Flowrate...

Pepsi, let me try to answer your question.

The quick and dirty answer is "No". Submersible or otherwise does not matter when talking about "pump heat".

A pump simply transfers the mechanical energy from the rotating shaft to the fluid being pumped. It does not matter what form the pump takes (submersible, centrifigal, positive displacement, etc.) as energy must be conserved.

If you are driving the pump with an electric motor, the electrical energy the motor uses ends up somewhere and it must all be accounted for. The Ehiem, 1048, for example, is rated at 10 watts. Since the pump is not operating at runout (max possible flow rate), somewhat less than 10 watts is actually being consumed by the motor. Let's assume the motor uses only 8 watts. A very small percentage of this 8 watts is dissapated as heat from the motor itself. The rest ends up as mechanical energy in the rotating shaft. All this energy (assume 7.5 watts - this is a guess) ends up in the water and it appears in the form of an elevated temperature. Again, I must emphasize, that the type or form of the pump does not matter. The energy still is transfered into the water.

As an aside, once this energy is in the water, almost all of it appears in the form of a higher water temperature. The pump does increase the mass flow rate and therefore the Kinetic Energy of the water. However, surprising to many, the increase in Kinetic Energy of the water only reperesents an insignificant amount of the energy contained in the water. Something called the "General Energy Equation" explains this but this is well beyond anything that should be discussed here.

More flow means better cooling in the heat exchangers. However, we have a double edged sword here because increasing flow via pump capacity means we are putting more energy into the system that must also be removed. You will find a balance, or sweet spot as some others have called it, where temperatures will be the lowest. Assuming your pumps are centrifugal pumps, I would suggest controlling your flow by throttling the discharge (certainly not the suction) of your pump. Even though it is still a big pump, the amount of energy a centrifigal pump uses is proportional to the flow. Therefore, if you throttle the flow, you will be putting less pump heat into the system. The big motor may be operating a little less efficiently but this will be a very minor effect.

If it is not a centrifigal pump, do not even think of trying to throttle the discharge. For example positive displacement pumps (piston, gear, etc) are going to pump at a flow rate determined by the speed of the motor. Attempting to throttle the discharge of PD pumps will only result in an ever increasing discharge pressure until something bursts. Also, if you try to throttle the suction of any pump, you seriously risk cavitating the pump.

I hope this answers your question.

Welcome to the thread Master of Twisters.

If you are generating flow, you are generating friction. Otherwise, you would not need a pump to keep the water flowing, it would just cycle through the system continuously. Where the friction occurs, that is where the pump heat is actually generated, and where mechanical energy is converted to thermal energy. The impeller, or whatever mechanism moves the water, generates very little heat from it's contact with the water, since the friction there is not very significant due to it's size relative to the total system.

The friction heat is partially absorbed by the water and partly by what is causing the friction(rads, tubes, blocks). There is also internal friction caused by turbulence within the water, though I believe that that amount of heat is very insignificant. RhoXS is right about the pump heat, inline or submerged, most of the heat is generated from the flow of water and the work done to move it and stop it, not given off by the pump itself. I would guess, and it's just a guess, that you only add 5-10% more heat from an inline versus submerged pumps, but probably less. Submerged pumps put all of the heat they generate into the fluid. That could be an extra watt or two to dissipate, but it would seriously depend on the pump. Most inline pumps are designed to use the fluid being pumped as a coolant for the pump anyway, so it gets in there with inlines too. Some inlines are better at that than others and would put in more heat energy, probably not much though.

adding restrictions to decrease flow will also cause heat. being that it is a low pressure system I'm not sure it would be noticeable or not. seems like lowering voltage by means of a dimmer switch or something similar would be more benifical.

Okay, this rates a sticky!

Hoot

Another great thread!

Thanks! to RhoXS and jimmytp for further explaining & clarifying flow rate in relation to our purposes. I've already printed this out for future reference, as I do w/ alot of really informative threads.

Also, the response about whether or not inline pumps add heat to the system immediately made me think back to when I had a salt water reef tank and I knew the answer was yes. Just ask anyone who uses mid to high flow pumps in their setup. Keeping the water temp. w/in a suitable range is in ongoing battle w/ many reefers, and, much like overclockers, reefers will go to great lengths to do what is necessary. The same applies to flow rate. Finding that perfect balance is an art all its' own.

