3 ms·

It's of course possible to design engines to be oil cooled, though water-glycol tends to be preferred due to about twice the specific heat capacity, meaning sma

by jabl 10mo ago

It's of course possible to design engines to be oil cooled, though water-glycol tends to be preferred due to about twice the specific heat capacity, meaning smaller coolant channels, radiators, and fans are required.

bluGill 10mo ago
I don't think it occurred to anyone in 1905 that a water/glycol mix might be good. They either used straight water with a warning to drain the engine when you shut down in cold weather so it didn't freeze, or they used oil. My 1939 tractor has instructions to start the engine and then pour water in the radiator when it is below freezing.
- jabl 10mo ago
  Not in 1905, no. I believe water-glycol mixes became widely used in the 1920'ies. But without glycol, water is an even better heat transfer agent. Shame about the freezing thing, though.
throwawaylaptop 10mo ago
Is it specific heat that we care about, or rate of heat transfer? Specific heat matters a bit, but if you make your coolant take twice the energy to change 1 degree, the same thing happens on the radiator side and you must release twice the heat to cool 1 degree. Rate of heat transfer in general if probably more important.
- jabl 10mo ago
  Well, it gets really complex. Yes, specific heat matters, but as you say so does the heat transfer coefficient. And the viscosity. And is the flow laminar or turbulent? Etc. There are various figures of merit, such as the Mouromtseff number https://doi.org/10.1109/JRPROC.1942.234654 https://doi.org/10.1109/JRPROC.1942.234654 or https://www.electronics-cooling.com/2006/05/comparing-heat-transfer-rates-of-liquid-coolants-using-the-mouromtseff-number/ https://www.electronics-cooling.com/2006/05/comparing-heat-t... for a quick overview. Some tables of heat transfer coefficients: https://www.engineersedge.com/heat_transfer/convective_heat_transfer_coefficients__13378.htm https://www.engineersedge.com/heat_transfer/convective_heat_... But, turns out water is just very very good also when you take these other factors into account. Compared to oil, it has, as mentioned, much higher specific heat, it has higher heat conductivity, it has lower viscosity which means less pumping power and more likely to see turbulent flow which helps with mixing.
- bluGill 10mo ago
  In the case of the oil pull tractor they needed to run the tractor much hotter than water boils - even if they had known modern antifreeze it still boils too hot. They also didn't know to make modern radiators - they usee very large radiators to make up for lack of surface area. (They also used exhaust to pull air through the radiator instead of a fan). All in all a very ingenious design - but there is good reason they don't make them like that anymore. (My 1939 is a lot more modern - and it lacks electric start)

by jabl 10mo ago

bluGill 10mo ago

I don't think it occurred to anyone in 1905 that a water/glycol mix might be good. They either used straight water with a warning to drain the engine when you shut down in cold weather so it didn't freeze, or they used oil. My 1939 tractor has instructions to start the engine and then pour water in the radiator when it is below freezing.

jabl 10mo ago
Not in 1905, no. I believe water-glycol mixes became widely used in the 1920'ies. But without glycol, water is an even better heat transfer agent. Shame about the freezing thing, though.

throwawaylaptop 10mo ago

Is it specific heat that we care about, or rate of heat transfer? Specific heat matters a bit, but if you make your coolant take twice the energy to change 1 degree, the same thing happens on the radiator side and you must release twice the heat to cool 1 degree. Rate of heat transfer in general if probably more important.

jabl 10mo ago
Well, it gets really complex. Yes, specific heat matters, but as you say so does the heat transfer coefficient. And the viscosity. And is the flow laminar or turbulent? Etc. There are various figures of merit, such as the Mouromtseff number https://doi.org/10.1109/JRPROC.1942.234654 https://doi.org/10.1109/JRPROC.1942.234654 or https://www.electronics-cooling.com/2006/05/comparing-heat-transfer-rates-of-liquid-coolants-using-the-mouromtseff-number/ https://www.electronics-cooling.com/2006/05/comparing-heat-t... for a quick overview. Some tables of heat transfer coefficients: https://www.engineersedge.com/heat_transfer/convective_heat_transfer_coefficients__13378.htm https://www.engineersedge.com/heat_transfer/convective_heat_... But, turns out water is just very very good also when you take these other factors into account. Compared to oil, it has, as mentioned, much higher specific heat, it has higher heat conductivity, it has lower viscosity which means less pumping power and more likely to see turbulent flow which helps with mixing.
bluGill 10mo ago
In the case of the oil pull tractor they needed to run the tractor much hotter than water boils - even if they had known modern antifreeze it still boils too hot. They also didn't know to make modern radiators - they usee very large radiators to make up for lack of surface area. (They also used exhaust to pull air through the radiator instead of a fan). All in all a very ingenious design - but there is good reason they don't make them like that anymore. (My 1939 is a lot more modern - and it lacks electric start)