Matt's 1970 Opel GT - Project Log

Update, Bodywork:

Keep in mind, I've very minimal bodywork experience. The goal here is to make the yellow piece fit the orange pieces, and no longer just in a rough way.

After months of staring at things and not knowing what to do, and it being too late one night to run a grinder, I eventually just grabbed tin snips and just started cutting to motivate myself into acting. My logic was: If you cut something tonight that you're not supposed to, you can fix it faster than you figure out what's perfect (and won't ever anyway).

Started with the passenger's floorpan:



Whew. I measured from the frame edge underneath and still got it wrong. It varies from 1" overlap to zero overlap. Oh well. First bodywork in 3 months, broke the seal. Onward next night!



That's the driver's floorpan. Good enough.



I'd been trying to figure out why my nested seat rails were making the floorpan so far off. I figured it had something to do with the added sheet metal width and not being perfectly flush. Nope. When I lined up the metal seam at the bottom of the rail, the floor grooves on the very first groove are already 3/8" off.

Why?

I figure, because they were hand assembled. Each car is a little bit different. Great news, it means that basically nothing matters precision-wise, if the factory was so variable itself.



This yellow blob has a vin. This yellow blob is "a car". The rest is all "parts".

Quite a compact little car, about 4' x 4'.

I was terrified of doing this, because it took 3 of us half a day in Phoenix to shove this yellow bit into the orange body. But, at the time there were 4 more floorpan pieces and 2 more seat rails attached.

I figure most of you have never seen these panels out of the car. Let's take a tour!



Check out the view!



Quick peak up her skirts from behind.



From headlights to diff, nothing but space.



I highly recommend removing all this unnecessary paneling when servicing your GT. It makes working on the car so convenient. Rolling from parcel shelf to engine bay without even having to lay down on the creeper.



Okay, back to work. This is from inside the engine bay, looking at the box where the cabin heater assembly goes (VIN plate is just over that yellow lip).

See the top of the yellow lip where the metal is torn?

In order to originally shove the yellow car inside the orange body, the three of us used prybars and my legs. It did not want to go. This resulted, I suspect, in bowing the top of the firewall upwards into an arc, and deforming some of the metal below.

This is going to be an ongoing issue.

The yellow panel is fit. That's how much it doesn't fit by. I have to go at some of the sections where it's being blocked and start chopping metal until it fits.



I scribe some lines from underneath the car to see where I have to clean the sheet metal to prep it for welding, and then flapwheel the grime, paint, and undercoat off.

This is about the best the three of us ever had it fit in Phoenix, and that was while the far side of the trans tunnel was still floating in air. It got much worse when we shoved it downward.

Good news, both sides of the dash arch need to move in the same direction.



Driver's side by the wiring harness lines up decently already.



A bodywork mentor once told me to make the firewall fit, it would help to remove one layer of the double-thick sheet metal along the edge where it attaches to the fender. This was my first attempt, I ground through it in some places by accident.

Someone pointed out you zig-zag a double-90 degree bend to make one layer of sheet metal cup under the next, so I used a cresent wrench to bend and rebend it gradually along the edge. Seemed to work. Doesn't show on camera.



Cleaned up the passenger footwell where the panels will overlap (by about an inch).

The driver's side, in contrast is almost butted flush (I got frustrated in Phoenix and ran out of time, so I just cut once).

I figure, on such a long seam through the air, some overlap is just fine. I'm not worried about water ingress like I am a bit on the floorpans.



For the first time ever, the entire dash arch of the yellow firewall fits inside the orange body. This is farther than we ever got in Phoenix.

Great success!

However...



Great failure!

Two issues:

1 - The inside of the lip (the edge) matches, but the main supporting lip does not.

2 - Harder to show on camera, but the angle of the lip is way off. Like, 10+ degrees off (The yellow is angled lower than orange).

Perplexingly, this was not always a problem. I remember it fitting better than this.

I think what has to happen is that I have to tack some of the yellow in place, in the right place (starting here) and then forcing it into place everywhere else. I care about the fitment of the windshield, I do not care about the fitment of the seats and trans tunnel.

I may, after keeping it all this time, chop the firewall into lower horizontal (trans tunnel, floorpan, seat rail) and upper vertical (dash) sections. If you think of them as an "L" shape, I'm sure that the angle of the "L" was deformed during all the prybarring and that's what's forcing the angle to be wrong here.

...

You're now caught up on progress.

...

I'm a bit stumped about where to go next, because I'm doing multiple processes at the same time.

