by Ben Prince

MG owners, T-car owners in particular, have been plagued by dripping oil. One serious leak frequently emanates from the rear main bearing. Many remedies have been tried to eliminate this problem, but the results have not always been satisfactory. The best solution was introduced a few years ago by Andy Bradley. His method consists of modifying the T-series engine so that a Chevrolet 350 rear bearing seal can be installed. His thinking was that "If Chevy engines don't have oil leaks from the rear main, then MGs should not leak either when fitted with the same seal."

Many who have Chevy seals in their MGs have experienced good results, but a few of us have had difficulties. A Chevy seal was installed in my TF about three years ago (8000 miles ago). The results were initially quite good, with only a few spots of oil on the garage floor. With time, however, the leaking has become progressively worse, a puddle often appearing below the rear main.

Before I embark on an engine tear-down to replace the seal, I decided to gather as much pertinent information and data as possible on the seal installation. The following is a compilation drawn from various sources who have applicable backgrounds and experience.

Andy Bradley, Mount Vernon WA
As mentioned above, Andy originated the Chevy oil seal installation. The crank is surface welded to build it up, and then machined to the proper diameter to use the seal that normally is found in the Chevy 350 cid engine. Andy specifies 2.437" for this crank dimension. A hole is also machined in the housing for the seal. The housing forms a tongue that slips into a notch in the seal. Andy recommends 2.775" -2.785" for this housing diameter, and specifies .187" for the tongue thickness. The crank and housing dimensions on my engine were machined in accordance with Andy's specifications.

The seal is a two piece assembly that is used on 1959 to 1985 Chevy 350 small block engines. Andy recommends the Fel-Pro seal, part number BS40013. Andy has installed a number of these Chevy seals in T-series cars.

Phil Marino, Riverside, CA
Phil is a superb machinist. Having installed a couple of seals, he feels that Andy's specified housing hole for the seal is .010" too large. If a hole in the housing has already been cut according to Andy's specification, Phil installs a .005" shim in the groove in each side of the seal to reduce the hole diameter. The shims are 3/16" wide.

Phil has made a fixture so the hole can be cut without sending him the block. His fixture is set up for 2.765" - 2.775". He also mentioned that the cap tube should be removed and no paper gasket should be used (as Andy has already stated).

Phil said that the four bolts on the flywheel might be too long, possibly tearing the seal. He had to shorten them by 1/8" on his engine. He removed the dowel pins, claiming that without them he could jiggle the seal back towards the crank before tightening the screws.

Phil replaced the three hex screws with allen screws. He claims it is tough to get a wrench onto the hex heads. When installing the seal, Phil says it should be rotated to about the two o'clock position, so to not line up with the joints of the cap.

Phil also recommends that a factory GM seal be used -- not Fel-Pro.

Federal Mogal, Southfield, MI
I talked to Federal Mogal. It is my understanding that they have bought Fel-Pro. They said that the crank diameter should be 2.430" -- 2.432". This is less than Andy's 2.437" figure.

Victor Reinz Co, Lisle, IL
This company makes seals for NAPA. They specified 2.431" for the crank, and 2.777" for the housing hole diameter. No tolerances or ranges were given.

Mike Goodman's Sports Car Service Ltd, Van Nuys, CA
Mike has probably performed more of the Chevy seal installations in T-series MGs than anyone. From Fel-Pro literature, Mike stated that the crank should be 2.430" - 2.432". He said that Fel-Pro also specifies the housing hole at 2.777" - 2.779".

Mike claims that the top-of-the-line Fel-Pro seal should be used; the part number is 2900.

Mike states that the Chevy seal is designed to run at 40 - 45 psi. When we have higher pressures in our MGs, everything must be right-on or we may have leaks.

Mike reiterated that the seal splits should not be in alignment with the cap splits. In some situations where the seal has been aligned with the cap, he claims that he has stopped oil leaking without pulling the engine. He removed the pan and merely rotated the seal to not line up the cap joint and seal joint. He also said that it should be possible to install shims without pulling the engine.

Mike claims that it is very important that the housing hole for the seal be in perfect alignment and centered with respect to the crank

Mike states that both seal halves, particularly the ends should be siliconed.

Kirkland Auto Supply, Kirkland, WA
A second Mike was the contact. He was highly recommended by the Chevrolet dealership in Redmond. He took measurements directly from a Chevy engine. The housing hole diameter was 2.775", the crank was 2.432", and the housing tongue thickness was .189". He said the housing thickness is a bit variable because it seemed to have a taper. He also said he doesn't think that the crank diameter is critical because his literature indicates that .010" can be ground off the crank and the seal will still work

This second Mike felt that my seal might be in backwards. He said the undercut area must go forward -- it must collapse into the oil.

He felt the pan should be pulled, the crank rotated, and one should look for pits in the weld. He also said that the built-up welded area might not be in alignment with the seal lip. If not, he mentioned that off-set seals are available.

Contrary to Phil and Mike Goodman, the second Mike also claims that it is all right for the seal and cap parting lines to be in alignment, as long as silicone is adequately used. He stated that the seal might get damaged if installed with parting splits out-of-alignment.

He said it is important to apply silicone between the cap and the block on Chevys -- at the beveled edge and at the seal. He did not know for MGs.

