Are adiabatic/ceramic engine concepts being researched nowadays?

[DaveShoe]

I understand that ceramic "adiabatic" engines were popular research and development topics in the 1980s. The 1979 oil crisis brought the ideas into vogue, and major engine manufacturers around the world were constructing and testing conceptual ceramic engines that required no cooling system. I have read some of the old sae.org technical papers on these experiments, and it seems the downfall of the ceramic engine is ceramic fatigue. At least that is my impression.

I am interested in learning more on the latest in published adiabatic engine technologies. I am having fun developing a concept, posted at

http://www.insulatedpulseengine.com,

and am looking to learn more and move beyond this level of understanding. My present plan is to purchase more recent technical papers from sae.org, but would be interested in finding a forum that discusses the topic. I just realized this eng-tips site is a great resource, and am now checking it out.

Thanks.

 

[BrianPetersen]

I briefly skimmed over your website. Interesting ideas. Take the following as constructive criticism.

I think you are going to find that by drawing in only one-quarter as much air per stroke, the amount of torque will be weak (which means the amount of friction will be high compared to the output of the engine - and friction is a good efficiency-killer, too). And the valve mechanism necessary to open and close a valve in such a short duration won't allow even moderate revs. Translation, big heavy engine with lots of friction and weak power output ...

BUT ... don't take that the wrong way because I think the idea is nevertheless interesting.

Extend intake closing later in the stroke (say, let in twice the air you are proposing) and you get more power output and easier valvetrain design. Yeah, perhaps it won't overexpand the exhaust quite as much, but I think "on the overall" it will be better to sacrifice a bit of overexpansion in the interest of letting the engine be smaller for the same power output. It's all about optimizing the design ... This has nothing to do with the use of ceramics, by the way.

Or even better ... use variable intake valve timing and lift (a.k.a. BMW Valvetronic) to allow full torque output when needed but operate in overexpansion mode at lesser torque output.

Thermal-barrier coatings are starting to show up. You can get it done aftermarket (I've had it done to one engine as an experiment with what I would say is moderate success - for an example see

http://www.swaintech.com

) and I think some Subaru turbo engines use them on the pistons. I would not say that these coatings are a revolution, though - it's a few percent max.

 

[DaveShoe]

Hi Brian,

I definitely seek and appreciate criticism. It need not be polite, as a jarring critique will often speed the learning process.

My Insulated Pulse Engine (IPE) concept does not represent any currently constructed engine. I am presenting it as a tool to help me learn more, and hopefully to trigger thought in others. To be sure, I cannot presently budget a build. Based on my current set of experiences I believe the IPE works, but I've been designing long enough to know that my goal must be to gain sufficient knowledge to learn why it doesn't work. When this happens, I will be able to take the idea to the next level (or to a conclusion), and I will gain knowledge in the process.

At present, I hope to find a technical paper which discusses the test results of a similar engine concept, but these are often not made available to the public.

Regarding the quarter-sized intake stroke, it is not immediately obvious, but I have come to find significant efficiencies in this approach. Just as modern 240 horsepower car engines are routinely tasked to provide good mileage while providing 20-50 horsepower during normal driving, the IPE might be considered a "240 horsepower" sized car engine that is only capable of producing 60 horsepower max, but at uncharacteristically high thermal efficiency. The IPE may compare poorly in a performance-oriented magazine comparison of conventional cars, but the IPE concept would seem to merit a different type of write-up, such as that which focuses on daily driving while avoiding 0-60 acceleration limits and such. Just as mechanical friction is not a severe issue in a 240HP car engine running at 40 horses, the same is true of the IPE. Friction significantly varies with load and velocity, not just size and surface area.

Valve actuation requirements are not particularly intense in the IPE. Quite the opposite. Granted, the valves are asked to operate for short durations, and valve head shrouding is notable in the shown IPE construction, but the IPE formula calls for restricted RPMs, OHC heads provide a low valvetrain inertia, roller-type cam followers allow for rapid valve ramps, and the IPE flows very little intake or exhaust volume, and seeks only a characterized flow that can be compensated for by the described camshaft geometry. Intake and exhaust actuation uses ordinary constructions in the IPE, nothing fancy is needed (this is my opinion, and may be flat wrong).

