As the name suggests, an internal combustion engine creates heat. The enormous potential energy from a few droplets of gasoline compressed and mixed with oxygen and a little spark helps propel us racecar drivers down the track, hopefully pushing us in front of our competitors. During this process, and especially at 6,000 rpm, all of this internal combusting starts to create an immense amount of heat. Too much heat starts to become a problem when temperatures go beyond the abilities of the engine itself.

Remember a steel engine block is just cooled off melted iron ore that has been cast into the shape of an engine. Bring enough heat to the party and you will find yourself in the meltdown zone. The good news is engineers already have solved this problem for us by creating a cooling system for engines: heat exchangers called radiators. These radiators use water to keep engines cool.

Instead of custom-building something ourselves, we hit the easy button and ordered what we needed from somebody who already figured it all out for us. Inside the box is a brand new Mishimoto aluminum radiator.

But if you remember anything about your seventh grade science class, you know that water boils when heated to 212 degrees Fahrenheit, or 100 degrees Celsius. It goes from a liquid to a gas. Essentially it becomes steam. Water cools engines, steam does not. As race engines go beyond 212 degrees, this becomes a big problem. But, in reality it isn’t a problem at all because of two things: antifreeze and pressure.



Your daily driver has coolant inside the engine/radiator that is 50 percent water and 50 percent ethylene glycol, commonly referred to as antifreeze. This simple 1:1 mix of water and ethylene glycol raises the boiling point of water up from 212 degrees Fahrenheit to 226 degrees Fahrenheit. Now we are getting somewhere. Except, actually, we are not. You see, NASA racers are not allowed to use ethylene glycol because of Club Codes & Regulations section 15.18 Engine Coolant, which says, “Glycol-based antifreeze and other additives that may cause a slippery condition if spilled on track are prohibited.” Yup, antifreeze on a racetrack is slipperier than moose snot and it’s hard to clean up … therefore you can’t use it. That means your boiling point is back down to 212 degrees, and if you live at high elevation, like Denver, believe it or not, your boiling point is already down to 208 degrees. We need pressure to solve this issue.

Out with the old to make room for the new. This copper-brass alloy radiator is from 1990, which is 30 years ago. It has seen a lot of miles, has had a lot of junk run through it, and was in dire need of upgrading.


Here is the good news. For every 1 pound of pressure added to a system, the boiling point of water increases 3 degrees Fahrenheit. Engine cooling systems are pressurized to allow for this increase in the boiling point for water. The amount of pressure is regulated by the radiator cap. To demonstrate how the math works, imagine you have a vehicle with pure water in the radiator with a 10 psi radiator cap. Since 1 pound of pressure raises the boiling point by 3 degrees, then a 10-pound cap would raise the boiling point by 30 degrees. Instead of the water boiling at 212 degrees Fahrenheit, it would boil at 242 degrees. Nice! Now we can run hot!

This Mishimoto radiator came with an upgraded radiator cap — 19 psi versus the stock 13 psi — and was designed to drop directly into a 1990-1993 Acura Integra.

Upgrading The Cap

A simple and inexpensive cooling system upgrade to any track car is to replace the OEM radiator cap with a higher-pressure-rated radiator cap. This will allow temperatures in the engine and cooling system to increase to a higher temperature before boiling the coolant and causing a big steaming mess under the hood. To choose the correct upgrade to a radiator cap, you first need to understand what your car came with from the factory. Radiator cap ratings aren’t always created for us dumb Americans. We understand things like 13 psi, which is the exact rating a 1990 Acura Integra came with from the factory. However, that radiator cap indicated the rating was 88 kPa which meant nothing to me. It turns out kPA represents kilopascals, which, again, doesn’t switch a darn thing on in my brain to anything useful.

For you visual learners, this is how a radiator cap works. The lower seal and the vacuum valve allow the system to expand and contract into the expansion tank (or reservoir). The upper seal and the pressure spring hold everything inside the system until the pressure goes beyond the pressure spring’s value and then things get steamy and are released through the cap.

