EXPERIMENTS WITH A VELOMOBILE BODY
AtomicZombie has decided to build a velomobile! The following will be an ongoing build log updated every 2 weeks, detailing every step of the journey that will result in the creation of a practical and easy to build velomobile. I am taking a different approach to building the trike body, so I am not sure what the end result will be, but I do hope you enjoy reading about the journey as things progress. There is also a great discussion on our Builder's Forum for this ongoing project, so feel free to drop in and share your thoughts and ideas on the subject. .
Every time I find myself standing at the gas pump, holding down the lever while the dollars spin past, I begin to wonder if there is a better way. Let's face it, the cost of using a gas guzzler adds up to a lot more than just a dent in your wallet each time you fill 'er up, although the pain you feel at the pump is certainly instant. How about all of the effects to the environment? Using a gas powered vehicle to pick something up from the store a few blocks away is certainly convenient, especially on a cold day when you can just press a button on your remote starter and let the interior heat up for you. But, with millions of people doing this, what is the net cost on the environment? Call me paranoid, but with the crazy worldwide weather we have been experiencing in recent years, I think the answer is obvious. From this point forward, I will use the word "car" to refer to all gas guzzling ground transportation vehicles. .
Environmental issues aside, there are many good personal reasons to be leaving the gas guzzler parked more often. My health has been impacted by the convenience of the car since the first day I passed my road test. How did I all get around in the days before becoming enslaved to my car? Well, besides begging for a ride, I got around on foot or by bike! I remember how simple things were back then. My main concerns were usually how long it would take to get from point A to B and making sure that my tires had air. I had no repair budget, no insurance costs, no parking problems, and didn't have to work overtime just to pay for fuel. Ironically, I had more free time even though it took a lot longer by bike because I didn't have to schedule in time for exercise because it came with the lifestyle! That extra body weight was a direct result of using a car, too. Sure, the car helps me get around in a hurry, but I end up either wasting more time and money to sweat over a treadmill or consulting with a doctor on how to fix my health. .
Seems as though in our later years we have things backwards, don't you think? "DING!"Oh, hold on a minute, the truck is filled now. I have to go give the attendant another $80 bucks! .
I have decided to get a grip on my shrinking wallet and ever expanding waistline, and find a practical way to leave the car at home as much as possible. Now, the key word here is "practical". Living in the country of Northern Ontario means that I will always need a reliable car or truck to move large cargo and to travel large distances to town in the winter, but having a few local stores within riding distance means that a bike could certainly be used for many journeys. For those who live in the city, a practical human powered vehicle with some cargo capacity could also be used for many local trips, such as grocery runs or social calls. For me, practical also means affordable and robust, which almost always translates to home built, which to us DIY types is great news. Of course, there are commercially available human powered vehicles "velomobiles" for those who can afford them, but since they tend to be as costly as a decent used car, they are out of reach for most. Here are a few commercially produced velomobiles that certainly inspire ideas.
All of these velomobiles are obvious works of art, but there is no way I would ever part with ten grand for something that I could build myself. Obviously, there will be tradeoffs between cost and aesthetics, but there is no reason why a very practical and sturdy velomobile could not be built using readily available parts by anyone with a few basic tools and a lot of motivation. In fact, I have seen some home built velos that are streamlined works of art, but often the cost of materials used and the skill set needed are beyond most of use weekend garage hackers, and the end product is more like a hotrod than a bike you would want to take out in traffic or ride around in the rain.
The yellow streamliner called "Barracuda" was designed and built from scratch by Warren Beauchamp, and has always been one of my favorite racing machines, capable of speeds normally reserved for cars on the highway. If speed was my primary goal, then my velomobile would defiantly take inspiration from the Barracuda, but this time I am after something more practical for everyday use.
My goal is to build a body using basic materials that is both aesthetically pleasing yet at the same time tough enough to live in the real world. Living in the real world means taking Mother Nature's wrath of rain, sleet, hail, wind, and constant bombardment of UV radiation. Living in the real world means surviving the odd ding, dent, or scratch from crowded urban environments, being able to bounce over a curb and take the abuse of a poorly maintained road without shaking to pieces. Living in the real world also means living in the urban jungle, so the vehicle will need to be visible in traffic and include the usual safety gear such as rear view mirrors, brake lights, head lights, turn signals and a horn. Living in the real world means offering the pilot some shelter from the elements without requiring any acrobatic maneuvers to climb in and out of the vehicle. And of course, living in the real world means that the vehicle must include some practical cargo carrying capacity for such things as groceries, a battery pack, and personal items.
So with all of these goals in mind, the first choice becomes - delta, tadpole or quad?
The type of base vehicle will determine the overall shape of the body as well as its load carrying capabilities, handling characteristics and aerodynamic advantages. Choosing one of the three configurations was actually quite a chore as they all offered advantages and disadvantages when it came to costs, aesthetics, practicality, and ease of building. In the end, I decided that a delta trike would be the most practical base vehicle, but I will discuss all three possibilities, as well as their strengths and weaknesses.
A quadcycle or "quad" is a four wheeled bike with two front steering wheels and one or more rear drive wheels. The obvious advantage of such a vehicle is stability, since it has four points on the road at all times. My initial idea was to start with a quadcycle like our StreetFighter and just bring up the seat height and overall track width a little more for visibility and stability. I had a lot of fun carving up the corners in the StreetFighter, and knew that it to be a very robust and stable vehicle for just about any kind of riding and terrain. But, after giving more consideration of the goals for this project, it became apparent that there would be no real advantage to having two wheels at the front and a quad configuration would make using an electric assist hub motor impossible. So, the quadcycle idea was abandoned in favor of a trike, which now left only two choices: delta or tadpole configuration.
Two wheels in front - The Tadpole Trike
A tadpole style trike places the two steering wheels in front beside the pilot and places the single drive wheel at the rear behind the pilot. This type of trike offers great handling and allows for the use of mostly standard bicycle components in the rear end and transmission system. As for aerodynamics, a tadpole trike is better than a delta trike because a teardrop body shape can be used, which places the tapered end at the rear of the vehicle for optimal travel through the air. Another nice thing about the tadpole trike configuration is that a full fairing looks nice as it wraps around the pilot, leaving the large end in the front with plenty of room to install a full view windshield. Most of these advantages are more aligned to a vehicle designed for looks and speed, and not so much to a practical velomobile designed for everyday use and cargo carrying. A teardrop shape leaves almost no room in the rear for a cargo area, and although rear hub motors area available, they are less common and more expensive. This leaves only one other configuration to examine, the delta trike.
