There is considerable speculation on the internet about the relative effectiveness of deriving thrust from the engine cooling air as postulated by the paper by F.W. Meredith in 1936. Much speculation revolves around the P51 cooling system layout and design and armchair estimates put the effect anywhere from 600 pounds of thrust to -200 pounds (drag).
The basic premise is that cooling air is decelerated and by a diffusion duct prior to the radiator, heat is added to the cooling air by the radiator and this causes the air to expand. The air is then converged and re-accelerated to exit faster than the free stream velocity which would give thrust.
Since we are talking subsonic flow here, I see one big problem with the premise. Once downstream pressure equals upstream pressure, flow must stop, therefore I see no theoretical possibility of deriving actual thrust as proposed. If we look at ramjet propulsion, efficiencies are very low until velocity exceeds about 300 knots. The ramjet involves burning fuel in the duct which raises its temperature far above what could happen with a radiator and 100-110C coolant. We might see an 80C rise if we were very lucky- about 1/10th of what a ramjet could deliver.
To be fair, Meredith's paper also discussed the possibilities of further heating the cooling air, post radiator, with engine exhaust gases. This might be able to effect enough temperature rise to have some useful effect. He also surmised that little could be gained at speeds much below 300 knots.
This topic has always interested me so when I had the opportunity to change the cooling system layout on my Subaru powered RV6A, I chose to fabricate a P51 "like" diffusion/ convergent radiator duct with variable geometry exit door and instrument it with inlet and exit pressure and exit velocity and temperature probes to try to quantify what was possible in the 150 knot ranges anyway.
Thoughts from any of the gifted aerodynamics experts here?
The basic premise is that cooling air is decelerated and by a diffusion duct prior to the radiator, heat is added to the cooling air by the radiator and this causes the air to expand. The air is then converged and re-accelerated to exit faster than the free stream velocity which would give thrust.
Since we are talking subsonic flow here, I see one big problem with the premise. Once downstream pressure equals upstream pressure, flow must stop, therefore I see no theoretical possibility of deriving actual thrust as proposed. If we look at ramjet propulsion, efficiencies are very low until velocity exceeds about 300 knots. The ramjet involves burning fuel in the duct which raises its temperature far above what could happen with a radiator and 100-110C coolant. We might see an 80C rise if we were very lucky- about 1/10th of what a ramjet could deliver.
To be fair, Meredith's paper also discussed the possibilities of further heating the cooling air, post radiator, with engine exhaust gases. This might be able to effect enough temperature rise to have some useful effect. He also surmised that little could be gained at speeds much below 300 knots.
This topic has always interested me so when I had the opportunity to change the cooling system layout on my Subaru powered RV6A, I chose to fabricate a P51 "like" diffusion/ convergent radiator duct with variable geometry exit door and instrument it with inlet and exit pressure and exit velocity and temperature probes to try to quantify what was possible in the 150 knot ranges anyway.
Thoughts from any of the gifted aerodynamics experts here?
