openinverter forum

cajamatt wrote: ↑Fri May 30, 2025 2:00 am
The equation Q=w*c*dT where Q us heat transfer, w is mass flow rate, and c is specific heat of air is used.
- mass flow rate is found from 200LFM and the open cross sectional area of the heatsink.
- dT is 42C (max battery temp from the other thread) - 27C (hot summer day) = 15C
- gives 450W of heat transfer out of the whole pack. This is not a conservative estimate and I plan to adjust the variables as i find better info for them, but 500W compared to the 129W targeted tells me there is margin for this type of forced air cooling directly on the packs!

Oops! This is not the performance of the heatsink, this is the amount of energy the air would absorb if the air changed by 15C.

The heatsink performance is:

q=h*A*(Ts-Ta) where R=1/(H*A). rearranged to: Q=(Ts-Ta)/R

Ts= 42C (max battery temp)
Ta= 27C
R=0.8C/W

Q=18.75W for a single slab of the heatsink.
Q=225W for the whole pack (12 individual slabs)

The surface won't actually reach 42C when the internals do. There is a gradient, and then another gradient through the thermal paste. So if the surface is actually 35C:

Q=120W per pack.

Hmmmmmmm...

One thing to consider before committing to aircooling - liquid cooling gives you the option to provide battery heating as well. I've had issues with my car even in the mild North Carolina winters because even above freezing is cold enough to affect my pack. I've got the cooling half plumbed but need to finish the job so that I can add heating.

P.S.Mangelsdorf wrote: ↑Fri May 30, 2025 12:04 pm One thing to consider before committing to aircooling - liquid cooling gives you the option to provide battery heating as well. I've had issues with my car even in the mild North Carolina winters because even above freezing is cold enough to affect my pack. I've got the cooling half plumbed but need to finish the job so that I can add heating.

I was wondering about this! you're currently switching from air cooling to liquid cooling? what is the heating device that you're adding? have you considered adding heating elements to your pack without liquid cooling? i was wondering about embedding something like a seat heater coil to the heatsink sandwich below my packs and adding a insulating cover around the packs for the winter season.

I have searched around a bit and haven't found a solid answer, has anyone experienced a hiccup with the cooling system when the PDM is oriented on either side? I might have a hard time fitting it in upright.

cajamatt wrote: ↑Sun Jun 01, 2025 1:41 pm I was wondering about this! you're currently switching from air cooling to liquid cooling? what is the heating device that you're adding? have you considered adding heating elements to your pack without liquid cooling? i was wondering about embedding something like a seat heater coil to the heatsink sandwich below my packs and adding a insulating cover around the packs for the winter season.

The Volt batteries originally were liquid cooled/heated and had a HV heating element in the head end of the pack. I've been running them without cooling (beyond whatever airflow they might get) and haven't seen issues with heat, just with cold. I see range degradation when they are below 70 degrees F (20ish C). The plan will be to use the OEM heating element, but I'm working on where I fill fit it, and how I'll control the power to it.

I couldn't get the PDM to fit in CAD so i went and tried to fit it in the drivers side front corner in real life and it didn't fit. So i chopped off some random little brackets and it kinda fit, but not with the giant radiator in. Next to do: see if i can get back with like a half size radiator for cooling a 2013 leaf stack with the 80kw inverter. Then going to consider rotating the motor and inverter about the transmission center to make more room for the PDM. That's what's shown below, and the most promising orientation for the PDM is fully upside down. After viewing Johannes' teardown and noticing the cooling channels are all planar i'm not concerned about air bubbles or anything regarding orientation of the PDM. It also looks like both the bottom and top covers of the PDM are oversized and maybe i'll consider making smaller custom covers for the PDM to do a honey-i-shrunk-the-PDM!

