Opel GT Forum banner

161 - 180 of 186 Posts

Manta B
16 Posts
Matt, once all these social distancing and isolation rules come to an end I would really like to visit with you and see your efforts first hand. I'm really intrigued by your project and admire your persistence and dedication to the project. Keep up the good work and keep posting the progress updates.


421 Posts
Discussion Starter #162
Buy all new brake lines, the three flex hoses and probably a master cylinder.
Well that's direct and simple. Thanks.

1 - Flex hoses - On sale, $59, regular $72: Opel GT Source

The flex hoses are for certain, desert car. Everything rubber on the car crumbles like Ramen noodles when you touch it. I don't think I'll even try to get it inspected without new hoses.

2 - Steel lines - $63: Opel GT Source

I'm tempted to just buy locally rather than ship across the border, and fab my own if it's decently cheaper. I'll have to be shaping them anyway and I probably have all the fittings. As a bonus, I don't have to wait for things to cross the border. Provided nothing else actually leaks, I can at least test the system first.

3 -

Master Cylinder & Brake Boost - On sale (25% off), $199: Opel GT Source
Master Cylinder only - On sale (40% off), $99: Opel GT Source
Master Cylinder Rebuild Kit $65 - Opel GT Source


So, cheapest, $187 US + shipping.
But at the current sale price, might as well pay the extra $35 and buy the new Master rather than a rebuild.

And that's if my booster is functional, which I can't really test until the rest of it is hooked up. That's as much as I paid for one of the cars!

Is there a compatable/close enough junkyard option for the master cylinder? They're $20 at Pick N Pull. Or is whatever a Master Cylinder does specially unique to the GT?

One hesitation I have on spending money there, is that in an electric car, you almost never even touch your brake pedal. Like, once a month of daily driving. You use regenerative braking through the driveshaft to the motor. So, clapped out Master vs. broken one starts to matter, since clapped out might be just fine for the rest of the life of the vehicle.

I'm in a place where I think I need to budget everything that needs to be purchased for the rest of the build, so I can be consistent in my cheapness. For example, if I'm going to spend $5000 on a paintjob, I might as well be spending another $3000 on getting lots of other stuff new so I have a consistent level of crappiness. If I'm doing a $100 rattlecan special, then scabbing it all together is fine. I bought the car to have a cheap project, not a good one, or I would've just spent more on better bones.

Brakes seem like a good place to not cheap out. But I don't want to sucker myself into "Oh, just spend another $200" another 4 or 5 times on a ~$1000 car right now, which I tend to do on projects.

Maybe that's what I should be doing right now regardless, a budget.

421 Posts
Discussion Starter #163
Boring "progress" without progress and without pics.

- Filled out a shopping cart of essentials on OGTS. ~$450.
- Wrote out a wishlist of salvageable stuff from the local co-op. Probably a couple hundred more there.
- Made a list of upgrades (heated seats, almost necessary in a Canadian EV, worth 10x their wattage in air heat), and junkyard stuff to find (auto windows, locks, stereo). Another few hundred.
- Tires.
- Sound deadening
- At least some minimal paint.
- Bulbs
- Some kind of heating system and all my actual DIY stuff.

Probably easily another $2000 over what I've already spent (~$2500 if you include $1000 in gas to go get it). So, $4000-$5000 ish car. Which is a lot more than I wanted to spend, (not sure I would have bought a $3500 car if it was available locally), but an amount I'm still okay with. So, that's my general "That's how crappy this car will be" range, to know what's worth spending on and what's not.

- No progress, but the community of people building this exact design has grown significantly, and more volunteers are figuring out more about it.
- The DC-DC converter that functions as the alternator to keep the 12v system stuff in line, is already onboard in the inverter, and using it is as simple as an extra 12v switch, and a 5v signal wire to some specific pins. Boom, 12v 100 amp converter for the 12v, powered from the main pack.

- Called everyone who recommended me to everyone until I had a closed loop. "Have you tried X?", "Yes, they told me to call you." No one does 1 5/8" shafts.
- Called a forklift place, they said what I'd need would be the transmission shaft, I asked ballpark how much before they look into it further. "Oh, thousands?" Heh. Thanks-no-thanks. Beats my $20-30 budget before I'd just grind the shaft smooth.
- Called tractor supply places. Some Ford and Massey-Ferguson tractors use 1 5/8" 25 spline PTOs, and you can buy a clutch disk for those, the center of which should match the forklift. Cheapest I've seen is ~$75 US + shipping. And I'd be buying a brand new clutch to destroy. There are tractor junkyards, none near me. I wrote to many. No replies in a week.
- JB Weld + Steel flakes + metal mesh may be my temporary solution. (Giant JB tubes were on sale). JB weld is 10% as strong as steel, but the motor is 400% the cross section of the driveshaft. Add in some steel to make it stronger, it might be in the right ballpark. Good enough for testing anyway.
- I've decided to try to machine my own coupler...

Last time I made a coupler I used a hacksaw to slowly cut the teeth. I hated doing it. I'll save that for later. I have worse ideas to try out first.

First I'll be making a mold of the shaft in plaster (if it holds up) or casting sand (sifted playground sand + crushed kitty litter + water). Then I'll melt scrap copper and pour it into that mold to create a copper driveshaft copy. Then I'll slowly lower that copper slug into a steel blank underwater connected to a high frequency power supply, and electro-discharge-machine it out of the steel.

It's a lot of work to save $100, but, I get to try out 3 new things: mold making, furnace making, and EDM. I've been meaning to do that for at least a decade, good excuse.

First i had to filter 50lbs of playground sand with a sieve. I did about 15 lbs before I remembered I had a scroll saw in the garage, set that up next to the sieve and powersieved by holding the handle against the scroll saw while it was on. Sounded like a woodpecker and wasn't the right motion (up/down just bounces it, side-to-side would have been better), so it wasn't any faster but it was a lot easier.

The first plaster mold I made, I somewhat-expectedly broke taking it off the shaft (silicone lube stopped it from bonding), it was very thin, just wanted to see if I can pick up detail. It picked up the fine detail in the recycling logos in the bottom of the plastic cup I held it in. Detail should be fine.

The second plaster mold I made I only waited 3 hours (vs 12), in a much larger cup, and when I pulled it free it had 1/4" of water still floating. I'm not sure how, it wasn't that wet. The mold is awful, useless with broken teeth, but good enough to practice casting in after placing it in the oven to drive off moisture.

I welded a piece of 1" black pipe to a steel base, and put that onto a 1000 watt electric stove element. Apparently you can cast copper in an iron pan on the stove. The element got red hot, the crucible did not, and nothing melted. I cut a piece of copper and directly laid it across the red-hot element. Never melted. Cut a strip of aluminum and did the same. Never melted. Not sure why, both are supposed to melt before red hot.

Then I dug out my carbon arc torch I've never used (what they used before oxy-ace torches, it holds two carbon rods in a "V" and arcs between them), and since the rods were only 2" stubs, I just grounded the crucible and dipped the electrode from a clamp into the crucible. It was incredibly difficult to maintain an arc without blowing a breaker. Just borderline possible with 15A. I got the whole crucible red hot, almost orange over a half hour of screwing up. It sort of, kind of, melted copper a bit. Enough to fuse but not enough to pour.

