Damper Histograms

[ben]

When I went on a Claude Rouelle seminar he explained damper histograms and that a symmetrical distribution was desirable.

At the time he mentioned a document he would be prepared to send us explaining in more detail the theory behind why this is the case. I've emailed him a number of times to get that document and received nothing.

Two questions really, if anyone has this document would they be prepared to send me a copy? Secondly, what is other people's understanding of why damper histograms should be symmetrical?

Ben

[ben]

When I went on a Claude Rouelle seminar he explained damper histograms and that a symmetrical distribution was desirable.

At the time he mentioned a document he would be prepared to send us explaining in more detail the theory behind why this is the case. I've emailed him a number of times to get that document and received nothing.

Two questions really, if anyone has this document would they be prepared to send me a copy? Secondly, what is other people's understanding of why damper histograms should be symmetrical?

Ben

[Lukin]

I think when Claude says 'e-mail me and I'll send it to you' he actually thinks 'well this will shut them up.'

As for the histograms, I can't tell you that either (don't know why Im replying really). I don't think it's vitally important, more just a point he is comfortable with and a trend he has come up with over time. After all, the stopwatch tells the ultimate story. If there is a huge difference in bump and rebound which we didn't expect, then I'd be a little suspect.

But overall I just look at the histrogram more to see the resulting differences after a shock setting change, not to look at it and say, 'we need more/less of .....' But then again, I'm not a race engineer either.

[ben]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Lukin:
I think when Claude says 'e-mail me and I'll send it to you' he actually thinks 'well this will shut them up.'

As for the histograms, I can't tell you that either (don't know why Im replying really). I don't think it's vitally important, more just a point he is comfortable with and a trend he has come up with over time. After all, the stopwatch tells the ultimate story. If there is a huge difference in bump and rebound which we didn't expect, then I'd be a little suspect.

But overall I just look at the histrogram more to see the resulting differences after a shock setting change, not to look at it and say, 'we need more/less of .....' But then again, I'm not a race engineer either. </div></BLOCKQUOTE>

Hi Jason,

I think you're probably right. There was a guy from Jaguar Racing on the course when I went, and Claude just couldn't convince him in any way shape or form why we should care how symmetrical a damper histogram should be.

Superficially the idea is appealing and I'm sure Claude wouldn't advocate it if he hadn't made a lot of racing cars go quicker using it. Claude talks about taking a scientific approach to racing and I'd like to know what the hypothesis is in this case.

Ben

[Chris Boyden]

What are they histograms of? damper position or velocity?

[ben]

Sorry - velocity.

Ben

[Storbeck]

I was there in 04, asked him for more explanation, he said he'd email me a more detailed mathematic explanation, I never recieved anything. I also didn't get the spreadsheet he said he'd send, but that's not so hard to make on your own, no big deal. I did try to email him about it and got no response.

I suspect it's his way of shutting people up. It's interesting to see that other people have reached the same conclusion.

[jack]

yeah, i emailed him (not about dampers) and never got a response, but i think someone else at WWU talked to him through email.

[Z]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by ben:
... what is other people's understanding of why damper histograms should be symmetrical?
</div></BLOCKQUOTE>
Dampers are "crutches" that you put on a car when its suspension is not quite right.

If you have a broken leg then you can run faster with crutches under each arm. If one crutch is shorter than the other, then you hobble around in circles. Symmetrical crutches are better...

Of course, if you have healthy "suspension", then you go faster with no crutches at all. http://fsae.com/groupee_common/emoticons/icon_smile.gif

(Hint: Would a really clever "active" suspension need any damping? Remember, damping = friction.)

Z

[Angry Joe]

What exactly is wrong with friction here?

And you're trying to be funny when you say dampers are crutches, right?

