How big is too big?

[man_utd]

You made your point, but if you could maintane a low head with infinite flow rate, he is still correct.

 

[clocker2]

Dude, stop trolling please, this is totally irrelevant.

from the original post...

At what point does a water pump get too big?

from SureFoot's post #23

You want the maximum flow, both in your waterblocks and in your rads.

Explain please why you think I am trolling.
You have made several unequivacal statements, the gist of which is "bigger is better".
When I question this I am accused of not reading the responses, failing to understand the responses and now, trolling.
How come?

 

[greenman100]

from the original post...

from SureFoot's post #23

Explain please why you think I am trolling.
You have made several unequivacal statements, the gist of which is "bigger is better".
When I question this I am accused of not reading the responses, failing to understand the responses and now, trolling.
How come?

you are accused for trolling for failing to read and accept information presented to you, as well as copping an attituide

an explanation was given for higher flowrates resulting in overheating, you failed to address it.

end result:

you are entitled to your false ideas

meanwhile

the enlightened will benefit from the truth

 

[uthungover]

"The reason behind why increasing flow in an automobile can cause it to overheat is this. First when you increase flow you transfer more of the engine's heat to the water. Since you haven't changed the radiator or the air flow over it, the radiator can't handle the extra heat being transfered. It's size and the airflow limit how much heat it can remove from the water. When you excede the limit of the radiator the water just continues to heat up until it's boiling point is reached. Overheating.

I think it's a bad analogy because of this:

Car: 125,000 Watt heat source --> 900 cubic inch radiator

PC: 70 Watt heat source --> 70 cubic inch radiator

AND

A car has a different design goal for the cooling system.

Ever had a car that didn't run right until it warmed up?"

From that thread - Tried running two of the rads you're using in series, rather than the one, with a higher flow?

Not the reason. If an engine is running at the proper temperature at low flow the system is in equalibrium, i.e, heat out the radiator = heat produced by the engine. Now if the flow is increased and, as you claim, more heat is trasfered to the cooling system the engine block would freeze. Simple enough.
A car engine may be a bad example because on the many varibles involved.
Here's some I can think of:
1. A good portion of the heat produced by an engine leaves through the exhaust.
2. Air to fuel ratios change with engine temperature changing the engines heat output.
3. Oils viscosity changes with temperature leading to more/less mechanical friction.
I'm sure there's more.
Computers are different. The heat output of the CPU is nearly constant (assuming a constant load) so if the the temperature of the system is in a steady state the heat out of the radiator is constant.
HEAT DISSIPATION IS CONSTANT, it's the equalibrium temperature that is impacted by flow rates.

 

[Sjaak]

Less temperature means that more energy has to be transported from the hot area to the cold area (cpu / water)

The speed at which this happens is dependant on the delta T, the difference in temperature (measured in kelvin, not fahrenheit or celcius). The higher that difference, the exponentially faster the transport of heat will go.

If more colder material gets flown in more quickly, the delta T will remain high, and the speed at which the heat is transferred remains high. When a fluid flows slowly, it gets heated up and the delta T decreases, and thereby exponentially decreases the efficiency with which the heat is transferred. Therefor, its better to have alot of flow, so that delta T remains high. Having the material longer in the hot area (slower flow rate, takes longer to pass the waterblock) decreases the heat transfer because delta T gets lower.

^^ I almost lost myself in that, but i hope you get the idea. English is not my native language, and saying what i know about such a subject is quite tough.

Yes, i am partially supporting SureFoot.

 

[uthungover]

1. Delta T can be in celcius, note the delta.
2. There are two, as you say, delta Ts that inversly effect each other, hence the resulting equalibrium.
4. "Less temperature means that more energy has to be transported from the hot area to the cold area (cpu / water)" This dosen't make any sense, the energy transfer is equal at all temperatures provided equalibrium is met.
5. I'm not supporting anyone, just supplying info.

Sjaak, When you say "The speed at which this happens is dependant on the delta T, the difference in temperature (measured in kelvin, not fahrenheit or celcius). The higher that difference, the exponentially faster the transport of heat will go." are you refering Newtons Law of cooling d(ΔT)/dt = -A(ΔT)? I'm guessing you are because the exponential dependance comes from integrating this law. This applies to objects that are cooling, not objects that are transfering heat at equalibrium.

