What an excellent mechanical experiment; I really enjoyed reading about the device. Too bad this comes way too late to cause any excitement in the automotive world...and about two decades behind the move to EVs.

Now, if we can just get the thing to seal in the combustion...

Rotary engines ended up in a particular configuration for a very good reason. It works. There have been no other practical alternatives. The age old "gear engine" would have produced similar efficiency as quoted above, if we could only get the thing to seal...

It is not too late.

It is too little.

It is a research-grade prototype that competes with the Scuderi - for lack of progress.

58% brake efficiency sounds somewhat (?) overoptimistic to me. One should also realize that the diesel cycle is more efficient at a given maximum cylinder pressure than the otto cycle. The geometric compression ratio for the diesel cycle has to be somewhat higher than the 18:1 CR mentioned for this engine to get equal efficiency as the otto cycle. Here is when we start to get outside conventional wisdom, since there is an optimum CR for highest efficiency and we know pretty well in what range that is for diesel engines. Higher CR increases a couple of losses that higher thermodynamic cycle efficiency cannot compensate for. If we, on the other hand, envision an otto cycle running at CR or 18:1, this is not possible with gasoline as fuel due to the “low” octane number of this fuel. However, using direct injection of methanol can give an “equivalent” octane number in the range of 150-200, as shown by MIT, so CR in the range of ~18:1 (or preferably, optimum CR if it would turn out to be slightly higher) could be used. This gives us an efficiency on pair with conventional diesel engines and maybe even marginally higher. However, we are still speaking of efficiency in the range of ~43% for engines of this size, or maybe a few per cent more if we are really optimistic. Overexpansion is a very nice feature but this can also be achieved to some extent in piston engines and it does not boost efficiency to those numbers. Even if the assumed “optimum” thermodynamic features and overexpansion are combined, I still find 58% brake efficiency very, very optimistic. I will leave friction, gas exchange and other losses for later discussion.

@SJC
If you are familiar with the analogies I hinted about, I just (indirectly) showed that they cannot achieve the efficiency goals. Do you want to bet? With the limitations regarding thermal efficiency I hinted about, the mechanical efficiency would have to be much higher than 100%. Do you think it could? Rotary engines do not have any inherent advantages regarding gas exchange and heat transfer compared to conventional piston engines. Instead, they have disadvantages in those areas.

I generally have an open mind about testing new ideas so, in essence, I am also positive about that this alternative engine is investigated. However, I do not like when inventors make unrealistic claims (or targets). A target of 50% break efficiency should be more realistic in this case, although even this level would be difficult to achieve.

How is this a new concept and how is it not a Wankel?. This is exactly one of the ideas Felix Wankel proposed,it was actually built and tested;The idea of a rotating outer ring was dumped after tests showed excessive heat warp of the outer ring. I guess these guys are just going through old patents and making prototypes to see what could/would work.Maybe that's what you do when you already get people to invest in a non existing concept.Correct me if I'm wrong, but I think this is the third different engine concept they already came out with,and non happen to be original to this company, just making prototype of other people's old ideas.

@Ben
I agree! This engine has much in common with the Wankel engine and potentially also share many of the problems. Felix Wankel’s original design made the engine fully balanced, which was a very nice feature. However, it turned out to be too complicated, so this design was abandoned by NSU (and others) in favor of the “wobbling” piston in all current designs.

I recall that Rolls Royce (division of diesel engines, not to be confused by the luxury cars) was working on a diesel Wankel engine. This was a two-stage design regarding compression and expansion, where a larger disk was used for the low-pressure stage and a smaller disk for the high-pressure stage. Thus, higher compression and expansion ratios than with a single-stage design could be achieved, which is essential for a diesel application. Connection between low and high-pressure stages was made via transfer ports. Besides the lack of the claimed overexpansion in the LPI engine, the RR design should have provided almost similar thermodynamic advantages as the LPI engine. Needless to say, however, the Rolls Royce concept did fail.

@SJC
Do you or do you not agree with the quote? If you do not, why did you quote the article in the first place?

"LPI has shown that if it achieve(s) sealing, heat transfer coefficients, and friction similar to today’s Mazda Wankel engine, it can achieve its efficiency goals."

I have no basis to agree nor disagree, this is their statement. It says that if they achieve the Wankel performance then they can meet their efficiency goals.

I an not trying to pick a fight, I am trying to clarify their position. This is one of the reasons I do not comment here, the usual suspects make personal attacks.

I challenge the claim that this engine would approach ideal isochoric combustion. Of course, the combustion is supposed to take place in a confined space that should not change during the duration of the combustion. However, no expansion – or work – takes place in a confined space; i.e. one must also consider admission of the gas to the chamber where expansion is made. Admission to the expansion chamber is not instant and during this phase, the volume of the expansion chamber does indeed change, making the process non-ideal. In addition, the expansion chamber has a certain dead volume to start with, which must be considered. This is more or less equal to the heat release in an otto engine – or stem admission in a steam piston expander for that matter – where combustion or steam admission cannot be instant. Thus, from a thermodynamic point of view, the LPI engine cannot provide an ideal isochoric process. Geometry decide how fast the process can be but I would assume similar conditions as in an otto engine, i.e. that this process last for about 40 crank angle degrees, or – in the best of cases – is only slightly faster. It would be very interesting to see cylinder pressure traces from the expander, since this would show the “true” behavior of the engine. I would also like to see real experimental data on brake efficiency for this engine.

@SJC
Why do you keep repeating their statements if you do not agree with them.

@Kevin,
Note that the Audi rotary has similar problems as all other rotary engines: notoriously bad efficiency.

A Wankel engine is the perfect compromise between all the drawbacks of conventional 2 and 4 stroke piston engines.

@Roger
Nice analysis! I have been thinking about the same problems, too, so I will add some comments.

Separation of compression and expansion might be good for volumetric efficiency, which appears to be a problem for conventional Wankel engines but the gas transfer between the chambers is an “unknown” area, so I would not expect any particular advantage over a piston engine after all regarding volumetric efficiency. As you indicated, heat will be a problem – I would say for any material, not only aluminum. One can read in the text (and between the lines) that they are aware of this problem. There are examples from piston engines (actually already Rudolf Diesel tried this), where water has been injected to cool the cylinder. If the amount of water is limited and the droplet size is small enough, they might be able to avoid condensation. If enough water could be used to manage the heat problem, I do not know. I suppose that they would avoid the oil sump, so this could help to avoid the problems of water-oil mixing experienced in 4-stroke engines. Careful control of water injection during cold start would have to be managed to avoid condensation directly on cold surfaces. The best solution would of course be to use water lubrication. This has been proposed for Rankine piston expanders but to my knowledge, nobody has been able solve the problems associated with tribology so far. If they would solve this problem then someone might immediately use the technology for a Rankine bottoming cycle expander on a conventional piston engine.

Regarding efficiency, it should be noted that water is “worse” than exhaust gases (and air) regarding the exponent “k” (according to their nomenclature), which would reduce efficiency. Furthermore, the heat of vaporization is high, so this loss would have to be compensated by reduced heat loss. I doubt that this could be accomplished, so (neglecting the gain from hyperexpansion to compare apples with apples) the result could even be reduced efficiency compared to a conventional engine.

I could agree that the potential for efficiency could be on pair with conventional automotive diesel engines, i.e. ~43%. With optimistic assumptions, why not aim at 50% as a future goal. However, the “moving targets”, i.e. conventional engines, also aim at this level.