With that Epilogics ratcheting mechanism cycling at over 80 Hz, I would imagine that this engine will have a lifespan of about 20 or 30 minutes at rated power.

Would the pair not be loading at 40hz as the back to back arrangement suggests only one is applied at any time. If the sequencing allows the next clutch to load as the previous one unloads there may be no shock loading.

This concept is flawed, the pistons don't have a sinusoidal type of cinematic, they will stop abruptly a the top dead center generating awful vibration or even shock.

Now is time for the resurection of the Pescara free piston engine, but with computer control. The output was a turbine run from the exhaust gas. We now have better control and turbines. A mechanical rectifier was used for a transmission by the guy who invented machine guns that fired when the airplane prop blade was not in front of it. If there is a good mechanical rectifier it should be used with resurected versions of his transmission. ..HG..

Gentlemen,

This is an ingenious arrangement for a completely mass balanced free-piston engine. The output of one cylinder is used via a seesaw pivot to power the compression stroke of the other cylinder, with the excess mechanical energy of the power stroke harnessed via a one-direction ratchet mechanism. Complete combustion via HCCI is virtually ensured, since provided that sufficient power is given to the compression stroke, the pre-mixed charge will be compressed until ignition occur. This engine is not fixed to any compression ratio, but can acquire any compression ratio sufficient for COMPLETE combustion and higher degree of expansion via HCCI mode, hence, efficiency will be much higher than current engines slaved to a fixed compression ratio as defined by the crank-slider mechanism. Of course, the compression ratio must be precisely regulated to ensure low NOx formation (HCCI with low-temp combustion) and low HC and CO formation at the same time. Low speed low power can use relatively lower compression ratio in the range of 12-13, but will have more time for complete combustion of the charge, while high speed will use much higher compression ratio in the range of 16-20 for complete combustion and near-isochoric (near-constant-volume) combustion to maintain high efficiency at high speed. Thermal loss via engine cooling will be minimized, since more heat energy will be converted to mechanical work due to the high-compression-high expansion HCCI isochoric combustion characteristic.

I can't comment on the durability of the "mechanical diode", or rotational movement rectifier, but the mechanical output of the see-saw mechanism can be harnessed electrically and converted easily into DC or AC current that can be used to power electric motor, thus a serial electric hybrid without requiring mechanical transmission. Even though serial electric hybrid is not as efficient as mechanical transmission, the overwhelmingly superior efficiency of the free-piston HCCI combustion will more than make up for this. Add a battery for recuperative braking and, bingo, you'll have and ICE-electric hybrid that can rival the efficiency of the FCV.

Who now can complain about the lack of efficiency of the ICE, eh?

All very nice but there is 1 little practical problem ! Because the upwards movement of the piston is only limited by the combustion of the fuel/air mixture, what in case of a misfire ? The piston will smash in to the cylinder head. Especially at the exhaust move where the air cushion is missing. Also consider that at the exhaust stroke the max piston height is only limited by the other pair's (at that time) combustion stroke piston position. So we may hope that the gear linkage between the 2 piston pairs has not to much "play". By placing the exhaust valve in the exact center of the cylinder head, it could be used as an emergency spring to limit the the piston stroke. Like somebody said above, the crank could be "an invention" in this case.

This engine sounds like a great idea but let's see if the inventors can convince Honda (who probably has more experience with highly-advanced engineering work with gasoline internal combustion engines than anyone on Earth) to take this engine to production level.

Is their claim of 60% thermal efficiency anything more than pure hype?

@Ralph,
60% thermal efficiency is a bit of a stretch. Expect real-life BTE of ~45-50% the most, which is still very good, considering a typical Otto-cycle engine at 32% peak BTE. You'll have above-diesel-level efficiency with low emission like gasoline engine. Maximum theoretical efficiency of Otto cycle at a CR of 20 is ~70%, and for a CR of 40, it is ~77% before friction and heat loss.

