Luke's 2001 NB MX-5

Post up a thread for your EV. Progress pics, description and assorted alliteration
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

I got these back today (busbars are tin platted and the bolts nickel plated), literally a 24 hour turn around and only $50 cash for the job. Weirdly I haven't handled physical cash in almost two years now.
Plated Parts.jpeg
Plated Parts.jpeg (972.47 KiB) Viewed 404 times
Next step is to pressure test the chill plates with the Davies Craig EBP40 pump I have. If they don't spring a leak then I can start the final assembly of the battery packs.
Bukes
Groupie
Posts: 41
Joined: Tue, 26 Oct 2021, 19:15
Real Name: Alex

Re: Luke's 2001 NB MX-5

Post by Bukes »

Man I am impressed with the pack build. Looking really tidy
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

Bukes wrote: Wed, 10 Nov 2021, 19:29 Man I am impressed with the pack build. Looking really tidy
Thanks! It's super slow going as there has been plenty of iterations on design and materials, but having everything (well mostly) drawn up in CAD helps a lot.
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

A little update on things. And a bit of a set back...

I received my Davies Craig EBP40 pump a couple of weeks back and got a few hours to run a pressure test of my largest chill plate. The pump runs at 37 liters per minute and just under 1 bar (12 PSI to be exact). This pump will be feeding four chill plates, three for the three battery boxes and one very small plate for the controller. So the pressure will obviously be lower than a single plate but I figured if a single plate can handle the pressure then four at once will be fine.

The idea with the battery cooling system is to run the three battery chill plates and controller chill plate off the same system in a parallel arrangement looped through a heat exchanger connected to the air conditioning system so I'm able to cool below ambient temperature. We get plenty of 30℃+ days and some 40℃+ days too. I most certainly don't want 50℃+ batteries. I may need to think about heating in winter (hah, who am I kidding, we're still getting ~8℃ degree mornings in spring). But that should be fairly easy to plump in line.

I had planned on using 19mm main lines for the cooling system, going down to 13mm lines for the battery boxes (1/4" BSP fitting) and 6mm lines for the controller (1/8" NPT fitting) but the lines seem a little small so I'll likely run 16mm for the battery boxes (I managed to find some locally produced 1/4" BSPT to 16mm hose tails, not easy to find) and 10mm for the controller. My thinking is even though the entry point is the same internal diameter for the 13mm and 16mm fittings as well as the 6mm and 10mm fittings there will be less restriction in the supply lines. A slight increase in weight though...

Anyway, this was the test setup:
Running the Pump.jpeg
Running the Pump.jpeg (828.23 KiB) Viewed 296 times


The hose fittings were dripping a little even though I'm using tapered hose tails. I didn't want to tighten them too much as these were the 13mm items which are already nickel plated (the 16mm need to be sent off for plating). I think a little bit of sealant tape should fix the problem. But the real issue is there is too much flex in the polypropylene lids on the chill plates, this was happening on the exit side of the chill plate. I think the restriction on the end fitting is causing some pressure to build up and.... Leaks 😓
Leaking Chill Plate.jpeg
Leaking Chill Plate.jpeg (473.44 KiB) Viewed 296 times


Soooo, it looks as though my idea to use a composite lid has gone down the drain. Plastics also have the issue of heat expansion which may have increased this issue. The problem with aluminium is at the same thickness the weight almost triples from 990g for the polypropylene. However, after sleeping on the problem I thought "Hang on, aluminium can be bent and by the same place that'll cut it for me...". Sooooo, instead of using 3-5mm thick aluminium I can use say 1.6mm and just have the ends bent 90° to create a tray of sorts. This solves both the weight (now 970g) and strength issue.

I've got PLENTY of spare 1.6mm aluminium heat sinks so I decided it was a great chance to test out my newly purchased Dremmel (for pulling apart my failed battery packs) to cut some aluminium up. Here's an example of a small section of heatsink turned into a lid. The markings are where the screws sit, as tight as I could go without damaging the treads. I think the heat sink is a bit too curved right at the edge to show the compression in the cork. But it is consistent at least.
Aluminium Test.jpeg
Aluminium Test.jpeg (788.32 KiB) Viewed 296 times
And what the end result should look like.
Aluminium Lid Render.jpg
Aluminium Lid Render.jpg (241.17 KiB) Viewed 296 times


Cutting two ends off a 280mm heatsink did result in a very, very, very.... Very worn down Dremmel cutting disc. It did allow me to very accurately and slowly (most important) cut through the heat sink. So that bodes well for pulling heatsinks apart.
Dremmel Disc 1.jpeg
Dremmel Disc 1.jpeg (244.72 KiB) Viewed 296 times
Dremmel Disc 2.jpeg
Dremmel Disc 2.jpeg (1.12 MiB) Viewed 296 times
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

I sent off an order to have the new lids made. I'm looking at a three week lead time though, and that's jumping the queue as this is a small job. I didn't want to do the final glueing of the battery packs until I'd performed the pressure testing successfully though so that leaves me in a pickle.

