EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: pipe2null on April 19, 2021, 08:35:05 pm

Title: How to calc Max PSI from pump flow/head height?
Post by: pipe2null on April 19, 2021, 08:35:05 pm
This is a little off topic, but not sure which board is most appropriate for this.  I think my question is relatively straight forward, so you can probably skip to the end without missing anything.

I'm trying to figure out the maximum psi I should be testing with for home-made water cooling blocks I'm using in multiple electronics cooling builds I'm working on.  Yes, the "electronics" in question are computer components, but this is not a cpu cooling question.  I do not have a CNC machine to do the job "the right way", so I've been fabricating custom water blocks for electronic components that do not have standard heat sink mounting patterns (or form factors) using aluminum tube and plate.  With my current brazing skills, the result kinda looks like a kid's school art project with globby metal filler and a disproportionate amount of JB weld.  But hey, IT WORKS.  Well, USUALLY it works.  There was a bug in my design of an earlier version that resulted in a small leak and a fried motherboard complete with smoke, char marks and slag.  The severity of the mobo damage was very odd to me, but can't really root cause it without schematics.  It was a bad day, but a very educational experience.  I've since fixed the design problem and added more leak protection, and I've added more rounds of testing of my home-made parts after analyzing how the previous failure occurred, but I don't know what PSI I should be testing with.

To test for leaks, I've been filling the parts with compressed air, dunking them in a bucket of water, and checking for bubbles, but the PSI I've been setting the pressure regulator to is a random number I pulled out of the air.

- The maximum PSI a water block will experience in an open-loop fluid cooling loop is roughly the same as the maximum pressure the water pump will exert if the fluid return leg/drain side of the loop is completely kinked, blocked or purposely clamped AKA zero flow.
- I THINK the maximum PSI in a non-compressed closed-loop would be similar to open loop as far as pump/hose blockage goes, but overall pressure might be higher due to vapor pressure/expanding trapped gas as coolant temperature goes up.  If the system is working decently, coolant temp should not go above around 50 degC, but the system "should" shut down long before coolant could get anywhere near 100 degC (CPU temp of 85 forces system shutdown.)  I do not think there would be a significant pressure difference in max psi between open-loop VS non-compressed closed-loop, but I do not know.

So, *I THINK* the maximum PSI a water pump will exert is equal to the water pressure at the bottom of a column of water equal to the "max lift/head/zero flow height" the pump is advertised at.  Specifically economy/consumer grade centrifugal water pumps.  I am not sure if this is the right way to evaluate this problem, and not sure how to calc PSI from head height (and/or advertised GPM/LPH pump flow if required).

Any input is much appreciated.  Thanks!

Title: Re: How to calc Max PSI from pump flow/head height?
Post by: jmelson on April 20, 2021, 06:53:21 pm
Yes, the max pressure should be about at minimum flow.  A foot of head height with water is about 0.5 PSI.

Title: Re: How to calc Max PSI from pump flow/head height?
Post by: pipe2null on April 22, 2021, 02:57:23 pm
So the calculation is plain-old-simple volumetric weight per area, regardless of whether the pump is pushing into a hose or an olympic sized pool...  So a pump with a head height spec of 7.5 feet, using a 1 inch x 1inch square column of water to keep it simple:

Appx: 7.5ft * (12 inch/ft) * ( 1 inch^2 area) * (0.0361 pounds/cubic inch (constant from random internet "source")) = 3.249 PSI max pressure from pump.

So testing my home-made parts with 10 PSI should give me a nice safety margin but stay within the appx max working pressure for the silicone tubes I'm using.

Awesome, thanks!