I need all the info about the crush zone!!!

[Chente(mcbo)]

Hi all you guys, i want to know about the crush zone!!! its very dificult find some info about it, if someone can tell me some info about it i will be so greatfull!! thanx!! I´m from Venezuela and this is our first time here!!!with this goal like all of you have it!! Help us!

[Chente(mcbo)]

Hi all you guys, i want to know about the crush zone!!! its very dificult find some info about it, if someone can tell me some info about it i will be so greatfull!! thanx!! I´m from Venezuela and this is our first time here!!!with this goal like all of you have it!! Help us!

[Arvin]

Follow the rules for the dimensions and make it as light as possible. Some type of foam typically works pretty well. If you want to do some base energy calculations drop a really heavy block on a mock up from a known height and see how much it deforms.

[Mi_Ko]

We've used many different crush zones. Starting from simple thin wall aluminium tubes, cutted aluminium plates, honeycomb material and a special foam.
I also remember teams using 5~6 beer cans supertaped together and bolted on the bulkhead.

[jack]

...a note on beer cans, the VRI has actually done extensive testing on beer can crush zones (i kid you not). if you go this route, make sure you use beer cans, not pop cans. i cant remember why, something to do with diffrent alloys i think.

[Chente(mcbo)]

Thanx all of you,the past saturday me an another partner of team went to visit a pilot that rece in formula ford! it was a really good expierience,he sshow to us every single part of his car made in italy!we are from venezuela, his crush zone was like Mi_ko said, with a thin wall but like a box! i will put in here some pics of that car, i will like that some of you post here some pics!! thanx again.
the idea of the beer can is very good!!

Chente
Universidad del Zulia. Venezuela(land of beautyfull ladies)
www.fsae-luz.org (http://www.fsae-luz.org)

[Greg H]

We used foam covered with carbon and did a little drop testing.

http://www.kengrimes.com/crashtest.jpg

Then, we just painted it and used it as the nose. It's worked for 4 years now, but next year we might try to get something lighter.

http://www.kengrimes.com/frame.jpg

[Chase]

it is good to note that most schools or atleast our school has a machine that will more accuarately test material(crush zone) strength. than a mass dropped from a certain height, but if you want to do it the phsyics 1 class experiment way go for it. usually the materials departement or textile/fiber engineering department will have them.

[James Waltman]

So I figured that I could spend a few minutes sharing some of the stuff we have done with crush structures and it turned into a long post. First my disclaimer – I am not an expert by any stretch. It is possible that I will make a mistake in some of my explaination – please correct me if you see one. Also – my apologies to our friends in the more civilized metric speaking countries.

Drop testing is great if your setup is instrumented. If it isn't instrumented you're missing ride down information – this is critical. The known weight/known height/deformation method will tell you about the energy that was absorbed but it can't tell you about the rate.

So an example:
Without instrumentation
You drop your known weight from 10 feet (is that about what you used Greg?).
You deform your crush zone 6 inches
You know how much energy was absorbed but you can't tell the ride down.
So the crush zone may have held 200G for the first 0.250"� of deflection then collapsed and held only 2G for the remaining 5.75"� of deflection. The total energy absorbed doesn't show that your driver died when his brain rattled around in his head at 200G.

So you need to know the ride down. We have been using 60G as our survivable limit. This is a rough estimate because many things influence how much the human body can take – how rapidly the forces are applied, in what location, and for what duration. Crash test dummies in Federal 30mph tests routinely register 40 to 50G's

We have been setting up our ride down for 30G. So our crush structure needs to hold a constant 18,000 pounds during the ride down (estimate 600 pound car and driver times 30G). So now how much area is available determines how many psi a crush material needs to ride down at. To keep it simple we can use a square 6"� by 6"� (not allowed by the rules) this gives 36 square inches of area for the crush structure. So 18,000 pounds over 36 sq inches means we need a material that rides down at 500psi. We have an instumented 75 ton hydraulic press that we have been doing our testing on.

So to examine some of the options I have seen teams using.

To follow Wolverine's suggestion: "Follow the rules for the dimensions and make it as light as possible. Some type of foam typically works pretty well."�

I have seen some teams using home insulation sheets. http://www.owenscorning.com/around/insulation/products/insulpink.asp

These are ridiculous. The graphs below show the relationship between the displacement and pressure.

http://dot.etec.wwu.edu/fsae/James/Crush%20Structure%20Stuff/Pink%20Insulation%20Foam%20graphs.JPG
The top graph shows that the ride down is not even close to linear. The bottom graph shows a detail view of the first 0.250"� of deflection. It takes a whopping 12psi to deform the foam at that point (remember that we are after 500psi). So when the foam gets crushed to about 0.750"� then it starts to pick it up a bit – the problem is that this sample was only 1"� thick. So somewhere just short of 1"� of deflection the graph jumps to about 1400 psi – or about 84G = dead driver. So no ride down from the pink foam but it absorbs substantial energy near the end.

Jack suggested beer cans. There was a time when the VRI messed around with beer can crush structures. I think that this was with Viking VI (late 1970's or early 1980's?). I think that beer cans were simply thicker then. The thing is: a full size car benefits from is more area for the crush zone. So to achieve the same G figure you can use a crush structure with a much lower ride down. I have seen some FSAE teams using aluminum cans so we tested those as well. We used pop cans but I don't think that there is a real difference between them and beer cans.

http://dot.etec.wwu.edu/fsae/James/Crush%20Structure%20Stuff/6%20cans%20graphs%20with%20detail.JPG
The graph shows that the cans were not really providing any ride down up until about 4.5"� of displacement then by the time they were displaced 4.85 inches the load is enough to kill the driver. The cans are about 5 inches tall. Again the goal was 500psi. The detail graph shows a spike at the start. This spike represents the load the cans can hold before the walls buckle. A massive 12psi. The first 3 inches of displacement takes less than 5psi (we are after 500psi). This situation is similar to the example I described above about 200G then 2G but on a different scale. The cans take about 0.02G to buckle and then 0.004G ride down. This doesn't seem like it is going to help very much.

So now some real candidates for crush structures. Aluminum honeycomb is fantastic if you pick the correct compressive strength. Basically, the strength varies with cell size and wall thickness. The best thing about aluminum honeycomb is the dead even ride down rate.

http://dot.etec.wwu.edu/fsae/James/Crush%20Structure%20Stuff/Aluminum%20Honeycomb%20500psi.JPG
The sample above has a ride down of right on 500psi from about 0.250"� of displacement until about 3.5"�. It climbs higher than 500 psi at the end because there is no space left to crush (it was a 4"� tall sample). If you want more than 3 inches of ride down you can stack sheets or blocks. Some problems arise from stacking. The most problematic is knifing – the walls of one layer cut through the walls of another layer. These problems can be solved.
One area of concern in the aluminum graph is the spike at the start. For a short time it held about 700psi. Since this was above the desired ride down rate it is a problem. It resembles the peak at the start of the graph for the cans because the same basic thing is happening. The walls of the honeycomb are able to hold more load before they begin to buckle and crumple. The simple solution to this problem is to pre-crush the aluminum before it goes in the car. If it were pre-crushed about 0.250"� then the ride down is near perfect for the entire range. The aluminum honeycomb will weigh more than very lightweight foam but it will also provide some ride down.
Most of our testing has been done on the press as I mentioned. I am pretty sure that the aluminum honeycomb ride down is not rate sensitive. It will crush in the same manner even if it is done at a high rate. One area of concern for honeycomb is off axis impact. It simply doesn't behave in the same manner. I don't think that the off axis loads can be as high but they could be significant. Maybe someone else has looked into this and is willing to share some information.

Aluminum is great but it can get even better. We have developed carbon fiber honeycomb for use as an energy absorber. The strength can be tailored by cell size and wall thickness (number of plies) like aluminum. It can also be adjusted by fiber orientation.

How awesome is this
http://dot.etec.wwu.edu/fsae/James/Crush%20Structure%20Stuff/V32%203-4%20inch%20Carbon%20Fiber%20Honeycomb%20500psi.JPG
One of our current setups rides down very evenly just above 500psi. The spike at the start is also present here but can also be eliminated by pre-crushing. The carbon fiber offers a similar amount of displacement for a given height. So the carbon fiber offers similar ride down characteristics but the real key is that its density is substantially less than comparable aluminum honeycomb.
http://dot.etec.wwu.edu/fsae/James/Crush%20Structure%20Stuff/DSC04760.JPG
http://dot.etec.wwu.edu/fsae/James/Crush%20Structure%20Stuff/DSC04221.JPG
It seems that aluminum skins are a popular crush structure for FSAE cars. Is anyone willing to share a ride down graph? I suspect that it won't look very good but I would like someone to prove me wrong. I would be happy to crush one on our press and share the info if anyone is interested and has a spare one lying around.

By the way - the safety judges don't seem to care what you have. I pointed our trick carbon fiber crush structure out to the design judges but none of them were interested and I don't think that any of them asked any questions about it.

This was just supposed to be a quick response but I have some exams this week and it helps take time away from them - sorry for the ramble.

[Travis Garrison]

It might be worth noting that one advantage of the CF honeycomb is that it, for the most part, gets out of its own way...turning to dust as it crushes giving it some more efective ride down distance for a given section of crush structure....


Travis Garrison

[Kevin Hayward]

James,

In case nobody else mentions it ... thanks for the info about crush structures. That has to be about the most informative post I have seen on these forums.

Kev

[Chente(mcbo)]

i want to thank you man...for you time expend it here, that kind of things motives me and sorry for my english!i will study your post, we are making the chassis structure with pvc in 1:1 scale to see all the mistakes but in other wau we are looking for all kind of info about crush zone!! and i´m really greatful four your help and known!we keep on touch!!

[Greg H]

This was done several years ago and I wasn't on the team. You will not make it to design finals with "it looked like we could survive a crash." I was just showing an option, you will of course need to do your own testing and be able to explain to the judges why you designed the crush zone with your method. Very thorough follow-up post though.

[Sam Zimmerman]

Great post, James.

It is unfortunate that many of the judges seem to care less about the engineering that goes into the crush zone. We have gone to competition with very well designed crush zones (but not near the testing you have done) and ones that just fit the rules. I don't think we won or lost points either way.

On a side note, I am heading your way on Friday morning and will call you when I get in town.
Denny, if you are going to be around this weekend, first beer is on me.

[Igor]

Maybe the judges should be forced to read James' post. I guess that most judges don't have a clue about crush zones (like most of us) and thus can't appreciate the engineering involved.

Igor

[Travis Garrison]

Sam,

I think we owe you a keg or two http://fsae.com/groupee_common/emoticons/icon_smile.gif

Travis Garrison

[vandit]

@ james waltman

thanks for such nice explanation. but still one thing is not clear to me .... i have not understood the set up at '30g' ?.... what i understood is , 30g is considered maximum value and calculation and experiment are done at that ... and the testing supports your 500psi value .... but how do we check the average acceleration value which should be less than 20g

[Tom W]

The rules were changed in 2006 to reduce the average allowable loading to 20g. The maximum peak loading is new for this year.

[vandit]

@ TOM

so this means, that while doing this 'hydraulic press test' we must have our ride down according to 20g now (average value, which was 30g earlier) .... or we must use 40g (maximum value) ...

[vandit]

@ TOM

i think i got the doubt clear. it is the average value of acceleration which is used to determine ride down ... james explained it in another thread...

[vandit]

@ anyone

is it to be understood that the slow 'hydraulic press test' is same thing as if we have slowed down our impact at 7m/s using each frame of high speed camera ? ... if not, then i am not able to understand, how does the speed factor is taken care of in this slow hydraulic test...

[Drew Price]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by shanky:
@ anyone

is it to be understood that the slow 'hydraulic press test' is same thing as if we have slowed down our impact at 7m/s using each frame of high speed camera ? ... if not, then i am not able to understand, how does the speed factor is taken care of in this slow hydraulic test... </div></BLOCKQUOTE>



It isn't the same. You can get away with claiming that the energy absorbed (and dissipated) during a slow crush will be the same as for am impact crush, and thus can show the average decel will be below a certain value (force over distance) if the slow crush data is a relatively smooth curve, I.E. no huge peaks

You can't do this for the peak decel value (40g for this year) because of the strain-rate dependency of whatever material you use for your IA. You will have to see what that is for your specific material, and see if you can claim that the dependency is a negligible enough effect that you can do the linear analysis from the slow crush.

If your material is highly strain-rate dependent in it's energy absorption (take a very brittle material, like glass for example, or ice) then you can't make that claim. That is when you have to use high-speed camera measurements, or accel data from an accelerometer or strain guage mounted to an impact test simulation.


http://www.sciencedirect.com/cache/MiamiImageURL/B6TWT-4N1JRTD-1-19/0?wchp=dGLzVtz-zSkWA


This is an example, for some material, of the strain-rate dependency at yield for three different speeds a few orders of magnitude in scope.

Strain rate effects in composites (http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6TWT-4N1JRTD-1&_user=1458830&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000052790&_version=1&_urlVersion=0&_userid=1458830&md5=0cc92b6b305538d8e2814cff2edfe5c8)



Best,
Drew

[vandit]

@ drew

thanks for a reply. cleared lot of doubts for conducting test and i might end up doing tests with some materials to understand things...

[JDan]

Well, I'm not sure about this approach but if we have that kinetic energy is 1/2(m*v**2) and the 'spring' energy can be 1/2K*x**2 then we may have that mv^2=kx^2 where K is the spring constant for a given geomtry and material and x is the deformation. The only problem I find here is that when you have a stress-deformation chart you know as well that there is an elastic part and a plastic part, so for the given x you must know which x is plastic and elastic, I don't know if this could be achieved by checking this particular chart for the material calculating the stress as you guy have been doing it. So it yield something like Ec=Ep(plastic)+Ep(elastic): mv^2=K(elastic)x^2+K(plastic)x^2. This done you may be able to calculate an average acceleration, but the peak can only be done either with a physical test or with a good FE analysis (maybe NASTRAN), Ansys might be able to give you this. I'm experimenting with ansys right now, see how things turn out. Any other contribution would be appreciated.

[EPMAl]

I am not to sure either about the slow speed testing. In our case, we use our composite nose as the crush zone, with no foam/honeycomb/other typer of crushing material. With this design, I'm pretty sure we have to go with the full instrumented crush test because of the impact effect on the composite material and potential (read assured) delamination between each ply. Also, since the geometry is curved, the crush is far from being steady all the way down, although it might work with honeycomb or more ''linear'' material.

[Drew Price]

I know there are some papers out there specifically pertaining to (the lack of) strain-rate dependence for composite laminate shells used as crush structures.

I think they are even SAE papers, because I think a lab at my brother's Uni published some, send me a p/m to remind me, I will see if I can get some info on it for you.

ALSO: keep in mine that when this thread started the revised rules for how the IA is scored and graded were not implemented. Doing only steady state crush approximation will not necessarily make the structure illegal pertaining to the rules, but you may lose some points for it, depending on how much the design judges take the IA report score into account.

Best,
Drew

[Adambomb]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Drew Price:

ALSO: keep in mine that when this thread started the revised rules for how the IA is scored and graded were not implemented. Doing only steady state crush approximation will not necessarily make the structure illegal pertaining to the rules, but you may lose some points for it, depending on how much the design judges take the IA report score into account.

Best,
Drew </div></BLOCKQUOTE>

Actually, there has recently been a revision to this. The IA report is no longer scored, but is only given a "go" or "no-go."

[twinky64]

This is what I do not understand. The 2011 rules state that a total mass is 300kg moving at a velocity of 7m/s. In order to do a dynamic test I need to figure out how much weight I must drop onto the impact attenuator. That being said, to find weight (W=m*g), I need to have something that weighs (661lbs)*(32.2 ft/ss)=21,284lbs of force? And if I want to achieve decceleration of 20gs then (F=m*a; (661lbs)*(32.2ft/ss)*(20)= some outrageously high number!) This doesn't sound right to me.

So then, should the 300kg be the weight instead of the mass?

[Jimmy01]

I think you are getting yourself confused with mass, weight and force.

The maximum average force(F)(to keep acceleration under 20G) which the attenuator can apply is very easy to calculate:

M = mass of 'car'= 300kg
a = 20 * 9.81 m/s2 = 196.2 m/s2

F = Ma = 58860N

[twinky64]

"I think you are getting yourself confused with mass, weight and force.

The maximum average force(F)(to keep acceleration under 20G) which the attenuator can apply is very easy to calculate:

M = mass of 'car'= 300kg
a = 20 * 9.81 m/s2 = 196.2 m/s2

F = Ma = 58860N"

I thought that 20gs of 300kg would be 6000kg? For instance, let's say you weigh 200lbs. At 2g's, you weight 400lbs, 3g's is 600lbs etc. Doesn't that same logic apply?

[twinky64]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Jimmy01:
I think you are getting yourself confused with mass, weight and force.

The maximum average force(F)(to keep acceleration under 20G) which the attenuator can apply is very easy to calculate:

M = mass of 'car'= 300kg
a = 20 * 9.81 m/s2 = 196.2 m/s2

F = Ma = 58860N </div></BLOCKQUOTE>

I thought that 20gs of 300kg would be 6000kg? For instance, let's say you weigh 200lbs. At 2g's, you weight 400lbs, 3g's is 600lbs etc. Doesn't that same logic apply?

[JasperC]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by twinky64:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Jimmy01:
I think you are getting yourself confused with mass, weight and force.

The maximum average force(F)(to keep acceleration under 20G) which the attenuator can apply is very easy to calculate:

M = mass of 'car'= 300kg
a = 20 * 9.81 m/s2 = 196.2 m/s2

F = Ma = 58860N </div></BLOCKQUOTE>

I thought that 20gs of 300kg would be 6000kg? For instance, let's say you weigh 200lbs. At 2g's, you weight 400lbs, 3g's is 600lbs etc. Doesn't that same logic apply? </div></BLOCKQUOTE>
6000kg = 58860N, Einstein. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Jimmy01]

Your logic is fine, but I don't think 'weight' should ever be used in an engineering context. Mass and forces are all you need.