descartesfool

Anyone see the Mark Ortiz column in the April 2009 issue of Race Car Engineering?

Question was about picking the right damper curve: "And what are your thoughts about setting the correct damping on a damper and about building it's curve from scratch"

I thought the answer was about as obscure as it could be. No answer given about anything to do to set a damper curve from scratch. Damping ratio (ratio of actual to critical damping) could be from 0.4 to 0.8 according to speed for damper to return to its initial position, so 0.6 to 0.7 might be good. But dampers have curves, with different slopes for low and high speed shaft motion, and different for bump and rebound, so knowing a damping ratio range is a bit of an incomplete answer I think. Anyone have an idea on how the damper manufacturers pick damper curves for a given application? Is there any science other than copying what worked for someone else? I just copied because I had no better idea. Anyone have an idea on how they choose?

Obviously there is a little more to dampers than a force vs velocity curve, but one has to start somewhere.

Andrie Hartanto

There are no set answer on what kind of damping curves will work for you. Depends on a lot of factor. If you have not already read shock absorber handbook, you should. Pretty interesting.

RR98ITR

Totally agree - No Set Answers.

And also totally agree with my own extrapolation of what Claude said - scotch yoke force / velocity curves don't tell the whole story.

Gooooooood Luuuuuuuck.

Scott, who has driven the same spring rates with as much as 800 lbs of high speed compression and as little as 100 lbs...very interesting...I still haven't figured out what I know besides that I was able to drive either pretty effectively however somewhat differently...and my preference was for rather less than more...

beanbag

Any way for me to read the article?

vinuneuro

I've read a few times over that .65-.70 is what you want for low-speed damping for most production cars. .30-.50 for high-speed, but I could see this varying quite a bit based on the track.

Low-speed bump and rebound are also supposed to be best if they're symmetrical for low-speed.

2-3in/s seems to be the cut-off for low/high speed.

descartesfool

I re-read the article a few times as well, hoping I could find something useful, but it wasn't there. (RCE don't usually have their articles online anywhere).

While I know there are a ton of issues to consider, the damper valving guys have to start somewhere. what data do they use and how do they choose the valving based on the data?

Basics:
FWD or RWD or AWD
Total Weight
CG location
Corner weights
Unsprung weights
Spring rates
Sway bar rates
Motion ratios
Wheel rates
Suspension travel
Aero
Bumpy tracks or smooth, curb riding
Roll center heights
Suspension geometry, SLA, Strut?
Driving style
etc.

If you had the above data, an empty damper, a bunch of pistons and shims and valves, and a damper dyno, what is your initial target?

RR98ITR

Two options:

One - start with some value and make as many calculations as it takes to satisfy you that you have arrived at "the answer". Then, unless you've got somebodies builder program or notebook full of curves and stacks, you get to iteratively build and dyno till you get the curve that corresponds to "the answer". Then you put it on the car and drive it, maybe even test it against alternate builds that must necessarily be "wrong".

Two - buy or borrow something known to work and clone it. From there you might test alternate builds to see if improvement lies down one path or another. Or just build something linear with some guesses for values at X inches per second...like 50 lbs compression at 3 inches per second and 150 lbs rebound at 3 inches per second. Then maybe try tripling those values and see how that feels in comparison.

From my very limited exposure I don't think there are very many club racers or amateurs that have the time or money to truly exhaustively build a mastery of this art/science. After a point the money you might spend on shocks and testing is probably way better spent on engine development which is an often overlooked part of a lower laptime.

Scott, who doesn't mean to be a party pooper and encourages anyone who really wants to party to party on...ain't no harm in it...

slammed_93_hatch

I'd say most "pro" drivers don't either...

They buy XYZ shocks, and XYZ shock guy is at the track. Driver/racecar eng/car cheif tells XYZ shock guy that A, B and C are happening and asks him to fix it...

beanbag

It's too bad that I don't have a copy of that Shock Absorber Handbook. Browsing thru the table of contents, it looks very useful. If I read it, then I could say smart sounding things backed up by SAE articles and equations. Until then, I'll have to go based on my experience revalving Bilsteins. Here goes:

The Bilsteins are basically 6 way adjustable shocks, only it takes 2 hours to make an adjustment. You get to choose low speed damping, knee position, and high speed damping on their COB digressive pistons. On their old skool digressive pistons, they are only 5 way adjustable, since you loose the ability to adjust either comp or reb low speed damping. They also have linear pistons, but I haven't run them yet.

First thing to note when building these shocks is that you will never get the smooth digressive curve that you want. The way these shocks work (and probably a lot of others) is that the dyno plot is made up of a low speed bleed, and depending on the preload of the main shim, there will be a knee at which the high speed part takes over. Then the thickness of the main and high speed shim set the slope of the high speed segment. The problem is that bleed behavior is progressive, so your "digressive" curves will always look progressive and then linear. That means that the differential damping (slope of the damping curve) rises, and then falls. (Some other shocks with different pistons designs might be better about this. For example, the Koni yellows seem to have a "smoother" looking rebound curve, but I don't know what their valving looks like. Even then, I don't like the behaviors of these shocks, but that's another matter.)

