A question for Mechanical (or Chemical?) Engineers

[wilkin250r]

I'm looking for an in-depth explanation on how coolant flow rates affect overall cooling efficiency.

Too slow is obvious. If you're not moving heat from your engine to your radiator to get dissipated, then what's the point of having a radiator? That one is simple.

But what about too fast? To me, it doesn't seem possible.

I've heard some people say "the coolant doesn't sit in the radiator long enough to dissipate heat", which seems silly to me. It doesn't enter the radiator at 180 degrees, and leave at 140 degrees. If you went by that train of logic, then it doesn't stay in the engine long enough to pick up heat, either, but I hear people claim that they're coolant is boiling over because it's moving too fast. Heat transfer is a two-way street, if it's too fast to dissipate heat, then it's too fast to absorb it, also.

It would seem to me that the only way it could be "too fast" is if the water going through the center of the cooling passages was going too fast to mix with the water at the edges of the cooling passages (the portion actually in contact with radiator tubes and fins, the portion actually being cooled). But with all the bends and turbulence inside a radiator, I just don't see how that would be possible.

Can somebody give me an in-depth explanation? A simple "it doesn't sit in the radiator long enough" just doesn't do it for me.

[hypersnyper6947]

Your guess at , if its moving to fast then it shouldn't pick up the heat either, i think is where the problem is. The engine temps are so high that the coolant does pick up the heat quickly but then the radiator or ambient temp is not low enough to make up for the extreme heat the coolant picked up inside the engine. So although the coolant only spends a short time in the engine it also only spends a short time in the radiator. Now if you sprayed some nitrous or something on the radiator to make it an extreme cold to combat the extreme heat then it would be more apples to apples. Thats how i look at it hope it helps.

[wilkin250r]

But if it was moving slower in the radiator, it would also be moving slower inside the engine. It would pick up even MORE heat, simply because of more time in contact.

[hypersnyper6947]

Move it faster through a bigger radiator with more fluid, this way it moves through the engine quickly but has ample time in the radiator to dissipate the heat.

[400exchick]

PM 44oEX, he's a mechanical engineer and will be able to explain it to you. He doesn't come on here often anymore though so by sending a pm he'll get the e-mail notification.

His name is Conrad

[wilkin250r]

Originally posted by 400exchick
PM 44oEX, he's a mechanical engineer and will be able to explain it to you.

His PM box is full.

[400exchick]

Sent you his e-mail in a pm ;)

[zach0199]

Here is my simple explination, im a mechaincal engineering technology student here is basics of what you are taught in thermodynamics. This goes to the change of energy in a system. In ideal situations it is considered a steady flow process so kinetic and potential energy is negligible. Coolant will not expericne a phase change ie liquid to solid to vapor in this application. Different types of coolant will have a different, entholpy, specific heat, entropy, at different temperatures etc... So within the system we are expereicning the temperature change causing boiling over. Exit temperature change can be caused by a change in mass flow rate, the cooling area, externat heat there are variable factors for every application. Im just going off of what i remember from class its been a couple months. Specific heat should remain the same in the coolant and water. However bad coolant could change mass flow rate causing slight change but shouldnt be a problem. We want the heat gained by the water to be equal to the heat lost by the coolant. Change in temp will cause a change in work done to the system causing an imbalance. Im finding it hard to anwser the question directly. Lets say the equation you can use to determine work into a system is work=(mass flow rate)*(Specific Heat)*(change in temp). Im sure someone more expericned can help you out some more. Pm me your email I can send u 2 examples that might clear things up for you if your still confused which i assume you will be i am confused just typing all that.

[Ralph]

I pondered over this for a bit and wanted to check my Heat Transfer book but I keep it in my office. Will take a look at lunch

[mxduner]

Originally posted by wilkin250r
But if it was moving slower in the radiator, it would also be moving slower inside the engine. It would pick up even MORE heat, simply because of more time in contact. wilkin, good subject. i feel your on track. I believe it's all about the balance here. In my time in the classroom for hvac (no mech. eng.) The biggest thing that made me understand cooling was temperature difference. It gets complicated for me lol, but in refrigeration, you want to absorb heat(head,cyl.) in a wanted space, and move it to a a place you are not concerned about (outside air, rad.). in refrigeration they BALANCE the flow by use of an orifice just like a jet, but like the head gaskets hole size differences. All this the engineers can break down more detailed, my wrist is injured so typing with 1 hand sucks.

I hope this contributed or 'got your wheels turning' on the subject lol

thermodynamics, entholpy/charts i bet will be later discussed i'm gonna get my popcorn ready:D

[wilkin250r]

Originally posted by mxduner
in refrigeration they BALANCE the flow by use of an orifice just like a jet,

And refrigeration is a completely different type of heat transfer. It's like a rock and a hill. Transferring heat from your engine is like rolling the rock downhill, you're going from high energy to low energy. But refrigeration, you are trying to pull energy (heat) from a system that's already LOWER than it's surroundings. You're rolling the rock uphill.