1 - What do I coat the cleaned metal with before welding? For example, in the overlapped section, should it stay naked? Do I use weld-through primer? (I've heard that's for aluminum or galvanized steel, but I've had it recommended before on bodywork). Do I use anything?

2 - I'm going to attempt to tack a few spots up top when I can get the dash arch perfect. Then weld in angle iron at the trans tunnel rails at the Yellow/Orange gap on one side. Then shove the bottom end in, and start welding it all up.

I don't know any of the obvious "Before you do X, make sure you Y" or "Don't do X" or "It is critical that you do X", and so on. So, any advice before I ruin things further, now's the time, because it's all about to be a lot more permanent.

brian_ said:

More importantly, there is zero chance that anyone manually formed the grooves in those panels, which were stamped by dies in a press. I assume that you are either seeing year-to-year variations, or they are not aligned at the point where you think they are.

Click to expand...

Ahh, no, not what I meant. The panels themselves are surely identical. Even year-to-year the GT only had one 4-year production run, almost everything between models is exactly the same.

So, the grooves are perfectly lined up in the floor pan. The issue is that the entire floor pan is 3/8" forward of where it should be because they weren't lined up exactly the same before welding.

I doubt that car-to-car hand assembly variations are the issue at all.

Click to expand...

I think you're wrong there, based on what some old-timers have been telling me.

The left line (where I say "lined up") is a weld seam. It's lined up perfectly at the bottom edge of the seat rail. But the floorpan immediately after that is shifted 3/8" forward. So, panels match but the exact point where they're spot welded do not.

Far as I know there's no obedience to a model year, they used parts and panels (and even VINs) from previous year and then switched mid way through.

...

Yellow and Orange are different years though, so, you may be onto something there.

Far as I've been told though, back in the day subtle variations in the panel placement is the norm. For example, doors back then were all adjustable, inward/outward, tilt, rotation, etc could all be tweaked by adjusting various bolts. By like, a half inch in any direction. And you'd need to use that, because each finished vehicle would be slightly different. Modern vehicles are so well jigged and use such larger panels that the doors are fixed, every car body is the same, no adjustment needed, you just bolt them on.

Had a 75-year old career bodywork guy show me his GT and the two of mine there and how none of the measurements were exactly the same. He said that back in the day when replacing panels or doors after a cosmetic crash (any vehicle), it would be standard procedure to get out the sheers and "trim" the donor door 1/4" if it was too long for the other body on a supposedly identical production run. With adjustment, you'd just aim for a middle of the road panel gap and that's what you get.

With how many smaller panels old cars have, that's supposedly normal. He warned me to watch out for it.

Just posting for the sake of posting, to keep things fresh in my mind...

Considering whether to have a transmission or not.

Transmission:
- Clutch. Good idea to have a safety disconnect on a DIY drivetrain, in the event of relays welding shut or, who knows what, can always stomp the clutch to disconnect power.
- Mechanical reverse. No advantage to me on an AC motor.
- Gears. Not much advantage to me, bottomless torque available.

Direct Drive:
- 100+lbs of weight savings with no transmission. Less weight, more batteries.
- No expensive or precise adapter plate to machine (I presume driveshaft is much more forgiving for alignment).
- Clears up engine bay a little bit for battery or storage.
- Moves weight further back (if that's still an advantage with all the batteries back there and engine gone).

...

Can I do direct drive?

- Rims are 13". Normal tires look to stick out 3" all the way around, so 19" total tire height. 19" * 3.14 = ~5 feet circumference.
- 5280 feet in a mile, 5 feet per revolution, means 1056 tire revolutions per mile.
- I want 80mph max speed (65 would be fine, but, 80mph is a ceiling I'd ever drive at even when passing), so 84480 revolutions per hour, or, 1408 tire RPM.
- 3.44 final drive/diff reduction ratio from prop shaft, means motor has to spin at 4843 RPM.

My 3ph, 6 pole, ACIM is rated for 961 RPM. So... this is 500% the rotation speed it's expecting.

That said, that would still be okay for a DC motor with a commutator, so, it should be plenty fine for an AC motor like this... right?

...

I think I'm going to skip the transmission and directly drive the prop shaft.

Thoughts? Anything I miss?

I suspect, as with your direct drive shaft drive motorcycle, you will be stuck with a less than ideal reduction ratio and no easy way to change it. You might make sure you have a rear axle that has a good selection of different ratios available. But, they may not have a high enough ratio and changing them could be expensive and time consuming.

My suggestion would be to keep the transmission for now. If you find that driving in the straight through, 1:1, ratio is fine for you, by all means go transmissionless. Another possibility is that one of transmission's intermediate ratios is best and could be matched overall with a change in the rear axle ratio.

