Instructables

Method for making 3D shapes (used to make a bike fairing)

Featured
Picture of Method for making 3D shapes (used to make a bike fairing)
Fairing_Shot1.jpg
whole_bike_side.jpg
This Instructable shows you how to make a fairing (enclosure) for your bicycle/trike/etc from a combination of sheet plastic and some plumbing items (PVC and some pipe for supports). It goes hand in hand with the wind powered bicycle project that I am working on (to be posted a little later), and represents a continuation of my work on Thinkcycle. The technique I explain here can be applied to a wide variety of different items: greenhouses, fairings for other objects, sculpture, boats, etc. For vehicular applications this also forms a perfect backplane for the attachment of solar cells, etc. I'll cover how to do that in subsequent instructables.

There are lots of reasons you might like to have an enclosure on your bike: they can improve your aerodynamics, help with thermal regulation, serve as structural attachment points (for hanging lights, cycle computers, etc), but in my opinion the biggest thing is that it can provide a solution to the thing I hate most about winter biking, which is rain. At least part of the reason I have pursued this work is that there were no good options for getting my hands on a nice fully faired HPV, and nothing spells freedom like being able to get where you want to go at any time.

The basic idea for how to build this thing is to decompose your ideal shape (for the example I use an elongated ellipsoid as per some of the photos in step 2) into crescent segments using a little math (don't worry, the attached spreadsheet does all that for you, at least for the simple shape used in this tutorial). Then you can cut sheet plastic (0.080" PETG or polycarbonate) into correctly-sized pieces and then tape the outsides of them together. If you're very careful about the outline of the segments you cut, you can generate a very good approximation to the large shape you want (similar to the way the almond-shaped segments of a football are stitched together to make that three dimensional shape). Once the segments are taped together, you can fill a turkey baster up with pipe glue and run a bead of ;pipe glue along the joint. You let it harden, clean it up with a Dremel tool, and use it! For the shape I'm building, I also show you how to attach it to your bike. This is accomplished by cutting PVC T-joints in half and then using pipe clamps to attach them to the bike. The mounts have proven to be kind of tricky to get right, however, and there's some room for improvement in this department.

The motivation for this work was that I wanted to be able to make big three dimensional shapes for a number of projects but was unhappy with the existing techniques. Either they required a big oven (necessary for most vacuum and blow forming techniques) or were messy and generated lots of extra stuff (ex: cutting foam molds for fiberglassing). If you still want to go those routes, the technique I show you here can be useful in building supporting structures for use in jigs and whatnot.

The one note that I should add is that for reasons I will detail in the wind powered bike instructable (I may also call it a "Side wind safe" bike) you have to be careful about the design of your fairing if you are using a front-steered only bike (whether upright or recumbent). Basically, large fairings are unstable in high winds (so you want to keep the cross section from the side low by adding openings or making it very compact) on a regular bike. In the wind powered bicycle project I show you how to build a bicycle that will get around these shortcomings (you have to steer the front and the back wheels), but you can also attach this to a tricycle (just keep the width of the vehicle as wide as possible and the center of mass low). The fairing I build in this example is a test piece and probably too large for everyday use (although you could try cutting holes in it).
 
Remove these adsRemove these ads by Signing Up
toadstool2 years ago
Man, dis stuff looks good !
Grant-533 years ago
Fine piece of work. I have been working on a similar approach for an upright bike. I use a four quadrant ellipsoid with the profile of an inverted NACA wing for some downforce. I use a skin of 1" hex mesh bonded between two layers of tarp material. The spars are aluminum angle stock attached to the frame with U bolts and the ribs are loops of 5/16" ID nylon tubing. The simple two cone version increased my cruise speed 50% and had little steering input with a 15 mph cross wind. I need to have the rear section slide or tilt to get my leg over the seat comfortably.
nickjohnson3 years ago
Thanks for posting all of this. I'm considering buiding the same.

Did you observe any improvement in aerodynamics? I doubt you took it to a wind tunnel, but does it at least feel more aero?

I Notice that you don't extend your fairing down to the ground. Why?

How heavy is your fairing, in total?
Vincer (author)  nickjohnson3 years ago
That's great!
So the big picture is that I'm still involved in a bunch of mathematics relating to fairings (including some Navier-Stokes tie ins), but I have not solved what I need to solve yet because it is super nasty.
My fairing, sans the supports, weighs something like five pounds, and is probably ten pounds with all of the bracketry.

I should caution you that no fairing of this style does well in high winds (the problem I am working on solving), so be careful when you take it out. If you're looking for something that is okay in wind, I'd recommend leaving the side panels off (or at least cutting some holes in them), so that there's some place for the wind from the side to go.

