shock absorber modeling 3

[zoghi]

As you know the actual shock absorber has two coefficients of damping, one in compression and the other in tension. I want to know which one is useful to calculate in the modeling (in mathematic and simulation software)

 

[mloew]

Zoghi,

Actually, dampers are highly non-linear and at best can be approximated by three linear regions in compression and two in extension.

Force vs Velocity graphs shows the compression and rebound adjustment range of the double- and triple adjustable damper:

There is a great deal of information on the internet on modeling and simulation. Try these searches on Google.com to start:

http://www.google.com/search?sourceid=navclient&q=damper+force+velocity+curve

http://www.google.com/search?sourceid=navclient&q=shock+absorber+force+velocity+curve

Also check out:
Measurement of Equivalent Stiffness and Damping of Shock Absorbers

http://www.me.mtu.edu/~mrao/et-shock-paper.pdf

Practical Application of the Empirical Dynamics Method (

http://www.adams.com/news/events/userconf/na/2000/04_MTS_susp_dev_edm.pdf)

SAE 2000-01-1625: Accurate Models for Complex Vehicle Components using Empirical Methods (Available from

http://www.sae.org)

Best regards,

Matthew Ian Loew

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[GregLocock]

For real simulation work you have to use a full non linear model, but for typical 1 to 5 DOF type models, ie ride modelling and so on, you can take the average (of the magnitude of thejounce and rebound values) damping at the velocity of interest. Of course one problem is deciding on the velocity of interest, and another is that the velocity changes with frequency.

Cheers

Greg Locock

 

[mloew]

Greg,

You are, of course, absolutely correct! I imagine you are using a piecewise linear damper model in ADAMS. Good enough?

One of the nice things about the Empirical Dynamics Modeling technique (see first reference) is that all of the dynamic properties (including amplitude dependence and frequency dependence) of dampers are captured for use in a numerical model. Unfortunately, I have no direct experience with this method. I am quite impressed by the results that I have seen, though. Any comments? Best regards,

Matthew Ian Loew

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[GregLocock]

The Barber paper - I can't say I'm a big fan of Powerpoint papers. He doesn't know his statistics. If you repeat a test enough times you can get any accuracy you want. It all seems a bit content-free.

The other one is OK, but is really aimed at road noise frequencies, ie 50 Hz and up, not for ride or handling models.

So, no, you can't really reliably replace a non linear model by a linear one. No surprises there.

ADAMS models the shock itself as a simple force vs velocity spline, but if we ever get into ride prediction in a big way I imagine we'll have to start modelling stiction and the valves themselves.




Cheers

Greg Locock

 

[mloew]

Greg,

I agree about the presentation -type paper. The really good information is in the SAE technical paper. Regardless, I was more interested in the ability to model the component at that level of fidelity rather than the statistical relevance of the inputs. Best regards,

Matthew Ian Loew

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[ajbarber]

Matthew & other suspension modelers:

If you're interested, you can learn more about neural network models for shock absorbers and rubber bushings via the following link:

A.J. Barber, Accurate Models of Dampers and Bushings using the Empirical Dynamics Method, Proc. 1999 European ADAMS User Conference, Berlin. Mechanical Dynamics Gmbh, 1999.
(

http://www.adams.com/news/events/us...ccurate Models for Bushings and Dampers...pdf

)

Also, I'd like to dissuade prospective readers from referencing the ADAMS conference paper you've listed ('Practical Application of the Empirical Dynamics Method '), as it was incomplete at the time of publication. Instead, please see this one:

A.J. Barber, G.Sandlass, Practical Considerations for Generating Neural Network Models, 2001 North American ADAMS Users Conference, Troy, MI, May, 2001.
(

http://support.adams.com/kb-files/id_9313/MTS_2001_NAUC.pdf

)

(Unfortunately, this is still a powerpoint. I'm planning to publish a full text article covering similar material in the near future).

For additional info re: advanced models for suspension components and other systems, you may find some of the bibliographic citations useful. Also, you may wish to peruse the following list:

http://www.neosimulation.com/publications.htm

Best regards,
A.J. Barber
Neosimulation Engineering

http://www.neosimulation.com