FRAME ARTICLE #5


Here is the last "catch-up" frame article for the BI, EM, and Ducati lists. I've included several return posts and my responses. Please read everything, because I was corrected on some less than clear and/or incorrect statements by either the sharp eyed readers, or by myself when I reread a referenced article. Even if I'm totally wrong, I think you will find some thought provoking topics. Others' posts are usually indicated with >, and my replies with >>.

REAR DAMPERS AND SWINGARMS: Before getting into swingarm construction tips (next installment) let us consider the rear damper(s), what they do and how many we need. Remember, they are dampers, not dampENers - they damp (diminish progressively in amplitude) the spring motion, they don't make it more moist (unless you blow a seal).

After the motorcycle industry in general realized that rear suspension was not a bad idea (early 1950's) most motorcycles used two spring/damper units mounted somewhat near the rear axle. Road/roadracing rear wheel travel increased from minimal to the 5 to 6 inch range today. In the early 1970s motocross bikes entered a period of rapid suspension development which saw rear wheel travel increase dramatically. Maico, AJS, etc moved their twin dampers forward on the swingarm and/or laid the top forward. Yamaha bought Lucien Tilkens' Monocross design using a single shock, and soon more factories followed with their own single shock designs. Kawasaki was the first roadrace bike I can recall to use a monoshock, mounting a Koni F1 car shock in what they called a "Unitrak" linkage.

Here is where I think people got confused about rear suspension. As far as I can tell, the big advantage the early monoshock suspensions had over conventional twin dampers was the quality of the damper. Regular Koni and Girling dampers work pretty well, but the build quality of the F1 Koni car damper was much higher, and the external adjustments were an advantage for tuning the damper. To rephrase - it wasn't the number of dampers or their location that was important - it was the damper itself.

Linkages: Much was made in the early days about the wonders of progressive linkages and how they could be varied. Specifications I've seen on current GP bikes indicate that they are running fairly linear rear suspensions with a mild rise in progressivity in the 5 to 10 percent range. Oddly enough, you can get about 11 percent rising rate in non-linkage mounted twin rear dampers. Hmmm. Ever fiddle with a monoshock rear end that has worn linkage bushings/bolts? That .030" slop at the bushing suddenly becomes about 3/16" free play at the rear wheel, not that OUR racebikes would ever find themselves with worn out parts like that. So if a conventional monoshock rear suspension with linkage doesn't have any more rising rate than can be designed into a twin shock suspension, there must be some other advantage, eh?

Leverage #1: Leveraging the dampers is a good idea on long travel dirt bikes, because a 12" travel damper ends up being about 30 inches long when extended and is pretty hard to fit on the bike. Plus, when you fall over in a pile of rocks it is easy to bend. Luckily, there are usually not too many rockpiles on our roadrace courses. A typical unleveraged twin shock damper will run a 45 to 75 pound/inch spring rate on each damper. If we go to a 2:1 leverage the spring is 4 times as strong, because it varies geometrically instead of arithmetically. Now each spring is 180 to 300 pounds/inch. Eliminate one damper/spring unit and we find that we are running 360 to 600 pound/inch springs. The poor damper has to work really hard to control this massively stiffer spring. This is going to cause the shock to get very hot, as it converts the spring energy into heat. Then you want to stick the damper in the dead air space behind the hot motor, possibly with an exhaust header pipe about 3 inches away? No wonder people go on about their shock getting hot and fading - the wonder is they last at all. It seems that maybe a pair of dampers might undergo much less stress, and be less prone to overheating and fading.

Leverage #2: Remember the example of how a little slop at the link becomes a lot of slop at the wheel? Along the same line, it seems to me that a near 1:1 ratio of wheel to damper travel would provide more sophisticated control than a 2:1 ratio. That is, the incremental movement of the wheel has a longer amount of shock movement devoted to controlling its action. Think of a sledge hammer - is it easier to control the movement of the head if you hold the handle right below the head and move your hand the same amount, or when you grasp at the very end and the head travel is leveraged over your hand/wrist motion? Front forks seem to do OK with 1:1 leverage. Since you can get rear dampers that are gas pressurized to control cavitation, the rear end should work even better than a fork damper.

