E28 spring rates

[Devinder]

Thanks to everyone who provided exemplar springs for measurement.

Devinder

[Mark 88/M5 Houston]

Thanks for testing, organizing and posting the data.

[sommy]

What about alpina springs? =]

[Nebraska_e28]

Got any specs on SPAX? From what I've heard they are among the stiffest.

[cgraff]

I have a few questions:

1. What equation did you use to calculate the spring rates?

2. What did you assume for your torsional constant?

3. What did you assume for number of *free* coils?

Thanks

-Chris

[Devinder]

Chris,

1. I don't know how to type it out here in ASCII so I'll just describe it as the "standard" equation for coil springs. There is only one and any Mechanical Engineering text should have it.

2. G=79.3 GPa

3. As you've surmised, the estimation of the number of dead coils is the key to the problem. Luckily, used springs have obvious witness marks that show the number of dead coils. Where I had access to used springs, I just measured the actual number of dead coils (tends to be about 2.6 for 6 coils and 3 for 7 coils). In all the dozens of E28 springs I've looked at over the years, the number of dead coils is very consistent. For new springs, I just assumed the same number of dead coils as a known spring of the same OD and number of turns.

I posted a summary sheet. The actual spreadsheet I used to calculate the rates has much more detail in it. I know that some of my findings are going be different from what people have heard or have been told. You should know that I only included in the summary springs that I personally measured and only springs where I could verify the provenance or the part number application. I actually measured many more springs than I've shown to check the numbers. I've also put several springs in my Instron.

I've run every spring combination listed here (in one form or another) except Dinan on my M5 on the track. I think Dinan is going in the right direction by keeping stock front stiffness and increasing rear roll stiffness.The E28 really needs to increase rear roll stiffness relative to front roll stiffness to deal with the understeer problems. I suspect BMW was trying to do the same thing by setting rear tire pressure at 40 psi on the M5. I hoping to try out Dinan springs on the track this year.

For those of you asking about Spax and Alpina -- I would need to get my hands on those springs to measure the dimensions and calculate the spring rates. If you want to send them to me, I can measure them and send them right back.

Let me know if you guys find any errors in dimensions or part numbers. It is possible that there are other variants of the same part number or application (like the ST).

Devinder

[cgraff]

Chris,

1. I don't know how to type it out here in ASCII so I'll just describe it as the "standard" equation for coil springs. There is only one and any Mechanical Engineering text should have it.

2. G=79.3 GPa

3. As you've surmised, the estimation of the number of dead coils is the key to the problem. Luckily, used springs have obvious witness marks that show the number of dead coils. Where I had access to used springs, I just measured the actual number of dead coils (tends to be about 2.6 for 6 coils and 3 for 7 coils). In all the dozens of E28 springs I've looked at over the years, the number of dead coils is very consistent. For new springs, I just assumed the same number of dead coils as a known spring of the same OD and number of turns.

Hey,

I just checked my Milliken text and they have G=11.0x10^6 psi. Which I assume is ~79.3 GPa. Of course we have to realize there will be some errors depending on the type of steel (granted not large, but there is some there).

Also, According to the spring equation in the text, you want to use mean coil diameter, not outside diameter of the coil spring.

As for number of active coils, you'd have to measure it at ride height, unfortunately. You can't really estimate it from looking at the spring marks. I feel that given the taper to the coils at the end, the number of active coils actually changes as you compress the spring. This adds another variability to the equation. I would argue that the number of active coils is slightly more than what you would get from looking at the witness marks. I pulled out my stock springs and I see a nice bit of rust at ~2.5 coils out of the 7. The better way I'd suggest is to measure the distance between the coils along the height of the spring. There you can find exactly the point where the taper begins/ends.

I posted a link some time ago to scans of the technical BMW documents of the stock springs and BMW M-tech springs. That is where I get the 166 lb/in and 194 lb/in spring rates for the front springs (and I have the rears up there someplace as well). I'd have to dig that up, but I'm sure a search will bring them up.

Anyway, going to an example stock 535i/528e front spring, using S = (G*d^4)/(8*(D^3)*N), and your measured 13.2 mm wire OD, 143.2mm coil OD (and thus 143.2-13.2mm coil mean diameter), a G = 11*10^6 psi, (yes I know, I'm using US units...forgive me), for N=4.5 (or 2.5 dead coils), I get a spring rate of 166 lb/in. That matches the BMW tech docs I have.

