22 replies to this topic

[Ali_G]

This has really come to my attention latley.

Why is it that all the teams run engines which have capacities of 2997 or 2998 cc when the legal limit is acctually 3000 cc.

I know 2 cc is tiny, but I'm sure it would give some bit of a performance advantage. And in the cut throat world of F1 any bit of a time advantage is an advantage.

Niall

[Ben]

Because the tolerance on capacity is zero above the 3000cc so if this was the design capacity any oversize in the bore or stroke would make the engine illegal so they build to just below the limit to ensure the actual capacity is never over the limit.

Remember engineering is never totally precise and any published dimension should be thought of as 'nominal' and subject to variation in reality.

Ben

[Ali_G]

I'm not very up in the maths of engines, but what sort of a difference would there be in the diameter of an F1 engine cylender with a 2cc increase in capacity.

Niall

[Jezztor]

Depends on the stroke of the engine...what formula / class / category of vehicle are you referring to? F1, or in general?

Jezz

[Ali_G]

F1 engine, if the stroke stays the same.

Niall

[H. Eckener]

To find the displacement of any engine with round cylinders, you take the (area of the bore)X(the stroke)X(number of cylinders). Or more mathematically stated.

Vd = pi*((b/2)^2)*s*n

Vd = displacement
b = bore
s = stroke
n = number of cylinders

You are essentially finding the volume of a cylidner.

If you want to determine the effect of varying the bore and or stroke on the displacement then the best way to do that is to take the partial derivative of Vd with respect to b and another one of Vd with respect to s, from there you will know how the variables in the function relate to the change in Vd.

But, that is a pretty big gun for a little problem, just write an Excel spreadsheet and do a little plug and chug work and you are there. I can tell you it won't take much to make up 2cc's.

[Johan V]

Well the difference between 3000 cc and 2998 cc is only about 0.067 percent in volume. If the bore and the stroke are more than 0.0222% off from the specs, 2998cc might well become more than 3000 cc.

Lets say the stroke is about 6 cm. The bore should be less than 8.3333333 cm. So they aim for 8.3277777 cm to get a total volume of 2998 cm3. The difference between 8.33333333 cm and 8.3277777777 cm is only about 0.00555555 cm, and thats about 55 microns ie. not a lot.

Johan V.

[marion5drsn]

Johan v. You must be the only one that has done very many actual bore and stroke relationships. As you stated it takes very little on the bore to put the capacity over the mark as it is the square of the radius whereas the relationshion to the stroke is directly linear. M.L. Anderson

[carlos.maza]

All the designers try to be as close as possible to 3000cc, but there are also practical considerations.
How can you measure and machine a 83.33333333mm piston?
Can you realize the kind of machines needed to achieve such precision?

Carlos

[Jezztor]

They're called CNC machines - very widely used in the industrial engineering & mechanical engineering. They are pretty expensive though. My old man has two of them. They work to micron precision. 83.3333333333 should be a breeze ;)

Jezz

[carlos.maza]

Jezztor:

As you say, they are very expensive.
Do F1 teams use these machines to build their engines?
Is the gain justified in $$$ terms?

Thanks
Carlos

[bbocaner]

Originally posted by Jezztor
They're called CNC machines - very widely used in the industrial engineering & mechanical engineering. They are pretty expensive though. My old man has two of them. They work to micron precision. 83.3333333333 should be a breeze ;)

Umm... aren't microns only 6 decimal places?

[david_martin]

actually it is only 3 decimal places if that bore measurement is in millimetres.

[lateralforce]

Originally posted by carlos.maza
Jezztor:

As you say, they are very expensive.
Do F1 teams use these machines to build their engines?
Is the gain justified in $$$ terms?

Thanks
Carlos

His old man has two of em. I think my school has two of em too. There are many kinds and price, depending on what kind of precision and automation level. Everybody uses them nowadays. It pays to be precise.

[lateralforce]

btw, regarding the question, I think the dimension is when the engine is cold ie not under operating condition. In operation the volume could be well expanding to 3000 cc limit. I think.

[MRC]

Ali_G, going from a 2997cc to a 3000cc would yield a 0.1% improvement in torque output, at best. If you have 850 bhp to start with, you have now increased your power by 0.85bhp, at best.

The bore and the stroke (and not the piston), will determine the displacement. You will be more likely to experience a higher nominal tolerance (as Ben explained), with the stroke specification than the bore specification. The stroke has more stack up tolerances, not to mention what changes to the stroke occur when the crankshaft changes temperatures. I am willing to bet that if you wanted to make a 3001cc or so engine that you could get away with it, and that they would not catch it. I don't know how much these guys are measuring things during an inspection teardown, though. Do they measure bore taper? Do they measure every cylinder? What temperature do they measuere the crankshaft at. Do they CMM the engine to inspect it, or do they just use convential measurement instruments? It all comes down to what Ben said in the second post in this thread. Why risk it?

Uh, about the 83.3333333333mm pistons. No CNC turning center is going to hold .000000003mm let alone .000003mm.

A one side tolerance of .00254mm or 2.54 microns (.0001") would be considered very tight for even a CNC turning center, especially at a diamater of 83mm. This is really getting into grinding territory.

A CNC machine is only as good and as accurate as the tools, the setup, the program, and the machine.

Just FYI, the piston is intentionally not made round, and the side of the piston is intentionally not made straight.

[Halfwitt]

Originally posted by MRC
Just FYI, the piston is intentionally not made round, and the side of the piston is intentionally not made straight.

