Forum Home Forum Home > CJ-2A Discussion Area > Tech Questions and Answers
  New Posts New Posts RSS Feed - Rear main seal - which one?
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Rear main seal - which one?

 Post Reply Post Reply
Author
Message
JeepSaffer View Drop Down
Member
Member
Avatar

Joined: 26 Sep. 2014
Location: South Africa
Status: Offline
Points: 963
Post Options Post Options   Thanks (0) Thanks(0)   Quote JeepSaffer Quote  Post ReplyReply Direct Link To This Post Topic: Rear main seal - which one?
    Posted: 11 Jan. 2019 at 3:39am
I have three different rear main seals available to use from my parts stash, and want to know if there is a clear winner that I should be using on my new engine build.



On the left is a seal that came in Walcks "engine gaskets and seals kit". Looks like neoprene rubber.

In the middle is a seal that is apparently NOS, that came from QTM a few years back. It appears to have some kind of fabric or material on the seal area, which would actually form the wearing part of the seal.

On the right is a rope type, which came from Peter deBella. 

Here is a close up of the NOS seal. Some part numbers and some installation arrows are visible.



That's what I have available. Is there a clear winner as to which one I should be using? My seal area on the crank is not brand new, but not terrible either.

Thanks,

Mike

 

1948 CJ2A #204853 in South Africa
Back to Top
swmoboy View Drop Down
Member
Member
Avatar

Joined: 17 Nov. 2014
Location: Springfield, MO
Status: Offline
Points: 133
Post Options Post Options   Thanks (0) Thanks(0)   Quote swmoboy Quote  Post ReplyReply Direct Link To This Post Posted: 11 Jan. 2019 at 8:18am
Following cause I'm fairly certain I need a new rear main seal.  Every other time or third time I park the ole girl she leaves me a puddle on the floor and sometimes it's a significant amount.  
1947 CJ2a #155307
Back to Top
JeepSaffer View Drop Down
Member
Member
Avatar

Joined: 26 Sep. 2014
Location: South Africa
Status: Offline
Points: 963
Post Options Post Options   Thanks (0) Thanks(0)   Quote JeepSaffer Quote  Post ReplyReply Direct Link To This Post Posted: 11 Jan. 2019 at 9:23am
I found this thread from some back, but it has some pretty useful information in it:

In summary, Sean (who else? LOL) indicates how to tell an original Victor Gasket seal from a cheap imitation:

  • Old-style Victor seals are marked  : "VG  49650  800093 Install -->"
  • New-style Victor seals are marked: "xx  VG  49650  800093 Install -->"
I don't have the numbers before the VG, so it seems I do have an original older style. I guess that is  good news. 

I also found this, which makes a lot of sense:




In summary:
  1. Some of the early L134 blocks provided more space for a seal than the later blocks, and were designed for a rope seal. Meaning that forcing a Victor type seal designed for a later block into an early block designed for a rope seal, can mean that there is not enough clearance for the crank to turn in the seal. This can lead to failure of the seal.
  2. Measure the seal area dia of your crank. If it is within the limits shown, you can use the later type seal. If it is larger than the limits shown, you need to go with a rope type seal.

I will wait for anybody else to chime in, but I think my way forward is exactly as above: measure my crank seal area dia and if it is within the limits above I will go ahead and use my NOS Victor seal. If my crank is larger I will use my rope type seal. My block is serial number 153955, so is not "early" for L-head, but certainly early for F Head.
I have also picked up that when installing your seal you should not feel more than a slight drag due to the seal on the crank, when the caps bolts are torqued down. If there is anything more than this, you are trying to force too tight a seal into too small a gap and you will likely have problems with failure in the not too distant future.

So I will go with the NOS Victor seal if my measurements on crank dia check out, unless someone thinks this is a terrible idea! Key being also to rotate the crank by hand after installation and see how much drag there is.

Any other good ideas?


