DG-FNR
20+ Year Contributor
- 234
- 62
- Oct 21, 2002
-
Geary,
NB, Canada
Alright, let's talk springs for a little bit. I'm a little hamstrung on this topic, because there are things I am keeping to myself in order to retain a competitive advantage, but there are some things I can discuss openly.
When determining spring rates for a vehicle that is primarily street driven, you need to consider the following things:
1) The motion ratios of the front and rear suspension (more on that in a second)
2) The maximum corner weights of the car as it is expected to be driven.
3) The amount of maximum suspension travel at the fully loaded ride height.
The aim of the exercise here is to make sure the suspension never bottoms out.
The "motion ratio" is an expression of how much mechanical advantage the wheel has over the spring. Car springs are not mounted directly over the wheel like a motorbike front fork; they are offset some distance inboard of the wheel, and often canted at some angle relative the the motion centre of the suspension arm. The effect is much like a wheelbarrow, where the car's wheel had ahold of the "handles" on the wheelbarrow, the spring is mounted in the "bucket", and the suspension pivot point is the "wheel".
The farther towards the "wheel" you move the "bucket", the easier it is to lift the load - which in this case, is the spring - and so the less force the spring can exert on the suspension to resist motion.
Specifying the exact instant motion ratio requires 3D gemometry software and some calculus, but on a 2G DSM it can be approximated at about 0.75 on the front, and 0.86 on the rear.
That being the case, we immediately know that we're going to need more spring on the front in order to match the rear.
Now we need to know the weights on each corner. On a race car, we corner-weigh the car all the time - before and after each race - but on a street car we can afford to be a little bit more approximate. DSMs are nose-heavy, on the order of 60% front, 40% rear. Depending on the individual car and how it is loaded, that can move around a couple of percentage points, but 60/40 is a reasonable approximation. Figure the car at 3350lbs fully loaded, 60/40 front split, and 50/50 left/right split, and you get 1000lbs on each front wheel and 675 lbs on each rear wheel - more or less.
Now we need to know static ride height. This is going to vary by individual taste and by how stiff of a vehicle you can tolerate - the lower the car gets, the stiffer it needs to be. 2 inches is a reasonable start point.
Now, a trick.
When the car rolls, weight transfers from the inside wheel to the outside wheel. At 100% weight transfer, the inside wheel comes off the ground and all the weight of that end of the car is on the outside wheel. This is unlikely to ever happen on a street car (not enough grip, for one) but if we pretend that it could, we want to make sure that the car doesn't run out of suspension travel before it hits 100% lateral weight transfer. So we can use the corner weight over the wheel travel to determine the wheel rate we need.
So for the front, 1000lbs/2in = 500lbs/in, and for the rear 675lbs/2in = 338lbs/in
That gets us wheel rate, but we're trying to pick springs here, so we have to account for motion ratio.
Wheel rate = spring rate * (MR^2) so that works out to 889 lbs/in front and 457 lbs/in rear
Now this is a pretty good guess for a car that has no roll bars, but we have roll bars, so we need to soften things up a little bit. If we assume that half the roll resistance comes from the bars, we can try 444lbs/in up front, and 228 lbs/in in the rear. Rounded up to springs we can actually buy, that'd be 450 lbs/in front, 250 lbs/in rear
And at this point I cheat, and I look at my little secret thing, and I confirm that's a pretty reasonable number. This is NOT the same process I use to choose springs for race cars, but I can cross-reference my rule-of-thumbing against that process and see that I'm within reasonable tolerences.
As it turns out, running roughly twice as much front spring as rear spring works out to be a pretty good rule of thumb. A heavier, lower car needs more spring, a lighter, higher car needs less spring, and the actual weight distribuition changes the ratio, but that ratio of 2:1 is in the right ballpark for most cases.
As the car moves more towards pure competition use, and less towards street use, you can get more agressive on the spring rate. The problem is that you'll need much better shocks to control those springs than are commonly availible.
The issue here is that a compressed spring stores energy. The purpose of the shock is to dissipate that energy. If the spring stores more energy than the shock can deal with, you get that "bouncy bouncy bouncy" thing going. A properly matched spring and shock package rides almost as nicely as OEM, even if the springs are way stiff.
