Howard Coleman

iTrader: (6)

very interesting as to the arm situation... i would place a premium on short links and ground clearance when designing my arms. i would actually not mind using the mazda concept, i e links at the end of the arm to promote ground clearance. i don't see being able to fine tune the bar w length adj holes as you get into too much angle.

let's not lose sight of the big picture. the tri point .188 wall bar appears to add 634 pounds/inch. best calculation (thanx Albert) for the stock bar is 590 minus a small discount for motion ratio (links are inboard of springs) we are adding say 450 pounds per inch to the corner w the bar. 475 times motion ratio at the spring is 285 at the wheel. per inch of bump.

so if we run 450 front springs and the stock bar we have 555 total rate per inch at the wheel. since the corner weighs, say 750 w another 250 dynamically transferred we reach lock up or no compliance at just over 2 inches of bump travel. the third inch of bump would be resisted by 1665 pounds!! of course we wouldn't get there as there isn't enough force to compress the bar and springs beyond approx 2 inches. 2 inches of travel is just right, you really dont want less or more.


a couple of suspension asides might be of interest...

during my racing days i found the right size Stock Car Products front hollow bar and did all the bar trimming w a cockpit adj blade rear bar. they were a load to engineer (ball bearings) and were tricky to adjust.

my favorite item was my cockpit adj rideheight. during a 35 minute SCCA race i would burn 65 pounds of gas. that's 65 pounds of rear inside weight lost. lots of my competitors would always complain about how their tires were shot and the track was greasy and they were sliding around. (funny that at the start of the next race the track had stopped being greasy) actually what had happened was the balance had changed in their cars due to the 65 missing pounds of rear weight. toward the end of the race i would crank on my weight jacker (hydraulic) and lift the RR ride height a half an inch and the track would stop being greasy.

as you can see from the above springs/bars thread it is really easy to do away w all your suspension by running too much spring and bar.

i am going to do the SCP straight hollow bar this summer though....

howard coleman

KevinK2

some quick notes:

oem bars are hollow

rich, Sway bars are primative, and RB/ST front bars are brutally stiff (30% stiffer than tripoint at neutral setting). I'd do some track days with the stock 93 bars and your coilovers, before changing bars.

It is nice to have bar adjustment at one end. Mazdaspeed (racer discount?) has a solid 17.5 with 2 holes, that would work well with a solid 28.6 addco front bar.

I have a modified Puhn formula for bars, and my numbers are close to alberts. I agree the SCP chart numbers look bad for solids. The Speedway tables for discrete lengths are more reasonable, and about 15% softer than Puhn numbers. My derived formula for A=C is a few percent softer than Puhn.

http://www.1speedway.com/sb_rates.htm

These rates are for single wheel joust. For cornering the rate is DOUBLED.

To extrapolate out to the spring mount at the a-arm, you need to square the motion ratio. if the bar link is 80% from the a-arm hinge vs the spring, then the rate at the spring bolt is 2 x .8^2 or 1.28 the calculated bar rate.

axr6

Are you saying that both Front and Rear OEM bars are hollow?

I looked at my rear bar in my junk pile today and from lifting it, I believe that it IS hollow. Which would lower my calculations from 220 to 186 if wall thickness was 0.188 or to 163 lbs if it is 0.125.

A = 41.000
B = 5.250
C = 5.250
D = 0.680
E = 0.430
163.10

I had the stock front bar off not too long ago to change bushings and to reinforce the towers and it felt properly heavy for a solid bar. Can someone confirm if it is solid or hollow? I think I actually weighted it but, forgot the measurement.

Albert

KevinK2

Mazda 93 service highlights spec's "hollow" fr and rr bars. I've seen similar specs for 94 and 95.

By wt, I estimated 4mm wall up front. I think rear is similar ... the ends are flattened tubing, about 2xt.

I have rare hollow front 32mm eibach. I think same layout as stock, RB, and ST.

For solid 32mm, K (puhn) = 1400 lb/in.
A = 30
B = 7.5
C = 9
D = 1.25
E = 0

motion ratio to spring is .76, so effective vertical rate at spring bolt is 2 x .58 x 1400
= 1630 lbs/in cornering, or 800 lb/in single wheel bump.

With the Eibach (1200 lb/in), I jack the side of the car up and the suspended front wheel hardly droops at all. That's stiff, and defeats the great F1-ish double wishbone suspension.

I'd slap on eiback bars for the summer. Before track days put on R4 pads and Kuhmo's on stock rims, head for summit point. best time was 1:30 with gravel. Otherwise stock suspension, and intake and exh mods, stock boost.

DamonB

They are.

Mark Ortiz wrote an article that brings up the point of remembering to differentiate between single wheel rate with the bar and the actual roll resistance of bar + springs as Kevin mentioned.

http://www.auto-ware.com/ubbthreads/...n=0&page=0#671

There's a hell of a lot of great articles in there is you haven't seen them before.

