Spring rates for Coilover system

[nitan]

Hi,

I have a small car that I am designing, with double a-arm suspension. I know the approximate weight, shock ratio, and a few other parameters,and want to calculate what poundage of spring that I need.

Thanks

 

[GregLocock]

The usual method I use for a road car is to guess what corner frequency is needed, and work out the motion ratio. Typically the rear corner frequency is 0.2 -0.4 Hz higher than the front, but can be less for RWD.

However for a circuit car the approach is different. There the ideal would be to use the softest rates possible that don't ground out.

If you've got aero then the rules are different again.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[nitan]

Thanks,

We are working a little more simply than that. We have alot of information on the basics, such as suspension motion, angles, ratios,etc, but the more indepth info we haven't gone into. What we have basically is the vehicle weight, spring rate of 1.67,( overall leverage ratio of the suspension arms, angles, etc.) and we want to find the pounds per inch for whatever spring will be a good starting point. We also know the shock info, such as oal, travel, and damping. The rest will be up to tweaking. This is an experimental car.

Any help is appreciated.

 

[GregLocock]

So, given the three scenarios above, which is it? That is, what is the intended usage of the vehicle?

And to be honest selecting a shock before the spring is a bit odd.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[nitan]

This is intended to be a road vehicle, with ability to go onto mild off road, gravel, sand, dirt roads, etc. We had a space consideration, and part of that is what determined our basic shock, but that is just the physical parameters. We have a weight of approximately 1100lbs, with a loaded weight of approximately 1300-1400lbs, and a 52r/48f weight bias. Is there a basic formula for calculating for the rate of spring that I need?

Thanks

 

[GregLocock]

You need to think about wheel travel, but, as a first guess design for 1.3 Hz at the front and 1.6 Hz at the rear.

So, working in SI, you've got 650 kg, split roughly between each wheel, so that is 160 kg. Less 30 kg unsprung, that is 130 kg

for the front wheel, 1.3=1/2/pi*sqrt(k/130)

k will be in N/m, you'll have to convert back to pounds/in. That is the rate at the wheel (wheel rate), you then use the usual rule to get the spring rate.

Now work out what the wheel force is at full jounce. Is that enough for your application? If not then you can think about jounce bumpers, spring aids, or non linear springs or stiffer springs.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[nitan]

I should be able to do something with that.

Thanks for your repeated help.

 

[apex944]

However for a circuit car the approach is different. There the ideal would be to use the softest rates possible that don't ground out.

This is where theory and practical application diverge. There are many race cars that I have helped design and/or worked with over the years that use substantially higher spring rates to work around a problem or to allow a driver to control the vehicle properly.

Some of the issues that require higher than theoretically optimal rates are:

Excessive roll, pitch or squat
This can cause issues with camber gain, toe changes, power delivery.

Roll or pitch rate.
A car can not negotiate a series of corners (chicane) if it is still recovering from the first turn .. Increasing the rate at which a chassis goes through a cycle can improve the driver's ability to deal with it. The same logic holds true for anywhere fast response required...be that cornering, braking etc.

Vehicle "balance"
Especially in FWD cars where the rear is lighter relative to the front. Often the rear of a FWD car is going to have a higher wheel rate (& freq.) than the front in order to help the car rotate and to help keep the front of the car "flat". Rear rates over 1000 lb/in are not uncommon.

A BMW factory designed 330i touring car we recently worked with came with 500# rear springs and 1600# (yes 1,600) fronts.

Now my question for those reading this is... how do you choose a wheel frequency for a race car? I know that modern damper technology has allowed the use of these high rate springs, but what is too high? Perhaps a new rule of thumb is to use the HIGHEST rate spring one can and still dampen it effectively ??

 

[GregLocock]

Yes, you are absolutely right, if you've got traction/braking problems that can't be solved any other way then use the other axle to hold the car level. None the less, the two professional race engineers I know are every much of the 'softer is better (so long as the aero is happy)' school.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Jrud]

I think it's hard to remain in the softer is better school because there are so many compromises with vehicle suspension. Depending on what you are starting with, ie production car turned track car, race car built from the ground up, etc, you will either have more or less boundary conditions to deal with.

