Talk Morgan Forums Morgan Sports Cars Morgan Three Wheeler & Super 3 A reason why M3Ws leak oil…

I am finally posting about my oil leak experience after ‘upgrading’ to the Centa for my 2013 M3W. I’ve delayed relating this story in an effort for Morgan to do the right thing and make a Centa journal with a proper oil sealing surface in-line with crank seal manufactures specifications; a machine-ground finish.

After ~180 miles on my new Centa a significant amount of oil began leaking from the lower bell-housing hole; more than the typical oil weeping many M3W owners tolerate. The rear main crank seal was determined to have failed with the surface of the Centa journal blamed. This is a pic of the Centa journal clearly showing jaw marks from the manufacturing metal lathe and a machine turned finish…



Morgan’s initial reaction was it’s my mechanic’s fault but then added that all their current Cetna parts look like this and their assembly crew just scotch-pads them before installation (a practice specifically discouraged by oil seal companies). Morgan continued to say that there is no other way of making these parts without the lathe jaw marks to which an instructional YouTube video of Soft Jaw Metal Lathe Technique was forwarded. This slowly progressed to Morgan saying they were going to meet with their Centa suppliers and admitting that the Centa journal surface was not previously specified as oil-sealing to its maker. During this time Morgan was reminded that the Centa journal surface should be a ‘ground finish’ for oil crank-shaft sealing success and longevity, not their machine turned finish.

After a couple months an unannounced Centa part arrived in October; no lathe jaw marks but still the machine turned finish. After questioning this ‘better but still not proper surface’ I was informed that new Centa parts will come in December, no January, no February 2019.

To my disappointment the last Centa journal received at the end of February 2019 still had a machine turned surface and my last attempt to communicate with Morgan’s leadership resulted in an email bluntly stating that while Morgan had talked about a ground finish they never formally promised a ground finish on the Centa part; because Morgan owners have not complained about leaking oil so this is not an issue (only for me).

I am now having my Centa journal modified to a ground finish before final reassembly. This will hone ~2/1000” off the OD in the process about 20% of the material that could be removed and still be within the seal manufacturer’s OD specs.

The first Centa part with lathe jaw marks was received in February 2018, the first replacement Centa part, with no jaw marks and a turned finish, was received October 2018. My expectation is that Morgan was using these questionable parts when making new M3Ws and in their Centa ‘upgrade’ kits during this timeframe.

I’ll add that for my experience Morgan is crediting $800 US; compensation that will almost cover my transportation cost to and from the repair shop. Specifically, Morgan adds this credit to the dealer’s account who then has to front the money to the customer (me) until they owe Morgan something in the future.

I guess the moral to this story is if your M3W leaks oil SAY SOMETHING and maybe, just maybe, Morgan will eventually do the right thing…

scotch pads.... amazing

Those marks were made with a hard jaw not a soft jaw and it would be a better man than me who could polish them out with a scotch bright pad.

Using Soft Jaws is not an industry seceret for metal lathe work but is was an epiphany to Morgan's engineers. This is the video that was sent to 'inform' them that there is a way to avoid the use of Hard Jaws.

Soft Jaw Metal Lathe Technique video

I copied these over from JV's Centa upgrade post back in February of 2016.

This first shot shows the marks if you look hard.



This is a blow up of the hub, clearly has jaw marks. So this gives you some time frame on how long they've been using them with the jaw marks. It would appear that it's just a crap shoot if your seal happens to fall upon one of the more pronounced jaw marks.



The other option might be to have the hub turned down and then press a polished sleeve over it. This is how lots of hubs are repaired that cost too much to replace.

I'm sure that with the small amount that you'll be grinding off, you'll be in good shape.

Really shocked to see such a poor quality of work being turned out. I wonder if Phil Bleazey has any comment on this he has had many of these apart yet I have not heard him pick up on this as an issue before. Perhaps it is a certain batch of Centa's
Ian

Good video 3G's, I used to bore soft jaws out to hold the finished end of components way back in the 70's as an engineering apprentice.

