Note de ce sujet :
  • Moyenne : 0 (0 vote(s))
  • 1
  • 2
  • 3
  • 4
  • 5
Efficiency 2
#1
The following post is an entire topic copied from the discontinued NuVinci Forum. The majority is written by members of the old forum and not by me.

by gsimpson on Sun Apr 22, 2007 4:35 am

Any testing been done in regards to mechanical efficiency of these in comparison to hub gears?

by NuVinci on Mon Apr 23, 2007 9:29 pm

Thanks for the post! As you can imagine, we get questions like that quite often. While it's a lengthy discussion, here's a quote from the whitepaper, "The NuVinci Experience", that may help:

Efficiency in Theory and Practice

Efficiency, while seemingly a straightforward measurement, becomes more complex when there is a sea change in technology. By conventional laboratory means of measuring efficiency of a single fixed ratio or a series of discrete ratios, no CVT, including ours, will ever measure up to a derailleur. Outside of the laboratory, however, other factors come into the equation.

• Serious riders spend a great deal of effort to learn optimum gears for every condition on derailleur bikes, and then still don’t always get it exactly right. The inexperienced casual rider is often bewildered by derailleur shifting – some combinations of the front and back sprockets can be duplicates and the path from the lowest to the highest ratio is not straightforward.

• Anyone who rides with any frequency has been on a hill where one gear was too high, the next one down was too low, and you simply could not hit a cadence that felt good until you lost some speed to match cadence to road speed. Finding the right gear on a derailleur is a challenge, for all the reasons discussed above. This problem is made worse by the wide ratio gaps found in many internal gear hubs.

• The NuVinci CVP allows the rider to continuously optimize cadence, and that makes you, the rider, more efficient. This is a substantial offset to any losses in the bicycle’s drivetrain efficiency because the rider is part of the total efficiency. To use an automotive analogy, major automakers are switching to CVTs to improve their total powertrain efficiency and hence fuel efficiency.

• The NuVinci CVP enables a simple, seamless, precise adjustment in ratio with no disruption in torque. The ratio change can be as small or as large as you need it to be without gaps – no more making an educated guess about what the right gear is. You just twist the NuVinci Cruise Controller until the cadence feels right. A casual rider can adjust cadence like a seasoned pro.

• If the chain falls off, efficiency over the course of a particular ride obviously drops dramatically. With NuVinci technology, as with internally geared hubs, the chain line is always straight. Even during the most demanding/rapid ratio changes, the chain will remain in line and will not jump off the chain ring or the cog.

• Because the NuVinci CVP can be shifted while stopped, it is not necessary to be pedaling to shift a NuVinci hub. If you come to a stop and either forget or do not have time to downshift, getting started again can be a challenge with a derailleur. This is particularly true for less skilled riders. The NuVinci CVP’s ability to be shifted at any time eliminates that problem.

Riders tell us that smooth shifting, the ability to find easily the right ratio and to be able to shift at any time really makes a difference.

A Fallbrook Experiment
In an attempt to determine how much the delta in traditional efficiency versus always being in the right gear would measure out under real world conditions, Fallbrook conducted an experiment.
Three different Fallbrook riders switched out between three nearly identical comfort bikes, one equipped with a 21 speed derailleur, one with an 8-speed internal gear hub, and one with an early prototype NuVinci CVP.

Four different routes were used, totaling 1,388 km, with all four routes traversing portions of the Central Texas Hill Country and its attendant rolling grade changes. Two of the routes had significant grades of 15% or more, the other two routes were more typical commuter routes with lesser grade changes.

We collected data on speed and ride completion time. At the end of it all, we found no significant difference in average speed or ride time between the internal gear hub and the CVP. Furthermore, the NuVinci CVP’s performance was very nearly equal to the derailleur-equipped bike’s, without the worry of struggling to find the right gear, or the associated noise and clatter.

The smoothness of the ride, the ability to find easily the rider’s best/most efficient ratio, and the ability to shift when not pedaling contributed to these results.

Based on the above experiment and for the reasons outlined above, we believe that the overall “ride” efficiency of a NuVinci CVP is potentially better than that of internally geared hubs and may in some circumstances exceed that of derailleurs.

you did not answer qestion. how much efishen c is lost in

by dug on Tue Jun 12, 2007 1:20 am

ATCNuVinci wrote:Thanks for the post! As you can imagine, we get questions like that quite often. While it's a lengthy discussion, here's a quote from the whitepaper, "The NuVinci Experience", that may help:

Efficiency in Theory and Practice
Efficiency, while seemingly a straightforward measurement, becomes more complex when there is a sea change in technology. By conventional laboratory means of measuring efficiency of a single fixed ratio or a series of discrete ratios, no CVT, including ours, will ever measure up to a derailleur. Outside of the laboratory, however, other factors come into the equation.

• Serious riders spend a great deal of effort to learn optimum gears for every condition on derailleur bikes, and then still don’t always get it exactly right. The inexperienced casual rider is often bewildered by derailleur shifting – some combinations of the front and back sprockets can be duplicates and the path from the lowest to the highest ratio is not straightforward.

