Accelerometer readings -> Static Force for FEA 2

[hygear]

I am currently working an implement for a piece of (slow moving) off road equipment and I am trying to determine the best way to find the forces on the implement mount due to the equipment traveling on a rough road. We normally use a 3g static load for checking items that are on the off road equipment (this is based on a company design standard) but we are getting peak accelerometer readings of around 5-6g on a mock-up of the implement. I know that it is unreasonable to use these accelerometer readings for a static FEA since it happens over a very short time period, but I'm not sure how to get an equivalent static load based on the accelerometer readings. I think my best option would be to use the accelerometer as a trigger to take strain gage readings and use those to determine the force.

Can anyone provide a little guidance on this?

 

[rb1957]

i don't know if i'd use s/g dat to rry and deduce an equiavalent static force on the equipment ... or maybe i doubt i'd believe the answers i got ?

sounds like you want to smooth out the real accelerations ... use a fast accelerometer, then post process with a time constant (maybe something like a running RMS) to smooth the curve. then apply these accelerations to your model.

 

[GregLocock]

Using accelerometer based data for fatigue and ultimate strength is very much by guess and by god. Put it this way, for cars we spend an enormous amount of money installing load cells and strain gauges, and using wheel force transducers, to measure in service loads, when it just takes a minute to stick an accelerometer on. There are circumstances where you can use accelerometers to generate force data, but they are rare.



Cheers

Greg Locock


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

http://eng-tips.com/market.cfm?

 

[IceStationZebra]

Based on my experiece doing almost the exact same thing, and without seeing your rig, what you are seeing is most likely "slap" due to the little clearance on the 3pt liks and pins. If you look at the data you will probably see a very rapid spike followed by a noisy sine wave. What I have done is use heave chain and binders to take as much slop out of the system as possible.

I did a project on a 6 bottom roll-over plow. I don't remember the exact loading anymore, but I seriously doubt it was over 3g. I tested it but putting it behind a QuadTrak.

ISZ

 

[GregLocock]

I'm a bit wary of the 3g figure... the other way. The front suspension of a passenger car sees loads of the order of BIGNUM g in events that it has to survive. BIGNUM >3

Cheers

Greg Locock


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

http://eng-tips.com/market.cfm?

 

[racookpe1978]

But aren't those extremely high front axle g force loads coming from the speed of the impacts for a car hitting potholes and junk at 60-80 mph?

If his loads are from off-road but slow speed impacts like a farm tractor or caterpillar tractor/grader/skip loader/non-road-qualified dump truck, then the speed is 4-5 mph and the impacts less jarring, right?

 

[hygear]

Ok, after discussing this a little further with my test engineer we re-ran the accelerometer tests with the data acquisition system in RMS mode rather than peak analog mode. Going this route the max acceleration was about 3.5g which isn't too far off from our company test standard.

 

[Twoballcane]

“Going this route the max acceleration was about 3.5g which isn't too far off from our company test standard.)


You have to be careful with RMS. What you found is maximum “RMS” acceleration load of 3.5g. Not maximum loads at 3 sigma. RMS is root mean square, in short mean or averaged loads. Thus the worst static load cases will be three times (3 sigma) RMS acceleration, which comes to 10.5g. In DoD, typical highway g loads are about 5g. So the 10.5g seems reasonable for slow off road equipment.


Tobalcane
"If you avoid failure, you also avoid success."
“Luck is where preparation meets opportunity”
"People get promoted when they provide value and when they build great relationships"

 

[zekeman]

"If his loads are from off-road but slow speed impacts like a farm tractor or caterpillar tractor/grader/skip loader/non-road-qualified dump truck, then the speed is 4-5 mph and the impacts less jarring, right? "

I agree and since the problem is random vibrations of multi-impacts, why not make a model of the worst case terrain and analytically model the response of the equipment.
This would yield fatigue life , which is the implicit requirement.

Another slightly more complicated method would be to get the analytical response (from the above model) to the input vibration spectra generated from a model of the terrain and the vehicle speed.

 

[GregLocock]

"why not make a model of the worst case terrain and analytically model the response of the equipment."
That is the gold standard for predictive CAE. For a passenger car we can do that up to about 5 Hz.



Cheers

Greg Locock


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

http://eng-tips.com/market.cfm?