Posted: Thu Aug 25, 2005 1:32 am
Just outta curiousity douglas, I remember you having problems with the belt slipping with the supercharger. Do you think you could adapt a belt with teeth.
A forum for 89-94 BC-BF(BJ) Legacy Owners and Fans
https://bbs.legacycentral.org/
Not necessarily.....it all depends on how much fluid flow (ie pump flow rate) you have going through the core, and how much acceptable pressure drop you're willing to accept.vrg3 wrote:napphappy - Is it necessarily bad that the water would have to go through narrow restrictive passages?
I agree; that makes sense to me.Too small and the fluid's not staying in the radiator long enough to achieve the maximum cooling that the atmosphere can do......
I don't understand why this is the case; can you explain a little more?or if stays in there too long it's not really gaining you anything, and you probably spent a lot more money for cooler that is essentually a pipe to flow water.
So what you're saying is that more water will be in contact with the metal tubing on the smaller cross-sect area compared to the larger cross-sect area when the flow rate is the same? If so, I'm not so sure because if the flow rates are the same the fluid will stay in the larger cross-sectional tube longer, therefore negating the effects of more fluid surface contact area.I figure that if you have a smaller cross-sectional area with the same flow rate, then you have greater surface area between the radiator and the water. That ought to allow greater heat transfer.
Too small and the fluid's not staying in the radiator long enough to achieve the maximum cooling that the atmosphere can do......
Ok let's say your ambient temp is 100 deg F. Your coolant temp entering the radiator is 130 deg F. Your cooler/radiator has a 10 degree approach temperature. (Meaning that the best the cooler can do is get the fluid within 10 degrees of the ambient. So the best you're going to cool that cooling water is 110 deg F. When ambient temps drop, so will the coolant water temp)I don't understand why this is the case; can you explain a little more?
Not quite.....it's delicate balancing act. Typically the radiator/cooler portion of the cooling system is sized larger then the heat source. This allows for more consistent operation over the life cylce of the cooler due to fowling of the cooler as it ages. If it was sized for exactly the amount of heat generated, as the cooler ages, it won't be as efficient and cooler outlet temps will rise. Also, it adds in a safety margin.It seems to me that -- as a very general and not too serious rule -- you want the water to spend very little time in the top-mount heat exchanger and a lot of time in the front-mount heat exchanger. Of course, these two goals are a little contradictory because it's the same water and the same pump.
Actually, your efficiency drops with closed loop non-direct contact systems like that. Remember from above I mentioned the 10 deg approach temp for the radiator? The heat exchanger that would have the intercooler water and radiator water also has an approach temp of 5-10 deg. So for a total system approach temperature, you now have 15-20 deg. So the actual intercooler water temperature will be hotter compared to if you just ran it directly through the radiator.If my logic thus far is correct, maybe there's another option that can help separate the variables: two water circuits! Use one lower-speed pump to move water from the reservoir, through the radiator, and back into the reservoir. Then have another pump that can push water very fast through the intercooler. That would allow as much heat transfer as possible to happen at the radiator, while preventing the water that's passing through the intercooler from heat soaking.
Yes, that is what I'm saying. It seems to me that it wouldn't quite negate the effect, though, because heat transfer ought to be roughly linearly proportional to surface area (i.e. doubling surface area should double transfer) but sublinearly proportional to time spent in the exchanger (doubling the time spent in the exchanger won't quite double heat transferred).Legacy777 wrote:So what you're saying is that more water will be in contact with the metal tubing on the smaller cross-sect area compared to the larger cross-sect area when the flow rate is the same? If so, I'm not so sure because if the flow rates are the same the fluid will stay in the larger cross-sectional tube longer, therefore negating the effects of more fluid surface contact area.
Yeah, I could believe that.Personally I think it's a wash as long as you're not comparing a 1/4" tube and a 2" tube.