• Anyone who rides with any frequency has been on a hill where one gear was too high, the next one down was too low, and you simply could not hit a cadence that felt good until you lost some speed to match cadence to road speed. Finding the right gear on a derailleur is a challenge, for all the reasons discussed above. This problem is made worse by the wide ratio gaps found in many internal gear hubs.

• The NuVinci CVP allows the rider to continuously optimize cadence, and that makes you, the rider, more efficient. This is a substantial offset to any losses in the bicycle’s drivetrain efficiency because the rider is part of the total efficiency. To use an automotive analogy, major automakers are switching to CVTs to improve their total powertrain efficiency and hence fuel efficiency.

• The NuVinci CVP enables a simple, seamless, precise adjustment in ratio with no disruption in torque. The ratio change can be as small or as large as you need it to be without gaps – no more making an educated guess about what the right gear is. You just twist the NuVinci Cruise Controller until the cadence feels right. A casual rider can adjust cadence like a seasoned pro.

• If the chain falls off, efficiency over the course of a particular ride obviously drops dramatically. With NuVinci technology, as with internally geared hubs, the chain line is always straight. Even during the most demanding/rapid ratio changes, the chain will remain in line and will not jump off the chain ring or the cog.

• Because the NuVinci CVP can be shifted while stopped, it is not necessary to be pedaling to shift a NuVinci hub. If you come to a stop and either forget or do not have time to downshift, getting started again can be a challenge with a derailleur. This is particularly true for less skilled riders. The NuVinci CVP’s ability to be shifted at any time eliminates that problem.

Riders tell us that smooth shifting, the ability to find easily the right ratio and to be able to shift at any time really makes a difference.

A Fallbrook Experiment
In an attempt to determine how much the delta in traditional efficiency versus always being in the right gear would measure out under real world conditions, Fallbrook conducted an experiment.
Three different Fallbrook riders switched out between three nearly identical comfort bikes, one equipped with a 21 speed derailleur, one with an 8-speed internal gear hub, and one with an early prototype NuVinci CVP.

Four different routes were used, totaling 1,388 km, with all four routes traversing portions of the Central Texas Hill Country and its attendant rolling grade changes. Two of the routes had significant grades of 15% or more, the other two routes were more typical commuter routes with lesser grade changes.

We collected data on speed and ride completion time. At the end of it all, we found no significant difference in average speed or ride time between the internal gear hub and the CVP. Furthermore, the NuVinci CVP’s performance was very nearly equal to the derailleur-equipped bike’s, without the worry of struggling to find the right gear, or the associated noise and clatter.

The smoothness of the ride, the ability to find easily the rider’s best/most efficient ratio, and the ability to shift when not pedaling contributed to these results.

Based on the above experiment and for the reasons outlined above, we believe that the overall “ride” efficiency of a NuVinci CVP is potentially better than that of internally geared hubs and may in some circumstances exceed that of derailleurs.

Re: you did not answer qestion. how much efishen c is lost

by NuVinci on Thu Jun 14, 2007 11:17 pm

dug wrote:you did not answer qestion. how much efishen c is lost in the hub not the dumy shifting gears

Hmmm...are you asserting in those last few words that only dummies bother with gears?

Straight up, the second sentence of the excerpt from the Engineering White Paper really tells it like it is, if you're looking to isolate the hub at a given ratio:
By conventional laboratory means of measuring efficiency of a single fixed ratio or a series of discrete ratios, no CVT, including ours, will ever measure up to a derailleur.

But, for a real rider who is changing ratios during the trip, the other efficiency benefits associated with the absence of derailleurs come into play, as the excerpt states.

Numbers please

by fb on Fri Jun 15, 2007 10:55 am

By conventional laboratory means of measuring efficiency of a single fixed ratio or a series of discrete ratios, no CVT, including ours, will ever measure up to a derailleur.
The problem is: NuVinci hub owners know all this, but sceptics (non-owners) use the lack of numbers to pick on the NuVinci hub. If we had numbers we could contradict this (say the efficiency of the hub is as good as a medium dirty derailleur).

The high weight is also a problem. But we can contradict that by saying that the added weight is the same as a good meal

by Harvey_Mushmann on Tue Aug 14, 2007 6:49 am

Regarding efficiency- the sheer weight of the NuVinci, when compared to traditional derailleurs, kinda' renders that argument a non-starter.. better to compare apples to apples

A NuVinci V. Rohloff efficiency comparison would be FAR more fair methinks.

_ _ _ _ _ _ _ _ _ _ _ _ _ _

Ride Bikes. Drive Culture. -Harvey

by Harvey_Mushmann on Wed Aug 15, 2007 7:40 am

Here's a big, spendy, apple to sink your teeth into..

http://www.sheldonbrown.com/harris/imag ... bild2.jpeg

..what better opportunity to let your big, shiny ceramic balls shine?!


