jonescg's NEW electric racebike BUILD thread!

[Hillhater]

...For any interested parties - the TTXGP rules have been updated. Maximum permitted system voltage has been increased to 700 V

Anyone have a suggestion as to the logic behind that decision ?
It certainly doesn't make it any cheaper !

 

[Arlo1]

...For any interested parties - the TTXGP rules have been updated. Maximum permitted system voltage has been increased to 700 V

Anyone have a suggestion as to the logic behind that decision ?
It certainly doesn't make it any cheaper !

Igbt's start to get a lot more usable with voltages like that. I'm sure there is some good reasons for some of the engineers to try hi voltage like that. I for one know I want to experiment with higher voltages for some of the BLDC controller drive problems we currently have.

 

[jonescg]

Yep, when I first enquired about the Evo Motors, they suggested I go for the high voltage version and run the controller up to 700 V. But they do make a motor with fewer turns which runs at lower voltages but pulls more amps. So I will go with the ~400 V version. 700 just gives me the heeby-jeebies.

Anyway, I have been sweating over this battery design for moths now, and I think I have a solution. It would expensive in the short term, but cheaper in the long term. This pack design will enable the construction of 6-10 kWh packs without difficulty.

It involves injection moulding. Yes, hell expensive to set up, but once it's done you can crank them out as fast as you like. I have opted for this as I can embed M4 brass nuts with a 6 mm diameter face that would provide sufficient surface for a good tight thread. The whole unit is thinner at the base and has a 2 mm slot cut the whole way across, meaning the cells can sneak forward an extra 4 mm, making it more compact and it means I don't need to add anything to the tabs. The tabs would be punched with a 4 mm hole that is bent over the brass nut, aligned, and the 38 mm x 42 mm x 1.5 mm copper plate is screwed in over the top with stainless screws.

I have made an enquiry with an injection mould place in Adelaide who make all sorts of stuff, and will see what's possible.

Something like this?

A list of nutserts:
http://www2.boellhoff.com/web/centres.nsf/Files/AMTEC-GB-0200/$FILE/AMTEC-GB-0200.pdf
Like I said, expensive at first, but to get such a plate machined out of acetal would cost thousands anyway. Might as well get the inverse of it made out of steel and make a reuseable mould.

 

[Hillhater]

best to brace yourself !!...to make a mould for a part like that would cost mega $$$'s . !

for small quantities, it would be much cheaper to have them water jet cut and CNC machined.
.. or maybe a 3D print job ?

 

[SplinterOz]

Anyway, I have been sweating over this battery design for moths now, and I think I have a solution. It would expensive in the short term, but cheaper in the long term. This pack design will enable the construction of 6-10 kWh packs without difficulty.

It involves injection moulding. Yes, hell expensive to set up, but once it's done you can crank them out as fast as you like. I have opted for this as I can embed M4 brass nuts with a 6 mm diameter face that would provide sufficient surface for a good tight thread. The whole unit is thinner at the base and has a 2 mm slot cut the whole way across, meaning the cells can sneak forward an extra 4 mm, making it more compact and it means I don't need to add anything to the tabs. The tabs would be punched with a 4 mm hole that is bent over the brass nut, aligned, and the 38 mm x 42 mm x 1.5 mm copper plate is screwed in over the top with stainless screws.

I have made an enquiry with an injection mould place in Adelaide who make all sorts of stuff, and will see what's possible.

...

Like I said, expensive at first, but to get such a plate machined out of acetal would cost thousands anyway. Might as well get the inverse of it made out of steel and make a reuseable mould.

Chris I was wondering if you could make that part in two layers? The first layer with the slots and the second with the separators for the connector bars. Your screw inserts could be placed into the lower layer with the slots. I presume there is some really good glues to join the parts together. The upper layer (without the slots) could even be made out of small strips of the material you are planning to use. Could save you a motza.

