skimmy problems

Shimmy Problems

At low speeds, 20 to 30 mph, the Quad had a tendency to shimmy.  The effect is exacerbated by rough roads or by deliberately wiggling the handlebars to bring on the effect.  The effect is most pronounced when in the process of moderate acceleration through this low speed range.  At higher speeds, the shimmy tendency completely disappears.  When riding in this shimmy prone speed range, it was necessary to keep a very firm grip on the handlebars … no waving!  The shimmy is manageable and predictable and at first, I thought that it might simply be characteristic of all trike designs and I would just have to live with it.  This demands rider caution when slowly accelerating through low speed ranges of 20 to 30 mph. and/or negotiating rough road surfaces at low speed.  It is possible that the use of smaller lighter weight tires might have better dynamic performance and have less of a tendency to shimmy.  Perhaps a different shorter torsion suspension system leading to narrower outrigger width would have beneficial effects as well.  Both of these alternatives would appear to be in the right direction.

The three and four wheeled Trike/Quad configurations are statically very stable but like the motorcycle itself, the ultimate handling characterizes depend on the proper balance of many factors, static and dynamic.  Manufactures spend much engineering talent and testing at arriving at the best compromises.  As I do not have such resources, a little trial and error adjusting had to do.

After doing some web research, I found that it is not uncommon for trike configurations (both standard Tikes and Quads) to have low speed shimmy tendencies in the 20 mph. range.    There are numerous posting on trike websites addressing this issue.  I recently had the pleasure of riding a custom fabricated Trike+Plus designed and built for a successful physician who purchased an earlier edition of this CD.  A beautiful fabrication job!  The suspension concept was considerably different, yet I was surprised to find that it too had a tendency to shimmy at about 20 mph.  The shimmy was mild and only noticeable if one relaxed their grip on the handlebars.  It was nothing of consequence yet it was evident, apparently inherent to the basic trike configuration.

Some riders have taken the approach of installing a damper on the front fork to change the natural frequency of the front end.  I was surprised to find that most of the modern high performance sport bikes now incorporate built in damping (either hydro-mechanical or electronic).  Adding a damper is a touchy move as too much damping could make the steering very stiff and unresponsive.  Fortunately, these dampers are adjustable to fit the situational need.  If a damper is employed in the Quad configuration, it may easily be readjusted when reverting to the basic motorcycle configuration.

Before resorting to a damper, I decided to first seek out the possible sources of the problem and make what adjustments I could. 

Shimmy problems appear to stem from several specific sources.

Frame Flexibility: The most obvious and easily dealt with question is lack of rigidity or misalignment of the outrigger frame.    Rigidity did not seem to be a problem as the outrigger frame is remarkably ridged, solidly attached and overdesigned for its purpose.  The outrigger frame and bike frame behave as one integral unit. 

I mentioned earlier in this narrative, that I had initially fabricated an outrigger lateral bracket to add additional rigidity.  This lateral bracket also bolts on at the passenger foot peg location.  I found it to be unnecessary on the lighter weight Honda Shadow and I earlier couldn’t tell any difference with or without it on the heavier Yamaha Roadstar.  I still consider this lateral brace to be unnecessary.  Nevertheless, in addressing the shimmy problem, I decided to reinstall it on the Yamaha just in case it might make some difference.  This added degree of rigidity made no difference whatsoever as rigidity proved to not be the principal problem.  The frame and its connections were quite adequately rigid.

Frame Alignment:  Careful measurements verified that the frame and axle line were in close alignment.  As I mentioned earlier, the part of the frame that should be most closely aligned with the motorcycle is frame element B which supports the suspension swing arm.  Alignment of this element can be quite easily verified by lining up a six foot level with frame element B and measuring the standoff distance from some symmetrical part of the bike (crash bars for example).  Right and left sides should of course be equal.  You can also place the same six foot level against the outside of the tires and measure to be sure both sides yield equal measurements.

