Air Water Liquid Intercooler for LGT

[NSFW]

Why think outside the box when you can run with the herd?

[praedet]

^^^Thinking outside the box is not always the best idea. Believe me, I am one for doing it when there is a benefit, but, sometimes there is a reason everyone is in the same box

[ClimberDHexMods]

As a side note, this system would need to come with the option of an ice box reservoir to really make sense to some buyers. Just a side note for all the drag racers wanting charge air temps colder than ambient, if only for several minutes, since CO2 and water sprayers don't do much for an air-to-water intercooler.

 

So: My dad was a physics major once upon a time, so we sat down to run the numbers. The intercooler core design is really what will make or break the cooling efficiency of the system. Everything else is ready to order. The key is that the incoming ambient temp water would need to flow the opposite direction of the charged air to hit the highest efficiency, but most of the cores out there run water perpendicular to the air flow, not parallel. I'm going to call some guys on Monday to see what price I can get on a custom core. Otherwise there's a big core readily available that might work just as well, though bigger costs more for the same result.

[NSFW]

^^^Thinking outside the box is not always the best idea. Believe me, I am one for doing it when there is a benefit, but, sometimes there is a reason everyone is in the same box

 

 

True. And sometimes the only reasons are herd mentality, laziness, and/or lack of creativity. For AWICs in Subarus I'm not willing to place a bet for or against, but I look forward to seeing someone take a good shot at it.

[bugblatterbeast]

The key is that the incoming ambient temp water would need to flow the opposite direction of the charged air to hit the highest efficiency, but most of the cores out there run water perpendicular to the air flow, not parallel.

 

reverse flow heat exchangers do result in the biggest temperature drop of the air but require the greatest amount of space due to the decreasing temperature difference between the two fluids along the length of the exchanger. on the other hand, if you run the fluids in the same direction, you get a large quantity of heat transfered (per unit length) but a higher outlet temperature. the perpendicular flow design is a compromise between the two. many commercial designs are constrained by space

 

you can find the equations and pre tabulated results you want in most first year texts on mass transport

[ClimberDHexMods]

I would figure it would be easier to find out about these

http://us1.webpublications.com.au/static/images/articles/i1087/108702_3mg.jpg

[mwiener2]

We already know about those

[GTkansan]

Bugbutter has it backward. Reverseflow (or counterflow) has a more uniform temperature differential between the two working fluids. The use of a parallel flow arrangement has a decreasing temperature differential along its length.

 

A cross flow heat exchanger is not really a mix of the two. The use of this arrangement is normally done when the thermal capacity RATE of the two fluids is DRAMATICALLY different. I.e. water to air, or in a case where the temperature differential (pre and post exchanger) of one fluid is constrained relative to the other. The reason our air to air intercoolers are in this fashion, is that the temperature differential on the outside air side is much different than that on the interior of the cooler. You move dramatically more air on the outer flow side than on the inside.

[GTkansan]

Climber D - Because the thermal capacity rate ( Cp_water*massflowrate_water ) is much larger than that for the air, the temperature of the water will not be so dramatically changed as the air side. This means a cross flow is still perfectly acceptable. I really think we can get away with a few finned cross flow tubes. Ive been working on modeling the exchanger for the last few hours and think it can be done quite easily. I've got a simulation running to test out a few flow and fin geometries and will get a representative solution.

 

Once I get it worked out (my PhD research is being nuts right now and absorbing my time), Ill have the cnc lab mock up a few tubes to test out the flow configuration.

[bugblatterbeast]

Bugbutter has it backward. Reverseflow (or counterflow) has a more uniform temperature differential between the two working fluids. The use of a parallel flow arrangement has a decreasing temperature differential along its length.

 

the above reason is exactly why it isn't baskwards (unless we are talking about different things). I'm assuming the coolant flow is low enough that it has significant temperature rise as it flows through the heat exchanger. If it didn't, we'd be stuck with the problem of getting the coolant to heat exchange at a reasonable rate with the ambient air. if we assume really high coolant flows and tank capacities, then the rest of the system will be so heavy that a FMIC makes more sense.

