My sister-in-law has an old (very expensive) Zero-King fridge. The door shelves are falling apart - cheap plastic end-caps are disintegrating. Seem irreplaceable (can't find any on-line).
My question: Is there a way of somehow scanning the existing end caps (complex shape) to allow new ones to be created using 3D printing?
Thanks in advance. Fowler

Simple answer is - yes, that's entirely possible. Unless you have the time, knowledge & a bug up your ass to do it this way, you could replace the fridge cheaper.

I'm an engineer & accomplished half-asser. I have a $400 3D printer & use cheap-ass material. I can fumble my way around CAD programs enough to make almost anything I've needed to make (I don't have a 3D scanner). Even if I ignore the already sunk cost of my equipment & work for free, I've come to the conclusion it's not worth doing something like you suggest for myself, let alone for a business.

The equipment & knowledge to do a professional job of it aren't cheap.
Scanning for mfg is as much art as science.
Depending on the shape, you might be able to get away with cheap material/printing or you might need some un-obtanium material & a printer costing 6-figures. (I seriously doubt that it's that bad, but you get the idea).

Rough guess, cheapest way to do it would take maybe $50 in materials ~10-15hrs of 'engineering time' to scan & make usable files + ~25hrs of print time. For someone to do this as a business, I'd say you'd lose money charging $100/hr.

As mentioned in the previous post, its totally doable but probably horribly not-economical.

The technology to scan a part comes in a form called a CMM. There are many models/types of coordinate measuring machines, but none are cheap and all require specialized skills to operate. It would be expensive.

Your best bet, if your not concerned with cost and want to keep the old fridge for sone numismatic reason, would be to have someone good with 3d solid modeling, to work up a solid model of your part the old fashioned way, with a caliper and micrometer, and produce a .stl file, which you could use to have the part 3d printed.

In the end, itll almost certainly cost more than a new fridge.

My degree is in Mechanical Engineering Technology, which is similar to Mechanical Engineering but with less focus on the analytics and more on the technology.

If it has some complex geometry, it may take a guy who earns 30.00 per hour about a day to work up a solid model of the part. You can see how this is fine for mass producing but not so much for a couple parts.

Im pretty competent with Inventor and Solidworks, and for a small part with complext geometry, paying a firm like, say, Everett Engineering, to solid model the part would cost the equivelent of a new refrigerator.

quote:

Originally posted by Ken226

My degree is in Mechanical Engineering Technology, which is similar to Mechanical Engineering but with less focus on the analytics and more on the technology.

Purdue? MET class of '05 here.
Does anybody ever let you get to '-Technology' before their eyes glaze over? I've almost given up not being lumped in with those dumbass MEs

Mines from a no name community college in the middle of nowhere, AZ. But, it was Summa Cum Laude, so its the equivalent of the 9th grade in any other state☺

And noone ive ever talked to had any idea theres a difference between ME and MET.

To the OP, if your a quick study, you could download a 30 day trial copy of Alibre Atom or Solidworks, buy a caliper and a micrometer at harbor freight and start watching youtube videos. Atom and Solidworks are surprisingly easy to learn once you learn the various ways of transitioning 2d drawing into 3 dimensional space.

Did you try emailing the company and asking if they have the parts?

quote:

Originally posted by Ken226:
The technology to scan a part comes in a form called a CMM. There are many models/types of coordinate measuring machines, but none are cheap and all require specialized skills to operate. It would be expensive.

There are more hands-off, automatic ways to "scan" 3D shapes.

Generally speaking the machinery works by projecting a line or grid onto the part, typically using one or more lasers, then rotating the part in place while using one or more cameras to measure the shape of the surface of the part.

Of course, the file that comes out is usually not perfect and requires some tweaking to turn into a usable solid model.

A couple of years ago I bought a kit for a simple one on Kickstarter. Still haven't gotten around to putting it together. The kit's a couple hundred dollars.

http://store.murobo.com/atlas-3d-kit/

FWIW, I have actually done something like the OP is talking about. However scanning was not my aproach. (I am a product development engineer with access to the tools and software to do this kind of thing)

We had an expensive printer go down at work because the gear that operates the paper elevator broke. The entire assembly cost about $500 and was a week away. I pulled the gear and with a set of digital calipers and the use of AutoCad Inventor and our 3D printer, had the machine back up and running within 2 hrs. We printed 2 spares and a few weeks later when another printer (same model) did the same thing, we had it running in about 10 minutes.

If the part can be measured and drawn, it can be printed. The only real issue I see with a refridgerator part is that you would want to orient the printed part in it's strongest direction as it relates to the loading on the part. There are some glass filled plastics that can be printed with but they are only strong in one orientation.

