Clutch

Improving the OSSA Clutch

Something that immediately impressed me about my Electric Motion Race is how well the clutch works – especially considering how similar it is to the recalcitrant OSSA clutch.  Specifically, I was shocked by how little finger effort is required to operate the EM's clutch.

When my EM clutch was apart for service, I noticed the spring force was so light that I could easily push on the “throw-out bearing” with my thumbs to disengage the clutch.   That's not at all easy to do with the OSSA.

EM's use of 4 friction plates versus 3 in other designs does give more surface area, so the spring force could be reduced for the same torque-handling capability.  Conversely, the EM's 45 Nm maximum torque is probably greater than any other trials motor.  But an electric motor's peak-to-average torque (per revolution) is nearly constant, so that makes life easier on the clutch.

Bottom line, the EM clutch works beautifully, and it seems the OSSA's clutch could be improved.

Clutch Pack Thickness

With all other variables being equal, the thicker the clutch pack, the greater the effort required to pull in the clutch lever.  Conversely, the thinner the pack the greater the propensity for the clutch to slip.

The first two XiU-rdi clutch packs I received measured 9.91 to 9.92mm, which is too thick.  I ended up using one of my original worn friction plates for a while.  In the XiU-rdi system, the three fiber plates typically measure 2.12 to 2.13mm thick, and the steels are 1.49mm and 2.02mm. 

1.49 + 2.02 + (3 * 2.13) = 9.90

See the section My Upgrades, Clutch Control Ring for more information.

As a side note, steel plates for the GasGas are available in four different thicknesses: 1.3, 1.4, 1.5, and 1.6mm.  This is a great tuning aid, but no such option exists for the OSSA.  Unfortunately, the GG plates won't fit the OSSA.

Clutch Release Arms (aka Fingers)

The 15mm test plug allowed me to determine that the clutch release arm height was not completely uniform. 

A Japanese repair video showed tweaking GG fingers with the tap of a hammer.  This looked like a good idea to me.  With each finger laying on a perfectly flat surface, I pressed on the wide end with my thumbnail.  If the narrow end moved up more than a very small amount, I clamped the finger in a vise and tweaked it with a tiny hammer.  I reasoned that if the narrow end did not move up at all, the finger was likely bowing up in the middle.

This process could ensure all eighteen fingers were as flat as possible.

Hydraulic Ratio

The clutch master cylinder has a 9.5mm bore.  The 2014 slave cylinder has a 14mm bore.  Hydraulic ratios scale on piston area, so the bore diameters must be squared to make a comparison.

2014 bikes: 14² / 9.5² = 2.17 (You can complicate this calculation by using pi and the radius, but the ratio works out the same.)

This means that the master piston multiplies the force applied to the slave piston by 2.17.  To do this, the slave piston must move 2.17 times farther than the master piston.  Thus force is traded for distance.  One way to remember which way the force multiplication works is to think of the large pistons on a disc brake caliper.  They move a small distance and produce the large force needed to scrub the pads against the disc.

But there is a limit as to how far this force multiplication can be taken.  The slave piston must move far enough to ensure the clutch discs can separate enough to completely release (slip). 

Although I don't have one to measure, the OSSA Factory models used a different diameter slave piston.  It was likely 15mm.

Factory bikes: 15² / 9.5² = 2.49  

This gives a greater force multiplication (making for an easier lever pull) but also a smaller separation of the discs in the pack.  This is probably good for a high-level rider who wants a more binary clutch operation, with a smaller slip zone.   The Factory model may use a different disc spring as well, but I don't have the part number to know for certain.

My notes are a bit unclear, but I think my 2011 TR280i probably used a 13.5mm slave piston.

2011 TR280i: 13.5² / 9.5² = 2.02

The 2011 bike definitely has a stiffer lever than the 2014/2015 bikes.  OSSA calls the slave piston a “clutch pump piston.”  The 2011 part number is 2620020211 versus 2620020214 for the 2014 models.

Disc Spring Engineering

The design equations for a disc spring are not as simple as those for a wound wire spring.  Typically, the manufacturers of such springs provide software tools to perform the calculations.  The screenshots below were created in about 2014 from a manufacturer's website.  Unfortunately, that website no longer exists.

Besides the elastic modulus of the material, four dimensions affect the spring's deflection versus force curve: material thickness, outside diameter, inside diameter, and overall height.  

In an effort to lighten the lever pull, I removed about 0.006" (0.15mm) from the inboard side of my 2011 disc spring.  I did this by sanding it on a surface plate.  A large cylinder of aluminum was machined to fit inside the spring's ID.  This gave me something to hold onto while sanding and assured even contact of the spring against the sandpaper. 

Removing this very small amount of metal effectively reduced the disc spring's clamping force by about 8 - 9%.

The standard clutch spring changed in 2104.  The old part number is 2420020211.  The new part number is 2420020214.