Kingpin Inclination, Caster, etc. for McPherson strut 1

[hutch325]

A tangential offshoot from another recently-started thread.

I'm doing some kinematics analysis on a front suspension with that's McP strut with the L-shaped control arm (i.e., not a two-link with a virtual steering axis). Scrub radius is near zero from the factory.

I'm investigating the differences in the regular and "motorsport" versions of the car, and what was changed through the life cycle of the models. I'm looking at it from a race car prep perspective, trying to figure out the effects of common modifications such as offset bushings, strut mount modifications for more negative camber, etc.

Any thoughts on increasing KPI on a strut front suspension? Some initial sensitivity studies indicate that increasing it makes the camber curve a little steeper but I have some concerns:
- What about negative effects for steering feel?
- Practical implicaitons of less negative camber on the outside wheel as a function of steering ? It seems like for smaller steering angles like on a road course, the improved camber compensation may win out?

Thoughts on increasing or decreasing castor? First, some observations:
- Increasing castor to the "motorsport" spec by moving the strut mount rearwards without making any other changes other than re-setting toe (i.e. not correcting bump steer) seems to reduce the steering feedback somewhat. I attributed this to greater mechanical trail.
- The "motorsport" version of the car has about 7-8 degrees of castor as opposed to 4-5 for the regular version. The manufacturer changed the upright to change the longitudinal relationship between the CA outer balljoint and wheel center. Presumably this is to achive the desired amount of mechanical trail, but I haven't completed the kinematics model so I don't have real numbers.
Next, some questions. Realistically,:
- How much of a benefit is the additional outside wheel negative camber with steering resulting from larger castor angles?
- Any recommendations on how much mechanical trail is reasonable?
- Thoughts on asymmetric jacking forces?

As a side item, a brief conversation with one of the engineering directors from the manufacturer seemed to indicate they have been reducing camber change as a function of travel in recent models in an effort to improve the behavior of run-flat equipped cars when going over bumps. I'm guessing tire wear, stability over bumps, and keeping it in the understeer regime are more important to the manufacturer than maximizing the cornering power of the front end.

Any thoughts, and particularly real-world experience, are welcome.

 

[GregLocock]

Answering that lot would take a book.

What is the purpose of mechanical trail? I think you primarily use m.t. to make sure that the lateral forces are always reacted behind the true steering axis.

During a manouevre the pneumatic trail point moves around, and with a shorter, wider contact patch, it is more likely to get too close to the steering axis, which would reduce the steering feel, and ultimately go over-centre.

Too much m.t. is likely to prevent the driver from feeling the change in pneumatic trail, ie he wil find it more difficult to approach the limit of adhesion.

I can't think of a good reason to have scrub. The original VW Golf/Rabbit I think introduced the idea of using it to compensate for partial brake failure. The only proviso there is that it is a bad idea to have forces connected to the steering wheel that can move either side of a neutral point.

Well those are the easy ones!







Cheers

Greg Locock

 

[hutch325]

Yeah I'm not surprised it would take a book, too bad RCVD doesn't seem to have the information I'm looking for.

Your assessment of the pneumatic trail and M.T. is consistent with the information I've come across through reading.

Typically on these cars people increase the negative camber and the castor angle by moving the upper strut mount as part of an effort to reduce understeer, without regard for other parameters (like bump steer, KPI, etc.). It seems like a less than ideal approach to me so I'm trying to quantify some of the trade-offs. But, I would prefer to have some design goals for an initial starting point. The strut suspension makes it tough though.

One of the myths I'm trying to bust or prove is "these cars work great with lots of castor." Well, maybe but does it work better with more castor or KPI? Or less of one or the other? Or other things like M.T.? So many questions. This could take a while to figure out. Too bad I don't have more advanced analysis capabilities.

As a side item, is Castor a relevant parameter to consider for a rear suspension that's trailing arm with upper and lower lateral links?

 

[GregLocock]

Last question first

No, I don't think so. Castor is a way of introducing a second order correction based on steer, in roll. No sensible rear suspension has much in the way of rear toe (say 1 degree max), so the castor is working off a very small base. But I could be wrong.

Front suspension castor is a complete gotcha. I do not know how to optimise it. FWD cars typically run 4 degrees, RWD 7 degrees. It affects yaw damping. It hurts steering feel.

KPI works a bit like castor in screwing up steering feel. Again, I don't know how big the tradeoff is. It is hard to imagine that it really helps anything.

So from all that I get:

zero scrub

zero kpi

some (but not too much) m.t.

some castor


Cheers

Greg Locock

 

[hutch325]

Interesting. Back when I did Formula SAE in school that was pretty much our design philosophy. Zero scrub, zero kpi, about 7 degrees castor, not sure about m.t. That was with SLA though so we could get a good camber curve by getting other things right. It seems like the whole McP front is a gotcha, unless you don't care about camber curves.

Thanks for the info on the rear.

Regards,

Hutch

 

[patprimmer]

Maybe I have a wrong idea about what King Pin inclination is, but as I understand it, it is the angel between the perpendicular and the axis of the king pin, or the axis that runs through the pivot points of the upper and lower ball joints, when viewed from either end of the car.

How can you get zero king pin inclination and zero scrub radius with a strut, unless the strut either fits inside the wheel, or is bent around the top of the tyre so that the pivot ends up over the centre of the tyre

Regards
pat pprimmer@acay.com.au
eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[red300zx99x]

I would think it to be very hard to get zero kpi AND zero scrub on any front suspension, but it sounds like a good guideline.