1948 CJ2A #204853 in South Africa
Back to Top
Spinnas View Drop Down
Member
Member


Joined: 24 Apr. 2018
Location: AZ
Status: Offline
Points: 190
Post Options Post Options   Thanks (0) Thanks(0)   Quote Spinnas Quote  Post ReplyReply Direct Link To This Post Posted: 11 Jan. 2019 at 11:02am
Personally I like the rope seal, it's more forgiving.
Back to Top
Mark W. View Drop Down
Member
Member
Avatar

Joined: 09 Nov. 2014
Location: Silverton, OR
Status: Offline
Points: 4509
Post Options Post Options   Thanks (0) Thanks(0)   Quote Mark W. Quote  Post ReplyReply Direct Link To This Post Posted: 11 Jan. 2019 at 11:57am
A key to getting this to seal up is the condition of the seal surface along with as the above bulletin states the diameter of the narrow sealing surface. My crank was a little messed up in this area and after the original machine shop determined the mains and rods were within spec to not grind the crank. I noticed they also did not mess with the seal surface. I ended up taking my crank to a true old school (been around since before WWII) crank grinder. There I was told that yes my crank was usable as is but it was on the last 20-25% of its life at that dimension. SO I had the crank ground to .010" under and then he cleaned up the seal surface to the min 2.302 to get as much of the rough surface out as possible. I am running a Fel-Pro Rope seal. Sometime this summer I will find out if its working.
Chug A Lug
1948 2A Body Customized
1949 3A W/S
1957 CJ5 Frame Modified
Late 50's 134L 9.25"clutch T90A D18 (1.25") D44/30 flanged E-Locker D25 5.38 Since 1962
Back to Top
athawk11 View Drop Down
Member
Member
Avatar

Joined: 17 Jan. 2012
Location: Arvada,Colorado
Status: Offline
Points: 3285
Post Options Post Options   Thanks (0) Thanks(0)   Quote athawk11 Quote  Post ReplyReply Direct Link To This Post Posted: 11 Jan. 2019 at 3:35pm
JeepSaffer,
I think you have a perfectly reasonable plan of action.  Based mostly on what I've read on the forums, and in Bob Westerman's book, my personal dividing line is...

If the engine was built in the period prior to the development of the M38, I will stick with the rope seal.
If after the M38, I will use the NOS neoprene seal... if the seal surface measures out good and the surface is not all beat up.  Of course this assumes the crank is original to the engine, and hasn't been 'improved'.

I am not lucky enough to have a later engine or an improved crank surface.  I use the rope seals.  And unfortunately, mine all leak a little.  Particularly after pushing the engines to their limits.


1- 1946 CJ2A   
2- 1949 CJ3A
Back to Top
JeepSaffer View Drop Down
Member
Member
Avatar

Joined: 26 Sep. 2014
Location: South Africa
Status: Offline
Points: 963
Post Options Post Options   Thanks (0) Thanks(0)   Quote JeepSaffer Quote  Post ReplyReply Direct Link To This Post Posted: 12 Jan. 2019 at 8:51am
Thanks Tim, you raise interesting points...

My block is from a CJ-2A engine, so was definitely built prior to development of the M38. According to your mantra that would mean "use the rope seal".

However the crank came loose in a pile of parts and was not installed in the engine (I bought a bare block), so it is possible the crank is later than M38. There is no way of knowing.

My seal surface is pretty good, but could do with a little polishing. The surface looks more "stained" than scored or grooved or chipped.

I did mic out the seal surface. The AERA Tech Bulletin says 2.302-2.312 = OK for neoprene seal. Mike mics out at 2.307. Can't get more perfect than that. 

My current understanding is that you have two options, each with their own advantages and possible downfalls:
  1. Use a rope seal. More forgiving, less chance of heating up and failing. But less chance of a "tight" seal and hence expect a few drops of leakage from your rear main.
  2. Use a conventional seal. Less forgiving, only fit in certain engines/cranks, seemingly some bad reproductions out there. Too tight will lead to heat and early seal failure. BUT, if you get it right, you get an improved seal over a rope seal and hence better chance of less leakage.