 

[adrian_sm]

+1. That is a very expensive design as you have it modeled.
- Large plan area mean big tool, and big pressures for molding, meaning it goes in the big expensive machine.
- lots of fiddly shutoff faces between the two halves of the tool, means a lot of high precision tool finishing, adds to the $$$'s

+1 for fabricating it from water jet cut parts if you can. You should still be able to use the nutserts but if you insert them from the bottom side and have them pull up into the hole and use the flange to stop them pulling through the whole you should get good clamping force. But I worry about how thin the plastic wall is at the narrowest between the nutsert hole and the tab slot. This is what will fail if you crank down too tight when clamping the tabs.

If you have you heart set on injection molding try and make a more modular design. The smaller the better, then just have interlocking features to assemble them into the size you want.

Another option is urethane cast parts from a silicone tool. Someone like Solid Concepts should be able to help you. I have used them in the past for similar stuff in the past.

Personally I would try and avoid the nutserts, and make that lower portion that clamps the tab out of a single bit of metal drilled and tapped, with plastic spacers or some such. I need to think about it a bit more. (so this is a 5p 10s configuration you have shown?)

Reminds me a bit of how cell-man was assembling his A123 packs.

- Adrian

 

[jonescg]

Yeah Tony if you look at the way I've drawn it you will see I have two separate bits - the separators and the base plate are made separately. The riv-nuts can be inserted after the fact. There are two extra holes in the middle that would hold them together. Works out to be a lot cheaper than machining the whole thing in one go. The water cutters can do a .75 mm cut fairly accurately, but I was quoted several thousand :? . Likewise I was quoted $3000 to machine the whole thing in one go. Just one plate! So no matter what way I go about it, a solidly designed piece of work like this is big money.

So there is a combination of 'too hard don't want to do it', and some genuine head-scratching time going into such a thing.

That's why I thought about injection moulding. If I'm going to spend thousands on getting one of these plates made up as a one-off, I might as well pay the extra dosh and get a template I can carbon-copy a few hundred out. And it it ends up being a winner, then I can sell some goods for the trouble. After all, one of these packs would make a sweet long-range 36 V 25 Ah pack for an ebike.

If I were to get a mould made, I would go for an earlier incarnation, one where there are few draught angles and I can make the slots thinner to fit the rivnuts. And if the riv-nuts still don't work I know that soldering still works just as well.

 

[jonescg]

And just to confirm that you can screw into the acetal over copper:

If the slot is wide enough, and long enough (48 mm long and 2 mm wide) you can stuff the cells right up there and fit enough tab over the threaded hole to tighten the copper bar down:

But I don't like it much - too easy to break something when you tighten down hard. Maybe solder isn't such a crazy idea?

 

[SplinterOz]

Man that is pricey. I am getting my 4 battery plates for less than $1000 in Acetyl.

 

[jonescg]

Personally I would try and avoid the nutserts, and make that lower portion that clamps the tab out of a single bit of metal drilled and tapped, with plastic spacers or some such. I need to think about it a bit more. (so this is a 5p 10s configuration you have shown?) Reminds me a bit of how cell-man was assembling his A123 packs.

- Adrian

You got me thinking Adrian...

I will come up with a plan that does just this. Reinforced epoxy would probably have the strength to hold the brass bits in place, but not the structural strength needed to make it a supporting member of the pack. So the trick will be to make the pack as self supporting as possible while the front plate serves only for terminations. It will add weight to the pack though. Given brass is only about half as conductive as copper, I might be able to get away with 1 mm copper plates...

There you go AJ, I did come around to fibreglass eventually

Edit:
Howzabouts this idea?

Each unit would be laser cut from 5 mm plate aluminum. I got no idea whether this would cost a bomb or is even practical, but I reckon if the slots are incorporated into each little unit it would make pouring the resin around it a hell of a lot easier. Then the anti-shorting ridges can be attached later, or incorporated into the mould.

