Battery Pack Disassembled (Frustrating Weekend)
- Thread starter Mark911
- Start date
Not without cutting one out and bench testing it (assuming the Mfg identifiers have been removed). However, I did note that the Positive button has an unusual vent hole configuration. Three rather large symmetrical ports. I've tried to match it to pictures of known cells but I just can't find representative pictures of all the potential candidates (approx. 2700mah, 10A) showing the top cap with enough resolution to make a determination. 
This is likely the cell they are using:
Sanyo UR18650-NSX
This cell would give you a decent boost in endurance (>20%) and still likely keep up with the discharge rate:
Panasonic NCR18650BD
Sanyo UR18650-NSX
This cell would give you a decent boost in endurance (>20%) and still likely keep up with the discharge rate:
Panasonic NCR18650BD
Those particular cells have not crossed my radar screen so far, I'll take a look. The only difficulty in cell replacement is wire bonding to the PCB. The tiny little connections could be from several "non-fusing" process. Ultrasonic possibly, but probably some kind of parallel gap welding process. The wires themselves are probably fuse wires, but the material, length and bend shape can all be replicated and the board has more than enough pad space for a replacement fuse. Just need to develop the weld schedule.
The short term solution is to get a second/third battery pack and swap them out before each race like the Alta guys did at last year's Slammer. Unfortunately, my dealer says he can't get any from Alta at ANY price (they don't have any stock). If anyone reading this knows of a source/dealer to buy a couple packs from let me know!
The short term solution is to get a second/third battery pack and swap them out before each race like the Alta guys did at last year's Slammer. Unfortunately, my dealer says he can't get any from Alta at ANY price (they don't have any stock). If anyone reading this knows of a source/dealer to buy a couple packs from let me know!
spinningmagnet
Member
- Likes
- 6
- Location
- Kansas, USA
As an option, I have seen pictures of fuse-wire being successfully spot-welded to 18650's with a kWeld. This was done in order to avoid the heat of soldering, without there being any way for the garage-enthusiast to use ultrasonic friction-welding.
- Likes
- 2,797
- Location
- Pennsylvania
Is this something that could possibly work to reattach a failed wirebond?
No need for just one/two cells. I'd just solder the darn things being carful to minimize the heat. Soldering to the PCB would be similar to soldering ANY component lead to the board. The pad are already gold plated, just tin and go. My guess is that it's probably a bad connection to the board (either a busted wire bond or lifted pad). If the pad's lifted, just solder a jumper to the same parallel trace on an adjacent cell. Remember, if the cell/cells been "open" for some time it'll be at a different voltage then it's partner parallel cells (you can easily see the PCB traces to determine parallel sets of 6 cells). Suddenly reconnecting it could allow the some cells to "self discharge" until all the voltages are the same again. This works great at very low delta voltages to keep them all balanced, BUT, if yours is off a couple tenths or more you'll probably need to bring that group of cells into balance. This avoids possible damaging current flow through those cells. There's lots of info on pre-balancing parallel cells on the web.
A shorted cell is a different situation and unlikely in this case.
A shorted cell is a different situation and unlikely in this case.
- Likes
- 4,217
- Location
- Lake Havasu City, AZ
- Likes
- 4,217
- Location
- Lake Havasu City, AZ
Should this thread stay in Electrical Powertrain, or be moved into Battery Repair?
- Likes
- 2,797
- Location
- Pennsylvania
It is appropriate in both areas, but this thread in my mind sets a standard for understanding how the battery is constructed, which is essential when talking about repairs. Perhaps let it here and link it in battery repair as a sticky?
- Likes
- 4,217
- Location
- Lake Havasu City, AZ
Done.
- Likes
- 2,797
- Location
- Pennsylvania
I have good reason to believe the R pack used Sony VTC6 cells. Not sure if both the R and A packs used the same. If the A packs used a different cell then I believe it to be the Panasonic NCR18650BD as you mentioned.
