Announcement

**B7ACKTHORN** · 03-29-2018, 05:21 PM

Re: Torque curve prediction from Stroke Length

Thread Approved

Crisp one! Keep it coming.

Cheers!
VJ

**Samarth 619** · 03-30-2018, 01:36 PM

Re: Torque curve prediction from Stroke Length

Nice thread. I'd like to add some points, for some of the observations:-

1. Longer stroke results in lesser revs, because piston travel speeds must not exceed a certain threshold when the vehicle redlines. Its about 16-22 metres/ second for streetbikes. Beyond 22-24m/ sec in a streetbike, we're looking at trouble.
So, short stroke bikes can rev higher safely, because for a particular rpm, their piston travels lesser.

You can easily calculate piston speeds, as piston speed = stroke value in mm x 2 x rpms/60.

At 60mm of stroke value, Duke 390's piston speed at redline is, 60mm x 2 x 10,000/60 = 20,000mm/ second = 20 metres/ second... Near the danger limit.

Hence, we conclude that as long as its a streetbike, we won't be seeing higher than 10,000 rpms, unless we shorten its stroke value.

2. On high engine capacities (like 1000cc or so), We don't have a choice for single cylinder, we have to go multicylinders due to piston speed...
For example, if you want to create a 1000cc single cylinder engine, it will have a huge piston, a huge bore & stroke valve, as a result it will rev only upto 3000 rpm or so. Not too exciting, right?

That's why we go multicylinder.
KTM is one of the rare manufacturers who make high capacity singles, like Duke 690cc. As expected, it is powerful for a single, but hardly revs too high.

3. Mostly people associate acceleration with higher torque. This is wrong. Torque is how forcefully the engine will continue to push the bike, despite tailwinds, headwinds, combined weight of the vehicle, higher gear, etc...
Acceleration is related more with max rpm and power output, plus a variety of other factors like aerodynamics, gearing, weight, etc.

4. When you use multiylinders, you're allowing a lower bore stroke values for each cylinder, this helps them rev higher. Smaller engines can rev higher due to mean piston speed mentioned above...
So, multicylinders have more freedom of revs, alongwith bottom end torque due to constant firing and overall engine capacity.
Also, even in multiylinders, less capacity bikes rev higher safely.
For example,
a 250cc 4 cylinder streetbike from 90's was known to rev upto 19000 rpm, because each cylinder is just 62.5cc.

a 600cc 4 cylinder streetbike does about 16000 rpm, due to bigger stroke value than the above bike, and each cylinder displaces about 150cc. (150x4=600)
a 1000cc 4 cylinder streetbike does about 13000 rpm,
a 1340cc 4 cylinder streetbike does about 10000 rpm,

5. The bore and stroke affect a bike's power-torque to maybe about 70%. Rest is determined by how the manufacturer tunes the vehicle.
There's a lot to be tuned in terms of fuelling, exhaust, valve actuation timing, spark timing/ types, etc.
Valve actuation timing is a neat trick for example. Different manufacturers use different names for it. Honda uses VVTi I think in cars.

So, what they do this, they electronically manage the valve's open & close timing, and change it in realtime depending on the engine rpm. This happens electro-mechanically, and this means you can keep a certain valve timing for low end torque and when you get to higher rpm, the timing is changed to favour high rpm's torque.
Like, a best of both worlds.
The latest example would be Yamaha R15 version 3.0. The earlier versions had quite weak low end torque, but this time, I've learned that around 4500 rpm, valve timing gets changed, so that you get good torque no matter which rpm it is.

6. Like I said, tuning matters a lot. Taking an example of 2 old models of similar year, Yamaha's YZF R1 vs FZ1, their bore strokes were the same. In fact, FZ1 had 5 valves per cylinder, instead of the normal 4 valves per cylinder..

But the FZ1 was just tuned weakly, due to its market focus, lower market price and maybe, cost of parts. The end result was that at high, middle, and even low rpm, the R1 was pushing out more torque.
Yes, even in low rpm R1 was better, where you'd expect FZ1 to perform better, given its street nature.

7. Generally speaking, the later bikes may have lesser potential due to need for compliance with stricter emission norms like Euro4, BS4, etc.
And they get heavier too, but that's a different thing as we're only talking about engine output.

8. Short stroke results in bigger bore, for the same cc level. This bigger bore allows bigger valves, which means making a rev happy engine where high flow pressure is obtained only at higher revs, unlike smaller valves.

The conclusion can be that a certain performance boundary can be drawn with bore stroke values, but the prediction can't be that accurate on these 2 values alone.
See, Duke 390 vs Dominar. Same bore stroke but huge difference in laid down power. Mostly due to SOHC in Dominar and the other modern world tuning at work.

