I pulled my trans (1993 MR2 Turbo) yesterday as I am swapping trans again (that's a long story, more on that below). I decided to inspect my clutch and flywheel at the same time. They have less than 4000 miles on them, and I replaced the clutch in May of 2011 (it is Oct of 2012 as of this post).
Flywheel surface, disk pucks, and pressure plate friction surfaces all look good. However, I discovered massive fretting of the clutch hub splines.
![]()
The input shaft shows indications of fretting in that the surface is coated in the oxidized powder. However, the splines appear to be fine, and I felt them as well. Feeling them reveals no abnormal wear on the splines, i.e.: I drug my fingernails from the part of the spline not engaged to the clutch disk hub to the part engaged, and there were no steps from wear. This is not obvious from observation, or from the photos, which is why I felt them. The photo below makes it appear as if there may be a step due to fretting wear, but there is none.
![]()
From my Machinery's Handbook, 24th edition, page 2053:
![]()
My summary: low amplitude oscillation movement between the male and female splines, results in fretting
Once you get a little bit of oxidized steel (rust, but not caused by corrosion), the powder increases wear of the splines dramatically, as it acts like sand in the works.
I also found this paper online:
"Surface Durability of Coaxial Splines" Nembered page 18 (pdf page 26) starts the discussion on fretting:
http://fcp.mechse.illinois.edu/reports/FCP_Report177.pdf
Background: My 1993 E153 w/LSD (original trans, which I will call Trans #1) had about 100 to 110K miles on it and has not shifted very well since I bought the car in 2005 with 85k miles on it. I have long thought that the synchros were worn. At 93K miles, I replaced the OEM clutch with a Spec Stage 2+ clutch kit. This is a kevlar faced disk, with a sprung hub.
In 2011, during some "too" enthusiastic shifting, I jammed up 5th gear. I started looking into rebuilding my trans, but discovered it was a more complex of a task than I wanted to tackle at the time. Its also exceptionally difficult to find all the part numbers for all the little parts needed.
Instead, I bought another 93 LSD E153 (trans #2) from a board member that claimed it was rebuilt. That was april/may of 2011.
I replaced the Spec clutch with the clutch masters, after the spec clutch had about 20k miles on it. I burned up the spec getting the car home when I destroyed 5th gear on trans #1. I still have the disk and PP from this kit. Inspection of this disk a few minutes ago reveals no fretting of the female splines. However, the hub springs have partially worn through, and any more wear by the springs would result in this disk throwing its springs (from what I gather, a common problem with specs). This is why I went for an unsprung hub when I swapped trans.
I installed a Clutch Masters Stage 4 (power plus 1 Pressure plate) with 6 puck unsprung hub disk at the time, and new throwout bearing. I also replaced the rear main seal on the 3s, but this is just for info, and is not the cause of any issues.
I pulled trans #1, and I put the trans #2 in, and drove the car up until a couple of months ago, when I discovered the input shaft seal seemed to be leaking oil badly. Before the leak started, the trans shifted great. After I had the trans in for 2-3k miles, I found that just after shutting off the car when parking (engine/trans are warm), it would dump oil out of the bell housing area--out the drain hole at the bottom of the bell. I stuck my finger in the oil and verified it was purple in color (Royal Purple) gear oil, and smelled like gear oil. Definitely not engine oil (recall I replaced the rear main seal as well).
I decided to roll the proverbial sleeves up and dig into rebuilding my original trans (trans #1) That was a huge project but I am almost done with it. I had to create my own guide with drawings and part numbers just to figure out which parts I needed and did not need. I did pretty good, and only ordered one extra part I did not need, but was able to return (extra inner synchronizer ring). It turns out that the No 2 clutch (synchro) hub had "sprung" 2 of its 3 springs that provide pressure to the 3 shifting keys. No 2 hub is for shifting 3rd and 4th gear. This explains why my trans never shifted into 3rd or 4th very well as long as I have owned the car. The springs were laying in the bottom of the case. These springs are very small, and look like pen springs, but are much stiffer. During my rebuild, I corrected this design defect, and the new springs are very unlikely to pop out of place.
At any rate, I pulled trans #2 trans yesterday in preparation for swapping trans, and discovered the fretting of the clutch disk.
Possible causes:
Input shaft "walking around":
Input shaft bearing is damaged or worn out. They obviously did not replace the input shaft seal of trans #2 (because it was leaking). It is possible that the input shaft was damaged or just plain worn out and they did not replace it either.
Input shaft bent
If the input shaft were bent, it would cause low amplitude oscillation movement between the male and female splines, resulting in fretting.
Damaged disk:
I could have bent the clutch disk during installation of the trans. This is super easy to do, as putting the trans onto the engine, while the engine is in the car is super difficult. If I bent the disk (remember its solid hub), then the female hub splines will "wobble" relative to the input shaft.
Insufficient lubrication of the input shaft splines, and the female splines of the disk hub: According to Sachs, its really important to lube this area at assembly time, to prevent fretting. I found this in their literature and tech bulletins by googling around. I am pretty sure I used a light coating of never seize compound when I put the clutch on.
Misalignment of the trans:
All transmissions have a slight amount of misalignment with the engine in any car. The dowel pins align the trans to the engine, but the dowels can be placed slightly off (but usually within manufacturing tolerances). If they are outside of tolerance, the axis of the trans input shaft and the axis of the crankshaft could be off by enough to cause fretting of the clutch disk splines.
Trans #2 was one I bought off the board, with unknown history. Seller claims it was rebuilt less than 6k miles before I bought it. I will be able to verify that when I disassemble to replace input shaft seal and bearing. Possibly the dowel pins are slightly off for this trans, possibly at the extreme end of the manufacturing tolerance range, or just plain outside of tolerance?
Trans #1 obviously had no misalignment issues, as the OEM clutch (0 to 93k miles) and the spec clutch (93k to 110k miles) had no fretting issues.
Flywheel surface, disk pucks, and pressure plate friction surfaces all look good. However, I discovered massive fretting of the clutch hub splines.
The input shaft shows indications of fretting in that the surface is coated in the oxidized powder. However, the splines appear to be fine, and I felt them as well. Feeling them reveals no abnormal wear on the splines, i.e.: I drug my fingernails from the part of the spline not engaged to the clutch disk hub to the part engaged, and there were no steps from wear. This is not obvious from observation, or from the photos, which is why I felt them. The photo below makes it appear as if there may be a step due to fretting wear, but there is none.
From my Machinery's Handbook, 24th edition, page 2053:
My summary: low amplitude oscillation movement between the male and female splines, results in fretting
Once you get a little bit of oxidized steel (rust, but not caused by corrosion), the powder increases wear of the splines dramatically, as it acts like sand in the works.
I also found this paper online:
"Surface Durability of Coaxial Splines" Nembered page 18 (pdf page 26) starts the discussion on fretting:
http://fcp.mechse.illinois.edu/reports/FCP_Report177.pdf
Background: My 1993 E153 w/LSD (original trans, which I will call Trans #1) had about 100 to 110K miles on it and has not shifted very well since I bought the car in 2005 with 85k miles on it. I have long thought that the synchros were worn. At 93K miles, I replaced the OEM clutch with a Spec Stage 2+ clutch kit. This is a kevlar faced disk, with a sprung hub.
In 2011, during some "too" enthusiastic shifting, I jammed up 5th gear. I started looking into rebuilding my trans, but discovered it was a more complex of a task than I wanted to tackle at the time. Its also exceptionally difficult to find all the part numbers for all the little parts needed.
Instead, I bought another 93 LSD E153 (trans #2) from a board member that claimed it was rebuilt. That was april/may of 2011.
I replaced the Spec clutch with the clutch masters, after the spec clutch had about 20k miles on it. I burned up the spec getting the car home when I destroyed 5th gear on trans #1. I still have the disk and PP from this kit. Inspection of this disk a few minutes ago reveals no fretting of the female splines. However, the hub springs have partially worn through, and any more wear by the springs would result in this disk throwing its springs (from what I gather, a common problem with specs). This is why I went for an unsprung hub when I swapped trans.
I installed a Clutch Masters Stage 4 (power plus 1 Pressure plate) with 6 puck unsprung hub disk at the time, and new throwout bearing. I also replaced the rear main seal on the 3s, but this is just for info, and is not the cause of any issues.
I pulled trans #1, and I put the trans #2 in, and drove the car up until a couple of months ago, when I discovered the input shaft seal seemed to be leaking oil badly. Before the leak started, the trans shifted great. After I had the trans in for 2-3k miles, I found that just after shutting off the car when parking (engine/trans are warm), it would dump oil out of the bell housing area--out the drain hole at the bottom of the bell. I stuck my finger in the oil and verified it was purple in color (Royal Purple) gear oil, and smelled like gear oil. Definitely not engine oil (recall I replaced the rear main seal as well).
I decided to roll the proverbial sleeves up and dig into rebuilding my original trans (trans #1) That was a huge project but I am almost done with it. I had to create my own guide with drawings and part numbers just to figure out which parts I needed and did not need. I did pretty good, and only ordered one extra part I did not need, but was able to return (extra inner synchronizer ring). It turns out that the No 2 clutch (synchro) hub had "sprung" 2 of its 3 springs that provide pressure to the 3 shifting keys. No 2 hub is for shifting 3rd and 4th gear. This explains why my trans never shifted into 3rd or 4th very well as long as I have owned the car. The springs were laying in the bottom of the case. These springs are very small, and look like pen springs, but are much stiffer. During my rebuild, I corrected this design defect, and the new springs are very unlikely to pop out of place.
At any rate, I pulled trans #2 trans yesterday in preparation for swapping trans, and discovered the fretting of the clutch disk.
Possible causes:
Input shaft "walking around":
Input shaft bearing is damaged or worn out. They obviously did not replace the input shaft seal of trans #2 (because it was leaking). It is possible that the input shaft was damaged or just plain worn out and they did not replace it either.
Input shaft bent
If the input shaft were bent, it would cause low amplitude oscillation movement between the male and female splines, resulting in fretting.
Damaged disk:
I could have bent the clutch disk during installation of the trans. This is super easy to do, as putting the trans onto the engine, while the engine is in the car is super difficult. If I bent the disk (remember its solid hub), then the female hub splines will "wobble" relative to the input shaft.
Insufficient lubrication of the input shaft splines, and the female splines of the disk hub: According to Sachs, its really important to lube this area at assembly time, to prevent fretting. I found this in their literature and tech bulletins by googling around. I am pretty sure I used a light coating of never seize compound when I put the clutch on.
Misalignment of the trans:
All transmissions have a slight amount of misalignment with the engine in any car. The dowel pins align the trans to the engine, but the dowels can be placed slightly off (but usually within manufacturing tolerances). If they are outside of tolerance, the axis of the trans input shaft and the axis of the crankshaft could be off by enough to cause fretting of the clutch disk splines.
Trans #2 was one I bought off the board, with unknown history. Seller claims it was rebuilt less than 6k miles before I bought it. I will be able to verify that when I disassemble to replace input shaft seal and bearing. Possibly the dowel pins are slightly off for this trans, possibly at the extreme end of the manufacturing tolerance range, or just plain outside of tolerance?
Trans #1 obviously had no misalignment issues, as the OEM clutch (0 to 93k miles) and the spec clutch (93k to 110k miles) had no fretting issues.