But back to the original, I STILL don't get it. Let's pretend it all simplifies to time. Does it really matter if I absorb energy for 5 seconds, then dissipate energy for 5 seconds, or if I absorb energy for 1 second and then dissipate it for 1 second? To me, it seems all the same. In fact, to me, it seems that the FASTER I can move it, the better, always, no exception.

What am I missing?

[400exrider707]

While the amount of heat dissipated by each measure of coolant passing through the radiator will in fact drop due to time constraints, The total heat dissipated will increase because the flow rate is increased. In other words, that same quantity of coolant does not have it's temperature reduced as much per time it cycles, but it cycles more often resulting in a greater net heat transfer.

You are right though, it can move too fast through the radiator. For instance, if you took your thermostat completely out of a motor, it would go too fast and wouldn't cool, it would overheat in fact. I've heard of some people stripping the thermostats of their guts, but leaving the "body" in order to slow the coolant but allow all coolant to flow at all times.

[44oEX]

Well I would say that since the water jackets in the cylinder and head are bigger then the ones on the radiator. So if your really going too fast it will still pull the heat but won`t have time to get rid of it. That would also depend on the size of the rad, the bigger the rad, the faster you should be able to go because you will still have time to loose heat.


My classes of Thermo are starting to get far lol

[wilkin250r]

Originally posted by 44oEX
Well I would say that since the water jackets in the cylinder and head are bigger then the ones on the radiator. So if your really going too fast it will still pull the heat but won`t have time to get rid of it.

I don't see how that's possible.

ANY closed system like that would have the same ratio, regardless of the actual speed. 6 seconds in the head, 2 seconds in the radiator to cool. I don't see how that is any different than 3 seconds in the head, 1 second to cool.

It's the same ratio, it HAS to be. So if it's the same ratio, why doesn't it have the same cooling efficiency?

[44oEX]

Your right it is the same ratio, i`m just saying since you have more time to absorbe the heat, then to get rid of it. It would all come down to how efficent your radiator is.

You do have a good question, and heat is not really my expertise, i`m more in machine design.

I`m thinking more along the lines of, even going faster your probably taking out more heat (of the engine) then the rad can loose. The radiator has a point of efficenty, if you trow to much heat at it, or your coolant is just going so fast that it just can`t transfer fast enough to cool it, every cycle the coolant goes back to the motor a little hotter, and it just gets worst and worst.

Beyond that i really don`t know.

[Fred55]

You also have to factor in that the heat transfer from the engine to the coolant is through conduction and the h.t. from the coolant/radiator to the air is through convection, which is much more inefficient compared to conduction. So in a sense the amount of heat transfer needed to properly lower the coolant temperatures does need a certain amount of time in the radiator. So "too fast" for the radiator is still far less than the "too fast" for the engine because of this.

[rbgnwa45]

I'm not educated on this, but this is how I see it: either or.

You'll either need a large reserve with a (?) computer-calculated in-and-out ratio or a fast flow of coolant that tips the scales of the linearity in the coolings favor (as in it cools more than the engine can produce heat, resulting in less-than zero degrees, lol).

Cooling seems linear, but if you can make a faster flow (vs. waiting) more efficient, that seems like your best bet, it seems more do-able, unless the wait period drastically cools more, for instance - 180 degrees into 100 degrees (waiting) vs. a faster constant change per, I dunno, second? With a faster & constant flow of coolant it'd have to be doing something inbetween the engine & radiator to cool it even more than the flow-rate is efficient, it'd have to cool more than the linearity it leaves the engine as, so it's tipping the scales of efficiency onto the cooler side.

Either or because there are two factors that you can change, but which is more efficient? If you wait for heat dissipation, you'll eventually have a differing actual heat because let's say your engine is (I have no clue about engine temps) 180 degrees, you're riding fast. You stop for 1 hour and the coolant in the radiator is not in-line with how much it needs to be for a constant & calculated dissipation when the engine is again started and vice versa, it's not as reliable if it can't be absolutely linear.

Waiting doesn't seem like it'd account for environment changes such as this, which is why I think the more linear you get it the better. Obviously the coolant has to stick around long enough to absorb heat, but couldn't you just as equally make the flow of coolant that-much faster for dissipation, outweighing waiting?

MX bikes have air-fins to help cool the radiator. That's great until you're riding slow, so it'd be better if the normal operational temperatures were calculated as if there weren't fins or any wind help and the engine is able to run either at normal-or-cooler temperatures, therefore almost no overheating issues.

Like a hot frying pan...

If you were to take a hot pan and poured in an inch of luke-warm water vs. small drops of nitro glicerine (that cold stuff, whatever it's called).

There's a bunch of other wind-cooling techniques that could be done to direct more air-to-radiator flow, for instance the front fender could have a narrowish tube through it's bend ending/pointing at the radiator, and the number plate could have fins it too.