_ _ _ _ _ _ _ _ _ _ _ _ _ _

Ride Bikes. Drive Culture. -Harvey

Efficiency...
by calfeenated on Sat Oct 13, 2007 11:19 am

Looks like nobody wants to talk specifics about the efficiency of these hubs and that is a real shame. None the less, I'd like to approach this from a different angle and maybe I'l get some info that might help me decide if I can use this hub for my Human Powered Vehicle.

If I put 500 watts of (human) power into this hub, at what ratio am I going to get the most power back out of it? I'm not talking speed, I am talking POWER here. speed is just a result of the other gearing that is optomized for the efficiency of the hub and the engine. It's clear to me from watching some of the video's of this hub that there is some SRM power data that has been collected on this hub but nobody is talking about it. Even the moderator has stated that "no CVP, even ours can approach the efficiency of a chain and deraileur system" but this hub might still be usable if we KNEW where it was most efficient.

It's a known fact that the Rohloff speedhub is most efficient at it's 1:1 ratio and the power losses go way up if you start using it to gear up which happens at the 11th gear. So you work around that by making the 11th gear your top gear and use the other 10 to get you there.

I'd like to use the NuVinci hub as a mid drive for a streamlined recumbent bicycle much like this team used a Rohloff speedhub

   

This system works great for them, it approaches the efficiency of an all chain driven system BECAUSE they know what ratios to stay away from. If I knew what the most efficient part of the NuVinci hub was that would give me an idea of how to best use it or if I should use it at all.
If it was succesfull it could be some great press for the NuVinci folks.

by tox20 on Wed Oct 31, 2007 9:45 pm

After debating this yesterday I have been convinced of one conclusion/ resolution. You can have the most efficient drive train on the face of the planet (single speed) but with poor gear selection your body will operate inefficiently. So the goal of every cyclist is ratio selection with at the highest efficiency. That does not mean that ratio selection is always better than efficiency. You have to know your application. Chances are if you are worried about efficiency this isn’t the hub for you. Just look at the weight. Anyone worried about efficiency but willing to take a weight hit of this magnitude, needs their head examined with a mallet. The Nuvinci has peak efficiency at 1:1 but I don’t know how it compares to Rohloff . it’s cheaper and way smoother though.

Efficiency
by LeakyDuck on Fri Nov 16, 2007 12:57 am

This question has been asked over and over in this forum and others. And always the answer is a bunch of double talk. Never any hard numbers for people to rely on.

In the Summer of 2001 in the Technical Journal of the IHPVA they did answer this question. They gave straight forward answers for each of the systems they tested. I fail to believe that your engineers can't do the same again and add the NuVinci to the test. Given the variable nature of the NuVinci the data for the highest and lowest would give some real life numbers to go by.

The fact that you won't publish any data on this subject makes me feel like you probably have already done some testing and your hub did not do well so you hide the fact in double talk.

I know about all the weight ect .... issues - they are not germain to this question. If I am to invest my time and money to acquire a NuVinci Hub and have it installed on my bicycle I need information to make my decision. If you continue to hide this information, my decision will be to not buy a NuVinci.

Re: Efficiency
by NuVinci on Thu Dec 06, 2007 6:17 pm

I understand you want to have all the information before you decide to purchase a NuVinci equipped bike. Please allow me to address your concerns as clearly as possible.

1) The best information you can have before you make your purchase decision is to test it yourself. Ride a NuVinci equipped bicycle and see what you think and what your body tells you about the efficiency. In Europe there are over 20 models for sale from over 15 brands reaching well over 2,000 dealers. In the US we have 4 OEM customers, one of which is selling a special edition Cadillac on the Costco website.

2) Every bicycle hub transmission is slightly less efficient than a derailleur bike. A derailleur bike starts around 91% and that is where all hub transmissions end (depending on gear).

3) We can say that the ATC NuVinci CVT for bicycles is comparable with all other hub transmissions in the market.

4) Most of those who ride a NuVinci bike think the efficiency is higher (including Peter Winnen (3 X time Queens (mountain) stage champion for Tour de France) and Uli Schoberer (founder of SRM). Please view our video on nomoregears.com or YouTube. Why?

− With NuVinci you are always in gear. Even when you 'change' gears. The power is never interrupted as it is with all other systems.
− It allows you to operate at your engines (heart, lungs, and legs) optimum cadence. You are never locked in between two gears. You are never too high or too low.
− A leading sports car compared a manual transmission (which on average is approx 10-12% more efficient than an automatic transmission, with a CVT and an automatic transmission. While the CVT is inherently less efficient than a manually geared system (just like a hub is less efficient than a derailleur), a car equipped with the CVT was still seconds faster than the other two gear systems.

As noted above, we have a video about the above where we asked critics from Polar Heart Rate monitors, SRM power measurement equipment and a well known Tour de France stage winner to test to NuVInci system. If you want we can forward this to you or you can view it on http://www.nomoregears.com or youtube.com. Unfortunately these miss the subtitles, but we are happy to send you the DVD with English narration and subtitles.

But again, don't believe us, nor the experts. Test a NuVinci bike yourself and decide for yourself. In this case we say, don't trust figures but listen to your legs, heart and lungs. They are part of your drive train. Other systems often forget that, we don't.

