OK, 50% better efficiency than a gas engine gets us to 45%, which is where we are with today’s diesel engines, a computerized fuel rack and a turbo. It doesn’t exceed Carnot limits, and I don’t see how that engine design could exceed Carnot limits either. It’s still a suck-squeeze-bang-blow engine when you get down to it. Cleverly done in many fewer parts, much lighter in weight, but still a SSBB engine.
If we make some use of Seebeck effect units plastered on the engine block, radiator and in the exhaust stream, we could capture some useful work out of the rejected heat on a diesel and go yet higher. There were some ag engineering students in the midwest who had a bright idea of using Seebeck effect modules in the exhaust stream of a diesel tractor engine to power Peltier effect units to run the cab cooling, which would then eliminate the need to an A/C compressor on a tractor. And there’s a lot less to go wrong than a typical freon A/C compressor system - no moving parts.
In general, most internal combustion engines use a third of the available thermal power in their fuel, they reject about a third of the thermal energy into the radiator/heat off the block, and about a third goes out the tailpipe.
A turbo captures some of the energy that had been going down the tailpipe. If we could use the heat rejected into the coolant and exhaust stream to do something useful, we’d be getting a good deal further down the road with existing engines or even new engine designs.
Getting better efficiency than a Otto-cycle gas engine is easy. Miller cycles or Atkinson cycles get us there, going to direct injection compression ignition with gasoline (ie, a diesel engine without the diesel fuel) gets us there.
Getting better efficiency than a theoretical best possible Carnot cycle at our current ambient temps (which would be in the low 60’s of percent)... not quite as easy.
I linked that engine because I just read about it last night and found it interesting. It should get better efficiency with much less weight.
How much of the time does the best designed engine spend in it’s peek torque band? Not much, which means that the engine will not provide advertised efficiency. Fuel cells are about 60% efficient now, absolute, not Carnot, and the energy not converted into electricity is rejected as high quality heat. Perfect for thermoelectric conversion since you brought it up.