Evaluating / Designing Suspension (KSport Examined)

[insite]

ok; since i got all this new suspension stuff, i though it would be cool to do the math and compare the KSports with calculated suspension designs for the boxster. this will be the place where i do that.

when i ordered my ksports, i requested dyno plots. they supplied me two graphs. one had 'F' written on it in pen; the other had 'R'. that would indicate front/rear, but the 'F' graphs were all higher damping rates than the 'R' graphs; i think they just swapped the F and R. that, or they really screwed up the damping.....

at any rate, the graphs they supplied each had six lines and no legend. i am working under the assumption that the lowest line is adjustment position 6 and the highest line is adjustment position 36. for the purposes of this thread, i programmed the graphs into excel with lines for min, mid and max, front and rear.

the graphs they sent are called displacement graphs or continuos velocity plots (CVP's). they graph shock displacement vs. force. continuous velocity is sort of a misnomer; the shock dynos use a crank that rotates; as it rotates, the shock is accelerated up and then down. RADIAL velocity is constant, but LINEAR velocity varies with displacement. there was no labeled max on my plot; the industry standard is to set Vmax at 10 inches / second, so i'm working under this assumption as well. using these assumptions, i was able to derive the far more usefull velocity plots which i will explain later.

basically, the assumption is that the dyno compresses the shock by 1" starting at zero in/sec and ending at 10 in/sec; then it extends the shock by 1" doing the same thing. there really are four strokes here: accelerative compression, deccelerative compression, accelerative extension, and deccelerative extension. it is helpful to examine all four phases because sometimes a shock will behave differently after a big rebound or compression stroke. this effect is called hysteresis.

i'm going to start with the fronts today; we'll move on to the rears later. here is the front displacement graph:

[insite]

for the above graph, compression is the top part and rebound is the lower part.

so why is rebound so big and compression so small? there are a couple of reasons. first, as the car compresses, work is being done to compress the spring. this reduces the damping force required by the shock. when the car rebounds, the spring is pushing against the shock, increasing the requirement for damping. second, think about what happens when you hit a bump: the wheel, brake calipers, rotors, etc. are pressed up toward the car. compression controls the UNSPRUNG weight of the wheel, brakes, 1/2 the strut weight, 1/2 the control arm weights, etc. the rebound stroke controls the weight of the SPRUNG mass, which is everything that sits on the springs.

looking at the graph, it's clear that adjustments higher in the range make larger changes than adjustments at the lower end of the range. i.e. changing from position 1 to position 2 makes FAR less of a difference than changing from position 35 to position 36. this actually makes sense; the relationship is roughly log linear.

there are some funky things going on at the upper right on the graph; it's almost like the shock wasn't sliding properly during the dyno run. this could be indicative of a problem or of the fact that the shock hasn't really been broken in yet. odd, none the less.

these particular graphs are ok for comparing adjustment 'clicks' on a damper or for comparing two dampers, but not quite so good for telling us what is happening in the design. for that, we want a velocity plot:

[insite]

ok, so now we have a velocity plot. let's examine what it tells us.

the job of a damper is to control the motion of the car. without a damper, a sprung mass will continue to oscillate (bounce up and down). the damper's job is to damp that motion, to gradually slow the controlled mass.

when something is critically damped, it will return to equilibrium without going past it. when something is underdamped, the mass will go past equilibrium at least once before returning. when something us underdamped, it will take too long to return to equilibrium.

in carspek, underdamped means more bounces when you hit a bump. overdamped means the car won't recover fast enough after a bump; this can cause a 'jacking' effect where multiple bumps in a row will cause the suspension to bottom out. critical damping means the car will hit the bump and immediately return to its proper ride height.

we want something that is slightly underdamped. the a damping ratio of 1 is critical; less than 1 is underdamped, greater than 1 is overdamped.

for road cars, the damping ratio is generally around .3 - .5. for sports / track cars, it's around .7. look at the graph below to see what these different damping ratios look like.

