Flowrate...

[BillA]

This post deleted by its "derogatory and insulting" author so it will not be (so) necessary
"to defend the little guys against the tyranny of the know-it-alls."

 

[robertm]

err, I believe I said higher temps with less flow, no ?
(exacerbated by pump throttling)

the various graphs around showing flow rate vs. temp are all mine, it seems that no one else (that posts) has the equipment necessary to generate good data
and though it be ever so slight at much higher rates (say 3 to 5 gpm), more flow always results in lower die temps
- when the flow rate is the only variable

but since WCing systems are an interactive set of components, it is not difficult to encounter situations where the substitution of a 'larger' pump did not increase the maximum overclock

much of this difficulty relates to the pump selection process, it is not performance driven - its cost driven
(high volume pumps are being used where a higher head pump would be more appropriate)
there is, for example, an Iwaki pump to fit EVERY circumstance; but even a high-roller like myself won't pay their prices - cost again
(but eBay is a great resource, got a new Iwaki 0.6gpm 45psi mag drive gear pump for $25 - $650 list)

I continue to believe that the expenditure of additional effort in the pump selection process is worthwhile
one cannot reduce the heat load from the CPU, but that from the pump is another matter

be cool

I agree with you I think My point in the last post was for most poeple a larger pump will not hurt there temps as long as they have a system that can support the larger flow.

Often I agree with poeple but am thinking from a differant standpoint from them so I with them for no resion LOL.

 

[Ortovox]

LVM Waterpump

Hi
I have planned to build my own watercoolingsystem for a long time now but money has been the biggest reason why I havent built it yeat. A few weeks ago I got a waterpump for free from a friend, its a LVM Amazon waterpump. I was satisfied by the performance but after reading this thread I got worried about the added temp from the pump. The pump spec are 12V and 4.5 Amp = 54W = bad , Is there anyone that has heard anything about this pump or tested it (even though I dont think so). Im thinking of running it at 5V instead, what do you think about that? The first thing Im going to do is to run the pump bye it self in a closed system and see the difference in the watertemp.

Link to the pump
http://www.lvm-ltd.com/water pumps.htm

Jonas

 

[army_man]

hey why couldnt have brotten this topic up a few months ago before i took my test on thermodynamics. my physics teach is a ***** and readying ur article made more sense to me than learning from that *****. thanks.

dont ya just love physics

 

[Barryng]

hey why couldnt have brotten this topic up a few months ago before i took my test on thermodynamics. my physics teach is a ***** and readying ur article made more sense to me than learning from that *****. thanks.

dont ya just love physics

Actually it would probably be best if your English teacher also had the same personality as your Physics instructor.

If you really want to learn this stuff at a reasonable level apply for a job as an Operator at your local Nuclear Power Plant. They will teach you from scratch everything you need to know about Thermo without using a single differential equation. They even start with basic Algebra and Physics. You will spend well over a year in class, at a good wage with reasonable benefits, before you do one ounce of usefull work. The public typically has no idea how well trained these operators are. Unfortunately, they will not help with the English problem.

 

[Owenator]

Actually it would probably be best if your English teacher also had the same personality as your Physics instructor.

If you really want to learn this stuff at a reasonable level apply for a job as an Operator at your local Nuclear Power Plant. They will teach you from scratch everything you need to know about Thermo without using a single differential equation. They even start with basic Algebra and Physics. You will spend well over a year in class, at a good wage with reasonable benefits, before you do one ounce of usefull work. The public typically has no idea how well trained these operators are. Unfortunately, they will not help with the English problem.

RhoXS,

Sounds like you're looking for new recruits Don't forget to mention that most Operations staff work 12 hour shifts, which can be good or bad depending on the person. Personally I like the idea of 4 days on 4 days off with shift work. My engineering group only works the regular five day weeks (8-10 hours a day)

O

 

[Barryng]

Owenator, Salem or Hope Creek? I am in Ops training (SRO Cert) at Turkey Point but my education is BSEE. By any chance do I know you? Your last name the same of one of our not to distant past Presidents? Please say hello for me to your licensing manager GS. He is an old friend of mine that I have been very negligent in keeping in touch with. If you tell him I am his old car pool buddy from Ivanhoe he will know who I am and my office number. Give me a call.

 

[Owenator]

Owenator, Salem or Hope Creek? I am in Ops training (SRO Cert) at Turkey Point but my education is BSEE. By any chance do I know you? Your last name the same of one of our not to distant past Presidents? Please say hello for me to your licensing manager GS. He is an old friend of mine that I have been very negligent in keeping in touch with. If you tell him I am his old car pool buddy from Ivanhoe he will know who I am and my office number. Give me a call.

