Battery Enclosure Design Considerations

[Ananas]

Hello, this is my first post on the forums.

I am currently designing and fabricating an e-moped using a QS 12" 3000w 260 V4 hub motor and Votol em-150 controller. I am welding a rectangular tube frame and swing arm and using pit bike front forks. I plan to make a 72V 20s8p battery with EVE ICR18650 cells from Battery Hookup and potentially an ANT bms.


Figure 1: E-moped (MopEd?) complete assembly


Figure 2: Battery box mounting to frame tabs and down tubes.

Figure 3: Battery box cutaway showing cell arrangement. Note cooling fans have been eliminated in current design plans.

I've been going back and forth on two potential battery box design methodologies. My current design plan uses a laser cut acrylic enclosure that is slotted together and glued, with acrylic tabs that bolt into tabs welded to the tube frame as shown in figure 2. This design will be simplest for me to fabricate and will have built in cooling slots to keep battery temps low. I have several concerns with this design: my first is that the mounting methodology is likely to fail, with cracks forming along the plastic mounting tabs, my second is that in the event of a battery fire the box will do nothing to slow/contain the flames causing potential damage. Lastly, I worry about water entering the battery pack if riding through puddles or rain.

I've calculated that under worst case acceleration draw of 100A (over twice normal operating conditions), I will generate about 750 watts of waste heat inside the pack assuming average cell resistance of 75 mOhms. I live in a very hilly area with lots of traffic lights to start in, so I will likely be using the motors 3kW continuous (43 A) and peak 6kW (86A) draw frequently.

Looking at designs for high performance battery packs, Electro and Co for example, I see that their batteries have airtight sheet metal enclosures instead of plastic like smaller e-bikes. I created a rough sheet metal design for a lidded metal enclosure that is riveted together and held to the frame with clamps and latches instead of bolts, since I can no longer tab the sheet metal as in my plastic enclosure. This design solves my concerns over durability and water proofing, but I think the lack of airflow increases the likelihood of overheating the battery as well as condensation forming inside the battery box and having no way of escaping. I have the equipment to fabricate this box but I have less experience with sheet metal.


Figure 4: WIP design for sheet metal battery box.


I would like to call on the advice and experience of the forums in choosing a better design for my enclosure, plastic vs metal, air flow vs airtight, etc. What cooling methods or lack thereof are generally used in bikes of this caliber? Are metal enclosures actually airtight or do they have hidden cooling holes that I am missing? Can I passively radiate more heat than a legal ebike is allowed to have power in my area? How can I make my battery box bombproof and element proof?

I appreciate the advice and hope to post more on this project to the forums soon.

 

[Dui ni shuo de dui]

Hi mate,

Instead of fighting the symptom (heat) I suggest to start with solving the cause, which is internal resistance.
This can be achieved using better cells or simply putting more of them in parallel. The cells you plan on using aren't really appropriate for this use, they are limited to only 7.5Amps discharge which is really low power. They don't even have a good capacity either. They are unlikely to last very long no matter if cool them or not. I'd really try to go with better cells if I were you, I suppose you got some kind of deal on these or something?

If you have no other choice but to stick with these cells, you seem to have a giant hole in the frame so you should be able to fit a lot more. I suggest you try using all this nice real estate, it would be beneficial in almost every way except maybe a marginal increase in weight. The battery will heat a lot less, you'll have more range, it will last longer before needing to be replaced and it should allow you for a lot more peak power.

The battery always needs to be the most powerful component of any build.

As for the enclosure, anything can work. I've seen people successfully doing the acrylic cases you describe, other doing metallic boxes which worked great as well. I chose to do mine with carbon but it was a whole lot of work so I dont really recommend going this way.
Venting and condensation is a real issue indeed, especially if you leave your bike out for the night. Someone from the forum suggested using gore tex vents, they are like some kind of check valve for humidity. I haven't tried myself, but it seemed to me like a great idea.

Personnally I tried to make the box as air tight as possible, because I don't want water to be able to find its way in during a storm or whenever I wash the bike.
A good compromize I think would be to shrink wrap the pack, then put it in the enclosure. That way if you have condensation in the enclosure at least it is less likely to drip on the cells. That's how I would do it if I had to build a new box.

 

[gromike]

You know, think I think Dui is right. Building a motorbike pack out of low amp 18650 cells is a tough row to hoe.
See this epic thread about building a 1989 Kawasaki EX500 (Ninja) - Second build, slow and steady .

 

[neptronix]

The weakest link in a battery pack is the cells in the center getting baked by all the other cells around them. No amount of cooling the external walls of the enclosure will help you.

It's better to not make the heat in the first place and either oversize the pack by a lot so that you get no more than 0.15v drop per cell under full load, or chose a way higher output cell ( Molicel P45's are pretty damn nice )

You will also be rewarded with lower voltage sag, much longer battery life, a much lower chance of fire, and closer to the full capacity of the cells if you make some generous overestimations for the output of the battery, because cells are typically over-rated.

 

[Ananas]

I must have overlooked my peak current values when determining if those cells were appropriate, 7.5 amps discharge is definitely far too low for my intended application.

I am thinking about instead switching to Molicel 21700 P42A cells with 4200 mAh capacity and 45 amps peak discharge. With these larger high power cells I’ll only need to run 20s5p so I can recover some of my budget from these pricier cells.

I should note I only intend to get 15-20 miles of range with this bike since this will be used for short class commutes and hops between campus buildings. I currently average 2-3 miles of travel daily on my gas powered motorized bicycle and I want this bike to fill the same role albiet quieter and torquier.

I’ve finished the structural welds on my chassis, in a couple weeks I’ll be able to water jet my tabs and finish the tab welding and swing arm. Unfortunately my spot welder and tig welder are 100 miles away so I am currently limited to designing and 3d printing my accessory components.

This is now a build thread, I will elaborate on my design goals, process and methodology in the near future.

 

[harrisonpatm]

Excellent, looking forward to the build