IanS

After reading the recent post on aerodynamics with interest, I thought I ask whether anyone has any scientific data on the power loss in transmission on a tandem.

As tandemgeek pointed out in the post on aerodynamics, the combined power of pilot and stoker delivered to the rear wheel on a tandem will never equal the sum of the individual rider's power output. The amount lost will depend on how well the riders work together. This is clearly the most significant factor in power loss on a tandem as can be seen from the fact that 2 well matched good riders can beat 2 poorly matched stronger riders. It's also evident when you take out most of the affect of aerodynamics and look at tandems up hill. A well matched pair can ride hills well but never as well as a good single bike climber.

What I'm interested in is whether anyone has looked at this affect scientifically? Or whether anyone has looked at power loss purely through transmission? My pilot has rigged his tandem up to his turbo and believes he's loosing 50-100 Wats riding on the front of the tandem as opposed to his single bike. This isn't scientific and based purely on his perceived effort but suspect there is a perceivable drop.

From the post on aerodynamics, a tandem typically experiences 130-150% of the drag of a single bike. But if you're loosing power in transmission through not riding well together, a team's total power may only be 175% or even less. Then when you consider the fact that as soon as one rider on a tandem is trying to push/pull harder than the other, they're trying to accelerate the entire bike on their own, working out of sink, no matter how little, is likely to be more fatiguing and result in a greater power loss over a even a 10 mile TT towards the end.

This is also bourne out by the results from 10/25/50 mile TT where tandems participate with single bikes.

Clearly a well matched strong team on a flat course with either a head or tail wind should be faster than any similarly matched single rider. But how often is that the case?

I think I've kind of argued that the power loss in transmission due to riders not working well together plus any mechanical loss is ultimately greater than 25% of the sum of the individual riders power?

Cheers

oldacura

My guess is that the transmission losses are very low (<10%). Also, I don't see how a team "working together" is any more efficient than a team that is not. Ultimately, its just how much power the two riders get to the ground.

Now, if one rider is truely trying to oppose the other rider, that is wasted energy. I doubt that we have ever done that for more than a couple of strokes or so.

merlinextraligh

The loss due to mechanical friction would be very low, perhaps even negative (while the tandem has the timing chain, it only has one drive chain, nd one set of derailleur pulleys. )

The loss due to teamwork issues depends on the team, but with a good team can be very low.

Thus two strong tandem riders can put the hurt on a group of singles.

For example our memorial day ride with most of the local racers, two tandem teams were driving the pace in the low 30's for prolonged sections, putting a serious hurt on a lot of people. Of course it helps that male half of one of the teams won the Tandem 90+ National TT championship, and the female half is the about the strongest amateur woman cyclist you'll find.

IanS

Whilst I can accept that mechanical losses should be low in theory by comparison this doesn't explain why my pilot believed he was having to put in a lot more effort to maintain the same speed/power on the front of his tandem as opposed to his single bike on the turbo. Maybe power was being lost having to drive the rear cranks around on his own but I can't think that would make such a difference. The other mechanical losses could be through chain strech/tension in the link chain but again, I can't see that being much either. Any other ideas?

I do still believe there is a much bigger loss through teams not riding well together. If anyone's not convinced of how significant this could be you can take it to it's limit and just get one rider to stop peddling. And if there isn't a signficant loss, why won't a tandem go up hills as quick as a good climber on a single bike. In theory, you've got less than twice the weight of 2 single bikes + riders and if no significant power loss, twice the power. So it should be faster. I don't care who you put on the tandem, it will never climb as quick as a good single climber.

Another way to think of it is that on a single bike, if on your trailing leg (the one coming up) you do not feel pressure on the top of your foot or feel any pressure on the base of your foot, part of the force you are exerting on your leading leg is being used to lift your trailing leg. So even on a single bike, if you are not peddling in perfect circles which nobody can, you're resisting your own force at some point in the cycle. When you add another person into the mix, the affect will be exaggerated.