The improvements in aerodynamics are pretty startling if you use the bike a lot, especially over flats.

I'll put up some of my simulation work at some point, I have to do all of this outside of work for patent reasons so it is going a little slow. I'd be happy to recommend some supports if you do go ahead and build it.
NYFRAPRES4 years ago
In actuallity the best type of glue* (*for all intensive purposes as it really isn't a glue) is a cyanopoxy.  It bonds more on a molecular level rather than on the surface of whatever you are looking to bond.  They tend to work on most any material even if they are dissimilar materials.  The one thing about them is that they are expensive.  However a little goes a long way with this product.  Cool Chem is one of the top brands on the market.
callmeshane6 years ago
With the "proper" PVC primer and glue, there exists a seemingly "Oooo it's mystical stuff that resists gluing by anything" mindset... And having been sucked into this limited perspective, I believed it for a while. Having designed some portable weather and crush resistant miniature amplifier cases out of it... I was shocked to find that some joints could be "tapped apart" from just the weight of the batteries (4 x AA packs).... I mean even model aircraft glue together HARDER than this... And I had some sheet steel fall guillotine style onto some PVC grey water piping... the joint and pipe smashed into pieces... - like the parts of the joint also separated cleanly along the glue line.... Given that I have "micro" hand planed the amplifier parts into precision fits... given them a light dry sanding, had given them a good wipe over with primer and had applied a nice amount of glue to the faces and had lightly clamped the parts... I was NOT impressed... I have made harder and tougher joints with shellac and paper... So after some extra attempts to stop the batteries dropping out of the bottom of the box's... (by their own weight only) I then tidied up the old joints, and used the contact cement. The contact cement's solvent/s REALLY dissolved the top layer of the PVC... about 1mm of it kind of went really gooey - the solvent evaporated and the whole joint welded properly together. Like in engineering terms, I am not super sure and have no results from conducting rigerous scientific testing of the joints... and the combination of PVC and contact adhesive - remains a chemical mystery to me.. and what the joints will be like under fatigue and time testing.... But the arbitrary tests of big muscles and close examinations etc... tell me that the PVC joints, welded with yellow cack contact adhesive.. ARE really strong, tough, not brittle or crackable, and they ain't coming apart... Yet all the properly prepared and glued joints, using the REAL PVC pipe glue and primer, could be peeled apart, broke cleanly under impact and fell apart in tension. I'd post some macro-photographs - when I have the time....
Gluing PVC is a piece of cake, if you want a real challenge try gluing polypropylene. I have been messing with Coroplast (like a polypro cardboard) for almost 4 years and I have found exactly 2 ways of reliably sticking it together: 1 Mechanical fasteners: rivets, bolts and for temporary joints: tape (esp gorilla tape) 2 heat sealing with a hot air gun or plastic welder. The heat sealing is harder, but once you get the technique it is very strong. I wish there were a glue that worked 1/2 as well as the heat sealing.
3M makes an adhesive specifically designed for plastics like PP that can't be glued with anything else. It's pricey though.
AdrianReef6 years ago
what if you smash yourself on someone else's car ?? I mean...what is the pain a vehicle of this kind can make to a human driving it ??
Vincer (author)  AdrianReef6 years ago
So the fairing has a lot of safety advantages: when reinforced properly, it's similar to a full body helmet. Basically having one is much better than not in a crash situation provided you handle properly the wind from the side (stay tuned...). Another way of thinking about your question is: would you rather be hit by someone driving one of these or by a car?
due to the number of cars on the streets It's rather difficult to be hit by one of these :-P ...so...I'll probably choose to be hit by a car this time !!! LOL anyway...jokes apart...I was trying to underpin the conversation on the dangerouness of the design. For example...your neck,in the image, seems to be very exposed in a possible collision with the enclosure , and we all know how painful a neck blow can be :-/ pay attention...I'm very pleased by your job..this instructable is awesome !! I'm only trying to give you some suggestions on the safety side of the vehicle ;-) no claims here :-P Best Regards - Adrian ;-)
cjbikenut6 years ago
Very cool I would really love to see more current photos as well as info on your bike. I've been collecting bike frames and such to build SWB recumbent commuter.
Vincer (author)  cjbikenut6 years ago
Will be a couple of months before I'm able to post that stuff. I'm working on the high speed testing at the moment, and there's a fair portion of rather tricky engineering involved. I don't want to post stuff that isn't safe, so I'm afraid you're going to have to wait a little. Having said that, I'd be happy to help with feedback on your design if you want to do a non wind-powered design.
codester6 years ago
I need to make one of these.........
callmeshane6 years ago
Sumfing you may wish to find out... I have used the "proper" PVC primer and glue and I have busted up a few things made from it.... It's crap. Best glue I have found is the Ubiquitious (pick own brand name) mustardy / yellowish contact adhesive. (Sellys, Bostik etc) It really dissolves the surface layers and makes for a TOUGH and impossible to get apart joint. And if the joint is smashed the glue won't allow the parts to separate.
Vincer (author)  callmeshane6 years ago
Cool, always on the lookout for ways to improve the project, I'll try some out. It might also be interesting to see if it does a better job than ABS for the segment glue. Will do up some tabs this week and update the site with my results...
dbarr156 years ago
will the spreadsheet work with different size sides(segments) keeping the length the same
Vincer (author)  dbarr156 years ago
Sure:
If you change the minor axis field that changes the width of the shape as a whole (the width of the ellipsoid you're modeling), you can set it to any number you want. I'm glad you made me look at this because I caught a bug: both of the axis fields in the spreadsheet at the moment represent half the width (this part is clear) of and half the length (I mistakenly have it as representing the whole length) of the finished "ellipsoid".

