How to select/compute spring rates for any car

[David Pozzi]

I'm starting this to get some tech discussion going on how to choose a good spring rate for a car.
It usually involves determining what frequency you want at the wheel, and calculating the spring rate that will provide it.

Here is a page on spring frequency and a spreadsheet to try out:
http://www.geocities.com/bry593/spri...teselector.htm

I'd appreciate any info on what frequency in cycles per second is desireable for:
1. mild street 1.25 cps?
2. high perf street 1.4 to 1.5 cps?
3. road course uses 1.6 to 1.7 cps?

I'd like to see some discussion on how to calculate the proper spring frequency/rate needed, and your comments on the above spreadsheet. Is it useful or even correct?
Thanks, David

[dhardison]

David, That URL does not work.

Dan

[nancejd]

Does this work?

http://www.geocities.com/bry593/spri...teselector.htm

[David Pozzi]

Dan,
try it now, I think I fixed it.

[dhardison]

Yep, sure does! Thanks.

Dan

[nancejd]

Interesting that rear spring rate decreases as speed increases, but the front srping rate stays unchanged. I would think the opposite would be true, that you would need a stiffer spring to control an increase in road (jounce rebound) frequency, and that consequently you would need a stiffer shock.

[David Pozzi]

James,
The concern is the front and rear suspension hitting a bump in quick succession and reacting to that bump causing a front to rear pitching vibration of the chassis. The faster you go, the higher the frequency input to the suspension. The trick is to keep the front and rear from vibrating in tune with each other. The rear is probably going to recieve a higher added load either from passengers or load in the trunk or even possibly a trailer, so a higher rate than the front was recomended in the web page.

I havent done much of this type spring selection, so I need to read up on it more. I've usually taken a proven spring rate compensated for the motion ratios.

One error I see on that web page is that he used the full spring rate, his car has McPherson strut suspension and although the coils sit right on the strut , they are on an angle and there should be a motion ratio reduction for that in his calculations.

[todd8541]

Ah, I think this should be retitled and rewritten to try and determine the "Wheel Rate" not "Spring rate". Just because two cars have the same wt (both sprung and unsprung) doesn't mean they need the same "Spring". Someone can change the spring location on the same car and it would require a different spring to maintain the same handling characteristics.

Todd

[David Pozzi]

Todd,
Yes I thought about that but really I'd like this discussion to include calculating the spring rate to achieve the desired wheel Hz.

I modified the spreadsheet from the above link to calculate the spring required for a given wheel HZ, and it uses some motion ratios if you have the needed A arm dimensions to input in to the spreadsheet. I was hoping someone with better math and spreadsheet skills would come by and do it, but that hasn't happened.
I need to finish it up and will upload it for you guys to try out if there is any interest.

[Chevy350]

I'm interested, and I think it could help others as well.

[David Pozzi]

My thought is, we could work out what frequency in Hz is ideal for:

(Quote from my first post)
"I'd appreciate any info on what frequency in cycles per second is desireable for:
1. mild street 1.25 cps?
2. high perf street 1.4 to 1.5 cps?
3. road course uses 1.6 to 1.7 cps?"

A member could select what his use is, enter the frequency in the spreadsheet along with motion ratio info, the spreadsheet would calculate the spring rate needed.

There may also be a way that would predict suspension travel for a given Hz, so if your car is lowered and had limited travel, you could select a higher Hz setup.

[Damn True]

Ouch. This makes my head hurt.

When the time comes I'll hit you up for help.

[Norm Peterson]

near-duplicate post

[Norm Peterson]

I was hoping someone . . . would come by and do it, but that hasn't happened.
I need to finish it up and will upload it for you guys to try out if there is any interest.

I'd worked through the most recent version of a FlatRide spreadsheet about a year ago, and didn't see this thread until just a few minutes ago. Give me a little time to go through yours, see how it compares/what it's intentions are. Mine almost certainly requires more inputs, and I'm sure has a couple of additional outputs.

I've occasionally considered offering it for sale (for beer money kind of $ to help defray the development time - suggestion of which I hope doesn't get me busted) if I could get past a couple of obstacles like guaranteeing the integrity of the equations. Based on some professional experience in developing engineering spreadsheets, I can tell you that people can get VERY creative when it comes to screwing them up. Perhaps I could let a "FlatRideLite" version go as a teaser "freebie".

Norm

[David Pozzi]

Norm,
I'm all for it if we can work something out. I'd kick in some cash to get something useable going. Perhaps larry could sell the spread sheets on the site thru his PT store.
I'm a hack at spreadsheets and not good at the math.