I know these Opel GTs are cute cars for a conversion. But, the lack of a trunk lid really restricts access to the inside rear of the car. If you plan on putting components in that area, you'll have to make sure the seats slide well out of the way, if they don't already. Or, are easy removable. And, practice your contortionist skills.

Cutting-up and welding a vehicle this much invariably throws off the body and door/hood/window opening alignment. I've found if you keep the doors (and if practical hood and lids) in place, with their hinges and latches, they can be used as a good guide for their openings. Pull the door glass(and all other glass, if possible) so it won't be damaged by weld spatter. On some vehicles, there are factory alignment holes on the bottom surface of the body. If not, you can make your own for checking the alignment by making periodic length-wise and diagonal measurements between the holes as the cutting and welding progresses, If you can, find two normally parallel flat surfaces or mounting points near the front and rear of the vehicle. Clamp long straight edges(square tubing is a good choice) on them. By eye, line-up the straight edges to check for twist in the body. If the straight edges and mounting points are stout enough, the can be used to correct the twist.

The hydraulic cylinder based body alignment tools of course are the best for correcting alignment problems, but the common mechanical Hi-Lift jack can also do a good job.

MattsAwesomeStuff said:

Transmission:
- Clutch. Good idea to have a safety disconnect on a DIY drivetrain, in the event of relays welding shut or, who knows what, can always stomp the clutch to disconnect power.
- Mechanical reverse. No advantage to me on an AC motor.
- Gears. Not much advantage to me, bottomless torque available.

Click to expand...

Near-maximum torque is available all the way to stall, but "bottomless" suggests "unlimited", which certainly isn't true.

MattsAwesomeStuff said:

- Rims are 13". Normal tires look to stick out 3" all the way around, so 19" total tire height. 19" * 3.14 = ~5 feet circumference...

Click to expand...

19" is a very small tire... Austin Mini sized. The wheel and tire nominal diameter is measured at the bead seat, not the visible outer edge of the rim; I think your tires more likely have about 5 inch sidewalls. Some sources list 155SR13 or 165SR13 as the Opel GT stock tire size, which would be 155/82R13 or 165/82R13 in the current system; if that's it, the tire diameter is more like 23" to 24". You might want to check the size and the calculations.

Many tire spec charts list the revolutions per mile or kilometre for each size, saving some calculations and accounting for the effective rolling radius being smaller than half the static inflated diameter.

MattsAwesomeStuff said:

My 3ph, 6 pole, ACIM is rated for 961 RPM. So... this is 500% the rotation speed it's expecting.

That said, that would still be okay for a DC motor with a commutator, so, it should be plenty fine for an AC motor like this... right?

Click to expand...

The speed that the motor can stand mechanically will depend on bearings and rotor construction. 10,000+ RPM is no problem for motors of that size which are designed for it, but I have not even a guess at what is reasonable for that motor.

MattsAwesomeStuff said:

I think I'm going to skip the transmission and directly drive the prop shaft.

Thoughts? Anything I miss?

Click to expand...

If you had more information (or really any solid information) about the motor performance as a function of speed and available voltage, I would suggest looking at this configuration to see if you will have workable or desirable levels of torque and power across your operating speed range. Without that it's kind of a gamble, but if it works out this is the most straightforward setup.

electro wrks said:

I suspect, as with your direct drive shaft drive motorcycle, you will be stuck with a less than ideal reduction ratio and no easy way to change it.

Click to expand...

I'll definitely be stuck, that's true.

"Less than ideal" is true, but my concern is whether it's contextually true, I.E. will I notice or care?

This car originally had a top speed of 117mph. I'll never take it above 80mph, and likely not above 70mph.

From what I understand of most DIY conversions, they never use 1st gear, almost never use 2nd gear, leave it in 3rd gear 95% of the time, and maybe 5% of the time pop it into 4th (1:1). And those would be on heavier vehicles.

So, me being functionally stuck in 4th (or 3rd, depending on whether 4th is an overdrive for those people) all the time is not a big deal.

My suggestion would be to keep the transmission for now. If you find that driving in the straight through, 1:1, ratio is fine for you, by all means go transmissionless.

Click to expand...

That makes sense, but I think I'll do the opposite because either way I am going to have a bunch of rework to do if I change whether to have a transmission or not, and I think it's slightly more likely to not have it. Also, making it work with a transmission is a lot of extra work in itself (adapter plate, etc), so that's the big part of the effort to avoid if possible.

