KW V3 Spindle/Hub Installation Problem

[PeteKz]

Jonny, thanks for the feedback. I'm aware of the need to correct bump steer. I currently have the spacer kit on my car. I know that's not enough for a 19mm raised spindle.

[Jonny042]

Originally Posted by PeteKz

Jonny, thanks for the feedback. I'm aware of the need to correct bump steer. I currently have the spacer kit on my car. I know that's not enough for a 19mm raised spindle.

I realize now that if I'd read your signature I'd know that you'll have no issue with 19mm raised spindles, with the 16x7's!

With the added suspension travel (and a decent bump steer curve) you could soften the front of your car up to 19's, and leaving the 26's in rear would be a great match. I just had to dig up this old post regarding the ride frequencies and the learning curve I went through with the Project:

Originally Posted by Jonny042

I've been mystified for a long time about the spring (torsion bar) rates that Porsche chose for the 911. They don't make sense. Perhaps it's because they didn't have the internet?

I'm not inclined to get into the nitty gritty of the math but we have to roll out a few numbers. Lets get the torsion bar stiffness out of the way. I found conflicting information on the wheel rates of various thicknesses of torsion bars so decided to calculate it myself based on the formula and data provided by sway-a-way:

https://swayaway.com/tech-room/vw-porsche-torsion-bar-rates/

They have actual data on that page for various diameters of torsion bars in "raw" form in inch-pounds per degree of twist. They also provide a formula you can use to come up with wheel rates in pounds/inch. I used 11" for an arm length in front and 19 in the rear to come up with wheel rates in pounds per inch:

Front

19mm - 126 lbs/in (OE -1989) *actually 18.8 (120-ish in/lbs?)
20mm - 167
21mm - 202
22mm - 244
23mm - 291

Rear

23mm - 96 lbs/in (OE -1977)
24mm - 113 (OE 1978-1985) *actually 24.1 (120-ish in/lbs)
25mm - 134 (OE 1986)
26mm - 156 (OE Turbo 1975-1989?)
27mm - 182
28mm - 210
29mm - 242
30mm - 277
31mm - 316

So far all is right in the world - lo and behold, thicker bars are stiffer.

What doesn't make sense is that for so long the OE wheel rate at the rear is less than the front? These calcs ignore the influence of the suspension bushings and shock nitrogen pressure so it's not than cut and dry, but only the torsion bars changed so we're going to conveniently ignore the other parts.

Suspension frequencies. You all remember high school physics when you learned about spring frequencies? All sprung systems will move at a natural frequency based on weight and spring tension. Thinner, tighter, shorter guitar springs oscillate faster than thicker, looser, longer bass springs, a heavy weight on a soft spring will bounce more slowly than a light weight on a stiff spring. The unit for this is Hertz. I don't think Mr. Hertz invented the spring, or mass (obvious that was Mr. Pound) but he must have found them interesting. 1 hertz is 1 cycle per second. 60Hz. is 60 cycles per second.

An interesting side note, the standard frequency for electrical generation is 60Hz but it has nothing to do with spring tension, it has to do with rotating machinery and the sine wave they develop when you create AC voltage and current.

Think of a 70's cadillac with blown shocks compared to a 90's civic riding on $39 ebay coilovers (or the bumpstops).

I stole this next bit from RCE. Not sure who they stole it from:

https://www.racecompengineering.com/blogs/the-apex-files/spring-rates-part-2-suspension-frequencies

1.0Hz - Passenger Cars

1.25 - 1.75 Hz Sports Cars

2.0 - 2.5 Hz Autocross cars and racecars with low downforce

2.5+ high downforce race cars

On to the subject at hand. Here's project heavy metal immediately after I've run out of gas (but before I've corrected the cross weight to 50%)



Notice the front weight is almost alarmingly light.... luckily the gas tank and spare tire are up there and someone has to drive, so lets arbitrarily add 232 pounds of gas and passengers and spare tire to that to bring it to an even 1000 (500 per wheel) pounds, and the remaining 168 pounds of passenger etc. to the rear to bring it up to 1600 (800 per corner).

The addition of the weight to the front end brings the weight distribution from a nearly scary 35/65 empty to a less scary 38.5/61.5). Makes you realize why they put lead weights in the front bumpers of the earliest 911s. It's true!!

