Spring Rates and Split Front/Rear

[Rad_ Delgado_]

I have learned through different car manufacturers, race shops and my own experience to split the spring rates according to the static weight distribution (Front/Rear).

For instance on a car with 50/50 weight distribution, I would use a 800# spring at all corners (assuming that's the spring that works for that specific car). So, for another car with a 40/60 weight distribution and the same total weight as the previous example, I would use 640#/960# springs (or the closest available, 650/950).

Then, I would increase/decrease spring rates to balance the car according to testing. Let's say that the 50/50 car oversteers a lot, then I would test 900F/700R springs. In these test scenarios I leave the sway bars in a neutral position. I developed my old 2004 GT3 with this approach and the results were decent.

Once the car spring rates are balanced, I would adjust with sway bars according to course layouts, grip levels, tires, etc (lots of variables).

I have found this procedure quite consistent in bringing good results. Ideally, once I have found the proper spring rates I should get the shocks re-valved to these springs.

I would like to read some feedback about the approach you follow to determine spring rates.

[Marshall Grice]

You need to look at the lateral load transfer distribution. In general I've found that a good starting point is to make the lateral load transfer distribution the inverse of the weight bias. meaning if you have a 60/40 (nose heavy) weight distribution you want to start with roughly 60% of your roll stiffness in the rear. Then you stiffen the end opposite of the drive wheels a couple percent to help with corner exit power application.

so for a 50/50 weight distribution on a RWD car you would start at approximately 53% front load transfer distribution and work from there.

you can go a little further and calculate the stability of the car and adjust spring rates, toe settings and tire sizes to minimize the static margin. You can validate/measure the stability with the data acq and adjust the balance accordingly depending on track conditions.

[Jason Rhoades]

Marshall is right on the money, that's great tuning advice.

A concept sorta in the middle to help bridge the gap is "wheel rate", dictated by suspension geometry. Wheel rate is what determines how the springs are contributing to LLTD. A car might have 500lb. springs all around but vastly different wheel rates. In production cars, struts will get you closest to 1:1, so a 500lb. spring rate might be a 400 or 450lb. wheel rate. A multilink/SLA suspension could be close to that, to a whole lot less, generally depending on how far out on the LCA the spring/coilover mounts. Could be a 250lb. wheel rate, or less.

Other thing with wacky cars like 911s and most FWD cars - once you have a tire in the air, you have "saturated" the lateral load at that end, and increasing bar/spring at that end won't (generally) do anything more to change the balance of the car in its 3-wheeling state - you're down to alignments and tire pressures and whatnot. You can make the wheel lift sooner and higher by stiffening the saturated end, or softening the non-saturated end, and do the opposite (delay and decrease wheel lift) by doing the opposite.

All this stuff gives you a good starting point only, and you're bound to need to change them from there. There are also lots of other variables, like what alignment you can get, what size wheels/tires you have available, other weaknesses, etc. For example, the Viper has a really crap stock diff - in an effort to make its life easier, I have a huuuuge front bar (never seen a Viper lift an inside front wheel before mine) to help keep as much load as possible on the inside rear tire, to help keep it from spinning. The car would "handle" better with a more neutral setup but it wouldn't put power down as well - the setup best for pure handling isn't always what produces the best times.

[Jason Isley BS RX8]

I have learned through different car manufacturers, race shops and my own experience to split the spring rates according to the static weight distribution (Front/Rear).

For instance on a car with 50/50 weight distribution, I would use a 800# spring at all corners (assuming that's the spring that works for that specific car). So, for another car with a 40/60 weight distribution and the same total weight as the previous example, I would use 640#/960# springs (or the closest available, 650/950).

Then, I would increase/decrease spring rates to balance the car according to testing. Let's say that the 50/50 car oversteers a lot, then I would test 900F/700R springs. In these test scenarios I leave the sway bars in a neutral position. I developed my old 2004 GT3 with this approach and the results were decent.

Once the car spring rates are balanced, I would adjust with sway bars according to course layouts, grip levels, tires, etc (lots of variables).

I have found this procedure quite consistent in bringing good results. Ideally, once I have found the proper spring rates I should get the shocks re-valved to these springs.

I would like to read some feedback about the approach you follow to determine spring rates.

The Fiat did not come with an engineer and crew? :unimpressed:

[Steve Ekstrand]

Fly. Wall. Digital recorder light shining bright.

Gentleman keep talking.

[John Coffey]

Basing spring rate on corner weights without taking into account the actual wheel rate is a bad starting point. It will work OK on a four strut car but on a multi-link, live axle, beam, etc. you're only lucky if you get close. You also want to avoid springing the car so the front and rear suspensions are at the same frequency which can lead to pitch oscillation.