@martinb,
Do a force vector analysis, and you'll see that at mid stroke, the side load will be quite little, and no side load at TDC and BDC. The pressure at mid stroke is but 1/8-1/10th that at TDC, depending on the CR (Compression Ratio), so the side force will be quite negligible.

Vibration can be eliminated with mass balancing by using a bundle of 4 piston-cylinder assembly like shown in the picture, with the pistons moving in opposite directions.

No need for throttling this engine. Just reduce the amount of fuel injected into it. No matter how lean a mixture is, it will ignite once the CR is high enough, and there is no limit as to how high the CR can go in this engine. Just keep on compressing it until it ignites. Practically, though, a CR of 20-24 is the practical limit due to the amount of heat and force involved, same as in a diesel engine. As such, direct fuel injection with stratified charge combustion at low power setting should be used to ensure ignition, while higher power can use stoichiometric pre-mixed charge.

Roger Pham’s analysis of the FP-FS engine is correct - for the most part.

Concerning efficiency, lean fuel mixtures increase the ideal Otto cycle theoretical efficiency. Plots of efficiency vs. CR are typically presented at Stoichiometric fuel/air ratios. However, as mixtures are leaned out, the curve shifts upward and a family of curves for different fuel/air ratios are readily calculable from proven formulas. As an example, with typical automobile fuels, at Φ of .25 and a CR of 50:1 efficiency is improved from 65% to 75%. Existing engines typically attain 70 to 80% of theoretical efficiency. Our analysis of losses in the FP-FS, lead us to believe the 80% of theoretical efficiency is realistic. Thus, at a 75% theoretical and with an actual of 80% of that, a 60% real efficiency is attainable. In the poster presentation, we include two graphs, one from Edwards work under GCEP at Stanford and the other by Van Blarigan at Sandia to illustrate these points. Edwards, et al. research work is ongoing and his latest report can be read at http://gcep.stanford.edu/pdfs/ryhz9TjH9TbEp9qJ290tWQ/2.6.3_edwards_2008_hq.pdf or go to the gcep.stanford.edu site and select “Latest Reports 2008” then report 2.6.3. He has endorsed the Van Blarigan work on a two-cycle two-cylinder free-piston engine prototype (that struggles with the scavenging difficulties typical of two cycle engines) as being a practical means of achieving the objectives of his research. We believe the FP-FS engines geometry is the most practical embodiment of his research (which we developed on our own concurrent with his work).

With high compression ratios, there is energy remaining in the exhaust that is practically recoverable through turbo-compounding - see the new Detroit Diesel engine and the development work by Deere - past DEER presentations. Using proven techniques, such as an exhaust driven turbine directly coupled to a high- speed alternator an additional 5% of power may be recovered (at rated load). See the work done by Bowman for Deere. The total 65% efficiency closely matches that of the full expansion Atkinson cycle.

The only way we know to reach such high compression ratios, with existing fuels, is to use HCCI/PCCI. Very high CR’s have been demonstrated repeatedly, by several other research groups, using Rapid Compression Expansion Machines (RCEM) in laboratory tests. RCEM are essentially single shot free-piston engines. Edwards “extreme compression” test rig is in fact a specialty RCEM with the objective of achieving 200:1 CR with HCCI/PCCI. There has been limited, but meaningful, R&D with HCCI in free-piston engines in addition to that of Van Blarigan. One of particular interest is the web site of Hans Aichlmayr at: http://www.menet.umn.edu/~haich/ (check out the combustion movies) and http://www.menet.umn.edu/~haich/engine/index.html. It has been clearly demonstrated that a wide range of fuel/air ratios may be attained with HCCI. Thus, “throttling” of the engine is achieved through fuel/air ratio variation - typically from about .25 to about .80). Note: It has also been clearly shown, by a number of researchers, that the Stoichiometric mixtures needed for slider-crank engines waste more than a third of the energy at the typical compression ratios that must be used. The accompanying expansion ratios are not adequate to extract the energy from the combustion and it dissipates as thermal losses in the engine and exhaust. Thus, to reduce these losses higher CR (and thus ER) are REQUIRED. Of course, operating at lean mixtures reduces power density (as does throttling an SI engine) and supercharging, with intercooling, can be utilized to compensate for this.