[Wayne Gross]

The reasoning behind having the symmetrical damper histogram is to be able to compare the balance of the vehicle. By comparing oversteer/understeer of the car, the weight transfer is usually easily monitored with the histogram. You can make adjustments easier when you know which damper is spending the most time in a certain range. Also, with dampers spending the majority of their time in the low speed section of the histogram, a symmetrical motion is often preferred to control weight transfer. Weight transfer issues can usually be traced to the damper/dampers which are spending an uneven amount of time between low speed bump and low speed rebound. On a side note, low speed is the most important area of damper control, and also the most easily controlled by the damper.

[mtg]

First off, if you email Claude and don't get a response, it is not because you are being blown off.

Claude gets so many emails with technical questions everyday it's not even funny. If he answered all of them, he wouldn't have time to give any seminars. You'd be surprised how much time we spend at OptimumG answering technical questions.

The symmetrical damper histogram question has come up quite a few times. There's a short explanation in a previous tech tip:
Tech Tip (http://www.optimumg.com/tech_july.htm)

This looks like a good topic to bring up at the FSAE seminar next month for those that will be there.

[ben]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by mtg:
First off, if you email Claude and don't get a response, it is not because you are being blown off.

Claude gets so many emails with technical questions everyday it's not even funny. If he answered all of them, he wouldn't have time to give any seminars. You'd be surprised how much time we spend at OptimumG answering technical questions. </div></BLOCKQUOTE>

I totally appreciate that, I wasn't taking it as getting blown off, I was just interested in getting an answer.

I hadn't seen the Tech Tips on the website before. Would it be better to put the spreadsheets and other things that he says he'll send people on the website? Would it also be possible for you to post the more detailed explanation of why the histogram should be symmetrical somewhere?

I can see that a simple 1dof vibrating system with some damping will exhibit this exponential decay, but I'm not willing to make the leap to saying I should want that on a race car without knowing what assumption are being made about the tyre.

A good example is Claude's explanation of load sensitivity in the seminar, which is based on Schallamach's work as presented in the DOT tyre design book. If assumptions like that are part of the damper histogram philosophy I'd be interested to know.

Ben

[Schumi_Jr]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
Dampers are "crutches" that you put on a car when its suspension is not quite right.
</div></BLOCKQUOTE>

Z,

I think you're off on this one. Dampers are necessary to "damp" oscillations of the sprung and unsprung mass (wheel hop). For FSAE cars dampers should be used to minimize grip disturbance over bumps. For aero cars there is the addition of maintaining a flat ride (minimizing aero disturbance). This is a big part of why race teams use 7-post test rigs- they can quickly measure the effects of damping on contact patch grip disturbance.

If anyone is interested in using their dampers "properly" there are some great SAE papers on the topic.

[Chris Boyden]

I think Z's trying to point out, please correct me if I'm wrong, features of active suspensions, like the Bose suspension.

There's no dampers in that suspension. Just high speed control processing and high force linear motors that keep the wheels planted and the car flat. Energy put into the suspension from a bump generates electricity in the linear motors/generators that is collected by the power amplifiers, kind of like a Class D audio amp. Class D audio amps are super efficient, like 97%. They're not wasting that energy by dumping it into a resistor.

[Schumi_Jr]

He said "damping". That system has damping, it's just built into the control algorithm. An active suspension still behaves like a spring-mass damper system, just with variable spring and damping rates.

[mtg]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by ben:
I totally appreciate that, I wasn't taking it as getting blown off, I was just interested in getting an answer.

I hadn't seen the Tech Tips on the website before. Would it be better to put the spreadsheets and other things that he says he'll send people on the website? Would it also be possible for you to post the more detailed explanation of why the histogram should be symmetrical somewhere?
A good example is Claude's explanation of load sensitivity in the seminar, which is based on Schallamach's work as presented in the DOT tyre design book. If assumptions like that are part of the damper histogram philosophy I'd be interested to know.