 

with the coolant flowing in the same direction as the air, you get the maximum possible temperature difference between coolant and air at any given point of the exchanger. while this allows the maximum heat transfer per unit length, it doesn't allow the maximum temperature drop in the air because the coolant at the exit of the exchanger has been heated above the coolant inlet temps. on the other hand, reverse flow allows the air to be cooled exposure to "virgin" coolant. in this case it is possible for the outlet air temps to be the same as the coolant inlet temps but this comes at a price. as you mentioned, the reverse flow setup has decreasing temperature differences between the coolant and air along the length of the exchanger. this means we need an infinitely long heat exchanger to get the air exit temps the same as the coolant inlet temps. in practice, one just picks a reasonable length or changes the flow direction.

 

again, the above assumes that the coolant temp rises significantly as it passes through the exchanger, which is a practical assumption if we are later faced with the task of dumping the heat to ambient.

[bugblatterbeast]

Ive been working on modeling the exchanger for the last few hours and think it can be done quite easily. I've got a simulation running to test out a few flow and fin geometries and will get a representative solution.

 

are you using Ansys?

[GTkansan]

A combination of ansys and NX nastran for the solid modeling and fluent and fluidsys for the thermal analysis.

[bugblatterbeast]

I'm jealous. Neither school or work will pony up for a full set of licenses.

 

I'm really glad to see that somebody has access to the tools and is running the simulations. Things get funny when you need to impose real world constraints on a problem and prototyping to test is bloody expensive. Best of luck with school.... I know the feeling....eats up all the car modding time.

[ClimberDHexMods]

I finished most of my calculations (done with the old man (physics major, not an engineer)). We figured high capacity (duh): If running at WOT ~6000rpms & 20psi (just for the sake of the calculations), and the exchanger is to drop the hot air charge to perfect ambient (impossible but again these are just to get some numbers to work with), the water would pickup ~33.6* F. We figured the turbo produces an increase in heat of around 1.56 x 10E6 J/min, though every turbo and all associated mods, tuning, heatsoak and weather/climate factors can of course change this. Designing the exchanger core for very high efficiency is beyond me, but engineers are abundant here or back home. Hopefully the numbers will be comprehensive enough to justify assuming the risk of building this. And at the end of it all, I don't know engineering, but I do know business, and all this is just a bunch of guys having some fun until it is installed and works.

[ClimberDHexMods]

Assuming for a moment that it works well, the $$$ costs of making it perfect may be the undoing of this.

 

Problems and Solutions:

1) Heatsoak from stopping at red light after WOT, bumper to bumper Florida traffic or silimar.

Solution: Radiator Fans on an adjustable thermostat.

2) Heatsoak from driving then turning off car, letting it sit long enough to bake the engine bay, turning it back on.

Solution: Wire an automated timer switch to automatically activate the pump on a time interval to move water into front IC radiator where heat soak is less of a problem and where more water volume will be stored anyways. OR just let it rest because it'll cool back down, just not as quick as an air-air unit can. Liquid IC technology may not be the best choice for a performance-minded person who run lots of errands .

 

Additional: A Pump failure warning light add-on and a quick-change backup will come at a higher initial purchase price. I also like the idea of a pump controller, which would allow a user to run it slower when commuting, for instance, if it would help the pump last longer.

 

But I am not a pump expert...

http://www.softduit.com/images/Blog_Folders/article_images/BloggingAboutSomethingNegativeNotMyBagBa_120CE/austin_powers_Swedish_Penis_Pump_thumb.jpg

 

Also an air-temp gauge would be nice, if you have any room left.

[NSFW]

Would it be practical to let the AWIC drain to the radiator (+ a small tank) while the car is off? At least then the water would be next to the engine, instead of sitting over it and right next to the turbo. (I'm assuming the AWIC core would sit where the TMIC core goes.) You'd need an extra valve or two to drain the reservoir and refill the AWIC when the engine starts again, but it should help with the 2nd heat soak scenario.

[rao]

So wait, now this won't work?

[Rika]

No it works.

 

Solutions:You need to plot out every place that sell dry ice on your commute to work.

[GTurabus]

Two cents time:

 

Adding (too much) circuitry adds complexity and increases the failure modes available. Assuming this device will be used by enthusiasts, and not by the subaru buying public at large, perhaps some good old fashioned engineering could work.

 

For instance, lets just say that there is adequate water pressure available to mount the reservoir (or additional back up reservoir) remotely. This can reduce some of the heat soak issues from the engine bay, but not necessarily those from the car being at idle.