The more threads I read about flow rate, etc., I can't help but to be amazed by the similarities I see in water cooling and reef systems. The hardware and application is different, but the questions are still the same. It really is interesting to see how much can be learned, and applied, from other hobbies and disciplines to our own.

Good answer,
I have to admit this is probably the first time I asked a question here and was overwhelmed with the reply. The pumps I use generate a serious amount of heat to the point they are usually too hot to touch this being the exterior part of the motor housing. I have considered using a portion of the system to water cool the pumps themselves, but after thinking aboout I decided not to dump that kind of heat into my system instead I went to air cooling them and trying to move the heat off of them that way. It seems to work I say this because when I went this route my coolant temps showed a change for the good. With exception of the motor it's self everything is plastic other than heat that actually went through the armature of the motor to the impeller it's hard for me to see how a large amount of heat would escape the air cooling and invade my system. Thinking about it a fun experiment would be to fire up the pumps with no air flow over them and monitor pump temps and coolant temps right at the exit of the impeller. Then do the same experiment trying my best to keep the motors cool.???????????
Thanks for the great info STICKY !!!!!Way to go Hoot !!!!!
Stay Cool
pepsi

I am glad that this information has proved to be of value. It makes it very easy to go to the effort to explain it when someone finds it useful.

Question: I have a Eheim 1048 pump inside my 4" PVC Bong. I am wondering if I could fit a 1250 inside 4" PVC?

The specs on Eheim's site say the 1250 is bigger, The 1048 was almost JAMMED in. I am wondering if anyone has done it before?

I would really like the extra head height, and flow rate without putting the pump outside my bong.

The ehiem 1250 is quite large. If your would like I can measure mine for you but I can say that it is much bigger than the 1048.

Before anyone goes off thinking that they can reduced
flow rate to simple equations for heat removal, this
is not freshman Physics 101.

It's just a bit more complicated.

I strongly suggest that you all go back and read
Bill Adams' excellent article on radiators.

Go directly to it:
http://www.overclockers.com/articles481/

Tecumseh said:

Before anyone goes off thinking that they can reduced
flow rate to simple equations for heat removal, this
is not freshman Physics 101.

It's just a bit more complicated.

I strongly suggest that you all go back and read
Bill Adams' excellent article on radiators.

Go directly to it:
http://www.overclockers.com/articles481/

Click to expand...

I'll second that. Although the individual physics above are correct, their application to optimizing equilibrium values in a complex multivariate system leaves much to be desired. Yes, increasing flowrate will generally increase the efficency of both the water block and the radiator. However, increasing the efficency may lead to higher temperatures at the die. It's a matter of balance. There is no universal answer other than to test your system in a variety of different conditions.

nihili

Faster flow is always better as long as nothing else is changed.

FAST
-----------
Say that you do increase the speed ... this means that it will have less time to cool in the radiator but it doesn't have to cool off much because it wasn't in the waterblock heating up very long. It will always have cooling power while it is in the waterblock because it hadn't heated up much. (Hotter water wont cool as well as cooler water ... duh)

SLOW
------------
On the otherhand say its moving slow ... this means that when the water first enters the block its at .... ohh lets just say 30*C ... by the time it gets to the center of the block it might be at 31*C (numbers are not accurate just to show example). This means that the hotter water will not cool the block as fast and by the time its out it might be at 32*C. However, the water will get cooled all the way back to 30*C in the radiator because its moving so slow.

I have not tested this but it makes sence to me. I'm sure in a real world test these results may not be the same but if you could keep all other variables constant then faster would be better.

...be sure you arent confusing water velocity (speed of water per second) with flow rate (volume of water per second). you can increase the velocity while maintaining the same flow rate, and vice versa.

(i know what youre saying, you just shouldve said "flow rate" instead of "speed" )

I have to agree increasing the volume of water to absorb the heat in question works better. The system has less work to do to get rid of it = less stress, but if you read up earlier larger volume = bigger pump = more heat from the pump it'self (I think that was the general drift of the conversation), but it seems to me be a trade off for better temps...........
I think this thread does justify one of us to do some experiments to see (this would only be valid for your particlar setup too many varibles) say pump temps for high and low flow, water temps for the same, processor temps add it all together and see how that came out in the wash.
Just a thought,
Stay Cool
Pepsi

I have major problems with the idea that these pumps when not submerged at that much heat to your system. The 28Watts that the pump is using/making depending on how you want to look at it will be spread over the outside as well as the inside of the pump. Infact water is in contact with a very small amount of the pump am sure most of this heat is not put into the water system but into the case temps. So going from a 10 watt pump to aq 28 watt pump is not going to add 18 watts more heat to the water in your system. Unless it is submerged.