Ben </div></BLOCKQUOTE>

The spreadsheet will probably be emailed out. Part of why it takes awhile to get this stuff out is because we'll be in the middle of working on it, and then you split town for the majority of a week to go do consulting for a raceteam, or we have a seminar coming up that needs preparation, etc. Those type of things take the priority, so sometimes the question answering takes awhile.

Ironically, the guy that's been working on the spreadsheet is out of town right now working with a raceteam.

If you have burning questions, the monthly Tech Tips are a good thing to check- a lot of times popular questions are answered there.

[Z]

Re: Dampers as crutches.
=======================

Angry Joe,

Actually I am being serious. There is nothing wrong with friction in itself. But when you use it to dissipate expensively acquired kinetic energy, which is what dampers do, then I think you should be looking for a better alternative (rather than looking for more expensive dampers http://fsae.com/groupee_common/emoticons/icon_wink.gif). You probably won't feel it on the millpond smooth FSAE tracks, including Detroit http://fsae.com/groupee_common/emoticons/icon_smile.gif, but high damping forces on off-road cars cause a noticeable drag, and they also wear out the car faster. Since dampers increase the tyre-to-road normal force on the uphill side of the bump, and reduce it on the downhill side, they also lessen overall grip (as well as increasing drag).

Schumi-Jr,

Chris has got the idea. I am not familiar with the Bose system, but it sounds like the type of thing I was thinking of. "Dampers", as usually defined, give a force that is ALWAYS in the opposite direction to the motion. So they always dissipate energy. (Whereas springs absorb and then give back the energy.) If you have clever springs (ie. "active") then by varying their rate the control system can suppress oscillations WITHOUT DISSIPATING ANY ENERGY. Smart springs can also maintain more even tyre loads than spring/dampers, because, as mentioned above, dampers always make tyre loads less even over bumps (ie. less even than the ideal case of no damper and no oscillation). As I said, dampers are fitted to cars because the springs are not smart enough to suppress oscillations by themselves. The dampers suppress oscillations (sometimes, if done right...) and help the car go faster. So they are sort of like crutches... http://fsae.com/groupee_common/emoticons/icon_smile.gif

It bemuses me that for most of the history of suspension design, the designers have used very simple (dumb) springs, and then tried to compensate for the simple spring's inadequacies with either complicated control arm systems (for camber compensation, etc. - eg. see "Dax" thread), or complicated dampers (eg. some modern active dampers). The suspension is there to let the wheel move up and down, so, to me, it makes most sense to concentrate the development on finding the best "springs" to control the up-and-down motion, and then use simple control arms and as little damping as possible. Multi-rate and fully-interconnected springing goes some way towards that, with relatively simple, passive components that don't waste energy.

Ben,

I guess the goal of "minimum damping" (also held by some other people!) goes some way to explaining the symmetrical histograms. You get the most damping (area under curve), with the lowest peak damping force, if the curve is symmetrical. This applies to a single wheel's curve and also to all four wheels (ie. you don't want one stiff damper trying to do all the work).

Z

[CMURacing - Prometheus]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Schumi_Jr:
If anyone is interested in using their dampers "properly" there are some great SAE papers on the topic. </div></BLOCKQUOTE>

care to recommend a few, as well as any someone might have on damper design?

[mtg]

If anyone wants proof of dampers causing "drag" like Z mentioned, go ride a full suspension mountain bike. Then ride a rigid one.

[RiNaZ]

mtg, been riding my bicycle to school for the last 6 years. I had a full on suspension and a really cheap rigid bike with no suspension at all. Both had 26" tyres.

My opinion is ... the rigid is much faster, and easy to control. So i think having suspension makes more "drag" than not having one.

Is that what you're trying to get at in your last post?

[mtg]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RiNaZ:
mtg, been riding my bicycle to school for the last 6 years. I had a full on suspension and a really cheap rigid bike with no suspension at all. Both had 26" tyres.

My opinion is ... the rigid is much faster, and easy to control. So i think having suspension makes more "drag" than not having one.