Made a third plaster mold, and it's been sitting with a radiant heater pointed at it for 2 days now, as I can't melt metal anyway. If I can get it off the motor shaft without breaking it, I'm not hopeful it'll survive copper temperatures (plaster holds up to about 1200'F, copper melts around 1100'F), so this'll be the last one I make with plaster. It doesn't have to be pretty or finished, as long as it's dimentionally accurate as it's lowered (some point of the cross section must be accurate). Green sand is quick, no pausing to dry, so I'll try that next.

Next my plan is to buy some new copper gouging rods (50% off) and use them in the actual arc torch, powered by my stick welder rather than clamped to my flux-core welder. Carbon arc is hot enough to melt steel if I pause for too long, so, being able to heat the crucible from the bottom should work better. If I had firebrick rather than a naked steel crucible this would be 10 seconds work.

If melting fails, I'll try aluminum instead of copper. All that matters is that the EDM electrode is conductive, but copper will get a better finish.


If this works, then I have to rig up some way to smoothly lower the copper driveshaft downward. Building the EDM isn't electrically intimidating, it's 2 lightbulbs and a capacitor. It'll have through-hole access, so no need for magical automated motor reversing to clear swarf and all that. Surface finish is a result of voltage used (lower is better), frequency (higher is better), and on-time (smaller capacitor, smaller spark erosion), but the better the surface finish the slower the speed. I'm expecting it to take an hour or so of plunge, and I'll be doing 1/4" or 1/2" pieces at a time (the thickest stock I have, I'll have to make a stack of pancakes and weld them when I'm done), so, it's not critical.

Jig-wise, maybe a drawer slider and a piece of threaded rod. Maybe I'll use a spare drill press.

... as I said, not really progress. There's nothing to look at.

1,204 Posts
If I didn't know better, I would say you are a scientist by trade :) Sometimes progress is hard to see, but eventually surfaces when you are persistent in your work :) Keep at it, I think you are doing great.

421 Posts
Discussion Starter #165
New milestone: I have now spent as much avoiding buying a coupler, to avoid buying a coupler, without having spent enough to actually make a coupler yet.

Well, time for an update on a series of failures that may someday lead to a success. Half of this is just so I have this written somewhere, it might not be that interesting.

There are 3 sub-projects to this, in increasing complexity:

1 - Make a mold of the driveshaft splines
2 - Make a furnace to melt the copper into a copper clone of the driveshaft.
3 - Make an EDM machine to use the copper clone to cut a female coupler out of steel.


Almost complete failure. I have now attempted to make 10 or so molds.

Plaster of Paris works okay, but it's fragile and you're supposed to have 2 degrees of draft angle. I have zero degrees of draft and a bajillion grippy sections (cylindrical shaft with splines). This almost always results in the mold gripping the shaft too well and shattering the plaster when I try to take it off. I've let it sit 2 hours, overnight, and 4 days. None held up.

I've tried mixing Plaster of Paris with sand and bentonite clay (kitty litter ground up using an immersion blender), for extra rigidity and temperature handling at the expense of surface finish. It also shattered taking it off.

I tried using green casting sand (sand with bentonite clay). The "right way" to cast copper. This is impossible, because you have to pack the sand around the object, you can't pack it first. The motor shaft is in a recessed hole, there's no way to pack sand around it. I tried anyways, not even partial success, there is no way. Loose sand won't hold, pre-packed sand won't crush. Neither sand will release even when I have an otherwise unuseable mold, even if I liberally cornstarch the splines and the sand surface. This is a dead end without having access to pack sand.


As earlier, a stove burner won't work. This is the garage equivalent of those Pinterest-mom drain cleaning or home remedy ideas, everyone keeps repeating it, no one knows what they're talking about.

I dug out both of my carbon arc torches that I've never used and bought some 3/16" and 5/16" carbon gouging rods (on clearance, 60% off). I'm not gouging, but no one has used carbon arc torches since oxy/ace has been around. Copper-coated carbon should just be copper-coated carbon, gouging is just when you direct air parallel to it to blow the molton metal away. I don't have a gouging head anyway, so should be fine.

Short vid:

My idea was to keep the steel crucible and waft the arc up and down around it like you would with a torch. I could not get my rods to light. They'd spark, but never light. I tried for an hour to feather it every possible way. I was moving the tips by a millimetre at a time, no luck. Even with an arc welder designed specifically for dual use as a carbon arc torch (you flip the ground clamp around, slide the stinger onto one side, and the clamp itself becomes the spring tension between the "V" to light the arc). I could get an arc, but only momentary. Not sure what I'm doing wrong.

Another short vid showing the torch, which I had to un-repair the previous owner's repair (drill out one side to fit carbons):

I cut up and machined a cavity and access holes into some fire brick. Normally fire brick is how you insulate your furnace walls, inside of which you put a crucible. I just used the fire brick itself as the crucible, which will consume it but, I'm not doing this 20 times. I replaced the carbon arc torch with a pair of jumper cables and lit the arc inside. It was so hot it turned the ceiling of the furnace into glass in about 3 seconds, which then dripped into the copper below. It's plenty hot, just not focused. I tried 5 different things and none worked, and it blew the breaker every 3 seconds so I gave up and took the lid off.

Still kept blowing breakers. Without a lid, no way to add energy fast enough to keep the copper molten without blowing the breaker.

So I stepped my game up. I took 3 beyond-junk car batteries and hooked them up in series. Then put a 15A car charger on Battery #1, and hooked my flux core welder (27.6v output) to Battery #2 and Battery #3 in series (13.8v per battery, about right). Was about 45v open circuit, 36v loaded. The idea being, the batteries would help pick up some of the load. I could leave them to charge for a half hour, and then use that stored energy combined with the welder and charger, and buy enough time to liquefy the copper while having higher average power output than the 1800w I could get from the breaker.

Still couldn't get the arcs to light for more than a couple seconds.

I changed methods. I stuck the positive carbon directly against the copper and then hovered the negative over top. Instant success. And I mean instant. Anywhere that arc moved, the 10g scrap copper melted like it was made of butter. Took seconds to melt the copper wires into a pool. I was having so much fun messing around with it heating and cooling that I eventually started blowing breakers again because the batteries were too low.

For scale: An entire 100 amp service to a home is 240v * 100a = 24,000 watts. A 50A outlet for your welder can be at most half that. 12,000 watts. A 15a 120v breaker is only 1800 watts.

Starter batteries are 12v, ~700 CCA. So, more like 1000A at room temp. 12 * 1000 = 12,000 watts per battery. 3 of them makes 36,000 watts. 50% more than the entire power grid to a house. 20x the max power of my welder. Even if they were underperforming junk batteries... ludicrious amounts of power.

I eventually just ignored the breaker being blown and ran the batteries into the ground to keep the arc lit, until the copper was almost boiling.

I poured it into the only mold that survived, the awful one I described in my previous post that was super wet and didn't transfer detail well. I'd only baked it dry for a few hours.