North Bend Auto Parts, North Bend, WA
The contacts were Wes and another Mike. This Mike also took measurements from a Chevy engine. The housing hole diameter was 2.777", the crank was 2.430" and the housing thickness was .190". This third Mike said that the housing thickness varies plus or minus .004". The third Mike also provided another measurement -- the housing diameter was 3.067" from shoulder-to-shoulder. The shoulders are at the base of the tongue on each side.

Like the second Mike, the third Mike said that he always lines up the parting lines since it is too easy to damage the seal otherwise. On Chevy engines, he also silicones the cap to the block and the seal, particularly at the ends of the seal.

This Mike also said he would leave the dowels in place. They are there for a purpose -- to properly locate.

RAM, Spokane, WA
RAM is a "mass-production" overhaul shop. The contact was Ken, who measured another Chevy engine. The housing hole was 2.765", the crank was 2.431", the housing thickness was .195" and the shoulder diameter was 3.080". Ken indicated that the housing thickness varied between .190" and .200".

Fel-Pro states that crank should be 2.430" -- 2.432" where the seal rides. All three Chevy rebuilders obtained measurements in this range. My crank was supposedly machined to Andy's number, 2.437", a figure that apparently is .005" - .007 too large. I don't know if this has any practical significance, but perhaps too thick a crank has burned up the lip on my seal. I plan on having the crank diameter reduced as needed to get it into the Fel-Pro specified range.

The desired size for the housing hole is a little less clear. Fel-Pro specifies 2.777" - 2.779". The rebuild shops measured 2.765" to 2.777". Andy's upper limit of 2.875" seems high, but once again I don' know if there is a practical problem. I am concerned, however, that too large a housing hole might cause the seal to distort or even wobble. On the other hand, Phil's fixture might make the housing hole a bit small, although one Chevy block measured the same as Phil's lower limit. I'll have my engine shimmed or bored as necessary to try to get it to 2.777".

The shoulder question is a bit of a puzzle. I don't know if the MG housing is machined with shoulders, but if not, could that be my problem? How critical it is to even have shoulders is not clear, but it seems that Chevrolet has them for a reason, perhaps to properly support the seal. Ken (from RAM) conjectured that they help keep the seal from turning. I have only obtained two measurements -- 3.067" and 3.080" (from the third Mike and from Ken). Without any further data, I think one should shoot for 3.073", the midpoint shoulder diameter. If there is no shoulder on the MG housing, then this diameter is of course academic. And without a shoulder, it seems even more important to select a high quality seal and to properly glue the seal onto the housing tongue.

On the question of lining up parting lines, I don't think it matters as long as silicone is properly applied. I would tend to not align the seal and the cap, provided the seal can be installed without damaging it.

Mike Goodman specifies Fel-Pro part number 2900 for the seal. Andy Bradley recommends Fel-Pro part number BS40013. Both are correct; the 2900 is for high performance engines, and the BS40013 is the regular version. I am inclined to use the high performance model, or one from Chevrolet (as Phil suggests).

Further Comment
I recommend that any interested readers drop by a Chevy rebuild shop and examine the block, cap and seal installation. A first-hand look will help provide excellent insight. If arrangements can be made, it might even make sense to have a tech session at a Chevrolet overhaul facility.

I would like to thank Oscar Stankov for urging me to prepare this piece, and his help in writing the introduction and editing.


The above article was a compilation of comments and suggestions from a number of individuals with experience and expertise with the Chevrolet oil seal at the rear-main bearing of a T-car engine. Particular attention was given to nailing down the dimensions for the crank (2.431") and the housing (2.778") where the seal is placed.

In response to the article I received a few email comments and questions. One of the best inputs was from David Edgar. He mentions that on his engine the weld did not extend all the way back to the flange. He points out that his crank was welded and machined with the journal separated by one sixteenth of an inch from the flange. This resulted in the rear dust lip of the seal not riding on the journal. As a solution, he recommends using an offset seal, the lips riding 3/32" farther forward on the journal. Victor Reinz makes an offset seal, part number JV1627. Napa also handles the seal, labeled Napa 1627.

Another comment cautioned that the rear main cap-drain tube should be removed. I wish I could add more to this suggestion, but it is not clear to me what purpose this tube serves. It apparently has to do with the original Archimedes screw "seal." Also, both Andy Bradley and Phil Marino mentioned that his tube should be discarded.

There is a groove in the outside of the seal that fits into the housing (tongue). It was pointed out that the thickness of the tongue may be insufficient, and that shimming may be necessary.

Lastly, when my engine was opened up it was discovered that the seal interfered with the flange. Contrary to David Edgar's situation, the weld extended all the way back, covering the ends of the bolt holes in the flange. When the flywheel was installed, the bolts pushed the weld (that was in the bolt holes) toward the seal, in turn protruding and touching the seal. The solution was rather simple. Shorter bolts were installed. The crank was also returned to the machinist to radius and polish the flange area. Just to play it safe, an offset seal (Victor Reinz JV1627) was installed.

The snags mentioned here and in my previous article should not detract from Andy Bradley's outstanding work. Andy was the first to come up with using the Chevy seal, and it appears to unequivocally be the best solution yet. Like all pioneering efforts, some growing pains are part of the process.