[insite]

so let's talk about the velocity plot above. there are a few things to look at. first, there are two phases to most automotive dampers: low and high speed. low speed generally is for handling while high speed generally is for bumps. low speed is usually a damper shaft velocity of about 0 - 4 inches / sec and high speed is over 4 or 5 inches / second.

the 'knee' in the plots above shows that the ksport's damping curves change from low to high speed slopes at 2 inches / sec. this is a little early.

without getting into discussions on transmissibility, etc, let's just look at the slopes. generally, the compression slope should be about 2/3 (.67) of total damping and the rebound slope should be about 3/2 (1.5).

the min plot has no knee in rebound. the slope multiplier is only 1.08; this means that the 'min' setting is overdamped in compression compared with rebound. the damping ratio is a comfy 0.5.

mid has an obvious knee at about 2 inches / sec. the multiplier is 1.41 and the damping ratio is 0.78. this is a pretty good curve. the only thing i'd change would be to move the knee to 4 inches / sec.

max has the same knee at 2 inches / sec. the multiplier is 1.73 and the damping ratio is 0.96. this is overdamped all around, leaning heavily toward rebound.

the middle setting looks like a good choice here.

[insite]

now let's look at the high speed damping. this is the slope to the right of the 'knee' in the velocity plot. ideally, we want a lower damping ratio of around 0.2 - 0.3 to reduce discomfort over high frequency bumps at speed.

the min setting gives us a damping ratio or 0.24, but it's a little underdamped in rebound compared with compression.

the mid setting has a damping ratio of .25; still biased against rebound

finally, the max setting has a damping ratio of .26 with an even greater bias against rebound.

so what would the ideal damping curve look like for this spring rate on the front of a boxster? well there is no true 'ideal'; it's all compromise.

for a streetable track car, i'll go with a low speed damping ratio of 0.7, high speed of 0.2, and a knee at 4 inches / second. look at the chart below for a comparison of my 'ideal' front shock curve for my 400lb springs to the 'mid' setting on my ksports.

[insite]

there's another damper performance indicator called hysteresis. basically, hysteresis is a difference in how the shock behaves based upon conditions. if a shock produces 200lb rebound force at 4 in / sec, but only 170lb rebound force at the same velocity immediately after hitting a bump, we have hysteresis.

essentially, hysteresis is degradation of performance that occurs during use resulting from internal construction. this can be caused by sticky seals, cavitation, or a host of other issues.

the following chart is a velocity plot of my front damper at 'max' setting. the outer-most lines are while the shock is accelerating; the innermost are while the shock is decelerating. the difference is hysteresis. it's in the neighborhood of 6%; not great, but not bad. probably average. for the price, i'd say it's pretty amazing actually; i expected much worse.

[insite]

Summary (Fronts Only!!)


the dampers curves are reasonably close to what i've calculated are the 'correct' values for this car with my chosen spring rate of 400lb/in. the minimum setting is underdamped and the max setting is overdamped; this is good, it means that the 'middle' of the curve is where we want to be.

by my calculations, for these springs, setting my front dampers between 18 - 24 should be about as close as i can get to my ideal.

look for a similar analysis of the rears shortly.

[RandallNeighbour]

holy crap, batman! That's some serious technical talk, replete with illustrations!

After you do the write up on the rear coilovers, put all the cookies on the lower shelf for this village idiot and type something like:

"Turn the knobs all the way to the right and then click back to the left X times for comfy street use and Y times for DE's..."

You're one smart feller.

[grantsfo]

Guess my question would be how long will K Sports hold those specs? Thats why I went with Koni tube coilovers in the KW's I installed. I just havent built up courage to go down market to try anything else other than Koni or Bilstein tubes. And its nice to have great Koni support here in the US.

It is cool to see a cheaper alternative being tested however! Keep up the good work!

[insite]

Originally Posted by grantsfo

Guess my question would be how long will K Sports hold those specs? Thats why I went with Koni tube coilovers in the KW's I installed. I just havent built up courage to go down market to try anything else other than Koni or Bilstein tubes. And its nice to have great Koni support here in the US.

how long they'll hold those specs is anyone's guess. time will tell...

as for US rebuilds? KSport USA has a shop in arizona that does rebuilds in like two days for around $100 per corner. they will custom valve as well, which is nice.

i think ksport still has some kinks to work out, but overall these things look a lot better than i thought they would. more to come.