Chassis loads: two dampers on either side of the frame running 50# springs are going to need a much less strong structure to resist their low level, symmetric loads than a single damper with a 500# spring. As my friend Craig says, if the rear subframe area is strong enough for the rider to sit on, it is probably strong enough to take twin damper loads. As an aside, I've noticed a number of monoshock installations where the upper damper mount is fed into the middle of an unsupported tube. If the tube isn't going to deflect appreciably under the heavy loads it will have to be very strong = thick = heavy. This area in a twin shock frame can be much lighter. Also, dampers mounted near the axle put much less of a bending load into the swingarm than a highly leveraged damper unit midway (or so) between the axle and swing arm pivot. This means that a twin damper swing arm can be lighter because it only has to resist the wheel forces, not the wheel plus leveraged damper forces.

Cost: According to my White Bros price list, the retail price on a pair of dual-adjustment WP dampers is about $200 - 250 more than an equivelant monoshock damper. It stands to reason; the internals are basically the same and it probably costs about the same to make each damper. But what if you can get by with a less sophisticated damper in twin shock applications? Then the price differential might lean toward the twin dampers.

High exhaust pipes: The single shock application undoubtedly allows a high level exhaust to be tucked in tighter, since the exhaust doesn't have to remain outside the rear dampers. I am a definite fan of high pipes on RR bikes - it seems like the low pipes always ground sooner or later.

M. Moriwaki, of Kawasaki engine tuning fame, has built some frames, and he thinks twin shocks have some advantages. When tests were done on the Norton rotary racers in which the riders didn't know which version of the bike they were on, they went faster on the twin shock chassis - B. Crighton, the developer of the bikes is sold on twin dampers. These guys are pretty sharp - why doesn't everyone agree with them? Fashion, maybe. Motorcyclists are a trendy lot - they tend to want to go with the herd. When in doubt, remember that physics doesn't change. Sit down and ask yourself "why is it done this way, what is the advantage?" Maybe it was a good idea at the start (see the beginning about Kawasaki), but technology has reduced the advantage. Or maybe something is done for a purpose different from what everyone thinks is the case.

I think there may be some justification for running a single damper in the normal twin shock position - something similar to what BMW does. You could save the money over the multiadjustable twin dampers (and have a little easier time finding and maintaining the damper) and since the damper is mildly leveraged at most, spring rates would be low and it wouldn't be prone to overheating. You could mount it on the side of the swing arm away from the chain, and tuck in the high pipe on the chain side of the bike. This will require a bit stronger swing arm and subframe, but the weight increase should be small. This does put an asymmetric load into the frame, which is theoretically bad, but I think the frame could be strengthened to withstand the offset load without a great change in weight.

Think about this, and let me know your carefully considered and reasoned comments.

Cheers,

Michael Moore


From: Randy Grein

Date: Wed, 28 Feb 1996 08:26:54 -0800

Subject: Re: Frame Making, Part 5

Randy Grein wrote:

>Michael,

>The original reason for running monoshocks was to increase leverage and reduce piston speeds. Cavitation inside shocks is a HUGE problem, and totally screws up dampening rates. Of course the early Yamaha monoshocks didn't really take advantage of this... .

Hello Randy - I agree that cavitation in the traditional twin tube damper was a problem, especially when trying to increase the damper stroke to increase travel. I would venture that the general use of emulsion/de Carbon style dampers has largely eliminated the problem on the amount of travel used on a road race chassis. A 1:! ratio damper on a RR bike will have a similar stroke to many of the leveraged off-road dampers, ie. 4 - 5.5 inches. It is interesting to note that the Koni 30 series auto racing single-tube high pressure gas damper is available with shaft travel up to 9 inches. This suggests to me that cavitation is probably not a big problem anymore, at least with any range of damper stroke we are likely to use.

>Heat dissapation is a problem. It will always be a problem. It's a rather simple function of how far the suspension travels over a given time period times the force applied; ie, integrate the total linear distance traveled by the shock in one second times the forces applied, you'll get the watts input as heat. FWIW, twin shock bikes had far MORE problems with fading than the then new monoshocks, mostly because it was fairly affordable to design remote reservoirs for a single shock. Cost for two would, of course be double.

Actually, I have quite a problem with my brain overheating and fading when confronted with calculus. Since the load on the damper varies as the square of the leverage, so that a 2:1 leveraged damper has 4 times the load of a 1:1 damper it seems to me that a heavily leveraged single damper could easily have 8 times the unit load of an individual damper in an unleveraged twin damper suspension (4 times the overall load divided by two dampers). Again, I think a lot of the troubles experienced with the twin damper setups were due to the less sophisticated dampers being used.

Thanks for giving the subject some thought, and taking the time to share your thoughts.

Michael Moore


From: Robert Kennedy

Date: Thu, 29 Feb 1996 15:34:12 -0800

Subject: Re: Subject: Re: Frame Making, Part 5 -dampers

{I've left all of Robert's text in, since it is fairly complex and I didn't want you to have to dig out the prior message to see what he said - Michael}

Michael started with:

>> Hello Randy - I agree that cavitation in the traditional twin tube damper was a problem, especially when trying to increase the damper stroke to>> increase travel. I would venture that the general use of emulsion/de Carbon style dampers has largely eliminated the problem on the amount of travel used on a road race chassis. A 1:! ratio damper on a RR bike will have a similar stroke to many of the leveraged off-road dampers, ie. 4 - 5.5 inches.

Robert replied to the previous post:

>Similar stroke, but probably a much higher peak travel speed. I suspect the peak suspension travel speed (measured at the rear axle) of a race bike, with its stiff-walled tires and very high road speeds, is probably *much* higher than that of a dirt bike, even in very rough terrain. Even if they're the same, the race bike's shock pistons mounted near the axle will have to move much faster than the dirt bike's suspension damper mounted near the swingarm pivot.

Hello Robert:: does the rear axle speed matter in the comparison as long as the damper speeds are the same? Since RR monoshock dampers seem to run about 2" stroke and MX dampers about 4", both at roughly 3:1 leverage, wouldn't a 3" RR bump at 80 be roughly equivelant to a 6" MX bump at the same speed as far as the damper is concerned? And a full travel bump on the unleveraged RR damper would seem to require the same shaft speed as a full travel bump on the MX bike, presuming the speed at which the bumps are hit is the same and the shaft travels are equal. I think when comparing RR to MX you need to use bumps that use the same proportion of suspension travel, not just bumps that produce the same wheel travel.

I said:

>> It is interesting to note that the Koni 30 series auto racing single-tube high pressure gas damper is available with shaft travel up to 9 inches. This suggests to me that cavitation is probably not a big problem anymore, at least with any range of damper stroke we are likely to use.

Robert said:

>This data is hard to refute. I guess cavitation is probably not such an issue, except that once again, a vehicle using a nine-inch travel in an automotive application is surely an off-road vehicle, and off-road suspension speeds would seem generally lower than on-road ones (though I have no hard data).

The 30 series is the Koni circle track line. The 9" models are listed for both asphalt and dirt oval applications. My impression is the de Carbon and emulsion dampers were developed because they wanted to put the damping fluid under enough pressure to prevent it from cavitating.

>> Actually, I have quite a problem with my brain overheating and fading when confronted with calculus. Since the load on the damper varies as the square of the leverage, so that a 2:1 leveraged damper has 4 times the load of a 1:1 damper...

Robert said:

>This is false. The load on the damper goes up with the square of *displacement speed* (assuming no fancy valving), and linearly with the amount of force. Mounting the damper nearer the swingarm pivot reduces the displacement speed (for a given wheel speed), and increases the force applied to the damper by the same factor. The result is that to get the same damping force at the wheel, you have to increase the damping rate according to the cube of the leverage ratio, BUT the load on the damper after you've beefed it up is LINEAR IN THE LEVER RATIO. Let's take it a little slower. Suppose we go from a 1:1 lever arm to a 2:1 lever arm. We have cut the damper speed by a factor of two (for a given wheel speed) and have increased the force on it by a factor of two (for a given force on the wheel). Our goal is to keep the relationship of wheel force to wheel speed the same as it was in the 1:1 case. Suppose for the moment that we keep the damping the same. Then the force on the damper at a given wheel speed is 1/4 of its original value, since the damper moves half as fast as the wheel does (and (1/2)^2 = 1/4). Moreover, since the leverage favors the wheel by a factor of two, the resisting force at the wheel is 1/8 of its original value. So to get the wheel to feel the same damping force as it originally did, we have to make damper's rate eight times what it was before. Teensy orifices, or tar where there was 3-in-1 oil.