Also the 10Watts on the Ehiem 1048 is the power consumtion of the pump not what its heat output is. The heat output will depend alot on how efficient the pump is. But lets forget that and say it does put out 10 watts of heat now the water in your system at best is in contact with 20% of the pumps over all serface. So it should get about 20% of heat so about 2 watts at best. I figure you very unluky if a Ehiem 1048 adds more than a watt of heat to your system.

Am sorry if I cause problems for poeple but there is no way a ten watt pump that is not submerged is going to add 10 watts of heat to the water in my water system.

Robertm, the amount of energy a pump puts into a system is independent of whether it is submerged or not. The heat dissapated from the motor is only a very small percentage of the total energy used by the pump so this heat is negligible and can be ignored.

The 10 watt rating of the 1048 Eheim refers to the maximum mechanical work done by the shaft. However, this mechanical work is ultimately converted to heat in the form of a temperature increase of the pumped fluid. The fluid undergoes a temperature rise because of the work done on it by the pump; NOT because it absorbs losses from the motor.

My problem is your are trying to make a 100% efficient motor and they dont exist. At best these electric motors are 60 maybe 70% efficient thats just the motor then you have to find our how efficient is a impeller water pump. After doing some looking for pumps that list there consumtion and output wattage most of these pumps are around 50% efficient. So I will assume a 10 watt pump is going to add 5 watts to the water. Where does the other 5 watts go? It turns to heat in the motor. So if the pump was submerged then all 10 watts would be added to the water. When its not submerged then this heat is not added to the water in the system.

The 5 watts for the small and the 11 watts for the larger pump. this is not even close to the output of AMD processors in. My T-Bird puts out 90 Watts

I dont dissagre that the pump adds heat just with the fact all 10 watts was getting added even if the pump was not submerged.

Even if 50% of the rating was converted to heat, most inline pumps are designed to use the fluid as a coolant, so it gets put in there anyway. You are right, and earlier in the thread I said that ALL the heat produced by a submerged pump is added to the fluid and that not all the heat from an inline is. There has to be some heat radiated to the environment by inlines, so not all the heat is added, but most is. It depends on your pump. If it is warm on the outside(compared to the fluid being pumped) then it is giving off heat to the outside world.

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."

BillA said:

let's put some frosting on the cake;
there is another variable not addressed in the preceding discussion:

the additional pump heat generated as a consequence of the head its pumping against

I'm not at all sure how the mfgrs' Wattage numbers are developed; but from extensive testing I can state definitively that the least heat will be added to the system with the pump at its maximum flow rate against a nominal head (back pressure)

as the back pressure is increased (and the flow consequently decreased), so is the heat generated by the pump increased, exponentially
- the maximum occurring when the pump is deadheaded (at shutoff)
I'm fairly sure this is not how the mfgrs rate their pumps, so users of big pumps that have them throttled back are adding a LOT of heat (unnecessarily) into their system

it is worthwhile to match the pump's flow and head capabilities/characteristics with the specific components of the cooling system, bigger is not necessairly better

be cool

Click to expand...

Your correct Bill that is a big issue I think alot of poeple overlook. Now the problem is we as overclocks need to find the best way to messure and predict what head and what volume we need in our systems. I dont think I want to try and come up with a standerd am to lazy and there are alot of variables. Also I think to often poeple forgot about the mass amount af variables in the the water cooling system. Thermial dynamics added with Fluid dynamics is very complex and you can not perdict preformace well without taking into account all the variables.

All that being said I also think larger pumps(to a point) over all can and will help most poeple. If someone can double there flow rate there is a better than average chance with hardware we are using that the gains in the system will outway heat added from the new pump size. To many poeple have proven this with the parts we can buy to make a water cooled system. If am wrong I would love to read the review where someone got higher temps with more flow. Would be something to ponder over.