Based on the only definitive measure out there that I have found, which is the AERO Bulletin, and the fact that I seem to have a genuine NOS seal, I plan to test fit it and check the drag on the crank. If the drag is minimal I will go for it. If the drag is concerning, I will swap out for the rope seal.

I've got a few more days before I actually try this out, so chime in if you think any differently. 
1948 CJ2A #204853 in South Africa
Back to Top
JeepSaffer View Drop Down
Member
Member
Avatar

Joined: 26 Sep. 2014
Location: South Africa
Status: Offline
Points: 963
Post Options Post Options   Thanks (0) Thanks(0)   Quote JeepSaffer Quote  Post ReplyReply Direct Link To This Post Posted: 28 Jan. 2019 at 3:21am
So I got the crank in over the weekend, fitted and torqued down all the caps. I wanted to get a proper feeling for how much drag the Victor seal added to the crank, so I installed it first without the seal and gauged the effort to rotate the crank. 

My plan was to put my torque wrench on and find the smallest torque where the handle clicked before the crank turned. Good in theory, but not in practice..... the crank was turning before the torque wrench clicked, even on its lowest setting. So I moved to the "how many free revolutions can I get" test. I could get 3 full revolutions of the crank if I grabbed a counterweight and spun it as hard as I could. I also polished up the rear main seal area using first 800 grit, then 1200 grit and finally 1500 grit sandpaper.

I then removed the caps and crank again, and installed the NOS Victor rear main seal, and gave it a generous wipe of engine oil. Torqued everything up and again I couldn't get a proper torque reading because it was too low to read on my wrench. I did the "spin test" and this time I could get about 1.5 free rotations of the crank with a good hard pull on one of the counterweights. So, drag did increase which is not unexpected, but it does not seem unreasonably tight. Everything turns freely by hand, just a question of how much torque does it require.

Then I had an idea. To go back to first principles and measure the torque using a weight on the end of a bar on the crank nut. I also have an old style torque wrench with the needle that reads off a scale on the wrench as you pull on it. Most usefully, it has a rubber grip on the end with indents where your fingers are supposed to go. This meant I could hang a weight off the end of the bar in exactly the same place every time, without the weight slipping.

I got a plastic milk carton, filled it with water, made a wire hook for it to hang off, and hung it off the end of the bar attached to the crank nut with a socket. I then reduced the weight in the milk carton by pouring out water a little at a time until I could feel I was getting light enough that I was getting close. Finally I got to the point that the milk carton was JUST light enough that the crank would not turn. I weighed the milk jug and measured its distance from the crank nut. I also weighed the torque bar, and located its center of gravity as the point where it would balance. This would also have contributed to the torque on the crank as I was locating it in the 3 o'clock position during the test. It was then just a matter of doing the maths.

End result: 1.98 ft.lb with 3 bearing caps torqued to spec and NOS Victor type rear main seal installed.

This explains why it was difficult to measure with the click type wrench! I suppose I could have gone ahead and removed the rear main seal and redone the test to get an accurate reading for the crank alone, but I didn't feel like undoing work at that stage. 

The whole point of this exercise was to give myself some comfort that I am using a seal that is not too tight. A lot of failures I have read about seem to be because the rear main seal (rubber type) appears to be too tight. People report after the failure that they remember the crank was hard to turn by hand. But there is no measure of how freely the crank should turn. One persons "slight drag" is anothers "tighter than it should be".

I realise this is a single data point and may not apply to all rebuilds, but it does give some sort of benchmark if others are wondering. 
  1. Freshly reground crank that is properly installed with correct bearing clearance and I got around 3 full turns on the crank spin test. Torque reading not measured, but certainly less than 2 ft.lb, and much more likely to be around 1 ft.lb.
  2. Above crank with NOS Victor type rear main seal and I required around 2 ft.lb of torque to turn the crank. This roughly equated to 1.5 turns of the crank on the "spin test"
It would be interesting to build up a library of corresponding readings on other rebuilds. It seems there are specifications on just about every aspect of an engine rebuild, but not in this important aspect! With enough data points a guide could be established that cranks should turn with no more than X ft.lb torque for certain conditions. Significantly more than X and you should be looking at your bearings for binding, or at your seal as too tight. Of course readings would have to be differentiated between Victor type seals and rope type. This might save a lot of people from having to tear back into their engines immediately after a fresh rebuild...