If the coolant in the radiator doesn't stay there long enough to properly cool than a larger everything would be needed.

This is confusing the chit out of me.

[400exrider707]

Originally posted by wilkin250r
I don't see how that's possible.

ANY closed system like that would have the same ratio, regardless of the actual speed. 6 seconds in the head, 2 seconds in the radiator to cool. I don't see how that is any different than 3 seconds in the head, 1 second to cool.

It's the same ratio, it HAS to be. So if it's the same ratio, why doesn't it have the same cooling efficiency?

Does the ratio stay the same when you put a different radiator on? What happens if you put a high flow pump on? Do you think it's just a ratio or is an exponential function?

Why is it a bike with a thermostat in it (even if it's just a thermostat that is stripped, so it's just the body) will not overheat normally, but if you remove the thermostat it WILL overheat...

Try it. We bought an 04 450R that we built from frame up. We stripped this bike down to the frame, and built it back up with top quality MX components. We did run a stock head that came on the bike for the first few races (until we could get a ported head from Dave @ HRE). Bike had a 14:1 baldwin piston and a stage 3 hotcam. It would rip for what it was, but if you let it sit and idle, it would start boiling over. As long as it was moving, it was OK. Tear back into it and sure enough, there was NO thermostat in the head... Put a new thermostat in, and whala, bike would idle all day long with stock radiator and NO fan on it, without issue. How do you build a bike from the ground up and miss that? Guess it's never safe to ASSume anything.

That leads me to another point. It's not always just a ratio because of ambient temps outside and how much air is flowing over the radiator. A radiator that is not moving will not work as well as one that has nice cool air blowing over it....

Just some more food for thought.


Wilkin, have you been pretty active on here? I remember reading a lot of stuff from you back in the day. I haven't been on here a lot myself, but starting to come back.

[wilkin250r]

Originally posted by 400exrider707
Wilkin, have you been pretty active on here? I remember reading a lot of stuff from you back in the day. I haven't been on here a lot myself, but starting to come back.

I took a little time off, I'm recently back into the lifestyle. :D




I agree there are always a bunch of other factors, like the size of the radiator, airflow across it, ect ect.

But like your 450r boiling over at idle, the thermostat is the ONLY thing that changed, right? (let's assume it is). So the flow rate is the only thing that changed, and high flow rate overheated, but lower flow rate did not. WHY?!? :huh

This is what I want to know.


Like I said in the first post, a low flow rate is obvious. The coolant is moving the heat, but if the coolant isn't moving, then neither is the heat. Obvious.

But if the coolant is moving the heat, then how can you overheat if you're moving the heat too fast? I DON'T SEE HOW THAT'S POSSIBLE!

[400exrider707]

Originally posted by wilkin250r
I took a little time off, I'm recently back into the lifestyle. :D




I agree there are always a bunch of other factors, like the size of the radiator, airflow across it, ect ect.

But like your 450r boiling over at idle, the thermostat is the ONLY thing that changed, right? (let's assume it is). So the flow rate is the only thing that changed, and high flow rate overheated, but lower flow rate did not. WHY?!? :huh

This is what I want to know.


Like I said in the first post, a low flow rate is obvious. The coolant is moving the heat, but if the coolant isn't moving, then neither is the heat. Obvious.

But if the coolant is moving the heat, then how can you overheat if you're moving the heat too fast? I DON'T SEE HOW THAT'S POSSIBLE!


You're on the right track.

Basically the coolant is going through the radiator way too fast, it doesn't have time to dissipate the heat it has absorbed in the head and the water jackets of the cylinder. At the same time, it probably isn't even absorbing as much heat as it should in the cylinder and the head. If it moves too fast it can not absorb or dissipate as it is designed to do.

[yam450_53]

The thing is that you must let the coolant absorb all the heat as efficiently as it possibly can and then flow it back to the radiator to be cooled down, all that in a given amount of time. If you change the cycle lengths the temps change.

The best demonstration I can give you is, say you're done welding a plate of regular construction steel, for example that is 3/8" thick. Obviously the plate will be really hot after that. Depending on the length and size of the weld it will be, say around 1000 degrees, which we can both agree takes a while to cool down. Now here comes the part with the water (the coolant); you put the plate in the water for 1 second and take it out. It is still way too hot to touch bare-handed. Now you put it in again but this time 15 seconds, which is more than enough time to cool it down to an acceptable temperature. There is a spot between that 1 second and the 15 seconds that has the most efficient ratio of temp absorption:time. This means the water has X conductivity that you must allow the right amount of time to get the job done. The time will also depend on the conductivity of the material that the engine is made of (radiators are made of aluminum for a reason). This determines cycle length.

If you were to do this again but this time waving the plate in the water, it would cool down even faster because you are encouraging heat dissipation, so reciprocally as the coolant is moving throughout the engine it is also helping the heat escape from the cylinder walls and the head.

