68Kw/140Nm DCFC capable street motorcycle prototype

[frk1206]

Hi All!
Wanted to share with you folks a crazy ambitious (to me) build to create a super fast comfortable street bike that could do 40-50 miles on a charge and be able to utilize the growing number of L3 CCS/Chademo stations to get 20-30 miles of range in about 15 minutes. Yet weighing under 400lbs so that its actually fun to ride. Calling this a prototype since if it actually works and there is interest i'm thinking of doing it as a startup/commercially. There is nothing else out there that does this - Zeros are heavy and only support L2 while Energicas are heavy AF and support 1C max. Neither is fully usable without incurring a lot of mental load in trip planning. I'm trying to build a tesla of motorcycles where you just don't think about the battery - since charging is SO fast and painless.

Currently trying to figure out if I use an existing frame (ninja 650 or so) or get a custom frame built locally. Thinking of going all out with the parts here mostly cuz i'm a nerd as well as a adrenaline junkie:

- Emrax 208 Liquid cooled motor
- Cascadia CM200DX
- Custom 94S5P Liquid Cooled Battery pack using 21700 LG/Pan/Tesla/Molicell cells. 360V nominal / 9kwh cap
- Orion 2 108-cell BMS with Chademo support
- Custom stuff for CCS support

Putting this out there for some comments / feedback on what you guys think of the build. I've done EV conversions before but never a motorcycle and never something with that much voltage.

My questions are mostly centered about battery pack creation:

- Whats a good spot welder to use? Should I get the amazon/chinese ones or is Kweld the best? Don't mind paying a bit for good stuff.
- Whats a good battery / vendor to use for 21700 cells that would give me 10C discharge / 2.5C charge? Molicell seems to be the only one thats available and touting 3C charge. Any experience with these? Seems like a lot of folks use them nowadays?
- I'm looking to copy the tesla modules in their construction / water cooling. Any tips for manufacturing this pack? I'm thinking of 3-d printing the battery holding structure and then getting water-cut aluminum plate for the connections. The aluminum plate will have holes which would let me weld fuse wires to this plate. Kind of exactly what tesla does.

- I'm also looking to copy Tesla's water cooling. I need water cooling since i definitely need to charge at 2-2.5C. How do I get Tesla style bent copper pipes fabricated? Whats a good material to cover the copper tubes with that'll isolate them from the batteries but still have thermal conductivity?
- How do I make sure I don't die while building this? Going to get rubber high voltage gloves and boots and tools - anything else?
- Does it make more sense to build the battery string in series? so something like 24S5P Battery modules wires up in series? Whats the best way to connect them in series? Just wires or solid bus bars?
- Do you recommend I put a circuit breaker in each module so theres no sparks flying while connecting them in series with bus bars?
- Do you guys use a thermal camera to check how hot the battery pack is getting? If yes is there one you recommend?

Here's a (bad) cad drawing of all the big stuff on a Yamaha FZ1 frame. This was hand measured (and hand drawn) by me so bear with me Future ones would be 3d scanned for accuracy. For orientation the motor is where the engine output shaft would be. Controller on top is where the tank would be. This one only has 4 battery modules. Most likely the motor is going between the swing arms though so there should be even more space:

https://cad.onshape.com/documents/5e50b1ea9e8aaf8a7f80a0ba/w/a083a3c302f8adf693aa27c8/e/23813a56fe78272c0e643803

 

[amberwolf]

FWIW, LiTi (lithium Titanate) cells can charge pretty fast without liquid cooling. Less energy dense, so you'd need a bigger pack (or the same size pack would have less range) but you can charge it quickly, which makes up for it if you have a lot of fast chargers around.

 

[dougf]

I have used a variety of cylindrical cells including molicels p42's and p26's. IMO at least for model planes the molicels are hard to beat for C rating during discharge, I tend to get more W/hr out of them compared to cells like vtc6's and 40T's. However Ive never been in a rush to charge them. All of these cells suck compared to lipo pouches in regards to c ratings though. Your 9kWh claim for 94s5p is a bit optimistic. I get ~6kWh from a 24s17p pack of p42's and I draw a peak of maybe 4-5c.

You can definitely get cooling/heating parts for batteries,

https://www.trumony.com/Pressed-18650-battery-cell-off-road-vehicle-3003-channel-flow-cooling-aluminum-water-cooling-plate-pd48032484.html

Ive also read that a lot of the heat from the batteries can be pulled out thru the end plates. It could be possible to machine battery sandwich plates that have watercooled bus bars.

Doug

 

[frk1206]

FWIW, LiTi (lithium Titanate) cells can charge pretty fast without liquid cooling. Less energy dense, so you'd need a bigger pack (or the same size pack would have less range) but you can charge it quickly, which makes up for it if you have a lot of fast chargers around.

Indeed! But they seem to be 4 times less energy dense than LiPo which is a giant bummer for a motorcycle

 

[frk1206]

I have used a variety of cylindrical cells including molicels p42's and p26's. IMO at least for model planes the molicels are hard to beat for C rating during discharge, I tend to get more W/hr out of them compared to cells like vtc6's and 40T's. However Ive never been in a rush to charge them. All of these cells suck compared to lipo pouches in regards to c ratings though. Your 9kWh claim for 94s5p is a bit optimistic. I get ~6kWh from a 24s17p pack of p42's and I draw a peak of maybe 4-5c.

You can definitely get cooling/heating parts for batteries,

https://www.trumony.com/Pressed-18650-battery-cell-off-road-vehicle-3003-channel-flow-cooling-aluminum-water-cooling-plate-pd48032484.html

Ive also read that a lot of the heat from the batteries can be pulled out thru the end plates. It could be possible to machine battery sandwich plates that have watercooled bus bars.

Doug

Thanks! You are right my math was wrong - probably gonna get around 7.8kwh at 100% charge. I actually did run into that trumony site and have inquired about exactly that part. Any ideas what thermal material I can electrically insulate it with from the batteries? Extracting heat from the end plates is a good idea and am hoping wrapping the aluminum around the outer casing works. Extract from the top plate is dangerous I would think.

 

[SlowCo]

Great project!
I can't help but hope you'll post your progress in this thread for us to follow. Good luck and enjoy designing and building :thumb:

 

[Haggan90]

Looks like a cool project!
If you tend to use aluminum bars as connection betwwen cells you can't use a spotwelder, you have to a use ultrasonic welding machine and bond-wire. I also think it would be better to col/heat the modules on either positive or negativ terminal, that's usually the part that get's hot

 

[spinningmagnets]

Before buying a welder, research the "copper nickel sandwich" method. Calculate the material to be used and the thickness needed first.

I am of the opinion that the kWeld is the best, and even though it costs more, it can easily be sold for a decent price after your pack is done and tested.

 

[larsb]

Are emrax motors really worth the hype? I’ve read a few threads where people couldn’t get the max power out of them and had controller issues.