Air to Water Intercooler?
- Thread starter zSP3CTERz
- Start date
IJ.;1393157 said:Or mount it outside the engine bay...
yeah, but that big green water holding tank doesnt look very aerodynamic......:sarcasm:
On a more serious note, how effective are the CO2 sprayer type setups for A2A IC for the heatsoak issue? Is it worth the time and money to install?
The meth setups have always had my interest for lower IAT's also....
lol7thousandpiecesMGTE;1393326 said:
Subscribed...
But just a question, keep in mind that I don't plan on running a W2A IC any time soon, do you have a gauge to measure either the temps of the IAT or the water in the system? If so, what would the ideal readings of either be?
Last edited:
Good God, so much to respond to... Kind of ironic since I'm probably going to go with a liquid to air intercooler on my build. Anyway, most of what has been posted here is in and of itself true -- just incomplete (this goes for arguments both for & against L2A).
So I guess let's start off with Newton's Law of Cooling:
Simple enough, no?
So when people say that a liquid to air intercooler is the most efficient, they are relying on the additional thermal mass of the liquid in the system. Using ice, they can drop the temperature inside the intercooler to below ambient -- which as we can see from Newton's cooling law, makes a larger temperature difference and promotes more heat transfer than otherwise...
Under steady-state conditions, however, a liquid-to-air intercooler is less efficient (provided both systems are comparably sized). We can see this from Newton's equation -- an air to air IC relies on heat transfer from the charge air to the aluminum intercooler, and from the aluminum intercooler back to the ambient air.
The L2A intercooler has additional thermal boundaries (and like everything, there are losses associated with each boundary). So we're all clear, the L2A setup goes like this: charge air transfers heat with the L2A's aluminum intercooler which transfers heat with the liquid coolant. The liquid coolant then relies on a 2nd heat exchanger which transfers heat from the liquid to the aluminum radiator, which then transfers heat to the ambient air.
The heat still ultimately gets transfered to the same place (the atmosphere) it just has to go through the coolant and a 2nd heat exchanger in the meantime. These additional thermal barriers mean this system is ultimately less efficient under steady-state conditions.
But but but IJ said his logs showed L2A was better for stop/go driving! I don't doubt IJ in the least, but stop-n-go driving is not steady state. In stop/n/go driving, the additional thermal mass of the entire L2A's coolant system acts as a buffer (especially when combined with a fan on the radiator).
One of my old professors recently wrote a somewhat related article on "green" home construction methods: http://www.newsreview.com/chico/content?oid=1137549
A liquid-to-air setup would be more more like a massed setup (steadier IAT's) while an air-to-air would be like a low-mass construction (rapidly sheds heat, but fluctuates more based on ambient temps/airflow).
Does that make one "better" than the other? No, just different and ideal for different constraints...
A couple other points:
Ceramic coating is not just for corrosion protection at high temps (otherwise people would just use high-temp paint). But ceramic coating is just a thermal barrier -- it slows (but does not stop) heat transfer. Given enough time, a ceramic coated part will reach the same temperatures as a non coated part (same goes for wraps). It is still certainly worthwhile, but it isn't a perfect insulation (nothing is).
On the same note, "thermal dispersent" coatings are kind of a joke. They add another barrier and inhibit thermal transfer -- just like anything else. the difference is they are designed to be better conductors than other coatings, so they insulate less than regular paint. They do not make a part transfer heat better than without it... The only time this isn't true is when looking long-term at a part that is likely to corrode (assuming the corroded material transfers heat worse than a thermal dispersent coated uncorroded part). Now if you are going to paint a part anyway, then by all means use a thermal dispersent coating. But don't use it just because you think it's going to be an upgrade...
Mounting IC inside the engine bay... I'm going to go this way on my setup, so no, I obviously don't think its stupid. Will it heat soak more than an externally mounted cooler? Of course, but for my needs a top-mount L2A is best. Here's why: The shorter piping means slightly less lag and better throttle response. The fact that the entire charged section of piping is now mounted solidly with the engine means less boost leaks (when the engine torques in the mounts, the only part of the induction system to flex will be the air filter pipe).