>Now let's suppose we've put tar in place of the shock oil, so the wheel feels the same force it once did. It's a simple matter of 7th-grade physics now to see that the force on the damper is merely twice what it was in the 1:1 case.

>> it seems to me that a heavily leveraged single damper could easily have 8 times the unit load of an individual damper in an unleveraged twin damper suspension (4 times the overall load divided by two dampers).

>No. See above.

Robert, I'm afraid your 7th graders do physics at a higher level than my stonage 12th grade course did. From your Stanford address I presume you are probably someone who has extensive math/physics background.

No question that the load on the damper is basically linear once the damping is increased. You are correct that I was not careful to point out that the damper moves half as much, although against a 4x more powerful spring. I should have said that for the same SHAFT travel the leveraged damper would be under 4 times the load, since it is working against a 4x more powerful spring.

My big question is: aren't you forgetting that the damping force is divided between two dampers vs one damper? Your illustration seems to presume the same dampers at the axle are just moved forward to increase the leverage, not that one of them is eliminated in the process. If this is the case and I didn't misread, then going from 1:1 twin dampers to a single 2:1 damper would increase the load on the damper by twice what you calculated, ie. 4 times as much. I think this is a significant amount of stress on the damper, even if it isn't 8 times as much as I originally said..

Also, I'm not sure if you are taking in account the negative aspects of smaller metering orifices/heavier damping fluid. Either one is probably going to cause the fluid to break down quicker, and I would think heat up more, due to the shearing of the long molecules common in oils. Shearing of the oil is one reason why engine developers try to keep oil away from the flailing connecting rods. As the oil is sheared it both breaks down and heats up. It will also be harder to prevent the fluid from cavitating when using the heavier viscosity/smaller orifice. I know some experiments have been done trying to use water as a damping fluid, since the viscosity doesn't change appreciably with heat, but they couldn't get a lubricant package that would keep the damper from quickly wearing out.

>> Again, I think a lot of the troubles experienced with the twin damper setups were due to the less sophisticated dampers being used.

>I think the biggest advantage of the single-damper setup is that there's plenty of room for a rising-rate linkage. Such a linkage gives you damping changes along with a progressive springing rate, and that combination is unavailable with simple twin-shocks that have progressively-wound springs.

> -- Robert Kennedy

Robert, I sat down a couple of years ago and drew out incremental wheel/damper travel for a variety of twin damper positions, from the damper being perpendicular to the swingarm damper mount to swingarm pivot at full bump (maximum mechanical leverage on the damper in this position) to having both damper mounts equidistant from the swingarm pivot. I calculated the rate of change for each setup (no fancy forumulae, I measured the actual incremental changes with AutoCad to the nearest .001", than figured the percentage change of each one with a spreadsheet). I was able to get a maximum rising rate of about 11 percent out of one of the normal, basically unleveraged twin damper placements. This rising rate is independent from the spring.

The trend the last 5 or so years in road racing (as far as I can tell from reports - I've not had a chance to measure any factory/National level bikes) has been to make the the suspension curve LESS progressive and more linear, reducing the amount of rise to about 5 to 8%. This is well within the range of rise that can be obtained without linkages.

Also recall that I was referring to rear suspension as an element of frame design, and that one of the advantages I found for twin dampers is you can get similar results from the suspension with a lighter, less complicated, less heavily loaded frame.

I'll run your post by my friend Craig - he has been dealing with the more theoretical/mathematical aspects of suspension design for much longer than I have, and has designed race car suspensions that have worked well. It is possible that I garbled what I thought he told me when I wrote the article.

Thanks for spending what was obviously a fair amount of time and thought on the subject. Even if I'm wrong (perish the thought, but it has been known to happen), this dialogue has probably enlightened a few people, and bemused many more.

