I’ve been running a tubeless rear wheel with Airliner on both velomobiles for a couple of months now. They’ve been working well but I haven’t had any punctures during that period of time. I’ve heard stories of Airliners shriveling up and of the liquid sealant drying up. I think that these cases might have been due to the use of a non-Vittoria sealant. I decided that it would be a good time to open up a tire to check on the state of the Airliner and sealant. I removed the Milan rear wheel and opened one side of the tire to have a look. The short story is that the Vittoria sealant was still in the liquid state and well distributed. The Airliner looked exactly the same as when I installed it. That’s the good news.
The bad news is how difficult it was to do the inspection. I was able to break the bead away from the rim pretty easily using the Vittoria tool. But stretching one side of the tire over the rim wall to remove the Airliner took a couple of hours of breaking tire levers and swearing. Then reinstalling the Airliner and tire took a couple more hours … and I was never able to stretch the tire back onto the rim.
Since I wasn’t able to re-install the tire with Airliner, I had to come up with a plan for a rear wheel setup on the Milan. After struggling for hours with the tubeless tire – Airliner set up in my garage with access to all of my tools, I realized that it would be impossible to deal with it on the road. The Airliner isn’t the solution that I had hoped it would be. I could just replace the Airliner wheel with my original non-tubeless rear wheel. However, I decided to install the tubeless tire without the Airliner. I was surprised at how easy it was install the tire without the Airliner.
I realize that I’m giving up the safety benefits of the Airliner with this decision. However, I do think that the use of a standard tubeless tire / rim may be safer than my previous non-tubeless setup. I believe that a tubeless tire is more likely to stay attached to the rim when a flat occurs.
One of the things velomobile riders fear most is a high speed roll over. Anything that causes the rear tire to lose grip at high speed can lead to a roll over. Riding with the rear tire on a roadside rumble strip can induce a roll over. Here’s an example of that:
Another common cause of a velomobile roll over is a rear tire flat. It seems to be even worse if the tire breaks free of the rim. Vittoria has come out with an interesting product, called the Airliner, that could bring a new level of safety to the rear flat situation. The Airliner is a light weight foam ring inserted into a tubeless tire/rim that compresses to a fraction of its volume when the tire is inflated, then expands to fill the tire if the air escapes the tire. Under these circumstances, the tire behaves as though it has a pressure of 25-30 psi. This should be enough to keep the tire on the rim. The fact that it is compresses out of the way when the tire is inflated makes the tire behave like a normal tubeless tire. You can read about it here… or watch this video.
I decided to give Airliner a try. This is my first time working with tubeless technology. The tubeless wheel consists of a tire with a special bead that mates to a rim with a compatible tubeless profile. The rim is taped with a special tape. Also a special valve stem is required. After all of that is put together a small volume of a sealant liquid is introduced into the air space of the tire. Once the tire is seated on the rim, the bead locks into the rim’s hook profile, making it very difficult to separate the tire from the rim. There’s a good overview of tubeless technology here…
Rear Wheel Build No. 1
Neither of my velomobiles had a rear wheel with the required tubeless rim so I set out to build one. I found a source for a new DT Swiss 559 tubeless rim on eBay. I had a conventional rear hub and a wide selection of spokes in the drawer. Since the Milan accepts a conventional rear hub, it was chosen to be the test mule.
This first wheel didn’t work out as I had hoped. I built the wheel with the DT Swiss tubeless rim and mounted the Schwalbe Pro One tubeless tire containing the Airliner. Surprisingly, the Airliner Road version fit the 559 tire without cutting it even though it’s sized for a 700c wheel. I applied soapy water to the bead before inflating it with an air compressor. The tire seated on the rim and held air just fine. However, when I then deflated the tire, simulating a flat, the tire’s bead separated from the rim, making it possible for the tire to escape the rim. This combination of tubeless tire and rim were not going to work.