Thermal barrier coatings are way cool, providing easy horsepower and longevity in high performance applications which may otherwise melt holes in pistons. While engine builds which use them are beyond my budget, I have friends who use them regularly. Quite inspiring. I've researched them a bit, as well. One thing I have noted is there are thermal conductivity benefits to using discrete ceramic components, because thin ceramic films lack certain insulating properties of bulk ceramic. It also seems that ceramic films thicker than .020" are difficult to control, yet there seems to be notable thermal conductivity benefit in ceramic thicknesses closer to 0.100" thickness.

The IPE would be so much simpler if a ceramic spray could take care of the insulation part of the concept.

The ceramic factor boils down to thermal latency: The less energy that is absorbed, conducted, and reradiated by the ceramic during each engine cycle, the more efficient the engine can be. With luck, I will learn of a conventional ceramic coating that provides both the thermal and longevity character I seek, but at present, I must lean toward discrete ceramics to get the math to work in the IPE. This is based only on what I've read, as I have "zero" direct experience with ceramics.

I am opinionated, but I also know I am very often wrong. I like to learn the right way, and talking it out is a great way to do this. Mainly, this is pure fun stuff to research.

Thanks for your time,
Shoe.

 

[DaveShoe]

Oh, by the way, I have found some cool discussions on ceramic engines archived in this forum, and expect to spend a good bit of time reviewing them over the next few days.

Shoe.

 

[BrianPetersen]

The trouble is that with such extremely short intake valve duration, if you want to use a practical cam profile, it will limit not only the duration but also the lift, and if you find some way of getting a normal valve lift (rocker arm with a really high multiplication ratio?) the accelerations involved in opening and closing it in 60 degrees will be the same at 1500 rpm as a normal 240-ish-degree duration cam would see at 6000 rpm if the lift were kept the same (which it can't be because it would hit the pistons if opened so quickly, but I digress). So now, not only do you have only one-quarter torque capability, but also one-quarter RPM capability. Now that 240-horsepower-sized engine (for example, Nissan VQ series 3.5 litre V6) is making not one-quarter but rather one-sixteenth power! Not very useful in my opinion. I think the frictional losses from spinning the engine are a lot more dominating than you think.

Assume you find some way around the RPM limitation so it is indeed one-quarter power. Now the question is whether such a concept of a large de-rated engine would be capable of beating the efficiency of a smaller but more conventional engine. Getting 60 hp from a 1.0 litre engine is practically a no-brainer. Do you really think a 3.5 litre 6 cyl engine that is effectively strangled to have the power of a plain ordinary downsized 1.0 litre 3 cyl engine is really going to be more efficient particularly taking into account the extra weight and space of the engine ? ? ? Most internal-combustion engines have best BSFC near their maximum torque output, under HIGH load, where the torque they are making most overwhelms the friction and losses.

Don't get me wrong, I think the idea is interesting, I just think you've gone too far. If you let in say 2/3 the normal amount of air then the engine doesn't have to be as oversized, and you are still making use of the biggest and most effective part of the overexpansion, and it will allow a much more rational cam profile that won't have such drastic RPM limitations.

This is not a new idea, by the way. It's called the Atkinson cycle, and the Toyota Prius uses it, and the current Honda Civic base 1.8 litre engine uses it on the "part load" lobe of its VTEC cam mechanism. I think those do it by LATE intake valve closing (pushing some of the air back out) in the interest of having a practical cam lobe shape ... same end result, though.

Don't get hung up on trying to make it a low speed engine; you need to spin it to make power. Automotive diesels are using roughly 4500 rpm redline nowadays.

Regarding ceramic parts versus barrier coatings versus conventional design, this is the interesting area. One thing to keep in mind is that ceramic with no cooling system does not necessarily stop heat transfer between gases and surfaces from occurring during the cycle. If the ceramic has a high surface temperature on account of being uncooled, it will transfer heat into the cycle at times when you don't want it (intake and compression). The thermal barrier coatings supposedly work by transmitting less heat in OR out between the gas and the surface in question. The temperature of the underlying parts (example, piston) stays the same, or if anything stays even cooler because of less heat absorbed during combustion and power stroke while the normal cooling system of the engine works as usual. You don't want to transfer heat into the air during the intake or compression strokes - that promotes pre-ignition or detonation, and you certainly don't want that. Diesels don't mind heat in the intake as much, but even there, transferring heat in during the intake/compression just reduces the power output and overall cycle efficiency.