The large battery box batteries weigh 50kg on their own and 42kg for the small boxes, which I can lift easy enough. But they're just a little awkward and getting them back onto the chill plates mm perfect sitting on top of a sticky thermal pad is going to prove, um, fun?!

Anyway, I prepared the end plates for the biggest box ready for glueing. I'm using polyamide as it's super thin and will allow disassembly with somewhat ease if I even need to pull things apart... I'm not entirely happy with the top one which I might redo, but you get the idea...
Preparing End Plates.jpeg
Preparing End Plates.jpeg (955.71 KiB) Viewed 296 times
Bukes
Groupie
Posts: 41
Joined: Tue, 26 Oct 2021, 19:15
Real Name: Alex

Re: Luke's 2001 NB MX-5

Post by Bukes »

Too late now you've sent off the order, but could you just put an aluminium or steel frame around the edge of the plastic lid, to clamp it flat?
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

Bukes wrote: Wed, 17 Nov 2021, 09:07 Too late now you've sent off the order, but could you just put an aluminium or steel frame around the edge of the plastic lid, to clamp it flat?
Honestly, that's not such a crazy idea... A little band-aid but I've done worse! I'm still waiting on a quote to come back from Mitcham Laser (who are usually pretty good on price) but it certainly gives me options. Though having the holes already counter-sunk and differences in thermal expansion do worry me a little.
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

So last weekend was one of discoveries and teachings. Given that I've still got a couple of weeks to wait for the seals to my chill plates to come back I glued the first battery box together. This involved taking the three cell pack groups (four cell packs each) applying thermal glue to the inner sides, then applying the same glue to the acetal end plates and also to the end cap for the final heat sink so all battery packs are fully enclosed in aluminium.

Now that I'm writing this I realized that I didn't actually take any photos of that process, just the final result which is a bit of a bummer. Anyway, the glueing isn't anything that I haven't shown before but I also put the thermal pad in place to ensure the correct final positioning of the batteries. One thing I didn't know was that the thermal pad was itself rather like glue and my cell packs don't exactly have handles. Needless to say getting them into position was rather tricky as one side was covered in glue, the other side had to be pressed up against the end plate and the bottom on the thermal pad.

I thought I'd be able to kind of get things into position and then shuffle the cell pack into place. Not so. I placed the first pack down a little askew and not completely on the thermal pad. It was a little stuck so I tried to pick it up and it lifted the thermal pad AND complete chill plate up. That's 26-30kg just on the stickiness of the thermal pad alone. Man, was I more than upset and a little stressed with all this glue drying and very little time to get everything together and clamped.
Thermal Pad and Drying Glue.jpeg
Thermal Pad and Drying Glue.jpeg (592.35 KiB) Viewed 190 times


I ended up using a large screwdriver to pry the batteries off and then very gingerly placed them in position. I was a lot more careful with the other cell packs and finally got everything together and clamped with 6 screws/bolts each side plus my 300kg clamp. Everything went together nice and snug, just how the CAD design looks.
Glued Result.jpeg
Glued Result.jpeg (830.26 KiB) Viewed 190 times

After reading some EV construction articles a couple of weeks back I realized that I hadn't thought of using spring washers or thread locks on, well, everything. Especially on my busbars. I had everything except the spring washers for the busbars actually and came across some rather well priced torx screw assemblies made by HOBSON Engineering here in Victoria. So I ordered what I thought were 8mm and 10mm M4 screws to use on the new seals on the chill plates and also to fix my busbars. But what I found out when receiving them was that the "assembly" part is included in that measurement, so I was about 2mm short and have had to get more, longer ones 🤦🏻‍♂️. They should save me a bunch of time with further assembly though.
Screw Assemblies.jpeg
Screw Assemblies.jpeg (985.67 KiB) Viewed 190 times

Once the new screws arrive I can affix all of the busbars for the first box. In the meantime I'm going to fabricate some acetal supports for the end busbars to attach them to the angle aluminium you can see in the photo above. Then I'll get onto the tray that will hold the BMS, contactor, fuses etc etc.
User avatar
jonescg
Senior Member
Posts: 3995
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Re: Luke's 2001 NB MX-5

Post by jonescg »

I found the spring washers to be unnecessary when biting into the copper, but if you have the space and they come pre-assembled, well that's still a win.
I hate working with time-limits like glues and resins. It mocks me every single time.
AEVA National President, retiring WA branch chair.
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

I'm really erring with a side of caution in regards to any fasteners inside the battery box. A screw coming loose could wreak havoc however unlikely.

