Newbie needs shocks/springs guidance

[aceswerling]

Hi, guys. I'm an engineer, but not an automotive one. I'm researching suspension changes for my 2000 BMW 540. I've found good information on "feel" for many options. But one man will describe a suspension as "firm" while another will describe it as "harsh." It's very hard to do a normalized comparison without real data. I'd like quantitative data before thinking about qualitative data.

For example, a suspension thread on another board did a great job talking about an installation. Then we get into a debate about Bilstein dampers vs. Koni and Intrax springs vs. everything else. I see arguments for and against all options. Another post argues for soft springs and stiff shocks. A suspension engineer once told me that's a recipe for disaster. It'll handle well but will rattle your teeth. He said BMW typically uses stiff springs and soft shocks, which is the opposite of Audi. What should I believe?

The threads discuss lower stance and stiff ride but nothing about spring or damping rates. Honestly, suspensions confuse the heck out of me. I don't *really* know how to interpret such numbers. But still, numbers allow you to do a straight up comparison. Subjective interpretations based on personal preference are interesting, but non-actionable.

To the specifics. I have a 2000 540 wagon with OEM sport suspension. It's tighter than the non-sport suspension but not at all harsh. It's a sophsticated ride but the shocks are shot. Like many people, I'm trying to decide between Bilsteins, Koni Sports, Koni FSDs, and OEM. I also wonder if I should install the Eibach springs matched to the Koni FSDs. I live in Florida so I drive on decent quality roads with no snow. I'm an enthusiast but this isn't a track car - it's a wagon. I don't care about lower stance. What are the definitive differences between my options?

Plus, I assume BMW spent a ton of time tuning the suspension. They made calculated decisions, and only then tuned on feel. Assume the OEM engineers really did target for the average driver. How would I know if I fall into the standard deviation for that average? Therefore, I want the engineering data *before* getting people's feel. I'm concerned about making things worse if I don't make a deliberate decision.

Any ideas? Thanks in advance.

 

[GregLocock]

BMW don't tune their shocks for the average driver. They tune their shocks for the average BMW driver.

That is not the same thing at all.

If you like it the way it is then just replace the shocks with identical parts.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[cibachrome]

When you are done with the shocks, go for the premier BMW feeling tire if yours are worn or some other brand: Continentals. Make sure they are the 'star' construction. Your dealer may not know or care about this tire deal but you should start off by knowing such things. Go search for what the " 'star' BMW tire" means.

Then you will be far ahead of most BMW owners...

 

[aceswerling]

I understand the star indicates OE tires approved by BMW and TUV. Basically, they're what shipped with the car. Is that correct? I don't see that I need to reinstall original OE tires. A lot of progress has been made since my car shipped in 2000.

Another point: Consumer Reports doesn't seem to like Continental. The BMW community tends to prefer Michelins, although I think they're overpriced. Wikipedia says Bridgestone is the largest tire company, followed by Michelin and Goodyear. That implies (but doesn't prove) that those manufacturers have more to spend on R&D. What do you like about Continental?

 

[cibachrome]

You said "feel". The Conti construction for the 2002-up 5 Series cars has 2 important measureable properties related to "feel" that ranks above all other brands used by BMW in that time frame. (Cornering stiffness is not one of the ones I'm referring to.

If you have a Control Engineering background, then this will have meaning to you, If not, then I can't help you.

P = 1/(p2s^2+p1s+1)
Q = qn3s^3+qn2s^2+qn1s+1/(qd3s^3+qd2s^2+dqs1s+1)

The Beamer Conti's mind their P's and Q's the best, followed by Pirelli and Michelin.

If you have even a student version of Matlab or Excel, the student should try these Conti params at 3.0 bar:
p2=0.0684
p1=0.528
qn3=0.00792
qn2=0.211
qn1=0.4051
qd3=0.00258
qd2=0.0593
qd1=0.6473

If your BMW is used to shuffle stock quotes around, this is not important. If you want to be one with the road, this is a magic key.

My question is whether BMW specifies these tire characteristics or do they float up from their exceptional subjective evaluation techniques and a lot of tire rides?

 

[hutch325]

Cibachrome, can you give me a nudge in the right direction with regards to your Ps and Qs? A hint as to what system these transfer functions drop into?

Sorry I'm slow to catch on but a long day of Matlab today has turned by brain into oatmeal.

 

[cibachrome]

p2=0.0684
p1=0.528
qn3=0.00792
qn2=0.211
qn1=0.4051
qd3=0.00258
qd2=0.0593
qd1=0.6473

fyn=tf(1,[p2 p1 1])
mzn=tf([qn3 qn2 qn1 1],[qd3 qd2 qd1 1])

step(fyn)
xlabel('distance (m)')

hold on
step(mzn)
title('The 16 inch BMW Conti''s are "Fast" tires.')

