@chrisp Hey, I'm right there with you, but also am not making patch panels from scratch. Superfastmatt has a video where he says 3D scanners are refined enough for you to have one in your garage. He later used one to check wheel alignment! I'd like to get electricity in mine! I could see stuff like a later dash scanned and narrowed to fit a Model A. Floorboards with the ridges planned and laid in. There are a lot of guys here that can form up a trans tunnel, but if they had 4-6 different sizes with flanges? Easy to repeat and sell. Additional cost to modify and run a one-off. Dry sump tanks or gas tank halves that fit a strange shape. All of this is semi-production stuff. So long as it didn't become a sea of look alike junk like the billet stuff, it could foster fixing and building all sorts of stuff.
Same with panels that were banged out on the big presses when they were forming sheet metal for the old cars we fart around with. You can't make a bubble in a flat sheet of metal without some stretching and thinning.
I see the want for technology like this but I'm also sad to see skills like yours go by the wayside. Hopefully in the future there will still be some people with the same attitude to keep the torch lit (pun intended). But I also hope tech like this won't push out hands on people out of business.
Ideal for a high-end build where the builder is in search of a part that cannot be found. More for them that could afford it and need the quality that can be produced in the amount of time. Most of us have the time, maybe the need but not the money to buy a part made by this method unless they are producing a mass part for inventory. Then the cost would go down but not exponently enough that we could afford to buy it.
I’d thought a 3D scan requires a part to be scanned? Ie no 35/7 Ford PU fenders in steel available. Only fiberglass, so could the fiberglass ones be used? Or say no panel for a scan, can it still be made?
You can scan anything But a copy of a copy isn’t the best but better than nothing The best would be to find the best example to scan
Dang.thats over my head On a side note. Had an uncle that built furniture as a hobby. A machinist by trade. He seen a picture of a piece of furniture he liked. Luckily it had a baseball in the pic. He measured the baseball to scale the furniture. I guess a smart programmer can use the glass and refine as needed. Or make sure the glass one fits and has crisp details before scanning. I’ve seen parts scanned for a CNC and the operator change the scale, reverse it, add or remove details. I kinda wonder about the depth of the parts made on these shaping tools. Could an entire fender be produced or in sections kinda like how you would by hand.
Yes in a sense. You don't just magically put a picture into a computer and have it transform it into a part. You would have to create a digital design, 3D in this case, just like you would for any other cnc machine. Not as easy as scanning but it can be done.
Yes. I have made parts from pictures. They need to be really good pictures. I also have a few 3D scanners, all of which are portable. If I can get to where an existing part is, I can get it into CAD. Not cheap, though.
This tech works great, but will not be available to small shops or individual users anytime soon, so don't toss those tools just yet. A Figur G15, as shown in the video, is just a tick under $500,000.00 USD. All tech drops in price over time, but a half-million is going to take a few minutes.
I know, that's why I pointed out that the money saved on the fabrication of the part will be spent on the file necessary to run the machine, so I think that for many years to come it will be out of reach of regular Joes to have a 1929 Hupmobile fender fabricated that way just like paying a panel beater. But if you have a really good friend or family member who's willing to do the rendering at a discount or for free it's viable because it's gonna be cheaper to fabricate the part that way than to pay a guy with a big hammer who thinks he can fix anything with it. Do you want to be my really good friend? Edit : they now can extrapolate 3d images from 2d using AI, how accurate it is I have no clue, but that opens new windows.
The perfect example of this is the Tucker prototype the Ida's are creating from just original renderings. But again it's massive amount of 3d renderings that is not free.
When the guys decided to do Cobra bodies in aluminum by east European aircraft builders, they scanned a few original cars. None were the same or even symmetrical. Using the computer models from the scans, they used the best side and mirrored it. That answers the scan question. Also, it was a huge undertaking 2 decades ago. https://www.kirkhammotorsports.com/about.shtml As for a missing part, it can be designed entirely on the computer. Say there is only one Budget 36 car and the LF fender is missing. You can either scan the right fender and mirror, or plot the mounting points and design the shape from scratch.
That computer cannot recreate the human flaw that is in every part to give it character, we drive and cherish cars of character not perfection.