**Sahil Goel** · 04-02-2018, 07:23 PM

Re: Torque curve prediction from Stroke Length

Originally posted by Samarth 619 View Post

Nice thread. I'd like to add some points, for some of the observations:-

1. Longer stroke results in lesser revs, because piston travel speeds must not exceed a certain threshold when the vehicle redlines. Its about 16-22 metres/ second for streetbikes. Beyond 22-24m/ sec in a streetbike, we're looking at trouble.
So, short stroke bikes can rev higher safely, because for a particular rpm, their piston travels lesser.

You can easily calculate piston speeds, as piston speed = stroke value in mm x 2 x rpms/60.

At 60mm of stroke value, Duke 390's piston speed at redline is, 60mm x 2 x 10,000/60 = 20,000mm/ second = 20 metres/ second... Near the danger limit.

Hence, we conclude that as long as its a streetbike, we won't be seeing higher than 10,000 rpms, unless we shorten its stroke value.

2. On high engine capacities (like 1000cc or so), We don't have a choice for single cylinder, we have to go multicylinders due to piston speed...
For example, if you want to create a 1000cc single cylinder engine, it will have a huge piston, a huge bore & stroke valve, as a result it will rev only upto 3000 rpm or so. Not too exciting, right?

That's why we go multicylinder.
KTM is one of the rare manufacturers who make high capacity singles, like Duke 690cc. As expected, it is powerful for a single, but hardly revs too high.

3. Mostly people associate acceleration with higher torque. This is wrong. Torque is how forcefully the engine will continue to push the bike, despite tailwinds, headwinds, combined weight of the vehicle, higher gear, etc...
Acceleration is related more with max rpm and power output, plus a variety of other factors like aerodynamics, gearing, weight, etc.

4. When you use multiylinders, you're allowing a lower bore stroke values for each cylinder, this helps them rev higher. Smaller engines can rev higher due to mean piston speed mentioned above...
So, multicylinders have more freedom of revs, alongwith bottom end torque due to constant firing and overall engine capacity.
Also, even in multiylinders, less capacity bikes rev higher safely.
For example,
a 250cc 4 cylinder streetbike from 90's was known to rev upto 19000 rpm, because each cylinder is just 62.5cc.

a 600cc 4 cylinder streetbike does about 16000 rpm, due to bigger stroke value than the above bike, and each cylinder displaces about 150cc. (150x4=600)
a 1000cc 4 cylinder streetbike does about 13000 rpm,
a 1340cc 4 cylinder streetbike does about 10000 rpm,

5. The bore and stroke affect a bike's power-torque to maybe about 70%. Rest is determined by how the manufacturer tunes the vehicle.
There's a lot to be tuned in terms of fuelling, exhaust, valve actuation timing, spark timing/ types, etc.
Valve actuation timing is a neat trick for example. Different manufacturers use different names for it. Honda uses VVTi I think in cars.

So, what they do this, they electronically manage the valve's open & close timing, and change it in realtime depending on the engine rpm. This happens electro-mechanically, and this means you can keep a certain valve timing for low end torque and when you get to higher rpm, the timing is changed to favour high rpm's torque.
Like, a best of both worlds.
The latest example would be Yamaha R15 version 3.0. The earlier versions had quite weak low end torque, but this time, I've learned that around 4500 rpm, valve timing gets changed, so that you get good torque no matter which rpm it is.

6. Like I said, tuning matters a lot. Taking an example of 2 old models of similar year, Yamaha's YZF R1 vs FZ1, their bore strokes were the same. In fact, FZ1 had 5 valves per cylinder, instead of the normal 4 valves per cylinder..

But the FZ1 was just tuned weakly, due to its market focus, lower market price and maybe, cost of parts. The end result was that at high, middle, and even low rpm, the R1 was pushing out more torque.
Yes, even in low rpm R1 was better, where you'd expect FZ1 to perform better, given its street nature.

7. Generally speaking, the later bikes may have lesser potential due to need for compliance with stricter emission norms like Euro4, BS4, etc.
And they get heavier too, but that's a different thing as we're only talking about engine output.

8. Short stroke results in bigger bore, for the same cc level. This bigger bore allows bigger valves, which means making a rev happy engine where high flow pressure is obtained only at higher revs, unlike smaller valves.

The conclusion can be that a certain performance boundary can be drawn with bore stroke values, but the prediction can't be that accurate on these 2 values alone.
See, Duke 390 vs Dominar. Same bore stroke but huge difference in laid down power. Mostly due to SOHC in Dominar and the other modern world tuning at work.

Excellent points [MENTION=24386]Samarth[/MENTION] !!

1) The peak piston speeds can be increased only upto a particular limit as you mentioned. Above those speeds the kind of materials used also come into play. Best example would be again D390 and D400.. revv limiters are at 10k and 8.5k respectively. Now duke is using much much better materials viz. pistons, valves, cylinder coating, head, crank shaft etc. Much lighter but still stronger than D400 to decrease the weight of reciprocating mass and withstand the forces generated at those piston speeds. Dominar ofcouse does not have to revv to that limit considering the application and cost factor.