Rear Tire Pressure:  Another possible source of problems stem from improper or unequal outrigger tire pressures.  The trailer tires I used are nominally rated for full load at 32 psi.  I chose to reduce this pressure to 20 psi as I was running only a very light load on each and the Quad tends to naturally ride somewhat stiffer and rougher than the motorcycle itself.  I set the rear wheel of the motorcycle at its lower recommended light load pressure of 32 psi.  After a few trails, I concluded that outrigger tire pressure was not a significant factor as I could see no change whatsoever in the shimmy properties at various rear tire pressure settings.

Wheel Imbalance can be another major source of such problems.  The tires were dynamically balanced when they were installed at the American Tire Store as I watched.  I had no reason to believe they did the job improperly but I nevertheless went back to the shop and had them check the balance once again.  To their chagrin, two of the Carlisle balance weights had come loose and had to be repositioned.  The rebalancing, however, made no difference in the shimmy properties.

 Wheel alignment:  I was able to easily adjust the cant of the wheels by simply tightening the adjusting bolts until the wheels were vertical (verified by using a large carpenter’s square and/or a level at the outer rims).

The forward tracking of the wheels was more difficult to measure.  I used three different methods to accomplish the track alignment task.  The first and perhaps most complicated required that I establish a centerline reference for the motorcycle.  To this end, I laid down a 25 ft. color marker snap string line on the concrete driveway leading to the garage wall.  I then carefully rolled the Quad into position such that the front and rear motorcycle tires were exactly centered on the line.

  I next made a simple plywood jig that fit on the 13 inch metal wheel rim, secured in three rim locations by strong magnets that I salvaged from an old bygone tank bag (I keep everything).  Once I was confident that the jig was mounted true and flat on the wheel rim, I attached my laser pointer flush to the wheel jigs one at a time, and projected the beam 25ft. on to the garage wall.  I marked the spots and measured first the total width between the dots and secondly, their relation to the centerline of the motorcycle.

If you go online, you will find a host of professional wheel alignment tools ($$$) to do the same job more elegantly.  Since I planned to do this job only once and perhaps check it again at some future date, the investment was unwarranted.

  The “laser” that I used was a simple red beam presentation pocket pointer purchased at Staples for a few dollars.  It projects a bright red beam from the body and also has a flex-neck flashlight as well.  Be sure the body of the pointer it is laying flat on the jig if you use it.   I had to grind off a small lip at the rear of the pointer to achieve this.

  When the wheel tracking alignment was correct, the dots were equidistant (L₁ = L₂) from the bikes centerline and the distance between the dots (L₁ + L₂) matched the width between laser pointer positions back at the wheel jigs (L₃).  I wanted to completely eliminate wheel alignment as a possible source of any shimmy tendency so I carefully aligned the wheels to a fraction of a degree (really paranoid).

You will find it handy to have both regular length and “short” wrenches when working under the frame adjusting the alignment bolts.

A second even simpler way to determine the track alignment is to simply use the good old carpenters square laid against the tire with the laser pointer attached, bypassing the need for a jig of any type.

Carpenter’s Square Technique

The third and simplest method of all is to lay a 6 ft level against the outside of the tires and measure the equidistant standoff distance from some symmetrical aspect of the motorcycle, such as the outside of the crash bar.  I repeat this exercise each time I reinstall the outrigger frame on the bike.  I originally used all three methods to exactly align the track of the outrigger wheels.  It is important to stand on the frame, “jiggle” the bike or roll it back and forth a few times before concluding that your adjustments are final.