 

[hutch325]

Duh. Thanks, my brain definitely was oatmeal. Got any other Ps and Qs for comparison?

 

[cibachrome]

Conti_300 = [0.068352 0.52803 0.0079216 0.2114 0.40508 0.0025843 0.059262 0.64729 ]
Conti_250 = [0.080874 0.62856 0.0098625 0.22809 0.41877 0.0038361 0.074553 0.75643]
Mich_300 = [0.080713 0.5788 0.0074966 0.21517 0.38234 0.0022762 0.051449 0.52003]
Mich_250 = [0.11184 0.77321 0.011092 0.28379 0.29163 0.0039091 0.08319 0.65384]
Pir_300 = [0.084909 0.64914 0.007267 0.1611 0.24578 0.0033068 0.065196 0.66481]
Pir_250 = [0.057435 0.55253 0.0065171 0.15747 0.29006 0.0025284 0.05082 0.59305]

f1=tf(1,[Conti_300(1:2) 1])
f2=tf(1,[Conti_250(1:2) 1])
f3=tf(1,[Mich_300(1:2) 1])
f4=tf(1,[Mich_250(1:2) 1])
f5=tf(1,[Pir_300(1:2) 1])
f6=tf(1,[Pir_250(1:2) 1])

figure('Name','FY Rollout','Menubar','None','NumberTitle','Off')
step(f1,f2,f3,f4,f5,f6)
xlabel('Distance (m)')
legend('Conti 3.00','Conti 2.50','Mich 3.00','Mich 2.50','Pir 3.00','Pir 2.50')

mz1 = tf([Conti_300(3:5) 1],[Conti_300(6:end) 1])
mz2 = tf([Conti_250(3:5) 1],[Conti_250(6:end) 1])
mz3 = tf([Mich_300(3:5) 1],[Mich_300(6:end) 1])
mz4 = tf([Mich_250(3:5) 1],[Mich_250(6:end) 1])
mz5 = tf([Pir_300(3:5) 1],[Pir_300(6:end) 1])
mz6 = tf([Pir_250(3:5) 1],[Pir_250(6:end) 1])

figure('Name','Mz Rollout','Menubar','None','NumberTitle','Off')
step(mz1,mz2,mz3,mz4,mz5,mz6)
xlabel('Distance (m)')
legend('Conti 3.00','Conti 2.50','Mich 3.00','Mich 2.50','Pir 3.00','Pir 2.50')

Legend shows brand and pressure (bar). This is NOT cornering g's, but estimated 'relaxation'. Even a BMW driver probably can not put a pure step input into the tire, but a computer can easily give it a go. There is some speed dependence to it, but walk the track before you run it.

 

[hutch325]

I'm assuming in your view the Fy response is more critical to feel than Mz since the Fy response of the Conti at 3.0 bar is relatively quick, while the Mz response seems to be in line with the rest of the pack?

 

[aceswerling]

FWIW, BMW people don't seem to like Continental's performance. This is regardless of ride.

http://www.m5board.com/vbulletin/showthread.php?t=102483

 

[cibachrome]

FWIW: The issue was road feel in a 2002 vintage chassis. The 2002 Conti Constructions have been matched by GDY, BRI and Hankook traits but I thought the question was about the 2002. Most forum members have never run 5 brands, 4 construction recipes from each brand, plus the reference tire through their car to see which gives them the best perception. They've only experienced a new vs. worn. Most probably with a non-OEM construction optimized for mileage. If the issue is grip, then this is a different solution. IMHO: If the subject is road feel, the 2002 5-Series delivers best on the 16" Conti's. And, these tires work pretty well on other high front caster, high KPI RWD leading steering cars with low friction gears and stiff I-shafts.

Ully

 

[hutch325]

Okay so on a tangent, what tire construction and performance characteristics seem to work well on a "high front caster, high KPI, RWD, leading steering car"? That configuration describes just about every BMW. Except depending on model, some have more compliance understeer built into the steering system than others. (FWIW, I have a BMW as my "weekend" car so I find it interesting.)

No need to name brands, I'm just wondering what your thoughts are with respect to what would work best with this configuration, and why. Also, why it might work better on a car with this configuration as opposed to something like, say, a double wishbone front / multi-link rear, rear drive car.

 

[cibachrome]

High sidewall stiffness from construction details. If cornering stiffness comes from these instead of tread/compound, transient response is better. Hence perception is better.

 

[hutch325]

Yes, that makes sense.