It's late for me here...I don't think I have many qualms with the measurements - I mean they are what they are. I'm trying to reconcile what I have in technical documentation with what's been measured.

In terms of the resulting 'balance' of the car (e.g. your comment on Dinan spring relative stiffness), it is a point well taken.

-Chris

[eurowannabe]

Devider, here are the part numbers as I have them from my recent order: Fronts: 31 33 2 225 645 ("short red" designation, it is listed as M-tech) Mine have Eibach in logo form written onto the spring and they came with BMW part numbered stickers.

Rears: 33 53 2 225 654 ("short red") The springs are not red, but the rears did come without markings, just the BMW part number sticker. If you look up the part number on Maximillians parts site ( http://www.bmwmobiletradition-online.com/bmw/ ) you'll see them listed as M-tech.

I hear the 5th digit in the part number indicates ///M parts, that's how I found the correct M-tech springs for my e28. I've had them on the car now for about 5 months and I really like the feel and the lower stance.

I hope this info helps.

[DaRedRocket]

Very Nice. I'll make sure to save this.

[Tammer in Philly]

Devinder already made the point, but to all those people who still claim the E28 has "oversteering tendencies" out of the box and needs staggered tires ... note that just about every aftermarket manufacturer AND BMW Motorsport increased the rates of the rear springs relative to the front (although the M5 is still slightly softer in the rear).

-tammer

[cgraff]

Devinder already made the point, but to all those people who still claim the E28 has "oversteering tendencies" out of the box and needs staggered tires ... note that just about every aftermarket manufacturer AND BMW Motorsport increased the rates of the rear springs relative to the front (although the M5 is still slightly softer in the rear).

-tammer

Tammer...I agree with you 99%.

The set-up of the E28 stock is understeering. In fact, stock, the thing understeers like a pig.

BUT. The one thing that all the mags and testers in the 80s found to be true was the fact that it would have 'snap oversteer' tendencies in situations where the coefficient of friction would vary very quickly in dynamic situations. And this was exacerbated by the piss poor TRXs.

I remember my stock car having completely random snap oversteer in typical driving but in rainy conditions. This was considerably reduced after I switched out of the Michelins to HP all-seasons - but it was still there.

And it was this combination of random, snap oversteer with piss poor TRXs that all the testers and mags in the 80s were raving about - and is probably what spread the myth that E28s oversteer.

-Chris

[C.R. Krieger]

And it was this combination of random, snap oversteer with piss poor TRXs that all the testers and mags in the 80s were raving about - and is probably what spread the myth that E28s oversteer.

That and the fact that a surprisingly large number of people who should know better experience understeer and identify it as oversteer.

[cgraff]

Devinder,

Do you think you could get a consensus from the people who've supplied the data on the number of active coils at ride height? As in, could we get some observations from 'as installed' on the cars?

I think that this is a question that really needs answering before we can go further with this data. The amount of error you'd get in the spring rate calculation from a variation of 0.5 turns (e.g. on the stock front spring rate) is ~20 lbs/in (which is > 10%).

-Chris

[Mr. E28]

So is it safe to assume that the euro M5 front spring rate is the same as the US?

[ElGuappo]

Thanks Devinder.

As for what Chris is asking for, we can measure both the ST's I have on my black car (that you measured already, light blue) and my Dinan that are on the wrecked car when I haul it up to Johns.

Thanks for getting the chart up.

Jay3

[Devinder]

Chris,

I do use the mean coil dia in my spreadsheet.

Automotive suspension springs are hot coiled from standard Chromium-Silicon alloy (0.25% Cr and 1.8% Si). Even if a spring is made from some other spring steel alloy such as Vanadium, the modulus will be within 2% of the value I used.

I don't think there is a way to "get consensus" on the number of dead coils. The best we can do is measure it the same way each time. That way any systematic error doesn't affect the comparison of rates. This is why I presented data only for springs that I measured myself.

