Thermal Siphoning (The Natural Pumping Action)

[Old Henry]

My engine was saved tonight by thermal siphoning.

Drove up the canyon to dinner. On the way back my generator wasn't generating and my temperature gauge was maxed out. Didn't worry too much about it cause it was just starting to get dark and I could make it down the canyon with just my parking lights on. The temperature gauge had me puzzled though because, although it was at the max going up the canyon, it should have been well below max coming down.

Anyway, made it home, lifted the hood, and, low and behold, my generator/pump belt had broken! That's why the generator wasn't generating and why the engine was so hot. Then the real mystery - why didn't the engine overheat enough to boil the water out when the water pumps weren't pumping any water through it at all? I checked temperatures and the radiator was 217, 20 degrees above where it usually is in the summer, but the engine block was only 208! Why? Obviously water had been flowing through the radiator enough to cool the engine some and nothing was at boiling temperature (with my 4 lb radiator cap).

Only one explanation - thermal siphoning - the natural tendency for cold water to flow down and hot water to flow up. (The same principle is used to keep the arctic tundra frozen for the Alaskan pipeline to sit on.) As the water in the radiator cooled down it fell down toward the bottom of the radiator as the hot water in the engine was rising up into the top of the radiator! Circulating it naturally the same direction the pumps would have! Certainly a lot slower but it still did the trick!

A mighty darn serendipitous event! Were it not so I might have wiped out my flathead on that one 30 minute trip.

Who'da thunk it?

I'll certainly be adding some belts to my trunk spare box.
I'll also be stopping to check things out next time the generator quits generating and the temperature gauge pegs to the max.

[cmbrucew]

Old Henry
Early four cylinder contintenal engines used no water pumps. they were called perculaters. later blocks were fitted with a pump.

Bruce

Works good
Lasts long time

[ford38v8]

Old Henry, you musta done gotcher ingineer skoolin' with the 'riginal Old Henry, cuz he done did that stuff in his Tin Lizzie!

[Old Henry]

Originally Posted by ford38v8

Old Henry, you musta done gotcher ingineer skoolin' with the 'riginal Old Henry, cuz he done did that stuff in his Tin Lizzie!

Thanks Mater!

[Mike51Merc]

Henry,

Early home hot water heating systems had no pumps either.

[fomocoloco]

Same thing happened on my 36 last weekend. Broke the belt and drove 20+ miles. Never got over 200 degrees. Now I know why. Thanks Old Henry.

[David J]

John Deere Tractors didn't have waterpunps [ the big Waterloo built ones ] until 1952 . You can actually see the water circulate under the rad cap . This system is pretty fool proof and very low maintenance . My Farmall tractors from the mid 30's don't have waterpumps either . No deez ain't fords but they are examples of what da counsellor's thread is about .

[Rebel Sympathy]

Hmmm! Does this reopen the old debate between: A) Speed up the water flow with better designed pumps, or B) Slow down the water flow with restrictors, or by knocking off every other blade on the impeller?

BTW: Farmall Super A tractors do not have water pumps either. (4-cylinder 113 cubic inch engines, 2200 rpm.)

[mrtexas]

I believe Model T radiators have a larger tube size besides having no water pump? This was mentioned in my college heat transfer class about natural convection(thermosiphon) to make then thermosiphon work.

[Old Henry]

Originally Posted by Rebel Sympathy

Hmmm! Does this reopen the old debate between: A) Speed up the water flow with better designed pumps, or B) Slow down the water flow with restrictors, or by knocking off every other blade on the impeller?

I've never thought moving water through the engine any faster than stock pumps makes any real difference in the overall average engine temperature. Here's why:

There is a physical limit to how fast heat can transfer from the engine to the water flowing through it and how fast the heat from the water can transfer to the air flowing through the radiator. That limit is reached when the temperature at the bottom of the radiator is equal to the air temperature flowing through it. That don't take much flow. Once that limit is reached the heat's not transferring any faster and yer at the optimal heat transfer rate and will find the temperature of bottom of the radiator rising as the speed of the water flowing through it increases. The only difference made by moving the water faster is that the difference between the temperatures at the top and bottom of the radiator would be less just like the differences between the top and bottom of the engine would less. That's why when you move the water faster the temperature measurements at the radiator hose neck of the heads is less as well as the temperature readings from the senders at the top of the engine. To think that's the temperature of the whole engine is as fallacious as thinking that the temperature measured at the bottom of the radiator is the temperature of the whole radiator. You know that ain't so. So, overall, if you measured the temperature of the bottom of the radiator and averaged it with the temperature at the top of the radiator it would be the same at any water speed above the minimum needed to maximize heat transferred to the air passing through the radiator by the time it reaches the bottom of the radiator.

Don't believe it? Do the testing. You'll see.

Here's how to conduct the test:

Beginning with the stock water pumps, get the engine up to operating temperature so that the thermostats are fully open and the water is flowing at the max (or, better still, just leave the stats out for the test). Record the RPM's to be able to duplicate it later. Now measure the temperature of the bottom of the radiator (which will be the same temperature as the water going into the bottom of the engine) and record it then measure the top of the radiator (which will be the temperature of the water coming out of the top of the engine). Calculate the average of the two temperatures and record it.

Now, install the fastest water pumps you can find and repeat the test exactly the same as before (including identical RPM's) and compare the average of those two temperatures with the average you got with the "slow" pumps.

As you will see, with the "fast" pumps, the high and low temperatures will be closer in value but the average of the two will be virtually the same as the average of the two temperatures with the "slow" pumps that will be further apart in actual temperatures between the top and bottom of the radiator.

Let us know how your tests go.

"But," you say, "my engine is cooler when running faster than when idling," and you think that's because the water is moving faster.

First of all, keep in mind that your temperature senders are at the top of the engine that will be hotter when the water is flowing slower, not at the bottom that will be cooler when the water is flowing slower. So, it's a little deceiving to think that the whole engine is hotter when the water's flowing slower just because the top of the engine is hotter.

Second, when you speed up your engine, what else flows faster besides the water in it? Right, the air flowing through the radiator. That's a variable outside of the speed of the water flowing through the engine that will change the overall engine temperature. The faster the air flows through the radiator, the faster the heat is transferred to it from the water inside of it. So, in fact, the engine overall temperature is cooler when the engine is running faster but not because the water is flowing faster but because the air through the radiator is flowing faster.

Make sense? Think about it.

Them's my musings on the subject for today.

(Watch out! Here comes you know who!)

[mrtexas]

Not to disagree Old Henry (except for the last comment) but consider this:

Heat transfer from an engineer's point of view is defined as:

Q = m*Cp*delta temperature

Q rate of heat flow
m flow rate
Cp heat capacity of fluid
delta temperature cool side minus hot side

This would be both for the air/radiator and the water/engine block

I don't see anything about time in that formula, rather the rate of heat transfer

Another consideration would be whether any air in entrained in the water circulating thru the engine. If air is entrained the heat capacity is lowered and the engine runs hotter. How much is entrained would be difficult to determine.

With no air entrainment, the faster the pump runs the more the heat transfer and the cooler the engine.

"So, in fact, the engine overall temperature is cooler when the engine is running faster but not because the water is flowing faster but because the air through the radiator is flowing faster."

It is true because of both factors, see flow rate in formula.

"There is a physical limit to how fast heat can transfer from the engine to the water flowing through it and how fast the heat from the water can transfer to the air flowing through the radiator. That limit is reached when the temperature at the bottom of the radiator is equal to the air temperature flowing through it. That don't take much flow."

Really? You have measured the water temperature at the bottom of the radiator? I doubt the water temperature is as cool as the air.

[Old Henry]

Originally Posted by mrtexas

Really? You have measured the water temperature at the bottom of the radiator? I doubt the water temperature is as cool as the air.

I did measure the temperature of the bottom of the radiator and it was, of course, hotter than the air going through it. But, there is a speed to which you could slow the water flow such that by the time the water got down to the bottom of the radiator it woud be as cool as the air flowing through it. That would be very slow flow such as the flow last night when my water was circulating by thermal siphoning only. That's what made it work. The water was so cold coming out of the bottom of the radiator going into the engine it then was heated fast enough to rise in the engine and keep it circulating. It is that principle that high efficiency furnaces use to "milk" the last drop of heat out of a flame before exhausting it. Instead of just using the flame burning through the heat transfer section then letting it go up the flue with still 40% of the heat unused, they use a fan to draw the flame slowly through a labyrinth of metal tubing that the air is blowing past until, by the time it leaves the tubing, it is only slightly hotter than the air it was heating. That's why it can then be exhausted out through PVC pipe instead of having to go up a metal flue. I've had both furnaces and the heat coming out of the vents in the house is much hotter and the air being expelled from the furnace is much cooler with the new furnace than with the old 60% efficient furnace.

[Old Henry]

Originally Posted by mrtexas

Not to disagree Old Henry (except for the last comment)

It's not you I'm watching for.

[mrtexas]

Originally Posted by Old Henry

It's not you I'm watching for.

Well, it's a friendly disagreement. I don't think the lower heater hose water temperature ever gets to the air temperature in a flathead, at least not in mine.

Originally Posted by mrtexas

Heat transfer from an engineer's point of view is defined as:

Q = m*Cp*delta temperature

Q rate of heat flow
m flow rate
Cp heat capacity of fluid
delta temperature cool side minus hot side

This equation also explains why adding antifreeze causes your engine to run hotter in a non pressurized radiator. The heat capacity of antifreeze is lower than that of water which means the rate of heat flow is less>hotter engine.

Originally Posted by Old Henry

I've never thought moving water through the engine any faster than stock pumps makes any real difference in the overall average engine temperature.

Now, install the fastest water pumps you can find and repeat the test exactly the same as before (including identical RPM's) and compare the average of those two temperatures with the average you got with the "slow" pumps.

Well this could only be true if the faster water pump was cavitating(doubtful) or entraining air in with the coolant and reducing it's heat capacity. Again, the rate of heat transfer is directly proportional to the flow of the coolant.

Anyway, pardon my engineers smart alleck-ness. I enjoy discussing engineering occasionally and being a retired chemical engineer don't get the chance often. Heat transfer is one of the core chemical engineering subjects.

[Old Henry]

Originally Posted by mrtexas

With no air entrainment, the faster the pump runs the more the heat transfer and the cooler the engine.

Perhaps that supports the use of pure distilled water to cool better than tap water that has other than straight H20 in it?

[Old Henry]

Originally Posted by mrtexas

Well, it's a friendly disagreement. I don't think the lower heater hose water temperature ever gets to the air temperature in a flathead, at least not in mine.

The water certainly never cools to air temperature by the time it gets to the bottom of the radiator at the speed even the stock pump pushes it through.

I think the bottom line still remains: Running the water faster may not cool better but it can't hoit nuttin' either. It would, at the very least, even out the temperature some. Might as well do it "just in case" (if you like puttin' yer money into a lot of "just in case" kinds of stuff. I admit I probably do more than some.)