Suspension geometry vs torque steer

[sierra4000]

Hello,
How much is usually acceptable king pin offset and scrub radius value for high powered FWD cars?
thank you for experiences
Radek

 

[BrianPetersen]

Your question prompted me to go out in the driveway and make some rough measurements.

VW and Honda have traditionally been proponents of a small but slightly-negative scrub radius (perhaps in the 5mm range). With MacPherson geometry that generally means rather high steering axis inclination, and that means a lot of camber change with steering movement.

Fiat seems to want to keep the tire as vertical to the ground as they can, via less steering axis inclination, still with MacPherson. I have two very different vehicles with Fiat design origins in the driveway (a 500, and a Ducato van) and I just made some rough measurements that suggest a positive scrub radius of perhaps 20mm give or take, for both of them.

The 500 doesn't have enough power to torque-steer. The van's steering will get skittish if you accelerate hard (ish) on uneven pavement although the stiff, high-pressure, load-carrying tires probably catch some of the blame for that. Neither have limited-slip differentials.

Limited-slip diff plus any non-zero scrub radius plus uneven left to right grip is going to impose a significant torque around the steering axis that will be different left to right (i.e. torque steer).

Lots of vehicles these days have electric power steering and the control module communicates over CANbus ... it can know how much torque the powertrain is producing, and can introduce artificial compensation ...

"Small but non-zero" are about all I can suggest. "Keep it between 20mm positive and 10mm negative" perhaps. Whether you use the Honda/VW design philosophy (slightly negative at the cost of higher steering axis inclination and thus higher camber change with steering angle) or the Fiat one (low steering inclination to keep the tire as vertical as possible, at the cost of having a slightly positive scrub radius) ... both seem to function well enough.

The fancy suspension used on Focus ST and certain high performance GM models, which separates the steering axis from having to go through the strut mount and thus allows the steering axis inclination to be kept lower, is perhaps worth study.

 

[GregLocock]

Two of the engineers in our department were part of the development team on the fancy Focus Revoknuckle suspension. I agree +/- 20mm is a good target for scrub. It is important under braking as well. Of course nowadays we just sim it to death until it is good enough, and as you say EPAS can auto compensate for it (I haven't worked on a high power FWD EPAS car yet).

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[sierra4000]

Thanks,
Yes, i know Revoknuckle , my interest is only about McPherson design

 

[BrianPetersen]

I believe the VW and Honda motivation for slightly negative scrub radius began with their use of diagonally-connected braking circuits in the early 1970s, the thinking was that if one braking circuit failed then the slightly negative scrub radius would help keep the steering stable despite one front wheel being heavily braked and the other one being not braked at all. The motivation for diagonally-connected braking circuits came from the highly front-biased weight distribution, which would leave the car with very little braking if it were done otherwise (conventionally, front and rear) and the front circuit failed.

I honestly don't know if either Fiat's brake circuits are connected diagonally or front-rear ... but they both have ABS, which was nonexistent in the old days. ABS can also introduce a situation where one side is braked more heavily than the other, without even having a technical failure in the car; simply hit the brakes on a split-coefficient surface hard enough to activate ABS on the slippery side.

My gut feel is that torque steer effects will be minimised with slightly negative scrub radius, but I can also say from practical real-world experience that my Gen 1 1970s-era Honda Civic without power steering and 60 hp on a good day definitely had a wee smidge of torque steer, and my current-generation Fiat 500 which is about as close as you can find in size and weight nowadays doesn't despite having more power (and positive scrub radius!), although it has power steering which may mask it. Neither one has equal-length halfshafts.

Bear in mind that ANYthing unequal left to right, whether by design (unequal-length halfshafts, which means unequal driveshaft and CV-joint angles), or by accident (alignment wrong, bent parts), or by wear and tear (busted and worn out stuff), could introduce torque-steer effects.

 

[cibachrome]

A lot of the agony of "torque steer" (torque induced steer) in high powered FWD cars comes from the loss of scrub radius during WOT activity. If you are "lucky", the brand/construction/pressure of the tires on the front will preserve what little negative scrub radius you have. I would bet the OEM supplied tires on your car do OK unless their Engineering department has only book learning and low bandwidth propeller heads in the office.

The trouble starts when you replace the OEM tires with something else. And, money/price is no guarantee that you will be happy with your choice or anybody else's.