RhoXS,

I sent you a couple of PM's. You can also email me from the Bay Fan article link on the main page under winners: Owen Stevens

 

[baker269]

So since flow rate is king you would benefit more from running duel rads in series rather then in parallel?

 

[Tecumseh]

So since flow rate is king you would benefit more from running duel rads in series rather then in parallel?

No, this is backwards. If you have more than one radiator you
want to run them in parallel. Why? Because rads in parallel offer
LESS resistance to flow than if they were in series. Also heat
transfer is proportional to deltaT. With both rads hot, you have
greater deltaT and, thus, more heat transfer. If they were in series,
the second radiator, getting cooler water,
wouldn't transfer nearly as much heat out of the system.

 

[Barryng]

Tecumseh, is absolutely correct. Many large industrial cooling systems use multiple heat exchangers to reject heat. They are ALWAYS in parallel for exactly the reasons Tecumseh gave.

 

[fan freak]

decreasing the hose length does most definetely not alter the flow rate.

after all the flow rate is determined solely by the pump.

the reason the temperatures increased is because the volume of water decreased.

 

[BillA]

This post deleted by its "derogatory and insulting" author so it will not be (so) necessary
"to defend the little guys against the tyranny of the know-it-alls."

 

[res0r9lm]

fan freak you are definitly wrong. Right on the box of most pumps it tell you what the flow rate is for different heads

 

[Barryng]

decreasing the hose length does most definetely not alter the flow rate.

after all the flow rate is determined solely by the pump.

the reason the temperatures increased is because the volume of water decreased.

Each unit length of hose presents a certain amount of resistance to flow. The more hose, the more resistance and less flow. Conversly, reducing the flow resistance in any system, with a centrifugal pump, will increase flow.

Most of pumps that have been discussed here are centrifugal pumps. Centrifugal pumps ARE NOT constant flow devices. The flow in a system with a centrifugal pump is determined at the point, on a head vs. flow curve, where the pump curve intersects the system curve. Increasing the amount of hose, increases the flow resistance and moves the system curve such that discharge pressure increases somewhat but flow decreases. Also, niether the pump curve or the system curve is linear so the response to a change in the system configuration is not directly proportional to the change.

Positive displacement pumps (piston, gear,etc.) do indeed force a contant flow rate as determined soley by the speed of the motor. Increasing the flow resistance will just cause pressure to increase such that flow does not change. However, I have not yet seen anybody state that they are using PD pumps.

This type of stuff is frequently not intuitively obvious. It is not necessarily difficult to understand but it does not always follow everyday logic. Making a statement of fact, based on what seems reasonable, as quoted above, is just going to confuse other people.

 

[nihili]

Another explanation of why fan freak is wrong.

1. THe pump provides power, but flowrate is the result of the power of the pump and the resistance that the flow encounters. Pumping a liquid through a tube creates resistance. The resistance is determined by the cross section of the tube and by the length. The smaller the tube diameter, the greater the resistance. The longer the tube, the greater the resistance. So even with identical pumps, systems will have a higher flowrate to the extent that the tube are bigger in diameter and shorter in length.

By the way, this is easy to test now that the weather is getting warmer. Make a milkshake. First try to drink it through a long thin tube, then through a short thick one.

2. The volume of water has nothing to do with the temperature at equilibrium (after it has stopped climbing). A greater volume of water will hold more heat, so it will take longer for system to heat up, but it will eventually get every bit as hot.

nihili

 

[Tecumseh]

In reference to nihili's second point: You always want to
design your system with equillibrium conditions in mind.

You guys who only run your PCs a few hours a day can get
away with more marginal setups. BUT I think all designs
should be done with the idea of running a system 24/7.
You may not currently run your system 24/7, but you
certainly want the option. Don't you?

fan freak, we aren't picking on you. You are wrong.
Actually the responses are more gentle than usual.

 

[pommie]

My first time so please be gentle

Been reading a lot on watercooling before making an attempt myself.
As far as I can figure it out,
You want a flow as fast as practical through the block to keep the temp difference as big as possible, as you are cooling the cpu not heating the water.
At the cooling part ( radiator, coil, whatever ) you want the flow to be as slow as possible, as you are cooling the water not heating the air.
The "sweetspot" therefore is a balance between the two rates, fast enough to cool the cpu and slow enough to cool the water.

Or to simplify, use a bloody big rad

I have deliberately ignored any and all technical stuff to make a general observation

Cheers

edited to correct grammer

 

[res0r9lm]

that pretty much sums it all up pommie.

 

[nihili]

That does sum it up, with one proviso. It's not that you want slow flow in the radiator, it's that you want the water to be in the radiator as long as possible. Radiators still benefit from the increased turbulence of fast flow.

So what you want is a bloodly big radiator with small pipes and a pump strong enough to keep the flowrate high.

nihili