One of the things that I've always had trouble with was the natural frequency ratio from front to rear. Two cars that I have been working a great deal with are opposite in their front and rear natural frequencies. One is a toyota supra and the other is a nissan 240sx. On the supra the rear has a higher natural frequency that than the front, but on the 240 the front is higher than the rear. Having them this way was the only way to get the TLLTD to work out the way I wanted without having to run too much of an anti-roll bar. I'm from the school of thought that doesn't like a big arb on a race car.

Now one of the reasons, that I didn't concern myself too much with the actual ratio of the front/rear natural frequency was these cars have a greater amount of damping than most, I'm running 70-80% of critical in both rebound and compression in the low speed range of the damper. So having a much lower damper natural frequency than it would if using less damping, I figured things would work out alright. I basically shot for natural frequencies in the 2-2.5hz range and tried to get the desired load transfer distribution and that was my baseline for both setups.

So how far would be safe to deviate from the rule of thumb on natural frequencies. I know that too stiff of a car is no good and neither is too soft. I mean just based on natural frequencies, I've designed setups for the 2-2.5Hz range for the 240 and the supra with no aero package and for the racetrack, but I've also done a FSAE car with frequencies much closer to 3Hz and also no aero package.

Am I putting too much stock into looking at natural freqencies and is still as much a good estimate as it used to be? Or are there just too many things to consider and that it less of a rule of thumb and more of just another variable to consider when putting together a suspension system?

Tim

 

[NormPeterson]

Since ride frequency incorporates both wheel rates and corner weights, it's a way of making a few general comparisons among vehicles that aren't all that similar otherwise. And it provides insight to the "flat ride" (pitch) performance, which ultimately affects contact patch loading.

But when it comes to transient handling, I'm not at all sure that ride frequency/frequencies - even damped ones - are easily related to any sort of frequency in roll and the time it takes for that to approach "equilibrium".

I know one autocrosser who has enjoyed considerable success over quite a few years by not running particularly high spring rates (and is adamantly opposed to them). And of others within that same competitive arena who have gotten faster every time they swapped in stiffer springs. Maybe a smoother driving style works better when mechanical grip is favored, and a point-and-shoot driver needs something stiffer?

Tim - similar front with relatively different rear geometric roll center heights between the two cars (or similar rears/different fronts)?


Norm

 

[Jrud]

Norm, the one car has a -4f/2r RCH and the other has 2f/4r RCH. The first one is a MacP strut car substantially lowered from it's designed height. And these are just estimates based on the information I have and was able to measure. I never made a true model for either of these suspensions because it just wasn't worth the time.

It is nice that you can compare two different cars with different suspensions, springs and weights with natural frequencies, but what does that really get you that you can't get through other, possibly more important, variables? I mean I am looking at this from the race car suspension side of things and I'm sure things are different in production cars.

Tim

 

[GregLocock]

Well it comes down to - what do you want. I'm just having some success optimising springs and shocks using a simple 2 dof model, as I have decided what I want as a result- in this case minimising the pitch angle and bounce amplitude after the first peak at 50 and 100 kph over 10, 25 and 50 mm steps, with various passenger/cargo loads in the vehicle.

For a race car that would not be an inappropriate goal.





Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[NormPeterson]

"It is nice that you can compare two different cars with different suspensions, springs and weights with natural frequencies, but what does that really get you that you can't get through other, possibly more important, variables? I mean I am looking at this from the race car suspension side of things and I'm sure things are different in production cars."

Perhaps for a competition car it's more useful as part of the logical approach than as a set of "hard" numbers to shoot for (assuming that other things in the ruleset aren't actually driving the design, a la NASCAR or certain SCCA autocross categories). By involving just one type of motion and one elastic component, it provides a starting point from which to iterate, and helps you decide what you're going to trade off as you go through it all. It's something that you can use to help make "judgement calls", assuming that you don't have access to a larger database than what most peoples' own personal experience represents. And even though the driver's perception of ride quality isn't perhaps very important during competition, ride and pitch as motions still have some effect on performance.


Norm

 

[Jrud]

Norm, that sounds like exactly what I've been doing with it. I start with natural frequencies, then look at roll rates and tlltd, and always check wheel travel/g and compare with camber and toe curves and keep tweaking it until I'm happy with how it looks on paper. And then real world testing until I'm actually happy with it.

I guess, thinking about all of this a little more, everything is just a guideline or essentially a starting point. There's no hard and fast rule for anything with suspensions.