"As they looked under the lid of the can that MRC had left for them, they saw it was full of worms...."

That'll send Ali_G's head into orbit

[desmo]

MRC raises some interesting questions about potential testing methodologies for measurement of bores and strokes! Might an engine that is illegal at room temp, be legal at it's operating temperature? "Screw it, 2998cc it'll be then." Is there a CMM of any sort than could be inserted through the spark plug hole without removing the heads that would be up to the task?

[PAD]

Please excuse my ignorance on this, but using H. Eckener's formula for calculating the displacement, is the stroke measured as the lowest point of the piston during its cycle (i.e. the opposite of TDC - let's call it bottom dead centre BDC) or is it the average of TDC and BDC.

Also in calculating the volume, how do we measure the position of the piston at TDC/BDC? As F1 pistons are not flat across the top (to allow for the valve intrusions and to improve fuel/air mixing), do we measure the highest point of the piston, the lowest point or have to calculate the volume of the piston shape and allow for that.

If the piston shape is important, then maybe that's where the 2 or 3 cc difference is allowed for?

[nicholasc]

I'd like to suggest an easier way to measure the capacity - not exactly, but to ensure it conforms:
take 3000mL and fill each cylinder in turn, making sure any excess from the previous cylinder(s) does not enter the adjacent one and if you've any overflowing at #10 - you fail. You will need to tilt the block appropriately of course.

[Ursus]

Originally posted by PAD
Please excuse my ignorance on this, but using H. Eckener's formula for calculating the displacement, is the stroke measured as the lowest point of the piston during its cycle (i.e. the opposite of TDC - let's call it bottom dead centre BDC) or is it the average of TDC and BDC.

Also in calculating the volume, how do we measure the position of the piston at TDC/BDC? As F1 pistons are not flat across the top (to allow for the valve intrusions and to improve fuel/air mixing), do we measure the highest point of the piston, the lowest point or have to calculate the volume of the piston shape and allow for that.

If the piston shape is important, then maybe that's where the 2 or 3 cc difference is allowed for?

I think you misunderstood the meaning of displacement. Displacement(or swept volume or cylinder/engine capacity) is the volume that is displaced by the piston as it moves from BDC to TDC not the actual volume of the cylinder. Thus the stroke is the distance the piston travels from BDC to TDC.

[MRC]

If you want to measure the stroke, via the piston's movement, you can use a dial indicator setup. You bring the piston to TDC (various methods) and assure yourself that it really is at TDC. Zero your indicator. Then move the piston to BDC (again assuring yourself that you really are at BDC) and then you have your distance (the stroke). Keep in mind that the indicator's plunger must travel parallel to the bore of the piston. Otherwise, you will read a higher stroke than what is really there. (For 2997cc, 96mm bore, 41.405mm stroke, if you were off by 1 degree, you would be off by about 0.5 cc. Although you'd have to be pretty sloppy to be off by one degree).

Then of course, you still don't have the real stroke due to bearing clearances. Of course who cares, beacuse it's so small it's not worth worrying about. (Although this will make the stroke measure smaller than the running stroke).

The other way to measure the stroke is by measuring it directly from the crankshaft. This will give a number less than the real stroke, by a very insignifigant number, due to the amount of wrist pin offset. If no wrist pin offset, than no error (except for the tools and the user). But like I said this change in stroke isn't even worth talking about unless you want to get ridicously anal.

So, I'd say that once again these F1 guys obviously know what's going on, and they don't want to get screwed by some inspector because he thinks that they are .5cc over or some BS like that. I still doubt that anyone gets crazy enough to detect 2 or 3cc over the rules. Has anyone seen any pictures of inspectors taking off cylinder heads to inspect engines?

On using a fluid to measure the volume. This can be quite accurate, given the proper setup. This is probably obvious, but you still will have to measure more than 10 volumes to get the swept volume. You can not just fill the cylinders with fluid and get the swept volume. That will give the total volume, not the swept volume (ie -the displacement) You must know the combustion chamber volumes, too.

Ah, cheating in motorsports. My favorite topic.

[DOHC]

Just back to Atlas after a week away, and I'd like to make a comment on Mr Eckener's formula,

V = pi*((b/2)^2)*s*n

V = displacement
b = bore
s = stroke
n = number of cylinders.

If you want to know what effect on the displacement a small change of bore would have, you don't need to be fancy and use partial derivatives. It's enough to differentiate V with respect to b. Then you get

dV = 2*V*db/b,

where dV = change in displacement, db = change in bore. From this it follows that

dV/V = 2*db/b,

which says that the relative increase in displacement (dV/V) equals twice the relative increase in bore (db/b). This is valid no matter what your bore and stroke are, or your displacement, or your number of cylinders.

In the special case of going from 2997 to 3000 cc, you have dV=3, V=3000, so dV/V = 1/1000 = 0.1%. Then it follows that db/b = 1/2000 = 0.05%.

For a typical bore of 80mm, a bore increase of 80mm/2000 = 0.04mm adds the extra 3 cc.

If you want to change the stroke instead, the formula is

dV/V = ds/s,

where ds = change in stroke. This formula too is independent of displacement, bore, stroke, and number of cylinders. It says that the relative increase in displacement (dV/V) equals the relative increase in stroke (ds/s). In the formula for changing the bore the relative change must be small, let's say below 1%, to retain its accuracy (the formula is approximate). The formula for changing the stroke is exact for any change, however, big or small.