 

[SureFoot]

Yes heat transfer (in J.s-1, or W) is proportional to the deltaT, with a fixed flow. (uthungover how did you insert those real delta symbols in there )

Also consider that, what holds true for a waterblock (more flow is better) also holds true for a heatercore. Heat is not picky about the direction it goes (water->metal or metal->water), il simply goes from hot to cold.

clocker2 you are trolling because you are steering the problem away from the reality at hand. Seeing water phase change into waterblocks or heatercores just out of pressure, is COMPLETELY out of reach, for anyone, without other serious trouble coming out first. As stated above, keeping the coolant in a car from boiling at high temperatures is an entirely different problem, we're *not* running our cooling water between 90°C and 100°C.

 

[uthungover]

I just copy those Δs from the Character Map in Windows.

 

[fafnir]

i can't believe you guys are still at it

 

[Etacovda]

Ill just add one more ******* comment;

When there was no thermostat in my car, it NEVER heated up.

More flow, more cooling, no regulation. SIMPLE REALLY ISNT IT.

 

[clocker2]

When there was no thermostat in my car, it NEVER heated up.

More flow, more cooling, no regulation. SIMPLE REALLY ISNT IT.

And you were running your car at redline for hundreds of miles at a time right?

Thought not.
Please try to remember that I am talking about race cars not your street vehicle.
Besides, if my automotive examples are going to be dimissed as irrelevant and non-germaine, then so should yours.

 

[Rpkole]

well if yall want i can sale yall some pumps from my work they have a flow rate of 300+ gallons per minute not hour and have a head of 300+ feet the only problem you might have is going from a 6 inch outlet to a 1/2 inch tube

 

[SureFoot]

And you were running your car at redline for hundreds of miles at a time right?

Thought not.
Please try to remember that I am talking about race cars not your street vehicle.
Besides, if my automotive examples are going to be dimissed as irrelevant and non-germaine, then so should yours.

Ok we'll end up with you car arguments now.
I think i understand what you mixed up: head and flow.
What one wants in a race car engine is:
* quick heat up of coolant to 90-100°C (for instance Mugen thermostats are rated for 92 or 93°C)
* NOT too much cooling, because if the engine is cooled down too much, **** happens (piston slack, bad lubrication, rocker arm to valve clearance problems, etc...), and there's a power loss.
Now, from your experience, a race car RUNS better with restrictor plates. This confirms the above: by restricting flow you're ensured that coolant reaches quickly >90°C and stays there (thanks to the thermostat..). Airflow on the radiator is often quite enough to compensate the lack of water flow, due to the high speeds.
Also, at these temperatures, water could quickly reach cavitation in the low pressure zone (== suction side of the pump, and a part of the loop on that side), which would be baad for cooling (gas layer). A restrictor plate will ensure that high pressure builds up before it, preventing cavitation in the zone that needs to be cooled down.

Note that the goals in a PC are quite different: we DO NOT want our CPU to run hotter. Moreover, even with perfect waterblocks, heat input to the loop could reach 250W in a huge system (big P4 + 6800 + NB + etc). Car engines put down 150 KILOWATTS (common for a street car..) or more. 250KW for a race car is an easy figure. That's A THOUSAND TIMES over !

Now your little troll about cavitation. So you reach 14m of head backpressure somewhere in your loop ? Insert a pump just before. No more cavitation. End of problem.

Note also that when NASA launches rockets into space (or other objects) they only use good old newtonian physics. They don't include relativistic speed adjustements. Why so ? Because there are several magnitude orders between our rocket speeds, and relativistic speeds, although relativistic corrections ALWAYS work even at 30kph in the street, simply they're not noticeable... See my catch ?
We're talking 300W (and that's an end-of-the-world system) heat input, 30°C max coolant temps (when things get very hot..) and pumps that starve at 2m of head..

Back on topic please.

 

[Sjaak]

well if yall want i can sale yall some pumps from my work they have a flow rate of 300+ gallons per minute not hour and have a head of 300+ feet the only problem you might have is going from a 6 inch outlet to a 1/2 inch tube

Im sure you can buy some piece of palstic to fit that .5" on the 6"


Yes, i was referring to the newton part. Im not an expert there, my dad is.