Two wheels in the rear - The Delta Trike
A delta style trike placed one or more drive wheels behind the pilot and has a single front steering wheel ahead of the pilot. This type of trike offers stable handling using standard bicycle components in the steering system and a great deal of cargo carrying ability. Having the wider end placed at the rear of the vehicle means that a delta configuration is less aerodynamic than a tadpole configuration, but this is hardly noticeable at the typical riding speeds that will be seen in city traffic. For a racing vehicle designed for optimal speed, this teardrop shape is the ultimate design factor, but for a utility vehicle, it will be of little concern. Having the load bearing wheels and drive system at the rear where the pilot sits is a great advantage of the delta configuration because it will offer the most traction and braking possible as well as keeping the load centered between two wheels. Since there is nothing except for the single front wheel and pilot's feet at the front of the vehicle, the body shape can taper towards the front and keep the total volume much lower than that of a quadcycle configuration. A delta trike is also the most robust of the three possible configurations because it uses standard bicycle components in the steering system with axles and industrial grade bearings in the rear. Having a standard set of bicycle forks in the front means that a commonly available electric hub motor kit can be easily installed to offer an electric assist drive.
So, it became obvious that a delta trike would offer the best base vehicle to use for a practical velomobile with load carrying capacity and an electric assist drive. Making the final decision on the base vehicle meant that I could now come up with my wish list of other features I wanted to have in my practical velomobile, which would include rear suspension, dual disc brakes, a suspended cargo bay, over-seat steering, long range electric drive, a trailer hitch, a head light, turn signals, rear view mirrors, and of course a full body with easy access when getting in or out. All of these features would be easy to integrate into a delta trike design.
Since winter has now set in around here and I normally work outdoors, I plan to build myself a velomobile in what would be considered reverse order - making the body first and then building the base trike to fit into the body. Typically, I'd build a body over a trike, but since I have to work indoors for the next six months and I already have a trike to use as a reference (my Aurora Delta Trike), I feel confident that this order of operations will turn out fine. The Aurora trike already has most of the features I want such as rear suspension and dual disc brakes. I only need to decide on a seat height and track width to use in the final design.
For my practical velomobile, I want a slightly higher seat height (similar to a car) for optimal visibility in traffic as well as a slightly wider track width for cornering stability and increased cargo capacity. Other than those changes, the Aurora trike is a good fit with its rear suspension, a recumbent seating position, and a robust transmission system. I'm able to remove the rear suspension spring and simply prop up the rear of the trike to the height I want, so now I have a base vehicle that can be used to take measurements and work out the size and shape of the body. Using a few photos taken from the side, I can make several mock-ups in Photoshop to help me decide on a general shape of the body.
With a base vehicle to take measurements from and a clear set of design goals, I can concentrate on the side profile of the body, which will involve some practical thinking as well as artistic abilities. The final design will become a combination of both aesthetics and ergonomics, allowing the rider to get in and out of the vehicle easily, while maintaining a sleek modern look with pleasing curves and proportions. Some of the other factors that will influence the final design will be the type and strength of the materials used, plus costs and complexity.
I printed out several copies of the Aurora trike side shot and then drew a few different body styles over the photo in order to get some ideas. Some of the important features of the body will include the rear cargo space, ease of access, and the size of panels needed to make the sides and top. If I can keep the maximum length to under 8 feet, then inexpensive materials can be used. So after a few drawings, it came down to the longer body that covers the front wheel, or the short body that ends before the front wheel. Once I decide on a final body shape, it will then be a matter of choosing the best possible materials to build a lightweight and robust body for my practical velomobile.
Now that I have come up with a basic plan, which so far includes putting a body around a sturdy delta trike, I can now move ahead into the specifics of the project. Winter has officially moved in around here now, so all of my good bike parts and the few bikes I still have together are now resting in a quiet corner of the basement, where I intend to build the velomobile body over the next few months. Since the Aurora Trike was one of the last projects I did in 2012, I still have it completely assembled, which was a good thing since it seems to be a perfect reference vehicle for the practical velomobile.
I had a few hours to spare, so I made a 3D model of the original Aurora with a few modifications such as a wider track, higher seating position, and over seat steering. I am still not sure if I will be using over seat or under seat steering, and will probably not know until I have enough of the body work done that will allow me to climb in and out to see how everything will fit together. I had this idea to allow the entire body to lift from the front, allowing the pilot to just sit down into the vehicle without having to climb through the side, and if this does actually work out, then under seat steering will be the better choice.
Using the 3D model, I scaled it for printing so that 1 inch on paper would equal 1 foot in reality, which will allow a cardboard model to be made so that I can test a few different body designs using construction paper. One of my key goals is the use of only common and inexpensive materials here, so I will need to keep the maximum side profile of the body to no more than 4 foot by 8 feet so that off the shelf materials can be used.
I printed out the 3D model and then traced out the trike frame bits onto some cardboard. Each of the parts were then taped together along with a copper wire rear axle in order to make a perfect scale model that represented 1 inch on paper for every 12 inches on the real trike. I cut out several 4x8 inch rectangles from yellow construction paper to represent the 4 foot by 8 foot panels that I will be using to make the body. A lot of strong and lightweight materials can be purchased in 4x8 sections from a hardware store, so as long as I stay within these dimensions, there will be plenty of options available.
The first shape I made for the side body profile was based on the previous Photoshop drawings, representing a simple curve that flows from the rear of the trike, up over the pilot, and then around the front head tube just above the front wheel. This body design just seems to look nice, and is just under 8 feet long, so it fits the requirements well. The longer body that encloses the front wheel would need to be almost 9 feet long, and it just doesn't look as nice as the shorter version.
making a the model body with the construction paper was as easy as taping it all together at the edges. The entire body consists of only 3 panels - 2 sides and a top strip, none of which exceed 8 feet in length or 4 feet in height. The body is wider at the rear, tapering towards the front so that there will be maximum cargo room and elbow space at the rear and just enough room for the pilots feet at the front. The body will also include the high brightness LED headlights, brake lights signal lights, and dual rear view mirrors. The front windshield will be made of thin lexan, travelling from the top slop of the body all the way down to the curve over the front wheel for maximum forward visibility.