I know thermal barrier coatings work, but the gain is small (few percent, possibly down to nothing depending on engine design). I don't have experience with actual ceramic parts.

 

[DaveShoe]

Thanks again for responding.

The cam ramp has much to do with the cam size. A larger diameter cam circle (larger hollow cam), combined with appropriately sized cam followers, can handle more exotic ramps. This is a simple mechanical exercise which I've only intuitively considered the details, as I senvision a resolution. A slightly less efficient alternative is to go the Nissan VQ route if required, but I will look at modeling my proposed IPE method to search out the issues you intuitively grasped on initial viewing. I appreciate this feedback a bunch.

Rather than the conventional high-lift 60 degrees for the intake you see, I think low lift and 90 degrees to achieve the quarter-fill, with an adjustably-timed intake cam to compensate for RPM. I see RPMs limited to around 4000 in the IPE, mainly due to flame front but also due to valve actuation. That is not much less than the normal operating RPM of an ordinary engine.

My take on the intake valve is: We know when and how it opens, as these ramps and limits are defined by piston clearance. It closes when sufficient air has been drawn to fill one fourth the full displacement volume. Large lifts are not required to achieve this. This eliminates all unconventional actuation forces from the equation, simplifying the valvetrain. Lift can purposely be limited to permit desired intake volume at a relaxed duration. This is quite different from engines with high volumetric efficiency (like the VQ), but works well with the IPE, particularly when controlled by a computer.

I will sidestep the frictional issues at this point, as I suspect the IPE and the VQ as being equivalent here, and will attempt to compare my 60HP variation of the IPE with a Nissan VQ operating at 60HP (roughly 75mph on the freeway in a smaller car?). Both engines have similar displacements in this discussion, but the IPE happily operates at full throttle all day long while the VQ is lazing along with another 180HP on tap, if needed. I am most interested in the steady-state 60HP, not the reserve power, so both engines are equivalent at this point, with exception of fuel consumption. I've never studied the VQ (my world is small), but assume it is an Otto engine, not a Miller or Atkinson engine. For this reason, I believe the IPE acts much like a VQ on the freeway, except the mileage is doubled.

You compare a 1 liter modern engine to a 4 liter IPE in your second paragraph. This is good. Power outputs are similar, except the 1 liter pumps out a whole lotta heat and pressure, while the 4 liter IPE is cool and quiet (even without a muffler). The cool, quiet 4 liter IPE logically consumes far less fuel than the hot and noisy 1 liter, just by listening to the exhaust and touching the radiator. This is a strange abstract, but is something that is sensible to me at this point. You use the term "strangled" to define the IPE, but I suspect you are not yet grasping the flow and combustion characteristics. I may be wrong, and hope you teach me where.

You advise changing from a 1/4 intake volume to a 2/3 intake volume, but that does redefine the IPE. Remove the ceramics and you've got a Miller or Atkinson engine now. I'm not looking in that direction. If you think of what a turbocharger or turbocompounder does, you'll see it is an inefficient bolt-on compared to using the pistons to recover this unclaimd energy. The IPE basically builds the turbocharger into the cylinder. When the exhaust valve opens, there is no energy remaining to be used by a turbocharger. You cannot get this if intake displacement is 2/3 the exhaust displacement. You can if the intake is 1/3 or 1/4 exhaust displacement.

Question: Do Atkinson engines or Miller engines ever use turbocharging? If so, does this bother you? It bothers me that Miller engines are turbocharged, and Atkinson engines are headed that way. It means there is too much energy being lost in the exhaust of these highly efficient engines. Rather than turbocharging, the IPE is simply looking to use the existing insulated piston to recover energy otherwise destined for a turbocharger.

As for the ceramic transferring heat, that is the "latency issue I previosuly referred. Latency generates a "deadband" of lost energy, both during the pressure and vacuum segments of the engine cycle, but this issue can be minimized by recognizing its presence, and then integrating it as if it belonged. Since I cannot get rid of the latency "deadband", I simply accept it as a minor efficiency detractor. No biggy, but something that must not be ignored.