 

[hutch325]

Our zero KPI and scrub radius was on a small formula car. We accomplished it with upright and wheel center designs that let us get the whole upright inside the wheel to get the desired geometry.

Pat, you're right - that's not possible with a strut as far as I know, unless maybe you use Ford's design with upper and lower balljoints on the strut but I guess that's not a true McP configuration.

Hutch

 

[red300zx99x]

did you have the rim with an offset far outside the wheel center? How well did the design work?

 

[hutch325]

It's been a while so I don't recall the wheel offset specification, but yes the mounting flange was outboard of the wheel centerline. I thought I would be able to remember more details since I had the uprights in my hands a dozen times.

From what I heard the design worked well but I never got a chance to drive it.

 

[GregLocock]

Sorry Pat, I meant those would be ideals, not necessarily achievable.

I think Audi run a virtual ball joint at both ends of an SLA, so they could have zero kpi and zero scrub. I don't have the hardpoints so I don't know if they have, but I bet they haven't gone to zero zero.

Cheers

Greg Locock

 

[MoreWing]

I'm a fan of all things in moderation. I think there is a place for KPI, scrub radius, caster, mechanical trail, etc.

The balance between KPI and caster is touchy, but I think in a lot of cars it comes down to the camber while steered. KPI tends to add positive camber to the laden tire while caster tends to add negative camber to the laden tire. Caster's camber influence works at about 2:1 when compared with KPI. If you run no KPI and some caster, then you get a significant camber change on the laden tire just by moving the steering wheel. For that reason alone, I tend to run KPI at around a 2:1 ratio compared with caster (i.e. 5 deg. of caster with 10 deg. of KPI)

I also think that some scrub radius is necessary in a car. The first reason is that a scrub radius is a constantly changing thing. It is based on where the contact patch of the tire is, and tire lateral deflection is a big deal. If you have a scrub radius that is smaller than the tire lateral deflection, then the torque about the steering axis that is sent to the driver will change sign as he goes through the corner. I just don't think that this is a good thing. It seems as if we could at least keep that torque from changing sign, then we would be sending a consistent message to the driver's hands.

Secondly, unless you run perfectly parallel steering, there is some jacking effect due to KPI/scrub radius. A pro-ackerman car will jack weight in the same direction as caster and an anti-ackerman car will jack weight in the opposite direction of caster. Regardless of which way you are set to go, it's something that can be used to contol various aspects of the car. I think that the jacking effect due to KPI is a very underappreciated phenomena.

Excessive KPI does several bad things. It puts some really serious bending moments through a strut suspension, which makes coulomb friction go through the roof and mechanical grip go to hell. It causes large camber changes or at least large static caster settings to counteract the camber changes. It causes excessive weight jacking which makes the steering heavy and a little unpredictable. And, at very high steering angles can cause the 'flop over' and be kind of self-energized so that you have to steer it out of the corner.

There is an SAE paper about an FSAE car in the mid 90's by Ed Gaffney and Anthony Salinas. They refined and implemented the zero KPI/zero scrub radius approach that me and another guy designed in their 1995 racecar. We left school before we could race it, and they were in teh class behind us. The car wasn't bad, but there were some issues. Inside front tire wear was one of the biggest, even when running 0 degrees static camber and 4 deg caster. The car also had a lot of oversteer at corner entry and in transitions, but I attribute this more to a highly inclined roll axis rather than than the steering geometry. The chassis design was very sound; however, and the car is still running after nearly a decade of college student thrashing.

It amazes me that even to this day, that car has a lot of influence on FSAE car designs. I would say that 1/2 the cars in the SAE competitions run something close to 0 KPI, 0 scrub radius, front roll center below the ground, rear roll center above the ground, just like that car. It wouldn't have happened if those 2 wouldn't have written the paper, but since they did, a lot of teams use it as a blueprint. I suppose you could do a lot worse.

 

[hutch325]

Thanks, MoreWing. Either by coincidence or design this particular car has about 14 degrees KPI and about 7 degrees castor (a little less castor or KPI depending on the version of the car). I suspect KPI is probably about 16 degrees if the car is lowered and the upper strut mount is moved toward the center of the car in an effort to get more static negative camber.

Interesting and good point about Ackerman causing jacking with KPI. I hadn't considered that - I was mentally stuck thinking KPI would only result in symmetrical jacking forces (i.e. I was thinking parallel steer).

The SAE paper sounds interesting. I'll have to look it up.

 

[patprimmer]

Sorry guys, I am not a suspension person, but I am still a little lost.

As I understand it, the stub projects at a fixed angle to the upright or knuckle or strut or whatever.

This angle is designed to give a certain kingpin inclination relative to camber. If you change one you can't help but change the other. The king pin inclination relative to castor is very significant in relation to the change of effective camber on lock, that is as the tyre to road sees it, not as the chassis sees it.

On most road cars I see it tends to go the wrong way, as I would expect most tyre distortion on tight turns, and therefore an advantage with regards to turn in or steering response if the outside tyre went into reasonably strong negative camber on lock. This statement is based purely on angles I visually perceive in parking lots etc

Regards
pat pprimmer@acay.com.au
eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[hutch325]

On the car I am examining it would be possible to change the angle between the strut and the stub without too much difficulty (relatively speaking). That way, provided there's no interference with the body or the wheel, I can set the KPI independently of camber.