The laws of physics dictate that if one ride is exerting more force than the other, they will be trying to accelerate the entire bike for however long that unbalanced force exists. This will be more fatiguing than on a single bike where you are only tring to accelerate your own weight.

merlinextraligh

The belief that tandems don't climb well is a misnomer. If you take two riders of equal power output they will climb just about as well as they each climb on single bikes, particularly if they have good technique.

The belief that tandems don't climb well comes from the fact that tandems are faster for a given power output on the flats. Thus a weaker tandem team can keep up with stronger single riders on the flat because of the aero advantage, but when the road turns up the aerodynamic advantage diminishes, and the weaker team gets dropped.

As an example Captain with an FTP of 4 w/kg, stoker with an FTP of 3w/kg, and a single rider with FTP of 4w/kg. The tandem team with an FTP of 3.5w/kg as a team (assuming equal sizes) hangs with and may even be faster than the single on the flats. But uphill the tandem team with a deficit of .5w/kg is destined to be dropped.

But if both tandem riders have 4w/kg, they drop the single rider on the flats and stay with him on the hills.

merlinextraligh

Couple of possibilities. One, your combined power measured as w/kg may be lower than his w/kg. If so its going to feel to him like he has to work harder than on a single bike. It's essentially likes he's gained weight and lost power.

Two, he may not be getting maximal effort from the stoker. Before we started training, my stoker didn't really have a good sense of perceived effort or how hard she was working. I often thought that we were not going as fast as we should be. Now that we're actively training together,a nd she's using a HRM, she has a much better feel for perceived effort, and she produces much more power.

Thus there can be losses in effiency from the 2 riders matching efforts. But if you've ridden with an elite tandem team, you can easily see that those inefficiencies can be overcome, at least to the point where the aero advantages of the tandem significantly outweigh any inefficency.

There's no other way to explain how a mixed tandem team can motor off the front of a pack being driven at the front by a number of Cat 1 and 2 riders.

oldacura

For a tandem climbing below about 10mph (where aerodynamic factors are insignifcant), speed is just power to weight ratio. A stronger, lighter team will outclimb a weaker, heavier team.

The only advantage a tandem has over single bikes is wind drag. When you are riding below about 10mph, that advantage disappears.

Hermes

From https://www.kreuzotter.de/english/espeedfaq.htm

Below is a list of variables. The formula will not print but the key variable, mechanical losses, for the question is cm which directly reduces the power e.g. 300 watts produced will be decreased by 300W/cm. The max for cm is 1+9% for all bicycles including tandems. The spread is 6%. He does not say on the website if tandems have the largest losses. I always assume we lose a few % of the captains power in the timing chain and rings. If I found that the captains power was much more than a few % different riding solo on the tandem than on a single bike on the trainer, I would suspect some other problem with the tandem.

y: Dependent variable: Travelled Distance [m]
P: Rider's Power [W] (zero for coast-downhill simulations)
cd: Air Drag Coefficient Bicycle+Rider
A: Frontal Area Bicycle+Rider [m2].
The Effective Drag Area cd*A can be obtained from the Bicycle Speed&Power Calculator
M: Total Mass Bicycle+Rider [kg]
cr: Rolling Coefficient (between 0.005 and 0.009)
Approximated values can be obtained from the Bicycle Speed&Power Calculator
cm: Coefficient for Mechanical Losses (between 1.03 and 1.09, dependent from bicycle type)
ha: Altitude of the starting point above sea level [m]
T: Air Temperature [°C] on the altitude of the starting point
p0: Air Pressure on sea level at 0° C (101325 Pa)
ρ0: Air Density on sea level at 0° C (1.293 kg/m3)
Vw: Wind Speed [m/s] (negative values: tailwind)
β: Slope Angle (negative angles: downhill).
If the slope in percent (%) is given, atan(s/100) yields the slope angle (either in radians or degrees, dependent from the calculation mode setting).
g: Gravitational Acceleration (9.81 m/s2)