Is that what you're trying to get at in your last post? </div></BLOCKQUOTE>

Yes

[Denny Trimble]

If you ride on a smooth road, that is true.

If you're riding offroad, there's a certain level of roughness where suspension is faster. I've been smoked up a climb by a skinny guy on a hardtail, then passed him on a flat, very bumpy section because I could sit down and pedal. He had to stand to absorb the bumps.

Also, yes dampers do dissipate energy, which slows you down, but springs store and release energy. With a suspended bike, the normal force on the tire is less on the uphill side of a bump, and more on the downhill side of a bump, which both make you faster.

Of course a bike rider can be considered active suspension. BMX racers don't need suspension because the tracks are smooth enough that the riders can work the bike over the rollers and jumps to their advantage, actually accelerating over the whoops (/\/\/\/\/\).

Oh, and I set my damping very low on my MTB, by the way.

[Z]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Denny Trimble:
BMX racers don't need suspension because the tracks are smooth enough that the riders can work the bike over the rollers and jumps to their advantage, actually accelerating over the whoops (/\/\/\/\/\). </div></BLOCKQUOTE>
Another place to see the advantage of "active springs" is at a skateboard park. The riders lift their legs on uphill sections, and then push down on the downhill sections, thus providing forward thrust, even though the wheels are freewheeling (similar to Denny's above quote). Likewise, snow skiers going over the moguls (this picture is often used to illustrate active suspensions). Also similar, though in the horizontal plane, is the mechanism by which ice and roller skaters provide forward thrust.

Previously I said that stiff dampers on off-road cars cause a "noticeable" drag. Perhaps a better word would be "considerable" or even "extreme" drag. It is really like driving through thick mud! And all that energy dissipation goes into bashing up the bumps, and the car as well http://fsae.com/groupee_common/emoticons/icon_frown.gif. Very soft springs and no dampers almost glides over the bumps. Trouble is, occasionally you get a bit of a resonance happening, either of a wheel or the body, and the cheapest, quickest fix is to fit some dampers. The softer the springs, the softer the dampers (critical damping being proportional to sqrt(K x M)).

One very simple form of "smarter spring" is longitudinal Z-bars (no relation http://fsae.com/groupee_common/emoticons/icon_smile.gif) connecting each side-pair of wheels. These provide stiffness in the 4-wheels' bounce and roll modes while offering no resistance to pitch or twist (aka warp) modes. A third soft Z-bar at the front or rear axle will give a soft pitch mode while still having zero twist mode stiffness. Softening the modes that don't have to be very stiff for circuit racing allows softer dampers to be fitted.

BTW, Denny, if you want to experiment with mountain bike suspension then you might try an interconnected suspension that gives stiff bounce mode and soft pitch mode. The soft pitch mode is great at absorbing bumps (which hit the wheels one at a time), while the stiffer bounce mode doesn't drain your energy when pedalling. A hose connecting the two dampers might be the easiest way to start.

Z

[clausen]

I haven't put a lot of thought into it, but doesnt a symmetrical histogram mean that the dampers arent jacking the car down or up?

[Cement Legs]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
BTW, Denny, if you want to experiment with mountain bike suspension then you might try an interconnected suspension that gives stiff bounce mode and soft pitch mode. The soft pitch mode is great at absorbing bumps (which hit the wheels one at a time), while the stiffer bounce mode doesn't drain your energy when pedalling. A hose connecting the two dampers might be the easiest way to start.

Z </div></BLOCKQUOTE>

Well that sounds similar to Giant's NRS(no resonance system) suspension design; however, the front and rear dampers were not linked. I believe they dramatically reduced the induced bob of the suspension from pedalling by the angle and geometry of the rear linkage while sill allowing full active rear tralvel for bumps comming up from the road.

[Schumi_Jr]

Mike,

Start with the paper 2002-01-0804 "Damper Tuning with the use of a Seven Post Shaker Rig". It also includes analytical methods for those that don't have a seven-post rig at their disposal!