Re: Efficiency
by PantoJack on Mon May 09, 2011 11:08 pm

"We can say that the ATC NuVinci CVT for bicycles is comparable with all other hub transmissions in the market."

Great. So how efficient is that then?
Shall I go and spend some time looking at how efficient other hub transmissions are, then come back here and post those figure here, and state that such figures represent the efficiency of NuVinci hubs?

I get the 'consider-the-bigger-picture' thing, and that is totally valid, but you are still holding back, not being open.

Why not publish some representative efficiency figures AS WELL AS making the bigger-picture argument?

People wanting some efficiency figures aren't "attacking" - I am supportive of your products / success, and am impressed overall, but you are being evasive / hiding some of the truth.

PS: "Test a NuVinci bike yourself and decide for yourself. In this case we say, don't trust figures but listen to your legs, heart and lungs." Mmm, actually, I'd prefer to be able to do BOTH!
Come on, don't be shy!

How efficient is the NuVinci N360 at a couple of well-spaced ratios to within a couple of percent?

Thanks in advance. ;O)

Re: Efficiency
by PantoJack on Mon May 09, 2011 11:32 pm

I've found some answers related to the NuVinci's efficiency here (but no figures):

http://forums.mtbr.com/showthread.php?t=663451
See posts #19 to #21 for what appears to be a good balance of the pros and cons.

These are very helpful to me as someone who enjoys attacking big climbs around where I live: as much as I enjoy the challenge of long steep climbs, I'm not looking for extra drag to slow me down on them!
(Your typical bike-shop test-ride wouldn't give me enough time to get out to such hills...)

Re: Efficiency
by mingoponta on Sat May 21, 2011 3:48 pm

Two days ago I was in Zurich and "rented" (for free) a byke from the City of Zurich. The bike was basically brand new and charaterised by a sizeable hub in the rear.
I started to ride it and, with my surprise, I noticed that there was not gear shifting but a continous change of the transmission ratio.
This has been my first encounter of the NuVinci 360. The ride was great, because of the exceptional qualities of the bike (and of the city of Zurich!).
I got very interested on it and I ended up in this website, looking for data on the efficiency of the gear.
I ride with pleasure both road bikes and MTB and I'm not restricting myself to the use of light bikes.
I was enthusiatic about NuVinci 360 but, since I am an engineer, I still am curious to know quantitave data on the energy efficiency.
After my casual test I'm looking forward to get a bike with NuVinci 360 but the fact that its efficiency looks like a well guarded secret bothers me a little and has prompted me to join the forum and post this message.

Re: Efficiency
by PantoJack on Thu May 26, 2011 10:20 pm

Dear NuVinci,

This is (at the time of posting this) the most read thread/topic in this forum. Clearly people are interested in the efficiency of your hub. (And why wouldn't they be? It's an absolutely fundamental feature of such a product!)

So can you please tell us all, has NuVinci actually done any proper testing on the efficiency? This remains the unanswered question that started this thread.

I know that efficiency will be a function of various input variables, but have you tested it at representative input torques (say 80% or 90% of rated) and at, say, two or three widely-spaced ratios at representative input velocities?

Either you're not providing representative efficiency figures because (1) you haven't carried out proper testing yet, or (2) the hubs are inefficient. Which is it? Maybe you haven't done serious testing because, well, they're selling well anyway, so why bother, which is fair enough.

People are just going to keep asking! And the more you resist, the more curious I certainly get! ;O)

-----------------------------------
PS: You said, "A derailleur bike starts around 91% and that is where all hub transmissions end (depending on gear)." Do they? Do Rohloffs end at around 91%?! I am only going from memory, but I thought they were in the high 90s in some ratios? (Someone can correct me if I am wrong on that.)
You also said, "We can say that the ATC NuVinci CVT for bicycles is comparable with all other hub transmissions in the market." So are you claiming that it is comparable to a Rohloff in efficiency then? If so, you MUST surely have done some testing? (I mean, other than a few people going for a ride and timing it) Because that's a pretty strong claim to make if it's based on guesswork...

Re: Efficiency
by PantoJack on Fri May 27, 2011 3:17 pm

This from http://www.rohloff.de/en/technology/efficiency/

"the working efficiency of the Rohloff SPEEDHUB 500/14 is ...96% (for gears #1 to #7) and 98% (for gears #8 to #14)."

NuVinci: I think your claim above - "We can say that the ATC NuVinci CVT for bicycles is comparable with all other hub transmissions in the market" - is seriously misrepresentative.

Re: Efficiency
by Twentyniner on Tue Jun 21, 2011 6:18 am

The Rohloff is in 1 : 1 in gear 11. In the other gears the power goes through 1 or more extra planetaries. Interesting they claim 98% for gears 8 to 14. In my opinion impossible.

Re: Efficiency
by PantoJack on Wed Jun 22, 2011 12:24 pm

In the absence of NuVinci's ability &/or willingness to provide data on the efficiency of their product (NuVinciSupport has rather conveniently ignored all my posts on efficiency despite responding to others who have posted since in this forum), I thought I'd post this chart, which throws some light on the subject.

