Highest RPM of any WWII Piston Engine

[Deleted member 68059]

.......See, that statement just baffles me. I've never encountered a team that said, "We'd turn more RPMs, but our conrods would break"......I'm aware of a motor that has a mean piston speed of 7540 fpm, 38.3 m/s. EN30B crankshaft ..... .

If you are baffled by F=ma thats an issue, as its been known since the year 1688 and if you`re going to spout engine figures like that you better have a reference for it. Quoting
mean piston speeds is fairly meaningless in terms of rod stress, as you need to know the instantanious acelleration figures over the complete crank cycle to calculate stress.
This requires that you actually know what you`re doing, which is why amateurs do things like saying "these pistons are good for such and such mean piston speed", this sort
of "I reckon..." design method descends from a time when people didnt have computers and didnt know enough maths to do it properly. Its ok to compare a speed
increase of the same engine but gets almost meaningless otherwise as the resultant stress is dependant on rod and piston mass, bore-crank offset etc etc etc,
which change radically from engine to engine.

I have already CLEARLY explained to you how compressive loading yield failure and also buckling failures
affect rod design, and explained the geometry limitation which is rod side face to bore base side clearance. One cannot simply just make the rod into
a solid blob because its self-defeating as the top half of the rod forms part of the reciprocating loading, therefore if you make the rod bigger
the loads go up as well ! This is why rod design is quite involved, and requires a very light compact component.

Its really very simple, and I`ve already shown you the theory, and the typical dynamic loadings, so feel free to proceed in ignorance if that is your wish.
Nobody can be forced to listen or learn, you have to choose to.


Secondly here is a selection of scientific papers on engine design. I really think you ought to read more before you post anything else claiming to be telling others about how
connecting rods impact engine performance, given that you have clearly never designed one.

http://www.ijeit.com/vol 1/Issue 3/IJEIT1412201203_02.pdf

https://www.sasft.org/~/media/Files/Autosteel/Programs/LongProducts/bar_connecting_rod_thesis.pdf

http://www.rs1234.wz.cz/img_m/ART_166.pdf

Making engines go faster is quite tricky, which is also why a man called Ricardo decided to write a book called "The High Speed Internal Combustion Engine" around
100 years ago (1922) which is worth reading.

The High Speed Internal Combustion Engine - Ricardo eStore

 

[eljefe]

If you are baffled by F=ma thats an issue, as its been known since the year 1688 and if you`re going to spout engine figures like that you better have a reference for it.

If that's a current photo in your avatar, it seems likely that I've been a degreed engineer for longer than you've been alive, so I think I have a nodding aquaintance with F=ma, thank you very much.

As for a reference, will this do? If you have a spare $120k lying around, I can have one delivered to your doorstep.

Mountain Motor: A Closer Look at the 1005ci - 2100hp Naturally Aspirated Monster From Sonny Leonard - BangShift.com

And here it is in action:

Quoting mean piston speeds is fairly meaningless in terms of rod stress, as you need to know the instantanious acelleration figures over the complete crank cycle to calculate stress. This requires that you actually know what you`re doing, which is why amateurs do things like saying "these pistons are good for such and such mean piston speed", this sort of "I reckon..." design method descends from a time when people didnt have computers and didnt know enough maths to do it properly. Its ok to compare a speed increase of the same engine but gets almost meaningless otherwise as the resultant stress is dependant on rod and piston mass, bore-crank offset etc etc etc, which change radically from engine to engine.

All of which matters if you're doing a detail design of a rod (or other engine part). It doesn't matter so much if you're trying to compare different designs. Does it really matter that team A uses alloy X to get 10% more stress at the same fatigue life than team B does with alloy Y? Espcially if team B has a clever design for the piston that reduces rotating mass. Overall, more piston speed = more hp, and more piston speed means that a certain level of sophistication has been developed to achieve it.

As for "didn't have enough maths to do it properly", the math has been around since at least the first edition of Charles Fayette Taylor in 1960, and I suspect that it was known in at least academic circles before that. It can be done and has been done with pencil, paper, and sliderules, albeit not to the precision that you can do with a decent workstation and FEM. You'd be amazed at how close to reality some of the simplifying assumptions that the old geezers used were to reality.

