Difficult Problems in Automotive Thermodynamics

While clearing the cars of frost Thursday morning, I was reminded of a difference of opinion Kate and I have about the best approach to a cold car. I tend to start the car up, and immediately turn the defrost on full blast; she prefers to let the car warm up for a while first, because it just blows cold air for a while. My feeling is that while the air from the defrost is not warm in an absolute sense, it’s still passed over the engine on the way in, and thus is warmer than the outside air. Which means that it’s going to increase the temperature inside the car somewhat, and thus will heat the interior to an acceptable temperature faster than waiting until the car is fully warm and just blowing in hot air.

Of course, this ought to be experimentally testable, but it would be a gigantic hassle. My car is much larger than Kate’s (2007 Ford Freestyle vs. 2003 Toyota Prius), so a direct comparison would be meaningless. And while I do have a colleague with a nearly identical car that we used for a previous experiment, he doesn’t live anywhere near me, and getting his car over here would necessarily warm it up, ruining the experiment (I guess we could try to coordinate our schedules to do a test after work some day, but I’m not sure I’d want to ask anybody else to stand out in the cold for SCIENCE!). And the morning temperatures vary widely from day to day making it difficult to do an extended series of tests.

I toyed with the idea of simulating this mathematically, but a few minutes’ thought show that it’s a horrible problem. I don’t know how the temperature of the defrost air increases over time, or how to model the mixing with the air that’s already in the car (which presumably exits in some manner, because running the defrost on full doesn’t overpressure the car to any significant degree. And I’m really not willing to spend a lot of mornings standing out in the cold long enough to collect data for this.

I suppose what I need is a USB temperature sensor that I could put inside the car, and use to log data from inside where it’s warm. Hmmm…

20 Replies to “Difficult Problems in Automotive Thermodynamics”

  1. I have no comment on the optimal thermodynamics. But as a teen, I was taught that when it’s very cold, it’s best for the health of the engine to let it warm up before turning on the heater. (Or before driving it, for that matter.) But that’s likely one of those things that hasn’t been important since the 70s or so.

  2. There are too many other variables here. Is snow/sleet/freezing rain falling now (and the type matters as well)? Is the stuff on the car windows snow, frost, or accumulated ice? Do you prefer to start the car first, or do you clear off the snow and ice before turning on the ignition (I subscribe to the latter school as I consider the former a waste of gas, but I see more people who adopt the former view)? Most likely there will be times when your approach is correct, and others where Kate’s method works better. What I usually do is turn the defrost on right away if precipitation is falling or I’m dealing with accumulated ice, but wait (or not bother with it at all) in dry conditions.

  3. At least turn it on low right at the start. Letting the heat come up to temp and then blasting it at a still-cold window is a good way to get a cracked windshield.

    Of course, it all depends; that might matter more here, where it’s currently -12 F.

  4. Also you have to factor in the “wind chill factor” of the moving air from the heater. Even if it has some warmth, the speed of the air has a chilling effect and it might feel colder than leaving the fan off.

  5. My goal is generally not to warm the car up as fast as possible, but to clear the windshield as fast as possible (because I’m close to running late and want to get going – but reasonably safely – as soon as possible). I’m also using a scraper, brush, or shovel, depending on what’s on the windshield.

    Does this change in goal change the answer?

  6. I am with Derek R on this one. You and Kate are probably both right, but trying to optimize two different things: you are trying to clear the frost faster and she is trying to maximize passenger comfort.

  7. A couple of notes on how your car heats up:

    1. There’s a thermostat in the coolant system, that prevents fluid from circulating until the coolant on the hot side reaches a specific temperature. It’s a very simple mechanical valve that uses thermal expansion to open.

    2. There’s a small radiator inside your car, just like the big one at the front of the engine, that is intended to transfer heat from your coolant to the air, either coming into your car from the outside, or that’s already in your car, depending on if you have the recirculation vents open or closed. There’s an attached fan that forces this air across the radiator. On recirc, there should be relatively little exchange with the outside, but on external air, it obviously comes in and goes out pretty freely.

    3. If the thermostat is closed because the coolant hasn’t heated up yet, you’re blowing air across coolant that isn’t circulating through the engine, and is thus pretty much the same temperature it was when it started.