[Marshall Grice]

Fly. Wall. Digital recorder light shining bright.

Gentleman keep talking.

about what? that's the end of picking spring rates.

ok i guess that's not true. you still need to set goals like what your target roll gradient is and your desired natural frequencies. That allows you to size your sway bars. But then that's the end.

[Aaron Goldsmith]

Fly. Wall. Digital recorder light shining bright.

Gentleman keep talking.

about what? that's the end of picking spring rates.

ok i guess that's not true. you still need to set goals like what your target roll gradient is and your desired natural frequencies. That allows you to size your sway bars. But then that's the end.

Haha, yeah.. turns out an excel spreadsheet can get you 95% there.

[Mako Koiwai]

You engineer types should have an addition Index applied to your and/or your car's results!

signed ... Jealous and Impressed

[Sebastian Rios]

[quote="Aaron Goldsmith"

Haha, yeah.. turns out an excel spreadsheet can get you 95% there.[/quote]
Really! I'm good at Excel...just really bad at tuning suspension.

Send me a spreadsheet!

[Doug Rowse]

Here's the frequency route laid out with equations and explanations. *Math warning*. It also includes bars and shocks.

http://www.optimumg.com/OptimumGWebSite ... hTips.html

If you've got some time there's a pretty smart guy that's going to be in your area next week giving a seminar on exactly this topic, though it's expensive for non-students.

http://www.optimumg.com/OptimumGWebSite ... minar.html

[Will Kalman]

Another thing about spring rates: You need rates that keep you off the bump stops in most normal corners (it's expected that you'll hit them in hard transient loading like a bumpy corner but you should really never truly bottom out on them).

[Marshall Grice]

Another thing about spring rates: You need rates that keep you off the bump stops in most normal corners (it's expected that you'll hit them in hard transient loading like a bumpy corner but you should really never truly bottom out on them).

It's a small difference but, I would say you need to set your ride height such that you stay off the bump stops with the springs you've chosen.

[Jonathan Lugod]

Another thing about spring rates: You need rates that keep you off the bump stops in most normal corners (it's expected that you'll hit them in hard transient loading like a bumpy corner but you should really never truly bottom out on them).

It's a small difference but, I would say you need to set your ride height such that you stay off the bump stops with the springs you've chosen.

i concur... i also have been taught that you should only really go as soft as you can get away with in regards to spring rates.

[Mako Koiwai]

Brian Peters had us raise our Miata a bit, plus further cut-down our bump stops, to get off of our BS's ... and it really helped. Now I'm only 1.9 seconds behind Jayson W and Ken Lord!

[Rad_ Delgado_]

Great tip on stiffening the end opposite to the drive wheels, to put power down better, I take advantage of this. Typically on a RWD car, you get double benefit, you put the power down better, and the steering becomes more crisp.

Marshall, interesting application of opposite static loads against spring rates, but it seems to me that doing that makes the body flex (twist) under lateral loads.

JasonR, I would use the Fiat as an example, it has a 0.7 wheel ratio (info I got from three different sources). It comes with way stiffer springs than my GT3 RS, but it rides softer and it moves much more. The Fiat uses 475/650 springs the RS 225/580, my old GT3 225/550. Both manufacturers Fiat and Porsche use springs proportionally stiff to the corresponding axle weight on their street cars.

The wheel rate would play in choosing the spring rates split, if the car has different wheel rates front and rear. In the case of Fiat and Porsche, the wheel rates are consistent front and rear. However, in the Porsche it's close to 1, in the Fiat just a 0.7.

These same two manufacturers have racing versions of their street cars (GT3 Cup and 355/360/430 Challenge). Interesting enough, I have the spring rates from a multitude of GT3 Cups and the 4 spring sets for the 360 & 430 Challenge cars. The spring rates split don't follow the static weight distribution for racing applications, both factories actually put stiffer front/rear springs, but they add even more spring to the front axles. For the example described, a F430 Challenge uses 1050#/1400# or 1200#/1650#. The latest GT3 Cup uses 1500#/1550#. The race cars have different weight distribution as well, better in the Porsche, worse in the Fiat compared to the street cars.

I ran a small exercise with the Fiat last week in Blytheville, I actually replaced the stock springs with 1100#/1200# and set an alignment that promotes less understeer. Well, everything worked well and according to my expectations, the car was still pushing. I had the privilege to co-drive with Matthew Braun on the 20 secs practice course on Friday, we played with tire pressures to defeat the horrible understeer, and it got better, but still no rotation, plenty of push, but much better than the setups I have driven in the Fiat so far. In this case I used the 45/55 weight distribution with driver (it's 43/57 without driver), but shifted another 2% to the front looking for a 47/53 split. I chose the rates based on the wheel rate and measured suspension travel with the stock suspension, I intended no more than 25mm of suspension travel, and got 1 1/4" at Blytheville. I'm moving to even stiffer springs from another set I got 1400#/1500#, interesting fact that these ones match the 3% extra split Marshall mentioned.