2) Well I'd like to differ here though not too much. Torque is still a governer of acceleration majorly, along with the other things you mentioned. But the thing is, if you look at torque curves, the moment peak torque comes in, the torque figure begins to decline rapidly thats why no acceleration. If the engine could still produce "that" peak torque on every rpm or on most rpms the bike would still accelerate no matter what . If there is no torque or less torque, no matter what the gearing, aerodynamics, weight it will not accelerate. Now since no bike ever produced a straight line curve(unless turbo charged bikes become commercial and achieve pure flat, flat and long lasting curves like diesel cars), gearing becomes one of the ways to utilise best, the torque band which is being produced. A more accurate way to determine the acceleration would be "average torque" produced by an engine over its full revv band. That ideally means bikes with very wide revv band like upto 18000rpm automatically will have higher "average torque" figure as compared to a same cc bike with a revv band of upto 6000rpm unless the toruq is very high..eg ducati monster with a redline at 10k accelerates faster than cbr650f with a redline at 14k. The honda might produce more bhp due to wider revv band but more torque on monster makes it faster.

**Samarth 619** · 04-04-2018, 05:32 PM

Re: Torque curve prediction from Stroke Length

Originally posted by Sahil Goel View Post

2) Well I'd like to differ here though not too much. Torque is still a governer of acceleration majorly, along with the other things you mentioned. But the thing is, if you look at torque curves, the moment peak torque comes in, the torque figure begins to decline rapidly thats why no acceleration. If the engine could still produce "that" peak torque on every rpm or on most rpms the bike would still accelerate no matter what . If there is no torque or less torque, no matter what the gearing, aerodynamics, weight it will not accelerate.

Yes, correct. In a particular gear, good torque helps us accelerate (if there are sufficient rpms' left ahead). If there's not sufficient torque, you downshift so that revs rise up and you get enough torque to lurge forward.

But, I would like to add that, fundamentally torque (FORCE applied) has no relation to movement or rotation. If I'm pushing a wall, there's torque, but that's not translating into movement.
At top speed of a vehicle, there is definitely some amount of torque from the engine that's involved in pushing the vehicle forward, but still the vehicle won't gain any more speed. For speed, the rpm's must move forward further...

Since power = torque x rpm, the power figure accounts for both revs and torque. This is why power and torque and related by, how high the engine revs.

Thus, its the power output that helps determine both acceleration and top speed of a vehicle.

Now since no bike ever produced a straight line curve(unless turbo charged bikes become commercial and achieve pure flat, flat and long lasting curves like diesel cars), gearing becomes one of the ways to utilise best, the torque band which is being produced.

Yes, although an electric motorcycle mostly has a straight torque curve and it has its maximum torque right from lowest rpm, 0. Only at around 70-80% of its rev range, it starts losing torque.

Thus, an electric is usually very torquey, but not that revvy, hence not that powerful.
The highest rpm I've known an electric motor touch is 10,000 rpm which is a liquid cooled motor abroad (Lighting LS218), but mostly they touch 4000-5000 rpm or so.

A more accurate way to determine the acceleration would be "average torque" produced by an engine over its full revv band. That ideally means bikes with very wide revv band like upto 18000rpm automatically will have higher "average torque" figure as compared to a same cc bike with a revv band of upto 6000rpm unless the toruq is very high..eg ducati monster with a redline at 10k accelerates faster than cbr650f with a redline at 14k. The honda might produce more bhp due to wider revv band but more torque on monster makes it faster.

That wouldn't be a very accurate statement...

The reason is, we refer to a vehicle's acceleration as its best performance under a drag race like run..

So, if I say that Duke 390's acceleration time for 0-100 kmph is 5.6 seconds, it is ONLY when I upshift above maximum power's rpm's, i.e. 9000 rpm.
Once you shift up at 9000 rpm, you'll reach about 7000 rpm in 2nd gear, from which you climb back to 9000 rpm and shift another gear and so on till the end of the test/drag.
You must have noticed that how we're constantly in the 7000-9000 rpm range, right from the 2nd to the 6th gear, right?

This is why top end power figure matters in both acceleration and top end. Duke's performance figures are great because it makes both good torque and power from 7000 to 9000 rpm.
A stock Royal Enfield won't be able to match Duke in acceleration, even though the 500cc Royal Enfield has obviously better torque.

Of course, we the commuters ride at varied rpm's, sometimes low and sometimes mid-high, so for us, average torque might be a relevant concept...

**sayanchakraborty89** · 11-08-2018, 07:33 PM

Re: Torque curve prediction from Stroke Length

We can not do this from only the Stroke length, Torque is a function of many other factors, and the curve, as i think, can be altered by using a different gearing setup.

Electric motors have linear torque carve, that is why they do no require gears, their torque is just a function of rpm.

And it is absolutely wrong that electric motors are not high revving, they are extremely high revving, they just do not need so much revs like petrol engines.

hope this helps: https://youtu.be/36H9BVeMYMI

Thanks.

Announcement

Torque curve prediction from Stroke Length

Torque curve prediction from Stroke Length

Comment

Comment

Comment

Comment

Comment