Nothing new, just good practice.

I must admit to being confused about this thread.
In all the preceding pictures, the bad scoring by chuck jaws is on the outside surface of the shell containing the centa mechanism.
What has that got to do with oil seal bearing surfaces which are internal / part of the component which mates to the crankcase ?
I am probably missing the point completely ...

Hmmm! Now then, this little component has multiple problems and the ones highlighted here are new to me and just add to the list. Trying to explain it all is going to be tedious and you may find it boring but here goes!

Firstly, I have never seen chuck jaw marks on any Centa coupling I have dismantled and I can only say that this is totally unacceptable and the component should be returned to Morgan for replacement. Polishing it with scotchbrite is just not going to work. The machine turned finish may not be too much of a problem as the Centa hub is of soft steel and there is a chance that the steel will polish under the seal ring ( mine did ) and that is why, I believe, that there have been few failures here. As to grinding this is one way round the problem but grinding is a technique for achieving great accuracy and the surface finish is a by product of a process which is expensive due to having to carry out a second operation on another machine. Ford used to use a process called superfinishing on their crankshaft journals and this was done using emery cloth in a giant clothes peg while the crank was spun round - seems crazy but it worked because the finished journals were selectively mated up with shell bearings so finish was more important than accurate sizing. My point being that a good polish in the lathe would be as good as having the Centa hub re ground and a good bit cheaper.

The greater concern with this component is the face at the end which mates with the crank/ flywheel. This is sealed with some silicone sealer at the factory. I have seen some with saw marks still on the face where the billet was cut off too short to clean up properly in the lathe! This face is one of great importance and it must be clean, flat and absolutely square to the axis of the part. There have been some failures of the female splines which are due to the part being made to the same spec. as the original Harley type compensator. The Harley one is glass hard and moves freely up and down on the splines. To make this a free running fit, it only fits closely on the flanks of the splines not on the major or minor diameters - correct for a hardened component. The Centa version is a fixed spline and made of soft steel this means that it has to fit on the flanks and at least one of the major or minor diameters - ideally both, sadly it only fits on the flanks. The upshot of this is failure, which Morgan and their dealers seem to blame on the nut being too slack. Take it from me, if the splines are loose then the nut will come undone in the first minute of running the engine, regardless of how tight it is! Tightening the nut will not make up for a poor fit on the splines. The point of explaining this is that if the nut is very tight and the end face is even minutely out of true or if there is any foreign matter ( silicone sealer?) in between the faces, then the crankshaft will be bent out of true. This can easily be demonstrated with a clock gauge on the Centa coupling while the nut is tightened and slackened - who ever checks that? This might explain a few of the broken crankshafts that have been reported especially as it will make the Centa coupling run out and this will affect balance.

My answer to this is whenever I have a Centa coupling out of the machine, I cut an O ring recess in the face and check the face for flat and true. You then slide the O ring over the shaft and follow it up with the Centa hub to push it home - no silicone nor any speck of dust - this is important.

If you get involved with the splines be aware that, to my knowledge, there are two different crank shaft sizes with about 0.2 mm. Difference in the diameters. The replacement, tough, alloy steel inserts I have made ( www.bleazey.co.uk/M3W.html#Image )have to be matched to the individual engine - a real pain in the bum when supplying mail order.

To answer nippymogs query the oil is held in by a seal on the outside of the Centa hub and by the silicone sealer on the end of the hub to stop it creeping through the splines.

Finally, if anyone should have engine bearing failure and finds bits of silicone sealer blocking an oil way somewhere - now you will know where they came from. I just hope that the oil filter gets them first. I am not familiar with the oil circulation path so I can't be sure how likely this is - sorry.

Here's another observation which begs a question. JV's pictures above show the hub as having a black finish which might indicate some heat treatment - does this mean they are now hardened? - that would be an improvement over the one I had. - sorry, I have searched but can't seem to find the original article.