Buyers Guide Forums Information Buy/Sell Communities Build/Tech Racing Site
John Tetz: Trike Foamshell Velomobile p2
What's New

About

Rides

Mailing List

Members

Photos

Projects

Library/Misc.

MARS Gear

Links

John Tetz: Trike Foamshell Velomobile, p.2

page 1
page 3


Another design change I made from the traditional tadpole trike was to raise the bottom bracket height so the heels can clear the bottom of the shell. The shell ground clearance is dictated by the lowest part of the trike frame. The holes for the feet are behind the cross tube further back from the nose. This helps maintain clean airflow along a longer distance. The airflow might even be laminar for a part of that distance because of the shape.

I like a BB to seat height of about 9 inches, so most of my vehicles have this ratio. My leg muscles don’t have the readjust going from one vehicle to the next.

In order to see over the top of my toes, the seat back is around 45 degrees (which I happen to like for a road vehicle - gives the ability to turn your head around to check for traffic).

The tie rods are in front of the king pins, rather than in the more normal rear position (except for Greenspeeds). This is to get the tie rods away from your feet when climbing on/off. My thanks to Peter Eland; I would not have been able to find this position without his trike steering spreadsheets:

www.eland.org.uk/steering.html

The front wheels lean in at the top to give the upper part of the shell a finer shape, reducing the frontal area. The wheels are flush with the sides of the shell. Eventually I will build fender spats covering the top half of the wheels to help the airflow across this part of the shell. The wheel spats will also reduce inside spray on wet days.

The main requirement of this vehicle - being light enough to stay within the power capabilities of more average riders - means that some of the wish list of car-like add-on amenities will be left off, such as interior weather sealing around the wheels. Without interior wheel sealing, the maximum steering angle is larger and the turning radius is shorter, often needed in town. This trike can make a 180 degree turn in about a 16 ft/4.8 m diameter.

However, I find that in cold weather the crosswind goes through the wheels and onto the rider. Even with wheel discs, the air flows inside the shell around the clearance between the tire and shell, so an interior sealed area is needed in the winter (removable in summer?).

Front suspension: I eventually developed a compact head tube suspension system. The ride was simply too rough without it since a trike is bound to hit more bumps and holes than a bike. Because the foamshell is not as stiff as a composite shell, it jiggled annoyingly left and right over bumps (no such left/right jiggle problems on a bike streamliner). Without the shell, the ride didn’t seem as rough, but inside the shell the jiggle gave the perception the ride was rougher. Because most of a trike frame is low, the upper shell support system is less stiff, which aggravates the jiggle problem.



As you can see in the photo, the head tube is a bit larger in diameter so a spring can fit inside. The internal parts layout consists of a lower bronze bushing, allowing the kingpin to slide up and down, plus turn. At the top there is a ball bearing with a center bronze bushing, allowing the kingpin to turn and slide. A ball bearing takes the vertical suspension load of the vehicle plus a bit of the side load from the angled head tube. At this high load position the bearing reduces steering stiction more than a simple bushing would, so that smooth micro-steering corrections can be made.

On rare occasions the system bottoms out on very deep holes, so I use a small external rubber bumper. A smaller diameter spring can also be placed inside the larger main spring. This spring can be used for heavier riders.

There is no damping. Between the small suspension movement of +/- 0.5 inches and the drag on the bronze bushing, I have not experienced any wheel hopping. This is a relatively lightweight suspension - it adds just over 1 pound total for the pair over a non-suspended head tube system.

You will also note in the photo that the tie rods are below the center of the main frame and go up to the king pin steering arms at an angle. This turns out to be ideal in the fact that they are completely out of the way when the rider is getting on or off the trike. On the non-suspended version, the tie rods were above the cross tube.

But the main reason for this important tie rod angle is to reduce toe changes vs suspension movement. A more vertical head tube would be ideal. This would reduce the amount of horizontal wheel movement during suspension travel. Because the disc brake rotor would come close to running into the head tube at maximum suspension compression, this is the steepest that the head tube can be. Drum brakes would allow a steeper angle. The head tube angle also has to be steep enough for the rider’s legs clear the top edge of the head tubes (which are a bit higher than regular head tubes), otherwise the track width would have to be increased. I wouldn’t want to go wider because at 29 inches max width at the wheel hub centers, the trike clears most doorways. Present track is 27.5 inches.

All the effort to design and build this head tube suspension pays off - it works very well. The ride is noticeably plusher. It takes out those hard hits, plus it reduces the side-to-side trike rotation when individual wheels hit bumps or holes. With these changes I’m pleasantly amazed that the handlebars do not vibrate across bumps such as railroad tracks anymore. I find I have stopped spending time looking carefully at the road surfaces in an attempt to minimize the road shocks (what a relief). Suspension is definitely worth the extra 1 pound.

Having fat tires and running them softer helps take out the high frequency vibrations over marbled road surfaces, yet the Crr doesn’t climb enough to be a big problem. Measuring Crr vs tire pressure on fat Comets (done on the non-suspended trike) results in a Crr of .0081 at 85 psi, .0083 at 70 psi, and .0089 at 55 psi. These tests were done on a relatively smooth blacktop surface at an air temperature of 85 degrees F. Temperature has a large effect on Crr; see my write-up on Crr vs Temperature:

www.recumbents.com/mars/pages/proj/tetz/other/Crr.html

www.recumbents.com/mars/pages/proj/tetz/other/Crrp2.html

I find that measured rolling resistance of fat tires to be as low, and quite often lower, than some narrow tires. Narrow tires have to be run at higher pressure, which means a harsher ride even with suspension. Tire loads when cornering are very high on trikes, another reason for fat tires. They seem to complain less than the narrow tires, and are also less prone to pinch flats.

