Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations GregLocock on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

LMS Simulation Vehicle Drift

Status
Not open for further replies.

morris9791

Mechanical
Feb 7, 2008
99
Dear Experts,

I am using LMS Virtual.Lab to simulate straight line motion of an 8*8 vehicle. However, this model is drifting significantly and I cannot find the culprit.

I have checked the following

• steering system (wheel yaw ~ 0.15 etc)
• weight distribution (rear bogie heavier loaded than front bogie due to position of CoG).
• The road surfaces is smooth and flat,
• same TSDA elements for each wheel station are consistent
• Tyre properties are very basic and only a ‘simple’ tyre element is required to define the properties.
• The driving torque to all wheels is the same.

What other factors can affect drift in a dynamic model?
Does the model have to be absolutely symmetric in every sense of the word to achieve straight line manoeuvres?
I believe this would be difficult because of the customer specified CoG for the vehicle etc

Thanks
Eddie


Thanks
Eddie
 
Replies continue below

Recommended for you

How much drift at what speed are you talking about? Does the latacc make sense if you look at the tire forces?

Without knowing any details, yes obviously there are many things in your model that can prevent it from tracking straight just like in real life.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Greg,

See attached plots of speed response, lateral acceleration and tyre forces. The manoeuvre is an 80kph straight run.

No, the lateral acceleration does not make any sense. The magnitude starts to increase as soon as the vehicle settling phase of 10 seconds is complete. (Vehicle is dropped from a height and allowed to settle) I cannot find the source of this behaviour.

Also the tyre lateral and normal forces don’t make sense either; both begin to increase when the settling phase is complete.

The tyre forces should remain relatively constant and obviously the latac should at least be very close to 0.

Your comments / opinions will be appreciated.
Thanks
 
 http://files.engineering.com/getfile.aspx?folder=b0f59321-ff6f-41ff-8e29-6999f05fc5cd&file=Animation_Trace.doc
In addition to my above message,
I did the following in an attempt to infer more information about the behaviour

• I ran a closed loop straight run 80kph forcing 0 lateral acceleration on the vehicle. The steering system ended up in full backlock as soon as settling phase complete and began traversing a circle and roll became excessive and overturned.

• In the tyre model, I turned off the lateral forces on the tyre ie, a plot of tyre lat forces read 0, in doing so the vehicle started ‘sliding’ across the road with very little forward movement.

I’m at a lost trying to understand how a vehicle would even move sideways when I specifically turned off tyre lateral forces…..
 
what happens if you lock the steering at 0 degrees?

if feedback control fails it may be that you have the wrong sign for the steering ratio.

if the latacc makes no sense when you look at the tire forces then you probably have an architecture problem.

Incidentally when debugging a model I might have 200 graphs, it is basically detective work.

If you just perform a static equilibrium does the resulting state look realistic, or has a joint flipped during the settling phase?


Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
It works fine under its natural equilibrium position. I believe the problem is the control system which I built manually by controlling the pitman arm of axle 1.

The pitman on axle 2 is linked via a ‘dot1’ constraint. I got it working perfect before where I could get it to go in a straight line by subjecting it to 0 acceleration and also doing 100m turning circles.

Now it is not working since I modified the steering geometry. I think my manual set up of the control system may not be stable enough and therefore I may use the advanced modelling elements available in LMS such as vehicle subsystems etc etc.

200 graphs…serious? Must be a complicated model!!! :-0
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor