Scalability of Impact Attenuator Testing

Greetings all,

As part of my university's Formula SAE, we are trying to develop a way to test the 'crush foam' (aluminium honeycomb structure, with cells having an approximate diameter of 8mm) to fix to the nose of the car. (To get an idea of the application, check out http://www.fsae.com/.) To prove that our design is the best and exceeds regulation, we need to develop a destructive testing method to gather data. Only one test is required, but we must show that our design can slow the car down with accelerations less than 40g. For a full scale test, we need to drop 300kg of weight (potentially weight plates from the gym) from 2.5m above the foam. We have access to any/all the sensors required to make measurements.

The difficulty we are having is scaling the test to still yeild accurate/meaningful results, since letting 300kg free-fall from such a height is... challenging. So the question is one of scaling. So far, here's what we know:

• The attentuator is strain-rate sensitive, so the rate at which the foam crushes is important.
• We are looking to scale the frontal area of the foam, while leaving the depth unchanged (i.e. so foam can be crushed to its extreme)
• There must be some sort of scaling that yields a 'critical surface area' where the size of the hexagonal 'pores' of the attenuator become a significant concern (assuming that above the critical area pore size is negligable?). Any thoughts to determine this scaling factor (i.e. if A=L^2, then L/d >= 10 should be alright for d = pore diameter, and L = length of side of test block?)
• Mass dropped, distance dropped and area of test specimen scale linearly with energy, but the energy absobed is porportional to square of velocity of the falling object (determined by its height).

Have with missed any factors that may skew our test results? Can anyone suggest a method to test a scaled version of the 30cm x 30cm x 30cm (approximate actual size) cube of foam that will be fixed to the front of the car? Ideally, if we could scale it to dropping say a 20kg mass from 2.5m or something, that would be much easier to achieve. How representative of the full-scale part would this test be though?

As with several of my other posts, I realize that this is not a homework site. I am not looking for an answer, just want to make sure we've covered all of our bases, and that our time and testing will yeild meaningful data. Thank you all for any suggestions and input you have.

Happy new year!

-MechMatt