[Garbo]

The reason that a plush full suspension bike is hard to pedal on a street is because it is not designed to be ridden on the street. The suspension is to keep the tires on the dirt when riding off road. Last time I checked, the Tour de France riders were all on rigids. This means that the cobble stones in Paris cause problems due to normal load variation on the tires (does this sound familiar?). Since the FSAE course is like a washboard, we have similar problems.

Also, the effect of 'drag' due to full suspension on an MTB is due to the fact that your legs (a significant portion of the overall mass) are making the power so, as your body weight shifts about, the bike will 'bounce'. Also, the tension in the chain can excite the rear suspension which is the reason for Giant's NRS and other VPP (virtual pivot point) designs. A recent mountain bike magazine explained this.

garbo

[racer_chick]

Hey guys,

The engineers here are working on that spreadsheet right now, and by the end of the week I should be sending it out to everyone who has attended the FSAE seminars.

FYI we are having another one at OptimumG on October 1-2. http://www.optimumg.com/seminar_fsae.htm

Karen

[John Bucknell]

Z,

Call me obtuse, but I just can't visualize your longitudinal z-bar. I can envision configurations that link opposite corners and provide stiffness in 4-wheel bump, little pitch stiffness but not something that does all the things you describe. An internet search provided zip - could you draw a schematic?

[Angry Joe]

It seems like the bigger issue here is this: a 'dumb' suspension is a compromise for many requirements and conditions because it reacts the same way to any input. A skateboarder, like an active suspension, can adapt on the fly for any bump, ramp or whatever. Replace his brain with a command that says "bend knees at 100 lbs/inch" and you end up with the same problem.

You could set up a desert racer to perfectly absorb a particular bump without upsetting the truck at all - but it will be crap everywhere else.

Friction may be a concern, but it seems to be secondary to other issues.

[Frank]

hey racer chick,

heres a spreadsheet

http://www.uq.edu.au/fsae/frank/Suspension_Statics.xls

regards to Mr Rouelle

[Lukin]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by racer_chick:
The engineers here are working on that spreadsheet right now, and by the end of the week I should be sending it out to everyone who has attended the FSAE seminars. </div></BLOCKQUOTE>

Any chance I could get a copy too? I did Claude's non-FSAE seminar July 25-29 in Melbourne this year.

[racer_chick]

Yeah, don't worry. We'll be sending it out to all of the seminar attendees, FSAE or not. I think it is done today, just waiting for the final OK before I send it out.

Cheers!

[Pavan Dendi]

Will that include previous years? I attended the Roulle seminar for FSAE 2004 and never received a spreadsheet. I never bothered to pursue it, but I would appreciate a copy.

monqy_at_mail_dot_utexas_dot_edu

[Z]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by John Bucknell:
Z, ... I just can't visualize your longitudinal z-bar... could you draw a schematic? </div></BLOCKQUOTE>
John,

A sketch would explain it, but difficult to post on this forum http://fsae.com/groupee_common/emoticons/icon_frown.gif ... I've got an idea I'll try later... Meanwhile, here are some words.

"U-bar" and "Z-bar" are shorthand names for spring elements that interconnect a pair of wheels. The central section of the bar is a torsion spring that is mounted on the chassis, and at each end it has a lever-arm, either bent in the same direction (U) or opposite directions (Z), with the ends of the lever-arms attached to the wheels.

A "U-bar", like a typical anti-roll-bar, resists differential movements of its ends (one up, the other down), but offers no resistance to similar movements (both up, or both down). A "Z-bar" resists similar movements, but allows differential movements.

A leafspring that is pivotted to the chassis at its centre, and connected to a wheel at each of its ends, is also a Z-bar (think about the central section of the Z getting shorter until it disappears).