However this type of suspension does have its own quirks. If the wheels are not rotating, such as when getting off the vehicle, the suspension doesn’t slide until the rider’s weight is off (so the wheels can move outward). This results in a surge of frame motion of around 0.4 inch. With the wheels rotating, the suspension and tires can easily move in or outward smoothly without that surge. The amount of horizontal movement (+/- 3/16 inch max suspension movement) does not seem to affect Crr noticeably, but I have not made rough road Crr measurements to verify this. This vehicle does roll right along - it out-coasts the few commercial trikes I have compared it to - but this is partially due to the straight-out arm position.



I used a lightweight small movement (3/4 inch) rear wheel suspension unit with rubber bumpers. There is no damping here and less swing arm bearing drag, so on some occasions the rear wheel can momentarily hop. I’m not sure what to do about this - might have to investigate a different suspension unit. I probably made the system too light, so it’s flexing. I hate to keep adding a pound here and a pound there; it easily adds up to a heavy vehicle.

Yet this suspension unit does a reasonably good job of taking out the hard hits, hits that would otherwise go directly into the back of the hardshell seat, and therefore into the rider. Most uncomfortable. This particular layout allows a structural support (carbon rod) to a lightweight (all-carbon) luggage rack, to which the tail of the shell is attached, so its stability is quite important. And yes, I do get a bit of suspension at very high pedal pressures, which my body doesn’t allow me to do very long. If you are using only the push muscles, then the suspension pogos a bit, eliminated by round pedaling.



Ideally, the shell and the vehicle should be designed together. All of the above design requirements for the trike have to be considered simultaneously along with the design considerations of the shell - challenging compromises. The shape was chosen for a low CdA. Many of my errands require going some distance, so reduced rider effort and decent speed are two of the requirements high on the list. This results in a smaller cargo area than on other velomobiles. However this satisfies my cargo needs. Total vehicle weight is a bit over 40 pounds (trike 33 pounds, shell about 7 pounds). CdA is around 1 sq ft.

See my Zotefoam Manual on how to fabricate the shell; all new techniques were developed, using only a male mold:

www.recumbents.com/mars/pages/proj/tetz/manual/0intro.html

The basic foil shape (top view) is dictated mostly by the length of the nose to the wheels at the hubs, the widest part of the shell, then a slow contraction to the shoulder clearance, and finally, a straight rate of contraction to the end of the tail. The tail is cut off to house a taillight and reflector material, and also to reduce overall length for parking and transporting – and a small reduction in weight. However, the short tail I believe aggravates crosswind instability. The popular feeling is the tail area should balance the nose area, hard to do with a long nose (wheels far back from the nose).

Side view: The toe/heel clearances and location of the trike within the shell have a large influence on the shape of the nose. The top view nose length affects the shape and rate of upsweep to clear the toes/heels. A compromise has to be made between the top view and side view nose shapes. The side view is surprisingly close to my VFS/Vacuum Foamshell with its upswept nose (its front wheel is further forward and totally inside the shell):

www.recumbents.com/mars/pages/proj/
tetz/VFS/projtetzVFS12finished.html

www.recumbents.com/mars/pages/proj/
tetz/VFS/projtetzVFS00intro.html


This upswept nose has been used on many vehicles, including Varnas, super-mileage vehicles, solar cars, and others. The theory is that the space between the shell bottom and ground forms a slot which accelerates the air, creating negative lift. Any lift is induced drag. And the much flatter bottom area of a trike shell aggravates this effect. The lift from an upswept nose can help cancel that negative lift. This shaped nose may also pull some of the turbulence out from under the bottom of the shell. Maintaining smooth flow along the bottom of the nose area helps keep the buildup of turbulence smaller for a longer distance.

Because of the larger shell width up front, pedal Q is not an issue.

I have chosen an open cockpit so the rider can hear traffic easily. This also makes the shell lighter, with the added benefit of less fogging of the windshield. I often flip open the canopy when climbing a long hill to get more cooling air. The very light small canopy simply hangs off the side of the shell, no restraints needed. Removing the windshield during the summer is another option. You get the feeling you are really moving fast with all that air coming up over the nose.

Mounting points are also similar to the VFS: a Y-connection to hold the area forward of the canopy up and out, attached to the tube from the handlebar to the boom; a carbon tube from the bottom bracket to a contact point on the inside of the nose (on the fiberglass nose, explained in my Zotefoam Manual under the Mold section); two upper connections at the top of the back of the seat; a tail support at the end of the luggage rack; the bottom attached to the main frame in the foot hole area and along the bottom of the frame with Velcro; and carbon tubes going from the head tubes up to the spray shields. These help prevent left/right movement of the shell in the area of the cutout for the wheels.



It takes about 6 to 8 minutes to get the trike in/out of the foamshell. The front wheels need to be removed (I have outside removable axles).

page 1
page 3