Tuesday, May 23, 2017

Efficiency (3)

Efficiency (3)
Velomobile efficiency is one of the best reasons to own a velomobile.  This is the third post, the first was on aerodynamics, the second on rolling resistance and today I will cover weight and mechanical efficiency.  Each velomobile is different so the discussion has to be general in nature. 

I received many good comments on my previous post regarding rolling resistance and some pointed to items I did not mention.  First, it is clear that a tire/tube combination at a given pressure may work best in specific conditions but may not be optimal in all conditions.  There needs to be experimentation to find what is best in your riding environment.  All the measurements will provide some avenues to explore or give you a quick gauge for comparing but they do not provide a definitive comparison for your conditions.

Weight is an important characteristic of velomobile performance.  Weight has an impact on acceleration and ability to climb hills.  There are significant differences in the weight of commercial velomobiles, are in the 22-23kg range while others are in the 40 to 60kg range.  This is something to consider when choosing a velomobile because the rider has to move that weight.  Currently a racing velomobile can be in the 12 to 15kg range and we could see a sub 20kg production velomobile in the not too distant future.

Weight in cycling is very important.  Several years ago the UCI has mandated that road bikes should not be lighter than 6.8kg, this is to ensure that riders do not have an undue advantage when climbing or accelerating and it was also to ensure that manufacturers are not cutting corners affecting rider safety in the process of making a bike lighter.  Nowadays, with better techniques and materials bike manufacturers can easily make safe bicycles that are less than the 6.8kg limit and the UCI could easily revisit that rule to help increase the performance of cyclists.

As you can see the weight of a velomobile is significantly more than the weight of a good road bike. Even the lightest racing velomobiles mentioned above are nearly twice the weight of a good road bike.

When climbing up a hill, a production velomobile is at a significant disadvantage compared to a road bike.  A rider in the best production velomobile has to push the equivalent weight of more than three road bikes up a hill and it could be almost 8 road bikes for the not so light velomobiles.  Since there is no aerodynamic advantage on steep hills and the weight and rolling resistance of the 3 smaller wheels makes the effort more difficult for the velomobile.  As a result, the rider has to expend more energy to climb steep hills and the velomobile will not be able to match the speed of a road bike.

Fortunately, the added weight will be an advantage on descents and increase momentum and that is desirable when the road has a number of short rolling hills.  The energy stored will enable the experienced rider carry the momentum of the descent into the next climb almost effortlessly and at great speed while a road bike rider would need to climb each one.

Of course if one could reduce the weight of the velomobile, there could be important performance gains to be made.  Top velomobile designers will use their skills to choose the best materials to reduce weight and improve performance.  In choosing a velomobile many people will place the looks over other important characteristics such as the weight.  There are a few things the average rider can do to reduce the weight of a velomobile including some choices can be made at purchase time like the type of drivetrain.  But unless you are racing, some things like lighter wheels and tires for the velomobile may not provide a significant improvement since the resulting savings would only reduce the weight by a few percent of the overall weight of the velomobile, the same change would be three or more times as important for a road bike.  A velomobile used on public roads may encounter several hazards and velonauts should be careful when considering lighter components since they can potentially trade-off weight for sturdiness, it depends on the material used and the type of construction.  A light wheel for example could fold more easily in a pothole.

However rider who would like to improve performance can decide if they should be installing optional equipment like a sound system.  An even easier way to reduce weight is looking at what they carry in the velomobile.  I’m probably guilty of carrying too much as I like to be prepared for most eventualities.  Leaving at home the kitchen sink, personal anvil or other stuff seldom used could probably help you get better performance.  It is very likely that you carry a kg of stuff that is not really required.

Mechanical Efficiency
Mechanical is the last element of efficiency; power needs to be transferred from the pedals to the wheel with the smallest loss possible.  Most of this is determined by the design of the velomobile and components but for the most determined rider, it is possible to make some improvements.   I make my comments based on velomobiles with rear-wheel powered tadpole trike configuration because they are by far the most popular, however some but not all of these comments will apply to other models (delta or 4-wheel). 