I took the spreadsheet and added inputs to calulate motion ratios for the car, and started on calculations to do the rear rates, but with a stick rear axle there is one motion ratio for a two wheel bump, another for a one wheel bump. Using the spreadshet formulas I'm getting very high recomended rates for the rear spring using the one wheel bump motion ratio method.

I haven't had much time to work on it since I've got other things going on.

[Norm Peterson]

James - understand that his analysis is only applying the user-specified frequency to the front suspension. The rear frequency is always higher, but as speed rises it will drop. That's where the 'intended usage' frequency numbers such as those in the original post come in. This doesn't mean that you can't use a 'comfort' frequency for ride quality at 125 mph, only that the resulting springs would be different than if you'd picked a more hard-core frequency.

Dave - I think my sheet is more along the approach that you've normally taken. But there isn't much effort involved to add the option of forcing solutions for spring rates based on user-specified frequencies. There are at least a couple of ways to handle this, but since it isn't possible to quantify things like individual ride preferences and tolerance for NVH I'd suggest letting the user specify his own target frequency from a table such as you've already provided, perhaps with a note suggesting the choice of a slightly lower frequency if ride concerns are particularly important.

I think there are more than a few auto-x and open-track cars running well above 1.7 Hz (I specifically know of one up around 1.9 f/2.4 r Hz).

I think this sort of analysis is best limited to the two-wheel bump mode; otherwise you're getting into roll effects and would need to include sta-bar contributions as well. It's not that you can't be slightly 'off' in your spring selections and still have an acceptable ride at the target speed.

Suspension type certainly enters into this, as you generally need to spring a McStrut stiffer than a control arm suspension for reasons of camber control.

After all that, it really gets interesting when you throw some front and rear damping terms into the mix and generate a plot. Note that damping helps "cover" for slight differences in available spring rates vs theoretical ones and nonlinearities in motion ratios.

Norm

[Norm Peterson]

Dave - this is what it looks like (I wasn't at all sure that it would upload in sufficient clarity). There is quite a bit of stuff that could be moved off-screen, which would free up some real estate for inputting a desired frequency (could set it up for either front or rear) or controls to run other iterative solutions. It wouldn't save much file size to delete the optional inputs, though.

Norm

[David Pozzi]

Norm,
That's quite sophisticated, I was thinking of something a lot more basic, something that would come up with just a front and rear spring rate for a given wheel frequency F/R.

The shock damping calcs etc are great, but beyond anything I've considered trying to work on, and probably above most of us here.

[Norm Peterson]

I agree that there are more input variables in that sheet than you need for a quick and dirty estimate. More than I would use if I was doing this with a slide rule and pencil/paper, actually. It's just that a spreadsheet makes it much easier to write the formulas once and avoids having to explain away the more obvious stuff that you didn't consider. And it provides a better record of what conditions were being evaluated. FWIW, all of the inputs and most of the outputs are defined/described on the 'HELP' page.

Granted that all of that stuff in the

ADDITIONAL INPUTS FOR ITERATIVE ESTIMATE OF "PitchKpsquared/ab" (OPTIONAL)

section really isn't necessary even though it results in some additional understanding of the overall motion. Truth be told, I haven't used that part all that much past crunching the formulas.

However, I do think you need to consider damping even for 'simplified' evaluations. Ignoring damping terms is a legitimate approach when the damping is a very small percent of critical damping (as it is for structural damping, where low-single-digit percentages of critical are typical). Or if you only want to investigate the condition where the shocks are almost completely shot.

But "proper" suspension damping involves much higher damping, 15% or so when ride quality has priority up to perhaps 45% for best performance. With that much damping, not only will the magnitude of pitch motion as the suspensions reach one full cycle be affected, but the theoretical "Flat Ride" speed or rear spring selection will also be skewed a bit by the slight frequency shifts involved.

Note that including damping terms allows you to investigate the change in ride (specifically pitch motion) as the shocks do deteriorate. A 15 mph difference between the damped and undamped flat ride speeds is entirely possible just by fitting new performance shocks to one end and leaving the dead players or even good ride comfort shocks on the other - with no change in the springs.

Basically, I'm suggesting that the undamped approach may be a little too basic for a sheet that could be used to look at an unknown variety of situations. There seems to be up to about a 10% difference in Flat Ride speeds, damped vs undamped, when the same percent damping is applied to both ends. That's enough to better suggest which spring to choose, given that there may be two or more that have rates that are close.

Norm

[David Pozzi]

Norm,
You are right as usual.
I guess it was wishful thinking that it could be more simple.

Your help files probably need to list some typical damping values to enter, for those of us who have not done this before.

I'm hoping the forum can generate some Roll center height, CG and motion ratio data to help get a good baseline to start people out.

I sent Larry a PM asking him to help make your spreadsheets available, for whatever fee you need, possibly through his PT.com store.
David