Keep in mind though that all this stuff is approximation. You will get MUCH better results by weighing your own car, measuring your own suspension travel, etc. There is no "one size fits all" spring rate.
Which brings us to "lowering springs".
A "lowering spring" is just a normal spring with different free length and rate than OEM. It is typically shorter than OEM, so the car sits lower, and the rate is normally higher, because it takes more rate to deal with the reduced wheel travel.
Let's look at that quickly - if you have 1000lbs of additional load on a corner and 2" of suspension travel, then you'll need a wheel rate of 500lbs/in to keep from bottoming. Cut that travel down to 1", and you need a wheel rate of 1000lbs/in to keep from bottoming.
ADVANCED - reducing travel normally implies lowering the CG, which in turn implies less lateral load transfer for similar cornering force, so the additional load will be some value less than the 1000lbs with the higher CG, meaning the spring rate increase will be somewhat less as well.
OK, so when you buy a "lowering spring", you are taking a gamble on two things:
1) The free height designed into the spring is going to generate the ride height you want.
2) The spring rate is going to be sufficiant to keep the car from bottoming out.
In practice, these two things almost NEVER work out. The ride height is either too high or too low, and the spring rate is almost always way too low (mostly because I suspect the springs are developed with OEM shocks in mind) As well, because lowering springs tend to be designed for a slightly different demographic than real racers, the quality on the springs themselves is often low as well.
To get around this, use coilovers. That allows you to use the high-quality racing springs, plus you get post-install ride height adjustment. ADVANCED: you also get the ability to corner-weight the car.
In racing, the tolerence level on free height on springs is 0.1" The only springs good enough to meet this is Hypercoils. Luckily, economies of scale work in our favour, and Hypercoils are cheap. You can use the exact same quality springs as Real Racers, and there is NO downside.
Now let's talk about balance.
First, we have to all understand one basic fact about DSMs - they understeer under power. NOTHING you can EVER do is EVER going to fix that - save one thing, and that doesn't exist .
The centre diff on a DSM is a 50/50 split. That means it is always going to try and send power forward to the front wheels in the same proportion that it does to the rear wheels. But longnitudnal weight transfer ensures that, under power, weight is going to transfer off the fronts and onto the rears. That means front grip is going to DECREASE, and rear grip is going to INCREASE.
That means power-on understeer. You can play around with how the car behaves at the understeer limit by changing diff types around, but ultimately, a 50/50 split centre diff means UNDERSTEER.
The only real way to get around this is to build a centre diff that has a rear-biased torque split, maybe 40/60. That would send more power rearward than forward, and if the bias was right, the rears would reach their limit before the fronts, and you'd have power-on oversteer, just like a RWD does. I think the WRX STi has one of these. Such a thing does not exist for DSMs, so we get to live with the understeer.
The common "solution" to the DSM understeer issue is to bias the roll resistance heavily to the rear with great big heavy springs and monster thick rear bars. This generates GOBS of lift-throttle and steady-state oversteer, but will never truly generate power-on oversteer - at least not for any reasonably sticky tire. It does, however, give away tons of overall grip. The end result is a car that wants to go around on lift throttle, dances around a lot, and is slow overall.
The better solution is to work with the car instead of against it. You still want a little bit of rearward roll resistance bias (so the car will still rotate on lift throttle, just not as abruptly) but instead balance the roll resistance more in line with the car's static weight distribution.
Conventional wisdom suggests that this would make an understeering car worse. Conventional wisdom is wrong. By doing this, what you are doing is correcting for excessive rearward roll resistance bias, and the car picks up grip. The trick is separating rthe car's power-on characteristics (which are always going to be understeer) from it's power-off characteristics.
What this works out to in practice is that 450lbs of rear spring is pretty much the absolute maximum, and the RM rear bar is pretty much as big as you'll ever want to go. I'm actually softer than that (400lbs of spring plus the OEM rear bar) but I'm also lighter (2804lbs w/o driver)
The rule of thumb is stiffen the front before the rear, and learn to drive around the power-on understeer. Good diffs (Quaife centre, then Quaife front) also help a LOT.
