[RCVTR]

I thought this might be of interest to some of the members here.

I've been working on a model of my chassis, to determine the effect of the various modifications I've made to the chassis geometry. It has turned into quite a science project. I've learned a bunch and have developed an accurate model that I can continue to develop over time and make informed decisions about where to go from here. My original intent was to get the bike measured at the nearest Computrack facility, in Roseville, CA, but they are no longer in the measurement business and have sold their equipment. I decided that I would learn more, by figuring out how to measure my own bike and model the evolution of the chassis geometry, from the stock configuration.

I purchased the Motorcycle Setup Analysis software, developed by Tony Foale. It has some useful tools for analyzing geometry and suspension. But my first goal was to determine if my measurements were reasonably accurate, before plugging them in to the modeling software. So I have done a combination of building a 2D model of the chassis and entering the data form that model into the analysis software, to be sure that I was working with valid data.

The first thing I did was to hang the rear end of the bike and support the swingarm in the rear stand, so that it was just making contact with the stand with the rear shock fully extended. Then I removed the rear wheel, rear shock, rocker and trailing link. I reinstalled the rear axle against the stops, so that it was in the installed position. I then had the swingarm in the fully extended position, with nothing in the way, so that I could measure the distances between all of the suspension mounting points. As a final measurement, I reinstalled the Moriwaki rocker and trailing link and measured the center-center distance for the rear shock mount, to get an accurate measurement of the “phantom” shock length with the suspension in its measurement position.

I measured carefully from edge-edge on the holes, then added the hole radius at each end to get the center-center distance. Where the holes were not on the same plane (rear axle-to-upper shock mount, for example), I put a 10mm drill bit in the hole to extend the hole position out to the measurement plane. I could then accurately measure to the drill point. I took the measurements carefully, but I really was not sure how much error I had.

I then drew the geometry from these measurements in AutoCAD. I drew circles with the center at one geometry point with a radius equal to the distance to another geometry point. The circles intersect at two points, one of which is a point on the suspension. I then drew lines from the centers to the intersections to get a vectoral representation of the geometry.

I had the position of the upper shock mount from 3 different locations. The difference of the intersection points for the 3 circles, should give a pretty good idea of the accumulated error in the measurements. The error was less than 0.5mm, so I knew I had a reasonably accurate model of the suspension attachment points on the chassis.

I could then add the stock rocker, or the Moriwaki rocker, the trailing link and the rear shock and establish a fixed, unladen position of the swingarm, relative to the main chassis:

Rear suspension with the stock rocker and 325 mm shock length:

Rear suspension with the Moriwaki rocker and 335 mm shock length:

 

[RCVTR]

I needed to know the distance from the swingarm pivot to the steering axis and the geometry of a stock RC51. I did a search and found Computrack data for a stock bike at RC51.ORG. That allowed me to draw the rest of the chassis and error check it against the Computrack measurements. When I put the chassis geometry I had developed on tires of the same diameter, with same wheelbase and swingarm pivot height as the published measurements, I got nearly identical results. I then plugged these numbers in to the Setup Analysis software.

From there I could implement each of the chassis modifications and check my geometry measurements with the ACAD model and the software model. The first thing I did was add 10 mm to the shock length in the software model. I used the new swingarm pivot (SAP) height and rotated the ACAD model around the front axle to put the SAP at that height, then rotated the swingarm to put the rear wheel on the ground. Then I redrew the rocker, trailing link and shock vectors and took measurements. The models again gave the same answer.

Next, I lifted my bike, so that the wheels were just contacting the floor and measured the swingarm pivot height, axle heights and wheelbase in the fully unladen position. These measurements agreed within 1-2 mm with the modeled chassis, with the same suspension modifications in the same position. So I have verified that I have an accurate model and reasonably accurate measurements of my chassis. I’m building a very flat, level workstand, so I will be able to refine my measurements.

The geometry modifications I have made are to the rear suspension only (Moriwaki rocker and +10mm shock length). The front ride height is unchanged from stock, with an SP-1 fork, installed at the first line at the top of the triple clamp. The fork has 25mm preload spacers installed to get proper preload and sag numbers in the middle of the adjustment range.

The drawings below indicate rake, ground trail, real trail and swingarm angle for various permutations of rear suspension geometry. The real trail is the measurement perpendicular to the steering axis to the center of the tire contact patch. It is the true lever arm that acts on the steering axis. The swingarm angle, rake and trail have changed dramatically.

I’m down to ~100mm of ground trail. I noticed that the bike wants to shake its head under hard acceleration. I think this is partly due to the short trail, but I also had a wheel misalignment, which was probably the major cause. I was not at the limit of the front tire, so I can’t say how the trail feedback through the bars was. The steering was very light on initiation and neutral midcorner. Every track day was rained out in the late Fall, so I have not been back for more testing.

The data from the Tony Foale software is shown below, for comparison with the ACAD plots. I’ll come with weight distribution, sag and suspension plots, later. They will show a comparison of the stock rocker vs. the Moriwaki rocker and the effects of spring rate changes.