No, the FIAT did not come with engineers and a crew. The last funny episode was when I ordered camber shims (it uses them at all corners), the parts manager asked me why I was ordering so many of them, and I told him I needed to test different alignment settings, his answer "your car already has the perfect alignment, it was track tested in Fiorano...". I told him that the car understeers a lot, and he spoke to a sales associate that was kind enough to explain me the Fiat Driving Experience (Driving School), they highly recommended that to me to fix the understeer.

[Jason Uyeda]

Great tip on stiffening the end opposite to the drive wheels, to put power down better, I take advantage of this. Typically on a RWD car, you get double benefit, you put the power down better, and the steering becomes more crisp.

Marshall, interesting application of opposite static loads against spring rates, but it seems to me that doing that makes the body flex (twist) under lateral loads.

JasonR, I would use the Fiat as an example, it has a 0.7 wheel ratio (info I got from three different sources). It comes with way stiffer springs than my GT3 RS, but it rides softer and it moves much more. The Fiat uses 475/650 springs the RS 225/580, my old GT3 225/550. Both manufacturers Fiat and Porsche use springs proportionally stiff to the corresponding axle weight on their street cars.

The wheel rate would play in choosing the spring rates split, if the car has different wheel rates front and rear. In the case of Fiat and Porsche, the wheel rates are consistent front and rear. However, in the Porsche it's close to 1, in the Fiat just a 0.7.

These same two manufacturers have racing versions of their street cars (GT3 Cup and 355/360/430 Challenge). Interesting enough, I have the spring rates from a multitude of GT3 Cups and the 4 spring sets for the 360 & 430 Challenge cars. The spring rates split don't follow the static weight distribution for racing applications, both factories actually put stiffer front/rear springs, but they add even more spring to the front axles. For the example described, a F430 Challenge uses 1050#/1400# or 1200#/1650#. The latest GT3 Cup uses 1500#/1550#. The race cars have different weight distribution as well, better in the Porsche, worse in the Fiat compared to the street cars.

I ran a small exercise with the Fiat last week in Blytheville, I actually replaced the stock springs with 1100#/1200# and set an alignment that promotes less understeer. Well, everything worked well and according to my expectations, the car was still pushing. I had the privilege to co-drive with Matthew Braun on the 20 secs practice course on Friday, we played with tire pressures to defeat the horrible understeer, and it got better, but still no rotation, plenty of push, but much better than the setups I have driven in the Fiat so far. In this case I used the 45/55 weight distribution with driver (it's 43/57 without driver), but shifted another 2% to the front looking for a 47/53 split. I chose the rates based on the wheel rate and measured suspension travel with the stock suspension, I intended no more than 25mm of suspension travel, and got 1 1/4" at Blytheville. I'm moving to even stiffer springs from another set I got 1400#/1500#, interesting fact that these ones match the 3% extra split Marshall mentioned.

No, the FIAT did not come with engineers and a crew. The last funny episode was when I ordered camber shims (it uses them at all corners), the parts manager asked me why I was ordering so many of them, and I told him I needed to test different alignment settings, his answer "your car already has the perfect alignment, it was track tested in Fiorano...". I told him that the car understeers a lot, and he spoke to a sales associate that was kind enough to explain me the Fiat Driving Experience (Driving School), they highly recommended that to me to fix the understeer.

Dude! Whateves... Just find a way to maximise in 3-runs... Braun = God (just ask KMo) but he likes push That car blows, you should go back to the RS

[Rad_ Delgado_]

Dude! Whateves... Just find a way to maximise in 3-runs... Braun = God (just ask KMo) but he likes push That car blows, you should go back to the RS

Actually I was showing people the difference between me hitting cones and God hitting cones. I rub them with the car sides (front, doors, quarter panels). God runs them over and hard. God likes push, but my Fiat has too much push for God. God is great setting up a car.

I placed an order for another RS before selling my old one, I'm going back to the RS, and yes the Fiat blows big time, but I like it.

[Marshall Grice]

Marshall, interesting application of opposite static loads against spring rates, but it seems to me that doing that makes the body flex (twist) under lateral loads.

it will twist more, which is why chassis torsional stiffness is so important.

