RPM -- F1 vs Road Car

[Reca]

v = v0 + a * T

Piston speed at both 0 (v0) and 42 mm (v) is zero => the average acceleration a is zero.
If average acceleration is 3400 g the piston speed at 42 mm is 3400 * 9.81 / 630 = 53 m/s.

[Monstrobolaxa]

Well what I've done is not 100% correct......you're right! The thing is that in the case of the piston (under the conditions I put) the positive aceleration when starting the upward is the exact same as the aceleration (in modulus) as the aceleration that the piston suffers while it "decelerates" to change the upward movement to the downward movement.....this is why the sum of the acelerations is 0....(upward and downward). So Reca you're correct.

In my case what I did was I only considered that the piston only acelerated on the way up....and didn't "decelerate"! If the piston only went up and took an instant to stop the average aceleration would be zero in any case! Now the thing that makes my calculations wrong is the fact the the piston doesn't stop in an instant like I mentioned...and I considered the calculations for the instant before it stops.....so more or less what I did was...considered that:

it took: 0,001587725 seconds to go up....only acelerating and it took 0 seconds to stop....since I only considered the 0,001587725 seconds I only calculated the average aceleration for a piston the during this time only went up and didn't loose speed (which isn't correct). So I considered the piston as a thing the only acelerated and stop in an instante.

Well in any case living and learning

[ReubenG]

A better (but still innaccurate approximation) of the piston movement is to treat it as a body undergoing simple harmonic motion - i.e. displacement is a sinusoidal function where the 1/2 the stroke gives the amplitude of the oscillation and the rotational speed gives the frequency of the sine function: x= L*sin(wt).
If you differentiate twice then acceleration magnitude is given by
a= L*w^2.
So if you assume a stroke of 42mm and a rotational speed of 19000rpm, then the peak acceleration is 83 000 m/s^2 or approximately 8500g!!

The innaccuracy of this lies in treating the piston motion as SHM, with a single sinusoidal function - it obviously is a more complex function with a few more higher frequency sinusoids thrown in. When you differentiate twice to get acceleration then the (w^2) term can get really big.That being said, hen we did Mechanics of Machines, we only looked at the first two sinusoidal terms when trying to balance a reciprocating piston engine.

[DaveKillens]

Please remember, the piston is connected to the crankshaft, which rotates. The piston never comes to a crashing stop at TDC, or BDC. Instead, because of the crankshaft, it's peak velocity decreases the closer to the limit of travel it reaches. In fact, the piston maximum speed is at the middle of the stroke, as determined by the simple geometry of the crankshaft, piston travel, and rod length. From the middle of the stroke, the piston begins to decelerate, until it reaches it's limit of stroke. If memory serves me correct, if you graph the crank location versus the piston velocity, it appears as a sine wave.

[riff_raff]

What kills F1 engine components is mechanical loads. Since force=mass x acceleration, its obvious that piston velocity is meaningless. It's acceleration that matters. And it's well known that, kinematically, a short rod produces higher peak piston accelerations than a long rod.

To make matters worse, a connecting rod (in a N/A four stroke engine) undergoes a complete load reversal once per revolution. It's in tension about TDC on the exhaust stroke, and in compression about BDC on the intake stroke. Reversing loads like that are critical with regards to fatigue life.

Ultimately though, the reduction in mass provided by a very short, lightweight titanium rod, more than offsets the increased accelerations produced by that short rod geometry. And that's the solution adopted by most high speed engine designers.

[Reca]

Ultimately though, the reduction in mass provided by a very short, lightweight titanium rod, more than offsets the increased accelerations produced by that short rod geometry. And that's the solution adopted by most high speed engine designers.

High speed engine designers appreciate a lot the reduction of acceleration given by a long con rod, but “long” in the right sense, ie looking at the ratio con rod / stroke. Since high speed engines have usually very short stroke, the rod is, kinematically, pretty long, even if, physically, is possibly a bit shorter than that of a low speed, long stroke, engine.

[riff_raff]

Reca,

I've worked with both low speed commercial diesel engines and high speed racing engines. For high-efficiency, low speed engines, you always strive for an "under-square" bore/stroke ratio. Minimizing piston area produces low heat rejection and low wrist pin/connecting rod loads. To minimize rod angulation/piston side thrust, and thus friction losses, you want as long a rod as will physically fit within the engine.

For a high speed racing engine, the opposite applies. You want an "over-square" bore/stroke ratio. A big bore diameter allows maximum valve size, and a short stroke produces low piston accelerations. A short rod and a slipper skirt piston with low deck height minimizes reciprocating mass. The lower limit for rod length is usually determined by by piston-to-crank interference at piston BDC location.