Getting power to the wheel efficiently requires stiffness.  The crank, boom and swingarm are all elements that contribute significantly to stiffness but the monocoque shell or the frame, as the backbone, is the glue that holds everything together.  When pressure is applied on the pedals, any torsion or unwanted movement of any components will result in a loss of power transfer.  The unwanted movement can be a very small lateral movement of the boom or the swingarm but the loss will be noticeable.  The movement can be so small that you would not be able to see it while riding.  Well-designed components made of materials like carbon fiber instead of aluminum normally provide more stiffness.

People realising the issue have modified their velomobiles to increase stiffness.  Being the most popular velomobile, many changes were reported for the Quest in particular in order to address this issue.  Some of the changes include the replacement of the aluminum coat hanger used to provide stiffness in the turtledeck area and the point of attachment for the rear shock.  The replacement made of carbon fiber material creates an attachment point at the top of the wheel well.  A second change is the replacement of the aluminum swingarm with the newer carbon fiber version.  A third change is the installation of a carbon fiber pillar to give rigidity to the aluminum boom.  A number of riders also have added carbon fiber ribs in several areas of the shell.  Many of these improvements could also be made on different models of velomobile with similar results.  These changes also have the potential of adding extra weight with minimal stiffness improvements so one has to be careful in making this type of modifications by carefully researching, planning and implementing them.   In the case of plastics like the Rotovelo, while I cannot confirm this, I expect that the plastic material would be much less stiff and may contribute to a loss of performance that may be significant, especially in the case of the Rotovelo where there is virtually no additional metal frame.

In addition to stiffness, the drivetrain can be a source of power loss.  Chain line rubbing, unnecessary chain tube, damaged or missing idlers, bad alignment of chain line components are possible sources of power loss.

Suspension can also rob significant power; every push on the pedal can contract the rear suspension to some degree.  This is particularly noticeable when climbing a steep hill at low cadence.  There could be a significant bobbing effect.  Instead of using the power on the pedal to turn the wheel, part of the energy is used to compress the rear shock.  The problem can be larger when the rear shock is soft.  Adding more air in an air/oil shock or installing stiffer springs or polymer in a mechanical shock could improve efficiency.
 Bearings play an important role in the velomobile efficiency.  Since all bearings are not created equal, for the same size bearing, there are several categories, some have looser tolerances and there is a bit more movement inside so they do not roll as well as others, I would call them cheap but there are others that are of tighter tolerance and are more efficient.  In the United States there are 5 classes of bearings (ABEC1, 3, 5, 7, 9 with 9 being the best) but since bearings are made in many countries, they may use another standard such as the ISO and DIN that have similar levels, see table below. Bearing classes explained
ANSI Standard 20
ISO 492
DIN 620
Class Normal
Class 6
Class 5
Class 4
Class 2

The different classes take into consideration several factors that will affect performance internal clearance, surface finish, ball accuracy, torque, noise, cage type, and lubrication Some racers use the lowest friction bearings to lower resistance and improve performance but the gain could be small relative to the increased cost for those bearings.  A 406 wheel equipped with 28mm tires will turn almost 10 times a second at 50km/h and small movement could make you lose significant power but the loss is even greater if those same bearings are worn or damaged.

It may surprise some people but bearings in the front wheels can be subjected to intense heat generated by hub brakes.  Excessive heat can also contribute to bearing failure.  Repeated long descents under load may cause bearings to fail.

There are bearings in wheels, pedals, bottom bracket, cassette and some can be changed some cannot because they are sealed inside the unit and you may have to change the whole unit (cassette, pedal, bottom bracket) if they are worn or damaged.  Idlers could have bearings or bushings and the bearings could be sealed too.  A damaged bearing or bushing will typically have one of many of the following unwanted movement (lateral or perpendicular), difficulty turning, noise (grinding, rubbing) or other damage like missing or damaged seal.