JasonR, I would use the Fiat as an example, it has a 0.7 wheel ratio (info I got from three different sources). It comes with way stiffer springs than my GT3 RS, but it rides softer and it moves much more. The Fiat uses 475/650 springs the RS 225/580, my old GT3 225/550. Both manufacturers Fiat and Porsche use springs proportionally stiff to the corresponding axle weight on their street cars.

when you apply the .7 motion ratio to the 475 lb spring in the fiat you get a 232 lb/in wheel rate. Depending on how the corner weight of the fiat compares with that of the RS that wheel rate could very well be softer then the 225 lb spring. The parameter of interest is the natural frequency.

These same two manufacturers have racing versions of their street cars (GT3 Cup and 355/360/430 Challenge). Interesting enough, I have the spring rates from a multitude of GT3 Cups and the 4 spring sets for the 360 & 430 Challenge cars. The spring rates split don't follow the static weight distribution for racing applications, both factories actually put stiffer front/rear springs, but they add even more spring to the front axles. For the example described, a F430 Challenge uses 1050#/1400# or 1200#/1650#. The latest GT3 Cup uses 1500#/1550#. The race cars have different weight distribution as well, better in the Porsche, worse in the Fiat compared to the street cars.

this agrees with what I said earlier. rear heavy cars are going to tend towards front stiff. Even though the actual spring rates are a lower number in the front I'd be willing to bet that when you analyze what their effect is on the roll stiffness you'll find they (the fronts) carry most of the overall roll stiffness.

[Jason Rhoades]

Not too likely Porsche motion ratios equal front and rear - with struts, probably close to 1 in front but most likely .6-.7 in the rear with the multilink.

Track car setups are often based around maximizing stability and aero contribution at track speeds, at the expense of low-speed mechanical grip and balance. Not necessarily a bad starting point, but probably not optimal for cone-dodging either.

Similarly, through rules or optimization, RWD track cars are likely to have a great proportion of rear tire to front tire. We need big yaw in autocross, and a great way to get it is what I'd recommend to you to cure the push - more front wheel and tire. This is assuming the alignment, spring/bar rates, and ride heights are in the ballpark.

[Marshall Grice]

so just taking some wild guesses with internet based numbers. lets say your car weighs 2975 in race trim (you in it) and has a 57% rear weight bias, and 100lbs unsprung mass.

3hz front spring rate is about 950#, 3hz rear spring rate is about 1400#. I wouldn't really try going much higher then that.


and assuming a whole bunch of other things, like equal track widths, etc, with 1000/1400 springs in the car you'll be near 50/50 roll stiffness distribution. on a car that is 57% rear heavy it will be LOOSE! so you put bigger tires in the rear, it's less loose and maybe even perfectly balanced if you have tire widths in proportion to your weight bias (ie ~30% wider in the rear). start throwing in sway bars...?

[Jason Uyeda]

so just taking some wild guesses with internet based numbers. lets say your car weighs 2975 in race trim (you in it) and has a 57% rear weight bias, and 100lbs unsprung mass.

3hz front spring rate is about 950#, 3hz rear spring rate is about 1400#. I wouldn't really try going much higher then that.


and assuming a whole bunch of other things, like equal track widths, etc, with 1000/1400 springs in the car you'll be near 50/50 roll stiffness distribution. on a car that is 57% rear heavy it will be LOOSE! so you put bigger tires in the rear, it's less loose and maybe even perfectly balanced if you have tire widths in proportion to your weight bias (ie ~30% wider in the rear). start throwing in sway bars...?

No, no, no... Just run a Lotus, It's marginally more reliable than a Fiat

[Rad_ Delgado_]

No, no, no... Just run a Lotus, It's marginally more reliable than a Fiat

That's not a Lotus, it is a Toyota with a Lotus body kit, seriously. Yes, Toyotas are damn reliable.

[Bill Martin]

I posted this on a RallyCross forum, hence the reference to rally suspensions. But some of you might find it interesting:
-----------------------------------------------------------------
http://www.rqriley.com/suspensn.htm" onclick="window.open(this.href);return false;

I was very surprised to find so much technical information in a single article. Of particular interest is his discussion of suspension natural frequency. This was mentioned at last years Ridgecrest rally school but no one knew how to calculate it.

I had forgotten how simple the measurement is. Measure your static deflection by jacking up the car until the spring just unloads. Measure that distance (SD) and plug it into this formula:

f = 3.13/(sq-rt SD) hertz

for example a 4" static deflection will give you 1.5 hz. That would be towards the soft end of a rally suspension. A comfy to slightly firm street ride.

3" SD gives you 1.8 Hz. Medium rally suspension

2" SD gives you 2.2 Hz. Bordering on too high for dirt but good for tarmac.

1" SD gives you 3.13 Hz. Too stiff even for track.

For those with not the slightest interest in techno-babble, skip the article.

-- Bill

Note, if something limits your droop before the spring unloads, you'd need to disconnect it to use this method.