Gear pump suction side cavitation

[Nich0]

While investigating a failed gear pump, I found signs of erosion/pitting matching damage you would expect from cavitation. However, this damage was found on the suction or low pressure side of the pump. Additionally there was a failed bearing which catastrophically destroyed the pump. So far I have assumed that the cavitation is a slower process and occurred first, potentially reducing the lubrication quality resulting in bearing failure.

All sources I have been able to find so far suggest that while cavitation is initiated in the low pressure region, the bubbles won't collapse until they reach the high pressure region allowing the vapour to transition back to a liquid (significantly reducing in volume and causing a shockwave with consequent damage). In my case there are no signs of cavitation on the discharge side of the pump.

My thinking is that as each tooth rotates, the change in volume between the tooth tip and the housing provides the localised increase in pressure required to transition the vapour back to a liquid, collapsing the bubble and eroding the suction side housing. I am unsure why there wasn't further cavitation at the discharge side when the rest of the bubbles reached the high pressure chamber.

Is this reasonable? Has anyone seen similar behaviour and knows what's going on?

 

[TugboatEng]

Cavitation on the suction side of pumps is very common due to low pressures. Turbulence can often initiate cavitation on the suction sides of pumps and piping leading to the pump.

Also, gear pumps are very noisy due to the rapid start/stop of fluid flow as each tooth cavity expands and then seals off. Noise is pressure waves so this may be enough to collapse a bubble that has formed on the suction side.

 

[georgeverghese]

There may be one or more high points on the suction side of the pump casing where vapor is trapped. Localised high concentrations of dissolved salts on the dry walls of the casing may then cause this corrosion??

 

[hydtools]

What kind of gear pump?
Generally the bearing fluid flow is from high pressure side through the bearing or bushing to low pressure return to the suction side.
Filling the space between teeth would be a low pressure event.
You may have contamination erosion.

Ted

 

[Artisi]

Photo's might add a clue.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[Nich0]

It was an external gear pump with aluminium housing.

 

[hydtools]

My first impression is that the housing is of poor quality. Your presumed cavitation damage is not symmetrical as I would expect it to be. The housing debris caused scoring as it was rubbed into the gear bores by the gears.

Ted

 

[LittleInch]

I suspect some of that may be due to the inlet chamber not extending the full depth of the gear housing and hence a lot more swirl and possible cavitation at that point.

Got a decent sectional drawing of this?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Nich0]

Unfortunately no, we don't have access to the drawings. The cavitation potentially did occur symmetrically however that evidence would have been lost when the driven gear bore down into the housing.

Another pump of the same design came off this machine with the same hours (~10k out of expected 16k) and there was no evidence of cavitation.

I think TugboatEng may have answered my question at the beginning. If the suction chamber conditions are close to the transition point, then small disturbances could initiate the collapse without having to enter the high pressure region.

 

[hydtools]

What would be causing the low suction pressure? How would air enter the suction? Is the pump suction flooded, i.e. below the tank fluid level? Do the two pumps have separate suction plumbing, separate or same tank?

Ted

 

[Artisi]

As hydtools asked, same supply source / pipework?
Initially looks like erosion damage from entrained solids, possibly from mechanical damage to the case etc?




It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

That damaged area looks like it will have high veleocities and swirl as it has a lot less surface area compared to the rest of the face. Not a great design IMHO.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.