For interest sake, I got as far as installing the first rod and piston, with new rings. All seemed to go well. Of course more drag again, for obvious reasons. I measured the torque on the crank again because it was quick and easy to do. 4.3 ft.lb to turn the crank with the #1 piston installed = roughly 2.3 ft.lb additional torque due to the addition of 1 piston to the crank.

I will continue testing as I install the rest of the pistons and see if it is a linear relationship. 


Edited by JeepSaffer - 28 Jan. 2019 at 8:53am
1948 CJ2A #204853 in South Africa
Back to Top
eestes1 View Drop Down
Member
Member
Avatar
Sponsor Member

Joined: 12 Feb. 2011
Location: Mineral, VA
Status: Offline
Points: 1041
Post Options Post Options   Thanks (0) Thanks(0)   Quote eestes1 Quote  Post ReplyReply Direct Link To This Post Posted: 28 Jan. 2019 at 6:18am
Wow ! That was great to read! Very creative. Thanks for taking the time to write it up.
Rick Estes
Back to Top
Gil View Drop Down
Member
Member
Avatar

Joined: 29 July 2016
Location: N.B.Canada.
Status: Offline
Points: 760
Post Options Post Options   Thanks (0) Thanks(0)   Quote Gil Quote  Post ReplyReply Direct Link To This Post Posted: 28 Jan. 2019 at 8:05am
Wow Mike,thanks for taking the time to post your finds.Nice to have something to compare if ever I rebuild an other motor.

Giles
1946 cj2a 59108
1998 Jeep Cherokee 2 doors
2016 Jeep Cherokee TrailHawk
Back to Top
willyt View Drop Down
Member
Member
Avatar

Joined: 31 July 2018
Location: Seneca,SC
Status: Offline
Points: 101
Post Options Post Options   Thanks (0) Thanks(0)   Quote willyt Quote  Post ReplyReply Direct Link To This Post Posted: 28 Jan. 2019 at 10:12am
Seems there are a lot of leaky seals out there in jeep land and one of them is mine. A '52 3A. I have documentation that this engine has been in the jeep since '55. Don't know if it has or how many times it may have been built.
I want to replace the rear main bearing seal without removing the engine. With the bearing cap removed can you get an accurate measurement of the rear crankcase journal?
 
1952 CJ3A (Lil'Green)
early M38A1(Ole Green)
1970 Jeepster Commando
Back to Top
JeepSaffer View Drop Down
Member
Member
Avatar

Joined: 26 Sep. 2014
Location: South Africa
Status: Offline
Points: 963
Post Options Post Options   Thanks (0) Thanks(0)   Quote JeepSaffer Quote  Post ReplyReply Direct Link To This Post Posted: 05 Feb. 2019 at 7:07am
Originally posted by JeepSaffer JeepSaffer wrote:


I will continue testing as I install the rest of the pistons and see if it is a linear relationship. 

I did get the remaining three pistons in and did some torque measurements. But what I realised is that there is a different amount of torque required to turn the crank depending on where a piston is in its cycle

At first I couldn't figure it out. But then I realised that the speed of piston travel is not constant as the crank turns. Pistons travel a greater distance for a given crank rotation in the middle of their stroke (hence more torque required to drive them) compared to at the ends of their stroke where they change direction, and have to travel very little distance for a given rotation.

Bottom line: I didn't record where the piston was in its travel when recording torque on piston #1, so torque results from pistons #2 to #4 were going to be meaningless as a comparison. I didn't end up measuring them. 

But, my reading of 1.98 ft/lb to turn the crank alone with no pistons attached remains valid. 
1948 CJ2A #204853 in South Africa
Back to Top
 Post Reply Post Reply
  Share Topic   

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.01
Copyright ©2001-2018 Web Wiz Ltd.