Awesome work Steve. 

 

I suspected but did not know as fact that the drive filled with oil, particularly on the C drive.  The B drive incorporates a pumping mechanism formed by the spinning ring gear and a sloped pocket machined into the housing.  How effective this is I do not know but it was not possible to do the same on the C drive since the ring gear is stationary.  I accepted the possible power losses from excessive oil rather than going for the weight and complexity of a scavenging pump. 

 

I should also add that I do not know for sure that the stacked needle bearings are actually needed in the C drive.  The single roller bearing in the B drive may be adequate and it is the same thickness as the two stacked needle bearings.  This would eliminate the shaft and race wear problem completely.   I plan to change over to the B drive bearing the next time I look inside the drive.  Even if the life of the bearing was reduced by 90% it would still be a 1000 hour replacement interval.  I realize that bearing wear is not linear with speed though.

 

Still stuck at the house for the last month due to the river flooding the road to Shady Bend.  Down to my last frozen pizza, hope it don’t rain and the creek don’t rise 😊

 

Tracy

 

Sent from Mail for Windows 10

 

From: Steven W. Boese SBoese@uwyo.edu
Sent: Saturday, January 12, 2019 9:41 PM
To: Rotary motors in aircraft
Subject: [FlyRotary] stacked thrust bearing

 

 

Tracy and Neil,

 

In order to prove to myself that the RPM is divided between the stacked thrust bearings, I installed a Hall effect sensor in the mounting plate and silver brazed triggers to the middle thrust washer.  An inductive pickup was installed to monitor the rotation of the flexplate (input shaft).  This allowed the determination of the input shaft RPM as well as the middle washer RPM under various conditions. The load on the drive was from a 3 blade 72" dia  Warp drive prop with the tip angle set to 20 degrees.   

 

All new bearing components as used in the -C drive were tested along with a used cupped middle thrust washer from a C6 transmission.  The cupped thrust washer was tested because it was possible that the cup overlapping the 3.5mm shaft side thrust washer would prevent the middle washer from contacting and causing wear of the input shaft.  After seeing the results from the used cupped washer, new cupped washers were obtained and tested also.

 

In the course of testing the thrust bearing configurations, it was found that the reduction drive always operates completely filled with oil.  The added triggers may have affected the behavior of the bearing stack due to drag between them and the surrounding oil.  To see if this was the case, the drive was modified such that the drive contained about 5 to 6 oz of oil during operation rather than the about 20 oz of oil when completely filled.

 

The results are shown in the attached plots.

 

These results indicate that, with all new components including a new flat middle washer, the RPM was reasonably well divided between the two thrust bearings and affected little by the amount of oil in the gearbox.  With the used cupped middle thrust washer, essentially all the RPM was seen by the mounting plate side bearing.  With the new cupped middle thrust washer, the RPM  distribution between the thrust bearings was affected by engine RPM when the gearbox was completely oil filled but little affected by RPM when the gearbox was drained.

 

All the results were obtained with an input shaft that didn't have a groove worn in it by the middle thrust washer.  

 

My conclusion is that with all new bearing components and unworn shaft the system behaves as desired.  With used components, the behavior is unknown.  The cupped thrust washer is not a solution to the shaft wear issue because of the uneven RPM distribution when submerged in oil and since shaft wear has resulted from the bearing cages as well as the middle washer. 

 

The internal gearbox pressure with -6 drain lines was about 4 psi when cold and about 2 psi when warm which may have implications in terms of input seal leakage and the seal coming out of its bore.  The completely oil filled gearbox consumes extra power which is converted to heat but apparently not so much as to cause serious problems.

 

All the above are simply results of my curiosity.  You are most welcome to draw your own conclusions.

 

Steve Boese