Suspension design priorities 5

[NeilRoshier]

In conversation with another motorsport enthusiast the priorieties of suspension design was discussed at length (and still is being discussed) and it was interesting to note that we had differing views, and infact that there was little in the way of concrete guidance in our books (including milliken/staniforth/smith). Could people look at the following two suggested vehicles and then rank the areas in what they see would be the priorities in highest to lowest (and add any that we have missed). What was agreed was that tunability of performance, servicability and predictable performance were all very important to us. Both cars would be using modified road car uprights etc (so not the lightest suspension components) and semi-slick type tyres (e.g. yokahama 032r etc)

1. A 'clubman' (lotus7) type car with a live rear axle. Approximate weight 500kg, 120kW, front engine RWD. Biased towards track use, with up to 90mm of total suspension movement. Expected close to 50/50 weight distribution.

2. A mid engine RWD coupe (ala' lotus esprit), weight 850-900kg, 200kW, for track and limited road 'targa' type events. Strong bias to handling/performance over comfort. Expected 110mm suspension movement, 40/60 (F/R) weight distribution.

Suspension Priorities
A. Outside wheels upright in roll at 1G
B.Vericle height of RC constant with regards to (wrt) CoG
C.Minimise lateral displacement
D.Minimise skewing of RC axis wrt mass centroid
E. Make links as long as feasable to minimise bump effects
f. Min Scrub
G. Min Offset
H. Damper ratio relative to wheel motion
I. Ackerman ratio
J. weight distribution/size of contact patch

 

[pmwltd]

Some other items you need to factor in.
tire data
design ride frequecies front and rear
maximum design roll angle
pakaging contraints
Compromise between ideal camber curve and stable RC, can be a challenge. Usually based on FVSA.
What gives you the maximum contact patch at the critical places on the track with even tire temps and tire wear?

Making the right decisions on the inevitable compromises is a challenge

 

[NeilRoshier]

pmwltd could you expand what you mean by tire data? I have an idea of what it means to me, but I would be interested in your thoughts.
Design ride frequencies have not been something that we have discussed. I would assume that they would be 100cpm+, naturally with a difference F/R to stop pitching.
Roll angle may vary depending on use but had thought around 4deg, though my friend may not agree!
Packaging constraints, well for the sake of this exercise we are prepared to overlook that in the search for more knowledge from the forum members.
Suspension is a challenge and those that seem to balance the variable do so much better. The more knowledge I have hopefully the better I will do!

 

[NeilRoshier]

In a discussion with the other party, he felt 4deg roll at 1G was way too much and tht 2deg was the better number. So again something for the melting pot! I felt that 4deg given the dual nature of the cars and our local road conditions, esp when wet was reasonable, he did not.

 

[pmwltd]

What I have done in initiating a suspension design in the past is go through Milliken's ride and roll calculations, the simplified version on Pg 601. I have put that on a spread sheet and play around with different inputs until I get some numbers I feel comfortable with. Milliken has some assumptions that you can use as guidelines. Then I layout the suspension based on track and wheel base dimensions, tire and wheel size allowed by class running in. On front SLA suspension, keep lower control arm as long as possible and use length and location of upper control arm to give desired camber curve and RC. I use Bill Mitchell's software to layout the front and rear suspension geometry, and steering geometry. Then based on packaging limits start to make necessary compromises.

Tire data such as slip angle sensitivity, camber sensitivity and types of race tracks car will run will dictate camber curves, ride rates, roll stiffness, etc.

One rule, the lower you can make the CG the better off you are.
Dave

 

[GregLocock]

I think 2 degrees at 1 G is going to overload the outer tyres too much, and will give a very uncomfortable ride, due to head toss.

A very high priority should be long lower arms parallel to the road at race design height (ie driver plus half a tank).

This will minimise your kinematic steering inputs, which is your main subjective priority.

I am getting more and more cynical about roll centre locations the more I read.

I'm puzzled by the upright outer wheel. The graphs I have seen (not many) indicate that slight negative camber (of the order of 0-1 degrees) is CRUCIAL for max lat acc performance, yet few people set their cars up with camber gain roughly equal to roll gain. I don't know why.

Do you mean offset? that's just the displacement of the mounting face of the wheel relative to its centreline, ie a (very important) packaging decision, not a geometrical kinematic consideration.

Scrub radius can be anywhere on a racer.

As discussed elsewhere Ackerman is important for low speed low power vehicles, but is irrelevant for a racer. Rod recommends parallel steer from his own experience.

It is good to get the shock ratio reasonably high, particularly in light of the low velocities seen on a circuit car. On the other hand anything better than 0.7 seems to work.

Don't know what you mean by J. If you are asking whether the control of contact patch geometry and forces are important, yes, the CP is the only reason you get round the track, so that is priority 1, everything else is fine tuning.