Other considerations: L2A is more complicated -- having coolant, a 2nd heat exhanger, a pump, and possibly a resevoir. There is "more to go wrong" although this stuff is all pretty simple/primative so its not a huge deal. It also adds more weight.
One last comment: a 450cfm intercooler is good for 300hp -- not 600+ The general rough approximation is 1.5cfm/hp (the precise amount will depend on efficiency but this is ballpark). It doesn't matter whether thats 450cfm through 4, 6, or 8 cylinders. I suspect you found something that estimated horsepower based on individual port flow...
So I guess let's start off with Newton's Law of Cooling:
From this equation, we can see that the amount of thermal energy transferred is controlled by 3 variables: Surface area of the heat exchanger, the temperature difference between the heat exchanger and its surrounding environment, and on the heat transfer coefficient (a coefficient that represents how well the exchanger transfers heat to its particular surroundings).Q = Thermal energy in joules h = Heat transfer coefficient A = Surface area of the heat being transferred T0 = Temperature of the object's surfaceTenv = Temperature of the environment
Simple enough, no?
So when people say that a liquid to air intercooler is the most efficient, they are relying on the additional thermal mass of the liquid in the system. Using ice, they can drop the temperature inside the intercooler to below ambient -- which as we can see from Newton's cooling law, makes a larger temperature difference and promotes more heat transfer than otherwise...
Under steady-state conditions, however, a liquid-to-air intercooler is less efficient (provided both systems are comparably sized). We can see this from Newton's equation -- an air to air IC relies on heat transfer from the charge air to the aluminum intercooler, and from the aluminum intercooler back to the ambient air.
The L2A intercooler has additional thermal boundaries (and like everything, there are losses associated with each boundary). So we're all clear, the L2A setup goes like this: charge air transfers heat with the L2A's aluminum intercooler which transfers heat with the liquid coolant. The liquid coolant then relies on a 2nd heat exchanger which transfers heat from the liquid to the aluminum radiator, which then transfers heat to the ambient air.
The heat still ultimately gets transfered to the same place (the atmosphere) it just has to go through the coolant and a 2nd heat exchanger in the meantime. These additional thermal barriers mean this system is ultimately less efficient under steady-state conditions.
But but but IJ said his logs showed L2A was better for stop/go driving! I don't doubt IJ in the least, but stop-n-go driving is not steady state. In stop/n/go driving, the additional thermal mass of the entire L2A's coolant system acts as a buffer (especially when combined with a fan on the radiator).
One of my old professors recently wrote a somewhat related article on "green" home construction methods: http://www.newsreview.com/chico/content?oid=1137549
A liquid-to-air setup would be more more like a massed setup (steadier IAT's) while an air-to-air would be like a low-mass construction (rapidly sheds heat, but fluctuates more based on ambient temps/airflow).
Does that make one "better" than the other? No, just different and ideal for different constraints...
A couple other points:
Ceramic coating is not just for corrosion protection at high temps (otherwise people would just use high-temp paint). But ceramic coating is just a thermal barrier -- it slows (but does not stop) heat transfer. Given enough time, a ceramic coated part will reach the same temperatures as a non coated part (same goes for wraps). It is still certainly worthwhile, but it isn't a perfect insulation (nothing is).
On the same note, "thermal dispersent" coatings are kind of a joke. They add another barrier and inhibit thermal transfer -- just like anything else. the difference is they are designed to be better conductors than other coatings, so they insulate less than regular paint. They do not make a part transfer heat better than without it... The only time this isn't true is when looking long-term at a part that is likely to corrode (assuming the corroded material transfers heat worse than a thermal dispersent coated uncorroded part). Now if you are going to paint a part anyway, then by all means use a thermal dispersent coating. But don't use it just because you think it's going to be an upgrade...
Mounting IC inside the engine bay... I'm going to go this way on my setup, so no, I obviously don't think its stupid. Will it heat soak more than an externally mounted cooler? Of course, but for my needs a top-mount L2A is best. Here's why: The shorter piping means slightly less lag and better throttle response. The fact that the entire charged section of piping is now mounted solidly with the engine means less boost leaks (when the engine torques in the mounts, the only part of the induction system to flex will be the air filter pipe).