Thanks,

Michael Moore


From: "fred (f.) farzanegan"

>Date: Thu, 29 Feb 1996 14:17:00 -0500

>Subject: re:Re: Frame Making, Part 5

>I'm interested in your belief of twin shocks being a better/ simpler design. I recently read in American Roadracing (11/12 '95) about a homemade center-hubbed FZR1000 made by Andy Stevenson. Alan Cathcart reviewed the bike and had general praise for it, but harshly criticised the decision to use twin shocks on the bike and cited them as the limiting factor in its handling.

>- -fred

Hello Fred - I dug out the article you referred to, it was about the second of three incarnations the ASP Yamaha has gone through. Bike magazine did four articles on the MK1 version. First off, let me make my bias plain: Alan Cathcart is a fine rider, and has probably ridden more different race bikes than anyone, and I've generally enjoyed reading his articles for the last 15 or so years, but he often makes statements relating to technical matters that seem to me to be completely out of left field. This is also the opinion of some other people who have proven to me to really know what they are talking about. Cathcart stated in the same article "There's no bump-steer whatsoever from the front end, in spite of the lack of any linkage to give a progressive suspension response" completely ignoring that suspension damping can have no effect on bump steer at all. Bump steer is when your steering angle is affected by variations in the suspension movement, ie the steering angle is 5 degrees at 1" travel vs 10 degrees at 2 " of travel, without changing the angle of the bars. This is due to steering rods/racks not being of the right height, length, position etc. Cathcart has made similar statements about bump steering in other articles, reinforcing the impression he hasn't got a clue about what bump steer is.

Cathcart ignores the fact that in the MK1 ASP, Andy Stevenson used the FZR linkage and single damper, has switched to the twin dampers, and feels it is an improvement. Stevenson said he used twin shocks "after reading why Brian Crighton did the same thing with the works Norton Rotary racers...If you have good shocks, ... same as he did, then you get adequate suspension response on a smooth racetrack as well as a stiffer rear end structure". Cathcart opined that the Norton uses them because of the heat coming off the engine and completely ignores the fact that Mr. Crighton has explicitly said in interviews that he feels the twin damper rear suspension has definite advantages over the conventional monoshock, and that he proved that to his satisfaction in blind testing where the riders rode both versions of the chassis without knowing which they were on, and they consistently went faster on the twin damper version. In a recent Harley/Metisse road test Cathcart praised the bike's twin Ohlins damper rear end as having "the same control you get from a longer-travel monoshock rearend...The ride quality is really impressive.." Granted this is a street bike, but it shows how variable his reports can be.

Any suspension can be set up wrong - you can take Doohan's NSR and ruin the handling by fitting dampers that are out of the range of what is needed, and you don't even have to convert it to twin dampers to make it work poorly. My impression is that many people don't spend enough time revalving the dampers to get them right, and then complain about how bad the rear end is (twin or single damper}. During the course of revalving WP dampers (front and rear) on a number of MX bikes, my friend Craig has discovered that the stock shim stacks in the dampers appear to vary with whatever shims the factory had sitting in a box next to the assembly line. They've been way off in a number of instances. So just because you have the latest trick damper, don't presume it is right. Even if you send it out for reworking, it seems like they often need several revalvings as the approximate specs are narrowed down to the right ones.

If Cathcart says the rear end didn't work well, he was probably right. But I don't give much credence to his being able to define exactly why it wasn't right.

Thanks for your comments,

Michael Moore


In my previous post concerning Alan Cathcart's article on the ASP 1000 I said Mr. Cathcart had ignored the fact that the bike originally had a single rear damper and was changed to twin dampers. He did in fact mention it, and included a statement from Mr. Stevenson to whit: "It was more comfortable (the original Yamaha suspension), but incredibly vague, because you never seemed to know what the rear tire was doing, so I switched to twin rear shocks". I apologize for the error on my part, although it doesn't really affect the substance of what I said in the previous post.

Michael Moore


Back to the top frame page
Back to the home page
© 1996-2005 Michael Moore, last update for this page 01 June 2005

For more information contact us by using

http://www.eurospares.com/frame5.htm