I used a suspension frequency calculator at:

https://www.racingaspirations.com/apps/wheel-frequency-calculator/

And came up with 1.6 Hz for front frequency and 1.25 for rear. The ratio of front to rear is 1.6/1.25 = 1.28

That calc uses the assumed wheel rate of 120 lbs per inch front and rear for the OE 911SC setup in the car at present. At least we are in the range for "sports cars" but there is an odd miss-match front to rear - there is definitely something strange in the state of Stuttgart.

For performance street/track use the usual advice from everyone and their Elephant is to go with 21mm front, 28mm rear bars so lets order that up and try them on for size.

This gives us 2.08Hz front and 1.66 rear (1.25 fr/rr). Which is racecar in the front and sportscar in the back if you ask me. I have this setup in my autoX car and have been fighting understeer but that's an aside. That car needs 29mm rear bars at some point or back to stock fronts.

My initial plan, then, was to go to 20mm front 27mm rear. I even placed an order. We get:

1.9Hz front and 1.53 rear (1.24 fr/rr)

After my recent road trip where I drove the car on all sorts of road surfaces (including (GASP!!!) a few miles of gravel, cement slab highway, broken pavement, etc. I concluded that the last thing I need to do is stiffen the front of this car. Also, when I got back from my trip my new torsion bars weren't waiting for me as promised.... so I cancelled that order and I have shiny new red sway-away 26mm rear bars for the rear of the car. This will give me:

1.6Hz front and 1.43 rear (1.12 fr/rr)

This last bit is important - I want to dial out a bit of understeer which I'm finding mid to late corner (don't forget the chassis is dealing with double the hp of a 911T!) and also want to limit unwanted rear suspension motion.

The car really squats when you accelerate (watch the dyno video for proof, and that not even with weight transfer!!) which brings with it bump steer, camber change, and ride height movement, none of which is really desirable. Not to mention there are times when the weight of the rear end is bobbing up and down over undulations and bumps for no good reason.

The rear spring plate bushings can also contribute rear steer influence under accel/decel, and not in the way you'd like. Adding the slightly stiffer HD rear bushings should help to tame the rear a little.

Funny thing is, I never gave the rear a bit of thought until I dialed out the front bump steer as it was so bad you couldn't even tell what was happening at the back.

Off to get out the tools!!

[Bill Verburg]

here''s a summary of street oriented effects from t-bars on a 2650# Carrera w/ stock 20/18 underbody sways>> track, sways and shock pressures all affect this, your effect will be reflected in that

[Jonny042]

Originally Posted by Bill Verburg

here''s a summary of street oriented effects from t-bars on a 2650# Carrera w/ stock 20/18 underbody sways>> track, sways and shock pressures all affect this, your effect will be reflected in that

Bill, it's hard to tell without having the excel file but are you including the nitrogen pressure effects of the shock in the spring rate? The effect of the nitrogen pressure is mostly constant throughout the range of travel and can't be considered as spring rate.

[Bill Verburg]

Originally Posted by Jonny042

Bill, it's hard to tell without having the excel file but are you including the nitrogen pressure effects of the shock in the spring rate? The effect of the nitrogen pressure is mostly constant throughout the range of travel and can't be considered as spring rate.

yes, Bilsteins all have ~42# of static push(this is not internal pressure it is the force extending the shocks), this increases linearly as the shock is compressed. I don't have those values other than static,
So the data above is a min.

Tell me why it can't be considered as a spring rate

[Jonny042]

Originally Posted by Bill Verburg

yes, Bilsteins all have ~42# of static push(this is not internal pressure it is the force extending the shocks), this increases linearly as the shock is compressed. I don't have those values other than static,
So the data above is a min.

Tell me why it can't be considered as a spring rate

It does increases linearly, but not at the same unit of lbs/in. ie: at full extension it takes 42 pounds to overcome the force, but 1 inch into the travel that force doesn't become 84 pounds, then 126 after 2 inches of bump, etc... like it would if it was spring rate. Which is why it's not considered in the spring rate.

The only increase in the force exerted by gas pressure through the shock stroke is due to the additional volume the shock shaft is taking up as the shock travels in bump, and it's very small as a proportion of the overall volume. If I had to guess it's a few pounds increase, if that.

That 42 pounds can simply be considered as preload that the torsion bar doesn't have to deal with and it's not a dynamic component.