A key aspect of power transfer is the drivetrain. There are many different types and all are not created equal.  Velomobile drivetrains borrow elements of several bicycle systems but they are different in several ways from traditional bicycles.  First difference is the chain length; velomobiles have the equivalent of 3 standard lengths of chain.  To function properly the drivetrain also needs idlers, tensioners, chain tubes to transfer the power from the pedals to the wheels, something only recumbent bikes use.  Second, because velomobiles have higher top speed than unfaired bikes, velomobiles need higher gearing but due to their higher weight, they also need lower gearing than most road bikes to climb steep hills.  As a result, the required gearing range is much greater than the gearing used by other bikes.  This means that components are often near or at the maximum capacity.  To meet the requirements, velomobile designers will mix and match different components made for different types of bikes (grupo) for example road and mountain bikes and even urban bikes.  They have larger chain rings some pushing 75 teeth, wider range cassette e.g.:  11-36 or more. 

Most often, high capacity long cage rear derailleurs will be chosen to handle the range.  I expect that designers will soon take advantage of the 11 and 12 speed rear derailleur developed for mountain biking with wide range gearing and install them in velomobiles soon.  It may not be a slam-dunk for velomobile applications there are several potential issues.  First, since the internal space required for the extra long cage derailleurs and the smaller chain may create some issues.  It is also not clear if the rear derailleurs designed to be used for single ring applications could handle the extra chain required for a 2 X11 or 2X12 configuration especially if the big and small ring has a big difference in the size of the rings e.g.: 62-34. 

When operating at or near the limit of components that were not designed together, one has to be careful that efficiency is not affected significantly.  Some of the new wide range cassettes may not fit standard hubs but if they did they may create issues.  There are cassettes with 9 and 10 teeth that are used to provide the speed for mountain bikes racing down hills or for trikes with smaller wheels.  In their applications they could be fine because these gears are only used seldom going down a hill for example but they are highly inefficient.  In velomobiles, we will tend to use the higher gearing even on flat ground and for much longer periods of time.  The angle of the tooth spacing increases and the chain does not sit in the grove very well; to illustrate this at the extreme, imagine a 4-tooth cog, the teeth would be spaced 90 deg.  The chain would get stuck or skip when you would try to turn the crank.  I have a bike with a Capreo hub with a cassette that has a 10-tooth cog and I find it awkward to pedal and I can feel the inefficiency.  Some road racers don’t even use 11-tooth cog on their cassette because they feel they are inefficient but I think they are OK in velomobiles.  As a side note, it is not certain that cassettes with 9 and 10-tooth cogs could be fitted on to velomobile axles as they are typically require installation onto smaller diameter hubs.

Velomobile with smaller rear wheels (20in/406 or 16in/349) at the rear will require higher gearing, to compensate because the smaller wheel means that it has to turn significantly more than a 26in wheel to achieve the same speed.  To address this issue, manufacturers will often install a mid drive; a secondary gearing that can be as simple as a 3-speed cassette and derailleur to Internal Geared Hubs (IGH) like a Rohloff.  Mid-drives add some amount of drag but can provide a much larger gear range.

As mentioned above, it is possible to use other gearing system like the IGH. There are different make and model of IGH with their pros and cons, there are also a few pedal-based internal gearbox and even hybrid systems that have an IGH coupled with a cassette and there is also crank based systems like the Schlumpf that multiply your gearing.  They each have significant advantages including providing a wide range of gears, some have the ability to change gears when at a standstill and they have low maintenance.  Unfortunately, some of those are noisy and can be heavy and may have torque limitations that can be exceeded by velomobiles under certain conditions and can’t be shifted under load.  As far as efficiency, some reports indicate that there is a 6% penalty for IGH compared to derailleur systems.  This can be significant for a performance-oriented rider and something to think about when making your choice for gearing.

There are other factors that could create power loss. In my previous post on rolling resistance, I mentioned ensuring proper alignment contributes to mechanical efficiency but so is truing of the wheels.  A wobbly wheel will sap your power because you will need to push harder to maintain your speed. While more rare, brake adjustment could be over-tightened and touch ever so slightly when the brake handle is released and this is another thing that could sap your power.  Brake adjustment should be part of proper regular maintenance. In short, if it rubs, squeaks or grinds it probably robs your power and needs to be fixed.