TonyWilliams
User asked to be "deleted"
- Likes
- 172
- Location
- San Diego, California USA
Good work here. I would be TOTALLY DOWN with redesigning the battery case for an external cooling system, and incorporating some type of rudimentary internal (onboard vehicle) cooling system.
Let’s chat! 760-798-0342
Let’s chat! 760-798-0342
- Likes
- 4,217
- Location
- Lake Havasu City, AZ
And you two are within an hour from each other too. And you have all those Sony batteries next door. Very nice!
Please build me a 10 kWh battery pack with a programmable controller!
Please build me a 10 kWh battery pack with a programmable controller!
TonyWilliams
User asked to be "deleted"
- Likes
- 172
- Location
- San Diego, California USA
One step at a time! My concerns are not a bigger battery (yet):
1) charge at 2C
2) cooling to charge at 2C
3) cooling to complete a PRO level moto
4) make the bike lighter
5) extend range
- Likes
- 2,797
- Location
- Pennsylvania
TonyWilliams
User asked to be "deleted"
- Likes
- 172
- Location
- San Diego, California USA
The fact that they left a bunch of room between the cells probably meant something like that.
Redwolf
My dog thinks I'm cool
- Likes
- 1,677
- Location
- Brinnon, Wa.
Tony, when you get to step 5, I would be an eager customer for an extended range for my EXR.
I spent about three months running ANSYS transient thermal simulations on my CAD model of the Alta battery pack. First a few caveats; There's little data on general thermal transport properties of the 18650 cell, particularly at or near maximum discharge rates (non-steady state). The required data of THESE specific cells (VTC6 and NCR18650BD) is non-existent from my research. The exact heat generation rate and physical properties such as specific heat and thermal conductivity are critical in any accurate analytic modeling are unknown. In addition, the physical properties of the Alta's four composite cell modules is also unknown. Also, the precise thermal contact resistance of the many interfaces between different components is unknown and highly variable depending on component material, surface profile, compression (bolt pressure), and the type of TIM (Thermal Interface Material) used, if any. Finally, what convection numbers to use for both forced and natural conditions for a "motocross" type environment is not well documented. Last but not least, the cell temperature(s) where Alta begins thermal limiting is unknown as is the method of measurement (surface, internal, physical, calculated, etc).
With that said, after much research I made what I believe to be reasonable "guesstimates" as to all the above and feel my degree of uncertainty is within an acceptable range for the purposes of my simulations.
After physical disassembly and study and literally hundreds of simulations I've come to the following observations and conclusions;
1) The Alta's pack arranges the cells as densely as possible. This is both good and bad. Good from a capacity per volume perspective but not so good thermally.
2) The Alta's "passive" thermal control system does little during an initial (from ambient) high amperage discharge cycle, less than 15 minutes to thermal limiting (an A level rider discharge profile, for example). Most of the generated heat is retained inside the battery and very little (approx. 10%) is rejected through the radiator. This is due to the relatively high heat capacity and very poor and highly anisotropic thermal conductivity of the 18650 cell plus the tightly spaced packaging and relatively little effective heat transfer area for each cell (essentially, the one end with closest proximity to the heat sink). This one interface is compromised further as direct contact to the heat sink is forbidden as this would effectively "short" all the 126 negative side cells together. The "non-conductive" composite cell module, a 2mm alum module plate and no less than three TIMs separate the cell end from the radiator. To add further, the total cell end surface area isn't used, as approx. 40% is non contact and used for routing hot gasses (due to possible cell over pressurization) away from adjacent cells that otherwise might lead to thermal runaway.
3) The Alta's "passive" thermal control system can help delay the onset of thermal limiting during less aggressive discharge drive cycles AND can help reduce cell temperature between charge/discharge cycles, but not by much. Once the cells are "heat soaked" it's very difficult to pull the energy back out, even with fans, etc.
4) Any "wet" thermal solution would need to flow axially, as the physical proximity between cells restricts efficient and effective radial flow. The nature of the packaging design pretty much makes an axial flow solution virtually impossible.
