2017-03-01 Battery Pack Design Review
Date
Attendees
Agenda
- Review design goals for the battery pack
- Improve Devon's understading of the electrical requirements of the pack
- Go over proposed design and discuss changes/alternatives
Requirements and Goals
Design Goals
- Easy accessibility to critical areas in the pack for testing, integration and re-flashing firmware
- Highly manufacturable design for reduction of build time and overall cost as well as increased "prototypeability"
- Highly reliable contacts with low resistance
- Safety (reducing need for use of HV isolated tools is a bonus)
Electrical Requirements
- Passive balancing means ampacity of balance wires is greatly recuded - max 300mA (likely more like 50mA)
- Max footprint of chaos and plutus is 100mm x 100mm. Height will be less than 20mm (extra height is mainly due to controller daughter boards)
- At least 4 contactors will be required in the master box, possibly 1 contactor or a fuse will be required in the slave
- The aux batttery may need to be stored in one of battery boxes
Design Proposal and Recommendations
Positioning in the Vehicle
- The original plan was to have the battery boxes positioned in the vehicle as shown below:
- The rear roof pannels will tilt up and the bottom pannel will drop out of the car
- In this configuration the box would be slid out on an angle to be removed from the car (with the person essentially standing "inside" the vehicle)
- Since the box may weigh up to 50 or 60kg, it may still be quite difficult to lift
- An alternative would be to have the boxes slid in from the side (i.e. normal to the view below)
- The HV connector could still be made accessible from the back as the battery box would have been in the configuration above
- It is probably best if HV exits the pack from the rear as this is where the motors are and it is easily accessible
Enclosure Design
- The battery box enclosure will be made of tecklam or an equivalent composite material joined with an adhevie
- The seams will be reinforced by Plastic 90 degree extrusions (possibly PVC or UHMW)
- A removable acrylic or poly-carbonate panel will be added on the box that houses
- The lid would slide off to allow access to the cells while still mounted in the vehicle
- It was decided that is it an unreasonable requirement to need access to the HV section of the battery pack accessible while it is mounted in the car (main lid does not need to be removable in position)
Layout
Devon's Vision
- All connectors and removable panels should be on the perimeter of the box (i.e. front, back and sides)
- The lid (top) and the bottom will be clear of connectors or auxiliary panels
- It less desireable to have connectors dangling off ar removable part of the box and it becomes much more difficult to retain structural integrity in the lid if you try to incorporate connector or auxillary pannels
- Key systems should not be stacked ontop of each other (you shouldnt have to remove all the boards to get access to the modues/contactors)
Proposed Layout
- The internal structure is based off an 80/20 frame
- This gives it good stiffness and makes it easily manufacturable
- The control boards would be housed in an isolated area on the side of the back and would be accessible via a transparent/removable cover
- Fans would be added to the front and or back of the box to facilitate air flow down the length of it (there are significant gaps between the cells in this orientation - best for cooling)
- Current would flow through the pack as shown below
- It may be difficult to run conductors between the HV section and the control boards - more thought needs to be put into routing
- We will likely need more space in each pack for contactors, fuses, aux battery and other power electronics so an extra row could be added to each pack and two modules could be moved to the slave (22 and 14)
- More thought should be put into the orientation of the boards for easy accessibility when in the car
- The orientation of the pack will also affect the placement of the air vents in the aerobody
Fixtures
- To prevent the modules from sliding axially, 3D printed spacers will "clip" into the 80/20. These will also help guide the modules into position when inserting them
- The biggest challenge with regards to fixuring the battery module is the "vertical restraint"
- Modules must be easy to insert/remove but also be held securely
- Possible designs:
- Make "hangers" for the modules - vacume formed plastic brackets that bolt into the 80/20 rails
- Build a flexure into the delrin pressure plate of the module that would clip into the 80/20
- Build a flexure into the 3D printed spacers that would clip into a knotch in the delrin pressure plates
- Add a threaded insert to the spacers and use a screw or a custom quarter turn latch to retain the modules
- Build preload into the lid
- None of the above solutions stand out a immidiate winners - mode design required
Modules
- The module structure is essentially a 3 layer sandwich - end plate, cells, end plate
- The end plates will come in two varieties (flexure and bus bar)
- The flexure are designed to preload against the bus bars when the module are installed in the pack (1mm interference = 15 to 20kg of preload per bus bar)
- The two end plates are essentially housings for the dimple plates and will be preloaded together using nylon threaded rods
- Compression foam will be used to compress the dimple plates into the cells
- We are confident the tolerance on the dimple height can be drastically improved with some rework to the tool and die
- An alternative solution to the bus bar - flexure series connections would be to have tabs taht extend off the busbards and bend back to extend axially from the cells. These tabs would be bolted together inside the battery pack.
- If the dimple plates don't work, we will be forced to go to go with spot welding
Bare Bones Test Plan
- 3 modules will be built and connected together in a mock up 80/20 frame to reflect the design as closely as possible
- The frame will will be mounted on a DIY vibe table and charged and discharged cyclically
- Some components of the BMS could also eventually be tested on this setup
Action items
- Devon Copelandwill finalize the design of the vertical restraints and order parts by Monday the 6th
- Devon Copelandto work with other members to potentially delegate design and build of the vibe table to another member
- Devon Copelandwill work with electrical to build up the first prototypes of the modules and setup test jig
- Taiping Lito continue work sourcing and purchasing batteries.