Looking for a decent triple crankset

[Albert 50]

Avoiding cross-chaining costs you nothing. It's a simple habit thing and one that can be easily adapted. There's really no value in defending cross-chaining as a habit.

The OP would do better to repeat over & over in his head, as he rides, sogoods above comment. The silly notion of trying to avoid front changes, at all costs, until you run out of rear cogs is daft IMO Using both/ all three up front efficiently will also 'even' out the drive train wear.

 

[artemidorus]

I don't have any scientifically measured data to back that up. But the logic makes perfect sense and is consistent with all engineering principles.

I agree with most of your points, but exactly how many watts go into my 39/12 combo at cruising speed (before you start, I don't actually use it!)? Misinformation spreads like American neologisms in web forums - we should do our best to combat it. Loss of efficiency with cross-chaining is an unproven assertion - it may well go the way of the rim rotational inertia sensationalism.

 

[sogood]

I agree with most of your points, but exactly how many watts go into my 39/12 combo at cruising speed (before you start, I don't actually use it!)? Misinformation spreads like American neologisms in web forums - we should do our best to combat it. Loss of efficiency with cross-chaining is an unproven assertion - it may well go the way of the rim rotational inertia sensationalism.

Ok, a very simple percentage ball park figure can be easily calculated by drawing a force vector diagram on the chain line geometry.

The loss of efficiency equates with the magnitude of the horizontal vector (force vector in line with the rear axle axis and directed towards the mid point of the axle) in the simple force vector triangle. And the greater the cross chaining ie. Greater angle from a straight chain line, the greater the horizontal vector.

No, this is not misinformation. This is simple physics and engineering that you learn even at high school level.

 

[gclark8]

Ok, a very simple percentage ball park figure can be easily calculated by drawing a force vector diagram on the chain line geometry.

The loss of efficiency equates with the magnitude of the horizontal vector (force vector in line with the rear axle axis and directed towards the mid point of the axle) in the simple force vector triangle. And the greater the cross chaining ie. Greater angle from a straight chain line, the greater the horizontal vector.

No, this is not misinformation. This is simple physics and engineering that you learn even at high school level.

EXACTLY!

This is why BB width is so important, chain centreline. See the degree of precision given to chainline in track (single speed ) bikes.

 

[artemidorus]

Ok, a very simple percentage ball park figure can be easily calculated by drawing a force vector diagram on the chain line geometry.

The loss of efficiency equates with the magnitude of the horizontal vector (force vector in line with the rear axle axis and directed towards the mid point of the axle) in the simple force vector triangle. And the greater the cross chaining ie. Greater angle from a straight chain line, the greater the horizontal vector.

No, this is not misinformation. This is simple physics and engineering that you learn even at high school level.

OK, let's keep going with the simple physics. You are correct in saying that the greater the chainline deviation, the greater the chain tension for a standardised torque on the crankset. But the vector that pulls directly forward on the cassette sprocket does not change with angle of cross-chaining (because the angular deviation is identical at each end of the working chain segment), only the sideways vector changes. The sideways vector does NO WORK because work=force times DISTANCE and the sprocket cannot move sideways.
So your vector argument is erroneous. You can only prove that more work is done if you can prove that friction is increased - indeed it may be the case that hub friction is increased by the sideways vector on the sprocket, or in the bearings of the angulated chain segments. But I would submit that this friction increase is likely to be a pretty damn small wattage.

 

[sogood]

OK, let's keep going with the simple physics. You are correct in saying that the greater the chainline deviation, the greater the chain tension for a standardised torque on the crankset. But the vector that pulls directly forward on the chainring sprocket does not change with angle of cross-chaining (because the angular deviation is identical at each end of the working chain segment), only the sideways vector changes. The sideways vector does NO WORK because work=force times DISTANCE and the sprocket cannot move sideways.
So your vector argument is erroneous. You can only prove that more work is done if you can prove that friction is increased - indeed it may be the case that hub friction is increased by the sideways vector on the sprocket, or in the bearings of the angulated chain segments. But I would submit that this friction increase is likely to be a pretty damn small wattage.

Here you go for some proper data.

http://scitation.aip.org/getabs/ser...00123000004000598000001&idtype=cvips&gifs=yes

 

[artemidorus]

Here you go for some proper data.

http://scitation.aip.org/getabs/ser...00123000004000598000001&idtype=cvips&gifs=yes

I can't get in to the article. The abstract is a little nonsensical - if inefficiency is not causing heating, where is that lost energy going? Sound energy? - that can't be too many watts.

 

[sogood]

Energy loss to sound, heat, material wear and others. And bear in mind that study wasn't primarily looking at cross-chaining effect.

Not too many watts? Well, for some, that's watts too much already.

 

[artemidorus]

Energy loss to sound, heat, material wear and others. And bear in mind that study wasn't primarily looking at cross-chaining effect.

Not too many watts? Well, for some, that's watts too much already.

Energy causing material wear is going to end up as heat, isn't it? Others?
Would 0.1W at 50km/h be enough to concern you? If so, fair enough. It wouldn't bother me.

 

[sogood]

Energy causing material wear is going to end up as heat, isn't it? Others?
Would 0.1W at 50km/h be enough to concern you? If so, fair enough. It wouldn't bother me.

Where did you get that 0.1W from? For material wear, a key issue in this thread, you don't need a lot of energy to effect it over time and repetition. And if your 0.1W loss is correct for a cross chained system, then a straight chain system may only have a 0.01W power loss. As such, that's a 10 fold increase in the energy loss which would directly contribute to wear. And even if it's down to 0.05W, that's still a 2 fold increase in the friction loss and a potential doubling of the wear rate on component. Would you still be happy with that? I know I won't as I don't want to unnecessarily half the service life of my drive train and make early contribution to Shimano or Campag.