It is a chart of the efficiency of the "Tilting-ball drive" type of CVT (the same type as the NuVinci):

   

Source: Chapter 13 of McGraw-Hill's 'Mechanisms & Mechanical Devices Sourcebook', which is freely accessible at the link below at the time of posting**. See relevant text on pg 343 and Fig. 15 on pg 344.

As you can see, the efficiency peaks at about 89%.

Now the [Output : Input] ratio of the NuVinci ranges from "0.5 Underdrive" to "1.8 Overdrive". (Source: http://www.fallbrooktech.com/docs/N360_D...nglish.pdf), so this chart suggests that the NuVinci's efficiency will lie in the range between about 84% and 89%.

(These sorts of figures appear, from my limited foray into the subject, to be about as good as it gets for traction-type continuously variable transmissions...)

-------------
**This chapter of the book can be accessed via this link without charge at the time of posting:
http://www.accessengineeringlibrary.com/..._ar013.pdf
Go to the 35th and 36th page of that pdf, or text-search "Tilting-ball drive" or "Fig. 15".
Last edited by PantoJack on Thu Jun 23, 2011 7:08 am, edited 4 times in total.

Re: Efficiency
by trike on Wed Jul 06, 2011 3:11 pm

Nice chart, PantoJack.

Interesting to see that efficiency on the tested device differed significantly 1/3x vs 3x, so its not symmetric. Overall it is similar to what I guessed but its nice to see numbers with it.

I imagine that different "tilting-ball" transmissions must have differing peak efficiencies but overall similar shaped curves. The efficiency of the N360 at its fastest ratio is of interest to me because in my application it would usually be used near the top of its range, with the lower regions being for hills. I wonder if they limited it to 1.8x out of concern for efficiency (poor efficiency being more tolerable on short hills than long distance cruises). This is potentially quite important for people who have no use for gearing that allows pedaling down hills (as opposed to coasting down).

Re: Efficiency
by PantoJack on Thu Jul 07, 2011 2:46 am

Cheers trike,

By the way, if the Log of the ratio, instead of just the ratio, were plotted along the graph's 'x' axis (which is arguably more appropriate so that the distance along the x-axis between say 1/2 and 1 - a doubling of ratio - was equal to the distance between 1 and 2 - also a doubling of ratio), it would be a lot less unsymmetrical.

I agree with you, "it's nice to see numbers with it". Or rather, it would be if NuVinci provided some...

------------------------------

I have had a thought since I posted the above chart. It is possible that the above graph actually flatters the NuVinci because, given how effective the traction oil used by NuVinci is reputed to be (acting something like a high-shear-strength glue at the point of contact between balls and discs), the NuVinci might suffer significantly greater "spin" losses than the system whose efficiency is plotted above.

Therefore, it is possible that the efficiency of the Nuvinci is only in the low 80%s or high 70%s.

I really LOVE the idea of CVTs for bicycles. I have wanted my bikes to be CVT-equipped since around 2002 or 2003. I really want to like the NuVinci hub. And I really want it to be more efficient than high 70%s / low 80%s. But it is possible that it isn't any more efficient than only around 78% or 79%...

Re: Efficiency
by Oran on Sun Jan 08, 2012 10:29 pm

This post is not going to be very helpful to those seeking efficiency data, it is instead an attempt to explain why I think Fallbrook are unwilling to provide it and how I agree with them on their decision.

Efficiency data from Fallbrook for their CVT is out there on the internet if you know where to look. However I am not going to provide its location or giveaway any more than the graph with efficiency figures removed for the following reasons. Firstly it was provided back in 2005 and may not be up to date but mostly because I agree with NuVinciSupport that it may get misquoted or misinterpreted.

The CVTs efficiency is not constant across the ratio range but instead drops at both lower slip ratios and slightly at the higher ratios in a similar way as PantoJack’s graph shows. Peak efficiency is achieved around the 1:1 slip ratio. I also suspect that efficiency drops as torque increases. The answer to the question, what is the efficiency, is not simple.

Quoting the lowest efficiency is clearly not representative of the overall efficiency. It is in fact impossible to state an overall efficiency because it depends on which slip ratios you are using and for what percentage of the time. For example someone who uses mostly the lower slip ratios is going to have a lower efficiency than someone who uses the whole range. With any other internally geared hub you could time how long you use each gear and calculate an average efficiency if the efficiency of each gear ratio is known. It is much harder to do this with a CVT because there are an infinite number of gear ratios.

As most people will be looking for a single value for efficiency not a graph I reckon Fallbrook are concerned that the lowest value will get used. This would not be a fair representation of the hubs true performance and as I have explained it is not really possible to come up with an average efficiency.

The most I am going to say about actual figures is that the efficiency is considerably better than the figures mentioned by PantoJack.

My advice to anyone seeking efficiency data is to except that it can’t be narrowed down to one single number. The way you use the hub and the type of terrain where you ride will alter the efficiency.