U-bars and Z-bars can interconnect any pairs of wheels. Standard ARB's are U-bars that interconnect front-pair and rear-pair of wheels, hence stiffen up 4-wheel roll and twist modes. If the Z-bars "that link opposite corners" that you mentioned linked diagonally opposite corners, then they would stiffen bounce and twist modes - not really desirable 'cos twist should be soft.

My suggestion of side-pair Z-bars, on a FSAE car, would be something like this. A torsion bar running down each side of the car, about 1/2" diameter (maybe tubular) and as long as the wheelbase. A lever arm about 2"-3" long at each end (and cranked in opposite directions), attached via pullrods to the upper wishbones. Thus each longitudinal Z-bar resists its side of the car going down, so resists 4-wheel bounce and roll. But no resistance to pitch or twist (ie. front wheel moving up, while rear wheel moving down). A third Z-bar, perhaps as a centrally pivotted, transverse leafspring, at the more heavily laden rear wheels would provide pitch (and bounce) stiffness (could also have similar at front, but not necessary).

The Citroen 2CV, designed in late 1930's, released post WWII, had this sort of system, although somewhat different pitch control. The 2CV "Z-bars" are actually coil springs and tie rods that link leading (F) and trailing (R) arms, but same principle. BMC cars of 1960's+ had a similar hydraulically connected system ("Hydralastic" - rubber springs, and "Hydragas" - pneumatic springs). There was a Packard pre- or post-WWII (?) that had, I believe, the actual longitudinal torsion Z-bars, although I've never seen it.

I converted a VW Beetle "paddock basher" (ie. no body) to side-pair Z-bars (with 2CV style coils and a rear leafspring-Z-bar) back in the 1980's. I have no idea why everyone else in the world is not using them??? They work well!!!

There are many other ways of doing interconnected suspensions, some better than the above 'cos more separation of the modes, and hence more scope for tuning, etc,. But the above is very simple (only 3 springs instead of 4 + 2 ARB's), has good packaging (with the longitudinal torsion bars), and is a long way better than the standard approach (which has a too stiff twist mode).

There are a lot of other details, and very little readily available information (probably why so few people use interconnected suspensions), but they do work well.



Angry Joe,

I accept that you need dampers on a standard suspension, and if they are crappy dampers then the car doesn't work too well. My point was that there are a lot of ways of doing the springing better, the above being one simple way, and there are more advantages to be gained by "smarter" springs, than with better dampers.

I think using a standard springing arrangement, and then trying to make it work well with expensive, hi-tech dampers, is a dead-end, no-win situation. Just my opinion though http://fsae.com/groupee_common/emoticons/icon_smile.gif.

Z

[romkasponka]

I would like to get a copy too
romkasponka @ gmail.com

[Lukin]

Is there an ETA on that file?

[Z]

Since I have been waffling on about interconnected U-bar and Z-bar suspensions above, I thought I'd add this here.

The recent Racecar Engineering (Oct. 2005) has an article on the "Creuat" suspension as fitted to the "Racing for Holland" Le Mans racecar (although not used in the actual LM race). The article doesn't go into much detail about how the Creuat suspension works, so I had a quick look at their patent (US Patent 6,942,230 by Josep Fontdecaba). Here's the gist of it;

In essence the system uses two "U-bars" that interconnect diagonally opposite pairs of wheels (so a bit like conventional ARB's but crossed over - LF to RR, and RF to LR). These provide all the stiffness for the car's 4-wheel pitch and roll-modes (conventional ARB's stiffen roll and twist-modes). This arrangement was shown by Mark Ortiz in his 1997 RE article "Interconnected Suspension - Part 2".

To these two U-bars Fontdecaba has added a single "Z-bar" that carries the weight of the car, ie. controls the bounce-mode. This Z-bar is connected to lever-arms at the middle of each U-bar, turning the U's into W's. Thus the bounce-mode stiffness is determined by the single Z-bar in series with the two W-bars. The actual implementation of U/W and Z-bars is done with hydraulics, pneumatics are used for the springing, and damping is via normal type damper valves in the hydraulic circuits.

