I do not believe the bushing, beam or tire properties could be changed enough to measureably impact the vehicle roll gain unless there are special cases: The VW "thinking Bushing" (remember that ploy?) Wedges in the bushings that were supposedly lateral load dependent. Trouble is, no real effect ever showed up in K&C tests of properties vs lateral load cases. Oops. Since the bushing stiffnesses control the amount of lateral force oversteer, a very soft bushing would have horrific cornering results. Yes their are voided bushings stuffed in in some cars because the ride is so bad with a solid bushing. Also, since these are generally FWD suspensions, the majority of roll gain (influenced by RCH obviously) is controlled by the front. Roll stiffness is another effect to consider, though. The beam properties have a very great influence on the steer by roll (roll steer) coefficient since the warping properties of the cross beam are the mechanics of the controll arm deflection. This situaton greatly depends on whether the beam is open or closed section, though. The Opel/ VW "V" beams with the nose down had roll oversteer, requiring special, high cornering stiffness tires to give the car reasonable handling. The US versions with the nose up, had a reasonable amount of roll steer understeer which allowed lower stiffness (and lower rolling resistance) tires to be utilized. The rotational orientation of the V-beam had some major structural and durability influences, which Terry and a few others were able to solve. Obviously the weight of the vehicle is a big factor in the RCH and roll steer and roll stiffness coefficients, because the need for a Panhard bar becomes evident when lateral deflections/loads become higher. The track bar, crossbeam, and control arm mechanism is overconstrained. so the arm flexibiulity now is a factor as well.
But, the ability to have a decent, load dependent roll understeer coefficient (as the payload increases, the total vehicle understeer can essentially be maintained constant) makes this suspension design quite attractive.
Still, the ability to sense and measure the effects on the amount of roll is so dwarfed by the influence the same parameters have on lateral force steer and camber and roll steer, make the study of such things a good academic exercise only, IMHO.
The coolest thing you can do to teach new handling evaluators is to connect the cross beam of a twist axle to the control arms using something similar to a pipe flange using, let's say 6 variable bolt positions. By rotating the V beam around these 6 positions and driving the car with the same tires, your Grandmother could tell the difference bewteen the best and the worst case. Anybody who claims roll oversteer is good for a car's handling rating will have to admit the error in their thinking.
I'm a firm believer that a twist axle would make a very nice package for a small pickup truck because of the load dependent geometric properties. Having a constant steering gain truck, no matter what the load, is a nice, safe package. Just don't put in Level Control to screw up the advantages...
