+ Reply to Thread
Page 3 of 3 FirstFirst 1 2 3
Results 21 to 29 of 29

Thread: Weight & Torsional Stiffness

  1. #21
    Senior Member
    Join Date
    Mar 2008
    Location
    Buffalo, NY USA
    Posts
    340
    Originally posted by Frank:
    The following statement applies to a chassis with no torsional stiffness: (case A)...
    Thank you for offering this excellent cheese, I don't believe the original poster is ready to appreciate such a gourmet experience.

    Nothing to do with FSAE, but here is a successful modern example where soft cheese is a good solution for a small utility vehicle:
    http://www.toro.com/grounds/ve...man/micro/index.html
    It has a central "twister joint" connecting front and rear frame. The pivot is a long rubber bushing with very low torsional rate (stiff in the other five degrees of freedom). The video linked under "Superior Ride Quality" has terrible marketing-speak in the audio, but does show the frame twist in action.

    Bonus points (not for trenchermen) -- what happens when the central pivot is raised or lowered relative to the front and rear chassis sections?

  2. #22

  3. #23
    There exists a roll centre at the front and rear, just as any typical suspension. Moving the height of the “Active In-Frame Twister Joint” above or below the roll axis would cause the chassis to crab. (front track not following rear track)

    When it is driving on smooth level ground....
    Moving the height of the “Active In-Frame Twister Joint” away from the roll axis will create some chassis torsional stiffness during lateral acceleration when the F:R Roll resistance ratio is not equal to the sprung mass weight distribution.

    On uneven ground...
    Moving the height of the “Active In-Frame Twister Joint” away from the roll axis will create some chassis torsional stiffness

  4. #24
    Doing the FEA correctly is not as easy as you may think.
    One of the primary sources of discrepancy between beam element FEA and reality is the fact that the tubes don't actually join at a single point in space, but at the welded perimeter of the joint. Because of that, the tube sections will not remain round when loaded. Beam end release values can be adjusted to improve the fidelity of the FEA model. The appropriate value to use can be determined by comparing the results of a single simple joint modeled with beam elements to the same joint modeled with shell elements that matches the real geometry of the tube joint.

  5. #25
    Senior Member
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    1,690
    Ziz,

    This topic has been covered countless times on this forum. Read this old thread for some different ways to go about it.

    Z

  6. #26
    With regard to FEA not being equal to physical testing

    I would say the primary sources are, in order of merit:

    Engine mounts not treated well in FEA( depends if engine is a major load path )
    Deflection of wheel hub bearings, inner race relative to outer race, due to moment at wheel
    Deflection of wheel hub bearings, inner race relative to outer race, due to force at wheel
    Contact stress of wheel hub bearings on upright and hub, due to moment at wheel (depends on fit)
    Deflection of rocker bearings, inner race relative to outer race
    Deflection of pushrod rod end bearings, inner race relative to outer race
    Deflection of shock spherical bearings, inner race relative to outer race
    Deflection of rocker bearings, inner race relative to outer race
    Deflection of pushrod mount on a-arm
    Deflection of rocker mounts, typically not modeled well
    Deflection of shock mounts, typically not modeled well

    It might be worth you time measuring these bearing deflection vs load curves on an instron.

  7. #27
    If you measure the stiffness of teflon lined spherical bearings or rod ends, used ones will give you more meaningful results. The lining compresses during the first day or two of use, so both backlash and stiffness increase. The suspension of a car with some miles on it will measure stiffer than a brand new car.

    Any assembly with more than a few joints will show nonlinear stiffness results, so either measure a range of forces and displacements, or use the same applied moment every time.

  8. #28
    I’m not talking about “bearing slop”, this does exist and will contribute to the histeresis of the chassis. This requires a whole new thread, and we will need to discuss chocolate and coffee in great depth.

    The elastic response (force vs deflection) these items all combined acts like there is another spring in series to the chassis.

    I agree that these items will not be linear elastic, they will have some non-linearity.

    If someone asked me to consider these items linear elastic and to give their “effective spring in series” spring rate, I’d say about 6000Nm/deg.

    This would mean a relationship between FEA and physically measured chassis stiffness would be: (values in Nm/deg)

    FEA Physical
    500 462
    1,000 857
    1,500 1,200
    2,000 1,500
    2,500 1,765
    3,000 2,000
    3,500 2,211
    4,000 2,400
    4,500 2,571
    5,000 2,727

  9. #29
    Originally posted by Dash:
    I'm with exFSAE on this one. I'd like to see some hard testing with results to show that it even matters. I think the main reason the production car world worries about it is because of the body panels having to line up all the time ( I've seen a commercial talking about it). Same thing with the unsprung mass. Maybe someone can bolt some huge chunks of steel to their uprights and observe the difference.
    We bolted in-wheel and then on-car 3kg per wheel of very high density cheese, made no significant difference in lap time. (Also looking for the viability of in-wheel electric motors). Maybe it would have made one in a top notch car, with top notch settings and top notch drivers. But with our car and drivers it made no observable change in lap time.

    For the frame stiffness goal, we chose a maximum anti-roll bar stiffness (Example, none in the front, stiffest possible in the rear) and chose a target in balance change versus same-stiffness rear and front. Using the same model as Frank but with tire modelled as springs also in an Excel spreadcheese. Therefore we have an idea of the range of adjustment in balance through roll stiffness we can achieve. The other point to look out would be natural frequency of the assembly, but I don't recall the results...

    P.S.: I can't believe no one suggested Parmigiano Reggiano as a top choice for high stiffness to weight, though very brittle I admit.
    :::::::::::::::::::::::::::::::::::::::::::::::::: :::::::::::::::
    2007-2012 - Suspension, chassis, and stuff (mostly stuff)
    Université de Sherbrooke

+ Reply to Thread
Page 3 of 3 FirstFirst 1 2 3

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts