Dyno mystery

[Corvus]

I'll have a go but probably need to understand your question better.

Rolling road dyno measures the torgue at the rear wheel (obviously)......



.....Not sure why you say that the figures are not rear wheel torque or isn't crank torque either.

Measures which torque at the rear wheel? This is my point.

I can easily prove it isn't rear wheel torque. Trying to define which torque it is, is tricky.



Edit: with hindsight I don't think my answer does justice to your reply. I apologise. I'll try again.

[Corvus]

They are getting confused with torque curves and tractive effort graphs.

Tractive effort = Force at rear wheel using gearbox to multiply torque.

'Torque curve' = 'Fudged' engine torque to remove gearbox effects.

Bike used to publish the TE graphs in reviews.

Woo hoo! Someone understands me!

I owe you a pint. I think superbike did for while too. I didn't know that was the term for the graphs, but I fully understand what you mean.

You and I are on the same wavelength torque wise. I have used the word "fudged" many times trying to get this point across.

It is fudged, plain and simple. It is useful to have, but you need to know what you are doing in having it.

Give me your definition of it (and don't nick mine!)

[conkerman]

As I put in my earlier reply a proper set of coastdowns will allow the dyno to be set with the transmission frictional losses over a range of speeds making this correction more accurate.
This takes time and is probably not worth the aggro for getting power and torque curves. So 'cookbook' figures are used which are a set of (generally OK) coefficients using vehicle weight.

Power/torque runs are generally done in 1 gear (3rd) so you can pich a set of numbers for a 'standard' 3rd gear and apply. TBH these are usually OK and are at least consistent, to the point where I would not be overly bothered.

Smile knowingly and say that's great in 90% of cases. Many more important things to be arguing about, like oil.

/Runs!

[Corvus]

They are getting confused with torque curves and tractive effort graphs.

Tractive effort = Force at rear wheel using gearbox to multiply torque.

'Torque curve' = 'Fudged' engine torque to remove gearbox effects.

Bike used to publish the TE graphs in reviews.

Woo hoo! Someone understands me!

I owe you a pint. I think superbike did for while too. I didn't know that was the term for the graphs, but I fully understand what you mean.

You and I are on the same wavelength torque wise. I have used the word "fudged" many times trying to get this point across.

It is fudged, plain and simple. It is useful to have, but you need to know what you are doing in having it.

Give me your definition of it (and don't nick mine!)

Done some quick searching. Tractive effort seems to be expressed as a linear force eg: N. And also road speed, mph etc.

That's fine as, ultimately, that's what having our rear wheel torque is for. My focus was the stage of the process immediately before tractive effort. ie: rear wheel torque.

But still, we seem to understand each other.

[Corvus]

As I put in my earlier reply a proper set of coastdowns will allow the dyno to be set with the transmission frictional losses over a range of speeds making this correction more accurate.
This takes time and is probably not worth the aggro for getting power and torque curves. So 'cookbook' figures are used which are a set of (generally OK) coefficients using vehicle weight.

Power/torque runs are generally done in 1 gear (3rd) so you can pich a set of numbers for a 'standard' 3rd gear and apply. TBH these are usually OK and are at least consistent, to the point where I would not be overly bothered.

Smile knowingly and say that's great in 90% of cases. Many more important things to be arguing about, like oil.

/Runs!

IMO dyno runs to compare motorcycles are misunderstood, misquoted, misused and ultimately overrated. Crankshaft figures are the best way of understanding the engine. We are given them by the manufacturer. Manufacturers seem prone to bending the truth somewhat so we are forced into testing for ourselves to check.

The only feasible place to make the check is at the rear wheel. Fine. But since the coastdown run should provide the losses it seems to me that, in respect to comparing bikes, we should convert or express our findings in true crankshaft figures. That would send a strong signal out that we are onto them. It would be the fairest way to compare one bikes' engine with another. Trying to express things at the rear wheel would seem to only confuse the issue. And as long as there's confusion, the manufacturers will keep getting away with it.

To demonstrate in figures how one complete motorcycle might perform against another, isn't tractive effort the best way?

With respect to engine tuners, surely they too are more concerned with what comes out at the crankshaft? Although that person would also want as much additional information as is available too, sure.

[conkerman]

More importantly, dyno curves are marketing, and are perfect for willy waving. A TE chart is a bit more oblique and can be changed by altering the gearing of the bike, whether sprocket size/gear ratio.

[Corvus]

More importantly, dyno curves are marketing, and are perfect for willy waving. A TE chart is a bit more oblique and can be changed by altering the gearing of the bike, whether sprocket size/gear ratio.

Agreed. All the more reason for pushing forward the suggestion that crankshaft figures probably better serve the willy waving punter? Or you could say that the TE graph fits the factory supplied machine and it's up to the person changing the ratios to understand the implications with respect to TE.

No one has yet ventured to discuss the mysterious torque figure shown on a lot of (most?) dyno sheets. The slippery one that defies easy definition.

[Corvus]

I'll have a go but probably need to understand your question better.