I have revealed the efficiency figures for this forum.
   
CVT Efficiency graph.jpg (20.06 KB) Viewed 181 times
Last edited by Oran on Mon Apr 01, 2013 10:57 pm, edited 2 time in total.
........__o
........\<,
......()/()

Re: Efficiency
by bigoilbob on Mon Jan 09, 2012 2:20 pm

Hello Oran;

You are indeed correct that efficiency at any operating condition can not be represented by one number. But Fallbrook has already said they did complete testing, so efficiency could be represented by a family of curves, with the variables representing anything that could change efficiency. I would think the main sensitivities would be, ratio, load, and trannie temperature. The "way you use the hub" changes the already referenced sensitivities, and the hub can't tell what terrain it is on (bumps might make the trannie skip from a change in fluid yield point, but that is not an efficiency change per se). The curves would be harder to understand than one number, but I think anyone interested enough to ask could figure out how to read them easily. Provision of the curves would be a good marketing move, unless they point to a problem.

Best to you;

Bob D.

Re: Efficiency
by Oran on Mon Jan 16, 2012 10:50 pm

Hi Bob,

I agree with you that it is possible to quantify efficiency with a family of curves. Ratio, load, and temperature would be the main ones. What I meant by terrain was hilly or flat as this would have an effect on which ratios are likely to be used most. However this just makes it even more difficult to understand. This is fine for someone who is really interested but for the average person where’s that one number they are looking for going to come from which they can then pass on to someone else. There is no real problem with the efficiency apart from at low slip ratios.


........__o
........\<,
......()/()

Re: Efficiency
by jakelewis3d on Fri Jan 20, 2012 6:29 pm

I suspect the data that Oran has graphed is derived from page 5 of the report at

http://www.nrel.gov/docs/fy05osti/38212.pdf

However, this paper makes no reference to which particular model of Fallbrook's CVT this data applies to, and considering it's focus on commercial wind turbines ( and the 2005 date), I very much doubt it's an N360. I'd imagine the size of the balls would have some effect on efficiency, with the larger radii of the automobile CVTs being more efficient.

On the other hand, progress may have been made in the last 7 years since this paper was published in increasing efficiency, maybe in the traction fluid for example.

Also these are Fallbrooks figures, and do not include losses in the chain drive, which are good for hub gears (in respect to derailleur) in general due to straight chain lines (and subsequent lessened chain stretch), more so for bikes with horizontal dropouts or elliptical BBs that can avoid a chain tensioner.

In regards to Rohloffs claims of 98% efficiency, (also excluding chain loss I presume), figure 8 in these tests http://www.ihpva.org/HParchive/PDF/hp52-2001.pdf tells a different story.

In response to Oran's claim "I also suspect that efficiency drops as torque increases" the efficiency graphs for other hub gears tend to increase with power, although the CVT would have an upper limit after which slippage would have a drastic effect on efficiency, (and possibly cause permanent damage to the hub). Hence the limitations of single riders and minimum gear ratios.

I won't be fitting a CVT on my performance road bike, but equally, I won't be taking it off my city runaround any time soon.

Jake
Last edited by jakelewis3d on Fri Jan 20, 2012 8:22 pm, edited 1 time in total.

Re: Efficiency
by Oran on Mon Jan 23, 2012 9:51 pm

I was wondering how long it would take someone to find the source of the data I was referring to. When I first came across the data I thought the same as Jake. Yes it is referring to a larger transmission and so is probably more efficient but that progress may have been made since then with the efficiency of the current bicycle CVT.
After doing some research I think it might be a closer representation of the N360’s efficiency then I first thought. It’s safe to say that it won’t be the efficiency curve of the N360 as the N170 hadn’t even been released in 2005. I know there have been some modifications to the internal workings of the N360 mostly to reduce the size and weight compared to the N170 but I don’t know about the traction fluid, in the end it is still a tilting ball CVT. The largest design improvements I believe were made by Donald C. Miller in 2000 when he came up with a better CVT design for his world’s fastest bike project. If you look at the graph on PantoJack’s post of Jun 22, 2011 it looks like Don managed to increase the efficiency by about 5 -7%. It's going to become increasingly difficult to gain the last few percent.

The data to me looks like it has come from tests on a real CVT not from calculations. After receiving more money in 2003 Motion Technologies had an independent testing laboratory test their CVT. It was these tests that proved the potential and showed that the CVT could be used in many other applications other then bicycles. In the early days I would have thought they were concentrating on a bicycle sized CVT. When NREL asked them about the efficiency of the design for writing up the report I expect they gave them data from the independent laboratory tests on what I would have thought would have been a bicycle sized CVT. Whether the data has been adapted to take into account the larger size of the wind turbine CVT it’s impossible to tell.

Another interesting point is that if you download the NREL report from Fallbrook’s website instead of the link on Jakes post you get a slightly different version of the report. The same data table is labeled as being from Motion Technologies Inc. instead of Fallbrook Technologies Inc. Motion Technologies Inc. is the former name of the company and they became Fallbrook Technologies Inc. in 2004. Could this be to remove the connection between the data and their current CVT?