I hope that makes sense.

[Tommy 17]

Heres my take on it... This is from memory so I might not be exact..

But the transfer of the heat from the coolant to the radiator is done through conduction or contact. Therefore you can use the equation for conduction

{Delta(Q)/Delta(t)} = K*A*(Delta(T)/Delta(X))

Constants are:
Q=amount of heat flux or flow of energy per unit of area and time
t=time
K=materials heat absorbtion rate (given for the material)
A=surface area
X=length
T=temperature gradiant going in vs going out (Radiator vs Coolant)


So in normal person terms you can use that equation because it reads like this:

the amount of heat absorbed over an amount of time = the absorbtion rate times the surface area times the amount of temperature change over the distance

So how does this explain the high flow rate = over heat is because your amount of time or the DELTA t is very small. It is a direct multiplier of the heat flux or amount of heat conducted from the coolant into the radiator. So if your in the radiator for 0.5 seconds it actually makes the number smaller. You are taking Q times something less then 1 making it smaller. If you are in the radiator for 2 seconds you increase the number by 2 times. ETC...


Now of the case of the coolant moving to slow. The reason is the temperature will have an extremely long time to dissipate. Lets go with infinity. Now if it is at infinity eventually T1= T2 or the temperature of the fluid will equal the temperature of the radiator. This will cause DELTA(T) = 0 which causes the entire Q=0 or 0 heat transfer. Granted you can never have infinity so it will always be some absorbtion but just not as much as you'd like it to be. Reason is the temperature of the fluid basically equals the temperature of the radiator. So your engine isn't over heating, your radiator is over heating.


Hopefully this helped some. Its been a long time since I've done this stuff so I can be way off haha.:huh

[yam450_53]

Originally posted by Tommy 17
Heres my take on it... This is from memory so I might not be exact..

But the transfer of the heat from the coolant to the radiator is done through conduction or contact. Therefore you can use the equation for conduction

{Delta(Q)/Delta(t)} = K*A*(Delta(T)/Delta(X))

Constants are:
Q=amount of heat flux or flow of energy per unit of area and time
t=time
K=materials heat absorbtion rate (given for the material)
A=surface area
X=length
T=temperature gradiant going in vs going out (Radiator vs Coolant)


So in normal person terms you can use that equation because it reads like this:

the amount of heat absorbed over an amount of time = the absorbtion rate times the surface area times the amount of temperature change over the distance

So how does this explain the high flow rate = over heat is because your amount of time or the DELTA t is very small. It is a direct multiplier of the heat flux or amount of heat conducted from the coolant into the radiator. So if your in the radiator for 0.5 seconds it actually makes the number smaller. You are taking Q times something less then 1 making it smaller. If you are in the radiator for 2 seconds you increase the number by 2 times. ETC...


Now of the case of the coolant moving to slow. The reason is the temperature will have an extremely long time to dissipate. Lets go with infinity. Now if it is at infinity eventually T1= T2 or the temperature of the fluid will equal the temperature of the radiator. This will cause DELTA(T) = 0 which causes the entire Q=0 or 0 heat transfer. Granted you can never have infinity so it will always be some absorbtion but just not as much as you'd like it to be. Reason is the temperature of the fluid basically equals the temperature of the radiator. So your engine isn't over heating, your radiator is over heating.


Hopefully this helped some. Its been a long time since I've done this stuff so I can be way off haha.:huh

Don't really use that everyday eh?

[Ralph]

The only issue i see is that a higher flow rate results in more heat transfer.

Conduction is based on time, but for internal flow it comes down to flow rate. If you think about each unit of water cooling the radiator. Its either longer, greater exchanges of heat or more smaller ones. The time drops out. It then comes down to length, which is constant for us. Lets define the problem a little more.

I need to know the specifics of this situation. How can you confirm the increase in flow rate. This is at a constant RPM correct? Right now this problem is not completely defined...

[wilkin250r]

Originally posted by Ralph
How can you confirm the increase in flow rate. This is at a constant RPM correct? Right now this problem is not completely defined...

And I don't really have any actual numbers, I'm not measuring the flow rate directly.

But the basic laws of fluid dynamics are undeniable. If you put a restriction in the flow path, flow rate decreases. It has to. The actual amount will vary, obviously, based on the specifics of the restriction, but the flow rate will always decrease by some amount.

I just constantly hear stories like 400exrider707 posted. He was overheating until he put some type of restriction on the flow path, then everything was fine.

I'll be honest, I STILL don't see it, but I haven't had time to fully analyze Tommy's post.

[TNT]

Tommy does not account for fluid flow convection. Here I think this may closer to answering your question...