Huh. I didn't know there was a such thing as an approach temperature.Ok let's say your ambient temp is 100 deg F. Your coolant temp entering the radiator is 130 deg F. Your cooler/radiator has a 10 degree approach temperature. (Meaning that the best the cooler can do is get the fluid within 10 degrees of the ambient. So the best you're going to cool that cooling water is 110 deg F. When ambient temps drop, so will the coolant water temp)
I see.With that in mind, and using the same design/operating/approach temps, if you get a cooler that's very large, and the water cools to 110 deg F half way through the radiator, then for the rest of the cooler, it's not really doing any additional cooling.
Right; the "extra" portion of the cooler is basically reserve capacity.However, with such a large cooler, if you increase the cooler inlet temp, you will still more then likely hit your 110 deg F cooler outlet temp. Also, if ambient temps drop pretty low, you're going to be able to cool that coolant more compared to a smaller cooler.
I guess so. As always, however, questions' answers beget further questions. Can you explain the phenomenon of approach temperature?Make sense?
So the approach temperature is calculated/measured/whatever assuming some very low fluid flow velocity?If the fluid travels too fast through the cooler, the cooler doesn't have time to cool to the best possible temp that it can do,
But once you change the cross-sectional area or tube length you're no longer talking about the same cooler with the same maximum cooling potential, right?To get the max cooling potention out of the cooler WITHOUT altering flow rates (I get into why I mentioned this below) you have two options, slow the velocities....only way to do that without altering flow rates is to make a bigger cross-sec area.....or make the cooler tubes longer.
What do you mean by "sized larger?" You mean it is designed to be able to reject more heat than the heat source is expected to inject?Not quite.....it's delicate balancing act. Typically the radiator/cooler portion of the cooling system is sized larger then the heat source. This allows for more consistent operation over the life cylce of the cooler due to fowling of the cooler as it ages. If it was sized for exactly the amount of heat generated, as the cooler ages, it won't be as efficient and cooler outlet temps will rise. Also, it adds in a safety margin.It seems to me that -- as a very general and not too serious rule -- you want the water to spend very little time in the top-mount heat exchanger and a lot of time in the front-mount heat exchanger. Of course, these two goals are a little contradictory because it's the same water and the same pump.
Sounds like common sense... I would expect that the engineers at FHI that designed the rest of the engine would have done the same with the engineers that designed the radiator, for example.We ask our vendors to add an additional 10% of cooling capacity to the coolers we purchase.....and typically we round up when we provide the vendor the heat duty the cooler needs to dissipate.
I wasn't talking about having a third heat exchanger... it would all be the same fluid. Like two pumps at the bottom of one reservoir. I was just thinking of a way to size the intercooler pump separately from sizing the radiator pump.Actually, your efficiency drops with closed loop non-direct contact systems like that. Remember from above I mentioned the 10 deg approach temp for the radiator? The heat exchanger that would have the intercooler water and radiator water also has an approach temp of 5-10 deg. So for a total system approach temperature, you now have 15-20 deg. So the actual intercooler water temperature will be hotter compared to if you just ran it directly through the radiator.
Yeah... the rated cooling capacity of the top-mount intercooler is 3954 kilowatts, and the rated cooling capacity of the front-mount radiator is 5408 kilowatts (hey, they did!), and pump's rated flow rate is 15 liters per minute at 12 volts with a 2-meter head. They don't really specify anything about the pressure drops across either heat exchanger or about what the pump's flow curve looks like at different pressure.Unfortunately, there's no way I'm going to get all the engineering data I'd like to properly size the radiator. I do have original heat loads from the factory manuals for the AWIC system. I also have pump flow rates.
Well, it seems like it won't ever hurt anything (except for potentially making the pump work harder)... it'll just provide more heat rejection and provide for a larger quantity of water in the system, which means more thermal mass.Basically, what I'd like to do is get as large of radiator as I can. It may not be beneficial at all times due to what I mentioned above, but when ambient temps are low or when more heat is generated, the system will be able to handle it.
