The battery will be constructed of 280AH cells. It will be installed behind the driver side rear wheel. Four rough concepts (A, B, C, and D) of the design are illustrated below.

Concept A

Concept B

Concept C

Concept D

Cell Electrical Insulation

According to the Will Prowes' forum, the aluminum cased cells have positive terminal voltage on the outer case. This means each group of cells will need to be isolated from each other and from any structure. This will be accomplished by Kapton tape.

Cell Compression

TLDR: Concept D should provide adequate compression and is the most compact and simplest design. It also eliminates the gaps between cells (vs concept A) potentially allowing for off the shelf busbars while minimizing overal length (at the expense of height).

EVE engineers recommend approximately 12 PSI, with no more than 17 PSI, to be maintained on the cells. Allegedly this increases rated cycles from 2500 to 3500. They also identify approximately 1 mm expansion between 0 and 100% SOC. Concepts A, B and C accomplish pressure and expansion allowance by threaded rod preloaded with belleville washers or springs arranged to give adequate pressure over the expected displacement. Aluminum or steel sheet are used used for the compression side plates. This forum post has lots of good information. Plan for about $100 in belleville washers.

However, a fourth concept (D) occurred to me while evaluating A, B, and C. The flex of the side plates themselves function as two springs in series with any other springs in the system. By designing the plates appropriately, they can be relied on exclusively for compression at high SOC and to accomodate the expected growth. To apply some compression at low SOC, single bellville washers, fully flattened, can be used. While this method will not be as calibrated as the others, it can guarentee some compression above 0 and less than 12 PSI at full SOC and is more compact, simpler, and cheaper. This belleville washer fully flattened will give about 4 PSI at 0% SOC for this design. This belleville washer fully flattened will give about 2.5 PSI at 0% SOC for this design.

Empirical Data

In this post, user @cinergi measured growth on 4 cells arranged in a similar fashion as they charged from 0 to 100%. His springs were compressed to 30mm. At 100% SOC he measured 1.0, 1.4, 1.4, and 2.0 mm growth over the four cells. Between terminals he measured 0.15, 0.14, and 0.19 mm growth.

His spring rate specification is 380 lbf/in per spring with a free length of 1.5 in. Compression to 30mm would correspond to spring compression of 0.32 in (1.5 - 1.18). That level of compression would correspond to 486.40 lbf (380x0.32x4) and 8.799 PSI initially. The final level of compression would be 571 lbf (380*((1.5-1.142) + (1.5-1.126) + (1.5-1.126) + (1.5-1.102))) and 10.339 PSI.

Design for Service

100% SOC Condition Compression

Cell dimensions are 6.85 x 8.07 in. Cell area is 55.279 in2. With four cells in the arrangement, the total area under compression is 221.12 in². Assuming 12 PSI, this yields a total force of 2653.4 lbf.

Assuming 15 points of compression, this yields a force per compression point of 176.89 lbf.
Assuming 10 points of compression, this yields a force per compression point of 265.34 lbf.
Assuming 8 points of compression, this yields a force per compression point of 331.68 lbf.

To achieve this compressive force the spring must have a load at solid height of at least the above value for each case. If this parameter matches closely, setup will be easier by simply tightening to the solid condition at 100% SOC.

0% SOC Condition Compression

Cell dimensions are 6.85 x 8.07 in. Cell area is 55.279 in2. With four cells in the arrangement, the total area under compression is 221.12 in2. Assuming 9 PSI, this yields a total force of 1990.0 lbf.

Assuming 15 points of compression, this yields a force per compression point of 132.67 lbf.
Assuming 10 points of compression, this yields a force per compression point of 199.00 lbf.
Assuming 8 points of compression, this yields a force per compression point of 248.75 lbf.

Growth Allowance

Assuming growth is 0.079 in (2 mm) from 0 to 100% SOC, the spring rate to achive the desired compression at each SOC is:

Assuming 15 points of compression, spring rate is 559.75 lbf/in ((176.89 - 132.67)/0.079).
Assuming 10 points of compression, spring rate is 839.75 lbf/in ((265.34 - 199.00)/0.079).
Assuming 8 points of compression, spring rate is 1049.74 lbf/in ((331.68 - 248.75)/0.079).

To achieve this range of compression the spring must have a spring rate of no more than the above value for each case and allow at least 0.079 in of compressive travel.

Candidate Compression Springs (must be 1 in or less in free length)

For 15 points of compression, the following spring meets the above requirements:
- LHL 625D 01 at $3.66 per spring from here. This could use up to 5/16 all thread. It would provide 12.394 PSI compressed solid at 100% SOC and 9.018 PSI at 0% SOC.
For 10 points of compression, the following spring (of many) meets the above requirement:
- LHL 750C 01 at $4.93 per spring from here. This could use up to 3/8 all thread. It would provide 10.786 PSI compressed solid at 100% SOC and 7.9458 PSI at 0% SOC.
For 8 points of compression, the following spring meets the above requirements:
- LHL 1000C 01at $6.40 per spring from here. This could use up to 1/2 all thread. It would provide 11.939 PSI compressed solid at 100% SOC and 8.7952 PSI at 0% SOC.

Candidate Extension Springs (Concept C) (must be ~11 in or less in free length)

For 8 points of compression, using doubled up springs at each point, the following spring meets the above requirement:
- 3630N389 at $10.06 per spring from McMaster. It would provide 12.31 PSI at full extension at 100% SOC and 11.75 PSI at 0% SOC.

Deformation (Concept A)

Calculations performed using clearcalcs moment of inertia and beam calculators.