How is everyone's impact attenuator testing coming along?

[Johnny B]

We just did 2 drop tests today of 770 lbf from 7 ft and def. didn't do to hot. Our first one gave a peak deceleration of 6gs and we thought that was great. The interior was pink foam, 12 cans, and some expanding foam.

Then the 2nd one gave us a deceleration of 80gs!!! But the only difference is the cans are replaced with pieces of aluminum standing up. These hardly crushed, but would that be a realistic difference?

I'm a bit surprised at the 6g rating myself so was curious how everyone else is doing on theirs.

[Johnny B]

We just did 2 drop tests today of 770 lbf from 7 ft and def. didn't do to hot. Our first one gave a peak deceleration of 6gs and we thought that was great. The interior was pink foam, 12 cans, and some expanding foam.

Then the 2nd one gave us a deceleration of 80gs!!! But the only difference is the cans are replaced with pieces of aluminum standing up. These hardly crushed, but would that be a realistic difference?

I'm a bit surprised at the 6g rating myself so was curious how everyone else is doing on theirs.

[Jersey Tom]

We'll see. Doing it Wednsday morning I believe.

[CMURacing - Prometheus]

http://www.andrew.cmu.edu/~msmiles/decel.jpg

[drivetrainUW-Platt]

the honeycomb is in the mail.....

[Jersey Tom]

Dude, just machine out some honeycomb from a solid block.

[CMURacing - Prometheus]

Johnny-

I'd be suspicious of the 6G number. Did your weight bounce? Is the material crushed as far as it will go? I'd suspect you're only dissipating a small portion of the impact, and the excess kinetic energy is going into bouncing the weight back up in the air.

[HenningO]

Anyone using the method of using the force/displacement data from a slow crushing and then calculating the acceleration from the data?

[Johnny B]

As of about 10 minutes ago we are! Unfort. we don't have the funds for any honeycomb structures so we're gonna have to get a lil creative on this one for some new design.

[Jersey Tom]

Try beer cans, Johnny?

[Johnny B]

Cans do not provide nearly enough resistance. And to get that much energy dissipation would take way more than we could fit on the front of the car. We are thinking about Red Bull cans though... got waaaaaay to many of those around the shop now.

[drivetrainUW-Platt]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Johnny B:
Cans do not provide nearly enough resistance. And to get that much energy dissipation would take way more than we could fit on the front of the car. We are thinking about Red Bull cans though... got waaaaaay to many of those around the shop now. </div></BLOCKQUOTE>

thats our backup plan, persnoally I think redbull tastes like crap, wasnt there a guy handin em out at comp?

[CMURacing - Prometheus]

what about full cans? the liquid's gotta be good for something...

AND, now that i think about it, carbonated beverages are canned under pressure...you could get a little reactive armor effect going on there.

[Johnny B]

Well it's 1:30am and we're running on Red Bull at the shop right now. I'm attempting to make a "honeycomb" mesh from sheets we have. Ghetto fabolous is probably the best description. If this doesn't work, then I'm just gonna use Chuck Norris as our impact attenuator.

[kwancho]

He'll glare at the offending object (cone, curb, brick wall) and it'll crush itself out of fear.

[James Waltman]

Chuck Norris you say? Facts about Chuck Norris. (http://dot.etec.wwu.edu/fsae/HostedPics/Junk/Facts_about_Chuck_Norris.pdf)
Sorry, couldn't resist.

Johnny B,
Do you have any pictures or video of your setup? Anyone else?
Search the forum and you might find some data on crushing cans...

[Johnny B]

Well your post was actually what informed me of how crappy cans & pink foam were. Our data just verified it. The video is frickin sweet, I'll post it within the next couple days. Just got back from the shop and it's 4am. I'm praying our homemade honeycomb works tommorrow w/ the Instron. Here's my ideas behind what the Instron data actually means...

Instron provides Force as a function of displacement F(x)
Using F/m=a we can find the instantaneous accelerations. Now the Instron will displace as far as we want it to (let's say 6"), but with an impact, it might only go 5". But how do we know that? I believe the answer is 'under the curve'

W=F*x, so finding the area under the Instron data curve provides the energy absorbed.
Find the x value that makes W = 1/2mv^2 (Kinetic Energy).

