Impact Attenuator Calculation (need help)

[Paul87]

Hi

Recently i did a compression on an actual model of the IA. We use Aluminium Honeycomb as the material.

There is a peak load of let say 15 KN before it start to deforms.

Now, there is a problem i have. I am trying to determine the peak deceleration which the rules states which is below 40g

F = M*G

F/M < 40G assuming G = 10

F/M = 400 , can i assume this is true ? The test i have done has a constant speed of 7 m/s, thus there is no change in velocity and the acceleration is 0 . We know that the car moves at 7m/s and weighted 300KG

Force = 10,000 N , G = 10 and M =300 kg

Thus 15,000/300 = 50 (equal to 5G) which is less than 40G. Which means that this Impact Attenuator will have a peak deceleration of 5G.

Is this calculation Okay ? I dont feel quite right about this calculation. What do you think ?

In the actual test, i actually include a 4 layer of 2 inch HC 100mmx200mm.
The compression test i need at 2m/s 4m/s has exact same peak load. and for test 6m/s the peak load are lower. Now what i assume is the characteristic of the HC is constant with velocity. And there is a force at the highest peak which is 72KN. I am trying to estimate if that peak load will be more than 40G.

So i use the above calculation 72,000 divide 300 KG = 240 = 24G which is less than 40G. Thus in actual dynamic test i can pretty much assume that my impact attenuator will not be more than 40G .

Can i safely assume that as long as I am sure that the total energy absorption is more than 7350 J ?

thanks for your help ! I wanna verify this before asking my proffesor http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[Paul87]

Hi

Recently i did a compression on an actual model of the IA. We use Aluminium Honeycomb as the material.

There is a peak load of let say 15 KN before it start to deforms.

Now, there is a problem i have. I am trying to determine the peak deceleration which the rules states which is below 40g

F = M*G

F/M &lt; 40G assuming G = 10

F/M = 400 , can i assume this is true ? The test i have done has a constant speed of 7 m/s, thus there is no change in velocity and the acceleration is 0 . We know that the car moves at 7m/s and weighted 300KG

Force = 10,000 N , G = 10 and M =300 kg

Thus 15,000/300 = 50 (equal to 5G) which is less than 40G. Which means that this Impact Attenuator will have a peak deceleration of 5G.

Is this calculation Okay ? I dont feel quite right about this calculation. What do you think ?

In the actual test, i actually include a 4 layer of 2 inch HC 100mmx200mm.
The compression test i need at 2m/s 4m/s has exact same peak load. and for test 6m/s the peak load are lower. Now what i assume is the characteristic of the HC is constant with velocity. And there is a force at the highest peak which is 72KN. I am trying to estimate if that peak load will be more than 40G.

So i use the above calculation 72,000 divide 300 KG = 240 = 24G which is less than 40G. Thus in actual dynamic test i can pretty much assume that my impact attenuator will not be more than 40G .

Can i safely assume that as long as I am sure that the total energy absorption is more than 7350 J ?

thanks for your help ! I wanna verify this before asking my proffesor http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[exFSAE]

What do you mean by deform? Plastic deformation?

Funny things happen with (a) working in the plastic range (b) huge strains (c) huge strain rates.

To be real honest.. I'd say your calculations are nowhere near sufficient. Not your fault, but this kinda material wasn't covered in any undergrad course I took.

Really the only way to do this stuff is full-scale dynamic instrumented testing. Come up with a handful of designs, build em, and test em.

[Tom W]

Thats the standard way of doing things. What it all boils down to is that your max force has to be below F(max)=m*a(max)=300*40*9.81 = 117.72kN at any instant. And your average force has to be below 58.86kN to satisfy the 20g rule.

Out of interest what type of machine are you using that is capable of giving you constant feed rates as high as 7m/s? Is it one of the fancy new Instrons?

[Paul87]

Tom :

Our university has a machine called Dartec capable with doing speed up to ...not quite sure but we are able to do it at 7m/s. Thanks ! you ease my worry about the calculation. Currently we are proceeding to the dynamic test but it takes hell lots of paperwork and time.

Exfsae :

We have tested a complete model that we plan to use on the real car. What i meant the deform means the the highest point of the force (in the force over distance graph) before the aluminium honeycomb starts to change it's shape. (well yea....deformation).

I tested the model under constant speed 7 m/s. I saw a peak load of 72KN in my force distance graph. So was just trying to figure out. http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[Tom W]

You shouldn't have any great dramas with that. Particularly as most people seem to say that honeycomb gets stiffer with higher strain rates. Somewhere on the forums there is a link to a fairly comprehensive Hexweb document if you get really keen, it has relationships for different density honeycombs stiffness with impact velocity, at our speeds they claim a linear increase in stiffness. If its too hard to find my e-mail address is in my profile.

If you feel worried about excessive g loadings being transmitted you could try pre-crushing the honeycomb panels past their yield point, this is a method sometimes used to remove the high initial peak in honeycombs. Obviously if you do go down this route you get to the densification region quicker...

