Front differential leak, should I just redo the RTV?
- Thread starter LJcooljay
- Start date
Most gasket RTVs state in the directions to go finger-tight, wait a period of time (usually an hour) and then torque the screws. The reason for this is to get the RTV to spread out and cure a bit, so that when you do the final torque, the gasket has some firmness to it so that you can develop some torque before you squeeze out too much RTV. The mating surfaces have small ridges and valleys, and if you immediately torque the bolts, you can have small pathways for leaks.
Try just barely tightening the bolts a bit more, but be careful not to over-torque (stripping the threads.) This isn't always successful, but it's worth a try. I just did this earlier this year on a Subaru rear diff cover. Worked like a charm, which is good because the rear diff cover on those is structural and part of the mounting system, so it's a half-hour's work or more to remove it. If that doesn't work, when you re-do it, be sure to read the directions, and if it calls for a wait period before final torquing the screws, heed those directions.
Actual squish-out:
View attachment 520105
Well under 90%.
I also used the recommended two-step process. Finger tightened the bolts until the RTV squeezed out a little, evenly all around, and let it vulcanize for an hour. Then torqued the bolts, squeezing out more of course, and waited 24 hours before adding lube.
Looks like 23.2%
give or take .08%.
But I’m just eye balling it! Ha
I agree with Brianj5600 on this. I've built dozens and dozens of engines, transfer cases, transmissions, and differentials over the years, and I never use anything close to that size bead. On the production line for Harley and Polaris (both companies I've worked for as a design engineer), the tech setting up the RTV gun would be in trouble using a bead that size. It's wasteful, and the part of the excess that squeezes into the case can plug filters, if present, when it breaks off.
Ok. The sealing surface width on a Dana 44 is about .75". As you can see from the picture below, the bead width looks to be about 1/4":
The area of a circle 1/4" in diameter is pi x R x R, or 3.14 x .125" x .125", which is 0.049 square inches. So, if 90% squished out, that would mean that the thickness of the cured RTV remaining on the .75"-wide sealed surface would be:
.049 square inches x 10% ÷ .75 inches, or .0065"
In my experience, that thickness seems to be realistic. I concur with Brian5600's estimate of 90%.
Additionally, the RTV on the outside of the holes serves no purpose. Again, on production lines, that would not be there to save cost (every penny counts in vehicle cost to the bean-counters I despised.)
I disagree. There's a similar amount on the inside that squeezes out, too. Together with what's pictured on the outside, that could very well total about 90%.
Absolutely! As I mentioned above, following the directions for the sealant you are using is critical. On my sealant shelf, I have many types of sealants, and some require an extended waiting period and some don't. Silicone RTVs are the ones that most commonly require it. I use a Three Bond product (can't remember the number) for aluminum case halves (engines, transmissions, etc), and it doesn't require a long wait time...
I am actually in the market for a front Dana 60 and a RJ60 would meet the need if you have one sitting around.
Brianj5600
TJ Expert
I use very little RTV and don't use a bead. I smear a thin layer of what ever RTV isn't dried in the tube on the mating surface, bolt down the cover and let it sit overnight. If your surfaces are flat you almost can't tell there is any RTV on the mating surfaces.
If it makes a mess, try less.
If it makes a mess, try less.