KevinK2

Most don't believe my 2x rate thing with bars in roll ... glad to see some back up.

Article was mostly for circle track, where you don't have to brake from 140 for a 40 mph corner. That puts the emphasis on springs, not bars (mabe z-bars ).

I have not used slotted links, but have used thin washers under the bushing bracket ears to soften the bar stiffness.

Although Mark mentioned computing rates at the wheel, he did not say how. Using a-arm hinge line as reference, measure to s-bar link, spring anchor, and spindle ball joint. To translate a rate to the wheel, take linear motion ratio vs spindle ball jt, and square it. (correct as needed for inclined strut). Rate at wheel ~ rate at spindle ball jt. This assume spindle motion is primarily vertical.

for 9" to spring, and 12" to ball joint, 500# vertical spring, rate at wheel is .75 x .75 x 500 = 250 lb/in.

Damon I need another link for back-up here.

DamonB

Sounds like a job for Milliken and Milliken. My book isn't in front of me right now

No formulae here either but another must read from Ortiz:

http://www.auto-ware.com/ubbthreads/...n=0&page=2#153

"An anti-roll bar is an interconnective spring. It generates forces based on its displacement, but its displacement depends on the relative displacement of the two wheels it connects, rather than their individual displacements. It generates equal and opposite forces in the two suspensions it connects....With springs and a/r bars, we have a wheel rate. That’s the rate of the spring, or the bar at the lever arm end, times the square of the spring-to-wheel, or arm-end-to-wheel, motion ratio. The wheel rate defines a simple relationship between force and displacement. Using the wheel rate, we can calculate the spring and bar forces at the wheel when we know displacement at the wheel."

Another interesting read:

http://www.auto-ware.com/ubbthreads/...n=0&page=0#703

KevinK2

Taking link 2 a step further, the new rear roll center at liftoff is the outside tire contact patch. If the rear tire can hold, and front holds too, higher g's will result in jacking up of the rear, increasing the car's cg.

Some mazda6 guys promote going from 19 to 27mm rear bar as only suspension change. Photos show lift off with minimal droop, and jacking. This is fun for street and autox, but would limit road course speeds and could be dangerous in emergency high speed lane change.

axr6

The hollow front bar changes everything. My original calculations were for my assumed- to-be-solid 1.13 diam. stock bar. The only reason for me to want to make this change this time was the expected weight savings, which should be substantial going from solid to tubular with alum. arms. Now that aspect is very much in question as I am wondering just how much weight I could save? Like I said, when I removed the stock bar recently it felt rather heavy, so much so, that I had no doubts about being solid. But, looks like I'm wrong.

From previous uses I know that the Tripoint setup is really light. With my first FD and subsequent GT race cars, I was sold on that setup on the basis of weight AND adjustability. For this present FD, that will only see street use, I am pretty happy with the front roll stiffness and really am not looking for adjustability. That leaves weight as the only factor for a change. Once I heal up a bit, I'll take that bar off again and weight it on my scale to see exactly what it is. I do not mind paying $$$ for weight saved off the car but, at one point it becomes prohibitive. If I can not save at least 10-15 lbs with the hollow bar, it will not worth it to me.

Thanks for the link provided. Interesting perspective and solutions from the author. I am somewhat disturbed to see that various sway bar rate tables, my software and other existing formulas, as presented by KevinK2, seem to produce widely different values for the same bar. I do have a lot of software for calculating pretty much everything on a race cars but, they are only as good as the inputs. If the inputs are in doubt than how can we have the confidence in our printouts?

For me, what it really always came down to was measuring my setup not so much by calculations and theory but, by lap times and that all important "feel".

I'll PM you back on the mountings.

Albert

DamonB

It's no different than anything else. There's theory and then there's application. Once a bar is installed into the car and you begin dealing with droop, jacking forces, linkage geometry changes etc the actual numbers never equal what the equations said they would.

All of that is no big deal though. If you know the rating of the bar on the car now and know how big a percentage change you wish to make in anti-roll bar rate then you just plug another bar in there. The actual numbers of the installed rate are very interesting, but not necessary. You don't need to know the installed absolutes of the bars on the car, merely the relative differences between them.

Coil springs themselves aren't linear (even though most think so!) and since the bar acts through the springs the installed bar rates are never going to abide perfectly by the numbers either...

http://www.stockcarproducts.com/spgtech.htm

Right. As long as we know what we're starting with and have an idea of where we wish to go all you have to do is stick another bar on there.

KevinK2

main stiff's calc diff for front bar is variables input. B=7.5 for front bar.

axr6

Kevin

Just wondering. You are suggesting that the front bar has different arm lengths?