DG
When determining spring rates for a vehicle that is primarily street driven, you need to consider the following things:
1) The motion ratios of the front and rear suspension (more on that in a second)
2) The maximum corner weights of the car as it is expected to be driven.
3) The amount of maximum suspension travel at the fully loaded ride height.
The aim of the exercise here is to make sure the suspension never bottoms out.
The "motion ratio" is an expression of how much mechanical advantage the wheel has over the spring. Car springs are not mounted directly over the wheel like a motorbike front fork; they are offset some distance inboard of the wheel, and often canted at some angle relative the the motion centre of the suspension arm. The effect is much like a wheelbarrow, where the car's wheel had ahold of the "handles" on the wheelbarrow, the spring is mounted in the "bucket", and the suspension pivot point is the "wheel".
The farther towards the "wheel" you move the "bucket", the easier it is to lift the load - which in this case, is the spring - and so the less force the spring can exert on the suspension to resist motion.
Specifying the exact instant motion ratio requires 3D gemometry software and some calculus, but on a 2G DSM it can be approximated at about 0.75 on the front, and 0.86 on the rear.
That being the case, we immediately know that we're going to need more spring on the front in order to match the rear.
Now we need to know the weights on each corner. On a race car, we corner-weigh the car all the time - before and after each race - but on a street car we can afford to be a little bit more approximate. DSMs are nose-heavy, on the order of 60% front, 40% rear. Depending on the individual car and how it is loaded, that can move around a couple of percentage points, but 60/40 is a reasonable approximation. Figure the car at 3350lbs fully loaded, 60/40 front split, and 50/50 left/right split, and you get 1000lbs on each front wheel and 675 lbs on each rear wheel - more or less.
Now we need to know static ride height. This is going to vary by individual taste and by how stiff of a vehicle you can tolerate - the lower the car gets, the stiffer it needs to be. 2 inches is a reasonable start point.
Now, a trick.
When the car rolls, weight transfers from the inside wheel to the outside wheel. At 100% weight transfer, the inside wheel comes off the ground and all the weight of that end of the car is on the outside wheel. This is unlikely to ever happen on a street car (not enough grip, for one) but if we pretend that it could, we want to make sure that the car doesn't run out of suspension travel before it hits 100% lateral weight transfer. So we can use the corner weight over the wheel travel to determine the wheel rate we need.
So for the front, 1000lbs/2in = 500lbs/in, and for the rear 675lbs/2in = 338lbs/in
That gets us wheel rate, but we're trying to pick springs here, so we have to account for motion ratio.
Wheel rate = spring rate * (MR^2) so that works out to 889 lbs/in front and 457 lbs/in rear
Now this is a pretty good guess for a car that has no roll bars, but we have roll bars, so we need to soften things up a little bit. If we assume that half the roll resistance comes from the bars, we can try 444lbs/in up front, and 228 lbs/in in the rear. Rounded up to springs we can actually buy, that'd be 450 lbs/in front, 250 lbs/in rear
And at this point I cheat, and I look at my little secret thing, and I confirm that's a pretty reasonable number. This is NOT the same process I use to choose springs for race cars, but I can cross-reference my rule-of-thumbing against that process and see that I'm within reasonable tolerences.
As it turns out, running roughly twice as much front spring as rear spring works out to be a pretty good rule of thumb. A heavier, lower car needs more spring, a lighter, higher car needs less spring, and the actual weight distribuition changes the ratio, but that ratio of 2:1 is in the right ballpark for most cases.
As the car moves more towards pure competition use, and less towards street use, you can get more agressive on the spring rate. The problem is that you'll need much better shocks to control those springs than are commonly availible.
The issue here is that a compressed spring stores energy. The purpose of the shock is to dissipate that energy. If the spring stores more energy than the shock can deal with, you get that "bouncy bouncy bouncy" thing going. A properly matched spring and shock package rides almost as nicely as OEM, even if the springs are way stiff.
Keep in mind though that all this stuff is approximation. You will get MUCH better results by weighing your own car, measuring your own suspension travel, etc. There is no "one size fits all" spring rate.
