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

Automotive NVH Test for rubber mounts on firewall

  1. Aug 7, 2017 #1
    Hello gang! Im dealing with a new issue at work and I though you might give me some much useful insight.

    Im trying to fix a component to a plate and using 4 screws and rubber bushings to attach said plate to the firewall. The component is around 5-6 kgs and needs to held in place firmly but cant be attached in a rigid way because it generates vibrations in operation.

    I need to run some NVH tests to see if the bushings do their thing. Problem is I dont know how to approach this:

    -Do I mount the the whole thing to a dummy firewall in which I introduce vibrations with a shaker and attach accelerators to the bolts to see what gets inside the cabin? Or do I shake the component and see what vibration I transmit to the dummy firewall? What is the best approach to check how well the bushings isolate?

    -How would a test bench look like? I know I need to do some tests with component freely moving (modal analysis) but how will I check for vibration filtering for example in the Z direction? If I attach the dummy firewall solidly to a test bench and shake the component wont that affect the results considering in real life the firewall moves too?

    Im at a standstill here, NVH is new to me so all advice is welcomed!

    Attached Files:

  2. jcsd
  3. Aug 7, 2017 #2

    Ranger Mike

    User Avatar
    Science Advisor
    Gold Member

  4. Aug 7, 2017 #3
    Thanks for the recommendation Mike, but we are building this from the ground up and will be making our own bushings. So I still have to test them :nb)

    ! Also in the previous test for transfer function it is reported they excited the component with the shaker being handheld... Is this normal?
  5. Aug 7, 2017 #4
    What type of vibration does the component generate? ie. Does it rotate so will induce vibrations at a given frequency? What frequency range are we talking about?
    What equipment do you have available to make and record measurements?

    There are several approaches to test this.
    Two accelerometers: one on the active side (ie the bolt head) and the other on the passive (ie after the mounting spring). You run the component as intended and plot the vibration transmissibility.
    Modal testing in this case is probably best done with an impact hammer. Tap it in Z then measure the response. To isolate the vibration frequency needs to be 1.41x the natural frequency of the object on the mounts.
  6. Aug 8, 2017 #5
    Thank you for the input Chris :wink:

    The component has a spinning motor, gears, ball bearings and hydraulic valves.... We haven't yet purchased the equipment as we outsourced our initial test.
    I was thinking about getting a shaker, an impact hammer and some ICP 3 axis accelerators. We have a pretty large budget, so we will also get the necessary capturing and processing equipment/software for future tests, do you have any recommendations?

    I cant place the accelerometers on the same bolt like you described because of space constrains.
    This sketch shows how the first test was run (blue=component, red=bushings, green=accelerometers).
    They suspended the "firewall" using a frame on some rubber isolating feet. Should I keep testing like this?

    Attached Files:

  7. Aug 8, 2017 #6
    I've mainly used LMS TestLab in the past and a bit of HEAD ArtemiS for data acquisition and analysis. It's very, very difficult to advise on data acquisition without knowing the specifics of the problem as the devil is in the detail.

    Effectively you want the transfer function across the bolt. Measuring further away will add noise factors (no pun intended) to this measurement.

    So if you are only interested in lower frequency vibrations then you can assume that bit of the casing moves exactly in phase with the bolt (i.e. the casing can be considered rigid). If you are interested in higher frequency vibrations than that setup may not be suitable as the casing isn't stiff at those frequencies. If the direction of interest is horizontal (in that schematic) sticking the active side casing accelerometer on the side of the case may be better than sticking it on the top.
  8. Aug 9, 2017 #7
    I know toyota mounts their "abs actuator" to the firewall on newer prius cars but have separated it into two parts.
  9. Aug 10, 2017 #8
    I redesigned the test stand:
    Now the firewall will be bolted down to the floor, it wont sit on cart with wheels and rubber dampers built in the frame.

    The shaker will be held securely in a frame as well. I intend to make 3 measurements, one for x, one for y and for z.

    Is this right Chris?

    Attached Files:

  10. Aug 10, 2017 #9
    Setup looks reasonable to me. All you really need to do is make sure that you aim to reduce and unwanted movement (eg make that shaker mount super stiff and heavy).

    To be honest if you are planning to buy a system to do this, there will be some level of training included in any package. It's probably worth using this as a real world worked example and asking advice for setup and acquisition.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted