chassis recipe, introduction.

[vanzen]

This thread will not be to everyone's taste.
It's contents will provide a comprehensive strategy to improve the handling characteristics of a twin-shock air-head –
a not entirely my own and not-so-secret "BMW Road Kill Cafe Recipe".
A plethora of information, mis-informatiom, and idle after-market advertising hype exists regarding this topic,
and together, maybe we can sort it all out.
All discussion is welcome, but please note:
My recipe, after all, is not intended to be THE definitive menu but rather just an example of one rather tasty "dish" ...
an inexpensive dish that would suit my taste, may suit yours, but will still prove too spicy for many.
So, if "spicy" ain't your style ... perhaps best to pass this one up ...
for the rest, let's get on with it then ...

The twin-shock air-head chassis was designed by the factory with then available technology
to be flexible in order to cope with a variety of road surfaces and conditions,
to be fairly competent in all possible riding situations, albeit extremely proficient in none.
As good and well received as this marketing strategy was,
it remains a compromise intended to appeal to the largest number of potential customers / riders most of the time.
The truth of this statement will be witnessed by the extensive chassis modifications that altered the Butler-Smith / Udo Gietl 1976 R90S
in order for it to be a successful competitor and Superbike championship winner.
The first point to note, then, will be that
modifications here are intended to improve chassis behavior when ridden fast on paved roads,
and that performance on dirt roads and marginal surfaces will likely suffer.
The resulting chassis will be more focused – and less versatile.
Next, THE FIRST & ONLY PLACE TO START WITH CHASSIS IMPROVEMENT
WILL BE A STOCK CHASSIS THAT IS OPERATING AT 100% EFFECTIVENESS.
If your riding skill and style does not approach or surpass the limits of stock chassis design capability,
any improvements to the system will prove moot.
And finally, improvements will be relegated to stock components as if we were considering building a bike
to (loosely) satisfy AHRMA Vintage 750 Sportsman Class rules relative to chassis modification –
i.e. no late model forks system transplants, custom frames, or mono-shock conversions.
Why do I emphasize "twin-shock chassis" ?
Because the factory effectively dealt with many, but not all, of the issues we will address here
with the introduction of the mono-shock platform and improvements continued
throughout the rest of the air-head model run.

Intro:
The motorcycle chassis and chassis dynamics must be considered as an INTEGRATED SYSTEM.
Having said that, I will attempt to order this recipe in hierarchical terms of (my perceived) importance.
I will deal with these three sub-systems individually –
Part 1, forks system and front suspension,
Part 2, main frame, and
Part 3, sub-frame and rear suspension.



stay tuned ... Part 1 to follow ...

[vanzen]

It is my understanding that the twin shock "airhead" frame is basically a copy of the Norton "Featherbed" frame, according to Hans-Gunther von der Marwitz - as quoted in "Bahnstormer" (p.141) by L.J.K Setright.

PT9766

Norton Feathebed:
http://i236.photobucket.com/albums/ff169/53vanzen53/misc%20pics/featherbed.jpg

The design is quite similar,
including the rather odd method of routing the main-frame's "spine" to the bottom of the neck-stem tube –
a concession to CoG in order to keep the weight of the gas tank and it's fuel lower in the frame,
and this to the detriment of neck-stem area rigidity.
Norton compensated for this flaw, however,
and this will be an important and major distinction between the two frames.
The Norton featherbed frame has an integral design feature,
what is referred to as a "head-steady" –
A rather stout bracket that connects the frame just below the neck stem
to the head casting of the engine.
This attachment, and the engine mountings on the bottom frame rails
allow the engine case to function as a structural frame element / reinforcement,
dividing the rather rectangular profile of the frame in half on a diagonal
from the neck stem to the swing-arm pivots.
This use of "triangulation" greatly increases the Norton frame's ability to resist distortion,
and is an important element in the success of the featherbed design.

BMW engineers were apparently absent that day ...
as no such attempt to triangulate the Type-247 frame was incorporated.
This will be the reason why the BMW frame flexes significantly at the neck-stem requiring reinforcement of this area.
And also why "frame-braces" have become so popular.

A relevant quote concerning the handling qualities of the Norton featherbed, considered excellent in it's day:

"What does allow the handling stability of most British bikes is the
normally minimal amount of power, relative evenness of its delivery and the
fact that the state of technology in the rest of the world during the
British heyday was worse."

[shire2000]

This should proove to be very interesting. I am certain that we can all learn something from this thread. I look forward to reading your "Recipe". Do you plan on any simple drawings or references to such so that we can follow along? Some of us need pictures to stay focused. :scratch

[Yarddog]

Lookin' forward to what you're gonna share...so far, in my one month of BMW ownership, I haven't been overly impressed about the engineering of simply things on my R100/7...maybe this will help me be impressed a little more...

[RecycledRS]

:lurk

Can't wait for more!

[Mika]

+1

:lurk

[Bigrider]

Well!!! We are waiting!:whistle

[lkchris]

Just think--after all this effort an Oilhead will still be better!

[RecycledRS]

Nudge, nudge. :stick

[vanzen]

Just think--after all this effort an Oilhead will still be better!

A statement that I have myself made numerous times,
but also one that will be irrelevant here, Ichris.
The subject is clearly stated:

"It's contents will provide a comprehensive strategy to improve the handling characteristics of a twin-shock air-head –"

... not intended to be a comparative analysis of disparate technologies,
or a philosophical and economic justification of pursuing the goal.
It is intended simply as a "HOW TO"
and a discussion of the practical merits of those procedures

Quite busy at present, re-roofing the shop ... then headed to Hunter.
Will continue with this ASAP.
In the meantime –
put the popcorn down. Anyone is welcome to share their ideas ... regarding the topic –

[Mika]

Just think--after all this effort an Oilhead will still be better!

This post made me chuckle.

Anyone who is a member and participates online in this forum is engaged in a past time which the vast majority of the world considers irrational at best. Therefore; as a member the idea of the thread as you laid it out makes perfect sense to me.
:nod

Your posts in the past have often included the why of doing some modification along with the how you did it. I don’t plan on running out and starting a twin shock BMW project but find that information very valuable for approaching other projects – real or imagined. I hope you will continue to focus on telling and discussing the whys of your strategy along with the hows.

[jamesdunn]

Just think--after all this effort an Oilhead will still be better!

And so will a number of other modern motorcycles. I love my oilhead but enjoy my airhead for what it represents. I will check in now and again to see what transpires. Perhaps I will gain an idea or two.

[crazydrummerdude]

I haven't even read every word yet, and I've already learned a few things.

This will be cool.

[BubbaZanetti]

Quite busy at present, re-roofing the shop ... then headed to Hunter.


hope to meet you there, i'll be the young doofus making a weathered R80ST look like a circus bear bike.:thumb

[lkchris]

The subject is clearly stated: ...

And you think you can tightly control the subject and the responses?

How about you start by listing your engineering background that qualifies you to do this ... you know, as part of the ego trip you're so kindly "volunteering" to afflict on the rest of us?

Oh, and please spare us the "reality TV" junk, like working on your roof. We don't care.

[BubbaZanetti]

And you think you can tightly control the subject and the responses?

How about you start by listing your engineering background that qualifies you to do this ... you know, as part of the ego trip you're so kindly "volunteering" to afflict on the rest of us?

Oh, and please spare us the "reality TV" junk, like working on your roof. We don't care.

MOD ON

Please let vanzen say his peace, then if you find things you don't like, you're free to criticize. One man's "ego trip" is another one's opportunity. Let him have his thread, k?

MOD OFF

[mmmalmberg]

Wow. Like, chill.

Vanz - I'm impressed - you've started a hot thread with only barely a word out of your mouth - Just the premise of the thread was all it took - good one!
:)

[RecycledRS]

Since when do you have to have an engineers degree to have input in this forum? If you feel his thoughts are being “inflicted” on you don't read them. I for one would like to hear his thoughts on the subject.

[jamesdunn]

Since when do you have to have an engineers degree to have input in this forum? If you feel his thoughts are being “inflicted” on you don't read them. I for one would like to hear his thoughts on the subject.
+ 1. Henry Ford did not possess an engineering degree.

[jlsmith37804]

One pompous troll wants to inflict his opinion on everyone who deigns to improve a bike whose time has passed some time ago. Stick to your own bike and leave the hot rodders be. Nothing would be improved if we all accepted the current state as the only possible way.

[shire2000]

I was brought into the light and will not attack Vanzen or any other person on the forum. I believe that everyone has the right to put forth their ideas in a manner that makes sense and can substantiate what they are saying, without attacking anyone else. As with all forums, there are always those that like to get on top of their soap box and spout off about many subjects, myself included. But it does not make any sense to attack any individual just because they are willing to share their knowledge and experience. A good rousing discussion is much different than outright attacks.

I am certain that any of us that are willing to answer questions, put forth interesting ideas, and pass along from our hard earned knowledge base would be willing to also pass along our credentials, be that degrees from a brick and mortar school or from the school of life.

Let's all try to play nice.

[BubbaZanetti]

Let's all try to play nice.

thank you

let's move on

vanzen..........oh vanzen, we're waiting!

[vanzen]

Fork Springs.

My ex has always said "the stiffer the better" and in some scenarios, this will be true.
Fork springs on a bike ridden on the road will NOT be one of those scenarios, however,
and yet the fallacy still has many proponents.

In the olden days of racing, teams chose springs with stiffer rates for these reasons:
1) Only "single rate" spring sets were available to them.
2) The high cornering speeds and rapid acceleration of racing impose greater loads on the suspension,
and these increased loads are better controlled by a stiffer spring rate.
3) The machine would be ridden exclusively on a racetrack.

The other side of this coin is the one that should concern those of us that ride on the street:
Using a stiff single rate spring will always be a compromise.
Fork springs with excessive spring rates cannot respond efficiently to the surface imperfections that are a common feature of our highways,
the bumps, holes, and debris that are typical of most roadways.
Road-holding ability will likely suffer as a result of using "heavy duty" single rate springs.
The question to consider will be this:
What practical benefit is a spring that effectively controls the greater forces of speed and cornering,
and yet is uncomfortable and has the propensity to cause loss of traction,
sometimes entirely,
when the tire encounters the inevitable bumps, cracks, a pothole ?

Dual-rate or multi-rate springs will be the better road choice.
These are available either as a single spring that is wound to have different spring rates,
or as a combination of springs that are 'stacked' together, each having a different spring rate.
It seems that technology has given us the ability to have our cake and eat it too.
A "compliant" spring rate to keep the tire in contact with the (sometimes rough, and certainly not track quality) road
AND a "stiff" spring rate to counteract the forces of spirited riding.
We want our tires to be in contact with the road in all situations and as often as is possible –
CONTROL and STABILITY will depend upon that fact.

Note that I have not used the term "progressive".
The true 'progressively wound spring' is not yet a reality,
and ... I do not wish to introduce confusion with that term and the Progressive BRAND,
which is a dual rate spring with excessively high spring rates for most applications.
A proper spring rate should be chosen primarily as a factor of combined mass –
the weight of the machine plus that of the rider (fairings and and any usual baggage to be considered),
second, as a function of riding style,
but never on the basis of an archaic bit of race trivia translated to street use.

Fork oil.

Fork oil controls the speed of a damping system's ability to respond to road imperfections.
Quite simply, heavier oil equals greater spring damping action and SLOWS the fork's ability to respond.
Lighter weight oil equals less damping action and a QUICKER response.
One more time:
We want our tires in contact with the road in all situations and as often as is possible –
CONTROL and STABILITY will depend upon that fact.
Compliant forks and responsive damping (and rebound) will make this happen.
One common mistake is to compensate for springs that are weak, sacked,
or springs of an incorrect spring rate,
by using heavier fork oil, and simply slowing the forks ability to react to the road surface –
and this just does not make any sense.
Using a heavier weight fork oil than that recommended by the factory is most often a "band-aid" attempt to address other suspension maladies,
counter-productive to the goal of good handling,
and in any event unnecessary with springs of an appropriate spring rate.

Damping & rebound.

Beyond a little fine-tuning with fork oil weight, damping and rebound on the air-head
might be considered a 'fixed' equation – with one exception.
On all twin-shock BMWs, damping and rebound are controlled by a number of holes drilled into the damper rod.
For whatever reason, throughout the years and, perhaps, depending on model,
the factory chose to vary this number of holes from two to five.
Five holes allow the forks to compress quite rapidly, increasing brake dive and wallowing at speed,
while two can be so rigid as to result in wheel chatter at speed.
The consensus seems to be that three is a good choice for spirited riding.
Holes can be drilled or welded shut and ground flush as your situation requires.

If money is no object, cartridge emulators or inserts are THE way to go.
I personally have little experience with either, except for those found on the 1973 Moto Guzzi Eldorado,
and yes they were stock equipment.
In essence, cartridges or cartridge emulators use a system of valving,
much like that used on quality rear shocks (dampers) to allow a broad choice of tuning,
including the use of differing damping and rebound rates –
a feat impossible with the drilled hole system.
The means to custom tailor a fork's performance to your own personal agenda will be the primary advantage of these.

Comments welcome.

Next up – the problem with telescopic forks.


STAY TUNED !

[RecycledRS]

I did not realize that progressive spings aren't turely progressive. Given this one should be able build your own custom spring combination to suit. Would one have to purchase each spring or could they be made by cutting up existing springs to length? I understand that changing a springs length greatly changes the spring rate.

Steven Bottcher's site http://www.stephenbottcher.net/BMW/springrates.htm lists a number of different combinations he has tried to his liking.

Thoughts, comments?

[vanzen]

... Given this one should be able build your own custom spring combination to suit.

YES !
I happen to know Stephen (http://www.stephenbottcher.net), have been following his bike's progress, but was unaware of the "fork spring experiments".
I'm impressed –
While there, be sure to check out the rest of his site, as Stephen has been involved in air-head modification for quite some time,
is very knowledgeable, and has detailed much of his work.
A good source of information.

Would one have to purchase each spring or could they be made by cutting up existing springs to length?

Either should work, but I would think that a reliable method to test spring rate would be a necessary bit of equipment.
Accessing reliable data for any spring may prove difficult and "used" springs may not perform to that data if it were found.
And then it's trial and error time unless one has some reliable point of reference with which to begin.

I understand that changing a springs length greatly changes the spring rate.

This article in SPORT RIDER (http://www.sportrider.com/tech/146_9510_tech/index.html) discusses the difference between spring rate and preload.

[535is]

Dual-rate or multi-rate springs will be the better road choice.
These are available either as a single spring that is wound to have different spring rates,


Note that I have not used the term "progressive".
The true 'progressively wound spring' is not yet a reality,
While you might be right in the context of springs available for Airhead forks, you're decidedly wrong if you mean to say that there are no progressively wound springs. All kinds of cars have them. Theoretically, unless there is a distinct kink where the angle (or diameter of the wire or winding; both pretty unlikely) of a dual or multi-rate spring changes, even it will have a very small progressive transition rate.

Further to Recycled's remark about spring length: Spring length and spring rate are unrelated except with respect to a single spring being modified. If you cut a given spring, then its rate will increase, but that is the only direct relationship between the two. It is perfectly possible to make two equal length springs with wildly different rates.

[vanzen]

While you might be right in the context of springs available for Airhead forks, you're decidedly wrong ...

– I'm wrong.
and in fact, the Progressive Suspension (http://www.progressivesuspension.com/prodSearchResults.aspx?yearID=1981&makeID=38&modelID=915) brand springs available
claim to be "progressively wound", and these are available for many of the the air-heads.

From the ad copy at AS Cycle:
"The rates of these progressively wound springs have been carefully designed to fit each model...
A Progressive Rate Spring has the advantage of a rising rate resistance to compression.
For example, a 15lb. to 25lb. Progressive Rate Spring will take 15lbs. to compress it the first inch,
then 17lbs. the next inch, and so on, until the end of the travel;
it will take an additional 25lbs. to compress it the last inch.
The benefit of this is that the spring can be soft enough at the start of the travel to offer a “plush” ride
yet be firm enough at the end of the travel to soak up the big bumps."

[ducatipaso]

And you think you can tightly control the subject and the responses?

How about you start by listing your engineering background that qualifies you to do this ... you know, as part of the ego trip you're so kindly "volunteering" to afflict on the rest of us?

Oh, and please spare us the "reality TV" junk, like working on your roof. We don't care.

another quality lkchris post. :thumb

not.

[orbitangel]

And you think you can tightly control the subject and the responses?

How about you start by listing your engineering background that qualifies you to do this ... you know, as part of the ego trip you're so kindly "volunteering" to afflict on the rest of us?

Oh, and please spare us the "reality TV" junk, like working on your roof. We don't care.

Ikchris:

Your post speaks for itself. I am completely underwhelmed.

[vanzen]

question:

Given a fork fitted with single rate springs.
and given that spring rate is measured with the springs in an upright vertical position.
Fork springs are not on a vertical plane when fitted to the bike,
but assume the angle of the chassis rake (let's say @30º) ...
would this angle introduce a rising rate ?
As does the same scenario on a rear shock ?

[orbitangel]

... start by listing your engineering background that qualifies you to do this...

Speaking of Burt, he didn't have any engineering qualifications either - but sometimes he just got LUCKY!

One day, at the Bonneville Salt Flats, he was blessed with a 150.67 MPH tailwind! Imagine that!

See? You don't really need any smarts at all, just blind luck will do the trick! :dance

[shire2000]

Vanzne wrote: question:

Given a fork fitted with single rate springs.
and given that spring rate is measured with the springs in an upright vertical position.
Fork springs are not on a vertical plane when fitted to the bike,
but assume the angle of the chassis rake (let's say @30º) ...
would this angle introduce a rising rate ?
As does the same scenario on a rear shock ?

I am not an engineer and only understand basic physics, but I think you would be correct in assuming that there would be a change in the rate. How you would measure it, I have no idea. Maybe we have someone out there that would be able to come up with a method to measure the change???

[ducatipaso]

question:

Given a fork fitted with single rate springs.
and given that spring rate is measured with the springs in an upright vertical position.
Fork springs are not on a vertical plane when fitted to the bike,
but assume the angle of the chassis rake (let's say @30º) ...
would this angle introduce a rising rate ?
As does the same scenario on a rear shock ?

I have a friend that's a professional suspension telemetry analyst/engineer/guru. I'll ask him this question.

[20774]

I'm having a hard time understanding how the angle would factor in. Spring rate is the force required to compress the spring a specific distance, right. Where does the fact that it's vertical versus horizontal enter into it? Gravity?? Seems like it would be a small order effect. I guess I'm not seeing it... :scratch

http://www.engineersedge.com/spring_comp_calc_k.htm

[crazydrummerdude]

You don't really need any smarts at all, just blind luck will do the trick!

It's worked for me so far!

In ikchris's defense; he's a wealth of BMW knowledge as BMW intended it. If you ask, "So, what year did _____?" or "What's the proper setting for ____?", he'll respond with the answer and only the answer.

It's also obvious vanzen knows a lot. If someone asked "How do I get the most out of my airheads _____?", he'd be one to respond.

..that doesn't mean I have to take either of their advice, though. I'm not going to go chop my frame up, or on the other hand, replace my carb diaphragms at the specific intervals that Bing recommends.

I think most of us strive to develop friendships based on this common medium; airheads. I know I've made friends based on the amount of people that opened up to me at the national rally when I said "I'm crazydrummerdude." I find the posts with the "reality TV" junk adds more character than clutter. Just think of 'Renegade.' My own thread is full of it, because my R90 is part of my life.

My one complaint about this thread is that there aren't more pictures.. :ha

[orbitangel]

question:

Given a fork fitted with single rate springs.
and given that spring rate is measured with the springs in an upright vertical position.
Fork springs are not on a vertical plane when fitted to the bike,
but assume the angle of the chassis rake (let's say @30º) ...
would this angle introduce a rising rate ?
As does the same scenario on a rear shock ?

VZ,

The way I see it, the fact that the rake angle is 30 degrees only tells me that the vertical force applied to the fork springs will be:

Applied Force x sin 60 degrees, in other words .866 (call it 87%) of the applied force.

I don't see how linear springs will become non-linear until you reach (or very closely approach) a coil bind condition.

I further do not see why the situation should be any different for the rear springs.

My solution neglects any front end "diving" effects which would apply HIGHER forces than predicted, making the springs compress even more.

I reserve the right to be wrong,

[RecycledRS]

I hadn't considered this before but here goes: give that the static position of the fork is 30' it will change in operation ie. braking.

The front end dive will result in reducing the 30' angle to something less. (for arguement sack if we assume 25'), this would infact result in a rising rate. Sine of the angle 65' = .9063. While not much different than .866 it is a design factor.

Similar rising rates can be built into the rear suspension using varying mounting locationds and levers.

In real life testing we would proably just keep trying available spring combinations that give us a better "feel".

[535is]

VZ,

The way I see it, the fact that the rake angle is 30 degrees only tells me that the vertical force applied to the fork springs will be:

Applied Force x sin 60 degrees, in other words .866 (call it 87%) of the applied force.

I don't see how linear springs will become non-linear until you reach (or very closely approach) a coil bind condition.
Agreed. I sort of intuitively knew that non-changing geometry of the head angle would not change the rate as the compression changes.
I further do not see why the situation should be any different for the rear springs.

I think it does. It seems to me, if the distance from the swingarm pivot to each end of the shock is not equal, that you get a falling rate effect. That's because the angle of the shock/spring would be changing more toward parallel with one or the other arm of the triangle. The closer it got to parallel, the less force it would take to move the other arm. Not being an engineer, I can't explain it as elegantly or as precisely as you can, but I think you get the idea ... so you explain it to 'em! ;)

[rpeckham136133]

It's worked for me so far!

snip

My one complaint about this thread is that there aren't more pictures.. :ha


lest we bash ourselves silly....:banghead :banghead :banghead


:love


If anyone is in doubt of Vanzen's abilities, thought processes, and wanting more pictures, I suggest you go to

Vanzen's home page (http://www.rockerboxer.com/garage.html)

and click on a coupla links about his frame project. Some incredible photos in there, may twist the knickers of the purists, but it really is Fascinating Stuff!

Go for it, I say!! :thumb There is room for all, here.... we have our own reasons for doing what we do, and our own reasons for being here, but we're all here !!

[Bob_M]

In ikchris's defense; he's a wealth of BMW knowledge as BMW intended it. If you ask, "So, what year did _____?" or "What's the proper setting for ____?", he'll respond with the answer and only the answer.

I second Dude’s validation of Kent’s knowledge. His is a lonely voice in our little club but in vintage car collecting, and in resale of vintage vehicles his POV is the one that carries sway. We tweak our bikes to make them look a certain way, or to re-fit them with modern technology, but when we are gone and our kids put our old bike up for sale they will surely wish that we kept them stock because our upgraded riders and hot rods will not have the value of an original bike.

I respect Kent’s point of view, but my 1982 R100RS would make him cringe.

Now back to our regularly scheduled programming

[mmmalmberg]

535is -

Bingo. I think.

The fork spring's in a (relatively) fixed geometry in its various states of compression. One could nit-pick and say that as it compresses, the front end dives and the angle relative to the road changes a bit, i.e. in the same way the steering quickens. But I think we could call this somewhat negligible for the spring's rate. I think...

Depending on the layout of the rear shock, it's angle relative to the direction of wheel travel under compression can vary substantially. In my recollection, earlier twin shock bikes had the shocks placed fairly equidistant top and bottom, from the swingarm pivot.

I seem to remember more angled shocks appearing in maybe the 80's, but I don't get what the benefit of that would have been. Maybe just to get longer travel out of a given shock length - ?

[AntonLargiader]

"For example, a 15lb. to 25lb. Progressive Rate Spring will take 15lbs. to compress it the first inch, then 17lbs. the next inch, and so on, until the end of the travel; it will take an additional 25lbs. to compress it the last inch."

If that's not Progressive's own copy, I'm not sure I'd take that as gospel. Unless the last coils are nearly touching you won't get a rising rate in the first inch of compression and I don't think I'd want that anyway. For most street applications I'd want a pretty linear rate during normal travel (going over normal road undulations and small bumps) and a rising rate once the fork leaves it's normal range. That's the basic principle behind dual-rate springs.

If the spring is correct for the load, it doesn't provide that much of the anti-bottoming anyway. Damping, especially travel-dependent damping and hydraulic anti-bottoming (like the R100GS, and to a lesser extent the previous generation of forks) do a great job at this without progressively wound springs.

As for the fork angle, if you want to figure in the change during fork dive the spring rate would actually drop slightly. Otherwise, no effect.

[shire2000]

Anton, just for poops and giggles, How would you measure the change in spring rate?

I pulled out my old Physics books from Uni. and can't get my head around the concepts anymore. Been way too many years since I barely scraped thru those classes.

[AntonLargiader]

The rate's a function of wire size, coil diameter and number of working coils. For a given spring, all that's going to change is the number of coils so you look at the coil spacing and as the spring compresses some of them coil-bind and go out of effect.

For instance, a spring with 30 coils. 10 of them are 1mm apart and 20 of them are 3mm apart. For the first 30mm of spring compression each of them is losing 1mm, so all 30 are in effect. At that point, ten of them become a solid block and only 20 coils are working, so the spring rate goes up by 30/20. After another 40mm of travel (20 remaining active coils times the remaining 2mm between them) the whole spring coil-binds. Total 70mm travel, not a very good choice for most bikes given that some of it gets used up in preload anyway.

A progressively wound spring loses coils one by one, so the rate constantly changes. But a progressively wound spring doesn't have to be linear or anything like that; the rate can follow whatever function the designer wants. He can make it gradually change at one point and then suddenly change at the end. The only limitation is that it has to always increase with compression, since you can't gain coils by compressing the spring.

[RecycledRS]

Nice clear explanation Anton.
Bill

[shire2000]

Thanks Anton, that now makes perfect sense to me.

[ducatipaso]

I have a friend that's a professional suspension telemetry analyst/engineer/guru. I'll ask him this question.

here's your answer

What you want to look at is the "wheel rate". This is the effective spring rate that the wheel sees relative to the chassis. To do this you look at the motion ratio for the suspension (inches of spring travel / inch of vertical wheel travel). The wheel rate is the product of the spring rate and the square of the motion ratio. All of these values are calculated at any given point; if any component is non-linear, it will vary throughout the range of suspension travel. The wheel rate can be rising, linear, or falling depending on the geometry.

In the case of a fork with a linear spring, the motion ratio is apparent through simple trig. The fork/spring axis is the hypotenuse of a right triangle. The vertical leg is the line through a point on the chassis along the hypotenuse, say the triple tree. This gives an easy visual of the proportion of vertical wheel movement (the long leg of the right triangle) and the compression of the spring (the hypotenuse). So, the more angle the fork has (relative to the perpendicular), the higher the motion ratio. The higher the motion ratio, the higher the wheel rate (for a given spring). Think of the extreme case where the forks are almost horizontal...discounting the fact that everything would be trying to bind, it would be very hard to compress the bike an inch.

Circling back to the original question, if the fork angle does not change as the bike is compressed, the motion ratio remains unchanged.

When you get into swing arms, you no longer have simple linear motion. Your wheel rate WILL vary...either rising or falling, or both (if it goes "over center"). You can tweak geometries around to minimize the change in motion ratio, but generally a good suspension that acts through any kind of a linkage will have at least a little bit of rising rate as the wheel goes into bump.

Clear as mud?

[vanzen]

"... the mounting location and angle of the shock determines
whether the system will exhibit regressive or progressive properties..." – Foale

More info on the topic of wheel rate.
A link to Tony Foale's discussion of the topic. (http://books.google.com/books?id=84hF-qoR5I8C&pg=PT258&lpg=PT258&dq=shock+angle+foale&source=bl&ots=FYA2xHL6Sk&sig=RlxH7Av0dNCZRE2BjDFuXJng1Vo&hl=en&ei=1TfLSoTOIMSFtgfV-OHhAQ&sa=X&oi=book_result&ct=result&resnum=1#v=onepage&q=&f=false)

[vanzen]

The Hunter rally was just cold and wet.
High point of the weekend:

A rumbling loud / awfully sweet sounding roar of an old hot-rod boxer
pulls up about 10 yards from the vendor tent where I sat.
Even just the SOUND of it commanded my attention.
From my vantage point, I could only see the front of the bike,
it's black Ducati-esque 1/2 fairing, single pin-stripe outlined white painted number plate,
a solitary offset 3" headlight tucked behind a similar sized hole in the right upper corner of the fairing.
I recognized the bike FIRST SIGHT & IMMEDIATELY ...
as it approached and before it ever stopped.
The rider dismounted, being without a stand of any sort,
he rolled the bike over to a stone pier of the Welcome Center,
propped the left valve cover against a stone that was just the right height there,
and quickly walked off.

I had seen this bike in person once before, years ago at the Finger Lakes rally – it's debut to the public ...
Now, so many years later, some may think the machine a little worse for wear,
not so "pristine"
with it's scraped valve covers, a nick here and there in the paint,
a dent or two in the alloy tank ...
but to me, much like a fine wine, better with the aging,
and all that much tastier knowing it has been safety-wired, ridden and raced.

So now, I checked out the bike with a personal agenda in mind.
I was armed with a length of yarn (thanks Helen)
and began in earnest to measure primary pipe length – marked that length carefully with a knot in the yarn,
the Y-pipe – also marked, the collector length, another knot, measured diameters,
and counted the number of discs in the SuperTrapp meg ...
By that time the rider had returned.

'you own this bike ?'
'yup'
'Then you must be Gregor, pleased to meet ya, I'm vanzen ...'
'vanzen ? Right. You've got a website ...been there ... need to update mine ...'
'Gregor, you probably haven't noticed,
but there's a socket-head bolt missing from the fairing bracket
and it seems to have lodged itself right there between the base of the right cylinder and the crankcase.
Bad news is the nut is long gone, nowhere to be found.'
'Thanks, I thought I had heard that fairing rattling about, got a zip-tie ?'
'Sure'

We talked at length about a number of things – boxer-building, chassis design,
the relative un-availability of hot-rod boxer info, websites ...
and I met Scott Kolb (who designed and built the frame).
After all these years
and the fact that this was THE very project
which had inspired me to design and build a bike
that was NOT limited by either available "stuff" or my fabrication abilities ...
It was very cool to finally meet and talk with Gregor Halenda and Scott Kolb.
Gregor: 'I'm thinking I'd like to fit an R6 fork and design a trellis swing-arm with an open drive-shaft'
I guess that a dissatisfaction with the status-quo and a need to go further and beyond
is a shared condition that we both understood ...

A couple of blasts from the starter and a few quick runs through the R's to warm the anxious beast,
a determined and demonic roar of which only a race prepped air-head is capable
was loud and then disappeared just as quickly as it had arrived,
back down the side of Hunter mountain and off into the curving tarmac of the Catskills.

http://www.caferace.com/CompleteBoxer.jpg

http://www.teamincomplete.com/Boxer/boxerhome.html

[BubbaZanetti]

i missed both you and that bike at the rally!

dang

also, good stuff in this thread, keep it going! :thumb

[ducatipaso]

balls! that bike is beautiful!

[vanzen]

The problem with telescopic forks is the inherent fact that they really don't work all that well.
The issue is "bending forces":
1) Lateral flex in the fork legs allows the tire contact patch to move away from the steering axis – a cause of an instability we know as "wobbles".
2) Sliders can move independently of each other – again, the contact patch moves.
3) When a fork is fully extended, there is minimum support for the sliders.
4) Braking causes the chassis to "dive" and chassis geometry to change radically and suddenly.
5) Because of the rake angle, bending loads induce stiction, and the forks must overcome stiction before they can react to the road surface.

This results in a system that is relatively flexible in all directions,
and the small diameter stanchions used by BMW on twin shock air-heads are particularly susceptible to ALL these bending forces.
Most motorcycle manufacturers have responded to these deficiencies by increasing the diameter of stanchion tubes over the years and then by introducing the USD fork.
BMW took a different route and began use of the Telever, a variant of the Saxon / Motodd design, and the Duolever based on the Hossack design –
and the performance of both of these systems put telescopic forks to shame !
(Ickris was right !)

So, what to do about the "al dente spaghetti" telescopic forks of the twin-shock air-head ?
Beyond refitting the forks with a later forks system thereby increasing stanchion diameter,
viable options for improvement exist:
Reduce the unsupported length of the stanchions.
Mechanically connect the two sliders to form a single "structure" –
thereby making the forks system "stiffer".
Stiffness will be defined as the ratio of the applied load to the deflection it causes.
Our ability to increase the telescopic fork's stiffness will apply primarily to lateral loads,
but will also impact longitudinal loads to some degree.
Given that lateral flex in the fork legs allows the tire contact patch to move away from the steering axis, a cause of an instability – this is good enough.

Method:

In any event, start with good internals and PERFECT forks alignment.
Duane Ausherman details the alignment procedure on his site (http://w6rec.com/duane/bmw/fork/title.html).
ALIGNMENT IS CRITICAL TO FORKS PERFORMANCE & SAFETY !

Fit a "dog-bone" brace below the lower triple-tree, and above the height of the front tire, fender, slider-brace
when the the forks are completely compressed.
This brace can be an actual lower triple with extraneous bits removed,
or a fabricated item made to suit.
By "tying the two stanchions together" at this location,
and reducing the unsupported length of the stanchions to a bare minimum –
the amount of deflection possible from any given lateral load is reduced.
This important option is often overlooked and although I have seen the strategy implemented,
unfortunately, I do not have a pic on file.

Install a fork brace to keep the sliders moving together as a single structure.
This can be the "old style" tubular type, paying careful attention to fit them correctly or stiction will be the result:
http://i236.photobucket.com/albums/ff169/53vanzen53/misc%20pics/r100s-forkbrace.jpg

or an adjustable "bar" style like the Telefix:
http://i446.photobucket.com/albums/qq188/orbitangel/800e_1.jpg

I am not convinced that replacing the stock "plate" upper triple tree with a thicker pinch-bolted tree provides any significant real world benefit.
IMNSHO, loads imposed on the stock item are adequately handled
**IF** the plate fits the stanchion caps exactly and is not distorted from maintenance neglect, impact, or other abuse.

Preeminent chassis designer Tony Foale, speaking of his experiences with the stock upper plate:
"I've ridden BMW twins for over 35 years, and started with the Earles' fork
models which also used the flat plate at the top. I've used them on the
road, in enduros and raced them. I've done this with the standard fork
arrangement and with machined yokes and noticed no difference at all.
On most of the leading link forks that I built I went for a similar flat
plate idea (although welded to the stantions), subconciously inspired to a
large extent by the BMs. no doubt.
When you consider the loads imposed on that BMW plate it is quite an
appropriate design for the job.
I think you'll find that most opposition to the standard plate stems from
the fact that it's not how others do it, and it certainly doesn't look
'trick'."

