Enclosure - Lessons Learned

Owned by Nicole Choe
Last updated: Feb 07, 2024
4 min read

Main takeaways + lessons learned from MSXV:

Design:

  1. More talks with chassis regarding space for box

    1. Would have saved a lot of time + hassle if we had more room to work with

    2. Less space for the box = harder to remove box = box had to be manufactured with much less tolerance = more extensive jig that had to be made for the box = wasted time = less space for the modules = more problems, etc.

  2. Find a way to better mount the modules to the enclosure

    1. Saves time.

    2. MS14 had a stand idea which works better, current method requires additional enclosure “feet” needed to lift the enclosure off the chassis which is wasting a lot of time

    3. Consider the tolerance/have more talks w other teams such as chassis, they already made a jig that fits to the chassis frame which would have saved at least two weeks

  3. Need a better way to make sure the air flows through all the modules in the box

    1. we can do better than the baffles solution which is due to the large amount of ceiling space the enclosure has

    2. Consider the fact that no inlet fans can be used. Liquid cooling is cool but not necessary and adds weight, but using pouch batteries may not be a bad idea…

  4. Make the box more to fit the battery pack

    1. Communicate with hardware to make sure everything can still fit in the car, but talks with Rohan (Formula Electric) regarding extra space on the front and back on the enclosure which adds more of a force is something we did not take into consideration

  5. When CADING the 3D printed fan holders for the enclosure, forgot to take into consideration the fillet of the CNC machine (minimum 1/8”), and forgot to tolerance them to fit (had to sand the edges down)

 

Manufacturing:

  1. Talk to the welding guys before making your jig

    1. Saves time - no need to laser-cut MDF for jig or machine holes in metals for weld screws

    2. They can easily weld 90 degree angles and such an extensive jig for the box was not needed

    3. For the enclosure cross-bar jig, less parts = better, and be more efficient, design the jig to do the welding send out once instead of twice

  2. Have everyone do machine shop training so things go more smoothly

    1. Create an extensive schedule. Machining took 1 1/2 weeks, could easily get that down to 3 days if we did not need to make such an extensive jig, cause there would be no need for milling holes in the metal.

    2. Machining - cutting angles in the metal - water bandsaw, cutting sections to make space for the fan-holders - vertical bandsaw, start from the edge and cut towards the middle (link here).

    3. Creating drawings and shadowing for the first couple of cuttings were good practices

  3. Procedure of manufacturing the fiberglass panels

    1. Better ratio of resin to hardener epoxy would have saved time since we did not have to remake the CNC layup and/or guestimate the amount of fiberglass left to work with

    2. Vacuum infusion is better than wet-layup since you can better control the ratio of resin which results in a stronger part. Downside are the holes in the foam, so when CNCing, make sure the holes are not exposed to the edges. In our panel, there was a section in the hand-layup that started detaching from the foam, guessing due to running low on resin towards the end. Also be careful so that no debris gets in (that also happened)

    3. TOLERANCING. Forgot to account for the 2mm (1mm per side) that the layers of fiberglass add to the panels. Had to remake panels as a result - wasted time

    4. Always account for double the amount of time you think it’s going to take you to get things done

  4. Validation

    1. Caught what would have been an awful mistake by testing if the enclosure fits into the metal frame with long M10 screws, need checks to ensure things are going smoothly

  5. Make the frame out of aluminum (with holes) not steel - lighter enclosure

    1. Steel was used due to difficulties with welding aluminum + expensive. Took time to drive- to-and-from the powder coating sponsor, and sanding enclosure before powder coating took a day

    2. Solution: Source out a sponsor.

  6. Perform FEA sims on the cross bars which attach to the chassis - more confidence in design I suppose

  7. Perform a more thorough water-proofing test

    1. Consider using attachments that are not just water resistant but water-proof, or create a really good shroud for the wheel-weld

    2. Soapy water test - any leaks would be identified through seeing air bubbles in the mixture, or we know formula uses the hose room so we could use that

  8. No need for two design reviews

    1. Eric’s enclosure design frankly was already good

    2. Too time consuming - Sprints spanned from September-April 2023 and the design was finalized all the way in July. The prototype was manufactured a year later in September 2023.

 

Other:

  • Fan testing was wasted time as the team already had an idea of which fans they wanted to use from the beginning

  1. More enforced deadlines and a vision for the end-goal would have saved delays and a lot of stress when manufacturing, and also give members more clarity for what needs to be done.