Is there a loophole in my thinking? I'm not positive yet on how to do average acceleration with a large amount of data. But I believe I'd just use the AVERAGE function in Excel on our data. Thanks for any help. I'll have Weed post the pics asap and the results.

[magicweed]

Basic Construction pics and test aparatus

Will be updated after lab

http://community.webshots.com/album/547868314ilysBj

edit: http://media.putfile.com/Quicktime-of-our-impact-attenuator-dynamic-test

[AK UT]

This reply may be pointless, but I had to get it out of me.

Jack Bauer could kick Chuck Norris' ass any day of the week.

[HenningO]

Yeah, Johnny B, that's how I've done it...

[Belo]

Johnny B

I'm doing it the same way too... This will give an approximation of the energy beeing absorbed in the structure. I doubt this will really show you what an impact would do to our structure tought...

It is possible to establish a impact factor to really have an idea of what would really happen during an impact, but that factor is only good if the structure does not go plastic... In our case, plastic deformation comes really fast and the factor would then be unuseful...

We have done a great effort on testing different combinations, I hope that judging on that will not be too severe since the equipment require to do a good crash test is not accessible to all of us...

It is also wrong to assume that the front bulkhead would not absord any energy during a 7m/s impact! Honeycomb seems to be the best way to do this

PROMETHEUS:

have you validated that you have absorbed enough energy to stop the car??? less than 6 inches deformation seems small... What was the thickness of the honeycomb? I've read that you could crush honeycomb to about 75% of its nominal thickness...

Im planning on using Alcore PAA 5052 5.7 3/16 cells...

thanks

[HenningO]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Belo:
It is also wrong to assume that the front bulkhead would not absord any energy during a 7m/s impact! </div></BLOCKQUOTE>

Still, with the new rule about node-to-node triangulation, I'd say that since the impact attenuator is not as near as stiff as the frame/or monocoque, that assumption would be fairly OK. I'm planning on (if I've got the time, and you all know the answer to that) to build the front structure of our frame in order to do test the impact structure with attached to the frame, or at least a part of the frame.

[magicweed]

Personally, I just enjoy dropping 700 lbs worth of water on metal objects. But thats just me. In our video, it seems to do its job. Good enough for me.

[Johnny B]

I got to do the test, analyzing the data right now. Can't quite figure out what 100 kN N means though for the heading of the force file. Can't tell if it's in kN, N, blah blah... oh well... more work

[CMURacing - Prometheus]

Now that you mention it, I've been thinking about throwing an energy calculation in my script too...the honeycomb sections are 8" thick (7.9). I'm at about 70-72% crush depth.

Its a pretty simple MATLAB script as it is, as I mentioned on the other post, it crushes very thin cross-sections one at a time, by finding the maximum force that area can sustain, converting that to an acceleration, and assuming that the accelerations are linear over each section. then it stops when v=0.

[GTmule]

We're doing it using "digital calculus" as one of my profs puts it. Put the material in the MTS machine, get a F to D curve, for each dD I'm finding the energy loss, calculating V, dV/dt is acceleration......average(c5:c1562)&lt;20g...hopefull y....... http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Bulldog]

Hey -

How does one use Alcore's compressive strength & crush strength to estimate energy absorbtion? i.e. can we assume a constant crush strength while the metal matrix is collapsing?

[Jersey Tom]

Dropped 660lbs of steel on ours today. Worked great.

[damon lemmon]

here is some last minute lo-tech shit: 1 digi-caliper, 1 scale, 1 arbor press, 1 hunk of metal, 2 zip ties. calibrate scale with hunk of metal, zip tie caliper to press, pull level, record data, make graph. http://img129.imageshack.us/img129/5908/crush0yl.jpg

[HyeZeus]

We used 4 pcf 2 part polyurethane foam. Its gonna be a bit heavier but its really cheap compared to AL/Nomex honeycomb. Ill probably experiment with some non certified honeycomb next semester.

[Jersey Tom]

Hint: aluminum honeycomb doesn't do jack.

[HyeZeus]

I pretty much had the Instron Compression machine plot force vs. deflection. The integral of the plot needs match the kinetic energy of the car before the compression machine hits (massx20g) force.