.... surprisingly, it worked. I'll make some videos of it when they have better context.

I'll have to run it through a bandsaw to cut off the top half, and, it doesn't matter how ugly the inside looks, it'll be getting mounted to an arbor regardless. Just the outside profile matters. Voids and such in the finish don't matter either, as I'll be plunging the whole thing through the workpiece.

I've been told you cannot cast copper in plaster of paris, it breaks down. Mine held up fine. The only errors were errors in how crappy of a mold it originally was.

My copper might have glass dissolved in it. Hard to tell. I didn't use enough copper either.

I'm 100% satisfied with my ability to melt copper in the "furnace" (an open bowl shape cut into firebrick). If my furnace survived (debatable, it started to crack) I just need a better mold to pour it into.

The current copper clone might be usable with some cleanup to make sure there's no extra material (extra material will cause extra material to be removed, and thus be missing teeth). It's ugly, but the electrode wears perhaps 10% as fast as the workpiece, so, it might be enough.

EDM Machine:

This is pretty simple. When you bring two electrical contacts together, they spark. A tiny dust fleck of material is vaporized from the positive side. But, only when the electricity first "lands" there. So you want to keep starting and stopping the spark as fast as possible. Around 5,000-20,000 times per second. The arc will happen whenever the two surfaces happen to be closest, which jumps around because you keep flicking a spec of dust off of what used to be the closest point.

You can do this with transistors and brains, telling it to start and stop. Or you can just control the rate that a capacitor charges, which eventually has enough voltage to ionize the fluid, jump the gap, get discharged, not have enough to sustain an arc, and charge up again. I went with the capacitor.

Old Popular Mechanics articles have shown this back in the 60s, when GTs were being made. It's just DC voltage to a light bulb (resistor) and a capacitor.

2 kinds of EDM: Wire EDM, and Sinker EDM.

Wire EDM is like a spark-erotion bandsaw that cuts through material and carves out a shape. That takes computer control. The electrode is very fine copper wire and, what survives out the other end isn't reusable. Wire EDM typically uses a trickle of distilled water as the bath.

Sinker EDM is just an electrode you slowly lower into the workpiece, and it copies the shape. The electrode is typically graphite. Sinker EDM typically uses kerosene as the bath. I'm building a sinker.

I tried using tap water on a test piece, lowing copper pipe into 1/8" steel. It sucked.

I've seen plans for EDM machines for over a decade, they usually have computerized stepper motors to control the sinking electrode, having it regularly back up to wash the swarf away and prevent it from sticking. I just used my drill press to lower it.

About every 3rd spark made the electrode stick to the workpiece. There's an instant visual indicator of this, because your light bulb stays lit rather than flickering. Then you have to back it off, then lower it again.

Also, because tap water is conductive, it electrolyzed the water, splitting it into H2 and O2. O2 goes to the positive side, forcibly rusting my steel workpiece. I had a frothy foam of rust that interfered with the sparking and soon rusted up the whole tank.

So I bought some Kerosene. 20,000 electrical arcs per second, in a tub of kerosene? Apparently, unlike the instant bomb this would be with gasoline, with kerosene this is safe. You can throw a lit match into a pool of kerosene (or diesel) and it won't light.

It worked... poorly, but it worked. I ran it for 1 hour before stopping to inspect.

My copper was cut crooked, so that's why it's a semi-circle, that half has always been closest to the workpiece.

The whole tank was black with iron oxide, I had no filter. I had a small pump, but my "small" marine pump pumps a gallon every 3 seconds. I don't want a kerosene fire hose, so I didn't use it.

I started off with perhaps 5 hz, and ended up feathering the gap better and achieving maybe 50-100hz. Not the optimal 10,000-ish, so it was very slow. Mostly due to setup I think. My capacitor timing is right, I just have no flushing of the worksurface so it regularly sticks and fails.

How much material did I remove?

Quick maths:

3/4" semicircle = 1.1775" length.
0.050" wide track.
0.050" deep, but in a V shape, so, average 0.025" deep.

Total material removed: 0.00147 cubic inches in an hour (24 cubic mm/hour).

In a proper setup, I should be doing 14-29mm^3/min that rate per minute, not per hour, but it's roughly in the same ballpark that my frequency is off by, and the first 30-40 minutes were much worse.

You can just add more power, 10x much even, to a point, as long as there's enough surface area that sparks are spread out.

How much do I need to remove?

After drilling out the coupler to the minor diameter of the splines (mechanical removal is thousands of times faster), I would sink the copper into the workpiece so only the splines themselves have to be spark eroded. On my splines, the major diameter is ~41mm (1320mm^2) and minor diameter is 37mm (1075mm^2). Sanity check, that's 4mm difference total, which is 2mm deep teeth on each side, seems about right. And call it a 50% that needs to be removed between those two diameters to create splines. (1320mm^2 - 1075mm^2) / 2 = 122.5 mm^2 of area needing to be cut. Or, 122.5 cubic mm per mm of depth.

If I want at least an inch (25mm) of depth, that's 3000+ cubic mm to be cut (1/5th of a cubic inch). I probably have 2 inches of splines if I want to cover them completely, but, let's no go crazy.

At my current average (really, worst-case) rate of 24mm^3/hour, that's 125 hours, or 5-ish days of cutting. At a proper rate, that would be about 2 hours of cutting. At a fast rate with aggressive settings and poor finish, 20 minutes.

I'm fine standing at a drill press feathering a cut for 2 hours, but I'm not standing there for 5 days. So, the viability of this depends on how good I can get my setup compared to a small commercial tap-burning rig.


So, plan moving forward:

Mold: Be less ambitious in depth, try to get 1/4" deep mold instead of 1/2" or 3/4", and hope it survives. I can make multiples of these and just swap them out if they get consumed by the EDM. Stick with plaster of paris. If that doesn't work,switch to green sand, which means finding a way to pack around the splines. I think I can add 2 more steps and create a low-temp mold first. Make an epoxy copy of the shaft (or a heated PVC copy), then melt plastic or pour epoxy into that female mold (with some kind of oil or release so it doesn't stick to the sides), then use the plastic male clone to create a proper packed green sand casting that can handle copper temps.

Furnace: Add more copper next time, see if firebrick survived, consider switching back to the steel crucible buried in sand, power isn't really a limit anymore.

EDM: I really want to avoid building a proper jig, with sliders and stepper motors and current sensing with computer controlled reversing. It might be inescapable to have to do that. That's a whole project in itself, ordering parts, troubleshooting, programming, etc. I'll try adding a pump, increasing capacitor size, and seeing if I can burn fast enough without doing anything else. I'll need to filter the kerosene too, coffee filter maybe, not sure. I have to remove 125x as much material, and the kerosene was already ink black.

... I'm almost content having experimented with the 3 side projects and just want to go buy a proper coupler now. The appeal was "Maybe it'll work easily, and I don't mind wasting a little time to find out". But I'm there now. Mixed results, and patience for side projects is wearing thin. A hacksaw would have had my splines cut by now.