You can include flaws in the 3D design. Fine features down to 0.005" can be cloned. I epoxied together a broken hose manifold for a vintage fire truck, scanned it, blended in the existing defects across the epoxy, and then machined a new, unbroken one, with all of the original wear lines, chips, and gouges. Artificial aging made it indistinguishable from the original, just not broken, with missing pieces.
I'd bet the bulk of the use of that machine will be in industry segments other than antique car parts. As mentioned previously, the cost in a one off part will be in developing the model and the code. While the idea if producing rare or non-existent body parts is attractive, the truth is 99% of the population doesn't have pockets deep enough to afford it. Brad Keselowski, the nascar driver, owns a business in Statesville, NC where they have multiple 3d metal printers as well as CNC machining centers where they can do both additive and subtractive operations on complex parts. There was a show about the printing side of the operation on TV a year or so back. Of course, you automatically think of race car parts given the owner of the business. But the man who runs the operation and was dong the tour said car parts of any kind make up a minor part of their volume, something on the order of 15% IIRC. The rest is primarily industrial and aerospace. I wouldn't be surprised if the same sort of percentages would apply here, assuming the goal is to keep the machine running as many hours per week as possible.
Any idea of a cost to scan a fender and have it made? I’m not trying to put you on the spot. I realize this might be something new, was just curious. Edit: Say I have a fender, have large $$ quote to repair it, vs the cost of sending it to a place and have another made.
Parts on demand. You place your order they put the file in the cue and out pops your fresh part. Part of the cost of each piece now is the amortization of tooling - if a die costs $5,000 and you get 1,000 parts out of it you add $5 to each part. You also have to pay for machine setup/changeover the storage space to stock the parts and eat a few of them that always seem to get dented and dinged while sitting on the shelf. The only downside I see is if you want 1,000 parts this process is much slower than a punch press.
Only if they are scanning your specific vehicle to make parts for it. I scanned 5 original 1936 Ford passenger side front fenders.. None of them matched each other.
That's not a thing, exactly. That's a whole bunch of things, all of which require knowledge, equipment, software, and time. Time equals money. Equipment equals BIG money. Assuming one owns a scanner, scanning it not a "click button, get 3D CAD model" endeavor. Scanning generates a point cloud. That point cloud needs to be processed by software to turn the space between the points into a polygonal mesh, and then turn that elaborate array of polygons into a parametric surface model. That process alone can stuff even the most powerful computer. Once that parametric surface model has been generated, it then can be imported into CAD software. Pictures of those steps are here: https://europac3d.com/products/polygonal-mesh/ Like most tech, this will drop in price, but the labor charge will still be there. One fender scan, inside and out, with image processing and export to most common CAD formats, depending on the complexity, is upwards of $2k, and that gets you a CAD file.
Yup. Vintage OEM Ford parts don't fit all vintage Fords. Their '32 is an averaging out of the samples they examined. Yes, may repair panels are truly terrible, but even the best ones may, or may not fit.
Yep. They’re not even the same side to side on the same car. So patch panels made with this technology could about as accurate as current patch panels.
Unless the exact vehicle to be patched was scanned. That is possible, and I can do that now, but it ain't cheap. If anyone needs to prove the variance for themselves, and you have a Model A, take tissue paper and wrap the dip in the front fender, ahead of the front wheel. Tape that tissue paper to make a mold of that area. Pull it off, turn it inside out, and go try to match it up to the other fender.
They did that at work with 32s. The fenders are stamped different from side to side. The 33 Plymouth I did was the closest to perfect side to side for patch panels. A friend of mine built a cool jig table. 10x20 metal floor with 1/2 inch holes cnc plasma cut on a grid across the entire thing. Has a measuring set up that he can compare side to side measurements. A lot of effort to learn these old bodies ain’t square. Still gotta fit the sheet metal no matter what the measuring says.
Back when the fiberglass dune buggy craze was at it’s peak, copies of copies created finished bodies that were actually wider than a VW floor pan!
I watched that video that had bad chad in it, glad he wasn't the main issue of that video. That being said, the technology that exists now, such as these cool metal forming machines, well they have a place no doubt, and they are not going out of style anytime soon. Just a friendly reminder to all of us, there was a time, when the simple toothbrush was a manually operated devise that didn't need to be plugged in or recharged! Everyone's results may vary greatly!
The Jalopy Journal