Convection is the transfer of heat from one place to another by the movement of fluids.[sf] The presence of bulk motion of the fluid enhances the heat transfer between the solid surface and the fluid.[2]
There are two types of convective heat transfer:
• Natural convection: when the fluid motion is caused by buoyancy forces that result from the density variations due to variations of temperature in the fluid. For example, in the absence of an external source, when the mass of the fluid is in contact with a hot surface, its molecules separate and scatter, causing the mass of fluid to become less dense. When this happens, the fluid is displaced vertically or horizontally while the cooler fluid gets denser and the fluid sinks. Thus the hotter volume transfers heat towards the cooler volume of that fluid.[3]
• Forced convection: when the fluid is forced to flow over the surface by external source such as fans and pumps, creating an artificially induced convection current.[4]
Internal and external flow can also classify convection. Internal flow occurs when the fluid is enclosed by a solid boundary such as a flow through a pipe. An external flow occurs when the fluid extends indefinitely without encountering a solid surface. Both of these convections, either natural or forced, can be internal or external because they are independent of each other.[citation needed]
The rate of convective heat transfer is given by:[5]

Q=hA(Ts-Tb)

A is the surface area of heat transfer. Ts is the surface temperature and Tb is the temperature of the fluid at bulk temperature. However, Tb varies with each situation and is the temperature of the fluid “far” away from the surface. h is the constant heat transfer coefficient that depends upon physical properties of the fluid such as temperature and the physical situation in which convection occurs. Therefore, the heat transfer coefficient must be derived or found experimentally for every system analyzed. Formulas and correlations are available in many references to calculate heat transfer coefficients for typical configurations and fluids. For laminar flows, the heat transfer coefficient is rather low compared to the turbulent flows; this is due to turbulent flows having a thinner stagnant fluid film layer on heat transfer surface

I think radiation is a factor on quads with recovery tanks close to the motor....

[Ralph]

Originally posted by wilkin250r
And I don't really have any actual numbers, I'm not measuring the flow rate directly.

But the basic laws of fluid dynamics are undeniable. If you put a restriction in the flow path, flow rate decreases. It has to. The actual amount will vary, obviously, based on the specifics of the restriction, but the flow rate will always decrease by some amount.

I just constantly hear stories like 400exrider707 posted. He was overheating until he put some type of restriction on the flow path, then everything was fine.

I'll be honest, I STILL don't see it, but I haven't had time to fully analyze Tommy's post.

I am just trying to fully understand the problem. So its say. Constant engine RPM. This is important because it means the engine is providing constant heat flux. So this limits it down to just the change in flow. Not how the system behaves with an increase in heat flux.

When talking about adding something into the flow, yes, it will be a resistance to the flow. Another important thing to consider is if the object is changing the flow. It could be creating turbulence or slowing the flow down to become Laminar (Non-Turbulent).

This could then change the way the fluid behaves as it enters the system. Fully Developed flow or non.

Ill keep poking around this as I get time to do so.

What specifically are we adding in line?

[Ralph]

Here's whats happening.

Lets define the components.

Engine: Can Provide a lot of energy. Possibly more than it needs to.

Radiator: Can remove LIMITED amount of energy. Restricted by airflow.

Increased Flow: INCREASES Heat Transfer

So, you unrestricted the flow and now the fluid is traveling faster. Coolant then passes through engine and picks up more heat than usual.

Coolant reaches radiator which is designed to only dissipate set amount of energy. Coolant leaves radiator at higher temp than normal. This cycles a couple of times and the coolant boils over.

You have to understand that you are cooling your engine to prevent it from overheating. What you are doing though is taking more energy out of the engine than you need to. Leaving you with a cooler engine for a short amount of time until the coolant boils over.

Case Closed.:chinese:

[wilkin250r]

Originally posted by Ralph
So, you unrestricted the flow and now the fluid is traveling faster. Coolant then passes through engine and picks up more heat than usual.

How does it pick up more heat than usual?

Let's say you double the flow rate. This means that every "unit" of coolant is now, on average, inside the engine for half the time. All other things being constant, half the time equates to half the energy transfer (per unit).

[TNT]

I have not been to college since 1984 and maybe I should stay out of this one but I thought the basic question is “I'm looking for an in-depth explanation on how coolant flow rates affect overall cooling efficiency.”….

To answer that question specifically would require an experimental test since there are many variables, pumps, piping, recovery tanks, thermostats that change turbulent and laminar flow rates…….

In general according to this formula,

Q= h A (Ts-Tb)

I posted above there is a direct relationship between surface area heat transfer and convective heat transfer Q…makes sense to me if you take an engine or radiator journal and decrease the journal diameter it has less surface area to cool, but flows faster, provided the pump CFM is constant.

There are additional factors to take into considerations as Tommy pointed out, convection heat transfer….. journal material e.g.: aluminum is more conductive than rubber so heat transfer will be better.

Radiation is another player here…

[Ralph]

Originally posted by wilkin250r
How does it pick up more heat than usual?

Let's say you double the flow rate. This means that every "unit" of coolant is now, on average, inside the engine for half the time. All other things being constant, half the time equates to half the energy transfer (per unit).

