Bosch VE pump shaft seal leak.

[greasemonkey]

I dont think that is right GM. The pump indeed turns at half crankshaft speed. It takes two crankshaft rotations on a 4 stroke engine to fire all the cylinders.
Therefore, the pump fires all four cylinders per revolution of the pump shaft (if it didn't,the internal distribution mechanism would have to be more complex).

That does seem logical. I still can't get it straight in my head. Not to worry.

[Head Womble]

Julian, as you didn't remove the timing cover and clean it out are you sure it's not bio left from the old pump ?

[Julian]

Julian, as you didn't remove the timing cover and clean it out are you sure it's not bio left from the old pump ?

Yes I'm sure.  There's too much and it's been going on too long.  I can't think it would be coming from any where else.  If it's not polar or monsoon weather over the weekend I'll have a closer look.

Lucky I bought two seals!

[1958steveflying]

If the bruise in the body of the pump is as slight as you say I cannot see a seal leaking, Hopefully when you remove the casing it will become obvious. I assume if it is leaking from the seal on the shaft the leak throws out from the pulley all over yet  leaking from where you say it will have just run down the back ?

[Julian]

If the bruise in the body of the pump is as slight as you say I cannot see a seal leaking, Hopefully when you remove the casing it will become obvious. I assume if it is leaking from the seal on the shaft the leak throws out from the pulley all over yet  leaking from where you say it will have just run down the back ?

My thoughts exactly the mark is minimal and I softened the edges with a tiny piece of 1200 wet and dry (without the wet!).  The lips (it's a double lip seal)were a good snug fit on what looked to be a perfectly finish on the shaft.  The taper to which the flange fits, gently expands the lips to working size, so there was no possibility nicking the lips during assembly.

Besides pressure testing the replacement, I'd like to try and pressurise this pump to find just where it is leaking from.

With regard to diagnosing on the car, the drive end of the pump sticks into the cam belt housing, and the leak is only visible from the drain plug at the bottom.

[Julian]

This was the stuff I was trying to remember earlier in the thread it's called Hylomar blue ... there's a rather poor quality video here ...

https://www.youtube.com/watch?v=yXiVZIBoHL0

What do folks think?

I've mailed the manufacturers asking about the application and if it's compatible with bio ... If i get one, I'll post their reply.

[1958steveflying]

Blue Hylomar as I know it is simply a gasket jointing compound that does not dry unlike red Hermatite which does, not suitable for your purpose at all in my honest opinion.

http://www.ebay.co.uk/itm/HYLOMAR-UNIVERSAL-BLUE-GASKET-JOINTING-COMPOUND-SEALANT-40g-FUEL-RESISTANT-/281159734843?pt=LH_DefaultDomain_3&hash=item41766d163b

[Head Womble]

Why not remove the cam belt cover and run the engine, then you'll see if it's the seal.
Lets face it you'll have to change the cam belt now it's contaminated.

I know it's a fare amount of work, and you'll need to be mindful of having no water in the cooling system, but if the leak is that bad it'll show itself well before any damage is done.

[julesandtash]

Just make sure you connect the oil cooler pipes to the cooler (you can sit the whole rad assembly on top of the engine) or block the ends of them before you start the engine.
If you dont, it will spray oil everywhere out of them. In theory it shouldn't unless the engine is hot as the oil thermostat should stop it but, from experience, it does.

But, Mark is quite right, you can run the engine without the front cover, serpentine belt and everything else on.
Just put the crank pulley and bolt back to ensure the cambelt drive dog does not slide on the crankshaft

[Julian]

I've briefly run the engine before without the belt cover when I had the same issue as Julianf.  It's a good thought, but I don't think you'll get a clear enough view to determine where it's leaking from and with the pulley and belt whizzing round the leak may get blown all over the shop.

I'm quite happy to pull the pump again ... it's a 5 minute job for us "old hands"!  I've had a look for sectional views of a VE pump and this is about the best I can come up with ....



Not even sure if it's a VE, many bits look very different, but if the fuel inlet is the same, then pressurising from the fuel inlet would be viable.

However, The nice man in the Bosch service place said that the IP will suck fuel through from the tank and that there was 90 psi on the seal which would imply that the fuel inlet was under vacuum and the seal pressure, therefore they can't be connected as per this drawing.

Getting confused!

[Head Womble]

In theory the shaft seal will be under negative pressure (the PSA XUD engine will pull air in the seal if it's gone), however with the lift pump before it that is not the case on a 300tdi.

[Julian]

In theory the shaft seal will be under negative pressure (the PSA XUD engine will pull air in the seal if it's gone), however with the lift pump before it that is not the case on a 300tdi.

From the design of the seal ... ie the outside face is concave, it looks to me that it's designed to seal pressure from inside the pump, not vacuum.

I thought the seal only saw the lift pump pressure and said this to the guy at the service center.  He said no it sees pressure form some part of the IP to the tune of 90 psi.

Still very confused!


Anyhow, I still have a spare IP that I know leaks.  If I try various means of pressurizing that and see it's leak from the seal, I'll know how to test the one that's currently on the car.

[Julian]

OK, up date ... I've discovered you can statically test a VE pump shaft seal with compressed air.

If you blank off the spill back connection (use the original bolt and a 14mm length of 15mm copper in place of the banjo) and connect a compressed air line to the fuel inlet, however the porting is arranged, it will put pressure on the seal.

I tried this with the old pump and it turned out that the seal itself was leaking.  There are a series of little dimples on the front face and air was weeing out of one of those,  So I can only assume that the metal and Viton make up of the seal had delaminated.

Armed with this information, I set about the fourth IP change.  Having got the pump off the car, the reason for the leak on the new seal I fitted was obvious ... it had popped out of it's housing and was sitting with it's front face up against the back of the drive flange.

This makes complete sense as I checked it several times after fitting and no leak was obvious.  Some time later it must have popped out causing the leak again.  At this point I decided that rather than trying to seal the seal in place, it needed "fixing".  After much thought, I whacked it back in with superglue!

After leaving it 10 mins or so I pressurised the pump to 80 psi again and covered the seal in gas leak detector fluid and not a bubble in sight!

I was in the middle of fitting the pump back in the car when bad light stopped play, so hopefully I'll get it back up and running tomorrow.  There may be a little issue to deal with though ... when moving the car on to the drive it wouldn't rev and was trying to stall all the time ... Oh that feeling of deja vu !!!

[Julian]

Back up and running again.  Half an hour of ticking over and no leaks ... wonder if it will last?

[therecklessengineer]

What's the stability of superglue in biodiesel?