Turbo compound engines extract some exhaust energy by using it to power a turbine connected to the driveshaft. Why are these engines not more widely used?
There are a number of factors that have limited wider adoption of Turbo-Compounds and it all comes down to how they actually work and where they receive their greatest benefits. In the early studies of the concept by NACA in the 1940's they determined that the most critical factor is the Exhaust Exit Pressure (Pe) to Intake Manifold Absolute Pressure (MAP, or Pm) ratio as this determines the Effective Mean Exhaust Jet Velocity, Ve(eff), available to the turbine power recovery stage. This means that a Turbo-Compound produces the largest benefits at low values of the ratio (low Pe, high Pm), making them ideal for use in conditions with either exceptionally high MAP or exceptionally low Pe. Excluding other exhaust recovery stages (such as a following turbo-supercharger), the Pe experienced by the Turbo-Compound turbine--the Power Recovery Turbine--is normally the same as the static atmospheric pressure (Pa). This means they make a lot of sense in Aircraft which operate at high altitudes with low Pa thereby naturally lowering the Pe:Pm ratio, and therefore increasing the Ve(eff) and the power available to the Power Recovery Turbine.
With this in mind, the limited efficacy of the Turbo-Compound in most road applications should be apparent. Since cars generally operate at or near sea level pressures they have a higher Pe, which increases the value of the Pm:Pe ratio and reduces the amount of recovery possible with the turbine. In order for a car to receive more than marginal benefit from a Turbo-Compound it would need to operate with a continuous boost in the MAP, and fairly significant one at that. This means it first must use Forced Induction (Supercharger or Turbocharger) and second, that the blower maintains a steady MAP of 30% or greater boost over Pa at normal operating conditions. Since most road cars operate at partial throttle under almost all conditions this means that it needs to have positive boost under those partial throttle conditions, which isn't a very common arrangement due to strain on the engine and the heat produced. This is why it is really only seen in Compression Ignition engines, Diesels, which operate within a limited set RPM range the vast majority of time at speeds which produce sufficient airflow for effective cooling (both charge-air and coolant). As others have pointed out, they could theoretically be of use in racing as well, especially long-track/oval racing, where the car is operated at continuous RPM and high-speed for extended periods. Another place where I could see good benefit to Turbo-Compounding is in Pike's Peak Hill Climb racing--short, full-throttle, high-boost sprints at high elevation.
All of these factors: the requirements to get the sufficient benefit from the system; the added cost, weight and complexity; the potential reduced reliability; and other available technologies; have relegated the traditional Turbo-Compound to a largely historical curiosity of "what could have been."
On the other hand, we are seeing variations of the idea. In the current F1 era they use something resembling Turbo-Compound which stores the recovered energy as electricity instead of applying directly back to the crank. Ferrari is developing a similar system for use in their next generation of sportscars as well but will use the recovered power to drive an electrical supercharger as a sort of uncoupled turbocharger set up. Either way, the power recovery is used more for electrical generation that anything and the electricity is then used to produce work through independent motors.
There are a lot of reasons, but the main reasons are increase cost and decreased reliability. The power they produced also cost engine efficiency as they increased the back pressure in the exhaust line.
So when you design an engine you have a few choices. For instance, you can either increase engine efficiency, or sap energy from the exhaust line. If you employ variable valve timing, for instance, you can extract much more of that exhaust energy and put it into the crankshaft without a turbo. If you instead add a turbo, you've added another set of gears, more moving parts, taken up more space, and added more weight.
Now the turbo might be slightly better at extracting energy than the variable valve timing, but does it offset the additional cost, weight, and reliability?
So most engine manufacturers don't employ turbo unless there's a reason for it.
For instance, some races have a limit to the number of cylinders and cylinder volume, and racers do care more about that extra 0.1% efficiency than they care about reliability since they have a team of mechanics working on the vehicle. So it makes sense there.
It doesn't make much sense on a standard consumer vehicle where the cost/weight tradeoffs relative to performance can be more effectively employed elsewhere in the engine.
Turbo compounding is used currently in truck engines by Detroit diesel and Volvo, but they are suited to long haul applications where the engine is at a constant revs/load, the power turbine it is said extracts up to 50 horsepower to put back into the crankshaft but the trade off as explained earlier is increased exhaust backpressure meaning the pistons have to push harder to get the exhaust gas out, so in transient (engine up and down the rev range) operation the power turbine costs energy, in a steady state it can reduce fuel consumption 2-5%
I limit my comments to aircraft:
turbocompounding, as used in the R3350, was a way of extracting the kinetic energy of the exhaust gases with a turbine and putting it back on the crankshaft. Exhaust gas backpressure buildup was not a design limiter for cruise at 25,000 feet. Today, the exhaust gas energy recovery turbine is used instead to drive the supercharger, which on the 3350 was driven off the crankshaft, making turbosupercharging less complex than turbocompounding.
The question is moot for aircraft engines bigger than 600SHP or so, where turboprop engines have taken over. they are far more reliable in the 1500-2000 SHP range than the R3350 or the R4360 as well as having better power-to-weight.
As such, turbocompounding was a bridge technology sitting between piston engines and turboprop technology. Once engines like the Allison 501 came along, 18 and 28 cylinder piston engines, with or without turbocompounding, were finished.
Modern compound engines use electric motors to translate the speed of the turbine to the speed of the wheel or propeller. For this, modern electronics is needed. It works in F1 and in the trucks and would work in hybrid cars with a small internal combustion engine. Turbines have a narrower working range than a piston engine. Cars most of the time idle or need full power for acceleration.
I wonder if it is possible to use a small compression ratio piston engine and a burning chamber (covered with a catalyst material, padded in insulation), where temperatures are low enough to reduce NOx, but still high enough to burn soot and CO. These temperatures aneal and thus flatten the catalyst's surface (so called pre-cat) and thus are avoided in usual catalyst converters. The high temperature, high pressure exhaust can be used to drive a two stage turbine. Note that this turbine housing costs more than a housing for relatively cold diesel exhaust. Also the first stage may need Nickel blades and only the second stage can have variable guide planes. A Cyclon (or even a filter) in the chamber may help to prolonge the duration of soot particles in the chamber.
All in all this is stuff for a premium hybrid car or truck. Agilty is due to the electric motor driven by a super-capacitor. FADEC enforces high temperatures at the filter in order to be never congested by soot. Excess energy is dumped into the batteries.
I’ve been doing a lot of reading on this and what we haven’t seen is a form of turbo compounding but instead using a small turbine engine capable of running itself when a steady torque is required. If married to a reciprocating engine with variable valve timing, the valves could be held open; the turbine generating the power. Under acceleration the turbine could operate as a turbocharger using bleed air intercooled as boost.