In this case you're saying Q is the heat rejection capacity of the radiator times time?Using the equation:
Q = mCp(T2-T1)
note: m reads m dot (or mass flow rate)
That formula can be used with volumes as well. But the general purpose of what I'm trying to show is that with an increase in m dot or mass flow rate, you can achieve lower T2's with the same Q, or you can absorb more Q with the same T2.
Hmm... Yeah, I wonder if we can get figures on how much the stock pump (or any other pump) heats up the water it moves.Just like everything, there's a balance between going too big. The pump puts work/energy into the fluid. If you get a massive pump, it's going to be putting more work/energy into the fluid then is gained by the increase flow rate through the radiator.
Yeah... I think pretty much all air/water intercooling setups (even factory ones) are more or less just thrown together.So unfortunately, there's no real fine answer to how big a cooler do I need, and how big a pump do I need.
That's how I likes it. =)I know this is rather lengthy and in depth
I should not have said negate the effect.....They'll be "close" but that can all depend on what you, myself, or someone else defines as "close"Yes, that is what I'm saying. It seems to me that it wouldn't quite negate the effect, though, because heat transfer ought to be roughly linearly proportional to surface area (i.e. doubling surface area should double transfer) but sublinearly proportional to time spent in the exchanger (doubling the time spent in the exchanger won't quite double heat transferred).
That last part's right, isn't it? As the temperature differential decreases, so does the rate of heat transfer, right?
I guess I should expand on approach temperature.Huh. I didn't know there was a such thing as an approach temperature.
What prevents the cooler from equalizing temperatures inside and outside the heat exchanger?
Approach temperature is basically set by whomever is designing the cooler. They can make it whatever they want, but as I mentioned, as approach temperature decreases, the size of the cooler will increase.So the approach temperature is calculated/measured/whatever assuming some very low fluid flow velocity?
CorrectBut once you change the cross-sectional area or tube length you're no longer talking about the same cooler with the same maximum cooling potential, right?
Correct again.What do you mean by "sized larger?" You mean it is designed to be able to reject more heat than the heat source is expected to inject?
I guess the balancing thing I was referring to was the sizing of the radiator to the heat source. So I guess I didn't answer your question directed about the amount of time the fluid spends in the heat source & radiator.Cuz, yeah, that all makes sense to me. But I don't see how that directly relates to my statement. What do you need to balance?
Ahh....I'm not sure exactly what that would give you....maybe more control over the system to gain some efficiency, especially in a retrofit environment where you couldn't resize a cooler. But if the system was sized properly from the beginning, I wouldn't think it'd be too necessary.I wasn't talking about having a third heat exchanger... it would all be the same fluid. Like two pumps at the bottom of one reservoir. I was just thinking of a way to size the intercooler pump separately from sizing the radiator pump.
Yes, heat rateIn this case you're saying Q is the heat rejection capacity of the radiator times time?
Actually, centrifugal pumps are very similar to turbos. If you have a smaller compressor wheel, and you try to move more flow through it, you have to increase pressure, which increases the temp. So with a pump, we could get a "comparitive" increase in the amount of energy inputted into the water by measuring pressure drop across the pump. The larger the pressure drop, the more energy inputted into the water.Hmm... Yeah, I wonder if we can get figures on how much the stock pump (or any other pump) heats up the water it moves.
Cool, yeah, I'm with you.Legacy777 wrote:I should not have said negate the effect.....They'll be "close" but that can all depend on what you, myself, or someone else defines as "close"
Yes, this is what I said, assuming a volumetric fixed flow rate.Also, for my clarification, in your previous post, you say that the smaller tubed radiator (aka smaller cross-sectional area) should allow for more heat transfer compared to a radiator that has larger tubes (more cross-sect area). Did I get this right, or am I mis-understanding what you originally said?
Yes, I was indeed saying increasing surface area would increase heat transfer. That's my reasoning behind the guess that a smaller-tubed radiator would transfer more heat. It's only at the periphery of the cross section that heat actually is exchanged, right?The reason I ask, is in the post right above, and what I quoted, you seem to say that increasing surface area, increases heat transfer. I guess I'm just a little confused on exactly what you're asking.
