What factors might account for slowing of ATF leak rate? - PeachParts Mercedes-Benz Forum

shertex · #1 07-16-2016, 08:58 AM

As I've noted in other posts, I have a 722.6 that leaks from the torque converter area. Tranny tech suspects front pump seal from looking at it up on the lift but can't be sure.

I thought I had the leak rate down to a science. Some variation, but roughly loses about 11 oz. every 2500 miles. So I simply check every once in a while and top off. I need to check at 25C because I have no confidence of being able to get an accurate 80C reading.

This morning I checked after 1300 miles and it doesn't look like there was any fluid loss at all. While I'm sure there are some inaccuracies and methodological flaws in my approach, I still strongly suspect that the leak rate has gone down. And my mechanic has told me in the past that sometimes leaks simply stop.

Why? What factors come into play?

leathermang · #2 07-16-2016, 09:20 AM

Without the introduction of some type of seal expander like Trans-x .... the rule is that leaks do not correct themselves.
Did you change the brand of ATF that you use... which might have a chemical like that in it ?

shertex · #3 07-16-2016, 09:28 AM

I had Redline D4 in there from when I bought the car three years ago. But I recently switched to Shell ATF 134 (drained and filled 3 qts.)....but I don't think there's anything in the Shell product that would cause seals to swell.

leathermang · #4 07-16-2016, 09:40 AM

Quote:

Originally Posted by shertex

I had Redline D4 in there from when I bought the car three years ago. But I recently switched to Shell ATF 134 (drained and filled 3 qts.)....but I don't think there's anything in the Shell product that would cause seals to swell.

Seems to me that you just reported the evidence in post one that it does have something to cause a slowdown or stop in leaking....

shertex · #5 07-16-2016, 09:43 AM

Quote:

Originally Posted by leathermang

Seems to me that you just reported the evidence in post one that it does have something to cause a slowdown or stop in leaking....

It would seem so. But, if that's the case, it's surprising that a fluid that meets the MB spec would have something in there that would slow/stop leaks.

Mike D · #6 07-16-2016, 10:01 AM

Redline is "full synthetic" whereas the Shell is "semi-synthetic" (whatever that is supposed to mean). I don't see why this would make a difference but as my old grand pappy always said, "It's hard to argue with success"

leathermang · #7 07-16-2016, 10:13 AM

Your leak was incredibly small in the first place....
it would not take much to make the leak less.... a few molecules here and there... bingo.

shertex · #8 07-16-2016, 10:21 AM

I will keep an eye on it. 1300 miles wasn't that long of an interval so I imagine different variables could skew things one way or another. But at least I won't have to worry about spending big $$$ on a reseal.

Zulfiqar · #9 07-16-2016, 11:31 AM

shell 134 is quite a nice fluid - it may have better astringent properties than regular dexron type ATF.

btw - checking at the 25C mark is pretty hard to get right.

snapped_bolt · #10 07-16-2016, 12:13 PM

How about an O-ring that has shrunken-sometimes it leaks, sometimes it changes position and blocks, or just slows the leak.
In this case the "self-repairing" feature is not necessarily permanent.

snapped_bolt

shertex · #11 07-16-2016, 05:20 PM

Quote:

Originally Posted by Zulfiqar

btw - checking at the 25C mark is pretty hard to get right.

I realize the cold reading is inherently less accurate. But, I'm at least confident that the temperature is right. I'm much less confident in being able to get an 80C reading.