Rear Wheel Build No. 2
Next I bought a 559 Alex tubeless rim on line at Universal Cycles. I disassembled the wheel with the DT Swiss rim and re-assembled it using the Alex rim. I mounted the same tubeless tire with the Airliner and soapy water and filled it using the compressor. The tire’s bead seated perfectly on the rim. This time the tire remained firmly locked to the rim when I deflated the tire. So I now felt confident enough to inject about 30cc of liquid Vittoria sealant through the valve stem with the valve core removed.
Testing the Airliner at zero pressure
I inflated the tire to 85 psi then mounted the wheel in the Milan and set off for a ride. At this pressure, the Milan’s handling felt normal. I stopped near a bike path and deflated the tire. The Airliner expanded within the tire to give it enough shape that the rim didn’t contact the pavement. So far so good. I set off for a ride with zero pressure. Vittoria says that the Airliner alone should provide the equivalent of 25-30 psi of air pressure. That’s what it felt like to me. The rear end of the Milan felt a bit squishy of course but it was manageable. I rode for about a mile on a combination of smooth bike path and rough pavement on an adjacent road. I kept the speed below 15 mph. I took some turns at a reduced but reasonable speed just to complete the test. The tire remained locked to the rim as I had hoped. I called that a big win.
Testing the Airliner’s Effect on Performance
There’s a hill near my house that I have used to do coast down tests to measure maximum speed with an onboard wired cycle computer. I do these tests to compare the effects of modifications and adjustments that I make to the velomobiles. I was curious to see if the presence of the Airliner would adversely effect the maximum speed of the Milan. I did a couple of runs with the original Pro One/latex tube wheel that I normally use and a couple of runs with the Pro One/Airliner wheel. All of the runs showed the same maximum speed within a few tenths of a mile per hour.
The Airliner performed as advertised once I got the combination of rim and tire correct. The fact that I was able to ride a mile on zero air pressure in the rear tire without the tire leaving the rim convinced me that the Airliner will work in the event of a normal flat. I’m also convinced that there is no performance penalty to pay running the Airliner.
On the negative side… I wasn’t able to simulate a blowout so I don’t know how well the Airliner will perform in that case. Another potential concern is that I will have a difficult, almost impossible, task of removing the tire from the rim to repair a flat or replace a damaged tire while on the road.
I’m convinced that the extra precaution against a roll over makes the use of the Airliner worth while. I will leave it on the Milan and build a tubeless rear wheel for the DF to run an airliner on it.
I’m still having fun building parts for the Milan SL. My last piece was the head out lid and shade cap. After building these pieces, I realized that I should be able to stow the cap inside the Milan. The single piece cap assembled with its pillars doesn’t fit in the available space in the Milan. Stealing some ideas from Wim Schermer’s splittable Quest hood, I came up with a quick way to remove the pillars and split the cap into 2 pieces without tools.
Here is the unpainted hood – split down the middle. At this stage, it still had the pillars mounted with screws.
One of the problems I had to solve was to come up with a quick way to join the halves without using tools. I settled on using 3M Dual Lock. The idea was to fabricate a flange or lip on the underside of one half of the cap, for the full length of the split. The lip would have enough depth to allow 2 thicknesses of Dual Lock. Here’s a cross section to show what I mean.
Laying up the flange didn’t sound like a lot of fun so I came up with an easier but less optimal way to accomplish the same thing by 3D printing something similar. Rather that printing a flange for the whole length of the split and trying to get the curvature right. I decided to print some 1″ long sections of flange, using carbon fiber/PETG filament, and epoxy them on one half of the cap along the split. Here’s the design of the piece. Also shown is the underside of the cap.
The next item to address was getting rid of the screws holding the pillars to the cap. I ended up using a slightly re-designed flange piece to attach the pillars using Dual Lock. Lastly, I wanted to get rid or the 2 allen screws on the clamps holding the lower pillars to the lid. I redesigned the clamp to use only a single M6 thumb screw. Here are some photos…
There’s a problem with the approach I’ve taken though. The split is not entirely sealed against the elements since there lip doesn’t run the whole length of the cap. If, over time, I find that this splittable cap is worth keeping, I’ll replace the one inch flange pieces with a full length carbon fiber flange.