Anyway, first thing I like to use as baseline is to set the low speed damping. For rebound, I really do like around 70%, and then add a touch more if you have a big swaybar and want to damp that too. The trickier part is setting the knee location. I look at shock velocity data to determine what are the characteristic speeds of large chassis motions, such as swerves, and recovering from big dips or wallops in the road. I found swerves to be about 5ips max, and big wallops/dips to be about 10 ips. A common pitfall is that some people set the knee too low (for example 2ips), and have a big digressive blowoff. Then they find that they have too little damping (because they are sampling over the 0-4ips range, for example) so they keep cranking up the low speed damping, until they are close to 100% critical.
Below I show two dyno curves, of Koni Yellow (orange) and 8211(blue) which were running the same springs and adjusted until they "felt right" on the car. Note that the much more digressive behavior and low knee caused me to crank up the low speed rebound.


I've found two ways to tell if you are running into this problem. One is if you feel like "the harder you drive the car, the better it feels". Then I might say you have too much low speed damping. That's because when driving "easy", you stay at low speeds, where you are overdamped, and something feels wrong. Then when you drive harder, you reach higher shock speeds, and the digressive nature of the curve "helps average down the damping" till it feels better.

The other way, is that after hitting a big wallop or dip (note I didn't say sharp bump), that the car seems to spring back really fast, and then suddenly stops. It feels a little weird, and I'm not sure it is necessarily bad.

Next up is setting high speed rebound. Sorry, I don't understand HS rebound, and I haven't messed much with it. I suspect that it is to help the wheels extend and track over a bumpy surface (unsprung mass). I have read somewhere that the bumpier the road, the less HSR you should use. I dunno since I haven't tweaked much with this, except to say that the Koni Yellows have much more HSR than my Bilsteins, and they are noticeably suckier at tracking bumps in the road.

Setting compression damping is trickier, and probably very application specific. I have heard that you should use more compression damping on your drive wheels, although I don't know why. A lot of setting compression damping has to do with the tradeoff of chassis stability (getting thrown around by bumps) vs ride comfort. I find that small bump ride comfort is determined by LS comp damping, chassis stability is determined by LS and medium speed, and bottom out resistance by medium speed and high speed. Perhaps LS C helps with turn in and HSC is to help prevent the unsprung weight of the wheel from getting pushed too far into the chassis after a big bump? I have read that you should stay above 1/2 to 1/3 of your rebound for compression, and that seems mostly correct, although I don't try to match the shapes.

Also, this is one area where twin tubes and monotubes differ. If you look at my last post on the compression damping thread (https://honda-tech.com/forums/showth...2374810&page=2, where I complain that Koni Yellows have too little compression damping, and instead everybody pisses on me for my spring rate) I show that Koni twin tubes have this behavior where on accelerating strokes, the compression valving actually blows off early and you get much less compression damping than you expect. In a way, this is a feature because you can set a large LS comp for stability, yet it's still comfortable over small bumps. Similarly, my 8211's stock have a LOT of LSC, and just as I have mentioned for rebound, I found the chassis stability to be good, but the ride quality to suck, until I drove the car harder. Again, if the car feels better the harder you drive it, you have too much low speed damping. When I revalved the compression side on those to decrease the LSC, I was able to set a higher knee point, yet retain both ride comfort and chassis stability. Those were very comfortable shocks.

Now try valving monotubes to have the same LSC as LSR, drive where I do, and you will be in a world of hurt. Although I have seen that some Miata autocrossers like this kind of valving, but autocross people do weird things with their shocks anyway, so who cares.

In any case, I am currently running a very linear compression curve on my Bilsteins, I find that they give good ride comfort and reasonable bottom out resistance. Even though if you look on a dyno plot, the Steins have much less compression damping than the Koni Yellows, the bottom out resistance is much better.

Finally, I have found that I only needed to adjust the rebound damping curve about 3 or so times until it felt about right, whereas I have played around a lot with the compression damping curve, and it's almost right, but not quite. Oh right, so driving style. Anyway, assuming that there is pretty much a "right rebound curve" then low speed compression damping also affects driving style in that it affects the speed of weight transfer vs chassis lean when you turn in. I don't really care about this issue since I don't saw the wheel back and forth like the autocross people do.

descartesfool

An interesting set of damper motion charts illustrating effects of resonance frequency values and % critical damping here (along with a few more articles):

http://buildafastercar.com/tech/Damp...Frequency-Rear

some other good info here on spring rates/frequencies: http://www.worksevo.com/Spring_Rates_1.pdf

and damper curve choices: http://www.worksevo.com/Damper_Curves_2.pdf

This from an unlikely source: http://www.theoryinpracticeengineeri...sic_damper.pdf
and this: http://www.theoryinpracticeengineeri...per_tuning.pdf

Andrie Hartanto

Don't hold the 2.0-2.5 Hz frequency like gospel. As long as the damper can handle it, you can have whatever high spring frequency.

