Dyno mystery

[SP250]

Sorry Steve went out just after my post.

We ended up at the Country Spice Indian in Hope.
Pretty good too.

[slparry]

Sorry Steve went out just after my post.

We ended up at the Country Spice Indian in Hope.
Pretty good too.

I know it, not ever been there so good to know it's OK .......

[Corvus]

No converts to the dark side today? I felt sure the xbr graph, with its three sets of curves would have clinched it.

[Herb]

Not sure about converted, but as far as I can tell, we are all arguing from the same side of the argument, as we all agree that the torque curves are representative rather than measured data.

I don't think anyone doubts that the graphs are not wholly accurate. They can only ever be indicative as they are derived from acceleration measurements with a whole lot of assumptions thrown in.

The dyno measures acceleration with assumptions for wheel slip, gearing etc fixed by the operator or the operating system.

Unless you specifically want to know what the acceleration of a known mass is, under power from your motorcycle rear wheel under WOT (and why would you?) then you only have derived data to work with. Good enough for most situations given a good operator and a knowledgeable tuner.

[Corvus]

Not sure about converted, but as far as I can tell, we are all arguing from the same side of the argument, as we all agree that the torque curves are representative rather than measured data.

I don't think anyone doubts that the graphs are not wholly accurate. They can only ever be indicative as they are derived from acceleration measurements with a whole lot of assumptions thrown in.

The dyno measures acceleration with assumptions for wheel slip, gearing etc fixed by the operator or the operating system.

Unless you specifically want to know what the acceleration of a known mass is, under power from your motorcycle rear wheel under WOT (and why would you?) then you only have derived data to work with. Good enough for most situations given a good operator and a knowledgeable tuner.

Hi herb

I'm sorry but you're missing my point completely. I'm not talking about inaccuracies. I wouldn't know about how accurate they are. I'm talking abstract. As in Lewis Carroll. Not actually there at all.

[Corvus]

.............They can only ever be indicative.........

The green torque curve on the xbr graph. What is it indicative of?

[Herb]

Not sure about converted, but as far as I can tell, we are all arguing from the same side of the argument, as we all agree that the torque curves are representative rather than measured data.

I don't think anyone doubts that the graphs are not wholly accurate. They can only ever be indicative as they are derived from acceleration measurements with a whole lot of assumptions thrown in.

The dyno measures acceleration with assumptions for wheel slip, gearing etc fixed by the operator or the operating system.

Unless you specifically want to know what the acceleration of a known mass is, under power from your motorcycle rear wheel under WOT (and why would you?) then you only have derived data to work with. Good enough for most situations given a good operator and a knowledgeable tuner.

Hi herb

I'm sorry but you're missing my point completely. I'm not talking about inaccuracies. I wouldn't know about how accurate they are. I'm talking abstract. As in Lewis Carroll. Not actually there at all.

The green line, is a calculation of engine torque measured at the rear wheel and factored for transmission losses.

Now you think about it, I don't know about abstract, but it is plain wierd. The torque available at the wheel will be higher due to the final gearing.

I concede you are right, although I still don't like the term.

[Corvus]

Not sure about converted, but as far as I can tell, we are all arguing from the same side of the argument, as we all agree that the torque curves are representative rather than measured data.

I don't think anyone doubts that the graphs are not wholly accurate. They can only ever be indicative as they are derived from acceleration measurements with a whole lot of assumptions thrown in.

The dyno measures acceleration with assumptions for wheel slip, gearing etc fixed by the operator or the operating system.

Unless you specifically want to know what the acceleration of a known mass is, under power from your motorcycle rear wheel under WOT (and why would you?) then you only have derived data to work with. Good enough for most situations given a good operator and a knowledgeable tuner.

Hi herb

I'm sorry but you're missing my point completely. I'm not talking about inaccuracies. I wouldn't know about how accurate they are. I'm talking abstract. As in Lewis Carroll. Not actually there at all.

The green line, is a calculation of engine torque measured at the rear wheel and factored for transmission losses.

Now you think about it, I don't know about abstract, but it is plain wierd. The torque available at the wheel will be higher due to the final gearing.

I concede you are right, although I still don't like the term.

Result!

Why don't you like "abstract"? I chose the word carefully.

I agree it is better to think of it as a version of crankshaft torque than rear wheel torque. But of course, it's not really that either. As I've said all along, it doesn't exist. Bear in mind I put the xbr graph forward because it is untypical in showing three sets of curves. Most times we only see one set.

You mentioned a constant (I think you amended your post as I was writing?). Yes, the constant is the power value. But then that's not constant in the literal sense of the word.

The next thing now is to think about what it means to have the losses expressed that way in torque terms. Is it right to say we loose xx ftlb from front to rear? The torque gets multiplied each time a loss is incurred. Surely the power value is the only way to express this?