The Technical Journal of the IHPVA (link on Jacks post) gives some interesting efficiency figures for other internally geared hubs. Efficiencies in the mid to high 80s appear to be not that uncommon in some of the gear ratios of other hubs. It appears that Fallbrook doesn’t have anything to worry about.

As I said in a previous post efficiency is a complicated issue and this one efficiency curve doesn’t say much about the actual efficiency you will get in the real world. I will put up another post sometime about my claim that efficiency drops as torque increases.
........__o
........\<,
......()/()

Re: Efficiency
by flatboarder on Wed Jan 25, 2012 10:06 pm

Love this discussion. Being curious about the N360 made me get one and try it out. I did several trips with my big dummy equipped with an Alfine 8 speed hub (34/20, 160mm cranks), lots of lengthy climbs and things, even offroad, and i will be doing it with the NuVinci as well, hopefully. To me, efficiency definitely is an overall estimation. How long can I stay on the bicycle without feeling exhausted, keeping up a *reasonable* travel speed.

Although I would definitely be interested in numbers, to me this would never be a show stopper. I have been riding sports bicycles for a long time and still have got one, but I would not go back to derailleurs for several reasons with my other cycles. Besides, I also run a streetstepper, which provides a totally smooth excenter drivetrain: while definitely being slower than a bicycle (probably too slow to set off for lengthy trips, but I am still working on this) I feel perfectly well after my rides, even when riding higher avg. heart rates than with the bicycle. You just do not feel the workout as much. Really strange thing, but I love it.

However, in case the NuVinci drive would make me feel exhausted and take a lot out of me _in the long run_, I would go back to cheap and easy Alfine 8-spd, which in my case proved to be sufficient for touring, even for steep sections, although I certainly know it being worse than Rohloff with respect to efficiency numbers and gear ratio. It also has got very rough gear distances, so you cannot always select your optimum pedalling rate.

Well, there are several NuVinci Reviews in the web confirming the good riding experience. These were not only short distance riders. All in all, I am looking forward to my own NuVinci experience and my personal efficiency rating. However, I need to resolve a minor but strange chain issue before being able to ride it as smoothly as intended.

Phil
Last edited by flatboarder on Wed Jan 25, 2012 10:08 pm, edited 2 times in total.

Re: Efficiency
by Oran on Sat Jan 28, 2012 10:12 pm

Efficiency drops as torque increases.
What made me say the above statement was that I find it a lot harder to pull a heavy trailer with an N360 equipped bicycle then a derailleur bicycle. Whereas without the trailer the difference is much less noticeable. My conclusion was that the efficiency is dropping when I apply more torque required to tow the trailer. Or perhaps it is because I am going slower when towing and so I am using lower ratios that are less efficient. But then why are the lower ratios less efficient, perhaps it is because torque transferred through the wheel side of the transmission will increase as the ratio decreases. If an increase in torque causes a drop in efficiency then what happens when you pedal hard and increase torque on the drive side as well.

I will attempt to explain a bit more about efficiency which I have leant from Rohloff’s website http://www.rohloff.de/en/technology/efficiency/ and the talk section of Wikipedia http://en.wikipedia.org/wiki/Talk:NuVinc...ansmission
I’m going to summarize them here, look at the websites if you want more information.

All internally geared hubs including the N360 will have more than one type of power loss. The first is power independent losses due to friction for example at the seals and these losses as the name suggests won’t change much with the amount of power transferred through the hub.
Then there are power dependent losses which are caused by friction between gears (if it is not a CVT) and bearings. Power dependant losses will always be a proportion of the power input for example if they are 5% then a power input of 100W will result in 5W of losses, an input of 400W will have 20W of losses. Therefore when the power input increases the power independent losses make up a smaller proportion of the total losses and efficiency increases. This explains the graphs in The Technical Journal of the IHPVA which Jake was referring to.

As explained further on Rohloff’s website the curve of the efficiency against power graph makes it difficult to determine efficiency because a cyclist produces a cyclical torque and so applying the average torque in a lab won’t be accurate.

A CVT has a third kind of loss which makes calculating efficiency a whole lot harder. This time instead of the power dependant losses being determined by a set efficiency the actual efficiency changes with power. However as power is a combination of rpm and torque and it is possible to increase torque without increasing power. It is therefore more correct to say that the efficiency changes with torque. This is important because torque determines the contact pressure between the balls and rings in this type of CVT.
The reason that efficiency drops with increasing torque is to do with the design of the CVT. If you think about the surface of the balls it is rotating at different speeds, this is how the CVT gets the infinite gear ratios. The problem is that if the contact patch between ball and ring is not an infinitely narrow line parallel to the rotation of the ring there will be some slippage. When I say slippage I don’t mean a large movement between ball and ring. I’m talking about the area above and below the centre line of the contact patch which will have an area of ball that is trying to make the ring go faster on one side and slower on the other side of this centre line. As torque increases so does the contact patch and slippage, therefore efficiency drops. There is obviously traction fluid which has a part to play here but what I am not sure about.

