Monday, March 26, 2012

Battery Pack design

Anyone who has followed my build has seen my previous battery plans. Initially, I had anticipated using 10Ah Headway cells. Then I switched to the blue 8Ah cells, since they yielded a lower internal resistance meaning less losses to heat and inefficiencies. Then, I switched to 8Ah cells with an even lower resistance, red 'high power cells'. 
The game plan was to mate 972 of these cells together, 9p/108s, for a nominal voltage of 345ish and 2000A bursts. 
I was almost certain the headways were the route I was going to take, however I decided against it in favor of some of the highly acclaimed A123 20Ah pouches. 

Alright, why? Because the cells are pretty awesome. Although their origins are unknown, they still boast some pretty impressive specifications. 

To get the same performance out of this battery pack, I will a 4s/416s setup, for 343v nominal, and 80Ah capable of 2000A bursts. 

Not only that, but the headway pack would weight about double what the A123 pack should weigh.

The only drawback to the A123 cells is their inability to be put into a pack easily. I originally designed a rather primitive pack, consisting of bus bars separating the cells, and the entire pack would be compressed via a threaded rod that would go through the cells. 

While this design would work, I had two issues;
1. The threaded rod would have to be some sort of nylon or ceramic. The nylon rod would stretch under heat, and the ceramic rod would be fragile.
2. The cost of the copper needed for the bus bars would be ridiculous. I calculated about $700 worth, using bars of 1/4" and 3/8" copper.

With that in mind, I went about designing another pack. In order to combat the expensive copper necessary for the first pack, I needed to design a pack that didn't rely on the cells connecting directly to each other, since that would call for thick copper to space the cells. I had always wanted to do something along the lines of the white zombie's pack, having the tabs bend over a copper bus bar, and then another bus bar being clamped down on these tabs. However, this wasn't practical using 4 cells in parallel. After a little finicking with my A123 test cells, I figured I would cheat, and treat two 20ah cells as one 40ah cell. Now I need 2 parallel (which is 4 cells), which would allow me to use the clamping system. 4 cells would need 8 slots, and I dont think the cell's tabs farthest from the central copper mounting bars would reach.

A couple of minutes in CAD got me this 

Simply by changing the design, I was able to reduce the cost of the copper significantly. With the new clamping design, I can use 1/8 copper bus bars, with more surface area. The total copper price for all eight packs will be around $223 (according to 

I then optimized the design, and entered some more accurate dimensions for the final battery pack file. 
I decided to make the pack with a 1/4" polycarbonate top, so that I can see the batteries. It also isn't much more expensive than other plastics. I wanted to use 3/8", however the cell tabs would not be long enough to go through the plastic slots, and attach to the copper bus bars. 
The sides of the box will be 3/8" on the short sides and 1/4" on the longer sides. There will be half an inch of space between the cells and the edges, so there is some room for error. Foam padding will be placed in these gaps to secure the cells. 

Here are some pictures of the final design

I still need to add some copper bus bars to the pack, but I just needed the box dimensions for now. 
The hole drilled through the center of the copper bars will hold them to the polycarbonate top plate, while two more holes will be drilled on the sides of the same bus bar to secure it to the larger bus bar which will hold 2 groups in series.

From there, I contacted a couple of engineering schools around where I live, and found a place to CNC the top plate and the individual sides of the battery box.

I still have more to update, I will update soon!

Thursday, March 22, 2012

Working out the final design kinks

Long time no update.
Truth be told, lots of informative and pertinent information regarding the build has been gathered,but I have been much too lazy to type it all up. So I will now dump it onto the internet!

I am still rather upset at the slow pace at which this conversion is taking place. While I work on the designing the car for countless hours each week, I can't help but feel like I am moving in place. I assume most builds feel similar, but darn. Oh well, with the motors nearly mounted and the battery packs finalized, the so called 'light at the end of the tunnel' can now be seen with a telescope. 

Anyway, here we go.....


A serious milestone was passed when I was finally able to order a coupler for the two warp motors. I called probably 300 different businesses, trying to find a reliable and trust worthy solution to coupling the litte red monsters. The solution to all the chaos finally came when I gave up on locating an appropriately rated coupler, and purchased a $20 generic set screw coupling. 

It basically is a steel oxide coupler with a 2 1/4" outer diameter, and  a 1 1/8" inner diameter, with a keyway. Not much to it, but it is only rated to 2500 RPM and 100 Ft Lbs of torque. My car will be hitting a 5500 Rpm redline, and have a max potential torque of 1500 Ft Lbs of torque. Quite a disparity between the two specifications, however I am hoping for the best. In the worst case scenario, the coupler explodes while on the freeway, sending shrapnel everywhere and destroying anything in its path. But what are the chances of that *knock on wood*??

With the coupler out of the way, I had a U channel made out of a 1/8" piece of aluminum. I then had two 1/4" arms welded to this channel, which would be situated in the original motor mount locations. The large U channel came out surprisingly cheap, ringing in at $60. Seeing how cheap a large piece of custom bended metal was, I returned back to the shop to have the 2 aluminum arms welded onto the channel. I was then hit hard with a $350 bill. I have no idea how two small arms took $350 of welding, regardless I was kinda upset and purchased a new tig machine.

I went ahead and spent a little extra to purchase a better 'starter' model. 
With the new machine, I carried on finishing up the motor mounting. Unfortunately, I soon discovered that having a nice welder doesn't ensure good welds.

My first project was to weld up some steel brackets to hold my differential in place. I used stick welding to do this, and was surprised by how straight forward and forgiving of a process it was. Needless to say, I welded up the two brackets, and had the differential mounted back into the car in a day or two. (more on this soon)

All my good luck was forsaken later that week when I began aluminum welding.  I practiced on some scrap aluminum with relatively good results, however when the time came to weld two of the motor end place onto the U channel, my welding suffered severely. I could not get the base metal to melt, the filler rod would always bead up, I would constantly spoil my tungsten. It was simply a bad start. 

I never really got any better at it, and as a result the welds look horrible. But I suppose it works.
I feel as though my failure was due to the large shape of the U channel, which acted as a giant heatsink wicking away all the heat I was producing. 

Regardless, I was able to horribly weld the front most motor holding place, and I simply tack welded the back plate, so now I can drop it off at a welding shop and have them actually do it professionally. Two more plates will be bolted to the U channel, but will be removable to allow for the movement of the warp motors. 

With the motor positions finalized, I will be able to put the motors in the car and get an accurate measurement for a driveshaft length. 
I hope to have the driveshaft ordered by tomorrow, since my custom driveshafts are ready to ship! (more about this soon also).