Because the two are not linear. Especially when considering turbulent flow, entry regions, thermal boundary layers.

Flow just does not behave as simple as you expect it to.

I believe it has a lot to do with how drag and friction works in pipes.

If you look at a moody diagram, look at the way friction changes with roughness and Reynolds number (a dimensionless measure describing flow which varies with velocity and diameter of pipe). note the logarithmic scale.

Things just are not linear. If this doesn't convince you, I am hoping the confusing visual will at least make me look smart :).
http://www.mathworks.com/matlabcentral/fx_files/7747/1/moody.png

[TNT]

Exactly! Friction....or surface area convection.

Way I read the Moody and correct me if I am wrong, is turbulent flow friction increases with slower velocity R#)…makes sense.

Laminar increase velocity friction decreases, just the opposite.

With increase friction, roughness, comes decrease heat transfer…

So what I gather is heat transfer depends on the type of flow, around a turbulent pump different not good than a smooth engine or radiator journal. Restrict these different types of flow depending on where in the system can have opposing effects.

[Ralph]

Originally posted by TNT
Exactly! Friction....or surface area convection.

Way I read the Moody and correct me if I am wrong, is turbulent flow friction increases with slower velocity R#)…makes sense.

Laminar increase velocity friction decreases, just the opposite.

With increase friction, roughness, comes decrease heat transfer…

So what I gather is heat transfer depends on the type of flow, around a turbulent pump different not good than a smooth engine or radiator journal. Restrict these different types of flow depending on where in the system can have opposing effects.

Turbulence has more to do with the velocity of the flow and pipe diameter.

And friction tends to level out with turbulent flow as opposed to laminar flow which keeps moving exponentially until the flow becomes turbulent. The reason the two don't blend together nicely is because its hard to predict when flow becomes turbulent. It kind of tends to happen in a range.

The overall meaning of all this is that its not linear. Increased flow rate leads to increased heat absorption. or efficiency if you will.

If you still aren't convinced maybe this will help. Even with convectional flow, you only ever transfer heat through conduction at the surface of your pipe.

When a flow is calm, the flow will develop a profile which tends to be the hottest at the pipe walls and cool toward the center. This means that the molecules toward the surface absorb the heat until they reach equilibrium. Then no more heat is absorbed.

But with increased mixing and turbulence, more molecules of coolant get to touch the walls and absorb energy. This way you are spreading the heat transfer among more molecules. So it is less likely that part of your flow is maxed out and not absorbing heat.

Now this is just another factor in the grand scheme of things.

http://www.engineering.leeds.ac.uk/ipse/yulong/images/research/p14_fig2.gif

[TNT]

That was a good explanation Ralph and it makes sense, thanks! It’s been too long can you tell what I’m looking at in the chart you posted now.

[wilkin250r]

Originally posted by TNT
With increase friction, roughness, comes decrease heat transfer…

Generally you get INCREASED heat transfer.

Since the heat transfer occurs mainly at the boundaries, the areas in contact, turbulent flow is constantly sweeping away hot coolant and bringing in new, fresh, cool coolant (on a small scale, just above microscopic scale).

If it was stationary, ZERO turbulence, you'd have very inefficient thermal transfer.

[wilkin250r]

Ralph, I agree that the flow patterns themselves aren't linear. (and I would like to see how velocity affects the boundary layer)

But the average flow rate HAS to be linear. I can't double the flow rate in the radiator, but not double the flow rate through the engine. That doesn't work.

[TNT]

Originally posted by wilkin250r
Generally you get INCREASED heat transfer.

Since the heat transfer occurs mainly at the boundaries, the areas in contact, turbulent flow is constantly sweeping away hot coolant and bringing in new, fresh, cool coolant (on a small scale, just above microscopic scale).

If it was stationary, ZERO turbulence, you'd have very inefficient thermal transfer.

Ralph seems right turbulant flow levels out and the Moody supports that, look at that zone....Laminar a straight expoed line.....then the transition area as he described and they don't blend well hence two different graphs.

[Ralph]

Originally posted by wilkin250r
Ralph, I agree that the flow patterns themselves aren't linear. (and I would like to see how velocity affects the boundary layer)

But the average flow rate HAS to be linear. I can't double the flow rate in the radiator, but not double the flow rate through the engine. That doesn't work.

You are correct. They both increase evenly. However, the radiator is still limited by the air flow across it.

So the problem is not the radiator accepting the heat. Its the radiator transferring that same heat to the air. So that extra heat is trapped in the radiator. And there's a point when the radiator temp is no longer cooler than the coolant. Shortly after your coolant will boil over.

[TNT]

Originally posted by Ralph
You are correct. They both increase evenly. However, the radiator is still limited by the air flow across it.

So the problem is not the radiator accepting the heat. Its the radiator transferring that same heat to the air. So that extra heat is trapped in the radiator. And there's a point when the radiator temp is no longer cooler than the coolant. Shortly after your coolant will boil over.