 

[Jay64]

I have been trying to get caught up with this thread, I missed a few pages worth of posts when I was gone on my trip. I also was going to suggest something like the EIG setup as I was reading along, but then saw that AW already suggested it by the time I got caught up to the last page. :lol: I will take one of my packs apart today and take some detailed pics. I was really impressed at how simple but efficient these pack designs are. I was a little concerned about trying to replicate it for different packs because of the cost of making an injected mold part. But if you are thinking of making something using that process anyways, doing it this way might be a lot cheaper, since the part is basically the size of each cell, and then the stack into each other. But this specific design only has them connect to each other in a straight up stack, they don't connect side to side stacks. But if you are custom making them for the a123 pouch size, then you might be able to do an addition to have them connect side by side as well. But let me go get a pack right now and take it apart for some pics. I know you have a pretty good idea of what you want to do right now, but thought I would just add this info just in case it helps you in some way. It will also be good to have the pics for other people's info.

 

[Jay64]

I have been trying to get caught up with this thread, I missed a few pages worth of posts when I was gone on my trip. I also was going to suggest something like the EIG setup as I was reading along, but then saw that AW already suggested it by the time I got caught up to the last page. :lol: I will take one of my packs apart today and take some detailed pics. I was really impressed at how simple but efficient these pack designs are. I was a little concerned about trying to replicate it for different packs because of the cost of making an injected mold part. But if you are thinking of making something using that process anyways, doing it this way might be a lot cheaper, since the part is basically the size of each cell, and then the stack into each other. But this specific design only has them connect to each other in a straight up stack, they don't connect side to side stacks. But if you are custom making them for the a123 pouch size, then you might be able to do an addition to have them connect side by side as well. But let me go get a pack right now and take it apart for some pics. I know you have a pretty good idea of what you want to do right now, but thought I would just add this info just in case it helps you in some way. It will also be good to have the pics for other people's info.

 

[Jay64]

Ok, here are some shots of the way EIG did their bat. stacks. I have more pics, which I can put up here if you want, but I don't want to clog your thread with them if you don't. I'll probably start another thread in the batt section.

Here is one 22s stack from the front. On the top in the center you can see the end of a long bolt that has a nut on the top to keep the front of the stack compressed.


Here is how they connect the the rear of the pack. There are 3 strips that all the plastic plates bolt through to keep it together in the back.


Ea. tab has a copper angle welded to them. One is bent up and the other is bent down. There is a copper bar that goes across the stack at each level and that makes the series connection with the pouch above and below ea one.


It looks like the brass angles were tack welded on, but it almost looks like a sort of press weld. Not sure what the technical term for that would be.


Like I said, I have more pics if you want, and I will be trying to put together another thread with maybe some more details.

 

[Hillhater]

Chris,
i dont know if it helps any, but if you are not aware , Jack Rickard over on his EV TV blog is also working on an optimum construction for a "pouch cell pack" and has started a competition for designs. But in the meantime, his tests has shown some problems with screw clamped tabs developing resistance after a while.
It worth a read of the blog 7 comments section, if you have not been there already.
http://jackrickard.blogspot.com/2011/11/elescalade-liftoff-and-a123-lifepo4.html#comment-form

 

[jonescg]

Cheers Guys,

Jay I like the look of those cells. The C rate concerns me though. If they were good for >300 A burst then maybe they would be good.

OK, one more point about cells - I think the corrosion / tightening issue won't go away, which makes me think soldering is a good idea in the long term. I found I could unsolder my 2 mm copper plates by holding a micro torch on the plate and lifting it off with a pair of pliers. Provided there is some means to hold them (say, via the balance lead) and torch them, it will come off without too much resistance. Abraham's idea of drilling holes in the copper where the tabs meet is a good one - as at least you will see for certain that the solder is going where it is supposed to. The idea is that I won't have to do this very often, though. That also means the material I use to make the front plate needs to be fairly heat tolerant.

NOW, something a little different.

Since I am in the mood for spending a lot of money on this bike, I might as well go all out and build a decent reduction for the jackshaft. I initially had the idea of a primary chain driving the jackshaft, which drove the rear wheel. I had several folks tell me that installing a fixed ratio chain drive is tough, and the slop will get annoying. I don't know if this is true or not, but it does sound plausible. So I thought about running a series of gears.
View attachment reduction gears.bmp
This is what Brammo et al. do in order to keep the bulk of the motor out of the way. I know this means getting expensive things made, but it sure would look sweet, and sound a bit quieter even though I'd no doubt lose some oomph through friction. Maybe even get a cast sump made for it?