Dynamic Instability:  This results from the interaction of the two suspensions and the mass distribution of the assembly … sophisticated stuff!  Each motorcycle/Quad combination potentially presents a different situation.  The motorcycle suspension itself has a natural frequency as does the outrigger suspension.  When the oscillations of each reinforce one another, a shimmy or dynamic instability will result.  Intuitively, one realizes that the substantial widely spaced heavy outrigger wheels lead to a “dumb-bell” effect that could enhance oscillation once it starts.  It is thus desirable to minimize the width of the outrigger wheel assembly.  This kind of analysis is beyond the garage mechanic level.  (See Section 7.2 “Shimmy Analytics).  It is clear that changing the stiffness of the front wheel steering by using a “damper” can help reduce a shimmy in most applications.  A wide variety of dampers are available on-line designated for specific motorcycle models and costing from $50 to $600.  Check YouTube for videos illustrating the simple installation of mechanical rod/cylinder damper types.

Damper Installation:  I purchased an inexpensive mechanical damper (from Thailand!!) which cost me about $50 including shipping charges to see if this would make a difference.  The damper theoretically is specific to each model of motorcycle; however, since in comes with no installation brackets; I suspect that this supplier actually sends everyone the same device.  The fittings are metric.  The damper has seven levels of adjustment which will change its stiffness.  When riding without the outrigger, it can be set to its softest position; however, I left mine at the “max” setting and found the ride totally satisfactory.

I made two brackets to attach the damper.  One is bolted to a hard point on the “freeway bar” (we called them crash bars in my time) and the other is bolted to the underside of the front fork assembly.  The final assembly looks a little dorky in my opinion but it does work.  Be careful to install the damper in such a manner that it is clear of the front fender when the fork is fully depressed and clears the front forks when the wheel is turned to the stops.

Results:   Success!  The notch shimmy mitigated!  I found that while slowly accelerating through the 20 to 30 mph. range, I could feel the very subtle tendency to shimmy but it is sufficiently suppressed that I no longer had to keep both hands firmly on the handlebars or even think about it.  I found that I was still able to force a very mild reaction by deliberately oscillating the handlebars in just the right manner in the 22 mph range.  The combination of adding a damper and extremely carefully aligning both the frame and the wheels appears to have made the difference.

Honda Shadow Adaptation:  I was surprised to discover that when I attached the outrigger frame to Charlotte’s Honda Shadow, I experienced the same shimmy effect between 20 mph and 25 mph.  This is a considerablly lighter bike and the effect was milder but nevertheless evident.  It appears as if there is something fundamental to the configuration that leads to this notch shimmy at a particulr low speed.  I added a damper, as I did with the Yamaha, and the shimmy was greatly supressed.  I now conclude that a damper is an intergal part of making such a conversion work and that the velocity at which the notch shimmey sets in is totally determined by the outrigger assembly configuration alone.

In the next section entitled “Shimmy Analytics”, I quantitatively explore the Trike+Plus shimmy tendency and how the design choices (suspension stiffness, wheel size and weight and outrigger frame width) effect the outcome.  Please read it for additional facts and insight.  The ultimate solution, however, is to install a damper.  If you are not thrilled with technical details, simply skip to the conclusions at the end of the section.

Edition 9 Insert to Section 7.1 “Shimmy Problems”:

As previously discussed, I orignially installed the damper in the manner illisturated in a number of youtube videos in a forward and aft position inline with the centerline of the bike.  Although it seemed to surpess the shimmy motion, I could still feel the underlining tendency and I wished to do even better.

Even the light weight damper that I originally employed noticabilly protrouded ahead of the front forks and certainaly did not look all that good.

The damper performs it function by moving a resistant cylinder down a central control rod.  Resistance is produced only when movement occurs.  The youtube illustrated mounting approach attaches one end of the damper to the bikes frame and the other end to the front wheel steering assembly.   As the handlebars are turned, the distance from the front wheel mounting point to the frame mounting point is reduced, forcing the cylinder to move along the control rod.

To be effective in surpressing  a shimmy type  motion, the damper must have its maximum effect when riding in the straight ahead situation with the front wheel at zero degrees turning angle.  The best mounting technique to result in maximum effect in the zero steering position would be to place the damper in a laterally rather than in a forward and aft position.  When placed laterally (side to side),  the damper is fully functioning in all steering positions and particularly effective straight ahead.