That is, I suspect I'll drive it straight through, and if it's irritable to me, then I'll look at adding the transmission back in. I'm not changing the frame or anything like that to fit the motor without the tranny, so it should be fairly reversible.

I know these Opel GTs are cute cars for a conversion. But, the lack of a trunk lid really restricts access to the inside rear of the car.

Click to expand...

It's my biggest pet peeve of the car, yeah. I've even considered modifying it to have a hatch, but, trying to narrow my scope.

If you plan on putting components in that area, you'll have to make sure the seats slide well out of the way, if they don't already. Or, are easy removable. And, practice your contortionist skills.

Click to expand...

I'm not too worried about that. It'll just be battery boxes and wiring. Motor will go into the front of the trans tunnel.

It's not the kind of thing that I'll be swapping out while grocery shopping. If something needs major work, the seats can come out to do it.

Cutting-up and welding a vehicle this much invariably throws off the body and door/hood/window opening alignment. I've found if you keep the doors (and if practical hood and lids) in place, with their hinges and latches, they can be used as a good guide for their openings.

Click to expand...

No pillars were or will be cut, so, that's good. No door frames are changing. Rocket panels are solid. It's just floorpan and firewall.

When I took measurements of the bottom of the windshield arch, it lined up perfectly (as I can with a tape measure) with another GT I took measurements from. So, I didn't deform it when removing the panel.

I'm not altering the shape, it appears as if the firewall and dash arch I'm putting in from the yellow vehicle is the correct size (although deformed from me shoving it into place).

So, that leaves the floorpan. And, I don't care about the floorpan. It'll be more than good, it'll be "good enough"!

I actually spent a lot of time worrying about critical alignment things but, from what I can tell from guys who've rebuilt them, and some of who've come to see it, nothing I'm doing requires any kind of critical precision.

Pull the door glass(and all other glass, if possible) so it won't be damaged by weld spatter.

Click to expand...

Too late for that, previous owner who cut away the trans tunnel ruined what might be a $700 windshield with direct grinder spray.

But considering I'm using flux-core, I was thinking of making two shields of thin sheet metal on a flexible tube with a magnet base, to basically shoulder every seam with and keep the spatter to a minimum. And/or adding a big rectifier to convert the flux-core to DC at least, cut down on that some more.

...

Brian said:

I was surprised how affordable hydraulic body jack kits are...

Click to expand...

Damn. There's a kit for $99 CAD.

https://www.princessauto.com/en/det...wer-autobody-and-frame-repair-kit/A-p8667826e

I don't think there's anything I need it for, all the edges of what I'm welding to are the strongest parts of the car, but, that's reaaaaally tempting.

Near-maximum torque is available all the way to stall, but "bottomless" suggests "unlimited", which certainly isn't true.

Click to expand...

Contextually bottomless. I think I have more torque than the axles can tolerate. There's obviously a limit, but for multiple reasons I likely won't even approach it.

The important context is "Will the lack of lower gears result in frustrating performance limits?" and I think the answer, as typical with EVs, is no. If most people who keep a transmission are still starting in 3rd gear, I should be okay.

The wheel and tire nominal diameter is measured at the bead seat, not the visible outer edge of the rim; I think your tires more likely have about 5 inch sidewalls. Some sources list 155SR13 or 165SR13 as the Opel GT stock tire size, which would be 155/82R13 or 165/82R13 in the current system; if that's it, the tire diameter is more like 23" to 24". You might want to check the size and the calculations.

Click to expand...

Ahh, my mistake. I wasn't around the car to measure. That's good to know. Circumference is linear with diameter, so, now the max revs I'd need are 3840rpm. Even better.

The speed that the motor can stand mechanically will depend on bearings and rotor construction. 10,000+ RPM is no problem for motors of that size which are designed for it, but I have not even a guess at what is reasonable for that motor.

Click to expand...

Bearings, I can't imagine wouldn't be fine at 4000rpm.

Rotor construction... likewise. I can't see that it would be built so flimsy as to actually fail anything close to the RPM that's listed on the nameplate. It doesn't actually say what that RPM is, expected, given under a certain circumstance, the actual max limit, etc. Just has an RPM on the badge.

It's a little bit frustrating that we have a good set of rules of thumb for DC forklift motors (I.E. Yes it's fine, any random one the correct size is fine, go ahead and use it), but not for AC forklift motors. However, the general construction of AC vs. DC motors makes me think it should be at least as good, if not superior. And, my limits are within what a DC motor could do.

If you had more information (or really any solid information) about the motor performance as a function of speed and available voltage, I would suggest looking at this configuration to see if you will have workable or desirable levels of torque and power across your operating speed range. Without that it's kind of a gamble, but if it works out this is the most straightforward setup.