[Jonny042]

And, even though I'm the last person that should be checking anyone's math, there seems to be something broken in the total/sprung/unsprung section that might be throwing off the frequencies as well.

[PeteKz]

Yes, the shocks also affect the suspension frequency. It's a "mass-spring-damper" equation, and because the shock damping increases as V^2, it's a non-linear frequency response. Same math as the RCL equations in electrical engineering.

[Bill Verburg]

Originally Posted by Jonny042

It does increases linearly, but not at the same unit of lbs/in. ie: at full extension it takes 42 pounds to overcome the force, but 1 inch into the travel that force doesn't become 84 pounds, then 126 after 2 inches of bump, etc... like it would if it was spring rate. Which is why it's not considered in the spring rate.

The only increase in the force exerted by gas pressure through the shock stroke is due to the additional volume the shock shaft is taking up as the shock travels in bump, and it's very small as a proportion of the overall volume. If I had to guess it's a few pounds increase, if that.

That 42 pounds can simply be considered as preload that the torsion bar doesn't have to deal with and it's not a dynamic component.

linear increase doesn't mean double, it just means that the slope of the rise is constant
PV = nrT

again I don't know what the slope is, just that it is constant

[Bill Verburg]

Originally Posted by Jonny042

And, even though I'm the last person that should be checking anyone's math, there seems to be something broken in the total/sprung/unsprung section that might be throwing off the frequencies as well.

I'll check into that, i've been doing a lot of consolidation and expansion of this in my copious spare time.

[Bill Verburg]

Originally Posted by PeteKz

Yes, the shocks also affect the suspension frequency. It's a "mass-spring-damper" equation, and because the shock damping increases as V^2, it's a non-linear frequency response. Same math as the RCL equations in electrical engineering.

damping rates are a separate issue

in the charts I did figure a nominal damping rate but it isn't used for anything, I was just curious how it compares to the published damping curves for the various varieties of Bilstein

[Jonny042]

Originally Posted by PeteKz

Yes, the shocks also affect the suspension frequency. It's a "mass-spring-damper" equation, and because the shock damping increases as V^2, it's a non-linear frequency response. Same math as the RCL equations in electrical engineering.

Yeah but we aren't even talking about damping, we're taking about the force exerted by the gas pressure in the shock on the rod which causes it to extend.

I'm not saying shocks don't affect the suspension, we're just talking about the simple calculation of how "stiff" either end of the car is. Like a weight bobbing on a spring, heavy weight on a soft spring bobs slowly, light weight on a stiff spring bobs quickly.

Damping is not figured in this calculation - it's something you add to manage the resulting characteristics after the fact.

This simple calculation results in a Hz value which is then commonly thrown around like Bill has at the bottom of his spreadsheet.

I'm not making this up. First hit on google:

[Jonny042]

Originally Posted by Bill Verburg

linear increase doesn't mean double, it just means that the slope of the rise is constant
PV = nrT

again I don't know what the slope is, just that it is constant

If you don't know what the slope is how can you include it in spring rate as though it's 42lbs/inch?

PS - I'll try to ignore the insulting and irrelevant dissertation on how linear doesn't mean doubling.... to double the compression on a linear spring the force needs to double. Can you please make an effort to understand what I'm saying instead of going straight to insulting my intelligence?

[Bill Verburg]

Originally Posted by Jonny042

If you don't know what the slope is how can you include it in spring rate as though it's 42lbs/inch?

PS - I'll try to ignore the insulting and irrelevant dissertation on how linear doesn't mean doubling.... to double the compression on a linear spring the force needs to double. Can you please make an effort to understand what I'm saying instead of going straight to insulting my intelligence?

The static force is easily measured w/ any shock on a bench

the compressed force can be measured but I didn't and don't have that information from any other source

canister pressure which is the source of the shock force is a tuning tool for shocks like MCS and JRZ

I use it because its there and the 42# is a minimum #, none of these #s are what one could call precise, there are all sorts of additional factors like the tire spring rate, scrub rate and stiction forces at play

[Bill Verburg]

Originally Posted by Jonny042

And, even though I'm the last person that should be checking anyone's math, there seems to be something broken in the total/sprung/unsprung section that might be throwing off the frequencies as well.