What do you guys think? Too much trouble?

 

[Hillhater]

I think your gear train is gearing UP not down ! .. need the small gear on the motor shaft .
Remember your gear drive would need tensioners , casing, oil bath, seals, etc etc
.. BUT, i would go with a belt reduction..for quiet, smooth, simple, clean, reduction drive. ?

 

[jonescg]

Ah shit, yeah. Well I blame MS Paint for that. You know how hard it is to get a circle to line up with another circle?? Just pretend the circles are the same size.

Belt drive would need a very deep profile to not slip.

Gears wouldn't need tensioners would they? They'd be fixed in place with a set of double row bearings, and the final jackshaft would still have a pair of bearings as per the original idea. Yes, it would still require a sump, gaskets and lube, but this wouldn't be too hard would it?

 

[Jay64]

Chris, I wasn't suggesting you get the cells, I was showing another option for holding the a123 pouch cells. Seems like the eig cells are impossible to get. I'm thinking that having holding trays like these made for the a123 cells would have a market, since so many people are using a123 cells.

 

[nieles]

why not make the gears usefull and pick a motor with a higher rpm, or reduce the size of your rear sprocket.

Gears wouldn't need tensioners would they? They'd be fixed in place with a set of double row bearings, and the final jackshaft would still have a pair of bearings as per the original idea. Yes, it would still require a sump, gaskets and lube, but this wouldn't be too hard would it?

the gears need to be meshed properly, some sort of tensioner could come in handy.
an other option is to experiment with different meshings on a test plate and use that for your end design

 

[jonescg]

why not make the gears usefull and pick a motor with a higher rpm, or reduce the size of your rear sprocket.

Gears wouldn't need tensioners would they? They'd be fixed in place with a set of double row bearings, and the final jackshaft would still have a pair of bearings as per the original idea. Yes, it would still require a sump, gaskets and lube, but this wouldn't be too hard would it?

the gears need to be meshed properly, some sort of tensioner could come in handy.
an other option is to experiment with different meshings on a test plate and use that for your end design

The jackshaft is due to the fact that I need to sling the motor as low as I can. If I did direct drive from the motor to the rear wheel, the chain would be too long. Also, it would take up the entire body of the bike if I did this. So my strategy is to hang the motor low and run a drive up to just in front of the swingarm pivot.

I would have thought if the sums were done right, then the gears would be fixed in place and you don't need to tension anything :?

Jay - yeah no worries. But I do like their moulded trays. However I think I'll solder these tabs in the end. Done and done.

 

[AussieJester]

There you go AJ, I did come around to fibreglass eventually

DiDn'T Doubt it for a minute CHRiS, fiberglass is..........your destiny Only thing that blows me away is
you haven't made the molds and laid up the glass already so you can start stackin' packs maaan!!!

KiM

 

[Hillhater]

Belt drive would need a very deep profile to not slip.

you would have to select the right belt, but belts are the way most modern transmissions are built.
Look at the big bikes , Harleys use belts for primary and final drives.
I know 8mm belts , 30mm wide are used for 50 hp drives.

 

[adrian_sm]

Just stumbled across this.

Source: http://www.carsguide.com.au/news-and-reviews/car-reviews-road-tests/holden_commodore_electric_first_taste_review?origin=hpc4

"We can buy a standard energy cell and then we have to adapt them to make it into a battery for us. So the first thing we have to do is attach a set of terminals to it. As a high-volume production cell they actually weld all the cells together. That's not a good idea for a small-scale battery because if something goes wrong we have to throw the entire battery away. So essentially, this (the terminal construction) is our technology we've had to develop. You lay a sheet of copper on the ground, lay a sheet of aluminium on top of it and then you put 50 grams of ammonium nitrate on that and explode it."

 

[full-throttle]

Just saw this in the Renew magazine: View attachment Renew118pg77.pdf

 

[jonescg]

Hehe, yes it's a bit dated but reads well nonetheless. I doubt I will race this year, unless someone else wants to race it for me (and pay all related expenses). I'd rather dump the money into my next bike.