To accomplish this type of installation, I fabricated an extension arm that protrouded approximately 9 inches from the centerline of the bike.   I made this from 1-1/4 inch angle iron and secured it to the bike by using two convienent bolts that attached the crash bars to the frame.  I then mounted one end of the damper to the arm’s outermost end and the other to a bracket from the steering head.

I also decided to use a more robust automotive type damper after reading posts on an Australian chat site dealing with stabilizing side car configurations.  Side cars are much more popular overseas and there was a wealth of experience exchanged between owners.  The most popular damper turned out to be a stock Volkswagen hydralic steering stabilizer.  I located one one at the J. C. Whitney site (www.jcwhitney.com) which cost $46 inluding shipping.  The part number stamped on the body is AZ1OHA .  The compressed length is 14 inches.   The fully extended length is approximatley 19‑1/2 inches.  The maximum diameter is 1-5/8 inches.  The long tubular portion’s  diameter is 1 inch.

As you an see in the photograph, it has some peculurarities with respect to the curvature of the shaft end and the presence of a bushing type head for  use as an end mounted strut.

To be useful for my application, I had to fabricate my own cylindrical clamp to grip the body of the tube.   I did this by bending a metal strap around a suitbilly sized pipe.   This clamp is attached to the steering head bracket with a swiveling bolt.   The swiveling bolt is necessary as the angular relation changes as the damper moves through its positions when the wheel is turned to the limit.  I also cut off the unneeded bushing end using my Makita hand grinder and carefully straightened the curved fitting at the end of the shaft after heating the part and being very careful not to bend the shaft itself.

As I mentioned, I rigidly attached the shaft end of the damper to the 1-1/4 inch angle arm.  For the geometry that I chose, this joint did not have to swivel.  At the other end, the clamp on the cylnder must be attached to the steering assembly at a point removed from the steering axis by several inches in order to create the necessary movement when the wheel it turned.  Rather than fabricating a complicated bracket, I decided to use the wheel lock as the attach point since I never use the wheel lock and have never used it on any bike I have ever owned.  To make this into the clamp attach point, I cut off the upper portion of the wheel lock, leaving only that portion that was a part of the front wheel assembly.  Your bike may not have this configuration and hence you will have to fabricate some kind of bracket to attach to the damper.

I next cut a about a six inch length of 3/8 inch diameter threaded rod and used a nylon lock nut at the bottom and jam nuts and washers along its length to rigidly secure the threaded shaft to the wheel lock while letting the lower connection swivel as the damper moved in its track.

I eventually painted all of the components to make them even less noticeable and better blend in with the bike.

Eventhough the damper is robust, I could not feel its presence when riding the motorcycle in its basic non-Trike+Plus configuration.  I would have expected to feel some resistance or sluggihness in the steering but that proved to not be the case.  The leverage between the force exerted by the damper to that felt on the handlebars is about 5:1 so the manual steering force easily overcomes that of the damper resulting in no feeling of steering resistance .

The most important aspect of this whole endevor is that the shimmy tendency was now so surpressed that  I was no longer aware of the problem as I rode.  I can now rapidly or gently accelerate through the 20 mph. speed range with no sense of a shimmy.  Shimmy has never been a problem at higher speeds and such remains the case.

As you can tell by the amount of space that I have devoted to the problem, it is difficult to get at the heart of a shimmy problem and difficult to come up with the correct solution for you particular application.  I would always first look to alignment problems and do such a good job that this factor is eliminated from further consideration.  It is worth experimenting with tire pressures (front and outrigger).  But alas, if these factors do not solve the problem, don’t hesitate to properly install a sturdy front end damper device. 

I hope that those of you who purchased my CD learned and benefited from my experience and will end up with a good looking safe design of your own.