Thermostats

[ardiesse]

Rosco,

"Fools rush in . . . " Here goes. Applied Thermodynamics 101.

The coolant has a certain specific heat (and specific heat is expressed in units of energy divided by volume divided by temperature. For water it's 4.2 kilojoules per litre per degree C.)
When the coolant flows through a radiator, there's a temperature difference between the inlet and the outlet. Outlet is cooler than the inlet, and heat energy is dumped into the air. If you have a flow rate of say n litres per second, and a temperature difference of T degrees, the amount of heat energy removed by the radiator is 4.2 kilojoules/litre/degree C times n litres per second times T degrees C. The "litres" units cancel, the "degrees C" units cancel, and what you are left with is 4.2 times n times T kilojoules per second, a.k.a. kilowatts. If the radiator has to dissipate a constant thermal power, the flow rate and coolant temperature drop must be inversely related to each other.

Let's plug some numbers in. For convenience, let's say that the radiator has to dissipate 42 kW. If your coolant flows at a litre per second, the coolant's temperature drop will be 10 degrees C. (4.2 x 1 x 10 = 42.) If your coolant flows at ten litres per second, the coolant's temperature drop will be 1 degree C. (4.2 x 10 x 1 = 42.) The faster the coolant flows through the radiator, the smaller the temperature difference between outlet and inlet. Conversely, the more slowly the coolant flows through the radiator, the greater the temperature difference between outlet and inlet.

Note that I am talking about the coolant temperature difference between the radiator's outlet and inlet here.

There is a second quantity to be considered: the radiator's thermal resistance. This is expressed in units of degrees C per kilowatt. It tells you how hot the coolant will get if the radiator has to dissipate a certain amount of thermal power.

More numbers. Let's say the radiator has a thermal resistance of 1 degree per kilowatt, and the engine it's cooling is dumping 42 kW into the cooling system. 1 degree per kilowatt times 42 kilowatts gives a temperature rise of 42 degrees above ambient. Using the example above, if the coolant flows at a litre per second, the water will come out of the radiator ten degrees cooler than it went in. If the coolant flows at ten litres per second, the coolant will come out of the radiator only a degree cooler than it went in, but still about 42 degrees above ambient temperature.

Finally, let's consider the effect that an 80 degree C thermostat will have on our cooling system on a stinking hot 40-degree day. You still have 42 kW to get rid of, the water pump can move ten litres per second, and the poor old radiator can only dissipate a kilowatt per degree C. The temperature rise above ambient (40 degrees) will be 42 degrees, giving a coolant temperature of 82 degrees. The thermostat will be fully open and the cooling system is operating at maximum flow, 82 degrees in and 81 degrees out. Now let's make it a gentle autumn day: 20 degrees. The coolant temperature would be 62 degrees if there were no thermostat (20 + 42 = 62). But the thermostat regulates the coolant temperature leaving the engine to 80 degrees by restricting the flow. The coolant temperature leaving the radiator will be still be 62 degrees (a bit of a fudge to preserve the 42 degrees above ambient), so the coolant temperature drop is 18 degrees, and because of the inverse relationship between coolant temperature drop and flow rate, the flow rate will be 0.55 litres per second.

It is all just a question of what the cooling system has to do to get rid of the engine's heat. Generally the limiting factor is neither the coolant's specific heat, nor the water pump's maximum flow rate, but the radiator's thermal resistance.

I am now running my FC without a thermostat. At highway speeds, the coolant is about 50 degrees above ambient, and the temperature drop across the radiator is about 12 degrees. Fifty degrees above ambient does not leave a great deal of headroom if I (say) go for a drive up to Brisbane in the summer.

Rob

[rosco]

Thanks for taking the time to study and explain the physics of this subject, Rob - it is very much appreciated.
I believe I have understood all of what you post - and understand that all of it is derived from applied science.

Thank you further for putting it into terms we can relate to - do you have any instructional training?... your post certainly reeks of it.

If I have understood what you have posted correctly, we have a number of variables which are determined by the specifics of their unique scientific properties.... ambient temperature, radiator efficiency and heat production being the main players... but also flow rates and efficiency, which is the bone of contention here.

I read, with interest that a reduction in flow rate achieves greater reduction in output temperature after cooling is effected by the element of the radiator... as in the case of perhaps the thermostat providing some restriction to the total ability of water pump flow rate.

I suppose, we would need to know if, in fact - the maximum flow rate of the water pump is by any way restricted by the thermostat when it is fully open in situ....

Further, I appreciate that this higher differential between admission and delivery through the radiator element in a restricted flow rate - may be inadequate (as in the case of your FC) to dissipate sufficient heat of which the engine is producing... and is why you run without a thermostat.

Those two statements produce an apparent paradox - but I understand exactly why you have gone to the lengths to explain relevance.

I believe I have understood what you post?.. and would be very much obliged if you expand on anything which you believe I have either misunderstood - or misinterpreted....