 

[SubSailor]

Interesting read. Can't wait for part 2.

 

[20_RC51_00]

cool info mang. good work!

 

[bensexcessbikes]

Very interesting, in-depth analysis. This is the kind of minutia hardcore enthusiasts enjoy. :woot:

 

[bmfgsxr]

great stuff. this is the type of info that companies like GMD computrack have when measuring bikes to straighten frames, etc...

 

[DesperateSP2]

Excellent read, thanks a lot. :notworthy

 

[RCVTR]

Once I had this, with cell-by-cell mapping of the part-throttle region, I am no longer in pursuit of more go. At 408 lbs wet weight, it goes just fine.
Now it's a matter of getting it to the ground.

 

[SubSailor]

Wow, did you manage to map out the massive lean spot from 3200 to 4200 RPM?

 

[bmfgsxr]

your not kidding, that is a dead flat hp and tq curve. mine has a fair dip around 5k that im hoping i can get tuned out at my dyno session

 

[RCVTR]

Wow, did you manage to map out the massive lean spot from 3200 to 4200 RPM?

Yes. I started with the map form the Dynojet site for an Akrapovic full system and the thing was falling on its face at WOT, below 4k rpm. On the dyno, it seemed like it was going to stop running.

The final torque curves are covered by the text, but you can see the HP curves. The response is very flat and linear everywhere.

I was running a race slick that is 30mm (radius) taller than the stock tire. I think the peak torque and HP would have been a bit higher with a standard street tire.

 

[bmfgsxr]

race tires always put less hp to the ground its just the way it is, but you always tune a bike the way its going to be ridden. my #'s are with a brand new 209gpA 190/55 med/hard compound tire on it.

 

[RCVTR]

Not a lot to add today. I measured the weight distribution a few days ago and calculated the centroid positions, unladen and with rider (me) suited and in a tuck:

I've looked at my sag numbers in the model and decided to change the rear spring rate. I'll explain more and post some images when I get the chance.

 

[bmfgsxr]

you must be an engineer im thinking, rcvtr.. great work. btw, on a side note when you get the cans can you weigh them so i can see what kind of weight loss im getting by changing to the gp style cans? thanks

 

[RCVTR]

Now it gets a little more interesting. I never really understood how a Moriwaki rocker changed the rear suspension characteristics from the stock rocker. I knew that the stock rocker provided higher leverage, to allow 2-up riding, without bottoming the suspension, but did not know how to characterize it.

Last winter, I had my Ohlins rear shock serviced by Stig Pettersen. I gave him my weight and talked to him about his recommendations. He recommended a lower spring rate, with more preload, so I dropped from a 100N/mm spring to a 90N/mm. But I forgot to mention that I had the Moriwaki rocker installed. My setup was a compromise between not enough unladen sag and too much laden sag. I was also measuring sag with the bike on the rear stand, thinking it was close enough. I have a good friend who knows a lot about RC51 chassis setup and he told me that the 90N/mm spring was too soft for my weight, with a Mori rocker installed. I believed him, but my error-prone measurement was giving me 5mm unladen and 25mm laden sag, which I knew to be good numbers.

I entered the suspension geometry into the setup analysis software, with a stock rocker and a Moriwaki rocker and entered my unladen and laden rear wheel loads, from the weight distribution measurements shown above. The software gives a pretty good graphical representation of the geometry:

I plotted the wheel force vs. suspension position and sag positions (black, vertical lines), both laden and unladen, for the stock and Moriwaki rockers, with 100 N/mm and 90 N/mm springs installed. It is interesting to note that the stock rocker produces about the same peak load with a 90 N/mm spring installed as the Moriwaki rocker with a 100 N/mm spring, but the Mori rocker does it with 140 mm of suspension travel vs. 125 mm with the stock rocker.

So that’s how it works. I’m going to reinstall my 100 N/mm spring. I’ll have about 30mm laden sag, but I have 15mm more travel to work with.
I’ll be doing more analysis later, but that’s where it stands, for now.

 

[RCVTR]

Plots for a stock suspension rocker:

Stock Rocker, 100 N/mm spring , unladen

Stock Rocker, 100 N/mm spring , laden

Stock Rocker, 90 N/mm spring , unladen

Stock Rocker, 90 N/mm spring, laden

 

[RCVTR]

Plots for a Moriwaki suspension rocker:

Moriwaki Rocker, 100 N/mm spring , unladen

Moriwaki Rocker, 100 N/mm spring , laden

Moriwaki Rocker, 90 N/mm spring , unladen

Moriwaki Rocker, 90 N/mm spring , laden

 

[SubSailor]

Good work!! Keep it coming.

 

[RVTMAVERICK]

Too Cool, Awesome Info.!

Hey RCVTR,

This is Awesome mang., One of the most interesting reads in along time here, So cool how you've gone through this step by step, explaining everything and potting it all out for us to see!:clapper :rockon

Thank You!


Peace Jeff

 

[DesperateSP2]

That's one of the best read this read.
I wonder if you and Bruchi could come up with a version to sell to the public or suspension workshop... could make you a few bucks. :rockon