Hoyt, for one, disagrees:
"The issue of course is whether the Beemer item is as good at holding its fair share of that torque
on the tubes, as would be a thicker clamp pinch-bolted to them. It's MHO
that that the BMW top can simply never measure up to the latter in that
respect, as a design method.
That is because no threaded fastener is secure against loads that tend to
rotate it in alternate directions. Even with very long fasteners (that
preserve a lot of preload with some loosening), the above is usually
sufficient to eventually un-load them, and the BMW fork caps aren't long
enough to even have that advantage."

Whichever position is taken, the billet or cast after-market upper-trees that mechanically "clamp" the stanchions
certainly will do no harm.
Unless ...
my negative opinion of these parts is partially based upon bad experience –
The after-market cast upper clamp fitted to my R100RS broke into 3 pieces
as the result of a low speed, low impact crash ...
perhaps a testament to the quality and integrity of the product
rather than it's theoretical or practical benefits to handling.

San Jose's billet clamp:
http://i236.photobucket.com/albums/ff169/53vanzen53/misc%20pics/32_lg-1.gif

stay tuned ...

[RecycledRS]

:scratch nudge, nudge.

[vanzen]

Try this experiment:

1) put the bike on your Harbor Freight lift.
2) clamp the front wheel securely in the wheel vice.
3) get on the bike and turn the handlebars left and right
4) ... and watch the front fender move back and forth across the top tire.

You'll be surprised how little effort it takes.

[RecycledRS]

No question that the fork assemblies of the day don't have much over limp noodles. Most manufactures were experimenting with longer travel suspension and rigidity suffered as a result.

Given this, we have two options to improve the forks rigidity. One is to strengthen the existing unit (which Vansen has touched upon in his previous post) and two is to replace the forks with modern alternatives. The sky is the limit if choosing option two but option ones effectiveness is limited by the original tube diameter. Are there other options?

[vanzen]

... Given this, we have two options to improve the forks rigidity. One is to strengthen the existing unit (which Vansen has touched upon in his previous post) and two is to replace the forks with modern alternatives. The sky is the limit if choosing option two but option ones effectiveness is limited by the original tube diameter. Are there other options?

Any possible option will exist –
from modification of stock
to fitting a USD fork
to designing and building a double link suspension ...
The sky is the limit

http://www.rockerboxer.com/i//2radical.jpg
Wilhelm Beirlein's radical hub-center steering Boxer.

There will be two limitations:
1) your ability to pull it off – successfully, and
2) the ability of the rest of the chassis to handle the load,
i.e. any front suspension improvements must be accompanied by
appropriate changes to the frame and rear suspension
in order to realize a maximum benefit to handling.

In defense of modifying the stock components of the chassis –
the results can be appreciable and quite gratifying.

[RecycledRS]

While for me the front suspension you've shown is far beyond both my skill set and resourses to acquire/reproduce. I have rebuild the forks installed a telefix, billet top clamp and will play with spring rates/oil to suit my taste. The additional brace below the lower triple clamp would present a problem with the fairing on my RS however.

Are we ready to move on to the main frame at this point?

[vanzen]

... Are we ready to move on to the main frame at this point?

Well, yes.
Now that the forks are behaving in a controlled manner,
the forces imposed upon them no longer being 'absorbed' by their flexure ...
the neck-stem will be a virtual "rocking horse" of activity.

http://www.gletter.de/rabruch2.jpg

The "flat" plates that are welded to the down-tubes of the main-frame and
at each side of the neck-stem are intended to control lateral and fore-&-aft flexure here ...
I've put the word "flat" in quotes because these plates are expected to be perfectly flat
but are often NOT.
They can easily be checked with a straight-edge.

Even with regular use and abuse on a 100% stock machine – these plates are easily 'bent'.
When it is discovered that they are not "perfectly flat" –
a check at an air-head friendly dealer that has the factory 247 frame jig
or a trip to the local motorcycle frame-straightening shop may well be in order
to be assured that the frame is still straight and within operating specifications.

Of the six frames in my garage presently, one has plates that are distorted.
Perhaps not a fair or representative example of the incidence of failure,
but still certainly a testament to the weakness of this area and the chosen method of reinforcement.
Stiffer forks, soft compound tires, and improved front brakes will only exacerbate this design weakness.

[vanzen]

http://www.rockerboxer.com/i//frameold.jpg

"The main "weakness" of the BMW series /5 /6 /7 frames is in the area of the SA pivot mounting. Side loads on the rear wheel will tend to make the upright tubes holding said pivot move fore and aft. This can largely be prevented by mounting a bracing tube on each side. These tubes can be attached to high up on the front down tubes or to the underside at the front of the mini backbone. At the rear I mount them above the SA pivot. Some others put the rear mountings below the SA pivot (see p182 of the Clymer manual 500-1000cc twins 1970-1979, October 1979) but the under side is better supported by the bottom rails which are tied to the rigid crankcase anyway. These tubes obviously need to be detachable for engine removal etc. and so the detail of the fixing can be very important to the success of the mod. I don't like the simple drilled flat brackets combined with a flatened tube end that I've seen. I've used mating bosses with one counter-sunk at 45deg and drilled and tapped, and the mating one turned with a male 45deg, the other side counter bored for the head of an allen bolt. These are mounted at right angles to the bracing tube and so the taper takes the main load and the bolt simply holds them together, as long as the bolt is tight then there is no slop, and I've never known the bolts to come loose.
Another though less important frame mod. is to put a short bracing tube from the rear of the top of the head stock back on to the backbone, this will reduce some distortion under braking."
– Tony Foale

The stock Type-247 "GUMMIKUH" frame was conceived, engineered, and developed with an image of the successful 1940s Norton Featherbed pictured clearly in the minds of the German designers.
First renditions of the swinging arm frame were used on Factory "works-bikes", most notably, the RS54 Rennsport racers.
This 1950 era, post W.W.II rendition, was then modified for production road fittings and usage, introduced to the public on the /5 series in 1969,
and used with no appreciable structural, engineering, or design changes until the demise of the Type-247 in 1995.
Outdated ? Ya think ?
As horsepower, contemporary suspension components, capable brakes, and sticky tires are added to an old Type-247 chassis, the stock frame will certainly show it's age and weaknesses, flexing like so much cooked spaghetti, unwilling to cope with the additional stresses without rearing it's head in retaliation.
Solutions would be forthcoming from most every source that sought to run the bike hard and fast.