Other considerations: L2A is more complicated -- having coolant, a 2nd heat exhanger, a pump, and possibly a resevoir. There is "more to go wrong" although this stuff is all pretty simple/primative so its not a huge deal. It also adds more weight.
One last comment: a 450cfm intercooler is good for 300hp -- not 600+ The general rough approximation is 1.5cfm/hp (the precise amount will depend on efficiency but this is ballpark). It doesn't matter whether thats 450cfm through 4, 6, or 8 cylinders. I suspect you found something that estimated horsepower based on individual port flow...
Last edited:
7thousandpiecesMGTE;1393326 said:
Are you actually getting heatsoak? If so they are fairly effective (and often used to lower the temp of an IC to sub ambient). Whether its worth the time/money depends on your needs...
mkiiSupraMan18;1393401 said:
Most cars already have an IAT sensor (though I believe the mk3's is in the AFM so it wouldn't help). But for those who go standalone its quite easy to log IATs. Many of the L2A intercoolers also have a port to mount a water temp sensor.
As far as what temps you should see: A well engineered intercooler should bring IATs within 10-30 degrees of ambient. For those of us in hot climates, this means we can see 150 degree IAT's on a 120 degree summer day at the track (a temperature that is high enough to warrant pulling some timing).
For the L2A's coolant temperature, you would want it to be as close to ambient as possible (and unless massively heat-soaked, it probably will be within a few degrees of ambient).
My point was only through real world testing in a daily driver the W2A was superior 
I logged IN/OUT AIT's and found the active nature of a W2A system coped with stop start driving being pump/fan assisted, A2A has minimal heat rejection if you can't generate forward movement, adding a pair of fans made a little difference.
On the HP/Sizing issue, I used a PWR 10"x6" barrel cooler and above 450rwhp it choked and stopped being effective with a huge temp rise.
Keep in mind most if not all IC HP ratings are in Crank hp, The Hardware involved in a 500+rwhp W2A becomes heavy and hard to fit in the front of a Mk3, even if you install it all inside the bay.
I logged IN/OUT AIT's and found the active nature of a W2A system coped with stop start driving being pump/fan assisted, A2A has minimal heat rejection if you can't generate forward movement, adding a pair of fans made a little difference.
On the HP/Sizing issue, I used a PWR 10"x6" barrel cooler and above 450rwhp it choked and stopped being effective with a huge temp rise.
Keep in mind most if not all IC HP ratings are in Crank hp, The Hardware involved in a 500+rwhp W2A becomes heavy and hard to fit in the front of a Mk3, even if you install it all inside the bay.
sk6471;1326786 said:
Actually
if the IC are identical and the only thing different is the medium for heat transfer (air or water), WATER will come out on top every single time. In short, waters thermal properties are 10x that of air. It is the same exact reason you go to the beach where it is 80 and the ocean is 76.You come out of the ocean freezing as the water is able to absorb copious amount of heat. Hence why cooling systems in all cars now a days are water/coolant cooling and not air
Doward;1327376 said:
and has been verified by Ian
It is no different than sizing an A2A IC. Size the IC for the right application.ma71supraturbo;1393479 said:So I guess let's start off with Newton's Law of Cooling:
From this equation, we can see that the amount of thermal energy transferred is controlled by 3 variables: Surface area of the heat exchanger, the temperature difference between the heat exchanger and its surrounding environment, and on the heat transfer coefficient (a coefficient that represents how well the exchanger transfers heat to its particular surroundings).
Simple enough, no?
So when people say that a liquid to air intercooler is the most efficient, they are relying on the additional thermal mass of the liquid in the system. Using ice, they can drop the temperature inside the intercooler to below ambient -- which as we can see from Newton's cooling law, makes a larger temperature difference and promotes more heat transfer than otherwise...