After finishing the head out lid for the Milan SL, I thought about those sunny days that I need the added air flow of the lid, but also want some shade. I looked back at the shade cap that I made for my WAW. It worked well and I liked the way that it looked. So I decided that I should make a similar removable shade cap to use with the head out lid of the Milan.
Here’s my WAW shade cap from a few years ago. The WAW has some nice clamps installed on the inside of the lid near the cockpit edge. These were great for mounting the clear plexiglas posts shown below. It only took a few seconds to loosen the clamps to remove the cap.
I already had a scrap piece of a hood from my Milan’s one piece hood mold. I just needed to come up with the plexiglas posts and some clamps similar to those used in the WAW.
Here is the Milan shade cap. Notice in the head on view below, how much open space there is for air to pass through.
Here’s the clamp that I designed to mount the clear plexiglass posts. I printed them with carbon fiber / nylon filament and embedded metal M5 threaded inserts.
By the way… I’m aware that the Milan’s aerodynamics are most likely compromised with this cap, but its only purpose to keep me comfortable on warm days.
The stock Milan opening is enclosed by 2 pieces – the lid which covers most of the opening and a small hood which attaches to the lid.
It’s nice to be able to ride without the hood on hot days. To do so, you remove the hood from the lid and ride with just the lid.
The problem that I have with riding head out with this setup is that the opening in the lid is so small that the front of the opening feels like it’s right in my face. So I decided to make my own lid to be used exclusively for head out riding. It provides a larger opening for a less confined feeling.
To build the lid, I started by laying up a hood from my own Milan race hood mold. I cut away the overhead portion of the hood, leaving a larger opening than the stock lid’s opening. There is also more of a lip on the front edge of the opening, forming a bit of a windscreen. Notice that I’ve added the same NACA duct and mirror covers as I used on my full hood.
The lid is held in place by the usual Milan cable housing hinge at the front. Also at the front is a single bungee that extends from the bridge up to a hook on the underside of the lid. Magnets are attached at the back 2 corners of the lid to mate up with magnets taped to the inside of the body shell.
I’m used to riding with the full hood, so riding head out is quite different. A lot more air flows past my face and around my neck and shoulders. That is the point of building this lid so this is a good thing. Vision out front and to the sides is excellent. The amount of air flowing through the NACA duct is not as noticeable with the open lid as it is with the full hood. I think that it probably isn’t worth the effort to add the NACA duct in any future lids that I may make.
As for performance… I didn’t take any time to do any actual testing. I’m not able to feel a slow down due to adverse change in the aerodynamics, but I’ll assume that the lid will slow me down a bit. Overall, I’m pleased with the effort. The fit of the lid is decent and the added airflow will be appreciated on warm days.
I’ve been tinkering with my Milan hood, making some minor improvements so I thought I’d do a quick update.
Closable flush NACA duct flap
I really like having a closable NACA duct on my DF – enough so that I’ve added one to my Milan SL.
The flap is adjusted using a lever mechanism similar to the one I used on my DF NACA duct.
Side Window moldings
I 3D printed these moldings in flexible TPU filament to hold the side windows in place. Formerly the side windows were just hot glued to the interior surface of the hood. It looked pretty messy.
Added Snoek-inspired Mirrors
I originally came up with these for the DF. They actually fit the Milan hood better on the flat sides of the knee humps. These mirrors replace the smaller Spin mirrors that were located closer to the visor where the plastic white plugs now fill the holes.
I’ve used various tools over the years to adjust toe in on my race cars, trikes and velomobiles. Mostly, I’ve used various forms of trammels like this nice one built for me by my friend Bill (A2naut on Bentrider).
or this gizmo that I put together using 80-20 T-slot profiles.