Okay, right. So that means that if you leave the two materials in contact for twice as long, you don't transfer twice as much heat. Right? But if you leave the two materials in contact with twice as much contact area, it seems like you ought to get nearly twice as much transferred heat.As for that last comment, yeah, you're correct. The heat transfer rate is what changes. The heat transfer rate will be greatest when there is a large temperature difference. As the temperature difference decreases, so will the rate of heat transfer.
Thank you for the explanation, Josh.I guess I should expand on approach temperature.
...or a much lower volumetric flow rate, right?Given enough time, the temperatures will come very close to equalizing. However like I mentioned, to do that, you need a much larger cooler so the fluid stays in there longer.
I guess that's kind of a problematic point with a lot of automotive radiators and heat exchangers... Nobody's actually trying to get their ATF down to anywhere near ambient temperature. Ditto with motor oil and engine coolant. So repurposing them for AWICs is a little weird.From what I've read.... 20-25 deg approach temps are common with air to water heat exchanger.
Your explanation makes sense to me except for the clarifications I'm asking about above.Does that explain approach temperature a little better? Or are there parts I didn't explain well enough.
Let's clarify what "it" is though. I read that sentence as:I don't think that you really want the water to spend as little time as possible in the heat source, and as much time in the radiator. The shorter the amount of time it spends in the heat source, the less heat it's taking away.
Right, I see and agree with what you're saying here. I guess I was thinking of categorically favoring the goal of a high temperature differential. I'm having some trouble wrapping my head around why that's not a good idea, though... My experience with Upstate NY's winter wind chills still makes me believe I'm right.Again, this all can be related back to the fact that the heat transfer rate decreases as temperature difference decreases. Let's say the fluid can reduce the air temp by 20 deg F by remaining in the cooler for 5 seconds. If we increase the time to 10 seconds we only decrease the air temp to 22 or 23 deg F. So it's a law of diminishing returns.....and maybe that was what I was getting at about the balancing thing. It really depends on your design conditions/temperatures. If you have high temperature differences, leave it in there longer and get more heat out of the system. If your temperature differences aren't that great, you'd probably be better off increasing the flow rate so you keep your temperature differences greater.
I'm saying it would allow you to have coolant flow very very fast through the intercooler and more slowly through the radiator, which I still think would yield benefits.Ahh....I'm not sure exactly what that would give you....maybe more control over the system to gain some efficiency, especially in a retrofit environment where you couldn't resize a cooler. But if the system was sized properly from the beginning, I wouldn't think it'd be too necessary.I wasn't talking about having a third heat exchanger... it would all be the same fluid. Like two pumps at the bottom of one reservoir. I was just thinking of a way to size the intercooler pump separately from sizing the radiator pump.
Is that really true? I can see why it's true for turbos and other centrifugal gas compressors, but a centrifugal liquid pump isn't compressing the fluid. Isn't the pressure determined pretty much entirely by the flow restriction downstream of the pump? I'm thinking of a fuel pump as I write this.Actually, centrifugal pumps are very similar to turbos. If you have a smaller compressor wheel, and you try to move more flow through it, you have to increase pressure, which increases the temp. So with a pump, we could get a "comparitive" increase in the amount of energy inputted into the water by measuring pressure drop across the pump. The larger the pressure drop, the more energy inputted into the water.Hmm... Yeah, I wonder if we can get figures on how much the stock pump (or any other pump) heats up the water it moves.
Heheheh... Thanks for indulging me in this discussion.Ok....that's enough for tonight. I need some veg. time before bed....haha
I hear ya, -K-. I didn't put very much thought at all into my two AWIC setups so far and they both do work quite well. It's still good to try to understand the reasoning behind stuff, though. One can always do better.-K- wrote:I enjoy the amount of thought that is going into this but I don't think it's that big of a deal to make it perfect.