P.S.Mangelsdorf wrote: ↑Sun Jun 01, 2025 11:46 pm The Volt batteries originally were liquid cooled/heated and had a HV heating element in the head end of the pack. I've been running them without cooling (beyond whatever airflow they might get) and haven't seen issues with heat, just with cold. I see range degradation when they are below 70 degrees F (20ish C). The plan will be to use the OEM heating element, but I'm working on where I fill fit it, and how I'll control the power to it.

I'll be sure to follow along with your findings on this!

I also had a hard time getting the PDM to fit in my build. But final solution looks like it was always meant to be there now.
I rotated 90 horizontally and ran the coolant flow through backwards and has all worked fine for me so far - 300 miles of testing so far.

I did consider a custom lid, but concluded it didn’t save very much. If you hunt through my build page you’ll see a link to Flying Tools YouTube where he briefly shows a custom lid in case it helps your thoughts.

The most important space saving I found for my particular constraints was how the cables exited the PDM. Once I’d sorted that, it worked out ok for me.

Sell the pdm and get an outlander obc, or Tesla pcs

since the inverter is not centered over the motor, it turns out tipping it to the right meant i could rotate it all the way to 90deg if i wanted, and create way more room for the PDM on the left side. Last night i did a test fit with the prototype battery box in place and it fits!
This means my adaptor plates need a rotated hole pattern. It looks like i could add a rotated set of holes for the NEMA B-face pattern in order to achieve this with you adaptor plates, @bratitude?

I haven't found any other conversions with the leaf motor tipped over like this. and tipping it this way puts the coolant ports at the bottom, which creates the possibility of bubbles trapped in the cooling channels in the inverter. i'm willing to take on this risk and monitor inverter and motor temps, and if there's a problem i'll replace the PDM with a smaller DCDC and OBC so i can rotate the motor back to vertical.

now that i've confirmed i can fit 1 battery 4-pack under the hood i can carry on with designing the remaining 3x 4-packs for under the chassis!

I have been spending less time on conversion and more time on restoration lately, with some big rust patches as well as evidence of an old collision discovered

Great project @cajamatt,
I am just starting a similar project (Leaf Stack, Zombi, Ioniq5 battery): viewtopic.php?t=6236
So much good information in your post and in others comments to you. Happily following your progress.
Would like to touch base with you.
Email me if you can spare some talking time.
jerryk48 at gmail dot com

cajamatt wrote: ↑Thu Jun 12, 2025 5:00 pm
This means my adaptor plates need a rotated hole pattern. It looks like i could add a rotated set of holes for the NEMA B-face pattern in order to achieve this with you adaptor plates, @bratitude?

Yup! Beautie of the nema pattern is you can clock it any angle you like

cajamatt wrote: ↑Thu Jun 12, 2025 5:00 pm
I haven't found any other conversions with the leaf motor tipped over like this.

This vw beetle build has the full stack mounted on its side. But could be worth asking them how it worked out longer term. Not sure if the build is finished yet.

The link might not take you to the right post - scroll to July 2023.

https://www.diyelectriccar.com/threads/ ... st-1113317

i've found conflicting info on whether or not the EA71 bellhousing has the same mounting pattern as the EA81, mostly searching through old engine/transmission swap threads on ultimate subaru message board. here is my hand measurements of the bellhousing i have, and i'm hoping i can get the measurements from the EA81 adaptor plate to see if the holes line up. This was made pretty hastily just to capture the rough hole pattern. i haven't found any markings on the transmission that would indicate what kind i have.

print at scale

unfortunately its not the same pattern

"just a little minor body work" -me to myself while buying the car...

i set off to repair the couple small rust holes, and the first one i tackled turned into a patch of most of a wheel well and part of the frame. the car had been crashed and filled in with body filler and repainted so you couldn't tell. I'm more or less back on track now, planning to clean and paint the wheel wells properly next year, but the engine bay, transmission mounts, and suspension components all need to be cleaned and painted for real before reinstalling everything.

my front battery box will be made once the waterjet at work gets repaired. i had to redesign it to house the hyundai CMU and make easier battery connections. also got a plastic box from polycase that will be an HVJB appendage to the battery enclosure since i can't make the battery enclosure big enough to house the HVJB components.