It's not just the final gearing by the way. It's the gearing at every stage in the transmission.

Fascinating stuff.

[Herb]

I amended my post after some further digging. The constant I was referencing was the factor for transmission loss, which is often not measured, they just allow a fixed percentage value.

This is wrong for a number of reasons, not least is that it is typically applied across the whole rev range and takes no account of increasing friction with RPM. It is also a bit daft, because the apparent transmission loss increases on the same engine as work is done, which is clearly not correct.

ie, if you test an engine at 100hp with 10% transmission loss the driveline loss is 10hp.

If you then work on the engine and increase the power to 110hp, the transmission loss has now increased to 11hp, despite the transmission being unchanged.

I used to deal with this kind of stuff for a living, but changed industries 5 years ago. I am struggling to get my head back around things.

[slparry]

I amended my post after some further digging. The constant I was referencing was the factor for transmission loss, which is often not measured, they just allow a fixed percentage value.

This is wrong for a number of reasons, not least is that it is typically applied across the whole rev range and takes no account of increasing friction with RPM. It is also a bit daft, because the apparent transmission loss increases on the same engine as work is done, which is clearly not correct.

ie, if you test an engine at 100hp with 10% transmission loss the driveline loss is 10hp.

If you then work on the engine and increase the power to 110hp, the transmission loss has now increased to 11hp, despite the transmission being unchanged.

I used to deal with this kind of stuff for a living, but changed industries 5 years ago. I am struggling to get my head back around things.

so a "relative constant" rather than a "fixed constant"?

[Corvus]

I amended my post after some further digging. The constant I was referencing was the factor for transmission loss, which is often not measured, they just allow a fixed percentage value.

This is wrong for a number of reasons, not least is that it is typically applied across the whole rev range and takes no account of increasing friction with RPM. It is also a bit daft, because the apparent transmission loss increases on the same engine as work is done, which is clearly not correct.

ie, if you test an engine at 100hp with 10% transmission loss the driveline loss is 10hp.

If you then work on the engine and increase the power to 110hp, the transmission loss has now increased to 11hp, despite the transmission being unchanged.

I used to deal with this kind of stuff for a living, but changed industries 5 years ago. I am struggling to get my head back around things.

So the losses are still
10%. That's how percentages work.

[Corvus]

I’m having a hard time understanding the question too.

But, if I’d be a rolling dyno owner offering people to dyno their bikes, how would I like to present the results?
Obviously, the most honest would be torque at the rear wheel. ‘cause that’s what the bike is putting on the tarmac.
Plotted to what? For the customer, I’d think engine rpm would be the best. It tells him something, as opposed to plot it to rear wheel rpm. What else relevant can you plot it at?

You can plot the torque against whatever speed you like, but then be prepared to accept that what you are doing is abstract. Abstract as in "not actual".

It seems to me that a rolling road is a highly convenient way of measuring (a bikes') output power. Rear wheel output power. It is also, to the same degree, a convenient way of measuring the rear wheel torque. Except we are not being shown the rear wheel torque! We are not being shown engine torque either. The moment you add a correction to 1:1 you get an abstract picture of events. Remember, we, the human element in the equation, decide to tell the computer the gear ratio is 1:1. So what we get is precisely that. A picture of our bike as if it had an overall gear ratio(s) of 1:1. In that abstract scenario then the torque figure aligns (not surprisingly!).

Why can't anyone else see this?

[Corvus]

Happy new year y'all.

[Corvus]

With respect to coastdown data (if that is the correct term) to account for transmission losses, it strikes me that the conventional boxer layout lends itself to more accuracy.

What I mean is (correct me, please) when pulling in the clutch to allow the inertia of the roller to determine the power required to drive the transmission and thereby quantify the transmission losses, in a typical UJM layout there will still be small doubts over the losses involved in the primary ratio and also whether the oil in between the clutch plates is absorbing a small amount of power and spoiling the accuracy.

In a boxer layout, once the clutch is pulled in, all the transmission ratios are accounted for as they all lay downstream of the clutch. Plus, being a dry clutch, there is no chance of the oil/multiple plate scenario.

Is that right?

The new water cooled engine is different, of course.

[Corvus]

I apologise for prattling on about this but I am very intrigued by this contrived torque value. I've been involved in power transmission applications in the past and everything was black and white. Give us the torque and rpm and we can get the power. But we dealt in actuals! None of this corrected mullarky!

It's doin me freakin head in!

Let me try and put it another way. A torque value must apply to a given shaft. It can't exist somewhere in between. In the gearbox aether!

We pretty much have four shafts to choose from (in a UJM) to calculate what torque value applies there. The crank, the gearbox in (same figure in our boxer), the gearbox out or the axle (rear wheel). That's it. So which one does our corrected torque belong on?

I can't put it any simpler.