Secondly here is a selection of scientific papers on engine design. I really think you ought to read more before you post anything else claiming to be telling others about how connecting rods impact engine performance, given that you have clearly never designed one.

https://www.sasft.org/~/media/Files/Autosteel/Programs/LongProducts/bar_connecting_rod_thesis.pdf

Well, lookee here, from the 2nd paper on your list, table 2-3. Peak resultant load on the big end at 5700 rpm is at 360 degrees (tension), and peak compressive stress is at 180 degrees. Who'd a'thunk it? This math is so hard! Peak resultant stress on the little end is at the moment of peak combustion pressure, 24 degrees after top dead center. I bet I could give a 3rd-year engineering student the bore, stroke, rod length between centers, wrist pin diameter, conrod mass, piston mass, and cylinder pressure, and he could calcuate the stresses to within about 5% of reality given all that. And you can also see from tables 2-4 and 2-5 that combustion forces become more important at low rpm and less important compared to inertial forces at high rpm...something Dr. Grandpa could told me back when I was running around in knee shorts, although it would have been a few years until I was old enough to understand him.

Don't get me wrong. I don't mean to belittle things like FEM analysis. You can definitely shave a few grams off with that, and the only way to get the same results the old-school way was to break a LOT of conrods, which wasn't really practical. But nothing I saw in any of those papers comes close to disproving the notion that mean piston speed is a better first-order means of comparison between different engines than simple rpm, which is how we got to this whole thing in the first place.

 

[Deleted member 68059]

When the worlds most advanced racing engine firms have invited me all over the planet to present my engine research to their engineers, they were only iterested in WHAT I am bringing to the table and have never asked my age before inviting me to their premises, so thats a new one!

Once again you've utterly misunderstood, misrepresentated / willfully ignored everything I've told you.

If you refuse to learn or listen, there is literally nothing to be gained in continuing, is there.

 

[pbehn]

When the worlds most advanced racing engine firms have invited me all over the planet to present my engine research to their engineers, they were only iterested in WHAT I am bringing to the table and have never asked my age before inviting me to their premises, so thats a new one!

Once again you've utterly misunderstood, misrepresentated / willfully ignored everything I've told you.

If you refuse to learn or listen, there is literally nothing to be gained in continuing, is there.

I like reading your posts, I actually think the forum is honoured that you take your time. I also suspect that as complicated as your posts seem they are watered down and condensed for a lay audience.

 

[eljefe]

When the worlds most advanced racing engine firms have invited me all over the planet to present my engine research to their engineers, they were only iterested in WHAT I am bringing to the table and have never asked my age before inviting me to their premises, so thats a new one!

Please take some of your own medicine. Literally, I've been doing F=ma in one variation or another longer than you've been alive, and you implied that I was confused because I didn't understand F=ma. This wasn't about you. Not everyone who disagrees with you is interested in making personal attacks. It was about pointing out that 1) I'm not an untrained layperson, and 2) I seem to know the basics far, far better than you're giving me credit for here.

Once again you've utterly misunderstood, misrepresentated / willfully ignored everything I've told you.

If you refuse to learn or listen, there is literally nothing to be gained in continuing, is there.

Well, then please make your case, rather than just throwing out a few papers that don't contradict my initial point, which was that mean piston speed is a better basis of comparing apples to oranges than rpm.

For that matter, they don't seem to contradict my secondary point, which is tangential to this whole discussion: namely that for much of history and certainly now in 2018, for a clean-sheet design and very low TBO requirements, availability of materials does not constrain mean piston speed. You seemed quite skeptical that an engine with the stats I quoted existed, to the point of requesting a cite, which is a polite way of saying, "show me or you're a liar". Well, I showed you...and I made sure to give you a detailed breakdown of rpm, stroke, and parts suppliers, through which you could find materials used...none of which were terribly exotic for "the world's most advanced racing teams". I also made a point of showing it running and making a pass, just to prove that it wasn't a shop queen that was designed to a specification that was never reached.

Now, I'm an engineer, and I'm much more interested in facts than hype or pontification. The Internal Combustion Engine in Theory and Practice really, really pounded into my head that Power = BMEP x Piston Speed back when I was young and thought I knew much better. If that's no longer true, could you simply explain it to me, a grizzled old engineer who is past his prime? Those were nice papers you linked to, and I enjoyed reading them, but I don't seem to be making any connection between the ground they cover and the points you seem to be trying to make here. Ernest Rutherford is said to have proclaimed, "An alleged scientific discovery has no merit unless it can be explained to a barmaid." You have a reputation here for being able to explain the fine points of engine design to a barmaid (or at least, to a few history buffs), and it would give me great pleasure to see you live up to that reputation.

 

[Deleted member 68059]

Well, then please make your case, rather than just throwing out a few papers that don't contradict my initial point, which was that mean piston speed is a better basis of comparing apples to oranges than rpm.

You have a reputation here for being able to explain the fine points of engine design to a barmaid (or at least, to a few history buffs), and it would give me great pleasure to see you live up to that reputation.

Neither are particuarly good, and I wouldnt use either for anything other than the first rudimentaty layout of what the boundary conditions of the engine are. When engines were first designed those with analytical techniques in slider crank linkages were few and far between (namely those who`d been working on locomotives and suchlike), so engineers didnt really know what the actual loads were in the pistons and rods dynamically. So you`ll find books from the 20`s and 30`s (and even beyond) giving things like tables of "acceptable mean piston speeds" for various kinds of engine, piston material type and so on. As rule of thumb analysis was common.

Once it was possible to analyse the loadings dynamically, and start looking at vibrations and so on, things like mean piston speed in the context of actual mechanical design of actual components was consigned to history, as all it is - is a very basic overview of what might work if the designer didnt really know what to do, but had previous sucessful designs to develop upon.

I can do in my programs what needed an entire stress department of the best dynamics engineers in Britain in the 1930`s, so its no wonder anyone less than Rolls-Royce or similar were not really looking beyond basic guidelines for what might work.

RPM isnt very useful by ITSELF either, because it does not directly relate to any stresses in the moving components (other than balance weight effects), because the accelerations in the moving components are what the designer requires to begin with to carry out the dynamic stress calculations. These accelerations are NOT directly related to mean piston speed, because if you double mean piston speed the acceleration does NOT double, it goes up a LOT more. So for example you cannot say if we double mean piston speed the piston stress doubles, it goes up vastly more and how much is dependant on the kinematics of the engine, the rod length, bore, stroke, pin offsets etc. Its just not useful as a value in any real aspect of stress dynamics, which is the core aspec of all engine mechanical design practise.

You cannot design any component of an engine using RPM or mean piston speed as core metrics, no engineer who is actually desiging real parts would say otherwise.

I do not understand why you think finding a dragster engine with high mean piston speed means F=ma doesnt apply to engines. It has a redline, just a higher one than would have been possible 50 years ago. The limit is simply higher, but it most certainly has a strict rpm limit above which it will throw a rod out the side of the block.

An engine with no practical worth whatsoever outside of the 1/4mile strip, can reach higher limits as the components such as con-rods can be run to their actual yields. Which is ONE reason they often use aluminim rods, as the fact alumiumium has no lower fatigue limit at all is alrigh when you throw the bits in the bin every few hours or even minuites (the other being that the very low youngs modulus prevents top-fuel-ers from hydraulic-lock destruction as they run so much fuel through it can actually fill the chamber during a run, and cause a steel rod to buckle as it wont stretch much elastically.

If you must make any engine other than for 1/4mile, you must run below not only the yield limit but below the fatigue life limit lines for each material, which designers now will have data-sheets for, at multiple temperatures called Goodman diagrams. In the case of steel parts a very rough approximation in terms of fully reversing stresses (as is the case in the rods) which are the most several fatigue cases, you can start by saying the stress must be less than 50% of yield to avoid fatigue failures. In fact once temperatures go up it is even less. Therefore once outside the 1/4mile strip, your rod limits HALF instantly as you must avoid fatigue damage. Toyota engine design department in Tokyo for example design all major engine moving steel components for infinite fatigue life, which is very extreme and explains their outstanding reliability reputation worldwide. Pistons and heads cannot be designed for infinite fatigue life as they are permanently damaged by each individual fully reversed load cycle. This is particuarly problematics for cylinder blocks and heads, which when made from castable alloys must be kept below 200MPa stress at operating temperatures to survive for any reasonable amout of time (to know stuff like this you need high temperature goodman diagrams, although Goodman published the concept over 100 years ago, only very well run, large and high-tech companies do enough testing to have real data to make such
information available. You`ll find a VERY hard time getting a goodman diagram for 200 degrees C out of most so called "Modern" aluminium foundries ! - without such data, its basically nearly impossible to actually calculate the lifespan of engine components at drawing board stage. Which is why things like sticking to known limits of mean piston speeds was a popular way of staying inside "known safe areas". Modern fatigue data and dynamics calculations merely reduces this level of ignorance one step further, to enable designers to get closer to the limit without having serious failures in service, or having to spend too much money on years of testing before releasing new engines into service.

Top fuel-ers dont even bother putting cooling into the heads, such is the irrelevance for fatigue life to them, as the parts will all be scrap long before the fail in fatigue, as they are running at almost yield constantly. The improvement in materials simply allows them to run more rpm and more boost, but this is simply a HIGHER limit, there are very strict rpm and boost limits even a top-fuel car is limited to.

The firms I visit to present to include Ferarri F1 Maranello, Renault F1 Viry, Renault F1 Enstone, Mercedes AMG High Performance Powertrains Northampton, and top engine design universities such as Modena. Honda are the only F1 engine constructor in the world who have not personally invited me to give their engine design departments lectures. Although I have to say that its a slight cheat for me to mention Mercedes AMG Northampton (where the engines of the last 4 world F1 world championships in a row are designed, machined, assembled, tested and developed) as having invited me as I was working there at the time, so at that point I was not really an external speaker.

Not very many people get invited to F1 teams to lecture on engine development.

Thats the best I can to do in any reasonable fashion to answer your point, and if that it doesnt work for you - you`re out of luck, as to do more I`d have to design an engine with you sitting here at my PC whilst I explain it step by step.

 

[eljefe]

Thats the best I can to do in any reasonable fashion to answer your point, and if that it doesnt work for you - you`re out of luck, as to do more I`d have to design an engine with you sitting here at my PC whilst I explain it step by step.

Ok, now I can see where you're coming from, and there is very little I would disagree with in that last post of yours.

If I haven't been perfectly clear about this, I'm not recommending piston speed as a prescription for engine design. If you are arguing that was how it was done back in the day, and time and technology has passed that theory of design by, I am in complete agreement with you. Rather, I am suggesting it as a metric for comparing finished work....whether I give the paper an A or a B. I still think it's a better metric than pure RPM, and that's why I brought it up here in the first place, simply as a response to the original poster.

And I think an even better metric, for both WWII aircraft engines and for modern race engines (all things, including rules being egual) is power/weight ratio. Add to power/weight "at a given level of reliability", and you almost have a complete metric. Throw in brake specific fuel consumption, and I don't know what else to add. Maybe some kind of packaging constraint.

But ultimately, I don't know why the original poster was interested in RPM, and I just put out mean piston speed to jog his thinking.

I do not understand why you think finding a dragster engine with high mean piston speed means F=ma doesnt apply to engines. It has a redline, just a higher one than would have been possible 50 years ago.

Well, that's exactly the point of our little side discussion on conrods. What determines that redline? There is a redline where parts will break. There is another rev limit, one in which even if all the parts were made of unobtanium, that the engine would not exceed. I'm simply suggesting that in the modern era, getting air into the engine at high mean piston speeds will be the limiting factor. The immovable object is currently "winning" the arms race against the unstoppable force. That doesn't mean that if higher reliability, or a lower budget, or some other factor comes into play that the calculus doesn't swing back the other way.

My observation is that induction advances to a certain point and gets stuck, and people develop parts techology to support the speeds associated with that level of induction technology. Then the state of the art advances, people start breaking parts and whine to the parts manufacturers, and lo and behold, within a season or two that technology advances and people now have more or less reliable parts. It just seems like induction is the driver, and that it always takes less time for materials to catch up to induction advances than it takes for those induction advances to happen in the first place.

The improvement in materials simply allows them to run more rpm and more boost, but this is simply a HIGHER limit, there are very strict rpm and boost limits even a top-fuel car is limited to.

I think we are both in agreement that drag racing occupies the extreme point on the performance/reliability curve, and that unusual considerations there drive them to do odd things that are not applicable to either F1 racing or to WWII-era aircraft engines.

And having said that, I maintain that the real limit on Top Fuel cars is the trap speeds that insurers are willing to cover. I honestly think that the NHRA is running out of constraints to put in the rules to slow them down, and the move from 1320' to 1000' is a blatant admission of that.

 

[gumbyk]

Every aircraft piston engine that I've seen that has had a failure due to overspeed has been in the valve-train, usually broken springs, but also valve seat damage.

 

[Deleted member 68059]

Ok, now I can see where you're coming from, and there is very little I would disagree with in that last post of yours.

Well thats an unusual sucess for the internet, two people eventually agreeing they dont have very much to argue about after all. Very good.

 

[fliger747]

Aircraft engines, at least the large ones we are discussing from WWII used rather large pistons. All the discussion about moving parts, weight etc all quite valid. C series R2800's had the rods quite nicely polished in order to relieve surface cracking along imperfections which concentrate stress. Yes, pushrod engines also tend to begin to have issues with valve float and other valve train issues as well.

However the large diameter of the pistons introduce another issue, the speed of flame front propagation outward from the spark plugs. This is one reason such engines use a dual ignition system, aside from redundancy. The flame front can move across the piston in less time if starting from two sources. This factor puts another potential limit on useful RPM. Inline engines may have had some structural advantages here and did tend to run a bit faster than the radials. The R2800 was much more tolerant of high MP than over RPM.

 

[fliger747]

The Merlin had a bore of 5.4" and a 6" stroke. The R2800 had a 5.75" bore and a 6" stroke. The C series R2800 engine had many improvements over the b series, including a reworking of the oil scavenging system. Max RPM raised from 2700 to 2800. The oil system improvements supposedly reduced losses from slinging oil around in the case by several hundred HP. The R2800 has about a 6% larger bore than the Merlin, but the latter engine could run about a 9% faster RPM.