Cheers

Greg Locock

 

[pmwltd]

Going back to roll centers. There is more focus as of late on FVSA location and length than on RC alone. As you shorten the FVSA you increase camber gain and visa versa. Also as the FVSA gets closer to horizontal the lateral movement of the RC increases. Also as you raise the roll center you cause an increase of lateral weight transfer through the geometry, which happens faster than if the transfer is through the suspension. Some solo event racers claim that raising the RC helps with quickness of turn in responce. Works very similar to anti-dive. There are some circle track racers setting their cars up with assymetrical RC's.

 

[GregLocock]

"Some solo event racers claim that raising the RC helps with quickness of turn in responce." Sure, that is the conventional wisdom for roll centres.

But

A racer generally talks about the geometrical roll centre, typically at design height

The SAE talks only about the force based roll centre.

The geometrical roll centre and the force based roll centre start in different places and move in different ways in response to hardpoint changes and wheel articulation.

Therefore they are not the same thing.

When the racers change RCH are they leaving ALL other parameters alone? I doubt it. It is very hard to design a strong realistic suspension in which you can change just one parameter reliably. Our hardpoint model for fine tuning SLAs runs to a page of equations, and will only work reliably for small changes. That tells us how to move ALL the hardpoints in order to change just one setting, and it is not always practical.

Since, in reality, we are concerned with the rate of force build up at the CP, it seems to me that a time domain model that includes ALL the paths across the car is the most likely to succeed in identifying which characteristic(s) is important. Milliken has a few graphs indicating this sort of measurement. This will only translate into a directly useful model when we understand the force vs time characteristic of the tyre, which is something that is only now being investigated systematically.



Cheers

Greg Locock

 

[NeilRoshier]

Greg thanks for your input. When you say that you are ecoming increasingly sceptical about RC's do you mean the geometrical RC or the force based RC as well?
I am also interested that you have seen few people setting their cars up with camber gain roughly equal to roll gain, it would seem a 'logical' thing to do. It would be interesting to know why no one has done such a thing. However I will take on board your thoughts on the outer wheel angle being between 0-1deg neg camber.
With regards to J, sorry it was poorly expressed. You did answer what I was trying to ask, so thank you.
Is the scrub radius potentially an issue with limited road use cars? I understand what you are saying regards competition cars, however both the proposed vehicles will need to see some road use. I have been told that minimising scrub is advantageous in such circumstances?
The long lower arm is interesting. I have been scouring the net looking at suspension pictures (not everyone's cup of tea) and have found it interesting that there some successful cars (e.g. those by Lee Noble-Ultima/M12) that use very short suspension arms, yet some others (e.g. the new farboud GT) do the opposite. Both designs seem to get good press for their handling. Am I correct then in thinking that the kinematic steering effects in the short link suspension has either been overcome or jsut not noticed/reported?

 

[GregLocock]

All RCs. It seems to me that geometrical RCs are useful as a first look, but that's it. Force based RCs are telling you something more important, but once you have worked them out you are so close to knowing the CP forces that you might as well work them out directly. Can you honestly say when you read the SAE deinition of a FBRC that it sounds as though it is important? Not to me.

Scrub radius on an unpowered axle isn't an issue, I think (It is of great concern on a FWD of course). One of the posters here has an early Mustang, he thinks it has 100 mm of scrub radius, stock (!). Nowadays we aim at around +10 to -10, to get reduced steering input from bumps, but that is not an important consideration on a racer. If you add more you will get more steering feel, but also greater pull when braking, and higher efforts when parking.

Long arms just reduce the drama. In practice so long as you can keep the angular motion of the bushes/rose joints/ ball joints within their design limits (20 degrees often), and can still fit sensible springs and shocks in, then there is no harm in letting the arms get short, so long as you have the freedom to move the hardpoints /exactly/ where you need them and can make everything stiff enough to cope with the loads. In practice it is easier to use long arms and desensitize the whole hard point geometry issue. However with your fairly short travel suspensions you might get away with 7-8 inch lower arms, particularly if you got equal length wishbones (which is not a good idea).


Cheers

Greg Locock

 

[NeilRoshier]

Thanks again for the replys Greg. I have only read a limited amount of information about FBRC and do not have anywhere near your level of expertise. So while some things have seemed important to an interested 'lay' person such as myself when reading soem bits of information, I know that I do not know enough and hence these questions on this forum.
I will not pay that much attention to scrub from now!
I went along to a wreckers today and had a look at a Honda NSX, Porsche 928/968 and Nissan 300zx (primarily looking for uprights). The Honda and Porsches had longer lower arms than the nissan and all were longer than Noble's cars by a fairly big margin. Again thanks, I will keep learning.