The side openings will allow enough room for the pilot to enter and exit the vehicle if I use over seat steering, but at this point I am swaying towards hinging the body at the rear to allow it to just lift right open at the front, allowing the pilot to sit into the vehicle and close the lid back over a front latch. This would allow those with less than perfect agility to enter and exit the vehicle easily. As for doors, I am have not decided if I want removable hard panels with side windows, or just a zip or snap up material type of door. Either way, the side panels (doors) will be fully removable and most likely only used during extremely cold months or when it's raining.
Considering the body is made from only simple 2 dimensional curved shapes, it still has a nice flowing look to it. This design is certainly not as sleek as those hand sculpted 3D teardrop streamliners, but it is certainly practical and will be inexpensive to manufacture from basic materials. I fully expect this body to weigh only a little more than some of the commercially available units, and have no doubt that I will be able to lay across the top without having it buckle under my weight. All edges will be made with 3/4 inch wide angle iron with a 1/16 thickness, so this body design should not need much reinforcement either - a problem that often plagues the ultrathin fiberglass / epoxy designs. It makes more since to me to go heavier right from the start rather than trying to build an internal skeleton just to support a flimsy shell.
I am happy with the look of the paper model, so next weekend it will be time to switch into high gear and actually start souring the materials that will be used to make the body. Currently, I am looking at using 3/4 inch angle iron to make a complete perimeter frame and then fastening 3/8 thick flexible plywood to this frame so that the metal corners create a full weather seal and protect the body from any accidental impact damage. I will probably roll a coat of epoxy over the completed assembly to just give it a full watertight seal and to allow a nice smooth surface for painting.
Now it's off to the steel supplier and hardware store to begin collecting materials. Hopefully the full sized body will be as easy to build as the 9 inch scale model!
To keep the materials both inexpensive and easy to source, I must constrain the overall length of any panel to 8 feet or less. Placing the Aurora up against a 4x8 sheet of plywood, I can see that this will certainly be possible, but only if I use the body design that exposes the front wheel rather than the longer body the encloses the front wheel. The next step will involve choosing some materials to build the body panels and then a full size layout onto a 4x8 sheet.
I have found some very lightweight, inexpensive and durable wood panels at the lumberyard that seem to be perfect for this project. These panels sell for about $15.00 each, and are referred to as mahogany, luan, and door skins. These luan panels are only slightly heavier than coroplast, and are much more stiff. According to my research, these luan panels seal up nicely and have been used to make such things as kayaks, canoes, and many other outdoor projects.
Coroplast has been the widely accepted material used to make velomobile bodies that don't include fiberglass and epoxy construction, but from my readings, this mode of construction requires an intricate ribcage underneath the panels so they won't buckle or flap in the wind. Many have claimed that once they have finished making the shell, the support system ends up weighing much more than the actual coroplast body. Since I intend to start with a robust support framework made of angle iron and just lay the panels in, coroplast would probably work, but would add nothing to the strength of the completed body. One of my goals is to be able to lay across the top of the completed body, so the side panels need to add some strength to the completed assembly, and I don't think that the coroplast is up to the task.
As show in these cross sections, coroplast is basically plastic corrugated cardboard. The panel I purchased for testing is 4mm thick and white in color, costing $25.00 at the lumber yard. Shown next to the coroplast is a cross section of the $15.00 luan sheet, which cost $15.00 at the same lumber yard. The luan sheet is slightly thinner, measuring 3mm in thickness.
Although weight was not as important as durability and functionality, it was still a factor to consider when choosing the side panel material. Luckily, the luan sheets compared very well against the ultra lightweight coroplast sheets. The 4 luan sheets I purchased all had slightly different weights, but were all close to 7.5 pounds each, some being slight more and some slightly less. being a wood product, this slight weigh variance in weight would make sense, so I will work from the high end and say that the full 4x8 luan panels weight 8 pounds each.
I already had the coroplast weight chart, but threw it on the scale anyhow, just to get a photo. To my surprise, the coroplast sheet weighed in at 6.5 pounds, not the 4 to 5 pound weight that was claimed all over the net. So perhaps my scale is not that accurate? Well if that is the case, then the luan sheet would only weigh 6 pounds, but either way, both 4x8 panels are very lightweight, with the luan sheet weighing only a pound or so more than the coroplast.
The panel material will need to bend to make the nice curved top section of the body, but also needs to offer enough stiffness so that the completed enclosure is strong enough to stand up on its own and not flap in the breeze or warp back and forth while turning corners. After handling the coroplast for a bit, I realized that it would not add any strength to my shell, leaving the angle iron perimeter to hold its shape all on its own. I can now see why coro built velomobiles bodies require such an intricate skeleton - the material has amount the same stiffness and strength as cardboard of the same thickness.
My strength testing was done 2 ways. First I placed the sheet on top of a ladder to see how far it would bend under its own weight. This would give me an idea of how well the longest side panel would fare in a wind. The coroplast bent almost right to the floor, so I can see it would definitely need some underlying support on any length over 4 feet or so to avoid rippling in a breeze. The second test was a bend test done by pushing down on the 4 foot width from the top as the sheet stood on the floor. This would give me an indication of how much strength the panels would add to the angle iron framework once installed. The coroplast did not do well in the bend test. In fact, I could warp it right over by pressing down with a single finger, and this became the deciding factor to not use it. I would probably use the coroplast for such things as wheel covers, wheel wells, or perhaps a tailbox for a 2 wheeled lowracer, but I can't imagine building an entire body from it. At least not one that will meet my durability and strentgh requirments.
