Quote:
Originally posted by gmat
Next, i think my pump is slightly cavitating. I have a parasite noise ("bzzzt") only when the pump is under load (ie in my PC). No such noise when the pump is free (ie in the kitchen sink).
The question is: should i port my pump inlet to prevent cavitation ?
|
Are you certain it's cavitation? It could also be simple vibration of the pump. Two things you can do to verify.
1) When it's in the sink, gradually blind off the intake with your thumb. As you restrict flow, the pump will begin to cavitate. Don't completely blind off the intake and don't do this for an extended period as cavitation will eventually cause permanent damage. This will allow you to know positively what it sounds like when your pump cavitates.
2) When it's in the computer, unmount it and either hold it in your hand or put it on top of a folded up hand towel. This will tell you if the pump is causing something in the case to vibrate.
There are a few things that can solve the first, though not all are practical. Ultimately you need to either increase the suction pressure or decrease the flow velocity.
Increasing pressure) You can put a water column open to atmosphere on the suction side (T it into the normal flow line). This will put atmospheric plus the height of the column on the suction side. Honestly, this isn't practical and would only help if you were "just barely cavitating" as you can only make a column so high.
If you really want to get funky, you can add an accumulator to your system (closed-loop only). This raises overall pressure throughout the system. All this amounts to is a vessel (with an air fitting) that you partially fill with water. Once the entire system is closed, you pump a few psi into the vessel. Obviously this has the potential to cause leaks around connections if they aren't tightly sealed. You really don't need much air volume in the vessel. Be warned, though, that overpressurizing an accumulator will cause fittings to blow.
Decreasing velocity) You can open the suction line as you have proposed. You can also restrict the flowrate, which lowers velocity at the suction. I suspect this second alternative isn't very attractive.