Click to expand...

In the event that it's not... a few choices:

1 - Try to get another AC forklift motor. Not likely.

2 - Switch to DC, which for sure can handle 4000 rpm, and which I can probably find and salvage cheaply. Very likely. The Prius controller can apparently control DC just fine too, so, small software change there.

3 - Scrap the driveline for a motor and transaxle from something else. Likely, if I get driveable but disappointing performance and want something better in the future.

Pre-Patching Primer Party!

Doesn't feel like a lot of work, but I think it's my last step before merging the two cars together.

Couple misc things first:

First up, had family coincidentally fly in to Arizona and New Mexico last week, where the remaining carcass of the yellow car still is. My mentor there spent some time cutting out the good rear passenger-side shock mount on the frame. The orange donor body had punched right through (and he had taken the time to temporarily patch it for me last time I was in town so it would still roll).



A different person, who bought the 3rd car from the estate where I got my 2, also had mentioned he had 4 hubcap centers that he thought were in slightly better condition than the ones on either of my cars, and a spare yellow tail light lens (I was missing one). Mine had no chrome left, rust eaten holes right through in some places, the new set he gave me polished up to a mirror finish with just the corner of my shirt. No longer embarrasing. And, he just gave them to me.


(white one is just lens flare, it looks the same as the others)

One guy was generous enough to drive all the way across town, meet the other at work, and then drop both sets of parts off at a hotel so they could be delivered to me later. Can't say kind enough things about either of these gentlemen often enough. I absolutely would have given up on my GT before I even got it home without their help, and they're still putting in effort for me from a couple thousand miles away. Great people.

...

Onto the priming party.

I broke apart the two car sections for maybe the last time to prime the joints (feel free to declare your ToldJaSos now, as I bask in my ignorance and hubris). It might not be a lie. Everything might go perfect.



Priming begins. Colormatching is coincidence, that's just the color of the chemical. Oddly, the can specifically says not for use on ferrous steels. Aluminum or galvanized steel only. I tried to look up why but found no reason not to use Zinc Chromate on steel.

I got this far before I realized, on some advice I should do the underside, and anything I'm not going to get around to painting later either so at least it's not bare metal. So I pulled the whole comparment out of the vehicle again.



Two coats. Last (maybe) look at the step-through frame car. Farewell convenience.



Funny how well matched the color of the primer is. I painted it and I can barely tell where it stops and the factor color starts. This is the underside of the trans tunnel, dash up top, seat rails at the mid.



I'm not a welder. I have at most a few hours behind the trigger.

I have the cheapest (and used) flux core welder money can't buy.

I was thinking last night, I have helped family and friends by welding things and fixing things and building things. I have volunteered. I have taught 30 or 40 people to begin welding. But I don't think I've welded on something for me in 7 years. Feels good.

In order to jam the two car sections together, because the pieces are so big they can't just move into place. They have to rotate into place. This means the trans rails had to be cut extra far to have room for the front end of the yellow car to swing through. That means there's a gap to fill.

It's ghetto and I know it's ghetto. It's angle iron from discarded bedframes. I cut it on a half-melted wood-cutting chop saw with an abrasive blade stuck in it. I tried welding inside the box a little, just because I could.



It's ugly, but hopefully it holds up. These box sections nest entirely into the engine comparment side of the trans rails, and then can be wiggled forward through the gap until they're also nested inside the trans tunnel section. Then if all's well, they'll get welded along their seams, welded to fill the gap, holes drilled and rosette welded on either side of the existing trans rails as well.

...

That's as far as I got. I think it's time to fit things as best I can one last time, and then assemble it all with self-tapping screws and start welding. There'll be a ton of cleanup from the horrible flux-core (and novice skills). I'm especially excited about upside-down welding the floor pans together under the car.

...

A few other tidbits:

- I own a key cutter. Me and another local with a GT put it to use making some spare keys tonight (they're double-sided, took a while for us to figure out how to abuse the jig to allow that). They key worked.

- I learned how to extract the lock tumbler and force it to fit a key so that the doors and ignition match (they used to, but my 2 were frankencars before I got them and don't have any door keys).

- Need to remove crumbled sound-deadening from inside the car? Cover it in bags of dry ice, then smack it with a hammer. It shatters and you sweep it off.

- Need to replace sound-deadening? For 10% the cost, use the stuff they use on roofs. Adhesive on one side and rubber on the other. About $40 will do the whole car. But it at any home store. "Peel and Seal" I think is one brand. A professional product like Dynamat for cars is outrageously expensive in comparison.