Hey guys, I’ve been trying to get a better understanding onthe balance we run between sway bar rates and springs rates to know what is anappropriate balance. If we double spring rates from stock, should we double barrates? If rates are limited by bumpy courses, how much will be lost in steadystate dropping spring rate and upping bar rate? To start understanding, I needto know what I’m actually running.

To start, I measured my stock FSB, estimated a re-drilledbar, and the whiteline 26mm bar all as well as I could. I had a resolution ofabout 1/8”. Using this equation:

500,000 D^4
K (lbs/in) = -------------------------------------
(0.4244 x A^2 x B) + (0.2264 x C^3)
B
-----_________________
A| /----------------------------\ C
| /------------------------------\
A - Length of end perpendicular to B (torque arm - inches)
B - Length of center section (inches)
C - Length of end (inches)
D - Diameter bar (inches)


The values I get are attached (Tried to embed them, but either I fail or the software does).


Some of the values, like A, I had to get indirectly bymeasuring the hole-hole difference and doing some math, but I feel comfortable they’re all within 1/8”.

I’ve heard people say the whiteline bar is 38% stiffer, but I’mcalculating nearly double that. The 38% I’m assuming is just a check on the 24to 26mm dia change, but I’m finding both a change in “A” and “C” also.

I’m sharing this stuff both so we can all learn and forsomeone to double check my work. Am I right the Whiteline is significantlystiffer than just the diameter would indicate?

 

your center section length (B) needs to be the measurement to the centerline of the car, aka half of what you currently have.

 

nice that you're taking the spring rate of the lever arm in bending into account. I'll have to add that to my spreadsheet.

 

Just pulled out the Puhn book to check, and its showing B at the full width

Yeah, At first I was only looking at either "A" or "C" but it turns out cause we actually have a pretty shallow angle, and its different by a few degree's on each bar that it ends up mattering.

 

what you've got is a torsion bar equation, i.e. you're getting lb/in if you hold one end of the swaybar fixed and move the other end 1 inch. if you're looking to calculate a wheel rate equivalent stiffness you need to take into account that both ends of the swaybar move.

or another away to think about it is you have 2 torsion bars connected together in the middle.

 

Oh, I gotcha. I was just multiply the final rate by 2 since deflection is happening in both directions but I see because of bar end deflection being separate from bar length its inducing an error. Ill change that in my spread sheet tomorrow.

Would you mind sharing what you have for swaybar motion ratio? If not, I still have to pull off springs to change rates and can take it through its range to make some measurements.

 

for the front i use .6 but it was a really quick measurement. If you're going to have everything apart i'd probably still measure it again.

speaking of, it looks like the stiffness of the lever arms is almost equal to the stiffness of the torsion bar which totally ruins my world. LOL so i guess i'll be doing some more validation myself...

 

Alright, Just took front suspension through a 4" sweep with the top being nearly completely compressed.

It started at 0.562 and dropped to 0.54. Average through the sweep was 0.555. So 0.555^2 = 0.316. Total effect the sway bar is contributing stock seems to be pretty small, but I need to change my numbers (left them at work) to get an estimated % value which is overall what I'm looking for.

 

Measured the rear last night and found some interesting stuff.

Rear swaybar motion ratio is 0.69-0.71. Average through a 4" stroke was 0.70. Now with the stiffness of the rear bar coupled with the motion ratio, Stock front and rear bar have almost exactly the same effective wheel rate of 170lbs/in.

 

nice, thanks for sharing Dallas. I dug out my stock rear bar and measured it on the bench the other night, here's what I found:



(500000*(0.866)^4)/((0.4244*10.25^2*38.5)+(0.2264*10.5^3))=142.21lb/in
which is less....

Hotchkis published their lb/in bar rates, also as a percentage over stock, using their 3 numbers and reverse mathing it (technical term) I came up with a claimed 150.14lb/in average.

[edit] and if I understand this right, wheel rate from my measurements, using your measured MR is:
((142.21*(.7^2))*2 = 139.37lb/in

 

I get a little different measurements than you, and I think you're measuring where the bar itself bends where Im measuring where it actually mounts. My thought is anything past the mount will be primarily in bending and not torsion. Of course real numbers are probably somewhere in-between.

I also measure C different. Wondering if you are measuring from the bar by the bend and I'm measuring to the center of the bar at the mount.

I dont directly measure A, I actually just measure the total width of the bar at the endlink then back calculate A given B and C. I guess the only way to know for sure is to make a fixture and load it on my bench to see where between our numbers the actual value is.

 

Oh, this looks fun!

With regard to what value you use for B, it's a question of whether you are calculating the effect of the bar on single-wheel bump or the effect of the bar on roll resistance. Bars are, as is well known, twice as effective for roll resistance than single-wheel bump. That's why there are two formula floating around, one that uses all of B and one that uses half of it.

 

It would be interesting to see how close the real stiffness matches up with theory. I suspect the portion in the center of the bar that is not inline with the mounts actually has a signifigant impact on stiffness.

All the more reason to figure out how to put a true spline drive torsion bar in there...

 

I don't know if I would say its significant, but I would say definitely not negligible or really calculable without modeling the bar and running the FEA. And even then, it would be quicker and easier to just bolt down the bar down to my work bench with an end overhanging and start adding weight.

 

That's why it would be interesting. You're thinking it's not significant, I'm thinking it is.

Not on the level of significance of the bending that takes place in the arm section, but ~20%? Maybe.

I agree though, no point in FEA work when it would take about an hour on the bench.