Due to my injury I did not spend the effort to scoot side to side under the car. I just measured the straight length (33"), the diameter, and one arm length. Of course there are small potential inaccuracies as it is difficult to exactly decide where the straight ends and where the arm begins as there is fair radius bend in between. I measured the pass. side arm to be 9.25". What you're suggesting that the driver side arm is only 7.5"? If so, I need to see sometimes what attachment differences would cause that?

Albert

DamonB

Just a thought. You did measure the lever length of the arm and NOT the physical length of the arm itself? Since the arm is not perpendicular to the bar the lever length is considerably shorter than the actual length of the arm.

The proper lever length for the stock bar will be the length of a line projected perpendicular from the torsion tube to the center of the bearing in the stock drop link.

(pic not of an FD bar)

axr6

Excellent thought!

I did measure the physical length of the arm itself and used it for both arms. Now, that I look at the software diagram for measurement guidance, it actually wants one arm measured for length and one arm measured for the lever. So, instead of putting in 9.25" twice, I should have put in 9.25 for one variable and 7.5 (or what ever the lever length is) for the other. Makes sense as the difference must be allowing for the angular position of the arms in reference to the main bar.

Blame it on the blood flow to the brain again. What would I do without such a convenient excuse?

Putting it in this way makes a huge difference. 590 vs 895 lbs/inch. Of course that still calculating it as a solid bar as we do not know the wall thickness.

A = 33.000
B = 9.250
C = 7.500
D = 1.130
E = 0.000
894.02

A = 33.000
B = 9.250
C = 9.250
D = 1.130
E = 0.000
590.07

Albert

DamonB

I've always used this spreadsheet from Auto-Ware:

http://www.auto-ware.com/updates/Aut...preadsheet.xls

It's 7 pages and does spring rates, solid and hollow sway bar rates, center of gravity height, g-force, weight distribution, compression ratio, engine displacement and gears.

KevinK2

for that input, Puhn formula gives 843 lb/in, vs your 894

your 590 lb/in case is 592 per puhn, and 571 from my formulas.

close enough

true value likely 10-15% softer, as formulas assume main tube twists, but does not bend.

I tested the '93 17.5mm hollow bar at 125 lb/in, but I thnk my work bench was flexing a bit. Then put on 19mm eibach hollow bar and it was 30% softer? noticed it really bent in the middle during test, and eventually found a 10" long crack along the central part of tube. No wonder I could not get the car to turn last sessions of last event.

axr6

Those sheets look like most you ever need. My files are also in Excel but with more graphic details for measurements (like I really pay close attention ha..ha.) and additional ones for really complicated full setup for formula cars with downforce, chassis rigidity tests, etc... I did not use those ones even when I raced Formula cars. Probably need to go through a training course just to be able to use those setup pages.

Yes, those rates are pretty close. You had oneset of tables posted which you claimed were 15 percent out from the rest of values and they certainly were.

Yes, that rear stock bar does not do much in car with this weight and CG. More like a decoration. Good thing you ran that test on the Eibach bar, after all, who would suspect a crack in a hollow bar when troubleshooting understeer?

Like I said before, I've always run much bigger rear bars than most people not only on my FD but, on every other fast road or racing cars. Even my wife's Olds gets a big, stiff rear bar against her protest, just in case I need to drive it. Then, she loves the added response and neutral handling. On adjustable bars I regularly ran the stiffest settings. Works for me, that's all I can say. For the present FD, I took advantage of a close-out sale and bought the 7/8" solid ADDCO bar for under $50. Had to do a little grinding on the thick mounting flats to make it fit with the stock end links but, the car corners beautifully on the road, with no tendency for oversteer. I have not pushed it very hard, yet, as I need to install my new shocks first. The old Touring shocks are not that great for these spring rates. Too much wobble, particularly in rebound mode. That is nasty! I have no complaints at all with this bar, despite some negative opinions that I had seen posted about the ADDCO rear.

KevinK2

Those sway bar formulas assumes strait back levers, and no flex in the levers. The Puhn formula assumes typical stock bar, all solid (or hollow) rod, with levers flexing. The aluminum levers look beefy, but aluminum has 1/3 the steel's modulus. Just a heads up.

from Albert:

"... I have no complaints at all with this bar, despite some negative opinions that I had seen posted about the ADDCO rear"

busted ....

axr6

Busted....? YES!

I always LISTEN to majority opinions but, rarely believe them as "gospel" without testing them.

Wonder how many of those posters who expressed negative opinions about this bar had actually tried it with various spring combinations?

On the other hand I had also read a number of posters who were happy with it.

Albert

DamonB

Understood. Using the effective lever arm length should be understood and as long as we're talking street cars with compliant bushings in the suspension I don't see a real need to worry about flexing etc. I think however we choose to dither the numbers in that case they'll still be wrong.