The 4x8 sheet of luan was put through the same 2 tests, and passed them both with flying colors. Placed on the ladder, the luan remained stiff enough to carry its own weight with only minimal deflection, so it would certainly remain stiff in a head on or side blowing wind. During the bend test, the luan held firm, and I could barely deflect it along the horizontal length without pushing down with a lot of force. Installed into the angle iron framework as side panels, I could certainly imagine the completed body having enough strength to take my weight over the top. So now that I have ruled out the coroplast as being a suitable material, the luan remained alone and will now be put through the most important test - the 180 degree bend test.
Since I decided to make a rounded body, the chosen materials would have to bend around these curves. I knew the coroplast could definitely bend easily, so much so that I could probably roll it up into a 2 foot circle without damaging the internal structure. Wood on the other hand will only bend so far before it snaps, and knowing that the luan sheets were fairly stiff, I did not know if they would pass the 180 degree bend torture test. The curve on the rear of my body design is actually less than 180 degrees, but I figured it would be best to push the materials just to be safe. I placed one end of the luan sheet up against the wall and then began to push the other end along the floor with a box made of 2x6 wood. I kept pushing the sheet until it was rolled into a perfect 180 degree arc, and it did not snap or crackle much. This was a good sign, but to be certain, I would have to test all 4 sheets. Luckily, they all passed the 180 degree torture test, and only one of the sheets showed a small amount of fraying on the edge, most likely from a glitch in the glue that holds the plies together. So now I have the chosen side panel material and can continue to move forward on this project.
Now that I knew what materials were to be used to make the body, a full sized side profile template was needed. This will be used to ensure that the 4x8 panels would be long enough to make the sides and as a guide to shape the angle iron into the curved framework. My miniature model had the lower section of the body placed at the same height as the axles (10 inches from the ground), and this looked good, so I copied this onto the final design. To raise up the panel to the require height, I stacked 6 small 2x4 sections on the ground and then placed the 4x8 sheet on top of them. Now I could sit on the Aurora and mark important points on the 4x8 sheet such as front and rear wheel placement, foot clearance, and head clearance. To mark these points, I taped a broom handle to a work light and used it to test clearances as I sat on the trike and pedaled in reverse. Headroom would be very important, and I gave myself a few inches extra to take into account tings such as alternate seat materials and a helmet. Of course, I do intend to build a brand new trike to fit this body in the spring, but the Aurora was a good start.
After marking the placement of both wheels and the clearance points for my feet and head, I made a half circle wheel template from cardboard so I could trace the rear wheel well and front wheel onto one of the 4x8 sheets. The wheel template also included the head tube since the front of the body would need to allow for clearance as it travelled over the front wheel. I also added 6 inches of extra length to the body just to be safe. Even with the extra 6 inches of length, the 4x8 sheet still had about 6 inches of extra length, so even a 7 foot tall person could make a velomobile body for a delta trike using a 4x8 sheet, so that was good to know.
I was trying to figure out how to draw the curved section of the body onto the side panels, and finally came up with a solution that takes care of 2 important issues; how can I make a nice curve?, and will the top section panel be able to bend into that curve? So I wrapped a ratchet strap around the length of one of the luan panels and then just cranked it into the curve I wanted. This process took care of both problems, making a nice flowing curve I could trace while at the same time proving that the panel could indeed be bent into that shape. I bent the panel into a curve that would span from back of the rear wheel well all the way around to the tallest point of the body, where the head clearance point was marked. I then traced this curve directly onto the panel using a marker. This curve was only about 100 degrees, so that was certainly safe since the panel already passed a 180 degree bend test.
To make the front section curve, I relaxed the strap until the panel was bent into a curve that would join up with the top point and then flow nicely to the front of the body, over the head tube. I once again traced a line onto the side panel, using the bent panel as a guide. This process worked out very well, resulting in a shape that was very close to my miniature model, and certainly bend safe for the top panel.
This side profile template will be used as a guide when I begin shaping the angle iron framework. Once I have made both metal edge sides as close as I can to this profile, I will then use the completed metal framework as the final template to trace out the side so they fit perfectly into the shape. In my design, the side panels create the wheel wells, so perhaps that coroplast can be used to create the wheel covers as well as the inside fenders later on. I know I will be able to use it for something here.
Here is the Aurora sitting next to the side profile template. it looks like the final velomobile will look very close to my original miniature model, so that is good. The overall length of the body is 90 inches, and it is 47 inches tall from the ground to the highest point. I did not add the side windows to the template yet because my new plan is hinge the body at the rear and gain entry by simply lifting up the front of the body. This will make it so much easier to get in and out of the vehicle and allow much smaller side openings to be used. I intent to build the entire body without the side openings and test the lifting system before I decide on how large to make them, just in case I have to alter the plan again.
The next step will be the angle iron framework, making it into a shape to match the side profile template.
I wanted to use 3/4" angle iron with a 3/32" thickness for the body framework, but when I went to the local steel supplier they said that size would need to be ordered. What they did have in stock was angle iron of the same diameter but with a 1/8 thickness and this was a regularly stocked item. I had a sample of the 3/32" thick angle iron with me so I went to the warehouse to compare it with the 1/8" thick stock item just to see how much difference there was. The 1/8" angle iron didn't seem too large for this project, so I purchased six lengths, more than I would need, but at $15 per length I would certainly make use of it elsewhere.
This particular angle iron weights about .58 pounds per foot, and since I calculated that I'd be using about 40 feet in total, that would add up to about 23 pounds for the framework. That wasn't too bad at all. Add the approximate 20 pound weight of the wood panels, and I could still make this robust velomobile body come in under 50 pounds, which is not bad at all considering a foam shell/Coroplast body can easily weigh 40 pounds if you include all of the extra support needed . I am now starting to think I may succeed in my goals of a super sturdy sub 50 pound velomobile body that is strong enough to take my full body weight across the top.
So, how does one go about bending angle iron? I mean, it is basically an open sided square tube, so you certainly can't use a pipe bender on it. My plan was simple: cut a few slots, make the bend, weld them back up, and then grind the face flush again. This may sound like a lot of work, but it isn't since the welding and grinding will happen on a flat surface, so it won't take much effort or time to get slots filled again. This method also offers the advantage of being able to make curves in two directions (dual axis). My design includes nice curves in the side profile, but also requires a gradual curve from the rear where the body is wider up to the front where it is most narrow. So, this method of cut, bend and weld will allow me to make controlled curves in two directions.