IanS

Completely agree with the above and that in theory, tandems should not be any slower up hills. Also agree with the reasoning above why tandems are perceived to be slower up hills. However, I'm just not convinced it is possible to ride together as a team so perfectly so as to eliminate any power loss through fighting against each other. The main thrust of my question is that I believe this is more significant that we might think. But even if the power loss isn't as significant as I believe it could be, the fact that when you're not riding totally together, no matter how little the difference might be, the fact that when this happens it is more fatiguing due to the fact that you're trying to accelerate the entire bike, that the overall affect in a TT is significant. That's what I'm interested to know whether anyone has studied or investigated.
To think of it another way, take any reasonable climb over 1000 feet. If, for the sake of argument the record for this climb held by an elite rider is 15 mins, I personally do not feel that the same elite rider on a tandem with a simlar (or even better) climber would be able to match that time. In theory, I agree, they should be able to. But I personally just do not believe they could for the reasons above. I'm just interested to find out whether anyone has scientifically looked at this and what the difference could be.
Alternatively, if anyone knows of a respectiable climb where an elite rider has also ridden the same climb with a similarly matched or even better rider on a tandem faster than on their single bike I would be interested to hear about it. Nothing like a good counter example to disprove a theory!

asu_gt

In our testing, there is about 6% power loss at the captain position and ~1.8% power loss at the stoker position. these tests were conducted in the lab on a stationary tandem with captain or stoker only pedaling. the power data from the srm was compared to power readings at the hub using a power-tap.

once we do some on-road testing we will submit a manuscript for publication.
jay

oldacura

It sounds like you're assuming that a single rider on a bike is perfectly efficient. This is not true. The greatest power loss in a bike drive train is in the rear chain and both single bikes & tandems have this. A correctly adjusted tandem timing chain should have very little loss.

A tandem is a compromise for both riders. Even a custom tandem that has been designed & set up optimally for both riders still requires that they ride at a compromised cadence. If each rider is found to have optimum power output efficiency at a significantly different pace, the compromised pace will be less efficient for both.

The idea that the riders will be fighting against each other seems strange to me. I think this could happen only if one rider were putting backward pressure on the pedals. For us, this occurs very seldom & only for a very short period of time. Why would someone pedal backwards? Wouldn't this say something about your relationship in general? So long as both riders were putting some forward pressure on the pedals, they wouldn't be fighting each other.

TandemGeek

What's the net power loss when both riders are on the tandem and pedalling?

merlinextraligh

Interesting data. And I think it's consistent with what you see in practice. Assuming it extrapolates to the road, On flat ground a tandem, with approximately a 4% average power loss, but a big aero advantage, would be predicted to smoke an equally powerful single rider, and in fact it does.

As for climbing, the 4% power loss makes the tandem a little slower, but not by much, with more of the explanation of tandems getting dropped on hills coming from the loss of aero advantage, which was keeping them up with stronger riders on the flat.


An interesting inference from your study is that a competitive tandem team should consider putting the more powerful rider on the back.

Curious how many teams you studied? Whether experience of the riders (in general, and on a tandem) made a difference, and whether size/power differences between the 2 riders made a difference.

merlinextraligh

Another question, doesn't the powertap (measuring power at the hub) already assume some power loss from power measured at the crank, and calibrate for that?

Do you know what the power loss is between the Crank (SRM) and the hub (powertap) on a single bike?

TandemGeek

That's always been the case and what you'll usually see when you look back at photos from when track tandem racing was an olympic event. Unfortunately, that rubs up against the reality that your average mixed gender tandem teams face.

asu_gt

interestingly, when two people are pedaling the power loss was slightly less compared to captain only pedaling. It was around 4.5-6%...this was not statistically significant (likely due to the small sample size of the preliminary study), but it was a consistent pattern.

from our paper:
When the captain position is occupied and the stoker position vacant, power values for the captain were approximately 6-7 percent more than those recorded at the rear hub. This difference in power is greater than the 2-4 percent that has been reported previously on single bicycles due to mechanical loss in the chain drive system (Gardner et al., 2004; Kyle, 1994; Kyle & Caiozzo, 1986). The greater length of the drive train for a tandem compared to a single bike may explain the increased power variation during tandem cycling. However, overall differences in the accuracy of the SRM and PT systems may also have contributed to this difference (Gardner et al., 2004).