The curves that you generate are placed onto four sides for regular segments. For the elongated segments you just add the equivalent amount of the elongation to the distance between the cut edges (basically insert a rectangle with a width equal to the elongation and a length equal to half the fairing circumference).

You can also change the Number of Segments Field (right now it's set at 8, which is the number of rulings we're giving our ellipsoid), you get more segments without changing the size of the fairing.
dbarr15 Vincer6 years ago
What if I wanted different segments to be different widths that would all come together. Could I just plot the curves for each width with everything else being equal?
Vincer (author)  dbarr156 years ago
I love this, you're definitely pushing what I have offered in interesting directions! The answer to your question is "yes!", but you are subject to a constraint that I will now tell you about. If you imagine slicing my ruled ellipsoid in half, one of the cross sections looks like an elongated stop sign (for your purposes let's assume we don't elongate it and it just looks like a stop sign), for the case where the number of segments equals 8 (the one I use in the instructable). Now imagine taking that same circle (whose radius is the same as the "Minor Axis" in the spreadsheet) and putting your new variable sized segments down onto it: your new requirement is that the segments (which now appear as lines connecting points on the circle) have to have widths that add up to the circumference of the circle so that when you put them edge to edge the circle closes. Since every edge when the thing is put together is an ellipse (with a major axis equal to the "Major Axis" field and minor axis equal to the "Minor Axis" field), It should still work, although you may have to tweak some of the fields on the spreadsheet so that the segment widths turn out right (as opposed to being automatically generated for you). An easy thing to do before you do anything in plastic is to build a scale model using the spreadsheet and some paper and scotch tape (send us photos! :) ). More advanced homework (since you seem to be up for a challenge!) would be to build a snail shell shape using your technique (it is possible!).
trebuchet036 years ago
What 'bent is that? I expect coast down tests within the next week :p Kidding Nice work :) Out of curiosity - does the rear portion of the fairing deflect much given the long PVC supports? I imagine if the fairing itself weighs too much, the bending moment would be significant (a problem I had with a cardboard fairing and super weak support) :) I added the HPV keyword as I have some related projects up with respect to fairing construction.
Vincer (author)  trebuchet036 years ago
Thanks! The rear portion does deflect a little more than I would like, but I think I just fixed it by replacing the 90 degree elbow you can see in the main support with two 45's (or at least, just helped it a lot). If I were to do it over I would probably use 3/4" tubing below the fairing and 1" PVC on the supports (I may well do that). The supports are probably going to have to get reworked at some point, however, I have some concerns about fatigue in the connections to them, probably the best way to do them is to glue strips of 1/4" plastic to the frame and then screw into that. One of the design challenges has been that I needed to make the bottom portion of the right fairing support detachable so that I can get in and out of the thing (this was accomplished by using two pipe ends and a hinge at the moment but is still being tweaked). The 'bent is a chinese special of some flavor purchased on ebay and then rebuilt from the ground up.
BobS6 years ago
Interesting project! My only reservations are about durability: wouldn't the plastic become brittle soon, after exposure to temperature differences and UV? A bike sometimes falls over, or bumps into things while parking.......
Vincer (author)  BobS6 years ago
Excellent question! The good news is that it turns out that certain types of plastic do okay with external exposure: for instance a lot of the windows used on motorcycles are made out of polycarbonate and PETG (although I have been told that PETG is more dependent on UV inhibitors, so check your supplier), and they do fine with time. It's true that scratches are an issue, but them's the breaks with this world, everything wears out with time... The scratch issue can also be addressed by making critical viewing windows replaceable.
rimar20006 years ago
Awesome, Vincer. Good work.