Not sure I agree with ya there...today I cleaned my AL water jug with hot tap water, reminded me how conductive aluminum is after one fill about 10 secs I could not hold the container in my hand....there cant be a lot of loss/time that the aluminum raidiator rids conducts itself of heat...ANd if you add airflow to the equations I say seconds, it's not storing heat long.

Old man TNT keeping up with the kids/MODS, gunna tell all my senior citizen freinds I still got it....can't wait to you and WILKINs are my age....hahaha! :D :confused:

[Ralph]

Originally posted by TNT
Not sure I agree with ya there...today I cleaned my AL water jug with hot tap water, reminded me how conductive aluminum is after one fill about 10 secs I could not hold the container in my hand....there cant be a lot of loss/time that the aluminum raidiator rids conducts itself of heat...ANd if you add airflow to the equations I say seconds, it's not storing heat long.

Old man TNT keeping up with the kids/MODS, gunna tell all my senior citizen freinds I still got it....can't wait to your and WILKINs are my age....hahaha! :D :confused:

The aluminum is not the issue. Its getting the heat to the air. Air is a poor conductor. This is why they pack fins around the radiator. Increase in surface area increases heat transfer on the air side of things.

It is completely independent of the flow rate internally though. Which means that the radiator can only handle what it was designed for. Air around the radiator is the limiting factor.

[Ralph]

Originally posted by TNT

Old man TNT keeping up with the kids/MODS, gunna tell all my senior citizen freinds I still got it....can't wait to you and WILKINs are my age....hahaha! :D :confused:

I have pretty poor long term memory so I don't anticipate this stuff staying in too long. You're probably doing way better than I will...

[TNT]

Originally posted by Ralph
I have pretty poor long term memory so I don't anticipate this stuff staying in too long. You're probably doing way better than I will...

Yeah man it's all good your training your mind, who knows where you'll end up using this stuff or not....I seen the post where the person ask if you use this stuff, well YES you will for the rest of your life in everything you do in life you will out think most because of your training/mind just be carefull of the wife let her thnk she's smarter lol....trust me my younger managers at a FORTUNE 500 company can't even manage me , I tell them what I will do for them like or not most of the time I make my salary 10 fold......so keep it up, be smart you guys as you are you will reep the bene's in many ways, just not ENG salary but life. Good Luck man you an wilk worn this old man out, TNT over and out.:D

[Ralph]

Yeah. Right now I am working in a field dealing more with hydraulics, fluids, and strength of materials. Not much heat transfer.

I am also doing some design incorporating the above. A great all around job with lots of things to learn.

[wilkin250r]

Originally posted by Ralph
It is completely independent of the flow rate internally though.

But that's the whole point of the thread. Somehow, it's NOT independent, and I don't understand why?!?

Somehow, someway that I don't understand, the internal flow rate of the coolant in the radiator and engine is linked to it's efficiency. Same exact system, same exact operating parameters (as far as I know), the only difference is the flow rate of the coolant.

High flow rate =>overheat

Restricted flow rate => normal operation, normal temps.

I don't get it, I don't see it. If it was an extremely SLOW flow rate, I would understand. But I don't see a fast rate being a problem.

[wilkin250r]

Every time I see an example of this problem, the explanation that's always given is "the coolant isn't in the radiator long enough to cool down" and I DON'T BUY IT. That's not it. It CAN'T be it, the laws of thermodynamics don't allow for that (as far as I understand them).

So there's something I'm missing, Somehow, the internal flow rate is related to thermal transfer efficiency.


Let's take an absurd example, but dealing with the same issues. I'll take a piece of hot aluminum, and dunk it in water, BIG water, like a lake. The aluminum will cool off, according to the temperature difference, the surface area, ect ect.

But somehow, if I MOVE that piece of aluminum real fast underwater, it will stay hot?!?!? It won't cool off?!?

[Ralph]

You really are missing it...

It is not that the plate will stay hot. Its the opposite.

You are focusing on the wrong end of the problem. Focus on the engine.

The increased flow is taking more heat out of the engine. So the coolant leaves the engine at a higher temp. MORE heat transfer.

The radiator can only dissipate a limited amount of heat. Thus the extra heat being taken out of the engine has no where to go. The heat transfer coefficient is limited to the flow rate of the AIR traveling across the fins...

It is like the cooling system is now biting off more than it can chew. More than it needs to take out of the engine...

--------

You are hung up on the radiator. There is no problem getting the heat into the aluminum. It is getting the heat from the aluminum to the air. This is the limited factor, bottleneck.

[wilkin250r]

Originally posted by Ralph
The increased flow is taking more heat out of the engine. So the coolant leaves the engine at a higher temp. MORE heat transfer.


I will be honest, I never considered that aspect at all.

Let me ponder that one for a little while...

[Ralph]

Originally posted by wilkin250r
I will be honest, I never considered that aspect at all.