RR98ITR

There's a magic frequency you don't want to hit a'cause it's the frequency of your inner ear and you will think to yourself that the handling of this car makes me want to vomit. Least that's what I 'Ear.

Scott, who doesn't know about any other frequency that makes you suddenly evacuate your waste storage tanks...

beanbag

There's also the "brown note"

Andrie Hartanto

I heard about that. But it is borderline myth, IMHO.

Seriously, the job of the damper is to dampen the frequency. In general, it doesn't really matter what are the frequency as long as the damper is up to the task.

You just don't want the front and rear frequency to be the same.

descartesfool

I am good on the 2.0 to 2.5 Hz not being gospel. As I am running 1100 lb rear springs on my ITR, so I figure I am around 3.5 Hz already without considering the sway bar. I would like to know for example why the dyno curve on a rear Mugen ITR damper for example is virtually linear while the front dyno curve has a ton of low speed bump and then levels off. Where did that come from? Some voodoo magic I guess.

Found a damping spreadsheet here for Univ. of Coventry:
Damping: http://web1.eng.cov.ac.uk/BPorter/damping.xls

others from the same author: http://web1.eng.cov.ac.uk/BPorter/downloads.htm

RR98ITR

Too Funny. I just read the inner ear thing pegged at 7.5 (?) from no less than Claude Rouelle fairly recently in connection with some FSAE stuff. Andrie - I think you are in ALOT of Trouble! Even with my ludicrous rates (which I'm decreasing before I drive again) I'm not close, but I think some of the F1000's and DSR's are close. At their g loads and ride frequencies I can understand how those aren't enjoyable rides apart from the absolute speed and performance.

Ouch, I hurt myself rolling on the floor. That thing about keeping front and rear differentiated is borderline myth in racing however having legitimate roots in passenger comfort engineering. But I think one would have to try deliberately to hit equal frequencies, and for what good reason I can't imagine. Ok, won't try to imagine.

Scott, who would answer Claude (our Claude) by suggesting an inherent bias toward transient understeer?...

solo-x

Some random comments....

I've set the ride frequencies on my car to roughly 2.3hz front, 2.5hz rear, following that old ride comfort rule of thumb. The car is notably easier to drive on bumpy courses, much more comfortable to drive to events, and feels more "natural". IE, when it does hit a large bump, the car isn't nervous, doesn't purpoise, etc. No change in dampers either, just springs. The old setup was 2.1hz/3.3hz and it wasn't pleasant to drive on bumpy courses. Not saying one was is right or wrong, just putting it out there.

Transient understeer is good. Especially once you get steady state balance to lean towards oversteer.

As ride rate increases, I think the damping curve should start moving towards linear. Lots of low speed + stiff springs/swaybars may cause spikes in the tire contact patch that could be avoided with less low speed. Less low speed without changing the high speed region would move the curve towards linear. I can't see any reason for a progressive curve unless you've got a geometry that causes some weird changes in motion ratio as the suspension deflects.

Here's my guess on getting a damper curve to start with. When starting with a clean slate, a velocity histogram generated on a shaker rig will get you your range of velocities that the damper will see. Combine that with a bunch of "rules of thumb" to get a starting point. Put it back on the shaker and look at the histogram again. If it "looks right", put the driver in the car and run some test laps. Season for flavor from there. Note that I'm not a shock engineer though. Were is GSpeedR? Wasn't he a shock guy for an FSAE program? IIRC, he built some dampers from scratch, designing the piston, shim stack, damper curve, etc. himself, but I may be remembering wrong.

RR98ITR

GspeedR (Chris?) is major thumbs up...hope he and whatever Nascar team he's with are doing alright.

Scott, who liked something in a post above relating to feeling the characteristics of a specific curve and driving around them...that's good training...

descartesfool

Nate.

Perhpas I need to get myself one of those "shaker rig" thing-ama-jiggies to start from scratch. Any chance I can get one on eGay?

RR98ITR

I got a 4 poster static setup. But for development it's hard to beat 4 tires on the track...that is if you love to drive.

Scott, who has had more than a week during which he could have worked straight on the car...instead I'm on my 4th book and trapped in Moscow in 1941...

solo-x

I did say it was my guess. I could be wrong, but my theory would at least explain why we can't find anything more concrete. I did my damper development by using a known good starting point, then figuring out what I want to change to improve something. Fortunately, my damper of choice comes from some people who've got a pretty damn good record and would appear to know what the curve should look like.

descartesfool

Lots of info on dampers in the Claude Rouelle seminar here:

http://www.theoryinpracticeengineeri...eminar_33M.pdf

and other stuff here:

http://www.theoryinpracticeengineeri...urces/dampers/

descartesfool

A nice paper here: http://www.theoryinpracticeengineeri...parameters.pdf

rrobb

Has anyone found one? I hear 7-post is the way to go though, not 4-post. We should be able to pick them up for, what, like 200 bucks, to the door? I mean if the F1 teams can afford 'em. Right?

Cory