Why is methanols flow rate so low???

Hi all
Hopefully somebody cam explain this mystery to me. Everyone keeps talking about sub-zero methanol cooling. Has anyone tried it?
Theoretically the thinner the liquid the faster it flows. Well not true if it needs to be pumped. I have been experimenting with methanol well actually mentholated spirits and found out the hard way that if you use it undiluted it becomes extremely difficult to pump. Especially at extremely low temps (-50). So its not actually the thinner the liquid the better. I did a little experiment to prove that too thin is not always better. Tests taken at room temp.

To fill a 1 litre container with water lakes 14.8 seconds.
To fill a 1 litre container with diluted antifreeze (50/50 concentration) takes 12.86seconds.
To fill a 1 litre container with diluted meths water (50/50 concentration) takes 18.74 seconds. The higher the concentration, the longer it takes thus the flow rate is slower. Which isn’t good when to comes to cooling down CPU’s. Don’t ask me for scientific formulas and stuff I don’t know.
At a temp of –50C undiluted meths is so thin that it is almost impossible to pump. It literally drips out.
Try it for yourself you will see what I mean. Maybe it’s just my pump but I doubt it. I have tried 2 pumps with similar results.

Regards.







my site

use pure h2o, 4 enough convection & hi flow rates

Liquid cooling properties - various liquids

If you're really serious about the differences and properties of various cooling liquids, this link will probably satisfy you:

http://st-support-cooling-electronics.web.cern.ch/st-support-cooling-electronics/CWGWeb/TABLE8-5.PDF

They apparently spent A LOT of time and money researching not only flow rates, but heat transfer properties of various liquids including solutions of various salts, achohols, etc. You may have to stare at some of the graphs for awhile to get it (at least I did), but the information is all there including freezing points.

I ran across this link about a month ago, and felt like I had hit the jackpot for information on below ambient cooling liquids.

Methanol has very low viscosity. (thickness) Most pumps are designed to pump a liquid with the viscosity of water. (Duh!)
So instead of pumping liquid, your impeller just twirls and does nothing. (or very little.) You could improve viscosity by adding water or antifreeze.

I agree. It is like your impeller pumping air. Well not that bad, but you get the idea.

That's about right, but not quite. The pump's are depending on water's surface tension to hold it together as it's flung off of the paddles of the impeller, alcohol doesn't have this property.
To pump alcohol as effectively, you'd need a pump that has a far better seal between impeller and housing, and maybe a larger impeller.

And Welcome to the Forums!

So would a solution be to use a higher rated pump, so that at reduced pump efficency, it would pump enough alcohol? Or would this not work for some reason?

JigPu

I think the solution would be to add a small amount of antifreeze. This would dissolve in the methanol and make it thicker.
Another solution would be to use a pump designed for lower viscosity liquids, like the fuel pump off a car.

P.S. It is NOT surface tension. Want proof? Soap essentially eliminates surface tension. A pump will pump soapy water just as well as regular. Don't believe soap breaks surface tension? Try this: gently place a small paper clip on the surface of tap water in a clean, rinsed glass. It should stay on the surface. Tap the surface gently with another paperclip or something similar. The first paperclip will not sink. Touch the surface gently with a piece of soap. The paper clip will quickly move away from the soap, (the surface tension pulling it back from the break,) and then sink.

Use a non-centrifugal pump.

Simple as that.

MajinSSJVegetto said:

Use a non-centrifugal pump.

Simple as that.

Click to expand...

I was going to suggest trying a diferent type of pump as well. The liquid may just be swirling around in the pump.

Welcome to the Forums Sparky003!

Hey CaffineHog, soap breaks up some surface tension, but does not get it all, not even close. Soapy water will still form a drop that sits still instead of being like alcohol and bleeding out all over the surface it's set upon. Maybe the same flowrate test should be tried with soapy water, I'd be willing to bet it would beat alcohol, but not plain water.

I'd agree that a different pump would be better at pumping alcohol, but fuel pumps (at least judging from the one on the front page article) aren't the best solution. Maybe a pump like the Rainbow would be more effective because of it's super flow rate and it's impeller design isn't an open one like most. It's fins are between two disks instead of being an open star shape like most.

One more thing to add that's a possibility, alcohol's molecules are very tiny compared to water, maybe it has more friction against the tubing walls as well as pumping troubles.

Cool… Thanks all for the advice.
Off the topic a bit now... but anyway
This is what I finally did to to increase the flow rate, I bought a new pump!!! My old setup pumped about 150 gph my new one almost 1500gph (5580 L/H). pumps site
Because the inlets and outlets on the pump are so big 1" I had to build a new W/B to prevent a bottleneck. I also used 1" hose.
After connecting everything and checking for leaks I fired the pump up.
It pumped the meths like it was water, at room thep that is… As the temp decreased the pressure did too, but not as much as the old setup and not nearly enough for the W/B to warm up like it did before.
My CPU now runs at a temp of between -11 and -7 depending on the cpu load. Coolant temp of -35.

This was an interesting experiment but definetly not worth the effort as I didnt achive much of a performance increase with my celeron. This is mainly due to the fact that I have crappy ram!!!!! When I decide to get myself a new cpu( probably a AMD 2200+) I will fill the freezer with meths again and see how far I can push it.

My site with slightly outdated pics