Awesomatix leaf spring rates lbs/mm yield gradient
 

I setup a test rig for measuring the leaf springs in a couple of my cars and I figured I would post the results separately from the main thread so that people could feel free to talk about them and not feel like they are cluttering the main thread. I have 1 chassis that just feels dialed and (a800xa evo) im trying to get the rest of the spring rates similar.
https://cimg6.ibsrv.net/gimg/www.rct...26d340ce3.jpeg
The top curve is what I want. There is something very magical about to low resistance and high travel this setup provides. The lower curve is a heavily bent eyelet on a 15% progressive PSS. I think the evo spring is probably a 98 degree. Before the bend in the mmx the curves we more like an s. All cars are set with screws in position 2. I’ll have to check the gaps on the hook and damper and add that value because it is not set at full soft. The track that it performs well at is 377 feet long with 12 seconds as a good lap time if you want to calculate the average speed. Surface is vht with bumps of 2mm spaced at 4mm.
https://cimg3.ibsrv.net/gimg/www.rct...d4ab9f58f.jpeg
This curve is unbent soft springs (all testing is on soft) brand new. All settings are at full soft and a ride height of 5.4 all the way around. Unless otherwise stated this is what all testing is at. This mid motor is an mmx on carbon.

I think the magic of the evo is the spring rate is very linear and starts very low so the shock probably sits in the middle of travel rather than biased towards one end or the other.

Im currently working on a method to soften the rate of the front spring to get the mmx closer to the evo and get the shock and springs into the middle of the stroke and much more gradual in the force it applies. Bends and shims are currently the methods im using. It could be that the soft springs on the evo are just so warn that the internal bonds that create the force have just broke down to this point and I may not be able to use bends and shims. Its odd the way the front curve on the evo is so shallow and although all of my chassis are 50/50 by weight and corner scales Im thinking that if I can get more weight distributed to the rear springs and balance the distribution with mass on the front I might emulate the soft rate of force the front applies. My logic comes from the evo front rate and the fact that it is the only chassis i have that im not fighting on throttle corner exit oversteer.

The horizontal green line in each curve is the amount of mass on each corner scale minus some misc mass to approximate where static ride height should be at. Tonight i installed the p09x on the mid motor front. i will set ride height and do some more data tomorrow,

as an aside I have always noticed something about setting up new springs on a car. If I put new springs on and adjust ride height to say 5mm i find that corner scale values can be 40 grams off per corner.. so then i adjust the ride height to get the mass correct on each scale and the ride height will be off by .2 or .4. This is on cars that are mass balanced and flat. I confused it for tweek for a long time. Its not on a new chassis that is flat.

I can tell you how I fix it and it will stay fixed until I change springs and then i will explain what it is.

When I adjust preload for scales i only adjust diagonally to distribute weight and I always do all 4 corners. If the front RT and left rear are heavy I lower the ride height on them and raise it on the opposite corners.

then the ride height is off. When i adjust the ride height I do 1 corner at a time and I do not adjust corners diagonally.

I do another preload adjustment on the scales with the same diagonal method.

now the ride height and spring preload are aligned. The corners are balanced and the ride height is flat.

springs have a difference start and end point. They even have differences in force per millimeter. If they are not synchronized from spring to spring it creates this force to ride height misalignment. It might appear as tweak or it might not even be evident except when the springs are in motion causing a rocking motion. When i do the static preload and ride height synchronization it has fixed all problems that looked like tweek but if the problem is in the curve of each spring while its in motion the corner scale/ ride height method wont find that. Messing each point on the spring curve can do that even though that is not the reason I setup this test rig. I simply wanted to copy the spring response on a good setup and put it on my other chassis.

I think the chassis becomes part of the spring preload. Its probably easier to visualize a twist running through the chassis even though its not evident in flatness measurements. You can see the twist when you look at what is happening at the corners on the preload measurement skewing the ride height and vice a versa. Once the twist is out it doesnt come back from a hit or breakage. I saw the same thing in coil springs and leaf springs. Its just a spring thing. Now that I work this synchronization out I never have to loosen the top deck to remove tweak.