I went back and looked at that

the unsprung &sprung isn't used elsewhere only the track is used to calculate roll rates

I did fudge the corner weights and used a theoretical 40/60 split instead of the actual measured #s from my car, the actual measured f/r ratio is 39.5/60.5 & 2594#

[Bill Verburg]

Originally Posted by Jonny042

If you don't know what the slope is how can you include it in spring rate as though it's 42lbs/inch?

PS - I'll try to ignore the insulting and irrelevant dissertation on how linear doesn't mean doubling.... to double the compression on a linear spring the force needs to double. Can you please make an effort to understand what I'm saying instead of going straight to insulting my intelligence?

no insult was intended
yes the 42# at full extension is a preload. the preload increase w/ compression and is proportional to the internal gas pressure described by the rule PV - nrT

if the gas chamber volume is halved the pressure and temperature go up, some of the energy goes into P inc and some goes into T inc, I don't know the ratio there. Internally there is a nitrogen gas chamber separated from an oil filled chamber, the gas volume is compressible but stays in the chamber, the volume, pressure and temperature of the gas in the chamber varies in proportion to the shock piston movement, the pressurized gas forces the relatively incompressible working fluid through various orifices to damp the motion and shape the damping curve.

the #s in this diagram are notional not specific to a 911


the gas pressure pushes the chassis upward just like a spring, the min. value for this upward push is ~42# the max is unknown to me

if it walks like a duck and quacks like a duck treat it like a duck

[PeteKz]

I tend to agree with Jonny about the pressure inside Bilstiens acting more like preload than additional spring constant. However the way to find out is to put a Bilstien on a scale and press down on it at the top of the stroke and the bottom of the stroke to find the spring constant. I don't have time to do that this weekend because I have a plumbing project to finish--house plumbing, that is, "honey do" list.

Jonny, the equation you found with Google is a first order approximation, good for comparing the suspensions when just changing the springs. As I said before, the damping contributes greatly to the suspension frequency response and "felt stiffness." You probably know that too.

For others following with a casual interest, consider the difference in response for a shock with very little damping, versus a shock with a LOT of damping. In the first case, the suspension will bounce up and down a lot before ironing out, at pretty close to the sping/mass frequency Jonny cited above. But with a lot of damping, that movement will be slowed considerably--thus necessarily decreasing the suspension frequency--even though it makes the ride feel much stiffer.

That brings up something I emphasize frequently: The harshness of the suspension that you feel in your butt depends more on the damping (shock stiffness) than the stiffness of the springs. You can prove this to yourself if you have adjustable shocks (like my Koni's) by setting the shocks at their softest and then comparing to the settings at their hardest. Big difference, with the same springs/T-bars. And, yes, shocks with digressive valves or other valve designs complicate these relationships further.

I find that for a street car, I prefer moderately stiff springs/T-bars, with moderately soft shocks to tame rough roads and maintain a reasonably comfortable ride without too much roll or wallowing. Race cars are a different case.

[Jonny042]

Originally Posted by PeteKz

However the way to find out is to put a Bilstien on a scale and press down on it at the top of the stroke and the bottom of the stroke to find the spring constant.

Great minds think alike! I took a good spare Bilstein insert today and did just that:

Just barely compressed:



Almost fully compressed:



So, the effective spring rate of the gas charge is less than 1lb/in.

[Bill Verburg]

Originally Posted by Jonny042

Great minds think alike! I took a good spare Bilstein insert today and did just that:

Just barely compressed:



Almost fully compressed:



So, the effective spring rate of the gas charge is less than 1lb/in.

Good to know, as I said I didn't have that info

still its a 42# shock rate * .9 to get wheel rate added to the t-bar rate

an 18.8mm front w/ wheel rate of ~110lb/in becomes ~148l b/in + ~4# at full compression for ~152#/in max

[Jonny042]

Originally Posted by Bill Verburg

Good to know, as I said I didn't have that info

still its a 42# shock rate * .9 to get wheel rate added to the t-bar rate

an 18.8mm front w/ wheel rate of ~110lb/in becomes ~148l b/in + ~4# at full compression for ~152#/in max

No, it doesn't. It becomes 111lb/in!

Which is the whole point I'm trying to make, and the reason the ride frequencies you're coming up with are so high.

Does it not strike you as odd that your calculated frequencies are double that of a typical passenger car?