[Reca]

Riff_raff, my point is that "the reduced mass more than offsets the increased acceleration" makes sense only if you relate it with the short stroke.
You say that on slow production engines you try to fit the longest rod you can, a typical value of conrod/stroke ratio for a slow engine is in the order of 1.5 - 1.7. In the 2.0 l Honda S2000 engine, a quite high revving engine for production standard, the conrod/stroke is about 1.8, the con rod being about 150 mm and the stroke 84 mm.
If now you look a F1 engine, the lower edge of the piston almost make contact with the counterweights, hence you can say that it’s physically very short. Still, that very short rod, exactly because of the short stroke, is kinematically very long. Take as an example the 049 Ferrari engine (2000) the con rod is little less than 120 mm and the stroke is 41.4 mm IIRC, that gives a ratio in the order of 2.7-2.8.
To further increase the length at that point is less beneficial because as you increase the conrod/stroke, the piston vs crank angle gets closer to a sinusoid and the potential in reducing the peak acceleration becomes smaller.

[riff_raff]

Reca-

Thanks for the reply and I enjoy this dialogue tremendously.

When I speak of low speed, production engines, I mean 1800 rpm diesels.

Also, you mention counterweights on an F1 crank. I've seen a couple of F1 cranks, but the ones I've seen typically are not counterweighted. The purpose of counterweights are two-fold. To reduce main bearing loads/crankshaft bending stress and to reduce engine vibration due to unbalanced inertia forces.

With the over-square bore/stroke ratio and lightweight reciprocating components of current V10 F1 engines, unbalanced inertia forces and couples are not seen as a problem. Also, with the very short stroke cranks used, they have lots of rod/main pin overlap in the crank structure. Thus the crank tends to be torsionally stiff, and has relatively low peak stresses at the critical fillet areas. F1 cranks are made of premium materials and have a very limited fatigue life requirement. So their allowable stress limits can be quite high.

Apparently, the added mass of counterweights is not worth the increased moment of inertia and the reduction in engine vibration they provide.

Everything in engine design is a study in compromise, right?

Regards,
Terry

[Reca]

http://www.nona.dti.ne.jp/~f-garage/item11.html

[West]

about 4000-4200 US

The site says something about testing; maybe it wasn't used in a race.

[Guest]

Hi,

I was wondering if anyone knew how f1 cars are able to reach such high revs? What would happen if a road car had this many revs? Is it the difference in revs due to the complexities of f1 cars?

thanks

Gabe

So, F1 cars are able to reach such high revs because all the internal components are much, much lighter (and made from far more exotic materials) than a road car. The only way a road car would have as many revs would be to make the engine more like and F1 engine and accept the fact you would need to change it often (and run it in very controlled conditions warm-up/cooling etc.)

If you try to rev a road car engine (any I can think of) to 19000rpm - then pretty soon some of the bits inside will join you outside and it will stop.

The difference in revs is not so much that F1 cars are more complex - perhaps more accurate to say they are very highly developed to produce lots of horse power with a limited life.

Make a road car engine smaller capacity with lighter bits inside and revs can approach F1 revs - think motorbike engine (still not all that close mind you!).

[riff_raff]

Reca-

Thanks for the picture of the Ferrari crank. Interesting indeed! Like I mentioned though, the last time I saw the internals of an F1 engine was 1995.

Ferrari always does some unusual things with their F1 engines. Some are good ideas and some aren't. Remember a few years back (1993?), the Ferrari V12 F1 engine had a cast iron block. And also, what about the Comprex supercharger they tried back in 1980?

Looking at the crank picture you linked, there are some design features I would question (as an amateur engine designer!):

First, is the extensive use of tungsten slugs in the counterweights. Apparently, it worked OK, but to me it seems a rather risky approach in an extremely high rpm race engine. One of those slugs coming loose at 17,000 rpm would do a lot of damage. A better approach to maximizing counterweight MOI at minimum mass, would be a more sophisticated counterweight shape (ie. a minimum thickness inner web and maximum thickness outer web, thus putting the counterweight mass where it counts).

Second, the distribution of counterweight mass along the length of the crank does not seem to be optimized. Usually, counterweight mass is biased towards the extreme ends of the crank as much as possible. While this may introduce small couples at the end-most crankpins, it also significantly reduces torsional oscillations in a long crank. The reduction in stress levels due to torsional vibration more than offsets the small increase in bending stress due to the counterweight assymetry about the rod pin. If you've ever seen a 180 degree crank from an old Cosworth DFV or DFX V8, you would note that it only has small counterweights on 4 of the 8 crank cheeks (the two inner most and the two outer most), for this very reason.

[Spencifer_Murphy]

Reading this Months issue of F1 Racing, I saw an article about the new V8's for 2006. Apparently they could run at the same power as current V10's. Some disagree, some agree.

In same article they talked about how Cosworth have already tested their 2006 V8 on the Dyno...at 20,000RPM!!!

[bcsolutions]

Lets hope that they are! This is the first year for a while that the cars are generally actually slower than a previous year. It seemingly has little to do with the 1 engine 2 races rule and far more to do with the new front and rear wing regulations implemented. The engineers are really earning their crust and it'll be interesting to see how next years regulations will affect lap times.