Here's the "expense report":

Kitty Litter: $7.98
50lbs Play Sand: $8.20
Strainer: $3.50
Plaster of Paris: $6.47
Cornstarch Baby Powder: $3.47
Containers: (free, had some)

Mold total: $29.62

Black pipe & plate: (free, had some)
Black pipe cap (didn't use): $2.12
Firebrick: $7.50
Flower pot (didn't use): $2.50
Stainless tongs: $1.25
3/16" Gouging Rods: $8.33
5/15" Gouging Rods: $9.13
Borax (sprinkle into molton metal to clean out oxides): $5.47
Car batteries, chargers, jumper cables: (free, had some)

Furnace total: $36.30

2 Gal Distill water (didn't use): $4.00
Tubs: $5.25
Lightbulbs: $4.25
1/4" tubing: $4.99
3/4" tubing: $5.45
Kerosene: $14.98
Pumps: (free, had some)

EDM Total: $38.92

Coupler experiments: $104.84

1,204 Posts
Wow that is incredible or should I say you are incredible. I think you have taken this Opel rebuilding to a much higher level :) :) nicely done and very well documented.

421 Posts
Discussion Starter #168
Well, it's now been a year since I first arrived at SLRacer's place and saw my GTs and begun working on them (cash already changed hands but I didn't legally own them for a couple more days after that).

A year in review... (no car stuff, safe to skip)

A bit bittersweet. I started what I thought would be a fun summer project. I had a lot of naysayers that said "Ha, you won't even get it driving in a year!!" which, as of today, have their moment in the sun. They were right.

Partially it's that things were more work than I thought, they always are, that's okay. Partly it's been that I chose to do other things for a few months, I'm okay with that too, that's why I did other things. But mostly it's been that I don't have a garage of my own, and I work long days 7 days a week. I'm still in the mindset of "I'll do that in the evening/I'll work on that on the weekend". Just, mentally that's how life normally has gone right from kindergarden through college then career. My brain subconsciously still books that "work on it on your next day off" time even though it's not there. It still gets me into trouble overcommitting because I still instinctively feel like I'm no more than 5 days away from a couple days off. That there's evenings and weekends to look forward to doing stuff. But for me now, by evening it's too late to make noise, and I've only taken 1 day off since I left Arizona with my cars last year.

So that's been a disappointment, not just because it didn't get done and I haven't been able to enjoy it, but because some of the choices I've made in my build were in line with "cheap, easy, summer project", where the goal was to have it done soon, not better. Shitty project done quick, that's the tradeoff I was okay with. But if I was stepping back and planning a 1-2 year build, I'd have upgraded some of those efforts and made different choices. I'm now going to end up with the same shitty end result, but that took me a medium amount of time to complete. Which is also my biggest source of frustration on the project, juggling all the options between slapped-together and done-right and never knowing how much more time to spend making something better. I get builder's block for days at a time.

On the flip side, part of the reason I was gunning for a summer project is that I've had a few years now of nose to the grindstone and I was burning out, and I wanted something creative to put effort into long term and look forward to. Conversation has been dead for me, "So what've you been up to lately?", "Nothing, I work." being the literal truth, have nothing to contribute because nothing changed. Not just socially, personally too. So in that respect it's been a success. Not as much of a success as driving the car and making little upgrades as I go, but, better than another year of doing nothing but work and get fatter.

We're more than half way through spring, and I'd really like it to be drivable by the end of summer this year. I've slacked most of the last couple months, wanted a break after I mostly-finished the nightly welding march.

Anyway, did have a bunch of car updates I might get to tomorrow, will save those for another post.

594 Posts
Throw money at it! I know, not for everyone, but it's always worked for me. I've learned that if I try to do a project a little at a time my motivation seems to go right out the door. Where if I buy everything I am going to need up front. Having my money tied up in a big pile of parts that are staring at me, makes me jump in and get it done.

1,808 Posts
I was one of those who said you wouldn't be driving your car by this summer .

I realize that makes me a "Na sayer" but it's also based on my knowledge and experience of my own car build taking 6 years and 2000+ hrs to get it back into painted, semi finished drivable condition... I also own a 20 x 24 ft garage, and don't work 7 days a week.

In your case, I knew such huge project goals in such a short period of time simply wasn't realistic......

1,204 Posts
Sometimes you have to just step back and access where you are at. Take a deep breath and look at your accomplishments and not what is still needed to do. In my opinion, you have come along ways and are doing remarkable work. If all you had to do in life was work on your Opel, then I would say you would have finished and be driving it. However, who really has that kind of life, we all have to live life and there is much much more to living than work on our Opel. As much as I enjoy working and driving it, there is a lot more to life than my Opel. Keep going and documenting your progress. I for one have enjoyed reading and watching your progress.

Detritus Maximus
2,135 Posts
Read thru the molding /casting issues you were having. If I get this right, you just need an electrically conductive casting, correct? The plaster and sand molds are not practical for what you are doing?

Instead of copper or aluminum, have a look at white metal. It's a lower temp casting metal that can be used with urethane or silicone molds. You can probably buy a small molding/casting kit, if not, hi-temp red rtv (Autozone or laps) should work. Not quite the same as a pourable mold material, but you should be able to brush it on for good detail and then squirt more around that. If you want a stiffer mold material, you can even try black urethane windshield sealant. For a mold release, I've even heard people use PAM baking spray.

I used to use Eager Plastics out of Chicago, but their website use to have lots of info.

421 Posts
Discussion Starter #173
Throw money at it!
No. I wanted a fairly cheap project, not an expensive project.

The money solution is pay a machine shop to broach internal splines and turn a coupler for me.

I was one of those who said you wouldn't be driving your car by this summer . I realize that makes me a "Na sayer" but it's also based on my knowledge and experience of my own car [...] In your case, I knew such huge project goals in such a short period of time simply wasn't realistic.
Oh, I wasn't talking about people on the forum. People I know, I meant.

And, I'm not doing what you did. You fixed everything right and your car looks amazing. If that's what you thought I was aiming for, you're absolutely right, there's no way that's a 1-year project. But I wasn't aiming to have what you have. I specifically bought a car that wouldn't need any exterior bodywork or paint. Interior bodywork is okay to kludge.

I think a summer project was reasonable.

If you look at how much work I got done at Doug's place with his help in the 3-4 days I was there, all the planning on where to cut, how to cut, how to nest the interior of one car into another, stripping both vehicles, etc... especially for someone with no mechanical or automotive knowledge on their first time doing anything more advanced than changing a bulb or battery, and that had only seen a picture of the car in advance... with anything I did being permanent and irreversible. Those 3-4 days feel like I got 3-4 months worth of work done (with Doug's help). 20 zip discs and a dozen sawzall blades and 1000 decisions about where to cut and how.

I also wasn't expecting to have to know as much car stuff. I didn't think I'd be stripping out a wiring harness, doing any brake work (just unbolting and rebolting new MC), touching the dash, rebuilding the shock mount, removing the windshield, etc. I thought what I was walking into was a mostly working car that needed a trans tunnel welded, a motor, trans, and speed controller added, then the interior redone. Just my ignorance of automotive things.

