AnsweredAssumed Answered

Cosmos motion capabilities

Question asked by John Sutherland on May 23, 2009
Latest reply on May 25, 2009 by Ian Hogg

Like • Show 0 Likes0
Comment • 3

SW corp marketing implies that a mere slip of a youth can analyse the suspension of an open wheeler racing car; but my reading of CM text User Guide (2001), Step by Step Guide (2008), hypertext Reference (Help) Guide (2009), Tutorial (2009) and Motion training syllabus has found no example of anything more complex than a 2D planar mechanism.

So I have jumped in at the deep end with a 3D mechanism and I am attempting a process of elimination by asking one question at a time.

Can ADAMS solve the 3D mechanism in the screenshot below?

Attachments

3Dmechanism.png128.0 KB

No one else has this question

Outcomes

3 Replies

Ian Hogg May 23, 2009 1:36 PM
Mark Correct
Correct Answer
Hi John,
I no longer work for MSC/MDI, but I can assure you that this is the main use of ADAMS in every auto company worldwide. Admittedly there are special modules that help automatically build the underlying models based on company best practices, but the underlying solver engine in Motion is the same.

As far as modeling these types of systems, the main challenge in 2008/2009 relates to determining the appropriate SW mate to get the mechanical joint. I have not played with the universal joint added in 2009 much, but that should help in defining a nonredundant kinematic system. This is the classic joint used in ADAMS/Car for eliminating redundancies.

When you are more interested in a more dynamic model, bushings are the way to go to add compliance to the system. I have not pushed the solidworks joint bushing representation to it's limit to see how well this addresses the needs. I would love to just have a bushing entity that I can define manually for this purpose.

If your bodies in the image are mounted with 3 ball joints, then there are redundancies (bogie has 6 dof and each ball joint tried to remove 3 for a total of 9) so either bushings are going to be needed or one of the ball joitns needs to be adjusted.

It's a little unclear to me still on the full setup of the model. For example, are the bogies free to translate/rotate with respect to ground in all directions? Do they stay parallel? etc. Are there shocks in the system to limit/dampen travel?

The first step I find is locking the engine in space and articulating the linkage system to ensure it is working as expected. Once that is confirmed, then enhance the model to provide ground inputs to the system and observe the response of the vehicle.

Cheers,

Ian
Like • Show 0 Likes0
Actions
- John Sutherland @ Ian Hogg on May 25, 2009 2:54 AM
  Mark Correct
  Correct Answer
  Thanks Ian for your reply.
  
  My present interest is in the steering geometry only. I will worry later about springs and dampers.
  
  All train components are constrained by ball joints. Relative rotation is allowed, but relative translation (e.g. a spacewalker losing her spanner) is not allowed.
  
  In the SW model, the linkages work as expected without locking/fixing the engine, but not in motion analysis.
  
  CM User Guide says all mechanisms have a rigid ground body. This sounds reasonable but is incompatible with a train mechanism rolling/floating on a fixed track.
  
  Perhaps the auto makers fix their car bodies and apply force functions to the wheels?
  
  I am wanting to analyse motion similar to a skateboard rolling in a channel. I graduated from childhood before skateboards were invented, but I suspect that they are not rigid bodies.
  Like • Show 0 Likes0
  Actions
  - Ian Hogg @ John Sutherland on May 25, 2009 12:10 PM
    Mark Correct
    Correct Answer
    Hi John,
    
    If you are doing something like aircraft/projectile studies, then obviously no ground is needed. But for most other systems it is somewhere supported by ground.
    
    In your case, the ground interaction with the bogies is the support path and then the body is suspended from those. This is where you need to think about how those bogies can move with respect to ground. With car tires, it's tricky to realistically model them due to the tire stiffness characteristics and the way the contact patch can shift around. If you are dealing with a traing wheel (or a skateboard wheel), it should be a safe bet to assume that remain parallel to the ground (but can roll on the ground plane). The one important thing to allow here is that the wheels can independantly slide aling their axis and also rotate about the axis normal to ground since depending on the rotation of the body, they will adjust their position.
    
    So I'd constrain the bogies to ground using a point/plane coincident mate at the center of the bogie (axles), and then add a parallel mate between an axle axis of the bogie (ie normal to your magengta planes) and the ground. That should minimally constrain them. You will probably need to add a coincident point plane mate between the bogie and ground in the 2 virtical planes) to give you a relative movement. Otherwise the parts can move anywhere on the ground.
    
    If there are particular rails keeping the bogies aligned (and parallel), then the restraints to ground will differ to reflect that.
    
    Sorry if the description sounds a little confusing, this is one of those things where it's easier to draw it or set up the model.
    
    Cheers,
    
    Ian
    Like • Show 0 Likes0
    Actions

Cosmos motion capabilities

Attachments

Outcomes

Related Content

Recommended Content