| jeramifrost |
So i am curious. I understand the general concept of conduction. So i guess my question for ex. would be. If using moving water at say 20C over an object that is say 45C. Will the water remove heat from that object more efficiently than water at 30C. Or a simpler way of asking is this...does the rate of heat transfer actually slow down as 2 temperatures meet and if so is it a noticeable amount in this kind of temperature range? My materials in question if needed are water and copper.
And another question while i am slowly pondering..should the temperature of the two materials meet in the middle..so say 20C water meeting 30C copper level out to roughly 25C? And does the fact that the water is moving have any effect? If so how would faster/slower moving water effect heat transfer? |
|
|
| jeramifrost |
| It is very possible the answer is in there, however i am not smart enough to understand much of what is stated there. I am in need of someone to dumb it down for me :) |
|
|
| Entropy |
quote: Originally posted by jeramifrost
So i am curious. I understand the general concept of conduction. So i guess my question for ex. would be. If using moving water at say 20C over an object that is say 45C. Will the water remove heat from that object more efficiently than water at 30C. Or a simpler way of asking is this...does the rate of heat transfer actually slow down as 2 temperatures meet and if so is it a noticeable amount in this kind of temperature range? My materials in question if needed are water and copper.
And another question while i am slowly pondering..should the temperature of the two materials meet in the middle..so say 20C water meeting 30C copper level out to roughly 25C? And does the fact that the water is moving have any effect? If so how would faster/slower moving water effect heat transfer? | |
Need more context - 20C water should remove heat more efficiently than 30C water.
That's all I can tell you. For the other answers:
How much heat (watts) is being dissipated by the copper? What are you trying to cool off? A CPU? An engine?
How is the heat being removed from the water?
Moving water will have a huge effect. Still water will heat locally to the temperature of the copper and then conduct heat away slowly. Moving water will heat up and then carry away the heat via water flow. This is why all watercooled PCs (with a few exceptions) use a water pump - the rest rely on convection to generate water flow but this is very inefficient compared to forced flow. |
|
|
| jeramifrost |
| I am looking to water cool a CPU. IT will be cooled via a triple 140mm radiator. My reason for this knowledge is my dilema of how to best distribute the liquid. Basically will the liquid be that much less effective at cooling the NB and GPU after cooling the CPU to make it worth cooling in between devices or will 5-10C increase in the liquid between devices not really effect the cooling ability for the most part of the water. |
|
|
| CNY_Dave |
quote: Originally posted by jeramifrost
So i am curious. I understand the general concept of conduction. So i guess my question for ex. would be. If using moving water at say 20C over an object that is say 45C. Will the water remove heat from that object more efficiently than water at 30C. Or a simpler way of asking is this...does the rate of heat transfer actually slow down as 2 temperatures meet and if so is it a noticeable amount in this kind of temperature range? My materials in question if needed are water and copper.
And another question while i am slowly pondering..should the temperature of the two materials meet in the middle..so say 20C water meeting 30C copper level out to roughly 25C? And does the fact that the water is moving have any effect? If so how would faster/slower moving water effect heat transfer? | |
A simple principle of thermodynamics (study of heat and how it moves) is that the transfer rate of heat is proportional to the difference in temperature. So, yes, the greater the temperature difference between the hot object and the cold water, the faster the heat will flow from the hot object to the water. That's not to say a 20C difference has heat flowing 2x as fast as a 10C difference, though, because it's all done with absolute temperatures (referenced to absolute zero, -240C).
The temperatures would not 'meet in the middle', the ending temperatures (or the temp of the outflow water) will vary depending on how fast the heat can flow from the copper to the water, and how fast heat can flow into the copper from whatever hot thing you're cooling.
If you move the water too slow, it will heat up enough that tjhe heat transfer will slow down because the water will warm up. I do not know if you could move the water 'too fast', I don't think so.
Dave |
|
|
| nipper |
quote: Originally posted by CNY_Dave
If you move the water too slow, it will heat up enough that tjhe heat transfer will slow down because the water will warm up. I do not know if you could move the water 'too fast', I don't think so.
Dave | |
You cant move water too fast when it comes to cooling. You can move it too slow, as you do not want the cooling medium to reach the temp of whatever is being cooled.
nipper |
|
|
| jeramifrost |
| Thanks guys, that answers it perfectly. Now i just have to figure out what to do about it :) |
|
|
| nipper |
Well let me quantify that, yes you can move too fast, but then you get into cavitation and erosion of surfaces, but on paper, and for what you want to do, you cant move it too fast.
