Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Impact Attenuator-Experiment Design!

  1. Nov 20, 2008 #1
    At the moment, my collegues and myself have been asked to test some materials for their crush strengths. Unfortunately our dept. doesn't have a decent method for impact/crush testing, so we've come up with a modification to a Charpy Impact test.

    If you're unfamiliar, it's basically a known weight/hammer on the end of a pendulum which swings into a material sample. Usually the quantitative result is read from a gauge mounted on the inside of the swing arm. However, in the case of 'crush' testing the sample isn't broken in two, but rather compressed. So we've attached a back plate so the pendulum swing arm hits a rigid wall. We've encountered some problems though, and I thought I could benefit from a lesson or two in mechanics as it's been a while since a-level physics.

    We need to record 3 different sets of data: Velocity (of the hammer), Time and G (althought this can be calculated from velocity and stroke - the change in length of the crushed sample). I.e: We have a known maximum for Peak G and Average G, hence a range of materials are being tested. The variable in the experiment is the Material, all we need is a decent way to set up the data logger or something else, so that we can record everything we test.

    We're thinking about using a data logger and potentiometer and using the gradients of the (time, potential) graph gradients to assosciate with velocity.

    Any help, ideas or other ways of doing it entirely would be greatly appreciated. Note that if you help, you will be helping a real-life project involving the materials selection of an Impact Attenuator (crash box) for a formula one car.

    Thanks for reading!


    Ps: The attatched document is not my own and is the produce of Hexweb, the basic ideas are explained far better then I ever could. It also includes the basic mathematical elements. Note: we cannot do a weighted-trolley-impact-at-known-speed-type test due to H&S regulations in the dept.

    Attached Files:

  2. jcsd
  3. Nov 25, 2008 #2
    just as an update, we're going to use a DEKRA type test, involving a weighted trolley travelling at a known speed to contact out various material types.
  4. Nov 25, 2008 #3


    User Avatar
    Gold Member

    Welcome to PF, Bert.
    This isn't a typical response on the site; I'm answering from the 'no education' camp. The numbers as shown in your link mean very little to me. What would make sense from my perspective, and might assist in your data collection, would be high-speed photography of the impact with some sort of scale visible in the action. Correlating the footage with the sensor data would be very helpful to someone like me.
  5. Nov 28, 2008 #4

    Generally what is carried out during impact testing of this sort is the following:

    1. Measurement of the velocity of the impactor (in your case a weighted trolley, although it could be a pendulum or a drop carriage) in the last 50mm before impact. As the previous poster suggested this can be done using an optical method; however, it's pretty time consuming. I personally use a single light gate that is interrupted by a carbon fibre section with two 'teeth' of equal width.

    By measuring the output of the light gate, you will see the voltage change between zero and full scale when the beam is interrupted. By measuring either the width of one pulse or more preferably the time between the rising edges of two pulses and relating to the dimensions of the section, you can measure speed before impact.

    2. By mounting an accelerometer on the trolley (preferably linearly in line with the centre of mass of the trolley and the axis of impact) you will measure acceleration versus time for the impact event. From this you can easily calculate the peak deceleration of the trolley and the average deceleration of the trolley.

    What is more complicated (but often desired) is numerically integrating the deceleration signal: once to obtain velocity and twice to obtain displacement of the trolley, during the impact event. You can then plot the deceleration of the trolley against its displacement in the test; by converting the deceleration of the trolley to the decelerative force experienced (using Newton's Second Law) you will then have force-displacement data.

    The light gate and associated circuit is extremely low cost and provides very good accuracy for the speeds you will probably be testing at (I seem to remember 14.1 ms-1). The accelerometer will require an amplifier, and you will need a storage oscilloscope or data logger to record the data.

    Hope this helps,

  6. Nov 28, 2008 #5
    Just out of curiosity - is this for a FSAE (or Formula Student) competition? I can tell you how we handled this problem when I was in school, using some fairly crude measurements and numerical simulation, if this is what you're looking for...

  7. Jan 20, 2009 #6
    Ridiculously late responce, but actually yeah it is for Formula Student Kerry....
    We've been struggling for ages now, and still haven't done any tests. We've been to all of the 'realistic' departments which would have impact testers...and despite have a few to choose from non of them are suitable.

    One is for a sample milimeters in diameter (not enough honeycomb cells present)
    One is drop weight tester with some weirdo dart fitting which goes straight through the holes in the honeycomb.
    ...and the last one has been broken since January and is far to similar to the previous one mentions.

    Getting pretty silly now, but with January exams looming as well as individual project deadlines, we might have to go with the manufacturers ratings...which is a real shame seeing as we can't do our own tests so that all of the materials undergo the same method, under the same lab conditions.

    look forward to hearing how you got around your problems Kerry....

  8. Jan 20, 2009 #7
    Hello Bert,

    We did some simple drop tests on a block of honeycomb and measured the displacement (plastic deformation) of the top face of the honeycomb. If you assume that the deformation happened at constant acceleration, you can get acceleration as a function of dissipated energy (weight * height * gravity). You can test different area and depth honeycomb blocks to see how acceleration is affected.

    We took data from these tests and scaled it to make some plots and drive a rudimentary simulation in Excel. We were able to show (with questionable accuracy) that our impact attenuator met the specifications of the rules. How accurate is this method? Does the weight bounce when you drop it on the honeycomb? How much energy was actually dissipated by the block? You should be able to come up with a few more questions, but I think the theory is sound. Of course, the accuracy of the simulation is another area for things to go awry. Ours was definitely off a bit...

    Hope this serves as a good starting point for you. Good luck!


    p.s. What university are you from? My team (Drexel U. in Philadelphia) took the trip to Bruntingthorp in 2005 but I've heard that the venue is now Silverstone...
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook