[vina]

There are several threads on engine design related stuff that inevitably boil down to diesel vs. petrol (e.g. this thread: http://www.team-bhp.com/forum/techni...ue-vs-bhp.html)

Many of these discussions boil down to max. rpm an engine can do. So here's my question (and I have no clue as to what is the definitive answer, or whether there is one):

What factors limit the max rpm for internal combustion engine and how?



May be the gurus can give their feedback on this one (please don't make this a diesel/petrol thread)

[bzr77k]

Thanks for starting this thread. I feel the same about threads leading to Petrol vs Diesel all the time...

Back to your question, i would like to repost what i had posted on another thread... as i think its linked to the title of the thread

why does the torque of an engine fall from a peak?
I understand that the torque would be low at very low RPMs as the amount of energy generated is low due to lower RPM. But then, as the RPM increases, what are the factors that cause the torque curve (and as an effect the power curve) to fall?

i can think of mechanical restrictions due to valve operations as one possible cause... the time taken by valves to open/close once the command is issued is fixed (a few microseconds (us)). This fixed time would be less important at lower RPMs as the overall valve operation time would quite large (a few ms against a few us) and at higher RPMs this time becomes more significant and imposes a restriction on the overall operation time of the valves...

is my understanding above correct? if no, then where am i going wrong?

also, any other factors that cause the torque curve to drop from the peak?


Would like to have the experts pitch in and share their knowledge. Lets keep the thread true to the title

[dhanushs]

Let me list the reasons known, as far as my tiny knowledgebase goes...

1.. One of the issues would be stress generated in the related parts. AFAIK, it increases exponentially as rpm increases.

2. OK, assume you have a perfect alloy capable of handling all those stresses, then the next issue would be, the valve mechanism. The valves NEED to stay closed even at high rpm's so that the piston can compress the air/mixture. At high rpm's this is very difficult.

3. OK, now assume you have perfect alloy and a perfect valve mechanism, then IMO, the Volumetric Efficiency or the output will become so low at a point that it cannot accelerate the crankshaft.

[Sutripta]

Mechanical limits:- result of inertial loadings because of the reciprocating nature of major subassemblies. Stress points will be con rods, conrod bolts, bearings, and the piston itself. Piston friction. Metric used is Mean Piston speed.

Valve float. Valves also reciprocate.

Gas dynamics.

Regards
Sutripta

[julupani]

What theoretically limits an engines rpm would be the ability to limit any higher. Basically the engine does not produce any more power to power itself, that is keep itself turning over. Thus at one point when the engine produces just enough power to be running itself, it will sort of reach an automatic rev-limiter.

This is also the reason why the engine's idle rpm needs to be suitably high to keep it from stalling. That is the power generated at idle rpm must be enough to power the engine itself.

The basic reason why the available torque starts to fall after is the amount of energy spent in keeping the pistons reciprocating, the crank/cams pulleys, alternator, turbo etc running rises higher.

With rising rpms, the energy produced by the engine rises at a faster rate than the energy required by it. But after a while the rate catches up and the situation is reversed. And at one point the actual energy required and produced also equalise, like I said before.

As an analogy compare the two curves Y1=X and Y2=square of X. Till the point X=1/2, Y1 rises faster than Y2, but after that Y2 rises faster than Y1. And till X=1, Y1 is greater than Y2, but after that Y2 becomes greater.

Its sort of a similar reason, the reciprocating energy depends on how long the reciprocation is and of how much mass. But the energy depends on the square of the distance, thus causing a situation like above.

Thus we have short stroke engines like those used in F1 and in motorcycles, which can rev easily beyond 10,000rpm. This is becuase the short stroke requires less energy to keep itself rotating. Though yes, shorter stroke has other disadvantages too, mainly decreasing the amount of torque made. But in a bike/racing car total weight is very low, requiring lesser torque, while higher rpms allow for higher speeds.

Of course, the above explanation is very basic. The true complication lies in the fact that how many different factors actually determine the amount of energy required to keep an engine ticking over at a particular rpm.

[bzr77k]

Originally Posted by julupani

With rising rpms, the energy produced by the engine rises at a faster rate than the energy required by it. But after a while the rate catches up and the situation is reversed. And at one point the actual energy required and produced also equalise, like I said before.

As an analogy compare the two curves Y1=X and Y2=square of X. Till the point X=1/2, Y1 rises faster than Y2, but after that Y2 rises faster than Y1. And till X=1, Y1 is greater than Y2, but after that Y2 becomes greater.

Its sort of a similar reason, the reciprocating energy depends on how long the reciprocation is and of how much mass. But the energy depends on the square of the distance, thus causing a situation like above.

So if i understand correctly, the energy required by the engine depends on the square of the rpm and the energy produced is proportional to the rpm (the latter i understand, but not the former).
If this is the case, the available energy for effective usage would be the difference between the energy produced and the energy consumed itself and hence the curve which falls from its peak

just re-phrasing your statements to confirm my understanding

[sgiitk]

Another limitation is the time taken for the stroke. The engine is called under-square if the stroke is longer than the bore, and overs-square if the stroke is shorter.

An under-square engine has better torque lower down the rpm curve, but will in general have a lower rev limit. An over-square engine will have a higher rev limit, and a narrower power band. The peak torque will be much higher.

Undesquare was popular due to the old RAC formula of 'Taxable HP' which favoured these.

Also, due to inertia a DOHC should be able to rev higher than a SOHC which will be higher that a push-rod design. Of course there are generalities and there will be exceptions.

Mobike engines tend to be over-square.

One classic case of a very short stroke engine was the FIAT 128 (1300cc) in the 70's It had a rev. limit of 8,300 (carb engine!).

[vijaycool]

The limitation of max rpm is be several factors like combustion time, stroke length, valve float. But max rpm is influenced mainly by one factor "Mean piston speed". Theroritical MPS is around 28m/s . The engine speed is directly influenced by the inertia forces on the piston. When the velocity of the piston crossed this value there is every chance of crack on the piston crown.
In simple terms: Slow speed engine with 90mm stroke @ 4500 rpm has the same inertia forces when compared to the high speed engine with 45mm stroke @ 9000 rpm.

Regards,
Vijay

[julupani]

@sg Sir

True, design of the engine is very important. Good example about engine dimensions there.

You gave an example of from old times in the FIAT 128, one from the newer generation comes to my mind. Its the Honda S2000, whose 2L 4-cyl naturally aspirated VTEC engine
revs upto 9000rpm. Though that is more thanks to the dollops of power the VTEC system allows to be generated at those high rpms.

@vijaycool

Right, MPS is another way of looking at it. But MPS is just another way of talking about the energy used up by the engine itself, of course you cant really compare MPS across engine designs. But for a single engine MPS is a great way of looking at it.

@bzr

Yes, you got the idea correctly.

[Born2Slow]

Originally Posted by julupani

The basic reason why the available torque starts to fall after is the amount of energy spent in keeping the pistons reciprocating, the crank/cams pulleys, alternator, turbo etc running rises higher.

With rising rpms, the energy produced by the engine rises at a faster rate than the energy required by it. But after a while the rate catches up and the situation is reversed. And at one point the actual energy required and produced also equalise, like I said before.

Its sort of a similar reason, the reciprocating energy depends on how long the reciprocation is and of how much mass. But the energy depends on the square of the distance, thus causing a situation like above.

Thus we have short stroke engines like those used in F1 and in motorcycles, which can rev easily beyond 10,000rpm. This is becuase the short stroke requires less energy to keep itself rotating. Though yes, shorter stroke has other disadvantages too, mainly decreasing the amount of torque made. But in a bike/racing car total weight is very low, requiring lesser torque, while higher rpms allow for higher speeds.

Of course, the above explanation is very basic. The true complication lies in the fact that how many different factors actually determine the amount of energy required to keep an engine ticking over at a particular rpm.

while I certainly agree with you that the energy lost in friction increases much higher rate as the rpm increases and the engine will hit the max rpm(if not limited by other methods earlier) when the energy produced becomes equal to what the engine consumes. But this is not the only reason why an engine power starts to fall after a while. One of the main reason is that as rpm goes higher there is less and less time available to fill the cylinder with air and the volumetric efficiency goes down. Also I think the main reason behind using larger bore in high speed engines is that for the same cubic capacity a larger bore design has lower piston speeds and allow larger valves which are needed because of the less time available for cylinder filling at higher rpm.

[Born2Slow]

Also can you give me an example of two engines with similar capacity (cc) were the larger bore design gives less max torque than a long stroke design.

[sgiitk]

@Born2Slow; We are talking of higher torque at low rpm, not across the band. At the top a more highly developed engine (most of these are under square) can have higher torque.

[julupani]

@born2slow

True volumetric efficiency decreases, but this does not rise to a level that engine will have no air. Just less air, thus less power. Like I said, what I presented was a very very simplified model, outlining the basic idea of why an engine cant rev any higher. Of course there are lots and lots of complexities in the actual power produced by the engine, actual power required by the engine etc etc. The actual torque and power curves are effected by a massive number of variables.

As for the requirement of large valves, you could just have a large engine and then use massive gearing to gain the speeds. Which is what racing cars of old times did, they had massive cylinders, with 8L or 12L V8 engines. But that just compromises on actual combustion efficiency.

[Born2Slow]

Originally Posted by julupani

@born2slow

As for the requirement of large valves, you could just have a large engine and then use massive gearing to gain the speeds. Which is what racing cars of old times did, they had massive cylinders, with 8L or 12L V8 engines. But that just compromises on actual combustion efficiency.

. But to produce more power from the same cubic capacity we have to go large bore and short stroke. Like some of the Super bikes produce 180 bhp or more from 1liter engines because it is difficult to put large capacity engines I guess.

[vina]

Originally Posted by Born2Slow

. But to produce more power from the same cubic capacity we have to go large bore and short stroke. Like some of the Super bikes produce 180 bhp or more from 1liter engines because it is difficult to put large capacity engines I guess.

Thanks to everyone for great and passionate responses. I have always wondered some things said in several places:

1. high rpm usually means oversquare design
2. diesels, and specially turbodiesels have lower rev-limits than petrols (and even turbo petrols)
3. high rpm engines do not produce good torque at low rpms

thanks to everyone, I'm beginning to get some idea of the conventional wisodm. (1) above may be due to (C) (D) and (E) below (keep reading).

(2) above may be because first of all most applications of diesel do not require high rpm is the first place and also because high compression requires long strokes leading to (1) above.

I can't particularly see any reason for (3) above so far except that an engine that revs very high and hence generates high power, generates (relatively) low torque all through its rev range - it is just that people notice at low rpm and say "what, no torque?" .




(A) I don't think friction losses increase so fast that they can swamp engine's power. Looking at the torque curves of MJD diesel and petrol engines - non-turbo engine torque remains substantially flat before suddenly and sharply dropping. If the torque remains substantially flat there is no way friction could have risen very very fast.

I'm not saying that the viscous friction behaviour is not non linear (I think it varies as the cube of Mean piston speed, not merely square). Just saying that even if it is non-linear, it may not get a chance of becoming so large as to dominate the limiting behaviour.


(B) Stresses will largely rise linearly, since as long as torque remains constant power also rises largely linearly, this should not be a problem. Though it can limit the engine rpm by compromising long term reliability.


(C) VE is another matter - if the cylinder can not fill up fully at the requisite pressure then it wouldn't be getting enough O2 (and not enough O2 means you can swamp it with fuel, it just won't generate any torque), and also higher rpm - higher power - higher temperature will make things worse.

(D) Valves require time to open and close - cams have smooth profiles and springs have their own behaviour. I guess this will be a problem at high rpm as people have suggested, and will make the problem of VE worse.


I don't see anybody taking about it, but I guess the following are other reasons (Correct me if I'm wrong):


(E) Time needed to burn the fuel completely will limit the minimum time in ignition/combustion stroke; flame velocity for a given fuel will limit max. piston speed for long-strok engines.
(F) At the end of the ignition/combustion stroke the gas still has some energy left - not extracting it will lose FE, while extracting it will take some more time and hence reduce rpm (and peak power and well as torque - the last part of the extraction is from relatively low pressure gas). Since race cars don't care too much about FE and emissions their engines can have a higher rpm.