Thank you again, for taking the time to explain this to me... and the forum.


frats,
Rosco

[Errol62]

So it is the size of the radiator relative to the heat load that is critical, rather than the rate of coolant flow. If the radiator is too small, or clogged, the motor won’t cool adequately. If it is too big the thermostat will restrict flow, keeping the heated water in the block.


FB ute fixer upper, EK van on rotisserie

[EK283]

Well,

Interesting ! What about air flow ?????

Greg He He He !

[Harv]

There is a second quantity to be considered: the radiator's thermal resistance. This is expressed in units of degrees C per kilowatt. It tells you how hot the coolant will get if the radiator has to dissipate a certain amount of thermal power.

Yes, and no.

In a past life, I was a chemical engineer. Chemical engineers don't play with chemicals (wish they had told me that before I signed up)... they play with fluids flowing in pipes. I got to study a couple of years worth of heat transfer, then went to play on industrial kit for a couple of decades. A lot of it is water or air cooled. I've dissolved a fair bit of that knowledge in the anaesthetic that comes in 375mL brown bottles over those decades (surprising what you can dissolve in ethanol), but lets see how much I can remember.

Your average car engine is not much different than a water cooled industrial compressor. Big heat source at one end (the car's cylinders), linked up to a heat sink (the radiator) at the other end. They are linked up by a closed circuit water loop (a nice green mixture of water and glycol). In an old car, you would let the water circulate itself (thermosyphon). In our snazzy modern FB, we have a water pump. Its an open impellor centrifugal pump. We have a control valve in the circuit (the thermostat). Its a modulating thermostatic valve... it will open and close gradually with temperature (i.e. is not simply on, off).

Now here is the part where you need to imagine (the chemicals would probably have helped this... I should have taken a chemistry degree). Imagine you are a water molecule. You're life is one big circuit... water pump, engine, radiator, water pump, repeat. You probably have a bad sense of deja vu. You get hotter in the engine, and cool down a bit in the radiator. You get pressurised by the pump, and lose a bit of pressure in your journey through the engine and radiator.

The pressure loss bit is important. You lose pressure because you rub up against other stuff... the rubber in the radiator hose walls, the steel and gunk in the radiator, even your fellow brother water molecules to some extent. You also lose pressure because you have to make sharp turns (like the 90-degree elbow in the radiator hose) and bang up against the walls of the turns. The longer the journey you have to make, and the more turns in it, the more rubbing and banging you do and the more pressure you lose.

As you are losing pressure by rubbing on things you are also slowing down. Fluids rarely flow as one big plug. Your radiator does not understand communism… not all water molecules are equal. If you are an unlucky water molecule, your journey takes you near the metal/gunk of the radiator (or the engine cylinders). As you rub up against the metal/gunk and slow down, most of your brother water molecules zip past you and leave you behind. You (and the other poor water molecules up against the wall) form a thin layer that is still flowing, but is sluggish and lazy. Because you are in a sluggish, laminar layer, you are not contacting lots of other water molecules... just the ones next to you. The poor contact in your laminar layer of brother water molecules means that you don't have much chance to transfer heat to each other, and you act as an insulator. The fast, zipping water molecules in the middle of the radiator passage can't get close to the radiator steel/gunk to give up their heat, and you won't pass it on either. The radiator struggles to transfer heat.

Now imagine if someone makes the water flow faster. You could put in a bigger pump, or a high flow thermostat, or make the radiator hoses bigger, smoother or more gently bent... lots of ways to increase flow. Suddenly you are being shoved along the radiator wall a lot quicker. You are no longer sluggish and lazy... you are now turbulent. Your laminar layer of fellow brother water molecules has gone... you are all now bouncing around into the main flow and off the radiator wall. If there was a chemical engineer looking at you, they would calculate your Reynolds number. When you were lazy and laminar, your Reynolds number was less than 2300. Now that you are a vibrant, turbulent water molecule your Reynolds number is more than 2900. The picture of the candle flame in the Wikipedia article shows the turbulent vs. laminar flow change pretty well:
https://en.wikipedia.org/wiki/Reynolds_ ... transition

This is why Rob’s explanation is kinda right, but kinda not. The radiator does have a thermal resistance, but it is not fixed. You cannot multiply the fixed (degrees C per kW) by flowrate, because it is flowrate dependant.

All that bouncing around means you are rubbing/banging a lot more, and losing more pressure. You are also getting a lot more chances to touch the steel/gunk, and your brother water molecules. This is lots more opportunity to pass along heat between your brother water molecules and the steel/gunk. The heat transfer increases markedly.

But wait… now that I am travelling so much faster, I don’t have time to transfer all my heat, right? Not really. Heat transfer is not massively time dependant… it is much more dependant on laminar vs turbulent flow. Clay got it right earlier in the thread… each water molecule spends less time in the radiator, but many, many more water molecules are going through the radiator in a given period of time. You are all working together to transfer the heat (hmmm… maybe your radiator is a partial communist state… especially in a red motor nyuck nyuck nyuck).