The conclusion from all this is that instead of the efficiency against power graph increasing and then leveling off as power increases which is normal for other internally geared hubs, the N360 will have a graph that rises, peaks and then begins to fall as power continues to increase.
I know there are people who know more about transmissions out there (for example staff at Fallbrook) so please correct me if I am wrong. Perhaps a section on Fallbrook’s website similar to Rohloff’s would be useful. This could explain some of the difficulties with determining a CVT’s efficiency when used on a bicycle rather than just saying when asked about efficiency data “that is not information that we publish to the public”.

Last edited by Oran on Sat Jan 28, 2012 10:14 pm, edited 1 time in total.
........__o
........\<,
......()/()

Re: Efficiency
by seansimp925 on Tue Feb 14, 2012 7:27 pm

I am a mechanical engineer and I'm equipped with a CycleOps Powerbeam Pro, heartrate monitor (prevent sandbagging), video cameras, and all kinds of bikes including but not limited to:

Giant Seek 0 with 8-Speed Alfine IGH
Electra Townie with Nuvinci N360
Scott Aspect 55 (conventional derailleur)

You outline the test and I'll post up the data. All I ask is you keep an open mind and let the data drive the debate instead of speculation.

Let's party!
Last edited by seansimp925 on Tue Feb 14, 2012 7:29 pm, edited 1 time in total.

Re: Efficiency
by bigoilbob on Tue Feb 14, 2012 7:38 pm

Great idea seansimp925, and a service to the thread. Some of these posts could have been written by Cliffie, the mailman on Cheers. Lots of pseudo erudition with no good way of comparing NV efficiencies with their competitive products, for applications we are contemplating. i.e. "News we can use".
Re: Efficiency
by seansimp925 on Tue Feb 14, 2012 8:49 pm
Using the tools at my disposal, it seems the best way to test each system would be to "max out" my power output in a series of tests selected in a manner so that my decreasing energy level can be understood and removed from the final efficiency estimations.

I propose something like that following:
1. Max out on derailleur system (in theory the ratio I choose shouldn't matter in terms of bicycle efficiency but there might be a speed/torque value that is optimal for my body which would change the results - I'll try to play with this to determine a good benchmark that will be used as a theoretical max available power)
2. Max out on N360 in ratio X
3. Max out 8-Spd Alfine
4. Max out on derailleur system
5. Max out on N360 in ratio Y
6. Max out 8-Spd Alfine
7. Max out on derailleur system
8. Max out on N360 in ratio Z
9. Max out on 8-Spd Alfine

And then repeat this for five consecutive days while changing the order around or something similar.

Going back to the derailleur and Alfine will let us see if my performance is being reduced from exhaustion and we can take that into account when assessing efficiencies.

I don't know let's hear your ideas.
Last edited by seansimp925 on Tue Feb 14, 2012 8:50 pm, edited 1 time in total.

Re: Efficiency
by jakelewis on Wed Feb 29, 2012 7:45 pm

I'm not sure what you mean by 'Max Out', but I'd suggest that you try to maintain a steady aerobic heart rate (power input) for a longer period of time and measure distance traveled via a rearwheel bike computer (power output).

However, I fear that the combined effects of different tire treads/inflation, frame stiffness, chain age, varying blood sugar levels, ongoing tiredness etc may incur errors greater than the transmission loss differences you aim to measure.

Still, worth a try until someone puts the N360 in a proper dynamometer like http://www.ihpva.org/HParchive/PDF/hp52-2001.pdf and publishes the data.

Re: Efficiency
by jakelewis on Thu Mar 22, 2012 4:06 pm

I think there are two factors that combine to make the overall efficiency:

1) Mechanical efficiency. This efficiency value will vary from model to model, ie n360 , Alfine 8, etc. This can be accurately measured using a dynamometer which, unfortunately, most people don't have access to, but an approximate value can be gained from comparing heart rate over a set time against distance.


2) Physiological efficiency - the efficiency of the rider given that he/she may be riding at a sub-optimal cadence. This could be due to steps between gears (not an issue with CVTs) or because the bike speed exceeds the max or min gear ratio of the hub. Perhaps this can be measured by charting heart rates against power at a variety of cadences. The power output can be adjusted by altering the resistance of the training roller.

The total efficiency is these two numbers multiplied together.
- Oran
Répondre
#2
Wow! Thanks for the posts Oran!

I tried to rip a part of the data on the Nuvinci site but there is a "robot protection" on the forum... It has to be done manually I guess!
-
Si ça a déjà été fait, je peux le faire
Si ça n'a jamais été fait, donnez-moi juste le temps de trouver comment !


Répondre
#3
Damn... I just checked and they really closed the forum... It says a lot about the company Sad
-
Si ça a déjà été fait, je peux le faire
Si ça n'a jamais été fait, donnez-moi juste le temps de trouver comment !


Répondre


Atteindre :


Utilisateur(s) parcourant ce sujet : 1 visiteur(s)