This because most tire and vehicle manufacturers have no specification on Mx (overturning moment) for the tires they choose to equip and sell. This property is related to the Pneumatic Scrub response of the tire. Mx over Fz gives you a hint at what the effective scrub radius actually is under static and dynamic conditions. This data comes from specific tire tests or road measurement wheels. (I'll have the Flat-Trak tests, please). Under WOT, if the scrub radius goes positive for any reason, you're gonna need some wrist protectors.

Why is this ? Simply: the Mx / pneumatic scrub can add or subtract from the static, geometric value. There appears to be a full range of DYNAMIC sub radii available to you, depending on brand, size, construction, pressure rim width, blah, blah, blah.

Here's a typical example from a test case of a V8 FWD pass car with a lot of steam. Some of these cars were 'campaigned' to fix the problem with torque induced steer (not the unrelated lead/pull issue).

Now you can see how much a tire can wipe out your static scrub radius, or wipe out your tread depth, depending on what tires you buy. Which tire would YOU choose ?

 

[sierra4000]

Very interesting and valuable Brian also Cibachrome!

of course often is "modified" car with:
lowered ride height (is KPI changed due to camber gain)
bigger wheel
wider tire (sometime even chinese semislicks!)
wheel ET changed to bad direction
etc.
then car becomes killer

from picture seems instant toe (slipangle) is quite significant factor
how way camber can affect?
due to slipangle change?
contact patch vertical load center is shiften?
how tire pressure?

some ideas:
please correct me if some is bad...........tire is unknown of course but generaly

bigger tire diameter is better
more tire aspect ratio is better
more tire pressure is better
more camber is better
narrower rim is better (can use more ET)
wider tire on same rim wide is better?
slight static toe-out is better
others?

 

[BrianPetersen]

"Better" in what way?

Often things can make one aspect of the suspension/steering "better" while making something else "worse".

A single-minded focus on "fixing" one particular aspect can make everything else worse.

The whole package has to work together.

Some of the stuff that so-called "tuners" do, is head-scratching. Perhaps in some cases they know about the bad side-effects but in other cases they don't even know about them. Then they build something and wonder why they just crashed into a tree.

If you're dealing with an existing vehicle design and it's front wheel drive and you don't want worse torque-steer than what it was built with, just make sure the scrub radius is somewhere near zero taking into account the sum total of all of the other effects (do the math), refuse stupid camber settings, refuse stupid tire sizes.

 

[BrianPetersen]

Correction, perhaps I should have said "stock" rather than "zero", to perhaps allow for situations as with my little Fiat, where the OEM has somehow managed to sort it out decently despite positive (but reasonable) scrub radius.

 

[sierra4000]

I'm sorry,
the "better" is meant less torque steer if we consider that the most common cause "worse" is a increasing positive scrub radius

 

[BrianPetersen]

"Bigger tire diameter is better" - why?
"More tire aspect ratio is better" - One would expect very low profile or high sidewall stiffness tires to be sensitive to how closely the tire is perpendicular to the road, whether due to static camber, or camber introduced by high steer angle plus high KPI, or body roll when cornering. One would expect high-aspect-ratio tires to be less sensitive to this because the sidewalls have more wiggle room when the tire isn't perpendicular to the road.
"More tire pressure is better" - why? The Ducato runs very high tire pressure because of the load ratings required, and it's worse for torque steer in my experience. Very high tire pressure reduces the ability of the tire to cope with being not quite perpendicular to the road ... see line item above. I can see how there could be competing effects. This vehicle (which is front wheel drive and appears to have slightly positive scrub radius) does not like accelerating hard on irregular pavement. (But it's no sports car, so perhaps it's better if it sends a "stop doing that" message to the driver)
"More camber is better" I would say not. I would say that the closer the tire is to perpendicular to the road, the less the center of pressure of the contact patch is likely to be off the centerline of the tire (i.e. affecting the scrub radius).
"Narrower rim is better" - that depends on how much clearance you have to work with, although this is in a way connected to the tire aspect ratio.
"Wider tire on same rim is better" - Why do you expect that to be the case?
"Slight static toe-out is better" - Toe-out usually has so many other bad side effects that it won't be worth trying. The car might end up so skittish due to excess toe-out that the driver won't notice torque steer ... If there is so much compliance-steer that the drive and braking forces are measurably affecting the toe to the extent that you are using static toe-out to compensate, then maybe *that* is what needs fixing.