This is how I understand the system from the patent. The system used on the RFH car may be different or have extra features (from the article I think it must).

Nevertheless, the main benefit claimed for the system is the soft twist-mode (called "axle-crossing" by Fontdecaba). However, in certain conditions, this is also considered a disadvantage for the very aero-sensitive Le Mans car (from what I call "tilting on the diagonal" - see the article - although I reckon this is easily fixed).

Regarding the advantage of a soft twist-mode the RFH engineer is quoted as saying "with shock absorbers [normal suspension] over the kerb you can really feel it, but with this system you don't know you are on the kerb."

Z

[John Bucknell]

Per Z's above comments, some 3,000 words (i'm just posting them as a favour, these are all from him):

http://i16.photobucket.com/albums/b33/john-bucknell/ZBAR2CVsmall.jpg

Citroen 2CV and BMC, from "New Directions in Suspension Design", Colin Cambell, 1981 (with some comments by Z)


http://i16.photobucket.com/albums/b33/john-bucknell/ZBARPKRDsmall.jpg

Packard, from "Chassis Design...", based on notes by Maurice Olley, pre-1960? (with more comments...)

http://i16.photobucket.com/albums/b33/john-bucknell/ZBARFSAEsmall.jpg
Z-bar sketches by Z

[Z]

Thanks for posting those John. (Note: Have sent John lower res. pics that should load faster and fit screen better.)

Some more comments ... (I have "enforced" holiday next week, so may as well get this subject out of the way! http://fsae.com/groupee_common/emoticons/icon_smile.gif)

The Colin Campbell book shows how little understanding there is of these longitudinally connected suspensions. And the book was written ~40 years after the 2CV hit the road!

Olley calls the Packard's a "compensated" suspension, and he talks of "skew" rates for twist/warp stiffness. I don't think his definition for "skew rate" is very representative of twist/warp-mode behaviour. I don't think the suspension as shown (Olley/Milliken's) would work - the car would fall on its nose (unless extreme rear CG) and there is too much rear RMD. Perhaps there was also a front axle Z-bar, or soft front springs? If anyone has more information on these Packard's (the last of the luxury car marque) I would be interested in hearing it.

The bottom-right of the third pic (my sketch) is only one of many possible twist-soft suspensions. Perhaps an easier approach would be to start with a conventional spring-at-each-corner suspension, fit low rate (ie. soft but long) springs, then add the two extra longitudinal Z-bars to stiffen the bounce and roll-modes, and use the Z-bars to tune the handling balance via their RMD. If done right this is a big advantage! http://fsae.com/groupee_common/emoticons/icon_wink.gif

Z

[racer_chick]

If anyone went to Claude's seminars and didn't get the Magic Number Spreadsheet, just drop me a message telling me your name and which seminar you attended. I just have the emails from the registration forms, so I'm guessing a lot of them have changed, especially if you had a student email.

Thanks!
Karen

[TG]

Z,

I started to understand what you were saying when I saw the pictures. It is an interesting idea but I don't think it would be advantageous to run your system as you have described on an FSAE car. A major problem I see with it is that when a wheel bumps or dips somewhere, all of the sprung weight of the vehicle would have to go half of that distance up or down, respectively (assuming the other two wheels are at a constant ground level).

I know this isn't a direct comparison , but take UTA's past cars. It shows the effects of unsprung weight. They were running with the full body aero under tray attached directly to the uprights. Now they have the front and rear wing forces (as well as some of the weight of the wings) fed into the rockers, reducing a lot of the unsprung weight they had. From what I've heard, they've said they have much better times since the switch. I'll let any of the UTA guys comment more on it, if they want.

You might be able to see benefits in your system if you had clutches in it that would engage when the vehicle is turning. But then again, you'd still run into problems with bump mid-turn...

[ben]

A single wheel bump is equal components of each of the four modes (heave, pitch, roll, warp) if you have no warp stiffness and the other modal stiffnesses are equal you just reduce the force from a bump by 25%.