http://www.rockerboxer.com/i//white_frame_cc_brace_frm_swng.jpg
CC Products "side-braces" featured conical mounts that were welded to the frame. The braces were then bolted to these mounts. Pictured also is a swing arm with triangulated bracing welded in place. A gusset has also been added to the steering neck area of this frame.
This frame and it's modifications, which are loosely based upon the 1970s Superbikes, provides a good model for a "low investment – maximum yield" project.
Conical mount weld-on frame braces, if no longer available from CC products are available from MOTORRAD UNLIMITED (http://motorradavl.com/item--Airhead-Frame-Braces--Airhead+Frame+Braces.html) in Asheville, NC. (no affiliation)
http://teammotorrad.com/media/images/product_xlarge/1_framebraceclose_pxl.jpg



http://www.rockerboxer.com/i//strng_stm_brcng.jpg
Ted Witting's photo highlighting the welded reinforced neck stem area on his "blue turbo boxer"

http://www.teamincomplete.com/Boxer/Resources/thebox4c.gifhttp://www.teamincomplete.com/Boxer/Resources/thebox6a.gif
Scott Kolb & Gregor Halenda's solution: diagonal tubes extending from the neck stem to the main-frame's "spine"
as well as an additional plate to connect the two stock plates.

http://i397.photobucket.com/albums/pp55/Dustdevil/hpn_rahmen_steuerkopf_unten_gr.jpg
HPN doubles the stock plates and securely welds them in place.

[vanzen]

This time Pommie John's race bike:

http://i236.photobucket.com/albums/ff169/53vanzen53/misc%20pics/pommiejohn_racer_bm.jpg

The frame:

http://i236.photobucket.com/albums/ff169/53vanzen53/misc%20pics/pommiejohn_bracedframe_side_rear_tu.jpg

Detail of trans mount to frame connection and diagonal reinforcement of the rear frame cradle:

http://i236.photobucket.com/albums/ff169/53vanzen53/misc%20pics/pommiejohn_chassis_rear_bracing.jpg

[vanzen]

Phreaky Phil's solution:

http://i236.photobucket.com/albums/ff169/53vanzen53/misc%20pics/more_neck_stem_reinforcement.jpg

http://i236.photobucket.com/albums/ff169/53vanzen53/misc%20pics/neck_stem_3.jpg

http://i236.photobucket.com/albums/ff169/53vanzen53/misc%20pics/neck_stem_reinforcement_2.jpg

[RecycledRS]

Pommie John's machine is very interesting and the bracing he employees looks to be top notch. I like the addition of the diagonal braces from the main backbone tube to the swingarm area.

At first I thought that Preaky Phills use of the square tube to reinforce the head tube was over kill untill I saw the blue frame failure.

Questions:
Why does HPN use a supplimental bratcket around the swingarm pivot and would it help on a road/sports machine?
Are there any other HPN braces that may be appropiate?

[vanzen]

I'm not sure I like Phreaky Phil's method of the square tube brace –
A better route to take would be the Halenda / Kolb plate IMO.
Lighter, simpler, and more effective IMO.


Udo Gietl's BMW racers:

1973 BMW MotorSport-framed Butler & Smith Formula 750 Grand Prix.
Rider #17 was Justus Taylor from Vermont:

http://2.bp.blogspot.com/_ao3uUOVmQo8/SQseM2rd-6I/AAAAAAAADw8/1LmaPuPc9sI/s400/86516413_405e04e3db.jpg

Reg Pridmore's ride, a 1974 version of the Rob North-framed F-750 GP bike:

http://4.bp.blogspot.com/_ao3uUOVmQo8/SQseMosE1AI/AAAAAAAADw0/23ymevTG2ho/s400/86516247_3b2b7e8753.jpg

The "model" for any successful stock Type-247 main-frame handling improvements
should be the modified 1976 R90S raced by Butler & Smith in the first season of AMA Superbike Class racing.
In that first season, the BMW team of Reg Pridmore, Steve McLaughlin and Gary Fisher dominated.
The AMA Superbike Class imposed strict frame modification limits,
and Udo Gietl bent them all "a little" to produce a race winner.
Gietl responded to the challenge with this excellent modified-stock frame solution:

http://home.ama-cycle.org/membersonly/museum/images/b41/classjan_400.jpg

Notice Gietl's anti-dive front suspension:

http://2.bp.blogspot.com/_ao3uUOVmQo8/SQsUzn-WFEI/AAAAAAAADwk/XJydSDENwkM/s400/cimg4972mediumna8.jpg


Pommie John's race-bike closely follows the frame modification tutorial introduced by Gietl,
and is, IMO, a fine execution of theory and construction methods.

[Dave Backmarker]

Anton, just for poops and giggles, How would you measure the change in spring rate?

I pulled out my old Physics books from Uni. and can't get my head around the concepts anymore. Been way too many years since I barely scraped thru those classes.

To measure the change in spring rates you could place a weight on the spring and measure the compression. Keep adding weight and each time you add weight, you measure the distance. Linear spring formula is F = Kx where F= force, K = spring constant and x = displacement. So, Divide the weight (force) by the compression distance (x) for each weight and when the ratio changes, you're discovered the displacement where the second spring rate or progressive nature of the spring kicks in.

[RecycledRS]

Any idea why the anti-dive front end on Gietl's bike (and a few others of the period) did not survive? It seems a very simple solution to the dive issue.

[lostboy]

The anti-dive linkage was actually developed by Todd Schuster, Udo's partner in crime. The parts are Cummins diesel rocker arms and pushrods!

Chris Hodgson had something similar on one of his street bikes, but he fabricated his from chrome-moly tubing.

[rpeckham136133]

Man, the things you learn here....

Cummins diesel!

:clap