I do believe that the thermal barriers you speak of are of the thinner film "protective" variety, and the discrete ceramics I refer provide far more insulation and efficency than you presently recognize. I pondered the thin vs. discrete ceramic issue a long while before becoming opinionated toward favoring the fat stuff. Not that my opinion is a good thing, but it is none-the-less my present conclusion. I can back it up with details of the Watts/Heywood papers, the Cummins/TASCOM papers, or a number of other sae.org papers on this topic which I've reviewed and have found abstract mention of in other threads of this forum.

Note that all sae.org ceramic engines I've reviewed aim to increase volumetric efficiency over an Otto engine. My IPE is the only one that decreases volumentric efficiency. For this reason, I feel it might be unique and uncharted territory. Frankly, the IPE is an ugly concept at first blush. It takes time to see how all the odd approaches add up to a uniquely fuctioning engine. That is, if the IPE works at all.

Thanks again for your time. You've given me a couple new things to research and ponder.

Shoe.

 

[TDIMeister]

I agree with Brian's constructive critique, particularly his detailed description of the challenges of valvetrain acceleration, friction, volumetric efficiency and rev-ability.

The OP performed a loss-split analysis and claims an efficiency of 65%, which I find to be a magnificent stretch. A proper analysis should begin with a calculation of the ideal-cycle thermal efficiency and then work out losses due to friction, heat transfer, dissociation, finite-time combustion, etc. It should be pointed out that an ideal air-standard Otto cycle with a compression ratio of 20:1 achieves a thermal efficiency of 69.8%. Granted, from following the sequence of cycle events in the website, the engine does not necessarily follow the Otto-cycle anymore, but an efficiency delta of only about 5 percentage points from any ideal case is not realistic. Even if you were to assume a Carnot thermal efficiency between the lowest- and highest cycle temperature (1-(300/2500)=88%) as your baseline, you are proposing achieving 65/88=74% of the Carnot efficiency, which will not happen.

As the Web page already mentioned, it is not constructive from the standpoint of brake work and pumping losses to expand the burned charge beyond the point where the cylinder pressure drops below atmospheric. Since the engine would operate naturally-aspirated and be heavily short on volumetric efficiency, I expect end-of-compression pressure around 10 bar despite the 20:1 compression ratio (you seem to neglect adiabatic expansion of the fresh charge to partial vacuum after IVC because you state a linear relationship between cylinder volume and pressure), and peak cylinder pressures to be in the order of 25 bar. If you want to fully expand this 25 bar during exhaust to atmospheric, the desired volumetric expansion ratio is about 65:1 (1/25)^(1/1.3) However, owing to the fact that piston pressure below that required to overcome friction is adding nothing to brake work, the result might be a very high indicated efficiency on paper, but pumping- and mechanical losses will be massive, certainly much more than has been estimated.

Anyway, from my very quick reading of the site, a couple of comments: You propose the combustion process to be spontaneous combustion of directly-injected gasoline or alcohol. If you want to run neat alcohol, the compression ratio will need to be even higher to have a sufficient ignition temperature, as Scania has done with its ethanol compression ignition engine. Power output will be further limited because you will run into a smoke limit (yes, even with alcohol) before you can run a stoichiometric AFR. Combustion occurring between 0 and 5 degrees ATDC is not realistic. Do a reading on the Wiebe function.

Some general notes: Some of the main weaknesses of adiabatic engines of the past have been already mentioned in earlier posts -- low volumetric efficiency and high friction- and pumping losses resulting in dimished specific power to name a few. In our current regulatory landscape, the increased NOx emissions that is inherently characteristic of insulated engines will be a big challenge. Furthermore, the OP proposes compression ignition, and an undisclosed method for an incredibly quick combustion process to boot. This will also have deleterious effects on NOx.

One item of loss not considered is the effect of dissociation. At higher process temperatures, dissociation increases; this decreases the polytropic exponent and therefore reduces the cycle efficiency.

If it seems I'm being extremely harsh in my critique, please don't take offense. The concept does incorporate some sound elements already in study or practise (early IVC, late EVO a.k.a. Atkinson cycle); direct-injected compression ignition; insulating of the combustion chamber; etc. I just think that some revision of certain minor details and tempering the claims with an injection of realism is in order.

 

[TDIMeister]

After proofreading my post, I realised an error in my calculation. In the expansion stroke, expanding polytropically from 25 bar to 1 bar would require a volumetric expansion ratio of about 11.9:1, not 65:1 as I originally posted (T4/T3)=(v3/v4)^(n-1)=(P4/P3)^((n-1)/n); rearranging, the volumetric compression ratio (v4/v3)=(P4/P3)^(-1/n); here, I've used a value of 1.3 for n owing to the composition of the burned gas and dissociation at high temperatures.

End of compression pressure of 10 bar was determined by assuming atmospheric in-cylinder pressure at IVC followed by adiabatic expansion at a 1:4 volume ratio to BDC (P2'=P1*(1/4)^1.4) followed adiabatic "rebound" as the OP called it and compression for the full 20:1 compression ratio (P2''=P2*(20)^1.4=9.52 bar.

Peak cycle pressure of 25 bar was determined by assuming constant-volume heat addition (P2''/T2'')=(P3/T3); P3=P2''*(T3/T2''). I took T3 to be 2500K and T2''=T1*(20)^0.4, with T1 (charge temperature at BDC) to be 300K. It is noted that when a gas is adiabatically expanded from ambient conditions, the temperature should fall below ambient, but this ignores in-cylinder wall heat transfer. I would argue that my 300K figure is already too low but it suffices for simplicity. This gave P3 of 23.93 bar that I rounded to 25 due to not carrying the rounding from the previous calculation.

My point is, expanding the burned gas past below atmospheric pressure at EVO may have theoretical advantages in exhaust gas temperature, but will mean re-breathing gas through the exhaust valves, increase pumping losses, kill volumetric efficiency, and may result in a high percentage of trapped residual gas.

 

[DaveShoe]

Hi TDIMeister,

Thanks for your response. You have math and calculation abilities which I can only seek to learn. I will work to understand the numbers and formulas you provide.

I would certainly not describe Brian's or your viewpoints as "harsh", as they both challenge the IPE concept using technical savvy from viewpoints which I lack. This is exactly what I seek.

I do grasp adiabatic compression factors, as I've long ago puzzled how my old 11.0:1 car engine can generate 230-235 PSI of pressure during a compression check, eventually realizing the temperature of the fuel/air mix measurably climbs with temperature, even in a gasoline engine. When I mentioned the "quarter atmosphere at BDC", I didn't realize there would be significant deviation. Except for ceramic reradiation losses, the actual vacuum trajectory is not too important, as the energy is reversibly recovered on rebound. The adiabatic pressure and temperature trajectory during compression is what I seek to learn more about.

You have provided me the first glimpse of a mathematical model to explore and learn from. I am not presently knowledgeable enough to do the math founded on an ideal IPE cycle and evolved through friction, etc. Intuitive guesswork has been my only recourse until now, and the loss-split method provides an eerie comfort level, since it can be derived from an otto cycle loss split.

I am distantly familiar with the basic thermodynamic laws and charting. I have never taken a class in thermodynamics (I am not an engineer), but I have sat down and studied some old thermodynamics books. Over the next days and weeks I expect to review these books in an attempt to plug in the info you have provided me. I will eventually get it characterized, and will then be able to move to the next level of the IPE.

I am learning, and I very much appreciate Brian's and your time in helping me fill in the blanks. I've long sought the value for "n" (as you mention) regarding air, fuel/air, and combusted gas - I know it is trivial stuff for many engineers, but I know of no one to ask these questions. "Dissociation" and other new terms you mention will also be reviewed and related to this project. To be sure, I will be focusing on the 20:1 and 69.8% numbers, as I know these are "otto" fundamentals which will provide clarity when I properly grasp them and plug them into the IPE. I do question whether the 80:1 expansion may actually be the number to plug in (or more likely a sliding scale value from 30:1 to 80:1 dependent on throttle position), and will try to figure this out.

I made a mistake when mentioning the Cummins/TACOM paper of 1984, as this is based on a diesel cycle engine, not an otto. Also, there is a Cummins/TARADCOM paper from 1981. I mention this to note these two Army projects are distinct, yet both relate to ceramic diesel engines. There are many cool charts in these papers which show ceramic and iron temperatures, stress levels, all kinds of charts I've never seen the likes of before. Most is way over my head, but, as time passes, some begin to make sense. I've based some of my ceramic understanding on them.

Thanks,
Shoe.

 

[DaveShoe]

I've been reviewing my Thermodynamics book since last posting, and it is more understandable than the last time I reviewed it, but it remains way over my head. I've now realized I am not yet able to step away from my split-loss approach, but I can stuff a bit more science behind my claim of how much heat goes out the exhaust. I intuitively claimed 10%, but that would require greater than 90% ideal cycle efficiency. Assuming an 80:1 expansion ratio (at continuous duty full throttle), I am now seeing this is optimistic.

Friction loss, thermal conductivity loss, and reradiation loss (heat soak?) are the other notable loss segments. Reradiation loss is energy that ends up going out the exhaust along side the "ideal cycle" losses.

I do wish I had a mathematical knowledge of thermodynamics. I will continue to plod forward, and will focus on learning the math. It all comes in time. Thanks for your help.

 

[black2003cobra]

This isn't at all technical Dave, but you might appreciate it.

"Thermodynamics is a funny subject. The first time you go through it, you don't understand it at all. The second time you go through it, you think you understand it, except for one or two small points. The third time you go through it, you know you don't understand it, but by that time you are so used to it, it doesn't bother you any more."
A. Sommerfeld

 

[DaveShoe]

Thanks, I needed that.

I look at good science more as an art than a rigor. It is so much fun to bang heads with others, as we all have our scientific strengths and together we can design stuff that cannot be conceived by one person alone. To be sure, this thread has provided a great deal of advanced insights which will help prepare me when I visit our local university in two weeks to talk to a mechanical engineering professor. I will better fathom what I don't know, as I know I won't be given any time if I appear too clueless.

I am reviewing each technical sentence of this thread, as it is critically impotant that I either factor such items as the Wiebe function, or else that I explain why the Weibe function may not fully apply. That I now understand a Wiebe function and dissociation exists will help keep me in conversation.

I remain optimistic that the IPE design is worth continued study. Until I learn of another similar experiment having been performed I will have to live with this puzzle, as it is such a fun challenge.

It was mentioned above that the IPE would only develop 10 bar during compression and 25 bar during combustion, but I don't believe it was recognized that the dynamic compression ratio is 20:1 (from IVC to TDC), and the dynamic expansion ratio at full throttle is 80:1. Because of this, I disagree with the 10 and 25 bar notations, and recognize the my compression ratio numbers were erroneously thought to be static compression ratio (i.e.: BDC to TDC). I claim 30 bar for compression in the IPE (plenty of energy for both gasoline or alcohol ignition) and 150 bar peak pressure during sub-stoich combustion to protect the ceramic (the IPE cannot run a stoich mix without pressure damage). These numbers would seem to support a highly productive 80:1 expansion ratio at full throttle, and would also support an intrinsically soot-free burn.

I have tentatively chosen 80:1 as the expansion ratio of the IPE, but that is mostly an ideal maximum at which mechanical tolerances/collisions become pretty fussy. A 50:1 expansion ratio might end up being more practical if issues such as friction become dominant (I don't believe they will). I targeted a "TDC to 05ATC" combustion window in my IPE description, knowing the IPE still performs well with a longer burn window. I hypothesize that a detonation-style combustion may burn 50% of the fuel between TDC and 05ATC and 90% by 12ATC, and find these numbers to be a functional compromise. Combustion experts can speed this up to improve efficiency, but the IPE does not require it. Whatever the flame rate is, I will simply aim for the shortest duration and run with whatever it is.

Valve actuation is not an issue at all, even in prototyping. While the prototype will have four camshafts per head for "proof of concept" versatility, two being servo adjustable to enable highly variable valve timing, I have described the "vacuum/rebound" method of the IPE as uniquely efficient and worth investigation, but ordinary valve actuation is also acceptable, allowing a relaxed 300 degree intake duration to replace the suggested 60 degree intake duration. Likewise, 270 degree low-throttle exhaust duration can replace the suggested 90 degree low-throttle exhaust duration. Port flow losses will increase with the elimination of vacuum/rebound, but port flow losses are not a large energy component. The idea is simply a fun design accent to consider, not a functional mandate. Still, I see "vacuum/rebound" working fine in the IPE, as port flow is so slight that valve lift on the order of .300 will likely be more than required.

I am not concerned with NOx emissions, as combustion is not that hot and duration is very short. I strongly doubt soot will be an issue in the sub-stoich gasoline CI environment needed to make the ceramic live without fracture, but soot would not be a show-stopper, just an emissions control cost adder.

I certainly don't have the math nailed, but I do have a strong feel for the IPE concept. I know the engine works, but don't yet know how efficently or reliably. My search continues until I either find a similar experiment that has been run, or until I figure out the math and present it properly.

Thanks,
Shoe.

 

[DaveShoe]

Hi again TDIMeister,

In your final paragraph you mentioned a concern regarding "expanding the burned gas past below atmospheric pressure at EVO", but this is another misunderstanding. I will review my paper to find the logical source of the misunderstandings, and will rewrite it to help assure others will not misread it. I am single-minded when writing without input, and actively search out these misunderstandings when I first present writings to others. My favorite subject of all is failure analysis. Just as I can find failure in mechanisms, I also search for them in my writings when provided the clues.

I do wish/hope you find a chance to re-review my paper, with the understandings that the dynamic compression ratio is 20:1 and the "vacuum" I speak of only occurs during a "dead time" prior to valve opening, for the purpose of reducing port losses. The vacuum is gone before the valves open, eliminating your (and my) concern of excess port wash.

I mentioned 30 bar for compression in in my prior post, but recognize your math suggests 66 bar (20^1.4) during a 20:1 compression. I can see this as a very fun possibility, but it is an unexpectedly big number, so I will have to learn more before it soaks in sufficiently that I can confidently print it. I do dig learning the 1.4/1.3/0.4 factors you mention for combusted gasses. If you can point to a textbook which might be of benefit in discussing these numbers, I would really dig buying it. My ancient Wark thermodynamics book may be great for a beginning college student, as I enjoy being in the dangerous position of having access to almost-understandable textbooks and basing research on them, but it lacks applied ICE formulas. If I could find a more topical book to base my conclusions on, I could be "that much" more dangerous.

Since I am a bit of a bookhound, I will review the sae.org library in the hopes of finding something that might whet my appetite on combustion formulas. First, I'll focus on a full understanding of the responses, and then I'll continue to review the many cool threads at Eng-Tips that pop up with the "adiabatic ceramic" search key, before looking for sae book resources. I also appreciate learning the IVC and EVO acronyms, as they are obvious but I've never seen them used before. I wll be using them in the future.

Thanks,
Shoe.

 

[carnage1]

Combustion occurring between 0 and 5 degrees ATDC is not realistic.

you need to aim for max pressure at 17deg atdc
the easiest place to find documentation for this is the naca technical documents server on nasa's website

Also the ability to operate at a wide range of revs translates to real world fuel efficiency. my max HP (150) is at 6000 rpm and it pulls strong all the way to the rev limiter at 7800
I get 40 mpg mixed driving and don't drive slow. at 1200 to 1500 rmp (where I cruse) I can run with the throttle open a bit as opposed to 2500 rpm where most people would run my 4 cylinder and I am sitting just off idle to choke the power down enough to stay at speed.

that isn't the most concise explanation of what I am trying to get at but look up bsfc maps and you should get an idea of what I mean.

 

[DaveShoe]

Thanks Carnage1,

Note that my target 4000rpm redline is simply an intuitive guestimate, as are all my numbers at this stage. Redline may be as low as 2000rpm or as high as 6000rpm. Better understanding of flame properties and inertial limits will define much of this.

The 17ATC notation is one I will look for in the NASA archives. If true for the IPE, it will certainly have a detrimental effect to the optimistic efficiency numbers I've posted, but reality and knowledge is what I aim toward. With luck, the 17ATC is optimal for a slightly different engine process, allowing my IPE process to be optimal closer to half that number. I look forward to learning how the number was derived.

It is fun to finally be at the math portion of this project. The conception part (in my case) doesn't get much exposure to formulas.

Your rpm numbers put a little smile on my face. While I've been immersed in the intricacies of my wimpy-powered IPE concept these past weeks, you reminded me I've got a neighbor/friend who is tinkering with my favorite engine, the old Ford 427SOHC. He has a cute little SuperFlow dynomometer in his milk barn that we have fun developing parts on when I'm not sculpting IPE theory, reference:

http://www.network54.com/Forum/74182/message/1210556768/

http://www.network54.com/Forum/74182/message/1206416859/

http://www.network54.com/Forum/74182/message/1199510914/

Thanks,
Shoe.

 

[TDIMeister]

My calculations do consider dynamic compression- and expansion ratios. Your figure of 66 bar with a compression ratio of 20:1 would be correct if the pressure in the cylinder is 1 bar at the beginning of compression (i.e. BDC). However, it is not.

As I explained, the moment you close the intake valve on the downward piston stroke, the charge will be expanded to a partial vacuum. If you assume an adiabatic expansion, the same polytropic equations would apply. At BDC, if you assume an expansion ratio of 1:4, the pressure would be 0.1436 bar, not 1 bar. From this point, when you compress the charge by a volume ratio of 20:1, you will get 9.52 bar.

As for reading, perhaps the definitive work on internal combustion engine fundamentals if by John B. Heywood of the Sloan Automotive Laboratory at MIT, or also that by Charles F. Taylor. The introductory text I used in my Bachelor days was Thermodynamics: An Engineering Approach by Yunus A. Cengel, Michael A. Boles.

I must have first read Heywood in 1996 or so, and I still have a copy on my bedside. It's not something that you read once and feel you've got a good grasp on everything.

 

[DaveShoe]

I am being a bit "figurative" and simplistic when advising the intake valve close at 55ATC, as the goal is to indicate a full atmosphere of air at 1/4 stroke past TDC. I will review a way to word this more correctly, as I wish my paper to be understood reasonably during an abbreviated first-time review.

I am excited to learn, assuming 1ATM/1bar at 55ATC equates to 1.4 bar at BDC, not the .25 bar I had stated. This makes the math quite easy to comprehend, though I also understand there are complications of heat soak, molar changes, and energy changes which frequently apply. I will start using the adiabatic compression values right away, though I will try to isothermally factor them a little.

I just remembered I have that Heywood book you speak of. It is a great one, but I didn't see it in my library searches of the past months. I just found it and recall this is the book I need to be reviewing right now. I really need to organize my library. Someday I will.

 

[DaveShoe]

Correction: 1.4 bar of my previous post should read 0.14 bar.

 

[DaveShoe]

TDIMeister,

I again see a conceptual misunderstanding you have regarding the IPE:

The static compression ratio from BDC to TDC is 80:1, but the IPE has a rather unconventional geometry, so I tryied to indicate the dynamic compression ratio is 20:1 and begins at 55ATC (or equivalently at 55BTC) and ends at TDC.

The 80:1 static expansion ratio (TDC to BDC) was intended to be the indicator of this factor, but it may be presented a bit too obscurely. Note that 0.14*(80^1.4) is the foumula I must present in the paper to clarify this.

Thanks for letting me see this.

 

[CCycle]

Interesting thread. Thanks to all who contributed.
The problem we are up against is the single containment container. This container must operate from the highest to the lowest temperature. There is most significant heat exchange with the container, all of which reduces efficiency. Insulating the container really is a moot issue because of the wide temperature swing.
It is helpful to study the development of the steam engine. As it developed techniques were applied to increase the expansion ratio. Early cut off with high steam pressure accomplished high expansion ratios giving high theoretical efficiencies. Steam engines can bottom well into the vacuum. But the single expansion steam engine still operates at less that 50% of its Carnot efficiency. It was not until multiple expansion engines with uniflow principles were developed that the efficiency improved.
A single wheel turbine is dismally inefficient Yet multiple wheel turbines when operating in their range can achieve high efficiencies.
No single expansion internal combustion engine, with one possible exception, can get above the 50% of its Carnot efficiency. The exception is very large engines that have high volume to surface ratios.
It is reasonable to expect that if and when a high efficiency internal combustion engine is developed it will be a multiple expansion device. In this case it becomes usefull to insulate the high temperature stage(s) to reduce jacket loss.