In regards to locking fasteners Chris, what do you use for the fasteners in the external walls/case of your battery boxes to keep them together? I recall you were using M3 screws, do you use a glue to keep them from coming loose?
User avatar
jonescg
Senior Member
Posts: 3995
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Re: Luke's 2001 NB MX-5

Post by jonescg »

No, just self tapping stainless steel screws into 2.5 mm diameter holes in the polycarb. Provided the cells have somewhere to swell to, they hold them pretty snug.
AEVA National President, retiring WA branch chair.
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

I managed to finish off the busbar brackets last week with some scrap acetal I had. They're not the prettiest but are functional and the best part of Acetal is it can take a thread, though I need to make sure there is some sort of mechanical lock on the screws/bolts (e.g. spring washer).
Scrap Acetal.jpeg
Scrap Acetal.jpeg (1.06 MiB) Viewed 98 times
Trail Fit 1.jpeg
Trail Fit 1.jpeg (564.65 KiB) Viewed 98 times
Trail Fit 2.jpeg
Trail Fit 2.jpeg (656.62 KiB) Viewed 98 times
Trail Fit 3.jpeg
Trail Fit 3.jpeg (454.33 KiB) Viewed 98 times


The original design as seen here had four bolts per bracket which I decided was complete overkill and would have taken all of my spare Acetal. So I cut the brackets in half and only needed to make three (well six) in total. They're now well hidden but my busbars are a little long now. Not a huge problem though.
Cut Down.jpeg
Cut Down.jpeg (1.06 MiB) Viewed 98 times
Final Fit.jpeg
Final Fit.jpeg (678.58 KiB) Viewed 98 times
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

I moved on to the fitting of BMS, contactor, fuses and other battery internals over the weekend. The basic layout is something like this (yes, I modelled the Zeva BMS and a bunch of other components to help with spacing).
Battery Box Internals.jpg
Battery Box Internals.jpg (340.54 KiB) Viewed 98 times

I swung by a aluminium supplier and picked up a ~1 meter length of 50x25x3mm C channel for $10 which is plenty strong and made for a good contactor bracket (though I had to turn it into an angle piece by removing one side).

I could probably find a suitable PWM contactor with a side mount bracket rather than the bottom but the Hyper 9 came with this and $10 is cheaper than a new contactor.
C Channel.jpeg
C Channel.jpeg (1007.35 KiB) Viewed 98 times
Contactor Bracket.jpeg
Contactor Bracket.jpeg (655.14 KiB) Viewed 98 times
Contactor Bracket Fitted.jpeg
Contactor Bracket Fitted.jpeg (907.15 KiB) Viewed 98 times


I also got to making some fuse mounts which to even get 3D printed would be around $20/30 each so scrap acetal it was again. I purchased a "slot drill" bit so I could create flat bottomed holes to secure the brackets to the tray they'll sit on. Those countersunk screws (M6) go up and sit flush under the tray. And the M8 bolts holding the fuse down go into the brackets which are threaded. Nice and neat, plus no chance of either screw touching (not that it matters in this case as they'll sit in plastic anyway.
Slot Drill.jpeg
Slot Drill.jpeg (634.19 KiB) Viewed 98 times
Fuse Blocks Mounted.jpeg
Fuse Blocks Mounted.jpeg (1.09 MiB) Viewed 98 times
User avatar
jonescg
Senior Member
Posts: 3995
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Re: Luke's 2001 NB MX-5

Post by jonescg »

How are the nuts restrained, if at all?
AEVA National President, retiring WA branch chair.
necrogt4
Groupie
Posts: 116
Joined: Sun, 28 Oct 2018, 16:49
Real Name: Luke

Re: Luke's 2001 NB MX-5

Post by necrogt4 »

jonescg wrote: Mon, 29 Nov 2021, 13:04 How are the nuts restrained, if at all?
The nuts for the M6 countersunk screws going up into the fuse brackets have flat and spring washers and to stop them turning when tightening a small flat screw driver that works well (I've also got an assortment of larger slot drills which I could have used but the screw driver worked well).
Post Reply