I have never tried this procedure on angle iron before, so I have no idea how many cuts per inch are needed for a given arc. I cut two 12 inch test sections from the angle iron and marked them to one inch and two inch points. I will be using a 3/32" thick zip disc, which is a little thicker than the ones I normally use for metal cutting.
Cutting out the bend slots was easy work; just lay they part on the bench and run the angle grinder along the marked lines until the zip disc cut was made through one side of the angle iron. So, now one length of angle iron is cut with 5 slots per foot and the other is cut to 11 slots per foot. This will help me come up with some basic formula to help determine the cut spacing for a given curve. Of course, it will be easy to make adjustments and just fill the gaps with the welder as well.
Bending the angle iron is easy once the slots have been cut. Grab each end with a vice grip and then bend it until it will no longer bend any further. The only thing that can go wrong is not having enough slots per foot for a given curve, so this is why I am doing these tests on short lengths of angle iron. Too many slots per foot is not so much of a problem as I can just back off the curve and fill the gaps with weld metal.
The length with 11 slots per foot bent into a very nice curve. In fact, it actually looked round, rather than segmented. Bending the 1/8" thick angle iron take a decent amount of force, but could be done by hand, so that made this procedure quite easy to do. The small arc that was made was quite impressive. I can see this system being used to make fenders and other curved shapes in later projects.
The other length of angle iron with the slots placed every 2 inches apart made a nice gradual curve that looked just about right for the curvature needed at the rear of the body. There was a bit more of a segmented look on this curve, but at a short distance, it still appeared to be round, so this process was a total success. It was difficult not to lose focus now and start plotting out all of the cool things I could make with the curved angle iron!
Both segments can now be used as a guide when working out the number of slots per inch needed to form the side sections of the framework. It will still take some "guestimation", but I could already see that the section cut at 5 slots per foot was about right for the curvature at the rear of the body. The angle iron can also be bent into an outside curve using the same method, and although it will take a bit more welding and grinding to fill the larger gaps, it will still be a fairly easy procedure.
The curvature on the 11 slot per foot section was about right for the wheel wells, but they will require the angle iron to be bent in the other direction, making an outside curve. No problem, this system works both ways, and now I know that about 10 slots per foot will work nicely for making the wheel well sections.
At five slots per foot, the resulting arc is almost perfect for forming the curve at the rear of the body. Since the rear has large curve that is basically the same from the rear wheel all the way to the center roof section, this makes it easy to figure out how to cut the slots in the long section of angle iron that will make up this part of the framework.
Now that I have a good idea of how many cuts per foot are needed for a given curve, I can start cutting the angle iron into sections that will make up the dual side panel frameworks. The longest section is the top rail that starts at the rear wheel and makes an arc over the pilot, continuing until the front of the body. To calculate how long this section needs to be, I have to measure along the lines made on the full sized template. Doing this measurement is impossible with a standard tape measure, so I will tape a length of rope to the template and then mark it so it can be used to determine the length of this top rail section.
The rope is taped to the side panel template, along the top rail section. The taping job isn't perfect, but it will be close enough, and I will cut an extra 2 inches of angle iron just to play it safe. The rope method will then be used to determine the length of the bottom rail and the rear wheel wells so I can cut all of the angle iron needed to make both sides, which will be 6 lengths in total. Once I have all of the angle iron sections, I will again use the rope to transfer the approximate slot cutting pattern to the angle iron by marking the rope where the different curves are needed. And, once the side rails have been bent to shape, I will weld them together to form the side profile and then use them to draw the final shape onto the wood panels so that they are a perfect fit.
The overall shape of the velomobile body will be determined by the two top rails that run over the pilot from the rear wheel to the front wheel. The top rails also determine headroom, trunk space, knee clearance and windshield length, so they are important to the overall design. Now that I have made a few test bends in the angle iron, I feel confident that this system will work to create the shape of the top rails, which should be close to the shape drawn on the wood panel. There are basically two curves in the top rails shape: the tight curve at the rear and the gradual curve over the front section. To make the tight curve at the rear section, the angle iron will be cut at an interval of every 2 inches, and then at every 3 inches for the front section. This should be about right as shown by the test pieces.
The rope that was taped to the outline is pulled back until the start of the tight curve, which begins just over the pilot's head. This section of rope is marked off and then used as a length template to determine how long the section with 2 inch spaced cuts needs to be. I determined this length to be 56 inches. The remaining 70 inches will be cut at every 3 inches.
All cuts are made using an angle grinder with the same 3/32 thick zip disc. After bending the rails, I realized that a thinner zip disc could probably be used as well, but this way, there is some room for error, and all gaps will be easily filled when the welding is done. After cutting the entire 126 inch length of angle iron, it lost most of its rigidity and would start to sag under its own weight when picked up from the center. Technically, it is just a piece of flatbar with side tabs at this point, but will become rigid once again after welding.
To shape the angle iron after all markings have been cut, it is laid down over the wood panel and bent one bit at a time by gripping between each bend. It would not be a good idea to just grab each end of the entire length and try to bend it all at once as this would create uneven bends, resulting in visible corners and errors in the flow of the curvature. The angle iron is still strong enough that it requires a fair amount of strength to bend it between each gap, which makes it easier to get the correct shape by bending it a little at a time. This photo shows the rear curve bent to follow the shape of the template drawn on the wood panel.
The first side rail worked out very nicely, and only took a few minutes to bend into the desired shape. As you can see, I went outside the template a bit, allowing for a little more headroom and a little less knee clearance. This new shape looks a little nicer and will take into account a taller rider and a helmet, just in case someone begs me enough let them take a test ride in my velomobile. The curved angle iron is not very rigid, so it must be handled carefully as to not bend under its own weight. I can carry the shape around, but it wouldn't take much to bend it out of whack, so it needs to be secured for welding.
Since it is winter, I am working indoors as much as possible, and have to travel up the stairs and across a snowy field to my welder, so the bent rails need to be made more robust for transport. I measure the desired length for the side rails from back to front and then marked this length on one of the leftover angle iron sections. The top rail is then tack welded to the angle iron so that it retains its shape and can be handled without risk of warping it out of round. This is also necessary because the welding process will try to distort the rails into a tighter curve as the weld metal cools and shrinks in each gap as it is filled. By tack welding the side rail to the angle iron, it can be safely handled and will not distort during the welding process.