Pressure Conversions

The simple conversion from psi to kPA is 1 psi = 6.89475729 kPa. OK, that seems simple enough. I have a calculator on my cellphone, but things get murkier the further you look into this radiator cap thing. You mostly will see a rating of “bar” on radiator caps. Bar is the metric unit used to measure pressure. The metric system loves the number 100, thus 1 bar = 100 kilopascals. Here are some examples of commercially available radiator caps, their bar rating and their equivalent psi rating: 1.1 bar (16.0 psi), 1.3 bar (18.9 psi), and 1.5 bar (21.8 psi).

Not all radiator caps are equal. This Mishimoto radiator cap is rated at 28 psi or 2.0 bar in metric. The cap that came with our upgraded radiator was rated at 19 psi, or 1.3 bar.

For our Honda Challenge 4 car, we wanted to increase the boiling point of our coolant, so we used a Mishimoto radiator cap rated at 1.3 bar, or 19 psi. Using the math that shows for every additional pound of pressure we increase our boiling point by 3 degrees we went from 212 degrees Fahrenheit — water’s boiling point at sea level — to 269 degrees Fahrenheit. That is a pretty easy upgrade that cost less than a tank of gas and took a total of five seconds to replace. We not only replaced our radiator cap, but we also replaced our OEM copper-brass alloy radiator with an all-aluminum Mishimoto drop-in unit.

To fill our new radiator, we used distilled water. Unlike tap water, distilled water doesn’t have minerals inside it that can become hard water deposits to clog up a radiator or water pump.

Coolant Types

Since we can’t use ethylene glycol in our radiator, thanks to CCR rule 15.18, then we are left with the obvious alternative of water. But just grabbing the hose outside of your garage and filling your radiator with tap water is not advised. This has nothing to do with the conspiracy theory about fluoride in drinking water having mind-control properties and taking over your car. This has to do with minerals in tap water that can harm your cooling system. Minerals create hard water deposits inside your radiator and water pump, which, can clog things up over time. To avoid putting these sorts of minerals inside your engine, you can use distilled water. Distilled water is created by boiling water and then condensing the collected steam back into a liquid. This process removes impurities and minerals from the water, which then won’t end up in your cooling system. A gallon of distilled water will set you back less than a dollar. This is a dirt-cheap upgrade, so it is an easy choice to make.

Here you can see our new Mishimoto radiator installed in our NASA Honda Challenge 4 Acura Integra. Because we increased our pressure for the system with a higher pressure radiator cap, we also decided to upgrade our hoses from rubber to silicone to sustain the added coolant pressure.

For our Honda Challenge 4 car, we took all of these deep dives into radiator caps, boiling temperatures, unit conversions, radiator types, and water distilling and decided to slap in a new Mishimoto all-aluminum radiator, a new lower-temperature thermostat, a higher-pressure radiator cap, a few gallons of distilled water and some silicone hoses. This combination of parts easily bolted into our 1990 Acura Integra RS and gave us the peace of mind that we would never have to think about this stuff ever again. Nobody wants to do more math than they have to.

From the front of the car, you can see we have the Mishimoto aluminum radiator installed, as evidenced by the big black “M” peeking out of the grille. I like to think of it as “M” for “Move out of my way.”

We went with the upgraded aluminum radiator, because by increasing the pressure to the cooling system with the upgraded radiator cap, we were not convinced our 30-year-old copper-brass alloy radiator could withstand the higher pressure and the beating road racing cars with stiff suspensions take on the racetrack. We looked at the system as a whole, as opposed to looking at just one component, which is important to do when doing any modification to a car.

With the hood shut and the AeroCatch latches closed, this upgraded cooling project is done! Time to hit the track.

As a team who races with NASA predominantly in Southern California, keeping our internal combustion engine cool is important to us at tracks like Willow Springs and Buttonwillow. Think, hot desert. Our Mishimoto components should get the job done. And now you know what a kilopascal is.

Image courtesy of Rob Krider


  1. THANKS SO MUCH for an EXCELLENT article!!!! I already knew about the effects the different things would have and the reasons you were doing the things you did, BUT!! I STILL truly enjoyed the entire article AND… I didn’t know what a kilopascal was!!! lol

    So thanks again Bob!!!