Cibachrome has a handle on the math and simulations needed to properly address this. I don't, I have only the school of hard knocks and a few thought-experiments and real-world experiments to go by.

 

[sierra4000]

thank you for react Brian,
i will try to better explain my thoughts

"Bigger tire diameter is better"..............because doing scrub radius less positive due to KPI?
"More tire aspect ratio is better"............maybe due to tire less sensitivity?
"More tire pressure is better"................presumption is that most car have non zero camber (negative), then more tire pressure will change contact patch shape? (contact patch load center is shiften to less positive scrub radius)?
"More camber is better"................same reason?
"Narrower rim is better" ..........can be use more wheel ET
"Wider tire on same rim is better"..................same reason?
"Slight static toe-out is better"................yes, due to compliance , and i think a big positive scrub radius increase longitudinal compliance again..........so eliminate toe-in under strong acceleration?

 

[BrianPetersen]

Okay, "bigger tire diameter in the absence of changing anything else", yes, this will send scrub radius towards negative due to the steering axis inclination. I really can't comment much further on the tire-related matters except I am thinking that if the tire pressure, camber, etc are in a plausible range (and are not set stupidly out of whack), slight but plausible differences in tire pressure etc just aren't going to matter.

If you have negative scrub radius and front wheel drive, forward drive thrust at the front wheels will impose a toe-out torque. If you already have static toe-out, that's not heading in a good direction.

Please back up for a moment and tell us "the big picture" of what you are attempting to accomplish and with what specific vehicle.

 

[sierra4000]

Well,
we have FWD car (Nissan Sunny GT) with 250Hp with brutal torque steer,
we build new uprights with original geometry to optimize bumpsteera and roll center, but his custom damper position not allows less wheel ET
now we have around 20-30 scrub radius
probably problem is also on side too much agressive LSD

 

[BrianPetersen]

Yup. That's not going to work. As soon as a limited-slip diff is involved, scrub radius plus the differing left-right torque is going to create large differences in steering torque reactions between left and right. I think you are now stuck with re-arranging this for zero scrub radius by whatever means are necessary. Zero camber, too (although perhaps adjustable near zero).

I know it has been stated that actual zero scrub radius is bad, but in this case, non-zero is probably worse. You can always design it for 5mm negative and use wheel spacers to tweak it.

High KPI plus the expected super wide ultra low profile tires that I suspect are also part of this, are going to result in badness when the steering is away from center (because of the camber induced by the KPI when steered away from center and the inability of tires of that profile to deal with that, they'll try to go up on the corners of the tread profile). If it's a roadrace car or similar, steer angles should be fairly low when the car is on track and it matters; the wacky torque reactions will only be when puttering around the paddock, which doesn't matter, "deal with it".

 

[sierra4000]

Yes,
225/45/18 tires with -1,2 camber, KPI more or less standard value , but problem is just also with straight ride under strong acceleration, when diff feels like locked and steering wheel require really firm hands to maintain straight ...........on low gears when one tire grip is lossen, the car becomes killer.
(rose joints are everywhere)

 

[cibachrome]

Engine mounts are players in Torque Steer, too. This because the powertrain is set in position under road load for symmetric joint angles, etc. But under WOT, your motor+transmission is certainly not in road load geometry. Take the hood off and see how much asymmetry occurs. That's why your Honda has a bit. As long as you don't dent the hood, your parts will be ok. BUT, the inboard angles are going be very different. Vertical and longitudinal transmission movement does some other things to the geometry as well (like asymmetric toe changes). Rig up a split screen camera and convince yourself that all this is going on.

 

[sierra4000]

Yes, this engine have main mounts on inclined axis (engine head mount upward and gearbox mount lower)
we will try to hold firm in place!

maybe important note:
the problem is symmetrical ............. once pull right, once left..............so i thing toe changes are symmetric, then depends which specific wheel have better grip

 

[sierra4000]

Why some tires (or worn tires) create strong tramlining?
is same factor who increase torque steer?
this can be significant torque steer contribution?

 

[cibachrome]

High Aligning moment camber stiffness in tires and high aligning moment camber compliance in the suspension. Clearly a marriage made in Heaven