Ahh, there is nothing like the sound of a crackling arc in the cold winter air! It can be a bit annoying to defog your welding lens every 30 seconds, but the job will get done nonetheless. I placed the angle iron base across a pair of saw horses and then propped up the top rail along a fence board so I could tack weld every gap. I wanted to tack weld the gaps and then do one last test fitting with the traced template, just to make sure that the shape would not be too distorted during the welding process. It was a bit of a challenge getting this part in an out of the house, but it all worked out.
After tack welding each gap, the part was moved back indoors for one last check against the wood template. There was almost no distortion in the shape, so it will be safe to complete the welding of each gap. The newly adjust shape is definitely much nicer than the original trace done on the wood panel using the curved wood sections as a guide. This shape is almost identical to the paper model now which is sitting on the window sill telling me to hurry up and get the velomobile built so I can ride it in the springtime. Hey, I'm going as fast as I can - it's -45C out there today!
To minimize my time in the frostbite zone, I decided to complete the welding on both side rails at the same time. Actually, tack welding them together insures that they will both be the exact same shape after welding, and that distortions will be kept to a minimum. The original side rails still has the angle iron base to keep its length secured, and the new rails are tack welded along the original rail every few inches so that they are exact copies of one another. All gaps are welded now, which is a process that only takes a few minute for each rail. This cut-n-weld process is really working out well, and I can already see it being used in future projects to create nice rounded shapes in angle iron or even square tubing.
After welding every cut on both side rails, they are then separated and the lower support removed. Both rails maintained their shape, so this process was a complete success. From here, the rest of the curved parts and the side supports will be based on the shape of these top rails, so it is smooth sailing from here. The strength of each rail section has also been greatly increased now that they are true angle iron sections again. With both rails placed side by side, I can lean my full weight over top and they do not bend out of shape, so my goal of making a lightweight body that I can lay across is going to be no problem at all. In fact, I would almost bet that the frame alone without the wood panels installed might be able to support my full weight over the top. Not bad for an 8 foot long body made from less than 20 pounds of metal.
A quick test using the Aurora Trike shows that this shape offers plenty of headroom for a helmet, and will leave a good amount of cargo space for groceries and a large battery pack. There is also enough room to make slight modifications like lowering the seat height and the front of the body over the head tube if I wanted to later on. Knee clearance is also fine, and the overall shape of the body will now be very close to the paper model, which I like. Looks like everything is starting to come together, and I think this was probably the most difficult part of the body assembly as it required that perfect curve. Working with lengths and angle is easy, but making a curved shape is more of an art form that requires a lot of guessing and trust rather than measuring and calculating.
It's so tempting to push on those pedals and see how fast I can go before hitting the end of the basement, but I am old enough to know better, right? Enough said. So, now that the two curved top rail sections are completed, the next phase of the body assembly will include the lower section and wheel wells in order to complete the side body sections. With the side body sections completed, they can be used as a template to cut the wood side panels, creating most of the bodywork. I will then have to work out the overall width, which will have to do with the under seat steering and crank widths. The completed body will probably be about 4 inches too wide for the Aurora, but I do intend to build a new base vehicle to fit the body, so for now I will continue to use the Aurora as a guide.
A week has now passed since the completion of the top rail, but the harsh Northern winter has made it very difficult to work on the Velomobile project. I intended to complete the bottom section of each of the side panel frames and then cut the wood panels to fit, but the snow buried my welder, and kept me busy for days just trying to shovel my way out of the driveway. When you open your front door and 2 feet of snow falls into the house, you know that the only tool you will be using all day will be a shovel or a snowblower if you have one that works!
After a few days of toiling away digging out of the snowfall, the weather actually warmed up to a mere -20 degrees Celsius. No, I did not type that wrong; it "warmed up" to -20 Celsius or 0 degrees Fahrenheit, depending on where you are from. Either way, it's probably colder than that out since my little thermometer only goes down that low, bust that sure beats the -30 and -40 temperatures we have been having all week long. I find that below -20, the welder has some trouble, especially if there is a wind. I have to crank up the amps to compensate, but that quickly causes burn-throughs as the metal heats up. Between that and the constant fog on the welding lens, it can be very tricky to work out in the open in this freezing climate. Ok, enough complaining, I just wanted to shock those who are lucky enough live in a place out of the evil grasp of Old Man Winter.
With the top rails now complete, there are two other sections needed to complete the sides - the bottom rail and the wheel well section. The rope trick worked perfectly on the top rail, so it will be used again to calculate the length of the other sections of angle iron needed. The completed top rail was laid down on the template and a section of rope was taped along the curved line representing the bottom rail. The rope was marked off at the wheel well joint with a piece of tape.
The curve at the front of the bottom rail had a similar arc to the rear of the top rail, so cuts were made every 2 inches from the front end of the angle iron to the lower horizontal section. The bottom rail section bent up nicely, and was adjusted to follow the line on the full scale template. Now, the only other section needed to complete a side shape will be the wheel well. This piece will require the most effort.
The wheel well sections are the shortest pieces of angle iron to be bent, but they will be the most technical and time consuming due to having a very tight arc and an oblong shape. On the template, the line only represents the top of the rear wheel but since the rear end is connected to a suspension swing-arm, the wheel will need to travel upward past the top of the draw arc as the suspension is compressed. This extra headroom will mean that the arc now has to stretch into an irregular shape above the top of the wheel. To figure this out, the original cardboard wheel cutout will be made into a mock suspension arm with a wheel attached.
Since I knew the distance of the swing-arm connection point from the rear axle, I added a section of cardboard to the original wheel template with this point represented as a woodscrew. I could then fasten the template to the wood and lift it upwards to create a tracing guide to draw the top of the wheel well. A bicycle suspension spring can compress about 1.5 inches total, so I marked a point on the template representing the suspension spring connection point and then pivoted the wheel template up by this amount to determine the maximum height of the wheel well. A new line was drawn around the top of the template and merger with the original wheel well at both lower corners. This new oblong shape is the shape that will be used to create the wheel well sections out of the angle iron.