Now, the FP-FS engine inherently has variable compression (as does all free-piston engines) which turns out to be a necessity for full load range HCCI. Ongoing work with HCCI in conventional slider-crank engines, even with complex charge tailoring schemes, has failed to extend significantly the operating range of HCCI - nor does it appear likely to do so. Minor variations in charge properties result in differing ignition properties - even from one cycle to the next in an operating engine. To accommodate this we adopted the variable compression of the free-piston engine and took this one-step further with the “floating stroke”. The FP-FS engine stroke is not controlled but allowed to reach whatever CR is required to reach HCCI and there will be minor variations from stroke to stroke. We utilize an adaptive ECU that compares the actual CR with the projected CR and “tweaks” the valve timing accordingly. For more details and background read the patent - easily available from GOOGLE.

We initially presented the FP-FS engine at the 2006 DEER conference to get a “peer review” type of feedback. The conference is an excellent event for this, as most of the world’s engine experts attend it. I have been chewed over by the best from industry, university and Government Labs and been able to address all of their issues. In 2007, in response to some of the feedback from 2006, we presented the FP-FS in a unique hydraulic hybrid application. Still, the major input we received then was “How about giving us rotary shaft power”. Pointing out that either hydraulic or electric power can be efficiently and cost effectively produced from a free-piston engine then converted to rotary was to no avail. So, we went back to the drawing board and researched, again, all of the methods for extracting rotary shaft power. We went all the way back to Otto’s first engine - which was an award winning free-piston engine - to the most recent developments for aerospace. The best we could find was the Epilogics mechanical diode. We procured two devices (produced as replacement parts for GM transmissions used in drag racing) and did some analysis of these units. It appeared to us they would do what was needed and subsequent preliminary design work by Epilogics confirms this. Their original mechanical diode was developed for a CVT, which required a 500,000,000 cycle life. With proper sizing and materials, unit life can be indefinite - although cost is a consideration for high performance materials. See their web site for more details. Incidentally, Otto had it right all along. By utilizing a free-piston engine and his ideal Otto-cycle, the most efficient engine geometry possible is attained. Unfortunately, discarding the crankshaft not only eliminates the combustion restrictions it also eliminates the best means of rotary shaft power production. Nonetheless, we believe the mechanical rectifier will be a close second to the crankshaft in that regard.

Now, in respect to the prospect of a misfire and the piston striking the cylinder head consider the following. Even without charge ignition (an extremely unlikely event with variable compression), there remains the “gas cushion effect”. Assuming rings and valves do the job intended there will be a “gas spring” that prevents contact of the piston and will rebound the piston over several cycles (if nothing else is done to intervene). This has been demonstrated in many two-cycle free-piston engines and is in fact sometimes used to start these engines. That is, they are motored until sufficient pressure and temperature is achieved for fuel injection (most have been Diesels). Therefore, the risk of piston contact with the head is strictly that of valve or ring failure.

Lastly, our engine is still alas, “vaporware” as Tom notes. It has taken seven years, many thousands of hours and considerable expense to reach the point we are at now. To build and test prototype engines, to the extent required to satisfy industry, is an expenditure in the many millions of dollars. Note: A presentation by Cummins took most by surprise because it advocated co-operation by industry partners to develop new engines - as the cost to do so has become so high. SBIR is inadequate for this purpose as is the Angel Investment community so we have carried the cost on our own (VC sees the technical risk as too high). We had several lengthy discussions with people from Ford, GM (the Volt is indeed a good prospect for the FP-FS engine) Cummins, Volvo, Deere and a number of labs and universities at this year’s conference. We are now endeavoring to set up partnering with at least one of each to go forward on prototype development and test. Stay tuned - or better yet come to DEER 2009.