  2. OH!!! And I forgot!! I ALSO didn’t know how many degrees the boiling point went up with added pressure!!! THAT is invaluable information!!
    THANKS again!

  3. How about a followup on the use of water wetter (etc) and it’s positive impact on preventing cavitation/hot zones where steam pockets can still form…

    And to that, why higher pressure also helps to fight off..steam pockets, but still needs help.

    Good one Rob!

    • Great Jeff: Water wetter reduces surface tension, like soap, but without the bubbles. It allows the water to make direct contact with the surfaces and reduces hot spots which cause cavitation. Cavitation is where the water then changes state from liquid (where it can make contact and remove heat) to gaseous (steam) and that steam has a low coefficient of heat transfer, this reduced cooling.

      Higher pressure increases the temperature at which water changes state from liquid to gas. This is shown through Boyle’s gas law. PV=nRT, where T is directly proportional P (Pressure).

      So higher pressure and lower surface tension both keep water liquid and in contact with the surfaces where heat transfer takes place. Another benefit is that the water carries more heat to the radiator. And the radiator’s efficiency is governed by delta T. That is the higher the radiator temperature, the more larger the delta (difference) in heat between radiator and cooling air. Or said another way, the larger the difference in temperatures, the more heat is transferred from the air through the radiator.

      A nice story might extrapolate from this comment!

    • Rob, Can you ask some questions to your race teams about Evans and or the other wetter’s that came up in the discussion. Additional pressure adds stress on the system, and if it’s a racing car that maybe OK with high end hoses, fittings and radiators. Bone stock systems may not be reliable over the long term with the additional pressures. Just asking.

  4. Yes, it would be helpful to know if any waterless coolants qualify for use. I know that some of them are actually flammable!!

    But also, what is water wetter, i.e. what is it made of, how does it work/help and what are the benefits/drawbacks!?

  5. I started using Royal Purple Purple Ice. I saw in their data sheet they want filtered drinking water to be used with this, NOT distilled water. I called their tech support line where that was confirmed. So that is what I’m using now. They said distilled is okay if it is with anti-freeze, water and Purple Ice.

  6. I use Blue Ice in my Goldwing. It is not waterless but comes premixed with ionized water. But that is not an engine that is driven extra hard like in a race car so im not sure of the effectiveness of that kind of coolant in a racing application.

  7. That is correct, Red Line Oil’s Water Wetter is allowed for use in NASA racing. It does not contain any ethylene glycol. We use it in our Honda Challenge cars to clean and lubricate the water pump seals. Seems to work great.

  8. Thanks for that great explanation of how water wetter works! Also the added bonus of the whole delta thing. I didn’t know a radiator was MORE efficient with a larger temperature differential EVEN when the water temp goes UP and the cooling air temp stays the SAME!?

    • Hi Keith: Regarding Delta T: Physics is such a cool thing. It allows us quantifiable ways to describe physical processes. The specific area of physics (thermodynamics) we’re talking about can be complex.

      So, Thermodynamics tells us that the amount of energy a system can remove is precisely based on Mass Flow and change in temperature (Delta T) of that mass. The delta T thing also gets into Calculus as in rates… dy/dt.

      But forget all that. Just think of it this way. Heat (energy) moves from higher to lower temperatures. So, for example, it’s obvious that 80 F air will not be able to remove heat from 80 F water. But 80 F air can remove some heat from 81 F water. And a heck of a lot more heat can b removed from 220F water.

  9. Mario,
    Yes, it is indeed the simple! It just sounded implausible when it seemed that you were saying that the radiator could remove hear more EFFICIENTLY when the temp differential between water and air b was higher! lol

    But I would think that the efficiency of the system components and process would remain fixed and (as you just explained), it is simply just a matter of quantity…

    i. e. The colder the air is relative to the hot water, the more heat it can absorb as the water goes through the radiator!

    So if the system operates with let’s say, 60% efficiency (Im just grabbing a number for reference), then it will remove heat at that percentage whether there is a little (low temp differential) or a lot (high temp differential)!?

    With one other factor thrown in… At dome point, the system will reach a saturation point where even if the temperature differential were to go higher, a particular system would not be capable of extracting any more hear, it would plateau, or “top out” as it were.



