wheel load transfer. idea for static test

[eilonoz]

Hello all
i want to check some parameters at our 2012 veihcle during steady state (constant velocity & radius) cornering
the parameters are:
1- wheel load transfer lateral and vertical
2- chasis roll angle and cg migration
3- true wheel camber when cornering

milliken offers calculation which neglect roll angle & cg migration, and wants chasis and suspension stiffness and roll centers

i would like to hear your opinion about this following static test:
1- placing each wheel on a weigh-scale
2- pull the veihcle with a winch from the cg point (like centrifugal force directed from corner center)
3- read true angles, vertical load suspension rods positions and loads (strain gauges) and get the answers to the questions above

do you think it will work and will it be reliable?

thanks alot
Eilon

*- im not sure about lateral force because i wont have actual sleep

[BillCobb]

Its actually a common industry procedure called a "Sidepull Test" You need to know the c.g. height of the vehicle (including the driver, fluids, etc). The way we did it was to place a bottle of fluid at the cg with the liquid level at its location. A clear tube allows the fluid level to be maintained relative to the body as the vehicle is pulled laterally and it controls a system to adjust the pulling height. The pulling point height is adjusted by an electric bidirectional winch so that as the vehicle body raises, the pulling point also rises. This mimics the sideforce acting on the cg during a rollover event (which is what the test is often used to study). You need to place the tires on very high friction or cross-hatched grip plates so that they don't slip or slide sideways.

Other than that (its done very slowly), you should be able to measure lateral and vertical forces and moments at the tires, even if one of the wheels becomes airborne (which is common). You can calculate load transfer and geometric ratios too if you have the proper instrumentation. You may want to include the effects of a steering wheel input during the loading because the caster changes from steer, sideforce and roll (resulting from sideforce generation) can be very significant.

The only problem you face is the doubters not seeing tires rolling. A couple of simple road tests are needed to convince them its a good tool for validatinge equations and sim models of cornering maneuvers. The idea here is to "test what you simulate and simulate what you test". This was a phrase I was always known for, as well as "Good Cars Test Well".

A strain gauged pulling cable is all it takes to read out your net lateral force. If you have an wheel alignment machine vendor office nearby, you might be able to borrow a couple of wheel plates from a modern alignment machine that can read vertical and lateral force and even aligning moments. You are essentially performing a simple inertia relief type of test.

[DougMilliken]

Originally posted by BillCobb:
Its actually a common industry procedure called a "Sidepull Test" ... You are essentially performing a simple inertia relief type of test.

The machine concepts shown in RCVD Figure 8.3 can run this test with tires rolling.

Bill--did you run the test to high lateral load and then back down to low load (check for stiction and/or hysteresis)? Or, did you add some "dither" (vibration) to remove the hysteresis as you varied the lateral load?

[BillCobb]

The test was always run from zero load up to 2 wheel liftoff as the slide plates at the contact patches released. The test was proposed as a low cost alternative to running a dynamic rollover event with all of its problems. (You would need a robotic driver and a roll velocity sensor to determine the steer dwell end point on the fly). That's. the way the industry test is now performed (by Federal law). But, the sidepull test went back to Corvair days as I seem to recall.

My only point here was to urge these guys to try out a novel idea and an effective and easy way to explore the effects of steady state side loading. Some cheap instrumentation and a data recorder (or a steno with a protractor, some scales, pencil and paper) can get some pretty useful information.

The problems with production cars are that you can't attache the pull point at the cg location. Its just above the center console usually. So, you wrap a belt around the car and pull at a point translated to the cg. This usually causes body damage from the loads but a FSAE car doesn't have a lot of sheet metal.

This machine can really run it, too:

http://www.youtube.com/watch?v=SgrY-LN4GUs

[eilonoz]

how will i determine the direction from which i'm going to apply the lateral force?

according to figures like 'rcvd figure 5.5' it seems like different corner radiuses have different load distribution, because the lateral force's line of action isn't realy perpendicular to the vehicle center line

[Warpspeed]

That's true, but I would expect the difference due to tire slip angles to be small enough to ignore for this type of static pull test.

The lateral component should be close enough to measure some very useful weights and angles.