I do not believe that CR of 40-50 is practical for current piston engine technology.
For a diesel engine, optimum CR is usually reached at 18-22 for large engines (trucks), and 16-18 for smaller engines (auto). Higher CR will lead to lower BTE due increase leakage thru the piston ring and valves, increase heat loss, and poor mixing of fuel and air leading to poor combustion. In HCCI engine, poor mixing of fuel and air is not a problem, but combustion will be much more rapid at higher compression and will severely stress the engine.

Hence, we should realistically expect a peak BTE for the FP-FS engine somewhere between 45 and 50%, depending on engine size and other factors.

Since the FP-FS engine must depend on the kinetic energy of the movable mass for the compression stroke, higher engine speed will result in higher compression ratio, (E=1/2MV^2). As such, in order to keep a compression ratio between 12-20, the speed range of operation of the FP-FS engine will be rather limited and will need a hybrid power train for automotive use. This will make the durability and practicality of the "mechanical diode" a moot point.

Gentlemen,

We appreciate your thoughts and comments and we will attempt to address them as best we are able, given the format at hand, i.e. lack of graphics.

Compression ratio is not the factor affecting the mass leakage of valves or rings - peak cylinder pressure is. The FP-FS engine, by utilizing lean mixtures (facilitated by HCCI), will not exceed existing peak cylinder pressure limits. Extensive testing, by others, has shown that a CR of 50:1 is reasonable - given the lean burn and resulting “normal” peak pressure. Mass loss due to ring blow-by or valve leakage will be no more than in existing Diesel engines. Thus, the operating range is not limited to a 12 - 20 CR. Further, the turn down range of the engine is about 4:1 - solely utilizing fuel mixture control (obviously throttling can extend this). However, we prefer un-throttled engines and instead utilize downsizing and supercharging (with good after-cooling) - which provides about a 3:1 turn “up” range.

Rapid HCCI combustion does produce a more rapid cylinder pressure rise and higher piston acceleration - which the free-piston engine has shown it can accommodate. Note: Free-piston engines operate at a higher speed (typically about 2:1) than conventional slider -crank ICRE, all other parameters being equal. Combustion noise is an issue but it appears to be manageable with existing technology (not our opinion but what other “experts” tell us).

As an aside, look at the EVE project to see where the future of peak pressure in engines may be headed - double that of existing engines (probably starting with the large Diesel engines for ships).

Nor will there be higher heat loss. The lean burn produces lower combustion temperature, thereby reducing heat loss (and emissions). In addition, the combustion chamber wall area is less, which also reduces heat loss. Overall, the heat losses are expected to be 18 -20% less than in a comparable Diesel engine.

From the outset, we intended the FP-FS engine to be a hybrid power train, either hydraulic or electric, for both mobile and stationary applications (there is a broad prospective market in distributed generation). We had to be browbeaten into coming up with the mechanical rectifier. However, in pursuing the subject we realized that converting the output to rotary shaft power opened the path to utilize high-speed alternators (developed for micro-turbines). We realized, from our past micro-turbine work, that at this time a cost effective and high power density drive train can be developed using a speed increaser gearbox and high speed alternator. The alternator output is rectified and inverted to whatever form desired. This is the configuration we anticipate we would use for Beta test engines.

We have begun the design on a 1.6 liter (2.5” bore, 5” stroke) proof of concept engine which would ultimately utilize this drive train (yes the Volt is the target). As we are still internally funded, it is uncertain if we will be able to build and preliminarily test this engine prior to next year’s DEER conference - which is our goal.

We do not expect the proof of concept engine to demonstrate cost effectiveness, as it will be custom built. It is not likely it will satisfy industry criteria in that regard. Here again, from our past micro-turbine work, we have learned that even though you provide a proof of concept engine the next question will be: “Can you make it at a competitive cost”. We know that question needs to be addressed by Beta engines, which is a very costly effort we cannot fund on our own.

We have also started work on a web site for the FP-FS engine, which we anticipate will be up, in its initial form, by the end of this month (URL of idealottoengine.com). That format will enable us to provide you with information that is more comprehensive.