everything in the engine bay and transmission tunnel that's gonna get hidden by the new components has been ground, cleaned, sanded, cleaned, filled, sanded, cleaned, rust convertered, primed, filled, sanded, cleaned, primed, over and over again! order of operations was a little off haha. new fasteners are showing up so i have started mounting things back on and MAN does it feel good to put things together instead of take stuff off. i have steering knuckles and brakes from a newer generation subaru en route, 1st of 4 battery boxes waiting to be made at work, planning to assemble the motor coupler this week. i rearranged the battery layout for the rear packs. instead of 2 4-packs under the rear seats which would have involved cutting a huge section of the floor out and remaking it, i realized i actually can fit more batteries under the trunk that i first thought. so new plan is to make a 2x2 saddle pack under the seats and 2 separate 4-packs under the trunk. this part of the project is frustrating, i'm putting tons of planning and buying in with very little tangible progress. but it feels like it will come together fairly quickly once i get past whatever tipping point!

photos below are some parts i painted, a pair of aluminum rails after i put riv nuts in that will be part of the battery pack, the 2x2 saddle pack for under the rear seats, 2 4-packs placed under the trunk, and an alternate layout placing 9 modules on one layer under the trunk. i was excited about this layout, but decided against it because i didn't like how it puts more weight on a cantilever behind the rear axle and i'd want to reinforce the frame more while removing material to make room for it. 2 separate packs bring the weight closer to over the rear axle and i don't have to remove a very convenient beam.

in setting off to figure out the cooling power i need i accidentally made a cool table that shows what gear i should put my transmission in. based on the leaf motor efficiency chart i plotted out the efficiency of my car in it's different gears at different speeds to see which gear i should stay in. 2nd gear is most efficient, but can't go 80mph. that's ok though i never intentionally go 80mph and that might be scary in this car!

Motor speed, power, efficiency and losses are shown at max power (80kw) in one column and at 50kw in another column. 50 kw is a random guess at the most continuous power i would need. the stock radiator in the subi is a similar size as that of the leaf, and i would like to get a slightly smaller one so i can push it closer to the grille and make more working room in the engine bay. 10kw cooling power on a hot summer day will be my lower limit. time to learn about radiators!

The biggest challenge will be the low delta temps you have with inverter and ambient. Inverters do not like getting hot and will start derating at like 50-60C.

Ahh right, thanks for the reminder about the smaller delta.

is anyone familiar with the thermal putty used in the hyundai/kia battery pack? is it actual play-dough?? i harvested a couple liters of it, and tried to use it between some aluminum layers of my battery enclosure and it is a very odd consistency that does not spread well. i only have experience with the kind of thermal paste you put between IGBTs and heatsinks and expected this to behave like that just higher viscosity. i'm worried about not making a very consistent thermal bridge between the bottoms of my ioniq modules and the heatsink as well as potentially damaging the very soft bottoms of the pouches if there's a big clump of the putty that doesn't smoosh well.

some photos of my meager progress lately.

- the hyundai PRA (power relay assembly) in it's new enclosure which will be an appendage to one of the battery boxes.
- battery box parts!

fun stuff, did you end up measuring the bellhousing pattern yet? i have a customer that wants to send me there bellhousing for me to probe and make a plate for.

Bratitude wrote: ↑Fri Sep 05, 2025 4:54 pm fun stuff, did you end up measuring the bellhousing pattern yet? i have a customer that wants to send me there bellhousing for me to probe and make a plate for.

i did measure, just with calipers, but close enough to where i think it'll work. i haven't made it yet because i want to double check the angle to tilt the motor. once i make sure it works i'm happy to share the pattern!

Spent the last week on the first battery box. This is 1 of 4 boxes i need to make, and this one goes under the hood. Once i finish this box i'll then flush out the designs for the remaining 3 with lessons learned in hand.