Under steady-state conditions, however, a liquid-to-air intercooler is less efficient (provided both systems are comparably sized). We can see this from Newton's equation -- an air to air IC relies on heat transfer from the charge air to the aluminum intercooler, and from the aluminum intercooler back to the ambient air.
The L2A intercooler has additional thermal boundaries (and like everything, there are losses associated with each boundary). So we're all clear, the L2A setup goes like this: charge air transfers heat with the L2A's aluminum intercooler which transfers heat with the liquid coolant. The liquid coolant then relies on a 2nd heat exchanger which transfers heat from the liquid to the aluminum radiator, which then transfers heat to the ambient air.
The heat still ultimately gets transfered to the same place (the atmosphere) it just has to go through the coolant and a 2nd heat exchanger in the meantime. These additional thermal barriers mean this system is ultimately less efficient under steady-state conditions.
But but but IJ said his logs showed L2A was better for stop/go driving! I don't doubt IJ in the least, but stop-n-go driving is not steady state. In stop/n/go driving, the additional thermal mass of the entire L2A's coolant system acts as a buffer (especially when combined with a fan on the radiator).
One of my old professors recently wrote a somewhat related article on "green" home construction methods: http://www.newsreview.com/chico/content?oid=1137549
A liquid-to-air setup would be more more like a massed setup (steadier IAT's) while an air-to-air would be like a low-mass construction (rapidly sheds heat, but fluctuates more based on ambient temps/airflow).
Does that make one "better" than the other? No, just different and ideal for different constraints...
A couple other points:
Ceramic coating is not just for corrosion protection at high temps (otherwise people would just use high-temp paint). But ceramic coating is just a thermal barrier -- it slows (but does not stop) heat transfer. Given enough time, a ceramic coated part will reach the same temperatures as a non coated part (same goes for wraps). It is still certainly worthwhile, but it isn't a perfect insulation (nothing is).
On the same note, "thermal dispersent" coatings are kind of a joke. They add another barrier and inhibit thermal transfer -- just like anything else. the difference is they are designed to be better conductors than other coatings, so they insulate less than regular paint. They do not make a part transfer heat better than without it... The only time this isn't true is when looking long-term at a part that is likely to corrode (assuming the corroded material transfers heat worse than a thermal dispersent coated uncorroded part). Now if you are going to paint a part anyway, then by all means use a thermal dispersent coating. But don't use it just because you think it's going to be an upgrade...
Mounting IC inside the engine bay... I'm going to go this way on my setup, so no, I obviously don't think its stupid. Will it heat soak more than an externally mounted cooler? Of course, but for my needs a top-mount L2A is best. Here's why: The shorter piping means slightly less lag and better throttle response. The fact that the entire charged section of piping is now mounted solidly with the engine means less boost leaks (when the engine torques in the mounts, the only part of the induction system to flex will be the air filter pipe).
Other considerations: L2A is more complicated -- having coolant, a 2nd heat exhanger, a pump, and possibly a resevoir. There is "more to go wrong" although this stuff is all pretty simple/primative so its not a huge deal. It also adds more weight.
One last comment: a 450cfm intercooler is good for 300hp -- not 600+ The general rough approximation is 1.5cfm/hp (the precise amount will depend on efficiency but this is ballpark). It doesn't matter whether thats 450cfm through 4, 6, or 8 cylinders. I suspect you found something that estimated horsepower based on individual port flow...
MA71
you are only applying one of many laws of physics and chemistry to what seems a simple system on the surface.
You really need to know
specific heat capacity of both (water is better by about 10x in all regards)
volume heat capacity of both (again water is about 10x better in all regards) and a slew of other thermodynamics functions for the metal use, thermal properties of the metal etc.
But, as Ian has measured. His numbers would coincide with physics almost spot on. The only short coming of the L2A is the additional mass for the additional components but of the two properly sized IC systems, the L2A will always trump the A2A system.
btw if anyone is remotely intrested, check out the system that Ford used in the F-140 lightning.