    4. Modern cars automatically set the recirculation system to external air when you defrost; the external air is drier (because you’re not breathing in it) and won’t fog your cold windshield (I’m not 100% positive I have this right, but it seems to make sense). This often can’t be changed manually. Just plain heat can be either external or recirculated air.

    Blowing cold air across you increases the rate you convect heat to the cabin, which warms the cabin up, but probably doesn’t make you any happier. Also, the best way to warm up fast is to put more gas into the engine, which means driving under load, probably in low gear. So if you’re going to do the test, you might want to do it under both driving and idling conditions.

  8. Agree with #4 and #6: Your method may give a head start in total time required to warm the car. But until it’s appreciably warmer, the occupants are sitting in a draft of cold moving air, experiencing greater wind chill and hence more discomfort.

  9. Until the coolant in the engine reaches a set point, at which time the valve opens and allows warm water to reach the heater core that feeds heat to the interior of the car, it really doesn’t much matter what the blower, defroster settings are. Having the blower, and air control door in the duct, preset and running, ready to direct the heat when the warm coolant is allowed into the heater core, may speed things up a bit by allowing the system to do its thing ASAP. If the goal is clearing ice as soon as possible this is the way to go but it may only speed things up a few seconds. Even then if you manually activate the blower and door exactly when the valve lets heated coolant into the heater core there is no difference.

    On the other hand waiting allows the system to have clear and obvious heat output when you work the controls and a sense of responsiveness that is lacking when the defroster is blowing air that is warmed but not so much that the temperature rise is obvious. If your desire is to feel in control waiting and manual activation is far more satisfying.

  10. Until the thermostat opens (usually when the coolant temperature reaches 200′-215′), coolant is circulating through the engine and the heater-core (the little radiator in your dashboard) and absorbing whatever heat is generated from combustion and not converted into energy or lost as heat through the exhaust. The coolant will immediately begin to rise in temperature, and transfer that heat through the defroster, beginning to defrost the windshield immediately.

  11. If you’re interested in heating the air, blast the defrost as soon as possible.

    If you’re interested in not leaching heat off your face, blast the defrost only after it warms up.

  12. Warm? This is physicist talk already?
    Go to the coop or Tractor Supply Co. and buy an RF temperature set, less than a USB rig would be, maybe $10-15. I use them to measure temperature in my crawl space to know when to ‘drip’.
    My recommendation is to let the vehicle heat up and then switch to defrost. Two reasons: the windshield ice will melt faster; and you won’t be shivering as you drive down the road waiting for the car to heat up inside. The latter is the critical factor.

  13. Kate is correct old sport, sorry about that.

    I took Automotive Engineering as an elective in attaining my BSME and taught a class in it, so unless my textbooks and professors were wrong here is the answer.

    First, how long does it take an automobile’s internal combustion engine to achieve steady state? Answer: half an hour. Although long before that, within say five minutes you’re close enough, say 85-90% give or take depending on the particular engine, which is why a five-minute starting time is usually enough to have heat flow into the passenger cabin assuming an idle engine start.

    So we return to the main issue which is heat transfer (conduction, convection, and radiation). If you turn the heater on right away you’re carrying air away from the engine compartment via convection. If you leave it off as Kate says the heat remains except for the small amount conducted then radiated away via the hood and sides, etc.

    Now, as a physicist you might think the Delta-T (change in temperature) would be aided by the convection. Congrats, you would be right! But the greater delta-T is offset by the fact that the engine heats up from the internal combustion chambers out, and by keeping the heater off at start-up, the retained heat is ALSO heating the engine from the outside (inside the engine compartment) in, and the engine heats up faster.

    The difference in how fast the passenger compartment heats up with the heater on at idle or not is minuscule compared to the faster time in engine warming. You might also wish to call a colleague who has had to run equipment at the South Pole or elsewhere in Antarctica to confirm this; the importance of proper warming procedures is especially critical there.

    I should very much like you to run the experiment however as I’m an Experimentalist myself and being from the original show-me state (NJ) it is always nice to see confirmation.