I am really amazed at how linear these leaf springs are. I have no evidence that is good or bad but I would have expected hobby springs to have allot more deviation than Ive seen. And the different slopes that can be created by a couple changes is extremely well thought out.

https://cimg7.ibsrv.net/gimg/www.rct...112d371b7.jpeg
Charts are easier. The blue is what I want and the other 3 are variations based on spring changes. The bottom right is the p09x at 4.2 and 5.2. I think you can see the point where it transitions from one screw to the next.
https://cimg9.ibsrv.net/gimg/www.rct...b96fe0c22.jpeg
This is a little math to compare the rate changes. Blue is still the target. Bottom right is the p09x set at 4.2 and 5.2.

Im sure some people are curious what slope means. If you divide the spring force or say .31 lbs by the distance you can see the rate of change per mm. So the slope a 1mm with a force of .31 is a slope of .31. As the spring compresses to 2mm it would be .62 lbs if the slope was to stay .31. That is the actual yield rate for every mm. Most springs are measured at 1 point for example 10mm. But as the spring compresses the rate per mm usually changes and measuring at 10 is an average of all the position up to that point.

a spring has to deal with compression and rebound and the mass in rebound is different than the mass in compression. That’s why I plotted ride height. 1 side of ride height is compression and one side is rebound. If you are accelerating the arm to return back to ride height it takes a different amount of force than the force required to hold the ride height of the chassis in compression. That is why I asked the question about where ride height should be when looking at the rate of change on the spring curves. The ride height is the green line and im not sure if its a coincidence that its very close to the tip over point on allot of the spring curves.

The bottom right slope and measurements graphs is moving in the right direction. It has the lowest slope so far and I have not softened the initial part of the curve with the 1st stage screw. I also haven bent the eyelet yet but considering its so low already I think Im working in the right direction. It appears I may have a little more control over the rates with the p09x design. However I have not tested the 25% pss. I checked everything and I was always testing 15% pss regardless of what some of the pictures say.

Interesting topic.
Do you have by any chance a picture of the cars and of the test rig ?

Also, it may be a dumb question, but what arm's length do you use on each cars ?

I'm currently doing the same type of research on my car too. I'm too trying to know the stiffness of the leaf spring. I tried to calculate it via the stiffness tables at the end of the manual and motion ratio, via simulation and via 2d beam calculation.... 3 different results... :(


Thank you

https://cimg2.ibsrv.net/gimg/www.rct...eeaf5162d.jpeg
here is the test rig. cheap drill press that I flipped upside down and 3d printed a bracket for the height gauge and strain gauge.
https://cimg3.ibsrv.net/gimg/www.rct...4cc50e25e.jpeg
This is the mid motor that im doing the testing on and the arms are mounted to the innermost position. Long arms I believe either +8 or +9.
https://cimg8.ibsrv.net/gimg/www.rct...f51b6f244.jpeg
This is the evo and the arms..+1ish are mounted on the outer set of screws. This is the reference car.

Thanks for the pictures!

So, I don't know your exact testing procedure but if you're measuring at the wheel hex you're getting the stiffness of the suspension at the wheel, and you're not getting the spring stiffness. I think the stiffness graph at the end of the awesomatix manual are at the wheel so you could compare to that.

Another thing, if you're in fact measuring at the wheel hex with your rig you won't find the same stiffness on your two cars when you are at the same A & B suspension settings. That's because you're using different arm length on your two cars, so it changes the motion ratio which could impact your findings. Same thing for the anti-roll bars, if one car is equipped and the other is not, you will have different results too. I hope you can find the solution to your problem. Cool project !

Yes sir. You are correct. Differences in camber gain or active toe, bump steer can all potentially skew the results. But I think there are some things that stand out like the slope of the evo compared to the slope of the MMX which should be opposite based on length if that were the driving factor. The results are more philosophical than empirical. Thanks for the info on the manual though. Makes sense.

By the way you are welcome to post your data. It sounds interesting.