 

One possible solution would be to add a bypass valve from the "main" reservoir, to a cooler, "remote" reservoir. When the temperature gauge (complex mechanical device likel to fail) of the main reservoir reaches a certain temperature (insert calulations here), the valve could be manually flipped to allow the temperatures to come down. I think back to one of my friends old 1960 VW Type 1 convertible with a "kick tank". When the gas gauge hits E (or when you run out of gas) you manually (with your foot) kick the lever down to open the reserve tank. Running the lines in the drive tunnel would naturally allow you to place the lever nearby, outside like the handbrake, or even inside the center console.

 

Before each and everyone interested in this thread tears this apart, please submit your alternate proposal ;-)

 

 

 

Now for the mad scientists among you, imagine the trunk filled with a big insulated storage container filled with enough dry ice to cool the remote reservoir for the length of an autocross event. You would of course then run the system full open to achieve maximum effect. Now this system works both for the stop light enthusiast and daily driver to those that go to the track.

 

If you would like me to actually design any part of this it will of course be for my usual fee. One million dollars.

 

 

 

Imagination: 0$

 

Foolish thinking: 0$

 

Over eight years experience in research and development and prototype design of missile guidance and control systems: Not quite priceless

[rao]

Before each and everyone interested in this thread tears this apart, please submit your alternate proposal ;-)

 

Air to air intercooler with 35% alcohol/Water injection. In operation for 4+ years, no issues.

 

 

Next.

[NSFW]

When the temperature gauge (complex mechanical device likel to fail) of the main reservoir reaches a certain temperature (insert calulations here), the valve could be manually flipped to allow the temperatures to come down.

 

How different is that from the thermostat (simple mechanical device, rarely fails) that keeps the car's water jacket from getting too hot?

[biz77]

Assuming for a moment that it works well, the $$$ costs of making it perfect may be the undoing of this.

 

Problems and Solutions:

1) Heatsoak from stopping at red light after WOT, bumper to bumper Florida traffic or silimar.

Solution: Radiator Fans on an adjustable thermostat.

2) Heatsoak from driving then turning off car, letting it sit long enough to bake the engine bay, turning it back on.

Solution: Wire an automated timer switch to automatically activate the pump on a time interval to move water into front IC radiator where heat soak is less of a problem and where more water volume will be stored anyways. OR just let it rest because it'll cool back down, just not as quick as an air-air unit can. Liquid IC technology may not be the best choice for a performance-minded person who run lots of errands .

 

Additional: A Pump failure warning light add-on and a quick-change backup will come at a higher initial purchase price. I also like the idea of a pump controller, which would allow a user to run it slower when commuting, for instance, if it would help the pump last longer.

 

 

So.... You are saying that after I add all the expense, weight and complexity of this device, it won't really work any better on my daily driven car than what is already available? Who'd a' thunk it!

[GTurabus]

How different is that from the thermostat (simple mechanical device, rarely fails) that keeps the car's water jacket from getting too hot?

 

One is controlled manually and can be opened at any time, the other is of course temperature dependent, and automatic. They both would perform the same function, diverting flow based on a temperature setting.

The normal temperature required to actuate most mechanical automotive thermostats is quite high, but if the coolant temperature in this system is within this range this is another option that is available. It may be that during the development process, the actual temperature for actuation may not be known, so this is something that may need to be added later in the development cycle, if it proves to be desirable for the particular application.

 

In normal development, you do the calculations to determine if something is feasible, then you model it, usually in software. Prior to cutting chips you try to optimize it, as the days of making multiple physical models ended decades ago. The first physical model is usually called the "protoype". Testing of course provides real world data to further refine your model. Before adding layers of expense it would be nice to actually know that the device works close to what was planned, achieving what is known as "proof of principle". Should the prototype work as expected, the next effort would be to add whatever else seems like a good idea that is affordable and meets the original design intent.

 

As we have learned in this thread, there is no further need to work on air to air cooling, as it has been optimized.

 

The OP wishes to investigate another potential solution to charge cooling, but is receiving nothing but flack for just making the effort. I applaud his efforts as a good exercise in the design process and putting to use in a practical application some of the "book learning" he has acquired. Should this project reach a dead end without any real advantage over existing solutions, it is only wasted effort if nothing was learned during the experience. Just because something has been tried before and didn't work doesn't mean it can't be successful. Sometimes looking at something from a different perspective, especially with modern analytical tools, can improve efficiencies to the point where something that was previously discarded may ultimately find some use. As an example look up Wankel engine and the Winner of LeMans in 1991. According to the "conventional thought" exhibited in this board Mazda should have never wasted their time with the rotary engine.