  14. This sounds like something the Car Talk guys would know. Yeah, I know it’s not the scientific way to settle the issue. But it would be pretty cool to hear Prof. Orzel ring in with a “My wife and I have a disagreement . . .” question while I was listening to Click and Clack.

    BTW, I’ve used a couple of their puzzles in my classes over the years as a demonstration of the admonition that it’s not enough just to know the formulas and techniques. Occasionally, you’ve got to use your head too 🙂

  15. I toyed with the idea of simulating this mathematically, but a few minutes’ thought show that it’s a horrible problem. I don’t know how the temperature of the defrost air increases over time, or how to model the mixing with the air that’s already in the car (which presumably exits in some manner, because running the defrost on full doesn’t overpressure the car to any significant degree.

    If only more people who played economists in the funny papers understood this concept. Just my two seconds of snark. But anyway, to add to the examples already given of confounding variables, I generally walk to work unless it’s raining (I don’t particularly mind the cold per se, but I hate squelching around in cold wet socks the rest of the day). Why does this matter? Because if I start up the defrost right away[1] and it’s a cold rain, the cold air will fog the windshield. While this doesn’t directly concern how fast the passenger compartment will heat up in either scenario, it does have a second order effect: since you’re not going anywhere until the windshield clears you can either wait for the engine to warm up, or you can start driving immediately with the defrost off, in which case the engine will warm up faster due to the increased load.

    I like this question – it turns out to be much more complicated than what a first reading would lead you to expect.

    [1] The default setting for the vents in my car in the winter is to have them directed at the windows so that the (theoretically) warmer air isn’t felt as a cold breeze by the passengers. After the engine warms up, the girl in the shotgun seat usually changes the vent settings so that a warm breeze is blowing directly on her.

  16. I’ll throw another couple of variables into the pot. Are you trying to clear up a fogged up windshield? When a modern car’s defroster is turned on the A/C also comes on. This is because it dries the air out before blowing it at the windshield and evaporates the collected moisture faster. Iced up windshield work the same way, just not as fast. I solved the problem by installing a remote start and leaving the defroster and heat on when I park. It’s much cheaper than a divorce.

  17. Don’t forget to RTFM; on some cars the “defrost” button turns on the air conditioner to dehumidify the air, while a few actually have electric heat.

    I don’t understand the comments about the air in the engine compartment. The air entering the car is drawn off of the front edge of the windshield (a high pressure area when the car is moving) in a place that is isolated from the engine compartment. Drawing air from the engine compartment is asking for death due to CO if you got a minor exhaust leak somewhere.

    IMHO, the correct answer is somewhere between @7 and @10, depending on how the heater core is connected to the system. My understanding is that coolant always circulates through the heater core and the only valve is the one to the radiator, but even that could vary with designs because of the importance of engine temp to pollution management. All I know for a fact is that the heater does cool the engine. I can see it happen on the temp gauge.

    What I do is run the fan on low with the temp control on cold until the engine temp gauge comes up off the “C”, then switch it over to heat. This is long before the engine gets to 215, by the way.

  18. For me the issue is rapid recondensation on the windshield (either inside or outside surface). I scrape or squeegee the windshield, but the first block I have to drive directly into the sun. Even a small amount of recondensation (can happen in seconds), with the sun shining on the windshield, but not the obstacles on the street can render visability too low for safety. Sometimes I limp to the intersection, drive 100feet stop, get out squeegee, dribve another 100 feet. Once I make the stopsign I’m home free, shade plus not going towards the sun solves the problem. If recondensation is on the inside surface, than the humidity of the defroster air is an issue. It can also help to not be inside the car breathing out humid air while it warms up.

  19. If your car has A/C or climate control, then when you set it to the heat setting it will not matter how hot the engine coolant is. This is because A/C works on the pressuraisation of R134a and its susbequent return to a gas. When pressurised to a liquid it gains heat which can be used to defrost a car. Put your and on the grill at the back of your fridge for proof that heat is generated.

  20. #10 is correct. There will be some heat added to the coolant very soon after you start the engine, and if you turn on the defrost/blower, this will be blown over the windshield. That will help significantly with defrosting, especially the interior, but also the exterior. However, it will feel cool to the occupants until the engine warms closer to its operating temp.

Comments are closed.