Let me ponder that one for a little while...

Ponder away. :) I think its time to put away the computer away. 9:41 on a Friday night and I am discussing heat transfer on the internet... It's the danger of engineering.

[Ralph]

Reminds me of you...

[Fred55]

Also, putting something in the hoses to reduce flow will also increase turbulence which can increase the rate of heat transfer.

[wilkin250r]

Originally posted by Fred55
Also, putting something in the hoses to reduce flow will also increase turbulence which can increase the rate of heat transfer.

Naw, I can't see that. That would hold true if it was one long, solid tube, but not for an engine. Too many bends and turns and whatever in the system already, it's plenty turbulent enough on it's own.

If your restriction is in the bottom hose, the next step is the water pump. Who cares what your flow pattern is up until then, the water pump is going to throw a massive amount of turbulence into it.

If your restriction is in the upper hose, the next step is a 90 degree turn as it hits the radiator to drop into the radiator cooling passages. Even if you had a perfect laminar flow through the hose until then, the drop into the radiator will throw it all out of whack.

[Fred55]

I guess I should have been more clear, but it all depends on where the restriction is placed, turbulent flow is better for heat transfer compared to laminar.

[TNT]

Fred look @ the moody graphs ralph posted seems more true for smooth pipes with low R #, the friction really spikes....with all that friction @ the boundry layer you'd think you generate heat if anything.....but as the guy's pointed out the molecules are bouncing around so much from the core coolant where it's cool to the boundry layer hotter, so Ralphs argument is that creates a balancing act in theory you can see in higher R# and roughness, if you go by the graphs which imo are a indicator in part that depends on velocty, pipe smoothness and diameter...

I say it's all theory proove would be in a test or a CFD model, placing different configurations in different locations noting results.

Other thing I'd like to see if the effect of a venturi both laminar and turbulant of different diameters, since it changes velocity and pressure.

I'm not sure I can agree that pulling coolant faster out of the motor generates more overall efficency, or air(push/pull fans), louvers design right are poor conductor of heat since there are other factors, barometric pressure, relative humidy, to name a few, that need to be considered, in most systems isolation and lack of a tested configuration gets you in thoery trouble. Take a head flowed on a bench for example, connect it to a intake trake and exhaust tract changes everything......:eek2: :huh

I do think we are closer to answer the OP, just not without specific test to proove all this theory. One could create a test w/ inline temp gages on thier quad, putting laminar/turbulant restrictions(hoses/baffles) of different diameters/roughness in and see. Put the gages in/out of motor and radiator.

You can tell been doing a little pondering of my own. :D

[TNT]

Originally posted by Ralph
Because the two are not linear. Especially when considering turbulent flow, entry regions, thermal boundary layers.

Flow just does not behave as simple as you expect it to.

I believe it has a lot to do with how drag and friction works in pipes.

If you look at a moody diagram, look at the way friction changes with roughness and Reynolds number (a dimensionless measure describing flow which varies with velocity and diameter of pipe). note the logarithmic scale.

Things just are not linear. If this doesn't convince you, I am hoping the confusing visual will at least make me look smart :).
http://www.mathworks.com/matlabcentral/fx_files/7747/1/moody.png

bump

[Ralph]

I think we have gone way too far.

You can read in a textbook that increased mass flow rate increases heat transfer.

There is not much more up for discussion here...

We are getting above the scope of the problem. And even above the scope of a standard engineering education.

[KevinAb]

Originally posted by Ralph
I think we have gone way too far.

You can read in a textbook that increased mass flow rate increases heat transfer.

There is not much more up for discussion here...

We are getting above the scope of the problem. And even above the scope of a standard engineering education.

Good stuff though.

To keep it simple- my first day in Thermo class years ago the teacher explained the basics this way- set a glass of cold water in a warmer room. The bigger and colder the glass of water is, the longer it will take for the water to reach room temp. The warmer the room is, the shorter the time will be.

You can start throwing all kind of different variables into the equations, but it really goes back to this simple principle. It's really common sense stuff.

Thus a bigger radiator will dissipate more heat.

[mxduner]

Originally posted by KevinAb
Good stuff though.

To keep it simple- my first day in Thermo class years ago the teacher explained the basics this way- set a glass of cold water in a warmer room. The bigger and colder the glass of water is, the longer it will take for the water to reach room temp. The warmer the room is, the shorter the time will be.

You can start throwing all kind of different variables into the equations, but it really goes back to this simple principle. It's really common sense stuff.

Thus a bigger radiator will dissipate more heat. glad you brought that up.

I really am terrible at explaining things in forums so i apologize in advance

1 lb. of water in a liquid state takes 1 btu to raise it 1*f. also the obvious, + temp differences= + btu change, like your water in the class room example.

Also maybe pointed out earlier, and may not mater, but heat travels from hot to cold, as temperature is a measure of heats intensity.

Hope i did'nt waste poste space :D it's been to long...