I didn't anticipate the coupler being such an issue. Every other motor has just had a $20 fitting you can grab at a transmission shop and be on your way. It's a 1-hour job for most conversions, a total non-issue.

Electric conversions can and are complete as weekend projects for first-timers. They're really, really simple compared to engines.

Read thru the molding /casting issues you were having. If I get this right, you just need an electrically conductive casting, correct? The plaster and sand molds are not practical for what you are doing? Instead of copper or aluminum, have a look at white metal. It's a lower temp casting metal that can be used with urethane or silicone molds.
Hmm, yes and no. Spent some time considering that after you posted...

The electro-discharge machining process requires a conductive "negative" of what you want to erode from the workpiece. However, conductive is not the only criteria.

1 - High electrical conductivity - So electricity can flow and so that heat isn't generated passing current through its resistance.
2 - High melting point - So that the electrode doesn't erode faster than the workpiece.
3 - High thermal conductivity - To quickly get rid of heat so that the electrode doesn't erode faster than the workpiece.

This is why copper (or copper-tungsten alloys) and graphite, sometimes brass, are the only electrodes ever used. And brass only gets used in a wire-EDM process because that adds a 4th criteria of tensile strength (it's being held taught like a bandsaw blade). Copper is second only to silver for electrical conductivity, has decently high melting point, and among the best thermal conductivity too. Graphite has a higher melting point than copper, which is the most important criteria.

So let's compare white metal (tin) to copper:
- Tin will make 6.7x as much heat as copper, because it's 15% as conductive.
- Tin melts at 450'F, copper melts as 1983'F, so Kelvin-wise that's 505k vs. 1357k. Tin is 2.7x as bad.
- Thermal conductivity at room temp for tin is 62, copper is 400. 6.5x as bad.

All combined, tin will have 117.6x the electrode wear compared to copper (not 117%, 117 times as fast). That's presuming things are linear, and I think they're worse than linear.

Copper will wear around 1-5% at the rate of the electrode. That's important because you need it to maintain the correct profile and not be consumed. That means if I want to EDM a 1" deep profile out of the workpiece, I would consume 0.01"-0.05" of electrode. About 1/16".

If I used tin, it seems I would need somewhere between 1 and 6x as much electrode as I would base metal. That means for every 1" deep profile I need EDM'd out of the workpiece, I would need 1-6 inches of tin profile.

That sounds bad (117 times as bad), but honestly, if I had a silicone mold and oven-temperatures could melt it, I could easily replace the tin over and over. I'd probably spend $100 on particulate filters though, that's a slab of tin the size of my fist that would be turned into dust and need to be blocked by the filter.

It's something to consider, but, even if I've only got 1/8" of copper electrode made out of the 8 or so casting attempts, it's probably good for 2.5" of workpiece erosion, which should be the whole coupler, so it's probably not my obstacle anymore.

And, I'm close to abandoning the EDM solution regardless. It's becoming a project in itself (which was why I was excited about doing it that way, I've had that project on a shelf for a decade and was happy to revive it), and switching focus is going to end poorly for me.

421 Posts
Discussion Starter #174
Not having immediate success with buying a coupler, nor casting, nor EDM, I've spent some time laying out my options. Organizing my thoughts, not much progress to show.

1 - Hire a machine shop to fabricate a coupler.
2 - Fabricate a coupler myself, using a machine shop.
3 - Buy a Massey-Ferguson tractor PTO clutch to match the motor side, weld it to something for the transmission shaft.
4 - Continue with EDM experiments.
5 - Grind off/fill in the splines, make a taper-lock coupler.
6 - Weld the motor shaft directly to the transmission shaft.
7 - Wait for the forklift yard to maybe have a matching coupler.
8 - Carve a coupler by hand with a saw and file.
9 - Cast a coupler from aluminum or zamak.

And in detail why/why not:

1 - Hire a machine shop to fabricate a coupler.

- It'll get done right.
- It'll be the most expensive.
- It'll probably take some time.

2 - Fabricate a coupler myself, using a machine shop.

- For 5 years, I volunteered at, recruited for, and otherwise built up a makerspace, from 30 members to 240 members, including about a $50,000 machine shop. A giant lathe, a manual milling machine, a Tormach CNC milling machine, etc. I'm still paying for my membership, I'm trained on all the machinery, I can let myself in 24/7 and use whatever equipment I want, make as much noise as I want.
- It's cheating, on a simple project, to go use $50,000 of industrial machinery. Even more so to throw a piece in a machine and click a button and let the computer do it (which I could do, like, tomorrow). In as much as my project might be educational, this is the antithesis of what anyone else could get done at home. It's like commuting in a helicopter.
- I tend to be an all-or-nothing person, and I miss being there a lot, but I can't afford to dedicate every day to helping the place run like I used to. It's responsible of me to continue avoiding it for a while.
- I tend to either volunteer helping everyone else with everything, or work on my own stuff, but not both. All my projects took a 5-year pause soon after I started helping out there. I tend to cycle in and out of volunteering at different places in my life, and now is still and out time.
- I enjoy the serenity of working on something by myself, with my own tools.

3 - Buy a Massey-Ferguson tractor PTO clutch to match the motor side, weld it to something for the transmission shaft.

- Ruining something brand new just for the splines.
- Kind of expensive.
- Still only a partial solution, still have to kludge the transmission half.
- Probably won't be very deep, (might need 2 couplers).

4 - Continue with EDM experiments.

- My EDM machine has been on pause for 10 years, it would be nice to finish it and use it.
- I'm most interested in this.
- I've already sunk a bunch of time into it.
- Casting hasn't gone well but is becoming passible.
- Solution might not be sufficiently accurate. EDM has small, but existent air gap between electrode and workpiece by function of the process itself. EDM can be about 0.0005" accurate (half-thou), but that would be around a circumference so 0.001" total. Automotive fit is usually around 0.001" for sliding parts, so I would have to be at the limit of the technology on a home-built machine. This isn't a rigidity constraint like on a milling machine, it's a configurable attribute related to voltage, current and frequency, but the construction of my plunge die has to be within those tolerances for that to matter. Alternatively I could undersize my electrode to compensate, but that means I can't just cast an image of the shaft, I'd have to machine it smaller, making the whole process moot.
- I risk going off on a tangent project that will sink too much time (for example, as it already has).

5 - Grind off/fill in the splines, make a taper-lock coupler.

- A taper-lock is just a barrel of aluminum with a narrow slot in it and some bolts to squeeze that gap (nearly) shut.
- It's apparently as strong as a weld.
- It's proven strong enough for DIY EVs in the past, it's the common machined solution.
- I don't like damaging the motor splines because I can't undo that later.
- I could fill in the splines with JB Weld and leave it large, it's just surface to squeeze on.
- 2" aluminum bar in short lengths is probably as much as a PTO clutch would cost, and I doubt my machining ability with steel.
- I could also/instead cut a keyway and use a keyed shaft. But I presume there's a reason that keyed shafts are never used like this even though that's simpler and easier.

6 - Weld the motor shaft directly to the transmission shaft.