With the new irregular wheel well lines drawn on the wood panel, the rope and tape method was used once again to make a length guide to cut the two angle iron wheel well sections. At this point, all of the curves required the cuts to be on the inside of the curve, but here, the cuts will be outside cuts. The wheel wells also have a much tighter arc than any of the other shapes, so they will require the most cuts, at only 1 inch between each cut.
The wheel well sections were only 38 inches ling, but require 38 cuts, making them the most labor intensive parts out of the all the others. Welding the wheel wells will also require the most filler metal because the cuts are on the outside, making their gaps open up rather than close up like inside cuts. But even so, cutting these parts only takes a few minutes, and so does the welding process. This process is still easier than trying to make sturdy edges using woodworking and glue.
Bending the wheel wells into the irregular shape was a fairly easy process and only required a few tweaks here and there. Even if the wheel wells were not perfect, they would still be fine as long as the margin of error was less than half an inch at any point along the arc. After checking the part along the traced line, I had them almost perfect with a few minor adjustments. The first bent part would then be used as a guide to make the other one an exact mirrored copy.
Since there is a left and right wheel well, both parts need to be made as a mirror image of each other. This was easily done by laying the parts against their flat edges as shown here, using the first part as a guide to bend the second one. To bend the wheel well section into the correct shape, it is worked from one end to the other, making any necessary adjustments after checking the arc using the other part.
After a few minutes, the second wheel well section was bent into an exact mirrored match of the first. At this point, I was getting good at bending the angle iron into a nice curve, so it was very easy to make both wheel well sections match perfectly. I will definitely be using this method of making curved angle iron shapes a lot more in future projects.
With all of the parts bent and matched, they were clamped together and tack welded to a length of tubing for welding, just like before. It was a very windy and cold day out at the welder, but things went well despite my welding lens frosting over and the constant fluctuation of the arc as the wind blew at my work. All of these joints will be ground flush with a sanding disc later on anyhow, so they don't have to look pretty at this point. The wheel well sections kept their shape very well considering that 38 gap filling welds were made on them, and this had a lot to do with them being clamped together and welded to a section of tubing to keep them from distorting while the welds cooled.
The sides of the velomobile was really starting to shape up now that they have been made into fully enclosed sections. I decided to drop the small 5 inch section that was originally planned for the front joint between the top and bottom rail, so now each side section is only made of three separate parts: the top rail, bottom rail, and wheel well. The removal of the 5 inch front section dropped the front down a bit, but there was plenty of leg room, so this won't be an issue. I just think the shape looks better this way, and now it matched the paper template almost exactly. There is still plenty of room to mount a headlight and air intake frill in the front, just over the front wheel, so the plan does not have to change.
The next step will involve welding the pieces together to create two identical sides that can be filled by the wood panels. Once the wood panels have been installed onto the side rails, the rest of the important measurements such as track width can be made based on how much room I need in between the sides when they are placed around the trike. I will have to wait for a nice bright day so I can set up a larger work area near my welder that will allow the sides to be placed flat and welded using the clamp together method.
With all of the pieces cut and ready for welding, I decided to do one more size test with the trike, since adjustments will be difficult to make once the welding is done. I clamped the three sections of angle iron together to form the side shape and then set it up over the trike so I could sit in the seat and check head and knee clearances one last time. Everything checked out according to plan, so the welding could commence. Welding the side rails together marks a milestone in this project as it defines the shape of the velomobile body, which everything thereafter will be based on.
To join the three rail sections together, a few cuts must be made so that the angle iron sections can be welded so that the corners continue through the entire perimeter. So, basically the L shape of the angle iron must run the entire length of the side so that it is sturdy and capable of supporting two wooden panels, one for the side, and one for the top. To join two segments of angle iron, you can either cut them at an angle or take a section out of one side as I have done here. In the end, either method will work the same, since the welds will all be ground flush when the metalworking is completed.
Welding the angle iron segments together proved to be quite a challenge. Not because the welding was difficult; it was the 2 feet of freshly fallen snow around my makeshift outdoor workbench and the fact that the side shape was much too large for the two old sawhorses I was welding on. It took an hour to shovel my way to the welder and dig it out, and then another hour of fiddling with boards to get the three segments aligned perfectly for welding. This is one of those jobs were having a nice flat and level workshop floor would sure have made things easier. Oh well, if I wanted everything to be easy, I would probably be riding a store bought bike and not building a velomobile!
After carefully shimming up the old fence boards on the sawhorses, I managed to get all three angle iron sections aligned and ready to weld. It was only -9 out today, so it was a real treat to be out welding without needing to scrape the ice of my lens every 30 seconds. The welding went quickly, and I then clamped the second rail along the first one to use it as an alignment guide, ensuring that both sides would be exact mirrored copies of each other.
Welding both sides went perfectly, but it was getting more difficult to move the parts up the stairs, through the living room and then into the basement. I am not quite sure if I will be able to join the two side panels into a full body skeleton and still move the frame into the basement. If the velomobile body becomes too large to get down the stairs, then I will have to shift focus to the base trike and wait until the weather warms up to complete the body. Either way, there will be a lot to keep busy with and spring should be coming in 60 days or so.
This photo shows how the sides have been tack welded together in a few places so that they will become perfect mirror images of one another. When I ground away the tack welds to separate the sides, I was surprised at how accurate they both were. I expected some deflection of the shapes, but when placed next to each other, they were no more than an eighth of an inch out at any place. So far every step of this project has worked exactly to plan, which is not unusual for a prototype with so many untested techniques and ideas into it. I hope this luck continues!
At this point, the velomobile exoskeleton is mainly completed, and will now have the wood panels installed so that they can be used as a guide to determine the optimal width of the vehicle. Once I determine how much elbow and leg room I need, the two side segments will be joined together with several horizontal rails, completing the entire body framework. Actually, it's a bit more technical than that, since the sides curve from the rear to the front, so a few more cut and weld steps will be made to form the slight back to front curvature of the body. But before that is done, the wooden sides will be cut out and placed into the framework in order to work out the optimal width for the front and rear of the body.