5) An externally cooled "cold plate" mounted between the individual module and the radiator housing does well up a point, but the delta temp between keeping the cells internal temperatures at or below my cutoff point and the ambient temperature is narrow. To work efficiently at higher ambient temperatures would mean adding very large external radiators and the weight hit and complexity just wouldn't be worth it.
Given 1-6, I've concluded that the best strategy (for me) would be the following;
1) Incorporate a cell with a higher discharge rating than the Sony and take the hit in capacity.
2) Throw away the Alta closed composite cell housing in favor of an "exposed cylindrical surface" support structure and design in a source of "forced air" cooling.
3) Design an alternative hot gas "venting" route.
4) Incorporate two "cold plates" for use between motos only to reduce core cell temperatures to approx. 10C below ambient (to about 50F). Along with a modest cell upgrade and some forced air cooling, this temperature should provide enough headroom to prevent thermal limiting even under extreme conditions. The rate of heat removal is as important as heat generation so further analysis is required in this area.
With that said, after much research I made what I believe to be reasonable "guesstimates" as to all the above and feel my degree of uncertainty is within an acceptable range for the purposes of my simulations.
After physical disassembly and study and literally hundreds of simulations I've come to the following observations and conclusions;
1) The Alta's pack arranges the cells as densely as possible. This is both good and bad. Good from a capacity per volume perspective but not so good thermally.
2) The Alta's "passive" thermal control system does little during an initial (from ambient) high amperage discharge cycle, less than 15 minutes to thermal limiting (an A level rider discharge profile, for example). Most of the generated heat is retained inside the battery and very little (approx. 10%) is rejected through the radiator. This is due to the relatively high heat capacity and very poor and highly anisotropic thermal conductivity of the 18650 cell plus the tightly spaced packaging and relatively little effective heat transfer area for each cell (essentially, the one end with closest proximity to the heat sink). This one interface is compromised further as direct contact to the heat sink is forbidden as this would effectively "short" all the 126 negative side cells together. The "non-conductive" composite cell module, a 2mm alum module plate and no less than three TIMs separate the cell end from the radiator. To add further, the total cell end surface area isn't used, as approx. 40% is non contact and used for routing hot gasses (due to possible cell over pressurization) away from adjacent cells that otherwise might lead to thermal runaway.
3) The Alta's "passive" thermal control system can help delay the onset of thermal limiting during less aggressive discharge drive cycles AND can help reduce cell temperature between charge/discharge cycles, but not by much. Once the cells are "heat soaked" it's very difficult to pull the energy back out, even with fans, etc.
4) Any "wet" thermal solution would need to flow axially, as the physical proximity between cells restricts efficient and effective radial flow. The nature of the packaging design pretty much makes an axial flow solution virtually impossible.
5) An externally cooled "cold plate" mounted between the individual module and the radiator housing does well up a point, but the delta temp between keeping the cells internal temperatures at or below my cutoff point and the ambient temperature is narrow. To work efficiently at higher ambient temperatures would mean adding very large external radiators and the weight hit and complexity just wouldn't be worth it.
Given 1-6, I've concluded that the best strategy (for me) would be the following;
1) Incorporate a cell with a higher discharge rating than the Sony and take the hit in capacity.
2) Throw away the Alta closed composite cell housing in favor of an "exposed cylindrical surface" support structure and design in a source of "forced air" cooling.
3) Design an alternative hot gas "venting" route.
4) Incorporate two "cold plates" for use between motos only to reduce core cell temperatures to approx. 10C below ambient (to about 50F). Along with a modest cell upgrade and some forced air cooling, this temperature should provide enough headroom to prevent thermal limiting even under extreme conditions. The rate of heat removal is as important as heat generation so further analysis is required in this area.
Similar threads
| Thread starter | Title | Forum | Replies | Date |
|---|---|---|---|---|
| R | crf450r 40kw with watercooled battery pack | DIY e-Motorcycles | 35 | |
| M | Alta battery pack rebalance capability | General Discussion | 8 | |
|
MSDS for battery pack | Europe | 2 |