Re: Re:
Nicko. said:
ScienceIsCool said:
I just showed that a naive implementation puts 30 Watts in that envelope.
John Swanson
I think I have asked you nicely seven times now to show how you would generate the torque, configure the gearbox, handle the temperature, interact with the freewheeling cassette, avoid public breakdown, etc etc, all within a non-conspicuous
road hub.
You keep doing a volumetric scaling of generic motors, never adressing the details (where the devil resides).
edit: removed insult. sorry.
Yes, yes. You want me to design something to an arbitrary spec. But that's not how you do R&D. First you do a feasibility study. Check. The physics is feasible enough. Then you have to do some of that R before you do your D. Then it's an iterative cycle of R & D.
First up is that there's lots of areas to explore:
- The motor has to work for only ~100 hours of operation. Not 100,000 hours like a generic industrial motor. What can you sacrifice?
- How are typical motors constructed? Time to reverse engineer
- Materials! How do you best use them? A glaring example is to sacrifice weight and use tungsten everywhere. That would allow you to shrink the thickness of everything and open up some volume.
- What happens if you ditch the gearbox and way over-drive the motor to get the torque you need?
- Again materials! What happens if you ditch copper wiring and go for performance rather than cost?
The name of the game is to gather empirical data that isn't publicly available anywhere. Use that data to generate trends such as more stator poles = better; or 320 V with gold wiring is the peak of performance; or wound rotors and stators > any permanent magnets or reluctance. Basically, find the peak of performance for every design element.
Now sit down and generate your first integrated motor design. Again, for bench testing and not necessarily something you'll slip into a DA hub.
Test and iterate design until spec is met.
Time for Beta! Build up your design into a hub and get out on the road. Attach sensors everywhere and start breaking the thing.
After a few iterations of that you have a product.
And just so you know, I do have experience in an R&D environment doing exactly these kinds of things. The biggest one was inkjet printing. A small team of us started from scratch and in a couple of years we demonstrated full color printing at 300 dpi at a paper speed of 10 m/s. Scalable to a 3 meter paper width. That tech became the Kodak Stream product line.
John Swanson