 

[matttheniceguy]

OK, I am a 4th year Mechanical engineering student, so I know my physics, but I am also a dye-hard tinkerer and know my way around a garage.

My last explanation was the sort of one I would give to people in my classes, I'll try explaning it like I would to my "salty language buddies"

When the water flows through whatever it is cooling heat is transfered to the water. The bigger the temperature difference the more heat is transfered. Whis is why cold water will cool something faster than hot water (pretty obvious ah?) So, when the water is flowing through whatever it is cooling, the water heats up as it flows through. As the water heats up, the temperature difference decreases, so the water doesn't cool as effectivly. If we push more water through the system, the "warm" water is pushed out faster, and the temperature difference remains high, so it will remove heat better. This is one reason a higher flow rate is better.

If you stick your finger under the tap while it is running on cold water, it feels really cold right? And then if you stick your finger into a cup of the same water, it doesn't feel nearly as cold. This is bacause there is more heat transfer with faster moving water, with more turbulence and all. This also makes higher flow rates better.

There are of course practical limits for a computer. The difference between a 300gph pump with 2m head and a 3000gph pump with 20m head will only be a couple degrees so why bother with the trouble. For most systems a good 300gph pump is plenty.

As far as the car questions, it could be from a lot of things. First of, you want the engine to be as hot as it can be without causing knock/detonatio/NOx emmisions, so it's not quite the same thing. As for your car overheating when you remove the flow restrictor... hmm... that's a strange one. This is a total guess, but is the flow restrictor put on a tube that will bypass the engine? I don't think your waterpump would appreciate you totally stopping water from going through it, so maby the system you saw this on was set up to that the engine and tube with the restrictor were run in parallel intead of series. With this setup, when you blocked the flow with the restrictor, all the water would go through the engine. If the restrictor were removed, most of the flow would go through the tube and not through the engine. I haven't seen a car set up like this before, but who knows.

 

[Sjaak]

[.......]

I tried to explain that 10 posts back, but i didnt succeeded as good as you, my english is breaking me up.

 

[SureFoot]

There are of course practical limits for a computer. The difference between a 300gph pump with 2m head and a 3000gph pump with 20m head will only be a couple degrees so why bother with the trouble. For most systems a good 300gph pump is plenty.

Ha, my thoughts, exactly, just in better english

As far as the car questions, it could be from a lot of things. First of, you want the engine to be as hot as it can be without causing knock/detonatio/NOx emmisions, so it's not quite the same thing. As for your car overheating when you remove the flow restrictor... hmm... that's a strange one. This is a total guess, but is the flow restrictor put on a tube that will bypass the engine? I don't think your waterpump would appreciate you totally stopping water from going through it, so maby the system you saw this on was set up to that the engine and tube with the restrictor were run in parallel intead of series. With this setup, when you blocked the flow with the restrictor, all the water would go through the engine. If the restrictor were removed, most of the flow would go through the tube and not through the engine. I haven't seen a car set up like this before, but who knows.

Good idea

 

[thorilan]

clocker2 did you read my post?
i edited it slightly just to make the important part more visible

 

[clocker2]

Yes I did.

It's well written and informative.

 

[uthungover]

When the water flows through whatever it is cooling heat is transfered to the water. The bigger the temperature difference the more heat is transfered. Whis is why cold water will cool something faster than hot water (pretty obvious ah?) So, when the water is flowing through whatever it is cooling, the water heats up as it flows through. As the water heats up, the temperature difference decreases, so the water doesn't cool as effectivly. If we push more water through the system, the "warm" water is pushed out faster, and the temperature difference remains high, so it will remove heat better. This is one reason a higher flow rate is better.

.

Not the reason. While what you say is correct, it dosn't apply here. The amount of heat dissipated is a constant. You will always dissipate the same amount of heat out of the cpu, always. The law of conservation of energy can't be broken either. It is the temperature that the dissipation takes place at, that is the concern. This depends on the radiator. The radiator is more efficient at high temperatures, less at cooler temperatures for the very reason you just gave. The cooling system will adjust temperature until the radiator disipates heat at the same rate as the cpu is producing it, which remember is CONSTANT.
I'm not adding anything to the flow debate. I can tell myself all day long that higher flow rates are better, which I believe, but I wouldn't argue it without experimental proof.