There are a couple of problems using these on velomobiles. You have to slide them under the velomobile which has minimal ground clearance. With the trammel, you’ve got to keep sliding the tool in and out to measure in front of and behind the wheesl. The other problem is that some velomobiles have enclosed front wheels so very little of the wheel is exposed below the bodywork.
So I looked at various commercially available toe-in gauges to get some ideas. Most are meant for use with race cars and involve lasers or measuring against an elevated string rectangle. Then I came across a simple type of product called Toe Plates. These are just 2 plates that are bolted onto the front wheels of the car. They use simple tape measures across the plates in front of and behind the wheels to determine the actual toe setting. These wouldn’t work with the velomobiles for a lot of reasons, but I liked the idea.
Looking around the garage, I found some 8020 T-slot profiles and some 8020 90 deg threaded nubs. Using these pieces, I concocted this low slung toe gauge. It’s similar to one that I had seen on The Velomobile Observer blog a long time ago. Each bar is placed along a front wheel. The 2 nubs on each bar hold M8 bolts which are screwed into the nub at the same length to form pointers to be pressed against the rim. Using these pointers eliminates any interference from the tires. The ends of the bars are pulled together by the bungee straps that run under the velomobile. These are adjusted with just enough tension to keep the pointers pressed against the rims. The tape measures are used to measure the distance across the bars at the same distance in front of and behind the wheels. So for example, if the wheels are pointed straight ahead, the bars would be parallel and the measurements would be identical and so on. The 4 red plastic pieces are just supports to keep the bars flat on the ground. I’ll probably replace those with some 8020 aluminum pieces that would be more robust.
Here you see the tool used on the Milan SL with its enclosed front wheels.
I used the tool today to check the toe-in on the DF and the Milan. I was pretty happy with the ease of use and the repeatability of the measurements. The 80-20 products are very precisely machined so once I had all of the pieces bolted together, I had a tool with no slop and very straight edges.
I’ve decided not to update the steering plates on the DF to the newer geometry plates. These plates really improve the straight line stability in cross winds. However, they increase the turning radius noticeably.
Recently, I was riding in some gusty winds and was reminded how easily the DF can be pushed around. Even though I’m not crazy about the aesthetics, I decided to add a tail extension. I had added one to my DFXL in the past so I knew how well it works. Here is the DF with its new tail extension:
I’ve since added a column of red LEDs along the trailing edge behind a cast resin lens and some single yellow LED turn indicators.
This is not exciting news but it may be of interest to a few people. I was looking at a photo of a DF seat mount built by Garrie Hill, when I thought that I saw a bag fitted to the back side of the DF seat. It turned out that it was not a bag. But I liked the idea of using that space for a lightweight bag.
This is the bag that I came up with. The material is a nylon mesh fabric that I bought on eBay. I don’t have a proper sewing machine so it’s put together with contact cement and held in place with velcro. It’s about 9 inches wide at the top, 12 inches wide at the bottom and 14″ from top to bottom.
When I sold my red DFXL, I included my shade cap in the sale. It was a copy that I made of the factory part sold by ICB. It worked OK, but I never really liked the way it mounted with the 2 straight struts holding up the front (shown below).
With spring around the corner, I need to replace that shade cap. But I want something that looks a little better than my previous shade cap. So I came up with a hood, popped out of my aero hood mold, that has a little more structure. It maintains the sections that interface with the cockpit so no straight struts are needed. I opened up the visor area and side windows quite a bit to promote airflow. The sides and visor area openings are contiguous so the cap area looks like it’s floating without any structure holding it up. It did require some interior bracing to stabilize the cap area. (By the way… I’m not sure I like the white cap paint job…)
I haven’t tried the cap out in warm weather yet. I think that there’s enough open area to allow air in, but may need some exit vents to allow that air to flow through the cockpit.
Update: After a short test, I noticed that as the speed increases, less air enters the cockpit through the front NACA duct. This may be due to chaotic air flow inside the cockpit. With that in mind, I’ve added a new air exit duct to see if that can improve the air flow out of the cockpit.