And remember use a 30% mix of antifreeze of some sort, and distilled water. |
|
|
| jeramifrost |
| Roger that on the mixture. I am gaining knowledge fairly quickly on the subject. Something that was bothering me on these systems though was that all i have seen look like they go from the main reservoir to the cpu, nb and then gpu before hitting the radiator and starting the cycle over. It just doesnt make sense that whatever is at the end of the cycle gets left overs so to speak. So i am trying to figure out if the water temps at the end of the cycle justify a smaller radiator in the middle of the cycle to components to keep efficiency up. I guess some testing will be in order for that. |
|
|
| nipper |
quote: Originally posted by jeramifrost
Roger that on the mixture. I am gaining knowledge fairly quickly on the subject. Something that was bothering me on these systems though was that all i have seen look like they go from the main reservoir to the cpu, nb and then gpu before hitting the radiator and starting the cycle over. It just doesnt make sense that whatever is at the end of the cycle gets left overs so to speak. So i am trying to figure out if the water temps at the end of the cycle justify a smaller radiator in the middle of the cycle to components to keep efficiency up. I guess some testing will be in order for that. | |
Actually thats fairly common, even in engine and cooling design. The idea is to keep the medium moving quckly so its not too big a deal. But if the coolant system suffers a failure of some kind, then it becomes a big issue quickly. If you can pick the least succeptable unit last in the cooling loop.
nipper |
|
|
| CNY_Dave |
quote: Originally posted by nipper
You cant move water too fast when it comes to cooling. You can move it too slow, as you do not want the cooling medium to reach the temp of whatever is being cooled.
nipper | |
Ayuh, I've just *heard* so many friggen times that you can that I almost believed it myself, despite knowing better.
Now time spent in the radiator, there you could circulate too fast, I think.
Dave |
|
|
| jeramifrost |
quote: Originally posted by CNY_Dave
Ayuh, I've just *heard* so many friggen times that you can that I almost believed it myself, despite knowing better.
Now time spent in the radiator, there you could circulate too fast, I think.
Dave | |
that would make sense as you wouldnt be allowing it sufficient time to release the excess heat. So it is a matter of finding a happy medium. Though with the size radiator i am looking at hopefully it will be a bit over kill and allow me to run slightly higher flow rates to get better cooling w/o moving through the cooler too quickly. |
|
|
| nipper |
quote: Originally posted by CNY_Dave
Ayuh, I've just *heard* so many friggen times that you can that I almost believed it myself, despite knowing better.
Now time spent in the radiator, there you could circulate too fast, I think.
Dave | |
Not really as it works the same in reverse too.
BUT you need a large surface area, and thats where it can get tricky. Water is a much more effecient cooling medium then air, so you have to work within that and take into account the max ambiant temp. That may take some experimentation. In your application, there is no such thing as too much cooling. If it was a mechanical widget that required lubrication, then there would be an ideal temp span that you would want to maintain
http://www.koolance.com/water-cooli...php?cPath=54_84
I dont know if that helps.
nipper |
|
|
| jeramifrost |
| I have been there. I am not sure if they carry the one i am looking at, i did not see it at first glance. I am figuring between 1kw and 1.3kw heat dissipation as i am looking at a triple 140mm setup as opposed to the triple or quad 120mm setup that i see there. And i guess the size of the reservoir also comes into play as the larger the tank the added benefit once any warmer fluid gets back to it...granted that will only work on the term as eventually i am sure the tank temps will rise to whatever the avg temp of the fluid is after it leaves the cooler. |
|
|
| tenpenny |
quote: Originally posted by jeramifrost
So i am curious. I understand the general concept of conduction. So i guess my question for ex. would be. If using moving water at say 20C over an object that is say 45C. Will the water remove heat from that object more efficiently than water at 30C. Or a simpler way of asking is this...does the rate of heat transfer actually slow down as 2 temperatures meet and if so is it a noticeable amount in this kind of temperature range? My materials in question if needed are water and copper.
And another question while i am slowly pondering..should the temperature of the two materials meet in the middle..so say 20C water meeting 30C copper level out to roughly 25C? And does the fact that the water is moving have any effect? If so how would faster/slower moving water effect heat transfer? | |
Simple answer is, as others have noted, yes, the greater the delta T (temp difference), the greater the heat transfer rate.
Another factor to consider in a heat exchanger is how the flow is arranged - in some units, the two streams flow in opposite directions, so that the exiting stream of the hot water is beside the entering stream of cold water - this arrangement will give the coldest exit temp of the warmer water.
There are reasons to use different flow patterns, depending on what the goal is, and how much cooling water you have, and what the temperatures are. |
|
|
|
