PROJECT SCOPE: Redesign an existing e-bike using CAD tools. Upgrade the looks and the kinematics. Create a CNC-ready 3d model.
The client had already built two bike prototypes when we started on this project. Basis and some geometry had come from Trek Slash. The bikes are equipped with a TQ engine and a 900 WH battery integrated into a visible carbon frame. The engine is not limited and can push the bikes up to 55-60 km/h. That delivers a unique riding experience. Two vertical columns and the huge battery give a robust look. Let's call them Thunders from now on. Although impressive to sight and fun to ride, Thunders were troubled by some technical and aesthetic issues.
Since the bike was not developed using a computer, there were some big oversights. The engine was not aligned to the bike centre line. Alternative holes for mounting were drilled in the engine body. Screws were placed inversely, not in the way the manufacturer envisioned it. Bottom bracket (BB) and the main pivot are at the same point resulting in the need for custom roller bearing (ø 170 mm) that encloses the entire motor. Bearing consists of 5 pieces and around 500 individual rolling cylinders, it is impractical and expensive to produce.
Thunder 22 rear triangle was reworked by engineering firm, so it can actually move for 180 mm. Still, whole setup is very complex and there are no competitive benefits besides that bike can jump from a plane. Analysis of Thunder 22 with Linkage1 software also revealed that suspension has some other issues. Anti-rise is around 45% and anti-squat goes from 55% to negative 80% in regards to the wheel travel. Those are not the most preferable values. Second shock point is fixed on to the chain-stay, making the leverage ratio nearly linear. The geometry was relatively off, since the modelling by hand stole a degree or two on the seat and head tube.
Vertical columns (seat-tube) had caused a challenging seat integration. That is why the bikes have a butt. Edges of the main surfaces are inconsistent and the form of the Head-tube is unclear in some areas. The down-tube is dull and squared.
The plan was to figure out a new kinematics system and then create a basic CAD model, with all the important points, proper geometry and elements (wheels, seat post, head-tube...). “Tech-model” would later serve as the basis and guide for the form modelling. The clients wishes were to preserve the spirit from the Thunders. Two vertical columns and the continuous line that goes around the nose. Size of the battery was to stay the same. Rear wheel must be able to travel 180 mm and the bike top speed around 50-55 km/h. Those were some design constrains and any design intent would have to work with that.
As a solution for the geometry and kinematics, a donor bike was found. Haibike’s Flyon, similarly to the Thunders, uses TQ engine and 4-bar suspension. Majority of points were derived from the Flyon images and data specs that were found on the internet. Budget and time wise, this was the best option. Starting size of the frame was XL since it would fit the most of people. Linkage Software provided a .dxf file with all the points. That gave the basis for the technical model that was later drawn in Autodesk Fusion 360. Parts like fork, wheels, seat ...etc were obtained from GrabCAD library.
I knew that a bike should look more like a car and less like a sports toy. At the time I played around in Speed-form and wanted to learn how to work in Alias. After about a month of playing in Speed-form, I figured that it will not be enough and that I should start modelling in Alias. Everything I read, heard, and saw on starting that path was very discouraging. Luckily something bugged, Speed-form started crashing. At the same time, Alias 2020 came out with a sub-D workflow. That made the transition so easier. Being an experienced Rhino user helped, but nothing could prepare me for Alias. I imported the technical model from Fusion and started modelling. It took me about 5 months to complete it (with some longer and shorter pauses).
I think I watched all the tutorials I could find over the internet. I was determined to draw it all as a class-A surface. I started with sub-D tools and then gradually moved towards more traditional methods of sculpting. The biggest challenge was to stop designing and move on, to finish it no matter what.
Battery and the passive cooling system.
*Update 2022 - Noticed that the battery design is faulty. There should be 14 S and 6 in parallel. This one is 14P and 6S.
At the start of the project, we obtained the Prusa MK3s printer kit and assembled it. I played with it and in a few days already knew a lot, even how to unclog the nozzle. As the model in Alias progressed, I slowly started making 3d prints of the pieces that were "done". At some point, I had half of the bike printed and another half not even drawn. So part of the Alias model stayed as sub-D and part is more or less Class A. I would export shells from Alias as a .step file format and then import it in Fusion. From there I could make solid models and later split them where and if needed. Fusion can directly communicate and send files to slicing software.
The finished 1 : 1 scale model gave valuable insights. Even though the print resolution was crappy (0,1 mm) it still gave a super clear sense of what the final surfaces will look like, and what needs remodelling. The 3d printed model helped to inspect many things. For instance: will the bearings fit? Will the kinematics work? I even printed out a model of Fox shock and put a large spring inside. "It worked like a real one".
This was my first time designing such a product. The process educated me about the transportation design, not to mention how much time can you lose if you don't think about tolerances. I got a good grip on why form and function must work together. It is totally clear to me now why do some companies often have teams of people collaborating on a project this size.
As of February 2020, the project rests as some other matters are being resolved. (+ this corona menace)
Thank you for Your time!