Lastly for a good velomobile, you also need efficient brakes to keep velomobile rider safe.  Most velomobile use drum brakes on the front wheels.  There are 70mm and 90mm drums and both will stop a velomobile efficiently on flat ground but when the road gets hilly, 70mm brakes will probably be unable to meet the challenge, the brakes will quickly overheat.  The larger 90mm version will fare better but they will eventually overheat and lose their braking efficiency.  There are 90mm drum brakes with added fins to cool brakes faster and some riders have designed simple water injection system to further cool brakes.  Riders in mountainous areas should consider these enhancements.  There are some velomobiles equipped with disk brakes that may offer better braking in some conditions but they require more maintenance and many have switched to drum brakes.

Through this three part series on velomobile efficiency I think that I have covered most elements affecting efficiency and performance. While I’ve touched on many things, the information is fairly superficial and there is much more to investigate to deepen your knowledge; an ocean wide but a foot deep.  There is still a lot to be researched and I would hope those who can add apiece to the puzzle would publish their findings through public forums so everyone can learn.  I hope that these post provided you a glimpse into what makes a velomobile efficient.


A new pedal-based gearbox designed for mountain biking is now available. EFFIGEAR's has a gear ratio of 444% with 9-speed providing a similar range as derailleur-based, double chainring 24/36t with an 11/34t cassette.  This system provides the ability to shift under load contrary to other IGH and gearbox systems.  The gearbox system weighs 2.8kg including cranks and shifters.  The gearbox can also be used with a rear derailleur to further increase the range.  Unfortunately to install in a velomobile would require the design of a new crank mount.  This system is similar to the Pinion gearbox that offers several systems from 9 gear 568% ratio, 12 gear 600% ratio to an18 gears and 636% ratio.



If you did not have a chance to attend SPEZI this year, there are several who reported on the annual show including The Laiback Report with an excellent Spezi video tour and interviews.  . Wim Schermer reports on SPEZI on his blog  and Ligfiets reports on several items of interest


  1. Dear Luc,

    I also carry lots of spare stuff in my velomobile, but this is mainly necessary to be able to do roadside repairs. A velomobile cannot so easily be transported home by normal car, so anything that can be fixed by the roadside is a big plus.

    That said, I much enjoyed this efficiency overview, so thanks for that!


    Erwin and Tante Lies (she's my Strada velomobile, named after my great-aunt)

  2. You probably went over most if not all resistances and factors now, your part 3 is impressive :-)

    Remains what tradeoff must be made for it's intended use, as the fastest most resistance efficient velomobile will be a very low, small and sleek racing model that is all but worthless for practical everyday use ! Much like a racing bike.

    Rider comfort, luggage/spare parts/repair tools carrying space will decrease efficiency yet make it a better vehicle to travel longer distances.

    The (more then a decennia old model) Quest was designed for comfort and is able to carry spare parts and tools, and even complete light camping equipment. That means it can never be optimum efficient but was nonetheless the (only) fastest model for many years and still breathing in the neck of modern racing models at competitions that often lack comfort or luggage carrying ability. Very impressive. It seems like a very balanced design, however aging it is.

    More modern velomobiles seem designed for more speed and somewhat less comfort and carrying space capacity. Yet not be spartan racing bikes only. But tradeoffs must be made.

    The guys that produce the Quest and Strada said designing a non compromising racing velomobile would be easy. It's the tradeoffs and compromises that make a velomobile usable that are hard to get just right.

    So efficiency, but at the cost of what ? There are always costs :-))

    1. Yes you are right, unless you are racing and have the lightest components, use a hood, plug every hole and crack, install very light and fast tires and leave with no spares,etc. you will not use achieve the best efficiency for your velomobile. My points were many:
      First when you make your choice of velomobile and components, you should realize that velomobiles are not created equal, what you buy may be a velomobile but it may not be efficient and no matter what you do, how strong a cyclist you are don't expect the great speed you see on the Internet if you picked form over function;
      Second, even fast velomobile will need tuning if you want to get the most of it;
      Third, be careful of what you you install and carry from big lights and cameras that stick out and significantly rob you of your aerodynamic features to carrying too many tools like a floor pump because it is easier of a crank puller just in case or even a sound system that weights one kilo because you want to blast your music.