Phantoj

One thing that causes decreased performance for tandem teams is a lack of syncronicity of efforts. If I am riding at "80%" (of, say, LTP) and my stoker is at 80%, then we are doing well. But suppose we're both at 80%, and we start to slip off the pack on a hill. I ramp up my effort to keep up with the pack, but my stoker doesn't feel any urgency and keeps the 80% pace. So I'm at 120%, stoker is still at 80%, and we're climbing like we are both at 100%. Except I'm wearing myself out a lot faster. If on another hill, I'm riding up at 80% and my stoker is at 120%, again, stoker is wearing him/herself out faster.

asu_gt

I am not sure that the PT has a built in assumption in that regard and makes some sort of adjustment to the power measured at the hub. Check out the study by Gardener et al., 2004...However, overall differences in the accuracy of the SRM and PT systems may also have contributed to this difference (Gardner et al., 2004).

the data do fly in the face of the traditional gender selection of captain and stoker...we had one team argue that it may be better to put the stronger person in front since they had more power, they could afford to lose more than the weaker stoker...i think he just wanted to keep the captain's seat.

jay

merlinextraligh

I think this has been a particular problem for us, particularly since my stoker has never ridden competitively, hasn't formally trained much, and doesn't have a good feel for "perceived effort."

When we've just been riding recreationally it hasn't been an issue (other than to frustrate me occassionally.)

Now that we're preparing to race, she's paying careful attention to hr as we train, so she can uderstand her zones,and what she can maintain,and then we can use that to pace efforts.

IanS

Very interesting. Accepting potential measurement issues and small sample size, I'm happy with an upper bound for mechanical losses of 7% which sounds reasonable to me. Probably a bit less but certainly no more. Interesting conclusion about whether to put your stronger rider on the back? However, as I hope I have explained above, if you're looking at optimal performance, both riders should be as well matched in terms of power and how it is delivered. So there shouldn't be a stronger rider. In my case, I'm blind so haven't got the option, but it does mean that I spend a lot of time on the back solely focusing on matching exactly how whichever pilot I'm riding with is riding. And they're all very different. Indeed, I suspect my pilots get a little p*ssed with me continually saying "think big circles" when I feel they're starting to stomp. It's amazing what a difference it makes though which is why I believe it's affect is far more signficiant than perhaps some of us might think it is.
It's this differences in peddling between pilot and stoker that I referred earlier to as fighting each other. It doesn't have to be much, the difference can be very subtle. But because a tandem is twice the weight of a single bike, any difference is far more fatiguing as another member descibed nicely above. This is a much harder problem to try and investigate though for a given team.
It might be easier (if it hasn't already been done) to research the physiological affects on the body of a rider trying to accelerate a large mass at various cadances over descrete time periods and how this affects overall performance. But it would be very hard to simulate subtle changes in peddling action in any controlled way to give an idea of exactly how this affects performance as a whole. My gut feeling is that it's significant.
In fact, an interesting comparison would be to look at the world best times over 10,15,25,50 and 100 mile TT on both singles and tandems. It's obviously not particularly scientific as course, conditions vary, but I suspect that for flat courses without any great decents, from 25 mile up tandem times will be slower, and possibly also for shorter distances. You could argue that the best riders in the world would ride single bikes not tandems and so you wouldn't be comparing like with like. But if a tandem does have a 40% gain in power after drag and mechanicl losses a least, this should easily be enough to compensate for small differences in individual rider performance. And if the tandem is quicker over any distance, I'll put money it won't be 40% quicker! Anyone know any stats?