The obvious answer that I just can't find.

Ad: This forum contains affiliate links to products on Amazon and eBay. More information in Terms and rules


Oct 31, 2022
I've been reading through the development of reciprocating aero-engines in the 1930's & 1940's & been left wondering why a particular development path wasn't pursued. To preface this I'll freely acknowledge that I'm not an engineer, so I realize there must be *some* good reason why it wasn't ... I just can't find out for the life of me what (was surely to the designers at the time obvious) that was.

The ability to remove heat was a critical limitation on the ability of parts to function within engines, with evaporative cooling examined then ethylene glycol, then pressurized ethylene glycol etc tried.

Designers were happy to use Sodium cooled exhaust valves, Sodium cooled piston designs were already patented for use in diesel engines ... did anyone think of casting Sodium heat pipes directly into the engine block/cylinder walls & plumbing that into a radiator?

Too costly? Too fire prone? Too corrosive? Too difficult to cast or too difficult to coat the channels within the cast with something compatible with each metal? Lack of sodium pump development? Doubts about adding another item to the logistics chain? Any of these seem plausible enough, I've just found a vacuum of information on anyone even considering it.

Thanks if anyone can shed some light on this.
The melting point of sodium is 209F so when your engine cools the sodium in the pipes would be a solid until you get it to
go to 209F + before it can run in pipes.

It is also dangerous if a leak occurs plus makes it hard to remove parts if the coolant pipes etc contain a solid.
I can not say for certain why the scientists/engineers of the time did not pursue the use of metallic sodium for engine coolant, but it was most likely due to the high melting point of 208°F/98°C.

In order to use a liquid to cool an engine it has to be able to move fairly quickly. Since sodium would be solid when the engine was first started, the various parts of the engine to be cooled would have to reach the melting point of sodium before the sodium could begin to move. The problem with this aspect is that while the sodium near the cylinders might turn to liquid quickly enough to be pumped away (thereby taking the unwanted heat away) the sodium a little bit further away would still be solid. The solid sodium would of course block the flow of the melted sodium. This means the entire engine would have to be above 208°F/98°C before the sodium could begin to flow through the engine. The problem also applies to the sodium in the coolant radiator and transport hoses. The heat from the cylinders would have to be conducted through the liquid sodium in the engine to the solid sodium in the transport hoses and the radiator and in turn melt the sodium in the radiator. This would take a significant amount of time, too long for the survival of conventional engine lubricating oil. (In general the temperature around the cylinders has to be kept well below 575°F/300°C so the lubricating oil does not boil in the transport channels, reducing/eliminating the lubricating value.)

A method for gradual heating up of the engine to operating temperatures would be doable (I think) but would have to be more sophisticated & complex (I think) than the relatively simple methods used in the water/glycol cooled engine. Also, a method of controlling the start-stop of the coolant pump would have to be implemented, different than the simple PTO used on the WWII engines, so that the coolant pump did not try to turn before the sodium melted throughout the entire system and was able to flow, and would stop the pump turning before the sodium returned to solid form again. Effectively monitoring/controlling a system like this would (I think) have been problematic with WWII technology.

The second part of the problem that I see is that while water/glycol coolant is liquid at room temperatures and remains so at temperatures down to about -50°F/-45°C, metallic sodium would begin to freeze whenever the temperature anywhere in the system went below 208°F/98°C. This means that the sodium cooled engine would have to be run at much higher temperatures (say about 250°F/121°C minimum to say about 500°F/260°C maximum) than the water/glycol engines, in order to prevent cold-spots and sudden blockage. This would require the materials used in the engine to be able to withstand the higher operating temperatures, so probably no plastic or rubber seals, and a different type of lubricating oil.

Disassembly, maintenance, and repair would also be significantly more difficult than for a water/glycol system.

I am not familiar enough with the chemistry of sodium and other materials what are commonly used in engines to comment on corrosion and such.
What ThomasP said plus there is the matter of the coefficient of expansion of sodium. It is over 60 years since I did year 12 so I have no idea of the actual rate but if it expands slower than aluminium then when the engine cools after the sodium solidified the sodium would cause the aluminium to fracture.
I always thought a Sodium valve worked by the Sodium taking up the heat of the hottest part of the valve and physically moving it to the coolest part many times a minute, the action of the valve makes this possible, would it work as well anywhere else? Sodium is both expensive and dangerous.
Part of the usefulness of metallic sodium as a coolant is its thermal conductivity value. While liquid water has a higher specific heat (ie how much heat energy the material absorbs per increment in temperature) of over 2x that of liquid sodium, liquid sodium has a thermal conductivity (ie how fast it absorbs/transmits/releases heat) of around 140x that of water. Incidentally, the thermal conductivity of liquid sodium combined with the mechanical movement of the sodium in the valve stem (as pointed out by pbehn above) is what allows the sodium cooled valve to survive the higher temperatures involved.

This means that liquid sodium can be pumped through the cooling channels much more quickly than can water while still absorbing the heat, and the heat absorbed by the sodium will be much more quickly conducted/transmitted to the radiator fins. Assuming the radiator materials and design can transmit the heat quickly enough this could be a significant advantage. It would allow a radiator with smaller cross section, a greater usable (efficiently) radiator depth, and consequently a significantly higher air velocity through the radiator - resulting in less drag due to the radiator cross section and the pressure head/reduction in air velocity in the radiator intake duct. You could possibly get by with significantly less coolant volume also, although since thermal conductivity of aluminum and copper used in the walls of the engine and the radiator are higher than that of sodium, and the density of liquid sodium is ~7% less than water, I do not think you would gain much in aircraft performance by reducing the weight of coolant.

If I am thinking about this correctly, you could end up with a radiator cross-section of about 25% of the equivalent water cooled radiator.

But the other problems mentioned in the thread would (I think) make the use of sodium not worth the potential trouble.

Users who are viewing this thread