    • Hi Keith: re: “At dome (some) point, the system will reach a saturation point where even if the temperature differential were to go higher, a particular system would not be capable of extracting any more hear, it would plateau, or “top out” as it were.”

      You’re good at making science common sense! Stated in scientific terms: I translate as follows. When the rate of heat removal is less than heat build-up, you get runaway overheating.

  10. Yep!
    I know that the air intercoolers on my Abarth can very quickly reach saturation and become ineffective on warner days if you increase the duty cycle. In other words: drive the car hard too long without a long enough “down cycle” where your foot is off the gas and the excess heat built up in the intercooler has time to disipate.

    Of course if you want to increase the capability of the system under those conditions, then you simply increase its capacity (intercooler size) or efficiency (change from air intercoolers to water for an increased heat transfer rate due to the higher density of water vs air).

    I took the cheap and easy way out since I had the nose off the car. I simply added patio mister nozzles in front of the intercoolers that week connected to a washer pump so I could spray cool water on the intercoolers with the push of a button when I was under hard acceleration. And or course, the evaporation of the fine water mist rapidly loweres the temp of the intercooler fins that the air is passing over AND the temp of that air as well!

    Increases the efficiency and performance of the intercoolers ONLY when you need it, without doing an expensive retrofit. Not an adequate solution for racing, but works great for performance driving on winding roads!! lol

    I never got into “hard core” physics…
    but applied physics (or practical physics? lol) and science and engineering in general has always been my “thing” since I was a young boy… The quest for the answers and understanding of “what” is it, “why” is it this way, and “how” does it work!

    Or to put it simply:a never ending quest (curiousity) to learn MORE about the world around me (and yes, that also includes psychology to understand the PEOPLE in that world!!).

    That quest always leads to one adventure (or frustration) after another!! lol

    Thanks again for your insight and explanations!


    • Keith that misting is incredibly effective. In terms of the physics, the latent heat of vaporization is an amazing thing. For example, it takes 1 calorie of heat to change temperature of 1 gm of water 1 degree C. However, the take 1 gm of water from liquid to gas (vaporization) it consumes about 500 calories of energy!!! So when that water evaporates on the cooling fins… you get extraordinary delta T.

  11. WOW Mario! I knew it was pretty darn effective from using that principle in many different ways over the years but i had no idea it was “extraordinarily” effective!!!
    That changes my to a lol.
    I’m the guy who put a hood scoop on my Fiat 131 Brava, ducted straight into the carb (ram air), only to find that when the outside air temp dropped below 50 degrees and you were at sustained highway speed of 70mph or more, the air/fuel mixture in the carb (which is always chilling from the evaporative effect) would ice up the carb and Id have to pull over to the shoulder and let the car idle for a couple minutes to melt it!! lol

    So, i solved the problem by mounting a switchback pattern of 1/4inch copper copper tubing in the opening in the hood where the air came down into the carb and connecting the tubing to the heater hose with a “tee” and a petcock (so i could turn it on and off)…

    And PRESTO!! lol I had “carburetor heat”!! (Im a pilot and we have carb heat on airplanes to prevent icing).

    Interesting thing, when experimenting on normal temp days (not real HOT or cold), I found that running the “carb heat” actually INCREASED my gas mileage above what it was both before and after adding the hood scoop, without decreasing my power a noticable amount (under normal.driving conditions)!!?? (In an airplane, carb heat dramatically lowers power output by about 25% but that is partially due to restricting the flow of the air to the carbs because of the different source/path of the air through the carb heat system. But the power reduction is not a problem because icing only takes place at low power settings associated with descent and landing).

    So does warming the intake air to the carb increase the efficiency of the venturi system and/or combustion process somehow?
    (Thermodynamics here we COME!! lol)

    (BTW I gotta get back to my work now!! lol)


    • There’s a whole lot of physics going on here. Warm air is less dense, and has less oxygen. Cold air burns leaner. So if you adjust the A/F ratio in response, there are a lot of things to consider. But icing up is devastating!

