Impact attenuator test

[Ebere]

Hello guys...
i was wondering if you have some information to make some physical tests to the impact attenuator... we´ve made some test to this years car´s IA in a compression machine in our university... but we have the problems to find a professor willing to help us doing this test...
I would like to know how else can we simulate the necessary energy.. because we have a project to build our own tester....
Thank you very much for any information...
I´m so sorry for my english.. i´m venezuelan hehe

[Brett Neale]

We had help from the University's Automotive Safety department for our tests, but the rig was basically a big heavy pendulum setup with a 2ft square flat board on it, with the attenuator bolted to something heavy on the bottom. Very similar to a Charpy test if you've ever done one for a fracture mechanics lab. It's not hard to calculate the height required for the energy absorption you're after.

Other than that, we just had a high-speed camera to see what was going on during the test.

[tool]

Hi,
Will a 1000fps camera suffice for the physical test? Also, does it make a difference if I mount the accelerometer on the specimen or on the hammer itself?
I am about to conduct one such test. Hence the query.

Abhishek
IIT Madras, India

[Brett Neale]

Yeh 1000fps should be fine... Ideally the accelerometer should be on the specimen, but it can go either way.

[Adambomb]

Yeah, we've thought this one out a bit now, and we keep coming back to the problem of finding something, anything, to safely absorb the energy in a manner that doesn't ruin the results. I don't think we have anything on campus; even if we did we would also have the problem of finding a professor that would get involved in such a potentially dangerous venture.

The "safest" method we came up with that would be simple, cheap, and reliable would be to use gravity as a prime mover...a drop tester. Then you just have to drop an instrumented 661 lb chunk of something from somewhere around 25 ft up. I'm not too excited about that, or what the university would do to us if we tried to do it. Of course a pendulum would be safer, but that would be a pretty big thing to build and have around, and the university would still probably lynch us.

At this point I'm thinking it would be much more worthwhile to find some sort of place that does crash testing and just pay them take care of the physical testing.

[Moke]

We tested a nose a few years back by dropping a large chunk of concrete onto it, the block had eyelets cast in it which ran on guide rails. The trigger was a bit of re-bar sticking out the top which a lab tech cut with bolt cutters. We did the test in one of the civil engineering labs where they like to break things. See if such a lab exists in your uni.

[McMasteRacer]

Originally posted by Adambomb:
The "safest" method we came up with that would be simple, cheap, and reliable would be to use gravity as a prime mover...a drop tester. Then you just have to drop an instrumented 661 lb chunk of something from somewhere around 25 ft up. I'm not too excited about that, or what the university would do to us if we tried to do it. Of course a pendulum would be safer, but that would be a pretty big thing to build and have around, and the university would still probably lynch us.

I am doing the IA as my final year thesis project, most likely going to be a carbon nose that acts as the IA and body. Anyways im curious as to why you need to drop this from 25ft? from my calculations its ~9ft.

what kind of instrumentation are you going to use to capture the data of the test. right now i have access to a high sped camera (going to film the test mythbusters style and try and verify the impact speed) as well as a 50g max accelerometer and a force plate that is connected to a datalogger.

[cjanota]

Is the physical test limited to a drop-type test with acceleration data or do you think we could crush them and use the force/displacement curve to calculate the acceleration?

[shampoo]

I did a drop type test with force data this year. Basically we dropped a metal block at 7 m/sec onto our structure and we had a load cell placed under it connected to a DAQ. The DAQ had a sampling rate of 20000S/s. If you assume the attenuator to be massless which is pretty much true in comparison to the block, the force/mass of block will give you the deceleration curve

[Adambomb]

Originally posted by McMasteRacer:
Anyways im curious as to why you need to drop this from 25ft? from my calculations its ~9ft.


...yeah, should have thought about that one more, we came up with 25 ft. at the bar, where our basic calculus skills apparently quit on us. The correct height isn't too far to achieve; close enough that I wouldn't feel too bad just doing it in an isolated area in backwoods Iowa, under the false sense of security that our university really has no idea what we do anyway. I like the rebar idea...

[blueflame898]

Is it possible to find the peak G's from a static test? I know you can find average.

[CappyUMD]

Record compression force vs. displacement. If the strain rate is low and/or the sample rate is high enough, you will resolve the force peak.

[Adambomb]

The biggest problem I see with doing a static test is it neglects strain rate sensitivity, which could give much lower forces than would actually be seen in impact. I haven't been able to find any info about the strain rate sensitivity of the hex-cell we use, but at any rate I don't imagine it would be very easy to convince the design committee that it's equivalent, nor would I want to risk it. Looks to me like the test standards are too cut and dried for that. I suppose you could always email them and see.

Another thought, since the test specs are so standardized, I imagine a "consortium" approach could be looked at? Have one facility set up the test, with a bunch of schools pooling in on it, then just crush a bunch of IAs?