- Shaft is permanently damaged if I screw it up.
- Can I create a jig to weld it sufficiently square and centered?
- Solution in general is pretty permanent, most other solutions I can just abandon and try something else with no consequence.
- It limits serviceability or making other changes.
- This has always been my backup plan, and tempting.

7 - Wait for the forklift yard to maybe have a matching coupler.

- Forklift yard mentioned they have 2 more machines like the one I took the motor from, that they're combinging into one. They'd pull the gearbox for me so I could recover the matching splined shaft. Though they're not sure when, or if they have a buyer for that. Or time to do it.
- This has been my backup plan for a while, knowing it might be "a month" or more, but, it's been 6 weeks not no call yet so, probably won't happen.
- I have no idea how long I might wait for that to happen.
- Is still only half a solution.

8 - Carve a coupler by hand with a hack saw and file.

- Seems like the worst choice, but it's so, so simple, and unlike welding, not a permanent commitment. It just takes time. And honestly not that much time compared to fancier solutions.
- I've done it before, and it worked then.
- I hate this.

9 - Cast a coupler from aluminum or zamak.

- Instead of casting the female image in a plaster mold, then pouring copper into this mold to create a male image, then using the male to EDM out a matching female, instead just cast the female directly.
- Make a silicone mold of the splines, use that only as a shape to pack a high-temp sand mold, pour metal into the sand mold.
- Aluminum has a lower melting point than copper, so if I could cast copper I can easily cast aluminum.
- Zamak (Zinc/Aluminum) is stronger than bare aluminum, melts at a lower temperature, and is considered "zero machining" surface finish quality.


As much as I hate throwing away my casting and EDM efforts of the last 6 weeks, and I'm confident it'll eventually work, I'm worried that I'll get it working but it still won't be good enough regardless of my effort because of physical limits. I guess it's time to put my EDM project back on the shelf.

I think I'm going to try casting a Zamak coupler. Opelbits' suggestion of using silicone mold and tin gave me the idea that, the only reason I can't cast a female directly is because I don't have a female mold. A plaster mold wouldn't survive being used as an image, but silicone would be fine to pack sand around and then remove.

If that fails, Plan B is an aluminum taper-lock.

If that fails, Plan C is to just weld it up solid. Or braze it so there's less warpage.

421 Posts
Discussion Starter #175
Onto less boring things, progress and pics!

Casting Progress:

- I cleaned up the other half of my borrowed garage, and built another pair of 8' long 2x4 shelves. So much room, for activities!

- The very last thing I did was move some cheap plastic wall-hanging shelves. Behind which I discovered... a 240v 50a outlet. Perfect timing, as I'm done pretty much all my welding with my crappy 120v 15a welder. sigh

- I finally got 2 molds to turn out decently. The secret was to be less ambitious about depth (1/4-3/8"), and, to mix yellow fiberglass insulation into the plaster so that it wouldn't crack across the whole mold when removing it (it worked).

- New casting (right) turned out great. On the outside. I forgot to pre-heat the mold so, steam boiled out of the mold when I poured, and volcano'd all the copper away. The surface is all that's left. Which is fine for EDM, I'm only using the surface. Surface finish is also great, only needs light touchup:

- Ruined my steel crucible by letting it get too hot. Top 1/3 of my copper poured out into the sand, bottom 2/3 is kind of stuck in a furance I can't use now. Oops.

- No problem, I'll go back to using my firebrick directly. Oops. I liquefied the firebrick into glass and then burned through it too:

- It probably makes sense if I explain that I upgraded my carbon arc furnace before this, it puts out a lot more heat now. Probably close to 500a at 24v.


- Let's have a closer look at the master cylinder.

- What about the brake booster?

I might have wrecked it, trying to unscrew the vacuum attachment. It now sloppily spins (but didn't come out of the metal enclosure).

- Front brakes apart.

Stripped one of the j-tube fittings (no problem, I'll screw it back together using vice grips, and just always disconnect the other end of the rubber hose in the future instead). Found broke the cross-shaped brake pad clamp on one side. Pistons okay on one side, other side the rubber was cracked.

- A transformation!

How did I restore them so well? I didn't. I bought them out of the local parts co-op, recently donated from another member who wanted to use them but upgraded to the big brake package instead. Ended up being free because of parts I've contributed that I didn't need (like a good gas tank from a desert vehicle).

I had no interest or skill in refurbishing brake parts so I just paid Keith to do it for me. Great value and peace of mind on something I don't care about doing and just want done.

My old reservoir had a matchbook-sized hole in it, and when Keith tried to pull it off, it half shattered, half crumbled in his hands. Desert car - win on the metal, lose on the rubber and plastic.

- My brake fluid reservoir cap was also cracked, so I took it to the junkyard and practised on 30 or so vehicles until I found one that matched. '04 and '07 Mazda 3 had screw caps (not tabbed caps) that seems to fit perfectly. $2.

Windows and Doors:

- I spend a couple days reading old threads about converting the Opel GT crank mechanism to power windows. Consensus seems to be 1996-2000 Honda Civic sedan, front window regulators are a good match. They are known to be durable unlike other brands.

Despite so many threads, there's not a consistent write up, nor labelled diagram, and conjecture is mixed in with proven results so it's hard to know what to door use. I'll try to write an article about that if I can colimate the advice correctly.

In brief:
- Flaw of original Opel mech is that both cables go to bottom, which is a hard angle to follow.
- Honda Accord openers both go to bottom, hence the Civic ones which form a bow-shape instead are maybe better. They're also close to the correct length.
- Using the actual slider seems to be abandoned. Even the Civic ones are a couple inches too long and arch is wrong and the attachment points are wrong. So instead you remove the slider from the donors and mount the motor and cables to the original slider mechanism.
- It might not actually matter what regulator you use because of this, I think this might be Xerox advice where each person copies the person before but have lost the context of why. Like the recipe that tells you to cut the chicken in half before putting it into the roaster, you eventually discover is because grandma's roaster didn't fit a whole chicken, it has nothing to do with improving taste, spice coverage, moisture, etc.
- The cables end up getting cut anyway (do not let them go loose, hold them with tape or they'll tangle in the spring mechanism) and then just clamped into the Opel slide mechanism.
- Opel GT windows get very tight towards the top. These little motors are weak and barely get the job done. Doors are often crooked without you noticing. The rubber sliders that the glass contacts is adjustable and many people don't know this. There is a procedure for this to help if needed.
- Maybe you should just get motors from bigger vehicles with heavier windows instead of Honda Civics, since the slider and cable length no longer actually matter, no one has ever re-used the slider mechanism anyway.
- These things are no cheaper from a junkyard than they are brand new. So just buy new ones.

- Quarter-panel automated window openers. From 1996-2000 Dodge Caravans. Or, 2000+ Caravans, but they'll be backwards so you'll have to flip left and right.

Don't try to disconnect a spring clip, if you want the ball-joint off, just look at the direction it's mounted and yank it off. The spring stretches to let it pop on and off. However, jab a screwdriver around the window cup (the whole circumference) and twist, and you can yank the suction cups too. Then you won't need to add couplers later.