      In the end, what we know is that the internal combustion engine is a heat pump… and again Delta T comes up. Delta T determines the power made. Cold in and hot out is total power. When you account for the waste heat (the heat that does not do work (i.e. turn the crank), which is mostly out the exhaust) you can calc efficiency. This is very difficult to do in reality. How do you measure all the waste heat? Difficult at best!

  12. Yes indeed, all that is true and i am aware of those things! But I am wondering how warming the intake air to the venturi, which raises the resultant temp of the air/fuel mixture to the cylinders could INCREASE the gas mileage (of course by increasing the efficiency or output of the combustion, but is that possible!?).

    I surmise that it could be because it improves the atomization of the air and fuel in the venturi and/or improves the combustion process in the cylinders.

    Id like to know if it’s a process you are aware of!

    In the 70’s gas crisis, when I was in college studying Applied Physics, Electronics, Machining and Television, they was a guy who had modified his car with a system to do just THAT, warm the mixture to the cylinders!

    Only be didn’t warm the air, be heated the fuel before it went to the carbs! He had conducted multiple tests and showed a marked increase in the MPG!

    So I wasn’t surprised when it happened in my car (and yes i conducted many controlled tests of my own), but I never quite understood “why” or “how” it was doing that when it seemed to defy the laws of Thermodynamics.

    Have you heard of that effect before? And/or do you know the principles by which that would occur? Or is it just a mystery?

    We know that raising temp reduces air density and thus power, but is there a sweet spot where warming the air/fuel mixture INCREASES combustion MORE thereby increasing power output for the amount of air/fuel mixture?!

    THAT is my question. I assure you my results were real and repeatable (about 2mpg)!

    Obviously this would only work under light load, i.e. cruise on the highway…


  13. Keith: I think you hit upon most of the reasons and questions surrounding warming the air so that the gasoline droplets vaporize before combustion in carbureted systems. The caveat to my answer is that engineering is the art of balancing compromises.

    Ideally, you want the best spray pattern in order to reduce fuel (liquid) droplet size so that the “liquid” droplets vaporize just before combustion. This is critical, since liquid droplets have less surface area for combining with O2 than vapor. Stated again, the spray is initially liquid droplets. After the droplets vaporize, they’ve completed their state change which soaks up heat. This state change drops the intake temperature! This is the start of the heat pump where work is about to be done. You want to start at minimal temperature.

    Then after that, the valves close, and the piston moves to compress the cooled vapor before the spark can ignite the cold vaporized air and fuel. Of course, as the vaporized mixture is compressed, it does heats up proportionally according to PV=nRT. You want to minimize this heat as much as possible by attempting to keep in the intake chamber cooled. Then with this compressed air/fuel you want it to combust at the optimal time to expand and act upon the piston to make it do work.

    Ideally there is enough O2 to fully combust the fuel. Then the burn happens. Temperature increase (change) causes expansion, and the work is extracted by moving the piston. The most ideal situation is caused by the max magnitude of Delta T. Given the above explanation, this should make sense.

    So as we know, today’s best performing engines spray the fuel directly into the cylinders where the fuel goes through the state change (from liquid droplets into gaseous state). This state change drops the temperature significantly where you want it to be coldest.

    Carburetors benefit from heating the air (when it’s too cold to evaporate) to ensure evaporation which is –> vaporization. But then, the added heat reduces the efficiency since Delta T is limited. It’s a balance between making sure you get vaporized fuel instead of solid droplets. This is an area of a tremendous amount of science and engineering knowledge to optimize vaporization without causing the mixture to ignite at the wrong time! We can get into stable fuels which allow more compression without premature ignition. Think high octane, and intercoolers…

    Ideal efficiency (using the least amount of fuel so that all of it combusts) results in very low unburned fuel (hydrocarbons) and optimal use of fuel to do work without wasting it. This efficiency is to save gas. However, performance engines run a richer (more wasteful) air/fuel ratio to make more power, in that there will be leftover fuel such that you get hydrocarbons (unburned fuel) in the exhaust. This extra fuel helps cool the intake charge, and contribute some additional fuel burn but wastes gasoline!

    So it’s very complicated.

Join the Discussion