Diesel Engine Overheating

[Acushla]

I have had the occasional cooling water temperature alarm this year so decided to investigate further. The engine is a 20hp Bukh DV20Me with direct sea water cooling via a Johnson pump (rubber impeller). The engine spec states the cooling requirement is 11 litres/min at 2000rpm. A new impeller was fitted 3 months earlier.
I measured the cooling water flow at 5.8 l/m at 2000rpm using a bucket and stop watch.
I next cleaned the strainer but this is a coarse strainer anyway and checked the supply hoses for restrictions (found none).
Next checked the the flow with the fridge/freezer heat exchanger bypassed and this did make a 0.5 l/m change but no more than I would expect. Conclusion that the Fridge /freezer heat exchanger was not the cause.
Next checked the cooling water into the engine block, this was slightly contaminated with a calciferous deposit. This was cleaned out and the pipework cleaned.
Next came the exhaust manifold cooling water outlets, the thermostat housing and the thermostat bypass loop. These all showed some deposit build up and were all cleared out.
Finally the pump casing was removed for overhaul. The back face wear seems excessive but the cover is good(recently replaced) as there is a marked step between where the impeller runs and the area under the vac cam which is untouched. The unworn section measures 0.748 ins so have assumed that the casing needs to be machined to recover this dimension. The new impeller width is 0.751 ins and with a new gasket on the cover of 0.010 ins would mean a final axial clearance for the pump of 0.007 ins. My thinking is that as these pumps function by creating a small depression then any axial clearance is undesirable, therefore I favour reducing this clearance to zero for the rebuild.

Do any of you guys have experience on these type pump builds and the axial clearance requirements?

[perchance]

Excess axial clearance will allow water to feed back to the suction side of the pump from the pressure side. A clearance of .007 would make the the loss of efficiency worse.

[Acushla]

Perchance
Thanks for the feed back. I have had comments suggesting that up to 0.020 ins interference with these impellers is acceptable but this seems a bit heavy to me. What are your thoughts?

jim

[perchance]

Originally Posted by Acushla

Perchance
Thanks for the feed back. I have had comments suggesting that up to 0.020 ins interference with these impellers is acceptable but this seems a bit heavy to me. What are your thoughts?

jim

Are you talking interference or clearance?

[sailinman]

if your measurments confirm you have correct flow thru the pump then I would move onto other possible issues...

is the riser clogged? that is the most likly cause of over heating... on a raw water cooled engine, I would then suspect the water jacket on the motor block... an acid flush will clean up that surface like brand new...

that was what ultimately worked for my situation..

g/l

[sailinman]

oh, just re read the op.. sorry.. did you check the belt tension??? make sure the water pump is turning at the correct rate...

[sailinman]

also, I would verify the size of the opening and exit of the thermostat cover...

I doubt that the thermostat cover can handle that much flow...my thermostat has a 1/8 " opening...

if that? I cant put a pencil thru it.. I was thinking of reeming it out but didnt know how it would affect other components downstream...

[Acushla]

Sailinman
The Bukh pump drive is a geared drive so no chance of belt slip. The flow for this engine is quite small at 11 l/m (2.9 g/m) with this small impeller (1.57 ins x0.75 ins width) and the thermostat housing is new.

However the block waterways are definitely on the list of possible suspects here. This a wet liner engine so any form of chemical attack of the "O" ring sealing needs to be carefully assessed before chemical cleaning.

[Acushla]

Perchance
My original post identified that there would be a 0.007 ins clearance between the end faces. I think that this can be reduced to zero.
A mechanic associate identified that this could be reduced as low as 0.020 interference which I think would be too tight.

[perchance]

Originally Posted by Acushla

Perchance
My original post identified that there would be a 0.007 ins clearance between the end faces. I think that this can be reduced to zero.
A mechanic associate identified that this could be reduced as low as 0.020 interference which I think would be too tight.

It may not be. The edges of the impeller are not usually flat but have a small raised rib down the mid-line. The small contact area reduces friction while still providing a good seal so .010 compression per side would not be an issue.
Maintenance tip: Lubricate the end cover and pump cavity with Petroleum Jelly or dish soap in a pinch (I use needle bearing grease ) to prevent a dry start. It takes 60 seconds to burn up a dry running impeller at rated rpm.

[perchance]

Originally Posted by Acushla

Sailinman
The Bukh pump drive is a geared drive so no chance of belt slip. The flow for this engine is quite small at 11 l/m (2.9 g/m) with this small impeller (1.57 ins x0.75 ins width) and the thermostat housing is new.

However the block waterways are definitely on the list of possible suspects here. This a wet liner engine so any form of chemical attack of the "O" ring sealing needs to be carefully assessed before chemical cleaning.

If your engine has been running hotter than the thermostat rating for any length of time you will most likely have deposits in the block if you didn't already.

[Acushla]

I have machined the pump casing to the same width as the impeller and rebuilt it at this. The flow through the engine has increased to 8.3 l/m and the thermostat is now controlling the temperatutre. I have also installed a "Saltaway" injection system to help clear out any disolvable solids in the engine and heat exchangers so will go with this for a few weeks to see how everything performs.