Rolling road dyno measures the torgue at the rear wheel (obviously).
Calculations are then made for the losses in the drive train, gearbox and the tyre slippage at the contact patch. Also for the reduction in radius of the wheel and tyre where the contact patch is strapped down onto the rollers.
They can either measure direct at the rear wheel or do all the calcs to approximate the results at the crankshaft - which is then roughly equal to figures obtained on an engine dyno. Don't forget any gear ratio used other that 1:1 alters revs and torque figures.

Not sure why you say that the figures are not rear wheel torque or isn't crank torque either.

Would I be right in saying that inertial rolling road dyno's measure the power absorbed to accelerate the known physical properties of its own drum? The computer has all the data needed to easily translate that into a torque curve based on the absorbed power it has just measured AT THE SPEED it has just measured. But, in most cases I can find, it doesn't do that. Instead it seems to take the absorbed rear wheel power but produce a torque curve based on engine speed. There are probably a few reasons why it does this, but I'll ask the question. Why?

The resulting torque figure is abstract. It never actually existed.

True?

[conkerman]

Not sure I understand you here. You can't have a power curve without generating a torque curve of sorts.

Torque = work

Power = work / time.

The at the wheels torque figure is correected and the power curve follows.

[Corvus]

Not sure I understand you here. You can't have a power curve without generating a torque curve of sorts.

Torque = work

Power = work / time.

The at the wheels torque figure is correected and the power curve follows.

Ok. Most dyno curves seem to produce a rear wheel power curve.

This is engine power minus losses. So it's not crankshaft power.

True?

[conkerman]

Agreed, but putting an engine on the dyno is not straightforward. And most motorcycle transmissions are broadly similar.

What we have now is a reasonable approximation. That's enough for pretty much anyone not doing homologation work or similar. In a lot of cases the absolute number may be of secondary importance.

I am out of my limited knowledge now. A dyno will give an estimated crank figure based on the torque applied to the roller by the wheel and the coastdown data in the dyno.

[Corvus]

Agreed, but putting an engine on the dyno is not straightforward. And most motorcycle transmissions are broadly similar.

What we have now is a reasonable approximation. That's enough for pretty much anyone not doing homologation work or similar. In a lot of cases the absolute number may be of secondary importance.

I am out of my limited knowledge now. A dyno will give an estimated crank figure based on the torque applied to the roller by the wheel and the coastdown data in the dyno.

My point is this. You agree the power measured is REAR WHEEL POWER.

Ok.

But the curve presented is at CRANKSHAFT speed.

That's where the abstraction comes in. (I believe)

Whatever the dyno "feels" at the drum in terms of torque, it is not (usually) presented on the curves. The power is right, but the torque isn't. You prove to me anywhere in the powertrain, from crankshaft to rear wheel where the torque expressed on the (usual) curves actually occurred. Prove it using the well known hp formula.

We seem to be given CRANKSHAFT torque, but with all the subsequent losses taken off. Of course that never actually happens because by the the time the torque finds its way to the back wheel and all the losses have ACTUALLY occurred, it has been multiplied many times over by the transmission ratios.

Want to know rear wheel torque? You'll need rear wheel power and REAR WHEEL SPEED.

Want to know crankshaft torque? You'll need CRANKSHAFT POWER and crankshaft speed.

[el-nicko]

.

Is anybody having any trouble understanding my posts?

[conkerman]

The dyno measures rear wheel torque.

To massively oversimplify.

Assume that transmission loss is 10% over the whole rev range in a single gear.

The bike does a run.

The Dyno also knows.

Engine RPM.


Once the run is complete the dyno correlates the wheel speed Vs engine speed and torque output and applies a this 10% correction, and adds 10% through the rev range, the power curve is then calculated.

In reality the coefficients are more complex and will include stuff like baro pressure. They can probably be described as a series of best guesses.

One thing is clear. (I hope). Dyno numbers can be manipulated and are a minefield.

Will google and see if someone can explain better than me.

[Corvus]

The dyno measures rear wheel torque.

To massively oversimplify.

Assume that transmission loss is 10% over the whole rev range in a single gear.

The bike does a run.

The Dyno also knows.

Engine RPM.


Once the run is complete the dyno correlates the wheel speed Vs engine speed and torque output and applies a this 10% correction, and adds 10% through the rev range, the power curve is then calculated.

In reality the coefficients are more complex and will include stuff like baro pressure. They can probably be described as a series of best guesses.

One thing is clear. (I hope). Dyno numbers can be manipulated and are a minefield.

Will google and see if someone can explain better than me.

Thanks. Maybe I'm simply misunderstanding how things are done. An example of a dyno sheet will help. One that shows rear wheel power.

We agree again. It's a minefield!

Your sentence : "Once the run is complete the dyno correlates the wheel speed Vs engine speed and torque output and applies a this 10% correction, and adds 10% through the rev range, the power curve is then calculated."

I don't follow the bit on the speeds and torque. That sounds suspiciously like the fudging of figures we talked about earlier.

Are we agreeing that the power curve is rear wheel power or not? Forget the torque for a moment, can we agree or disagree on that question?