The strength vs. orientation has to do with the way the printer lays down the plastic in layers. Loads perpendicular to the layers will fail from the layers pulling apart. Loads in-line with the layers are much stronger.

Ken

quote:

Originally posted by maladat:

quote:

Originally posted by Ken226:
The technology to scan a part comes in a form called a CMM. There are many models/types of coordinate measuring machines, but none are cheap and all require specialized skills to operate. It would be expensive.

There are more hands-off, automatic ways to "scan" 3D shapes.

Generally speaking the machinery works by projecting a line or grid onto the part, typically using one or more lasers, then rotating the part in place while using one or more cameras to measure the shape of the surface of the part.

Of course, the file that comes out is usually not perfect and requires some tweaking to turn into a usable solid model.

A couple of years ago I bought a kit for a simple one on Kickstarter. Still haven't gotten around to putting it together. The kit's a couple hundred dollars.

http://store.murobo.com/atlas-3d-kit/

I love that DIY coordinate measuring machine. I may buy that to play with. Very cool.

I bet the OP could learn some CAD, buy a few tools and a 3D printer and with a little youtube study time, end up with useable parts. And a very cool new hobby.

I could CNC some for you out of 6061-t6, but it would be a long ways from economical.

Not your exact answer. I don't have a 3 d printer, nor a 3 d scanner. But I do have a very old fashion milling machine. I have made a zillion parts to save old stuff. Yes they don't have the svelte lines of the originals, but most of these type of parts are very simple in context. You can probably find someone to make a workable part for not too much money.

Get a resin casting kit.

$20 sees you doing all you want until the stuff runs out.

tac

I have been training in 3D technology at work.

I have been using a 3D scanner and printer for a few months now.

What you want is a serious undertaking that will require a lot of time to scan, clean up and scale.

The printing is the easy part once the art has been generated.

There is a good chance the part might be available on thingaverse or some other 3d parts site.

quote:

Originally posted by KenS:
FWIW, I have actually done something like the OP is talking about. However scanning was not my aproach. (I am a product development engineer with access to the tools and software to do this kind of thing)

We had an expensive printer go down at work because the gear that operates the paper elevator broke. The entire assembly cost about $500 and was a week away. I pulled the gear and with a set of digital calipers and the use of AutoCad Inventor and our 3D printer, had the machine back up and running within 2 hrs. We printed 2 spares and a few weeks later when another printer (same model) did the same thing, we had it running in about 10 minutes.

If the part can be measured and drawn, it can be printed. The only real issue I see with a refridgerator part is that you would want to orient the printed part in it's strongest direction as it relates to the loading on the part. There are some glass filled plastics that can be printed with but they are only strong in one orientation.

The strength vs. orientation has to do with the way the printer lays down the plastic in layers. Loads perpendicular to the layers will fail from the layers pulling apart. Loads in-line with the layers are much stronger.

Ken

There's also things like Raptor PLA that gets stronger after you print it by running it through a dishwasher. I don't know the relative strength vs fiber-filled, but it corrects for a lot of the layer-separation weaknesses of the cheap 3D printing materials. I've used a heat gun to make sure some craptastic high-warp ABS didn't separate at layers....but hey, it was $12/kg & used to make light-duty racks with 1/2" PVC.

I would guess that ABS or PETG, maybe PLA would work for this, but you might want to beef up the weak points vs the injection molded originals. 3D scanning or CMM would replicate the object as a point cloud, but then it's a pain in the ass to modify.
It probably wouldn't be that hard to draw up in CAD, but you either A) have to invest $ in someone that knows CAD & 3D printing or B) have to invest time/$ in learning CAD & 3D printing - again back to cost/benefit, which is way upside down. I could see it taking months to do in spare time, plus the learning curve if necessary. $20 worth of epoxy & screws would look like shit, but could probably fix it in an hour or 2.

Autodesk Fusion360 is also free for non-commercial use. I've been using it for a couple years & it's pretty easy once you get going (I started with AutoCAD LT on 3.5" floppies, so I can't really judge it from a true beginner). Autodesk is pushing it to colleges pretty aggressively, so in 5-10 years, it will still be around & it's designed to be taught to/by idiots (A shameless dig at a friend who teaches it).

Lot of cheap 3D scanners about. Still more toys than usefull. We're a long way from desktop object replicators that anyone can use.
3D printed 3D Scanner on the Cheap

quote:

Originally posted by RAMIUS:
Did you try emailing the company and asking if they have the parts?

I’m sorry, but this made me laugh