Road Load data without Wheel Force Transducers

[dhoadley]

I am trying to figure out whether there is a way of collecting RLD on a vehicle WITHOUT wheel force transducers (way too expensive) that I can then put into an ADAMs model to extract component load time histories. I am guessing accelerometers are not much use as there is no way of being sure of the mass the acceleration is applied to...

I'm sure people must have done this before WFTs, but I am struggling to think of a way that captures the loads in all directions.... Any thoughts would be appreciated.

 

[GregLocock]

That rather depends on your suspension configuration. If it is a sensible suspension then your arms are loaded axially, so you can just strain gauge them to measure tension.

It is an order of magnitude more difficult than using WFTs, and very hard to be confident of your results.

Probably worth asking what you are actually tring to achieve?

Accelerometers are pretty neat, as a cheap correlation channel, but they are not very sensitive to modelling errors in my experience.




Cheers

Greg Locock

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

 

[dhoadley]

I want to get data that I can carry out fatigue analysis in NASTRAN/MSC Fatigue on suspension components (e.g arm, which I could do as you suggest) but also the BIW. The suspension is mcpherson struct at front and leafspring/axle at rear....

Sounds like you are confirming the conclusion I was coming to, which is there a reason why people use wheel force transducers, and they are so expensive...

 

[GregLocock]

Nothing wrong with measure the axial strains in the arms and feeding that into the BIW as a force.

You could take an enormous brave pill (about the same size as a pineapple), measure the triaxial acceleration each side of the body side mount, multiply the double integrated displacement difference by the (frequency dependent) dynamic rate of the bush and hey presto, a force.

I wouldn't trust it.





Cheers

Greg Locock

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

 

[dhoadley]

Measuring strains in arms gets me loads in the arm itself, but the angle at which those loads are applied to the BIW changes, which means I would somehow have to tie the two together, to turn the vectors of the loads into those seen by the body.

PS I like the double integrating idea; like so many methods in this area it sounds great in theory, but ......

 

[GregLocock]

As you say, but.... It won't work because the dynamic rate of the bushes is an N dimensional property.

Yes, we measure the suspension configuration by measuring the spindle location vertically, with stringpots, and the steer angle, with stringpots. That sets the arm geometry and you can then apply your forces correctly.

This is at once rather crude and perfectly sensible. It's not worth bleeding over road loads - run to run variation is of the order of 20%.

Cheers

Greg Locock

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

 

[dhoadley]

So then the next question is why bother try to do a fatigue analysis based on loads that may be 20% out....

Maybe I'll stick to 3g bump and hope for the best...although the more complicated analysis are always more fun!

Thanks for your help

 

[GregLocock]

Do lots of runs and build up your statistical confidence of the loads. Then get fatigue life analysis of your welded joints. Then multiply the uncertainities in each together. /Then/ wonder why you bother!

I do agree, if I were designing a car for my own company I'd design to a series of static loads. 3g in Z is a bit light on in my experience for a road car. It will work for a race car.

Cheers

Greg Locock

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