The very fact that the number of dead coils changes as the wheel moves up and down indicates that the spring rate is going to change with wheel height. If you can see rust on 2.5 coils, then you will probably see marks on the paint beyond the rust too. I chose the higher number because I'm interested in the "loaded" spring rate. In another words, I'm interested in the spring rate when the car is in a turn, rather than the rate when the car is going in a straight line.

As far as comparing to BMW's numbers, that is going to be difficult to do across the board. You can match a particular spring by choosing the number of dead coils that gives the same rate. But then you have to chose a slightly different number of dead coils to match the rate for a different spring. This is because BMW likely measures suspension spring rate according to SAE J1121 sec 2.11: "...one half the difference between the loads measured 25mm above and 25mm below the specified loaded length". Basically, they take the average spring rate over 2" of spring deflection around the nominal.

I think the best way to compare springs is by taking the same easily-repeated measurement for each spring. In the end, if you want to use 2.5 dead coils for 7-coil spings, that's OK. It's just going to change all the numbers by a comparable amount. Keep in mind that the manufacturing tolerance on hot coiled springs is pretty bad. You can see winding differences up to 1/3 coil in springs from the same set. Even SAE J1121 only requires 10% manufacturing tolerance (at best) on spring rates for automotive susension coil springs.

I have looked into the rate of pitch change at the ends of these springs. Maybe we can talk by phone ... too much to write.

Jay,

I'm not sure how I would measure the number of free coils on a car. Both ends (especially the upper one) are hidden when the spring is installed.

Mr. E28,

I would suspect Euro M5 front is the same since US M5/535i/528e are all the same. I can't say for sure since I haven't measured one.

Devinder

[cgraff]

Chris,
<SNIP>
I have looked into the rate of pitch change at the ends of these springs. Maybe we can talk by phone ... too much to write.

You make all good points. I guess it's one of those things that is just tedious (but not necessarily difficult) to reverse engineer, and tricky to try to get a reasonable comparison to aftermarket springs due to all the differences and variables.

I'm not sure how I would measure the number of free coils on a car. Both ends (especially the upper one) are hidden when the spring is installed.

You make all good points. I guess it's one of those things that is just tedious (but not necessarily difficult) to reverse engineer, and tricky to try to get a reasonable comparison to aftermarket springs due to all the differences and variables.

I'd love to chat at some point on the subject. Just been a tad busy with work et al. I'll PM you my cell in case you feel like chatting.

Mr. E28,

I would suspect Euro M5 front is the same since US M5/535i/528e are all the same. I can't say for sure since I haven't measured one.

Devinder

I would venture to guess that the euro M5 is going to be *slightly* different - and the reason I say this is because there were (I believe) 3 different spring PNs for the euro M5, based on a 'points' system as shown on the ETK. And all of these springs were different than the US M5/stock 535i.

Also, most reports from Europe at the time tended to make mention of lower/stiffer springs on the M5 - but I take it with a grain of salt.

Perhaps it might be useful for someone like Rich in Houston to measure his 'sprangs? At least to get comparative measurements.

-Chris

[Shawn D.]

Great data, Devinder! Thanks for the effort.

I likes me my Dinan springs!

[Mr. E28]

I would venture to guess that the euro M5 is going to be *slightly* different -

Also, most reports from Europe at the time tended to make mention of lower/stiffer springs on the M5 - but I take it with a grain of salt.

That's what I always figured. I know the springs differ quite a bit as far as height and thickness goes compared to '85 eta springs. My avatar is of my euro M5 springs next to my stock.

[cgraff]

Devinder,

Just FYI, I know this is not the most scientific method, but I went out and measured my front RD springs - on the car at ride height - the best I could:

~2.8 *active* coils
Coil OD: ~144 mm (I'm assuming same as stock)
wire OD: ~0.53 inches
mean coil diameter: ~5.14 inches
G = 11*10^6 psi

Spring rate = G(d^4)/(8(D^3)N) ~= 285 lb/in

That is not that far off what Ed Walters measured the RD springs way back when in the late 90s at 274 lb/in.

-Chris

[MShimon]

Does anybody have any spring rate data for the Eibach Sports? I'm just curious to get another comparison.

[tschultz]

Found this shared on a different website during some research. Enjoy.

[Bonsaibacker]

Thank you Tschultz. Had seen it recently and couldn't remember where I found it.