      The next thing the person thinks of adding an e-assist because it is too hard to get that velomobile up small hills

      At the end of the day, all these choices add-up and maybe taking a few minutes to a few days to look at your choices and see if you can improve things.

    2. Still "blasting your music" could deliver many more watts then nominal and negate the effect increased weight has on rolling resistance and gravity pull when climbing :-)))

      After all much is "between the ears" when it comes to performance....ever seen Theo van Andel ( one of the vm.nl producers ) race in a Quest ?

      Him being hot on the tail of much faster designs and sometimes exceeding 70 km/u or 44 mph is not because the aging and "fat" Quest design is particularly fast, even though it is a special very light built racing model, it is mainly because the guy in it is exceptionally strong and highly motivated :-)))

      I know you set out to address resistances working against the rider, not variable generated power that depends on the rider. But it is a fact that even something with the "streamline" of a bus stop shelter can go fast if it has enough power to work against the resistances :-)))

    3. It is my understanding that Theo has indeed done many of the things I mentioned to improve efficiency but he is limited somewhat by the basic design.

      I think you would agree that if the same rider were to ride in a lighter, stiffer and more aerodynamic velomobile, he/she would have much better results.

    4. Oh yes, anything that absorbs energy that is not going to drive the vehicle forward is wasted.

      Yet if that is to result in a extremely stiff velomobile without any suspension as that absorbs energy too one should not be very surprised the result is very bouncy (micro wheel slip too because of that) uncomfortable and wears out the driver due to continuous vibrations and bouncing.

      For such a vehicle to be of any use it needs a very smooth surface to ride on. Real world roads are not often like that. So there need to be compromises and tradeoffs, and the results are the velomobiles you can buy.

      You can tune them somewhat but with that you will make your own tradeoffs and compromises. The skirts on the DF are a good example of that. Great for races. Not so great for everyday 90 degree turns...Quests will take wheel fairings. Good for races, but they will touch the real world roads in too many situations and collect dirt. All velomobiles take hoods....but they are not so great in adverse weather ( oddly enough ) due to poor vision. I crashed my Quest because of a damp visor...

      That list goes on...but it does explain why we get consumer velomobiles the way they are. They are the best they can be without going into things that adversly affect their everyday use. Unless they are pure racing vehicles.

  3. When considering weight do not forget the rider. He is usually 3, 4 or 5 times heavier than the bare vehicle and a little effort, or a season or so of riding, can produce more weight saving than spending much money on lighter equipment... :)
    I've been there and had 10kg to drop... ;-)

    1. That would be true for any bike, many of us could lose some weight. We could also say that a rider in shape will provide better performance and combined with weight loss would increase power to weight, however, it is beyond the objective I had set for my posts.

  4. Weight can also hide in unexpected places...
    I was looking at a new pair of pedals and looking at options to save 50g at a cost of $50.
    Thinking that I needed to allow for the cleats I decided to weigh my riding shoes. 1.5kilo!!! Shopping through 'last years colours' I found some racing shoes for $100 that saved me a full kilo.

  5. I am wondering what your thoughts might be on this article about planing and flexing and how it may relate to velomobiles.

    1. Hi Brad: This is interesting but I think there are several differences between a velomobile and a road bike. First, one would have to compare the stiffness of a velomobile vs a road bike. The swingarm in most velomobile is single sided and I suspect the rear wheel moves significantly more under load than a regular bike. Second, additional flex in a tadpole velomobile IMO accentuate side to side movement instead of propelling the velomobile forward. Third, there may be an optimal level of flex/stiffness for bikes and velomobile but while road bike bike manufacturers may have to dial back stiffness, I doubt that velomobiles have passed the level of stiffness that requires dialing back. Even the best streamliners at Battle Mountain are looking to increase stiffness to improve their performance.