, read though a few more posts on inertial relief, really for troubleshooting only. High chance of affecting the results and making them whacky. Better to restrain properly.

Actions

5 Replies

Billy Wight Oct 1, 2013 2:53 AM
Mark Correct
Correct Answer
Your model is underconstrained. Particularly in axial translation and rotation. Think about the real life problem - you have at least one constraint in the axial direction with a shaft clamp, circlip, or something and you have rotational constraint with either the brake applied or the torque to the drive sprocket. Add in the additional constraints and it should solve fine. Be sure to check your solution with some simple beam calculations.
Like • Show 1 Like1
Actions
- Jared Conway @ Billy Wight on Oct 1, 2013 4:05 AM
  Mark Correct
  Correct Answer
  Slam dunk billy. The other thing is that you're making the assumption that the loading is perfectly balanced and hat the shaft won't rotate. Because the mesh isn't perfect, numerically it won't be stable. So you will also need to constrain rotation in at least one place. There are a bunch of threads about bearings that you can learn more about this. No specific tutorials unfortunately.
  Like • Show 0 Likes0
  Actions
Adam Fink Oct 2, 2013 8:38 AM
Mark Correct
Correct Answer
BiIly and Jared, thanks for the explanation.

Btw, I came across another forum thread suggesting to use "inertial relief" to cancel instabilities due to mesh imperfection. Would this make sense for me to use this option or should I definitely try to add some explicit constraints to prevent shaft rotation/translation as per Billy's suggestion.
Like • Show 0 Likes0
Actions
- Billy Wight @ Adam Fink on Oct 2, 2013 11:05 AM
  Mark Correct
  Correct Answer
  If you were to constrain one node to stop rotation and axial translation, then the results should be the same as with inertia relief enabled. If you were to constrain the part as it is with circlips/shaft clamps, and the sprocket/rotor interface the results would likely be a bit different, but also more applicable to the actual situation. I only use inertia relifef when there is no other reasonable option to stabilize the model.
  Like • Show 1 Like1
  Actions
  - Jared Conway @ Billy Wight on Oct 3, 2013 3:15 AM
    Mark Correct
    Correct Answer
    Adam, read though a few more posts on inertial relief, really for troubleshooting only. High chance of affecting the results and making them whacky. Better to restrain properly.
    Like • Show 0 Likes0
    Actions

How to make a simple flex simulation of a go kart rigid axle?

Outcomes

Related Content

Recommended Content