The enclosure fit into the car just right and got the mounts match drilled. Then i had to do a handful of small but time consuming reworks (slotify some holes, move a riv nut, paint, add gaskets, etc.) The big picture design looks good still and i think everything is going to fit!

Then i built the battery heater. It's a super flexible silicone jacket 12V heating wire from amazon, used for things like making custom heated gloves or footbeds, and i used thermally conductive epoxy to glue it in between the fins of the heatsinks. The heat will conduct through the base of the heatsink, thermal paste, the battery box enclosure, thermal paste, a thick aluminum filler plate, and a thick layer of thermal play-doh from the ioniq battery. So i think the heat should disperse pretty evenly by the time it reaches the bottom of the cells. Vents will be plugged and i may "wrap a blanket" around the battery box in the winter time.

Remaining on the first battery pack:
- ingress test and plug leaks with RTV
- roll out the thermal play-doh that goes beneath the modules. It's a very annoying consistency but i'm going to try wax paper and a rolling pin and two rails of a set height to try to get an even surface.
- build bus insulation
- print the gland plate

I also visited my friend at the University of VT machine shop where we did a shrink fit for the motor coupler on some practice steel stock and it was suuuuper satisfying! heading back again tonight to hopefully do the real thing! Bratitude, did you learn what steel old subaru splines is made of and did you take any precautions against accidentally heat treating the splines while welding?

got the motor coupler made! (haven't welded yet)

a couple notes:

referencing the machinery's handbook we targeted the tightest fit, FN5. We didn't quantify how tight the grip would be, but reasoned that if we can't press it together we'll reduce the interference until it fits. we machined a practice run since my machinist friend hadn't done a shrink fit before. practice run had .0023" interference. we heated the outside up to 400 degrees F and they pressed together like butter and it was ultra satisfying.

for the real thing we turned down the OD of the inner piece (subaru splines) to get a larger contact area, since it was originally tapered. measured that OD and then machined the bore to .0028" smaller than that OD. This is technically out of spec with the FN5 fit, we were .0005 too small. We heated it up to (allegedly) 500 degrees F (found out afterwards that out IR thermometer was wonky) and the inner piece dropped right in! lesson learned: 200 degrees F probably would have been fine.

I am planning on a 1/8" fillet weld. we put the chamfer for that in the larger cylinder so more material is left on the tiny subaru part.

I test fit the motor again to double check the rotation angle. I had planned to rotate it 90 degrees towards the drivers side, but i really didn't like the cooling ports being totally hidden underneath. so i flopped it back to 45 degrees towards the passenger side and convinced myself the PDM and contactor assy will fit in this configuration and it gives me better access to inverter cooling ports. the tricky part is it BARELY fits between the battery box and the engine bay wall. my tilt measurements were trustworthy so i actually slotted the leaf mounting holes and left one hole out from cutting. that way i can place the motor and inverter at the angle that fits and match drill the last hole to hold it at the right angle.

bought the 1/2" aluminum plate and got the parts cut and they fit great! then i hit some roadblocks:

Big delays on getting my motor mounts and battery pack parts cut and bent
which is pushing me into working in my unheated garage over the winter or shelfing the project until the spring
and my restoration battle continues with the steering knuckle/brake assembly. my trust in the old subaru forum is at rock bottom

I powered up my brake booster yesterday and it doesn't have fluid but it made some sounds that make me feel like it's working, which would be really cool

got my motor coupler welded by a friend of a coworker and it looks like he did a great job. bringing that back to the machine shop to trim off the lumps so it's closer to balanced while spinning at 1 million RPM.
motor tilted 90 deg towards the drivers side.
motor tilted 45 degrees towards the passengers side
leaf/ea71 adaptor plates! thanks for the leaf pattern, bratitude. note the slots for adjusting the mount angle and corner in the top of the plate where the last hole will be match drilled.
motor adaptor plates adapting motors
motor mounts mounting motors! it shoonked right in to the transmission pretty easily
motor coupler welded!