 

Let the development continue then compare the dyno specs, that is the only thing that will satisfy the skeptics.

[ClimberDHexMods]

I love this thread

 

Would it be practical to let the AWIC drain to the radiator (+ a small tank) while the car is off? At least then the water would be next to the engine, instead of sitting over it and right next to the turbo. (I'm assuming the AWIC core would sit where the TMIC core goes.) You'd need an extra valve or two to drain the reservoir and refill the AWIC when the engine starts again, but it should help with the 2nd heat soak scenario.

 

I ran this idea by some people after I read it (really like it)... in theory it would solve the problem of not having the fully cooled charge for the first [x] minutes of operation if it can be made to work very easily, as this is a very small problem, though my thought is everything is good to explore.

 

Two cents time:

 

Adding (too much) circuitry adds complexity and increases the failure modes available. Assuming this device will be used by enthusiasts, and not by the subaru buying public at large, perhaps some good old fashioned engineering could work.

 

For instance, lets just say that there is adequate water pressure available to mount the reservoir (or additional back up reservoir) remotely. This can reduce some of the heat soak issues from the engine bay, but not necessarily those from the car being at idle.

 

One possible solution would be to add a bypass valve from the "main" reservoir, to a cooler, "remote" reservoir. When the temperature gauge (complex mechanical device likel to fail) of the main reservoir reaches a certain temperature (insert calulations here), the valve could be manually flipped to allow the temperatures to come down. I think back to one of my friends old 1960 VW Type 1 convertible with a "kick tank". When the gas gauge hits E (or when you run out of gas) you manually (with your foot) kick the lever down to open the reserve tank. Running the lines in the drive tunnel would naturally allow you to place the lever nearby, outside like the handbrake, or even inside the center console.

 

Before each and everyone interested in this thread tears this apart, please submit your alternate proposal ;-)

 

...

 

Not at all to put your idea down, but this is unnecessary for this application (at least the one I've been designing). I agree it's good to minimize failure points especially considering an AWIC would have several more than a normal FMIC. The radiator fan thermostat feature I mentioned would simply be an extra that may be completely unnecessary. For better or for worse, it's partly me wanting to build the most fully-developed "super system", which is unnecessary to most people.

 

How different is that from the thermostat (simple mechanical device, rarely fails) that keeps the car's water jacket from getting too hot?

 

and if it does somehow fail, nothing critical is lost. Drive around until the replacement arrives.

 

I also had some fun ideas for the pump. One would be a quick-change housing for if it dies. Just a convenience feature. The other is having two pumps and having an electrical switch easily available for when one dies. As if that's not overkill enough, one could turn on the second pump for the track or for other increased load, though we found that a pump like the Bosch Cobra is far from the limiting factor of the system, lasts a looong time, and is named the Cobra

[ClimberDHexMods]

The normal temperature required to actuate most mechanical automotive thermostats is quite high, but if the coolant temperature in this system is within this range this is another option that is available. It may be that during the development process, the actual temperature for actuation may not be known, so this is something that may need to be added later in the development cycle, if it proves to be desirable for the particular application.

 

Yes the only existing options I found previously were the adjustable units, which make sense because you want to adjust it relative to the ambient temperature. Having it digital would be the best. Having it automatically activate at measured ambient+20*F would be yet another awesome though excessive feature that may be better for a concept or show piece.

 

As we have learned in this thread, there is no further need to work on air to air cooling, as it has been optimized.

 

The OP wishes to investigate another potential solution to charge cooling, but is receiving nothing but flack for just making the effort. I applaud his efforts as a good exercise in the design process and putting to use in a practical application some of the "book learning" he has acquired. Should this project reach a dead end without any real advantage over existing solutions, it is only wasted effort if nothing was learned during the experience. Just because something has been tried before and didn't work doesn't mean it can't be successful. Sometimes looking at something from a different perspective, especially with modern analytical tools, can improve efficiencies to the point where something that was previously discarded may ultimately find some use. As an example look up Wankel engine and the Winner of LeMans in 1991. According to the "conventional thought" exhibited in this board Mazda should have never wasted their time with the rotary engine.

 

My exact thoughts, thanks!

 

Let the development continue then compare the dyno specs, that is the only thing that will satisfy the skeptics.

 

YES, lets