[Hub]

some info page 18(pdf):
http://www.hexcel.com/NR/rdonl...sorptionBrochure.pdf (http://www.hexcel.com/NR/rdonlyres/96FE250C-7BB1-4295-82C4-461A31CC97A0/0/HexWebHoneycombEnergyAbsorptionBrochure.pdf)

[Adambomb]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Tom W:
If you feel worried about excessive g loadings being transmitted you could try pre-crushing the honeycomb panels past their yield point, this is a method sometimes used to remove the high initial peak in honeycombs. Obviously if you do go down this route you get to the densification region quicker... </div></BLOCKQUOTE>

+1

The easy and fun way to do this is with a ball-peen hammer. Got rid of our initial peak. We just did a slow crush, and F vs. D was impressively flat.

If you're doing the whole test at 7 m/s it should look excessively stiff if anything, as it maintains a high strain rate throughout rather than slowing down as it goes.

BTW, what did the test look/sound like? I can't help but imagine the sound of 400 beer cans being simultaneously stomped by Optimus Prime. http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[Paul87]

Yup.... because the control are done all using the software in the computer. All i do is click "test". And the sample is flat.

U mention using a ball-peen hammer....lol. I was thinking about how you hit every corner of the flat HC...haha.

I did a test on a 50mmx50mm 2inch HC sample. At 2m/s and 4m/s the high load before it begin crushing is 25KN. But at 6m/s it is 20KN . ????? i assume it is weakest during high speed and strong at low speed.

[Adambomb]

Hmm, that is odd. From what I've read material strain rate effects shouldn't be that large with Al at the speeds we're going, but if anything there would be a bit of strain rate hardening. In addition to that, the other source I would expect to provide higher force is in inertial effects as each little element of each wall rotates as it bends over. As for why it would get softer at higher strain rates, really don't know.

I've run into problems with test data (specifically shock dyno data) where the machine wasn't able to keep up its specified rate, ie test was supposed to be 15 in/s but looking at dx/dt in the raw data it varied between like 7-15 in/s; was clear the mechanical portion of the machine couldn't keep up. Might be worth looking into.

Also, FYI, from what I've heard they're not too keen on scaled tests...

[Paul87]

Hi,

I face another problem. I personally would like to do the dynamic test.

And i ask a few technician and professor about some advice, they said it would be expensive which i know but something i cant understand is they said the accelerometer will be damaged in the process.

I search around google and found many similar method were used such as
-using tapes and stick the accelerometer on the 300KG mass and attach the IA on the bottom of the 300kg mass and release it using a catch.

So i cant think any ways that the accelerometer could be damage which i suspect the accelerometer cant accept high vibration.

Any solutions or specific accelerometer that can be used ? Or perhaps attachment of the accelerometer could be modified abit.

would be grateful if I get more info and ideas as my group project mates are pretty much useless and a kind of ask and do type.

thanks !

[Adambomb]

Most vibration accelerometers should be able to handle it. If you've got something that will read 50g then it should generally be good damage-wise to like 100 or maybe even 200g. At any rate the spec sheet should say. We got some from Omega a few years ago that met roughly those specs, pretty sure they were around $150-$200. Biggest concern I imagine would be something going catty-wampus and crushing it.

[Paul87]

haha....thanks for the reply.

I was just wondering with the formula student rules. It mention not exceed average 20g and a peak deceleration of 40g.

Subjected to the condition that the 300kg do not bounce and the IA absorb all the energy.

Does that means if i had an average deceleration of 10 g and the peak less than 40g, the IA would pass ? It never mention the minimum so kinda wanna recomfirm it. As i saw many people strive to reach an average of 20g.

Cheers !
Paul

[Mikey Antonakakis]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Paul87:
haha....thanks for the reply.

I was just wondering with the formula student rules. It mention not exceed average 20g and a peak deceleration of 40g.

Subjected to the condition that the 300kg do not bounce and the IA absorb all the energy.

Does that means if i had an average deceleration of 10 g and the peak less than 40g, the IA would pass ? It never mention the minimum so kinda wanna recomfirm it. As i saw many people strive to reach an average of 20g.

Cheers !
Paul </div></BLOCKQUOTE>
There is a minimum amount of energy in the rules that the IA has to absorb. 300kg at 7m/s is 7.35kJ. If you drop a 300kg mass in a full dynamic test, and it comes to rest, then you've absorbed that energy. If you're doing a quasi-static test, then you can calculate how much energy you've absorbed (area under force-displacement, for instance).

[Adambomb]

The main reason people want to get an average of 20g is to minimize the size/weight of the IA. You could make one like 30 inches long that never gets over 5g, and it would be legal...but it would also be 30 inches long.

[Paul87]

thanks for the reply.

I did a slow compression test. using the data i get and subtitute to the eqn of motion. fill in the data , intergrate it

My IA has been compress till it pass it's yielding point to remove high peak force thus my deceleration are pretty low. So i was curious if it is actually ok ...haha. My IA is at 250 mm high.