The next step in the creation of the VeloMobile body is to cut the wooden side panels and install them into the angle iron side rails. Filling in the side panels will allow me to set them up over the trike and decide on how much elbow and leg room I will need so that I can then decide on a front and rear width for the new trike.
Since I will be using just a handheld jigsaw, I knew that cutting the thin plywood without splintering the edges would be a bit of a challenge. Cutting along the grain would be fine, but crosscutting would definitely splinter the plywood sheet badly since it is nothing more than two layers of thin veneer glued over some type of filler wood. I did a few test cuts on a scrap piece of wood using several jigsaw blades ranging in pitch from 20 teeth per inch all the way down to 36 teeth per inch. The fine blade with 36 teeth per inch made a much cleaner cut, but there was still a significant amount of splintering around the edges when cutting across the wood grain, but I had a plan.
The splintering only occurred in the topside of the wood when cutting, so I figured it was due to the blade pulling up in the edges and causing the fraying. In an attempt to stop this fraying, I added a strip of packing tape over the area to be cut with the idea that this would prevent the blade from fraying the edges when making curved cuts against the natural grain of the wood. I cut along two random curves, one with the tape and one without and the results were perfect with no splintering at all on the taped section. After the tape was removed from the wood, the completed cut looked clean enough to be a laser cut. Problem solved!
Tracing the wood panel to fit into the angle iron side rail was going to be an easy job. I just laid the side rail onto the wood panel with the welded side out and could now trace around the inside edge with a pencil to draw the cut lines. Since the traced shape would be made slightly smaller that the inside edge of the angle iron, the panel should drop right in after cutting directly along the traced line. The panels will also make the sides very rigid once installed. At this point, the angle iron frames can be distorted if pressure is applied, so the addition of the wood panels will make them extremely rigid.
The 4'x8' panels are much too large for my workbench, so I had to do the cutting on the basement floor. To make this job easier, I placed the wood panel over a pair of 2 inch Styrofoam insulation boards to kneel on and make the cuts with my jigsaw, keeping the blade high enough so that it would not hit the concrete floor. The Styrofoam boards were adjusted as I cut along the line so I could kneel close to the saw and not cut into the Styrofoam. It's important to support the work near the blade to avoid oscillations that could splinter the wood, and this system worked well for that.
Cutting the sides from the wood panel went very well and the tape kept the edges completely splinter free and smooth. The tape peeled from the edges very easily without taking up any of the wood. This tape and cut procedure was a total success. I completed the job by lightly sanding the edges with some fine grit sandpaper just to take off the sharp edges so I wouldn't get a hand full of slivers from handling the panel. Now, the big test: will the panel fit into the angle iron rail?
Yes, the wood panel fit into the angle iron rail just like a glove! Now, I could simply place the cut side shape over the next wood panel and trace around it so that both wooden panels would be exactly the same size. Having both wooden panels exactly the same is important so that any slight malformations in the shape of the angle iron sides can be removed once the panels are inserted into the rails.
Tracing out the second wood panel was easy since it was only a matter of laying the first one on a new piece of wood (face down to create a mirror image). This time though, I would need to cut just slightly inside the traces line since I had traced around the outer edge of the first piece. Once the cutting line was drawn, the entire outline was taped with the packing tape and then cut on the floor, using the Styrofoam boards to support the panel. After cutting, both wooden sides were exactly the same size.
I now had two very strong and lightweight side panels to work with! The wood panels will not be fastened into the angle iron frames just yet, though, as there is still a lot of cleanup work to be done on the welded sections, and I will want to remove the panels for cutting again when I decide on the size and placement of the side window cutouts. The next stage of this build will involve choosing a front and rear width by sitting between the panels to decide on how much elbow and foot room I need. From there, the adjoining rails can be installed and then the roof panel cut. Unfortunately, I don't think I will be able to move the completed body up and down my stairs, so this might present a challenge because it is still much too cold outside to move out of the basement.
Two weeks have past and the snow is really holding on this year. Although temperatures are getting above zero during the day, there are still 2-3 feet of snow out there, so getting to my makeshift workshop is almost impossible. The next step of this build will be determining the overall widths of the body at the front and rear so that the horizontal support rails can be installed, making the body into a single shape. Since this full skeleton will not fit around the stairwell to the basement, I have to switch to outdoor building in the old trailer at the bottom of the hill. To see where I build bikes, check out the mini webisode entitled, "Dodging the Rain", under the Tutorials section. This waiting game between spring and summer is always difficult as I am just itching to get out and build bikes!
I placed the panels over the wheels on the Aurora Trike just for a test fitting. The new trike will be about 3 or 4 inches wider than the Aurora, but I can still use it as a guide to determine the profile of the two sides as they run from rear to front. The widest part of the body will be at the rear, which will be about half an inch wider than the under seat steering on each side. It seems as though I could probably squeeze into the width of the Aurora, but it would be so tight that my knuckles would probably rub on the side panels during a tight turn. With over seat steering, I could probably squeeze into a body this wide, but my goal here is not to make a slipstream racing machine. I would rather have the extra width for comfort and stability.
The difference between the rear width (A) and front width (B) will be at least 12 inches and must begin to taper ahead of the rear wheel wells so that the wheels aren't at a different angle than the framework over them. To create this tapered profile, more slots will be cut into the angle iron framework, but this time on the other part of the angle iron, 90 degrees from the initial cuts that make the side profile. I will probably make eight or more cuts just ahead of the rear wheel and make a slight arc so that the transition is gradual, making a curve in about a 2 foot length.
Sitting between the side panels, it would appear that the optimal rear width would be shoulder width plus 2 inches on each side, and about the same at the front, having about 2 inches between the end of the pedals and the sides. So, now I will have to wait for at least another week for the snow to melt so that I can get the body out of the basement and into the outdoor workshop to complete the horizontal adjoining sections of angle iron. The rest of the velomobile body should come together quickly now and I will be moving on to the base vehicle soon.
2019... Due to my work day commitment and loss of workspace, this project has been put on indefinite hold. Hopefully the info presented here will be of use to those looking for various methids of building a velomobile fairing.
Check Out Some Of Our Featured DIY Bike And Trike Plans...