Visit our Popular Forums

daveNZ · 03-03-2019, 13:21

I’ve had continual leaks from the Johnson brass pump on a 40hp Lombardini. Always assumed it was scratched shaft, replaced that, added a speedi sleeve when I wasn’t sure the new one was prefect enough, use o-rings instead of springs in the water seals; etc.

I’ve made a soft funnel and tube to collect the leak so it doesn’t drip on to the engine, and is easier to monitor.
It definitely leaks more at higher revs, but that stands to reason as there will be more pressure from the impeller.

There is a custom stainless takeoff immediately after the pump, for the drive-shaft seal, and a siphon break after that.

I’m thinking of measuring the pressure after the pump, and after the heat exchanger.
What sort of pressures should I expect, or am I barking up the wrong tree altogether?
Many thanks for any comments.

captstu · 03-03-2019, 13:35

Wrong tree. Coolant pumps always operate with pressure, typically 12 to 20 psi.

DougR · 03-03-2019, 14:20

I disagree with captstu.
I've see lots of water pump">raw water pump leaks as the result of rusted/ corroded exhaust elbows creating too much flow resistance in the raw water circuit.

Take your measurements before and after each component in the raw water circuit......
- raw water pump
- heat exchanger
- gear cooler if there is one on the pressure side of the circuit
- exhaust elbow

There shouldn't be more than three or four psi drop per component. You may see more than that at the exhaust elbow, but there you are measuring the flow resistance as well as the exhaust back pressure. If the pressure is high at the elbow, temporarily place a "T" in the circuit before the elbow and divert half of the flow overboard, or into a sink drain or back into the exhaust circuit down stream of the muffler, and see what that does to the leak.

DougR

daveNZ · 03-03-2019, 23:24

Thanks for the comments.
I should have said, the exhaust elbow and heat exchanger are as clean as I can get them, cleaned every year. Exhaust hose and water lock muffler are 16 years old but it is a good installation and I’m hoping they are ok.
If the pressure is in the order of 20 at least I know what sort of pressure gauge to get. And thanks Doug, that’s the kind of process I was thinking of. The gap between the heat exchanger and the exhaust elbow/mixer is less the 1”, only enough for a short connecting hose, and no room for a tee. There is a blanking plug on the outer cooled jacket of the elbow, not sure what it is for, but if I can get it out I’m sure I could measure the pressure there? But yes that would be exhaust back pressure and heat exchanger outlet pressure.

01kiwijohn · 04-03-2019, 09:21

Don't remove that plug in the elbow. It's there to facilitate removal of casting residue, during manufacture.
Is that pump belt driven? If so, you may be running the belt too tight. That will cause excess load on the two bearing, which are very close together and not designed to handle side loadings. If the pump is correctly assembled, it will not leak. Good, customary, workshop practices should be sufficient.
Usually, there's not much pressure in the system as it relies on unobstructed flow thru the cooler and into the exhaust elbow. Any pressure is limited by the impeller blades bending.
The fresh water side of the system should have pressure as the centrifugal engine cooling pump requires 13 to 20 psi to work properly.

FULLCIRCLE331 · 04-03-2019, 11:06

Do you have a vacuum break in the system. On my Volvo with a Johnston pump I was checking for a missing impeller blade and found the vacuum break had three different size niples that could be cut to the size you needed only the 1/4 inch hole was open. This was causing back pressure and breaking impeller blades. I opened it to a larger hole no more problems.

daveNZ · 04-03-2019, 11:30

Re the plug. It is a stainless elbow, so could the plug still be a manufacturing artefact?

I don’t think any of the fittings are small enough to be restricting flow??

Maybe I should bench test the pump, and slowly restrict the outlet, measuring pressure, and see if I can make it leak!!?

I’m no mechanic but service everything, and am as careful with assembly as I can be.

One differencing piece of advice is whether to apply marine grease to the lip of the seal.

(I’ve had same advice to use grease on the lip of a water lubricated StrongSeal shaft seal)

Horus · 05-03-2019, 02:17

I guess you mounted the water / oil seals with the inside to the outside. The springs must be on the pressure side! So on the water and oil side.

daveNZ · 05-03-2019, 11:52

Yes, thanks, seals are correct! I also have the original 1970s? engineering drawing for the pump.
One unusual aspect is the impellor end of the shaft is larger for about 20mm, so the seals have to be eased over this. The current seals were installed from the other end, prior to pressing on the bearings, so did not have to be eased over the larger end.

Capt Gill · 05-03-2019, 23:27

There's some misinformation in this thread, e.g.,
"The fresh water side of the system should have pressure as the centrifugal engine cooling pump requires 13 to 20 psi to work properly."
The OP is referring to the Raw water pump, also the fresh water pump does NOT require pressure to work properly, although most closed cooling systems such as my Perkins 4-108 have a pressure around 10 psi at operating temp. regulated by the coolant tank cap.
The OP asks about "too much back pressure?" which is a v good question and one I've been trying to find out for some time. I've contacted Bowman Heat Exchangers Mfg., they gave me flow rates, but were silent re. pressure, I asked because I've had a chronic issue of blowing the seal between the heat exchanger and rubber end cap resulting in sea water in the coolant (only happened when over 2K RPM), I tried many theories, too many to mention here, to remedy this problem, but kept returning to the idea of too much pressure (couldn't find press. spec's anywhere), then hooked up a press. gauge at the aft end cap and got 35 psi @1.5 RPM which seemed pretty high high enough that at higher RPM it could overcome the 10 psi coolant counter pressure, however what was causing the excess pressure assuming it is excess pressure ? Well, as serendipity would have it, I got a raw water leak where the rubber hose clamps onto the exhaust elbow due to rust buildup. I removed the elbow from the engine to clean it up, then checking the nipple where the 5/8 hose for the raw water injection attaches I couldn't push a #2 phillips screwdriver past where the nipple is welded to the elbow, it wasn't carbon buildup it was excess weld! There was only about a 1/4 " hole for the water flow – hmm, I really missed that one. Anyway, drilled out the hole and no more problem. Still would like to know what a normal pressure range should be. I believe these positive displacement pumps are capable of producing high pressures.

daveNZ · 05-03-2019, 23:48

Interesting!! What that reinforces is that I need to check the whole pathway, although not that many pieces. Maybe even the final skin fitting at 36mm ID (1 1/2”) is not big enough ?
This pump obviously leaked for the previous owner (we’ve owned now for 6 years) as there was considerable evidence of salt water damage from the pump which must have been leaking for some time.
Maybe the silty mud at low tide is repeatedly scoring my new seals?
I’ll find it!

Capt Gill · 06-03-2019, 17:40

" Maybe even the final skin fitting at 36mm ID (1 1/2”) is not big enough ?
This pump obviously leaked for the previous owner (we’ve owned now for 6 years) as there was considerable evidence of salt water damage from the pump which must have been leaking for some time."

I don't know what a "skin fitting" is, however my r/w pump did leak a bit from the weep hole prior to relieving the excess pressure then stopped after the fix, maybe coincidence, but I did rebuild the pump twice within a couple years; the first time it needed repair, the second time no.