As for the energy loss due to cross chaining, that remains a physical fact. And the fact that one can hear the crunches of a cross chained drive mechanism over and above road noise, that's more than enough evidence of significant energy loss.

 

[artemidorus]

Where did you get that 0.1W from? For material wear, a key issue in this thread, you don't need a lot of energy to effect it over time and repetition. And if your 0.1W loss is correct for a cross chained system, then a straight chain system may only have a 0.01W power loss. As such, that's a 10 fold increase in the energy loss which would directly contribute to wear. And even if it's down to 0.05W, that's still a 2 fold increase in the friction loss and a potential doubling of the wear rate on component. Would you still be happy with that? I know I won't as I don't want to unnecessarily half the service life of my drive train and make early contribution to Shimano or Campag.

The 0.1W is an arbitrary figure - I was simply stating the kind of figure that wouldn't bother me in the slightest.
As I've already said, I think that the wear concerns are valid - I just don't wish to contribute to the inefficiency mythology. I'm always conscious of cross-chaining while riding.

 

[artemidorus]

As for the energy loss due to cross chaining, that remains a physical fact. And the fact that one can hear the crunches of a cross chained drive mechanism over and above road noise, that's more than enough evidence of significant energy loss.

Significant energy loss is unproven, at least in terms of data we can get our hands on - sound power can be very low.
Perhaps we can get JTE83 to do a study?

 

[j.r.hawkins]

I have a triple on my MTB, and the extra noise at 22T inner chainring / 11T outer is so loud it can be heard over any amount of wind noise. Can't be good for either the chain or sprockets/chainrings.

Also, I've found that if I have the cogs at any further out than 4, the chain is many times more likely to miss the inner chainring and wrap itself around the bottom bracket on the downshift under even mild loads. I suspect the reduced tension from the rear derailleur is partly to blame, and I've had a chain tear apart from the damage.

Just some things to bear in mind.

 

[dhk2]

Significant energy loss is unproven, at least in terms of data we can get our hands on - sound power can be very low.
Perhaps we can get JTE83 to do a study?

Don't know about friction losses from the bending of the chain, but I'm pretty sure that the only component of chain tension doing work on the cogs is the perpendicular one. It really doesn't take a study to see where the vector loss comes from.

From my 9 sp triple, would estimate the chain angle in 30/12 (worst-case) is 5.3 degrees. The perpendicular component of 5.3*(cosine) is 0.9957, meaning .43% of power is lost due to the chain angle. At 300 W output on a hill climb, it's 1.3 W lost.

Not huge certainly, but about the same slowing effect as carrying up a second small water bottle, or an extra 390 gram on the frame or wheels.

 

[artemidorus]

From my 9 sp triple, would estimate the chain angle in 30/12 (worst-case) is 5.3 degrees. The perpendicular component of 5.3*(cosine) is 0.9957, meaning .43% of power is lost due to the chain angle. At 300 W output on a hill climb, it's 1.3 W lost.

I argue that the perpendicular vector does no work, as the sprocket is not accelerated perpendicularly, and is consequently not a power loss, except in so far as it increases hub friction.

 

[sogood]

Significant energy loss is unproven, at least in terms of data we can get our hands on - sound power can be very low.
Perhaps we can get JTE83 to do a study?

Ok, at least we agree that the wear issue is significant and cross chaining is not a good practice on this account alone.

As for the power loss, our subjective reference on "significant" seemed to differ. We'll wait for harder data on that.

 

[matagi]

The following was lifted from discussion in another place. I offer it here as food for thought ......

In recent years proper research has been applied to cycling by many universities around the world and many of the old cycling myths have been found to be not relevant to modern cycling hardware.

Cross chaining was once considered to be harmful to the chain and it was thought that it would cause excessive wear on the cassette/freewheel and chainring teeth. It probably did cause problems on the older five and six speed freewheels and the conventional wide chains used in the "old days".

Some notable research was undertaken by Kyle and Berto that suggests that cross chaining is not an issue with modern equipment. (Kyle C.R., and Berto, F. 2001. The mechanical efficiency of bicycle derailleur and hub gear transmission systems. Human Power 52: 3-10.).

An overview of this research is described by Chester Kyle in Chapter One of "High Tech Cycling" Second Edition, edited by Dr Edmund Burke, Human Kinetics 2003. Available at the larger bookshops in Oz.

Of course individual bike setup plays a big part in how successfully you can "cross chain". Drivetrain alignment and Chainstay length play a big part in the equation as does the relative sizes of your chainrings and the range of your cassette freewheel. As a general rule, if your drivetrain is silent when cross chaining then you are probably not putting any undue stress on the system or causing undue wear.

The current guidelines put out by Campy and Shimano about derailleur capacity etc etc assume that you WILL be cross chaining at some stage.

Interestingly, Kyle and Berto did observe that smaller rear cogs did cause noticibly more friction and their recommendations include that for maximum efficiency you should not use a rear sprocket smaller than 14 teeth.

 

[sogood]

...except in so far as it increases hub friction.

That's another source of power loss isn't it?

 

[artemidorus]

That's another source of power loss isn't it?

Yep, but I reckon it'd be insignificant.

 

[sogood]

The following was lifted from discussion in another place. I offer it here as food for thought ......

I think there's some truth in it in that component makers would improve the design of their system to permit greater operational envelope and reduce the differential of wear b/n straight vs cross-chain conditions. But the fact remains, there's still a differential. It would be interesting to know how the study defined "not an issue". The other point worth considering is what's the effect of environmental factors on these, which I haven't seen any simulated lab studies. Would those fine sand particles disproportionally worsen the wear rate under cross-chained condition than when straight? I can just picture particles trapped b/n chain's outer plates and the cog teeth and subjected to a compression force in the direction of the axle.