Michael Mackay Mclaren and James Nugent show off their 3-D printed bicycle.

In the winter of 2014, the Discovery Centre opened up their two 3D printers for student use. In need of student helpers, Mike and I were recruited to be involved in all things 3-D printing. In the early phases of working with the Makerbot 2 Replicator and while trying to solve teething problems, Mike came up with a crazy idea. Why not build a full bicycle out of plastic?

There were definitely some difficulties to overcome for the project to succeed. Plastic is not a great structural material. In addition to low strength giving it limited structural applications, the printed plastic  also gains built-in weak points where layers are laid down. Thanks to a small build-plate of the Makerbot Replicator 2, the maximum part length is 26 cm. At the time, we were still first year Engineering students with only ECOR 1101 as our limited experience of force analysis.

Mike and I began the research phase by looking at frame shapes and how force travels in a bike. We came up with a plastic bar design we could modify for most of the frame and began to test it for strength. Shape was not the only constraint; the printing method gives some problems as well. We discovered that the Makerbot prints in a cross hatch pattern only on the top and bottom of the part, while the sides are just individual layers of PLA stuck together with weakening lines all the way down the length of the part.  The printer hollows out the part and fills it with a hexagonal infill structure.  We learned that the infill, although visually appealing, had almost no structural strength. This had a large effect on strength due to the number of layers we set the walls to have. The largest contribution to the overall fracture resistance was the crosshatched top and bottom layers sandwiching the easily separable walls.

On to the design phase: Once we had the rectangular bar dilemma resolved, the time came to design the frame.  As it turns out, bike frames are not designed well to be made from 26cm bars.  Who knew? To address this issue, the frame was constructed of two large triangles with a moment arm holding the front wheel. Mike addressed the challenge of attaching the front wheel to the rest of the frame so that the large moment generated at the lone pivot point would not break the brittle plastic. James got to work on the bike’s backside, which comprised of assuring all elements of the frame were attached and sturdy.

The frame was designed by the end of the summer 2014, when we started to produce parts on the printers. We made a modular design that could be snapped together and taken apart if needed. This was a good idea as it turned out!  Then a problem arose we had not foreseen or been prepared for; the effects of thermal expansion on plastic. Due to the thickness of the walls we set for our parts, they did not cool down as fast as otherwise expected. As a part cooled it bent and assumed the shape of a Pringle. It quickly became obvious that a bike made from deformed, Pringle pieces would never work. Our solution was to use alligator paperclips to clamp down the parts as they printed. With that, production flew until the whole bike was finished before the Fall reading week.

Once we had attached wheels and a seat, we were ready to sit on the final product to see if it could withstand the weight of a person.  Ignoring the known weaknesses, like the part where the printer had run out of plastic and not finished the print or the one with a small crack, we decided to go for it.

We performed a weight test in which subjected the frame to a large deformation. Nonetheless, the bike was able to withstand 120 lbs of force. The front wheel was the main cause of the frame deformation as it moved forward a few inches during the weight test. With the frame being tested we now had to examine the wreckage to identify the failure methods. As engineers we take failure as the optimal change to refine and redesign any flaws we find in our original design. In the near future we will be testing again with the hope of finishing before the looping makes us too dizzy to think.

By James Nugent and Michael Mackay Mclaren 2nd year Mechanical