Surely this means the sprung mass will move vertically less than a warp-stiff equivalent?

Ben

[Z]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by TG:
A major problem I see with it [Z-bars] is that when a wheel bumps or dips somewhere, all of the sprung weight of the vehicle would have to go half of that distance up or down... </div></BLOCKQUOTE>
Tim, no, not "half", but half-of-half!

Firstly, the Z-bars are springs, so they can absorb bumps like any normal suspension. Ie, a conventional suspension with soft springs that hits a single-wheel bump at speed will have very little movement of its "sprung mass". The bump compresses the spring, but this only increases the spring force slightly ('cos the soft spring is "low rate"), and the short time of this slightly increased force means little upward acceleration of the body. A Z-bar suspension with soft springs works the same way.

But consider a conventional suspension with very stiff springs (say, a typical circuit racecar). If, say, the LF wheel hits a 4" high bump (eg. the kerb), then the spring barely compresses, and the centrepoint of the car rises ~2" (2"/4" = 1/2 the bump height).

If the car is rear heavy (eg. rear engine formula car), then because of the stiff springs the RF wheel rises ~4" off the ground, with the two rear wheels still on the ground. This is equivalent to the wheelprints moving with Bounce-mode=2" (all wheels up 2"), Pitch=2" (front wheels up a further 2", rears down 2"), Roll=0", and Twist=0". If the car is front heavy (eg. front engine "saloon" racecar), then the LR wheel rises ~4" off the ground, with the two right wheels still on the ground. This is B=2", P=0", R=2", T=0"

Look in any racing mag and you see these "action photos" of the cars on the kerbs (although often with 3 or 4 wheels off the ground!).

Now consider a Z-bar suspension with very stiff springs (ie. stiff Z-bars). Again the LF wheel hits a 4" high kerb, the springs barely flex, but the suspension can move freely in twist. Now the wheelprints have Bounce=1", Pitch=1", Roll=1", Twist=1" (as Ben said). So the sprung mass (~body CG) only moves up 1"/4" = 1/4 the height of the bump! And more importantly, all wheels stay in firm contact with the ground. In fact, they all push down slightly harder than their static weights (they share in the upwards acceleration of the body, though exact amounts of extra load depends on the body's pitch/roll MoI's).

Isn't that better than a conventional suspension with stiff springs?

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">You might be able to see benefits in your system ... But then again, you'd still run into problems with bump mid-turn...

Tim Gruhl </div></BLOCKQUOTE>
The thing that really surprises/bemuses/disappoints me is that twist-soft suspensions are extremely common in the overall "land vehicle" industry. Every farm tractor ever made, most earthmoving machines, the "load sharing" linkage on truck bogey axles, railway bogeys, etc., etc., etc.... There are countless different types of land vehicles that have a "soft twist-mode" for their supporting structure (wheels/tracks/etc...) and they all work much better for it!!!

The only section of the industry that doesn't use it, and they are a miniscule minority, are the dead-heads in motorsport!

As I said earlier, I tried it in the 1980's and it works well (very bloody well!). I had lost interest in cars about then, but I still used to tell anyone interested in suspension/racing about it. Always the same old piss-weak excuses - "Well, if it is so good why aren't the other guys using it...", "Hmmm, it looks very difficult to package...", etc., etc., etc... No one ever pulls their finger out and tries it!!!

Can you see why I sound like a cynical old fart? http://fsae.com/groupee_common/emoticons/icon_confused.gif

Z

[TG]

Sorry, I was thinking in the 2-d linked wheel system there, not 3-d as I should have. I still think there are other possible solutions, like delphi's magneride that a test driver for ferrari has been trying to steer me towards, that can retain the independence of each wheel while managing the vehicle dynamics at the same time. Now, if I could just get some time to spend on designing the car instead of trying to find sponsorship to get our team going...