Have not gone out and checked my rivets on my R1C, will do that this weekend.  However, there is one possibility for the cause of the smoking rivets that I have not heard addressed and that is vibration nodes of a rotating disc.  There is the possibility that if the disc were in some oscillatory mode that the smoking rivets might be caused by with the rivets  being at the nodes of certain modes (like that?) of vibration which could be causing the holes in the discs to be sliding up and down the shaft of  rivet.

 

In the end, it may be important to know what is causing this.  The 20Bs should be smoother than the 13Bs, on the other hand, they do have a different firing frequency.  For the same RPM the number of firing pulses from the 20B are 1/3 higher than the 13B over the same time period.   That could possibly induce vibration nodes in the disc that the 13B (which has longer duration between firings pulses).  The long time between firings in the 13B might permit the mass/structure of the assembly to dampen the amplitude of any such induced vibration  between firings pulses – where  not so for the more frequent pulses of the 20B.   The higher the frequency of the input, the stiffer structure that can be caused to vibrate.  A vast over simplification, but thought I would throw it out just incase somebody had a finite analysis package sitting on their computer.    Just a WAG.

 

Here is an URL that provides a drawing of how such nodes might form on a disc

 

http://demonstrations.wolfram.com/NormalOscillatoryModesOfRotatingOrthotropicDisks/

 

 

Al,