Nothing fancy to access them either just yank the plastics out. Two 10mm bolts hold the actuators on. $5 apiece at my junkyard.

Caravan switches are lighted switches, DPDT momentaries. Mine were $15 for the driver's assembly and $5/switch for the passenger. $15 extra for the wiring harness (I should have just cut it off a few inches after the plug, would've been free).

Many vehicles have shielded, press-fit rubber connectors to the doors. Caravan ones kinda suck, they're very short. The Civic ones were quite long, comparatively. Didn't look at enough vehicles to recommend any particular vehicle as being a good match for the access room in a GT'd door.

1,204 Posts
Well Matt I would say you are back in the driver seat and driving :) Progress is progress no matter how far you go as long as you are moving. Looking good and still on track with what you set out to do. I like it :):)

Detritus Maximus
2,135 Posts
9 - Cast a coupler from aluminum or zamak.

  • Instead of casting the female image in a plaster mold, then pouring copper into this mold to create a male image, then using the male to EDM out a matching female, instead just cast the female directly.
  • Make a silicone mold of the splines, use that only as a shape to pack a high-temp sand mold, pour metal into the sand mold.
  • Aluminum has a lower melting point than copper, so if I could cast copper I can easily cast aluminum.
  • Zamak (Zinc/Aluminum) is stronger than bare aluminum, melts at a lower temperature, and is considered "zero machining" surface finish quality.

I think I'm going to try casting a Zamak coupler. Opelbits' suggestion of using silicone mold and tin gave me the idea that, the only reason I can't cast a female directly is because I don't have a female mold. A plaster mold wouldn't survive being used as an image, but silicone would be fine to pack sand around and then remove.
I wasn't sure of the temp requirements of the EDM, so white metal does sound like a bad choice.

You might consider the lost wax method. Instead of having to remove the silicone item from your sand mold, make the master (image) piece out of wax. The wax can be either be cast from a mold or you can machine a chunk of wax as if it was metal. The idea is that the wax stays in the mold when you do the pour. The wax supports the sand and will melt and boil out as the molten metal goes in.

421 Posts
Discussion Starter #178
I wasn't sure of the temp requirements of the EDM, so white metal does sound like a bad choice.
It's done cold, at room temp, in a bath of kerosene.

But, the spark erosion (the ED or Electro-Discharge of EDM) is, at a microscopic level, still a thermal process. An infinitesimally small particle is instantly vaporized for the few millions of a second that electricity is flowing there. The electrode will also generate heat in the same momentary timeframe as it's the other point of the arc, and when power shuts off, hopefully the electrode just barely survived before getting hot enough to vaporize itself. If it did, no erosion of the electrode. Tin would basically guarantee constant erosion on the workpiece, maybe to a point where it would extinguish the arc and never wear the workpiece.

The idea is that the wax stays in the mold when you do the pour. The wax supports the sand and will melt and boil out as the molten metal goes in.
As I understand, that's not quite correct.

  • Lost-wax casting is baked/burned out ahead of time until a hollow cavity remains.
  • Lost-foam casting might be what you're thinking of, you leave the foam in the mold and let the pour burn it out.
Wax won't burn up anywhere near fast enough to just pour the metal onto it and hope it disappears (imagine trying to burn a whole candle in 2-3 seconds, it has a tremendous amount of energy) and it doesn't shrink, it just liquifies and burns. Wax takes a couple hours to burn itself out of a mold, you have to do it in advance.

Foam satisfies the burn out constraint, if you've ever pointed a heat gun at styrofoam, you'd know it vanishes almost instantly. There's almost no energy in foam. Even then, most people will burn out their foam because the expanding gases will resist the flow of the metal and ruin the pour.

That said, doing an oven burnout ahead of time is not an issue for me. So that would be fine.

But, that's not my constraint. What I need is something that automatically forms (i.e. liquid at some point).

Foam does not automatically create itself, I would have to machine it into the correct shape. That's almost as difficult as machining the metal itself, I have to make ultraprecise splines into the foam. So foam isn't helpful to me.

Wax satisfies the automatic form requirement (liquid wax can be poured onto the driveshaft to copy its shape flawlessly).


That's not normally what lost-wax is used for, or the problem that it solves.

Lost-Wax is useful if you have to cast an impossible shape. For example, if you were trying to cast an hourglass shape vertically, there is no way to create a 2-part mold (cope and drag) that could release the object being cast. It's trapped. Anything with both concave and convex features at the same time will create an impossible draft no matter how you orient it. So instead of casting the object, you machine the wax and then leave it in the mold and burn it away in an oven to create the cavity.

This is common for complex parts, or jewelry. You create the object in wax, then use casting to create it in metal.

My issue isn't that I have an impossible shape, not directly. My issue is that I don't have the shape at all and can't create it manually. I need to use a liquid form just to create the shape I want to cast. If I could accurately carve the driveshaft image out of wax with a knife, then yes wax would would solve the shape issue. But I can't.

Wax has the same challenges as plaster and sand. I need it to survive the forming process, and it probably can't.

The challenge I'm having is that I want to cast something that is not designed to be cast. There is a whole science to die making and casting that has to do with how you get the object out of the mold. Mainly, what's necessary for draft angle. Not only do I have zero draft angle (it's a cylindrical, not conical shaft), the shaft has many fine details and ways to be gripped. This is why my plaster and sand molds keep breaking, because they're brittle materials trying to unstick themselves from a perfectly flush surface with no clearance gained when they begin to part (the mold doesn't depart from the image until it's slid completely off of the shaft, it's touching perfectly tight the whole time).

I don't know that wax is any stronger than plaster, if I tried to pry a wax mold off of the shaft, it'd likely just break too.

Silicone seems to be the perfect solution because it's not brittle, it's flexible. So I can yank it off of the shaft without damaging it.

Silicone has another challenge, it's sticky and it bonds to everything while it's curing. Yeah I can make it match the driveshaft perfectly, but it would be like bathroom caulking, bonded to the surface and not let go.

The way to resolve that apparently, is to mix it with cornstarch. So that's what I tried last night. I squirted out a half-tube of some left over silicone bathroom caulking, into a basin with some corn starch. I neglected to remember to use a stir stick to do most of the mixing and tried to mix it with my hands. Disaster.

Until you have enough corn starch mixed in to mess up the stickiness, it's like mixing tar. You can't even use your fingers to scrape it off your fingers, it just gets stickier and stickier until there's nothing left in the mixing bowl, it's all stuck to your hands.

But eventually the surface stopped sticking and it was like a lump of chewed gum, soft but not sticky anymore.

So hopefully what that means is that it's malleable enough to capture the shape of the driveshaft splines, but also not sticky enough to actually bond to the metal.

Takes a while to set, so, I'll find out tonight if it worked.

The next difficulty is how do I use the silicone to create a mold for sand or plaster?

Typically you just cut a zig-zag pattern into the silicone so you can clamshell it out of the part when you're done.

But I'm not using silicone as a mold, I'm using silicone as a image to create a high-temperature mold. Ideally I'd be using sand, but, will a 2" tall delicate tower of Green Sand survive me pulling apart on a silicone mold around it? I doubt it. It's just sand.