Which brings us to "lowering springs".
A "lowering spring" is just a normal spring with different free length and rate than OEM. It is typically shorter than OEM, so the car sits lower, and the rate is normally higher, because it takes more rate to deal with the reduced wheel travel.
Let's look at that quickly - if you have 1000lbs of additional load on a corner and 2" of suspension travel, then you'll need a wheel rate of 500lbs/in to keep from bottoming. Cut that travel down to 1", and you need a wheel rate of 1000lbs/in to keep from bottoming.
ADVANCED - reducing travel normally implies lowering the CG, which in turn implies less lateral load transfer for similar cornering force, so the additional load will be some value less than the 1000lbs with the higher CG, meaning the spring rate increase will be somewhat less as well.
OK, so when you buy a "lowering spring", you are taking a gamble on two things:
1) The free height designed into the spring is going to generate the ride height you want.
2) The spring rate is going to be sufficiant to keep the car from bottoming out.
In practice, these two things almost NEVER work out. The ride height is either too high or too low, and the spring rate is almost always way too low (mostly because I suspect the springs are developed with OEM shocks in mind) As well, because lowering springs tend to be designed for a slightly different demographic than real racers, the quality on the springs themselves is often low as well.
To get around this, use coilovers. That allows you to use the high-quality racing springs, plus you get post-install ride height adjustment. ADVANCED: you also get the ability to corner-weight the car.
In racing, the tolerence level on free height on springs is 0.1" The only springs good enough to meet this is Hypercoils. Luckily, economies of scale work in our favour, and Hypercoils are cheap. You can use the exact same quality springs as Real Racers, and there is NO downside.
Now let's talk about balance.
First, we have to all understand one basic fact about DSMs - they understeer under power. NOTHING you can EVER do is EVER going to fix that - save one thing, and that doesn't exist .
The centre diff on a DSM is a 50/50 split. That means it is always going to try and send power forward to the front wheels in the same proportion that it does to the rear wheels. But longnitudnal weight transfer ensures that, under power, weight is going to transfer off the fronts and onto the rears. That means front grip is going to DECREASE, and rear grip is going to INCREASE.
That means power-on understeer. You can play around with how the car behaves at the understeer limit by changing diff types around, but ultimately, a 50/50 split centre diff means UNDERSTEER.
The only real way to get around this is to build a centre diff that has a rear-biased torque split, maybe 40/60. That would send more power rearward than forward, and if the bias was right, the rears would reach their limit before the fronts, and you'd have power-on oversteer, just like a RWD does. I think the WRX STi has one of these. Such a thing does not exist for DSMs, so we get to live with the understeer.
The common "solution" to the DSM understeer issue is to bias the roll resistance heavily to the rear with great big heavy springs and monster thick rear bars. This generates GOBS of lift-throttle and steady-state oversteer, but will never truly generate power-on oversteer - at least not for any reasonably sticky tire. It does, however, give away tons of overall grip. The end result is a car that wants to go around on lift throttle, dances around a lot, and is slow overall.
The better solution is to work with the car instead of against it. You still want a little bit of rearward roll resistance bias (so the car will still rotate on lift throttle, just not as abruptly) but instead balance the roll resistance more in line with the car's static weight distribution.
Conventional wisdom suggests that this would make an understeering car worse. Conventional wisdom is wrong. By doing this, what you are doing is correcting for excessive rearward roll resistance bias, and the car picks up grip. The trick is separating rthe car's power-on characteristics (which are always going to be understeer) from it's power-off characteristics.
What this works out to in practice is that 450lbs of rear spring is pretty much the absolute maximum, and the RM rear bar is pretty much as big as you'll ever want to go. I'm actually softer than that (400lbs of spring plus the OEM rear bar) but I'm also lighter (2804lbs w/o driver)
The rule of thumb is stiffen the front before the rear, and learn to drive around the power-on understeer. Good diffs (Quaife centre, then Quaife front) also help a LOT.
DG
Cusco makes a 35/65 for our cars and at least one frequent poster on this board has one in his ESP car.