For the doubters… imagine you are an air molecule (now where did I put those chemicals again?). The same thing happens on the air side of the radiator, except now you are trying to transfer heat from the steel/bug guts of the radiator to air, instead of between water and steel/gunk. Got a heating problem? The old theory says you need more “time in the radiator” to get better cooling. Try slowing down your fan, or putting in a smaller one and see what happens.

So what was the story in the high revving Cheving? Could be a couple of things. The centrifugal water pump is a basic bit of kit. It’s also not a communist state… not every water molecule is treated equally. Some water molecules flow through the pump and get pressurised. Some water molecule however get churned around inside for a few spins of the pump shaft before going out. The faster the pump spins, the more churn. If you get enough churn, the pump will start to cavitate. The water molecules in the pump get upset with all that churn, get hot, and boil off into little steam pockets. They then collapse back into water. The pump does not perform very well at all, and flow becomes erratic. In big industrial kit it sounds like someone has put a bucket of gravel down the throat of the pump… but very hard to hear in an automotive engine. Cavitation can also occur from high flow – if the fluid loses enough pressure on a tight bend, it can get to the point where it boils locally. Not as likely though in an automotive system, and does not affect the bulk water flow anywhere near as much as pump cavitation. By fitting a thermostat you slow the flow down, the pump comes back onto it’s pump curve, you have enough nett positive suction head (oooohh!! I can remember some stuff ) and does not cavitate. Slowing the pump down can do the same for cavitation.

So what is the thermostat for? It slows water flow right down (almost stops it) when the engine is cold. This lets the water around the cylinders heat up (very little flow… very, very laminar around the cylinders… crappy heat transfer… cylinders warm up) to give you better starting on cold days. As the water gets warm, the cylinders get too hot (you risk boiling the water and getting almost zero heat transfer… steam is a crappy conductor). The thermostat starts to open, you get flow, you get heat transfer and the system cools. If the radiator is big enough (and clean enough) it can transfer enough heat to keep up with the engine. The thermostat will modulate (open a little, close a little) to maintain constant temperature in the circuit. Great for performance, and emissions. If the radiator is undersized or clogged, the thermostat can’t keep up and stays fully open. You-get-what-you-get for water temperature, dependant purely on engine load and air flow past the radiator.

Cheers,
Harv

[Errol62]

There are many variables, and we tend to deviate from manufacturer spec in search of some form of improvement. I had a 192 in my sedan for many years of hard driving. 186S block bored 60 thou over. It had many woes that I worked on solving over the years. It would always run hot on the highway. I fitted a massive radiator, but the temperature never behaved quite properly.

One night I was towing the 6x4 with a precious upright piano in the back, from Adelaide to Kadina, about 140km. I didn’t want to jolt it so I sat on 40mph all the way. Despite the extra load, including my wife and father in law, the temp gauge sat on 82C, the thermostat rating, the whole way. Never happened before or since.

Eventually got sick of the gudgeon pins rattling away and replaced with a reconditioned standard bore 202 blue motor. Rock stock bar a crow towing and economy cam and disabled egr and pollution gear. Same radiator, in fact I reused the water pump as well. It struggled to get up to 82C at all on the highway.


FB ute fixer upper, EK van on rotisserie

[rosco]

Thanks Harv - I was hoping someone with formal qualification would come on board in this thread - you have explained everything very, very well...
And thank you for taking the time to put so much together in layman's terms (mine).

Between you and Rob, I believe you have both dispelled the myth that by having a thermostat in situ, the radiator works more efficiently... those molecules need to travel... reducing the flow rate allows the lazy little blighters to find somewhere out of mainstream to rest up (and get hot).

thank you both for your factual and scientific input... we can put this one to rest now, and I aplogise to the forum if I have put sway into the minds of those who place more faith in my knowledge than is warranted.

frats,
Rosco

[FireKraka]

Just read this and now I've got a head ache

That said what Harv is saying is very true re water pumps and flow, cavitation and water hammer/steam are much more apparent in the large shipping 2 stoke engines that I deal with. Cavitation and incorrect heat within the cooling water system creates issues like cold corrosion of cylinder liners and if temperatures are not right erosion of the water passages inside the engine blocks, of course it is easier for me to see when I'm dealing with an engine 4 stories high that I can climb inside to inspect

Very interesting discussion.
Neil

[Harv]

That said what Harv is saying is very true re water pumps and flow, cavitation and water hammer/steam are much more apparent in the large shipping 2 stoke engines that I deal with.

Have played with some Frame 9 GTs, and some 30MW diesels, but always wanted to play with the bigger marine stuff. Boats is cool

Cheers,
Harv

[ardiesse]

Harv,

So that's where the Reynolds number comes in.

I must admit that I was doing an "Ohm's Law"-level analysis, and then you weighed in with the equivalent of Maxwell's equations. (At work, the computational electromagnetics people used to say, "If you think electromagnetics is computationally intensive, you ought to try fluid dynamics.")

And in answer to Greg's question about air flow through the radiator: the more air you pass through the radiator, the more heat it can dissipate.

Rob