My plan earlier was to EDM a bunch of 1/4" steel plates and then weld them together like a stack of pancakes to create the 2" tall coupler. But I can't weld a casting and keep any strength, I have to cast the whole part all at once.

So that means I should create a plaster mold of the internals first, let it harden, then remove the silicone.

I can't remove the silicone if it's the full size of the coupler, so I have to try to create the plaster tower inside first, then add a tube and do the outside after.

I suppose if I have 2" of grip on the motor splines, I should have 2" of grip on the transmisison shaft's taper-lock, so now I'm having to cast a 4" tall tube. That's a lot of metal, and a lot of things to go wrong.


I would have liked to cast iron, but I don't think plaster will hold up to cast iron temperatures, and I don't think Green Sand will survive silicone peeling away from it. So, probably sticking with Zamak.

I don't have a graphite crucible, I don't have firebrick large enough to hold that much volume, and I'd just use a soup can or a skillet but apparently Zamak is very, very sensitive to iron content (as in, better be below 0.1%) else it autocorrodes. Not as badly as it would if there was lead or cadmium contamination (0.004%), but, yeah, iron ruins the alloy. So I wonder if a steel vessel would be unacceptable.

Detritus Maximus
2,135 Posts
You are correct about the lost wax having to burn out. I haven't done that, but my wife has. But you're right I was probably mixing lost foam/wax.

I wasn't terribly detailed in my thoughts on the wax master.

The silicones and siliconized caulks can vary quite a bit in viscosity. I find the tub and tile clear 100% silicone (GE is my favorite at work, but that is for a different application) flows better, but is still not more like a liquid. I have used it and others have as well in making open face molds. The trick I use, although I hate to call it a trick because I do this even with pourable silicone or urethane, is to brush it on and into the shape to be molded. I generally do this to avoid having voids and bubbles at the surface of the mold. Voids will produce warts on the surface of the casting, bubbles will do the same if pressure is used while making a casting in the finished mold. I have had extremely good results brushing on a skin of silicone, letting that dry or cure, then backing it up with additional silicone. If you are making a thin wall mold and want to save material cost, a plaster mother mold can be poured around/over the silicone to provide a naturally fitting support for the silicone mold. Aside from the material savings, this creates a thin wall, stretchy mold that can be peeled off with ease, even with lots of detail or dramatic drafts.

If you like the idea of the wax master, this is what I would do (I am assuming you are trying to make a mold of a female shape in order to make a male master for the EDM?):
-get some real molding silicone....find a hobby shop, a decent one. Hobby Lobby might have the required molding silicone, but more of a dedicated serious hobby shop. If they sell photoetch detail kits for model planes and tanks, you should be on the right track.

-do some more research on making a mold with an outer support (mother mold or a mold box). For what you are doing a single piece mold sounds best. Two piece molds have part lines that are almost always visible in the surface of your casting. Not much fun to clean up, especially in a splined drive coupler where tolerances might be critical.

-have you been using mold release? Even silicone will stick to metal and tear when you try to remove it if you don't use mold release.

-make a mold of your original. The wall thickness of the mold will depend on the dimensions of the original piece. Say you want to replicate the end of a splined axle shaft 2" dia.but only 3" of it's length. The easy thing would be to take 2.5" plumbing PVC pipe about 4-5" long (nice flat cuts on the ends, no angles!) and slice it open one end to the other (this allows you to stick a screwdriver in it and pry it open) then cover the split with duct tape. Position it over the shaft until the pvc extends past the end of the shaft by about 1/2". As parallel to the shaft and as evenly spaced as you can get it. You can use modeling clay on the bottom end to hold it in pace on the axle shaft and seal the bottom of the pvc to the axle. Make sure you have mold release coating the splined shaft and the inside of the pvc.

-mix the silicone. Mix it really well. I use two containers, one for the initial mix and then pour it into the second. The reason for this is the first component into the first container creates a skin on the inside surface of the container and does not necessarily mix. When you pour that into the mold the last bits in are not mixed and and can create a sticky uncured area that makes more problems. I solved that by mixing, then pour into a second container, mix some more then pour.

-as I said, I like to brush the first coat on if possible. Molding silicone can be really runny, so it is easy to tell if you have bubbles or missed areas. Hopefully you project is shallow and can be reached with a brush (the cheap metal tube ones with stiff bristles work well and are disposable. Slowly pour the silicone into the pvc tube (hopefully, the duct tape and modeling clay will not leak out the silicone). Make sure it is flowing well. Too fast and you might trap air. Fill to the top of the pvc.

-let cure.

-remove duct tape and modeling clay. Use a wide blade screwdriver or stiff putty knife to pry open and remove the pvc. This may take a little gentle wiggling and prying as you want the pvc to release from the silicone without really disturbing the silicone to axle contact. Once it is off, work the mold loose starting at the bottom (where the clay was) and work your way up. The silicone should in this example would have a wall thickness of 3/8" maybe? 1/4" to 3/8" is decent for something this size, flexible but able to support itself if you stand it on end once it's off the original piece. Some people will cut one side to create a split line, but I like to avoid that. It creates a split line in the copy and the mold can start to tear at the bottom of the cut.

-now, if you just want a wax copy of the original male axle shaft you are done except for pouring in the hot wax.

-if, however, you want a wax female of the male axle shaft, you need to make an intermediate casting. There are two ways to do this, one is risky.
First the risky one...coat the inside and top of the female mold with mold release. DO NOT MISS ANYTHING! Mix, brush and pour more silicone into the mold and let cure. The silicone will stick to the silicone mold if you missed getting mold release anywhere. If you got it everywhere, the silicone male should pop out of the silicone female. his

Second way, get some casting urethane to pour into the female silicone mold. Mold release will help with demolding, but the urethane will not stick to the silicone. Urethane is a higher durometer so it is stiffer and more rigid. It is not flexible like the silicone.

At this point it is a matter of creating a second mother mold/mold box to support the male casting while making a wax female casting.

The silicone male would pull out of the wax female easier than the urethane, but at the risk of a failed casting sticking to the original silicone mold...back to square one.
Urethane eliminates that problem, but may be harder to remove the female wax casting.

the above gives you an idea of a male original to female wax copy. With some thinking, it pretty much applies to female original to male wax copy. It's just a matter of rethinking the mold boxes
I think your shape is a relatively easy one to do.

Attached is a mold I did for an exhibit on trransplants. I had to make molds of BioLike organs (same weight and feel as real organs. Floppy and squishy is hard to mold...) to be cast in urethane with plastic inserts to hold barcodes. Another set had 1/4" stereo jacks in them to plug in cords that would trigger a video showing info on that type of organ.

SLSC Transplant009.jpg

Transplant Mold6.jpg
Transplant Mold5.jpg

1,204 Posts
That is incredible work you guys are doing. Opelbits and MattsAwesomeStuff, you two are really putting the meaning of science into restoration :) I must admit I'm lost on some of it, but very intriguing though. I do enjoy reading how you are using other methods in restoration.
161 - 180 of 186 Posts