They went one step further and used the AC system to chill the L2A at certain times.There is an ENORMOUS amount of science and variables in this and why I posted my "observations" as the theory/math behind it would reduce my brain to a puddle
Yeah, and I could see how daily driving in a hot/metropolitan area would yield those results (although I'd recommend using the fans in a puller configuration if you could fit it -- let the front surface area be exposed to as much cool air as possible).IJ.;1393573 said:My point was only through real world testing in a daily driver the W2A was superior
I logged IN/OUT AIT's and found the active nature of a W2A system coped with stop start driving being pump/fan assisted, A2A has minimal heat rejection if you can't generate forward movement, adding a pair of fans made a little difference.
My car will be weekend/track only where it'll be most comparable to steady state (loads will obviously vary, but it'll rarely be below 40mph).
It would seem that an A2A would be "best" for me, but I'm willing to give up a few hp to have a more responsive engine/precise throttle. Perhaps I should just get it over with and do a V8 swap
Yeah, I'm only shooting for 400-450rwhp, so this 700+cfm L2A IC should be adequate:
I'll probably lengthen the notch on my valve covers and have to reroute the ISC, but hopefully I can sneak this between the PCV ports and the coilpack...
With a 12x24" radiator, 1250cfm fan, and another few gallons in a remote reservoir, it should be adequate (if not a pain to fit)
I also found tuning was easier with the W2A as you don't have such a wide variation of intake temps to program compensations for.
ie: pull in to fill up with fuel come out and try to restart the car with an A2A that's heat soaked into the twilight zone of AIT Compensation that's pulling fuel and messing with timing and it can take a couple of minutes of cranking to get it to fire.
I ended up adding a cranking compensation to get around this.
ie: pull in to fill up with fuel come out and try to restart the car with an A2A that's heat soaked into the twilight zone of AIT Compensation that's pulling fuel and messing with timing and it can take a couple of minutes of cranking to get it to fire.
I ended up adding a cranking compensation to get around this.
<== wonders how big an L2A would fit between the bumper support and the AC condensor........ hmmmmm
figgie;1393596 said:
That mainly is due to the energy absorbed during the phase change when water evaporates from your skin. For our purposes, the coolant remains in a liquid state, so this phase change never occurs.
They are liquid cooled because it was a cheap/effective way of maintaining the engine at a desired temperature range. But you're neglecting one very important thing...Hence why cooling systems in all cars now a days are water/coolant cooling and not air
The coolant still needs to be cooled by air! All of your statements would be valid if we were using an infinite capacity of ambient temperature coolant (like a marine application where we pump water directly from the lake). But we're not. We still need to use a radiator to cool the liquid back down as close to ambient as possible. Our main limitation then is how much heat we can transfer from the front heat exchanger to air. With a L2A setup, you have to transfer this heat an additional 2 times (from the coolant to the radiator and from the radiator to the outside air). Every time you erect a thermal barrier -- no matter how efficient -- it still has associated losses.
The mass & capacity of the liquid to hold heat still is greater -- so Ian's observations when idling between runs is perfectly valid. But under steady-state conditions, a comparably sized A2A setup would have better thermodynamic efficiency.
My first setup was far too small in capacity (pic 1) I then upgraded (pic 2) and was still too small
Heading out to work will check back at lunch!
Heading out to work will check back at lunch!
ma71supraturbo;1393619 said:
thinking when you get out. get in for the first time. also why it is easy to even get hypothermia in water that is just 65 degrees.
ma71supraturbo;1393619 said:
Since you insist.
The thermodynamic losses you speak off (entropy) even with three energy state conversions of the L2A setup is less than the A2A setup. Why? Because of the Specific heat capacity of water.
In the end it is about energy transfer efficiencies at each point which the L2A has advantages not to include mass (which is a EBC twist away anyway). Particularly, position of the W2A heat exchanger can be better used for disipiation and or optimization of the air flow around the heat exchanger which means that it can also allow for better thermal rejection from radiation/convection of the asphalt.
Throw in some dry ice or a system exactly like Ford has on their Lightning, and then there is a no win for A2A.
IJ.;1393622 said:6"x10"
refer to my pic, I always have AC!
that how big the one you have is but there should be more room there no?