Apple Experimenting With 3D Printing To Create Devices

According to Bloombeg's Mark Gurman (paywalled), Apple is experimenting with a new 3D-printing manufacturing process for some device production, starting with the upcoming Apple Watch Series 9 models. MacRumors reports: The new manufacturing process that Apple is testing would use less material than the large slabs of metal that are needed for traditional CNC manufacturing, plus it would cut down on the time that it takes to make new devices. With a technique called "binder jetting," Apple is able to print a device's outline at close to its actual shape using a powdered substance. A second process uses heat and pressure to squeeze the material into a substance that feels like steel, and it is then refined with milling.

Gurman's information echoes what we've already heard from Apple analyst Ming-Chi Kuo. Back in July, Kuo said that the upcoming second-generation Apple Watch Ultra will include 3D printed mechanical parts. Specifically, he claimed that Apple is "actively adopting 3D printing technology," and that some of the titanium components in the new Apple Watch Ultra would be 3D printed. Gurman claims that Apple plans to use this new 3D printing method for the chassis of the stainless steel Apple Watch Series 9 models rather than components for the Ultra but either way, it sounds like Apple is more actively testing this manufacturing method as of 2023. Gurman says that Apple plans to 3D print titanium devices in 2024. The report notes that the shift to 3D printing would also "allow Apple to improve manufacturing times and potentially cut down on costs."

they will find an way to block 3rd party parts wit

[Joe_Dragon]

they will find an way to block 3rd party parts with this

No laser scintering of metal?

[OneOfMany07]

A few years ago that was the assumed path forward here. Where you 3D print a structure instead of molding them or whatever.

Re:No laser scintering of metal?

[monkeyxpress]

A few years ago that was the assumed path forward here. Where you 3D print a structure instead of molding them or whatever.

The additive manufacturing proponents didn't factor in that costs would also fall for traditional manufacturing technologies. CNC machining is absurdly cheap now, and that is something that was not the case 20 years ago, and this has meant direct machining (like Apple does) is very economical, and other derivative processes, like mould making, are also a lot cheaper than they used to be.

Take this binder jet technology for example. One of the big things it will make (and is already making) cheaper is plastic mould tools. So in some product designs it is already competing with itself.

The big problem with 3D printing though is that it is slow, and there aren't any really clear paths to getting a step change in that. This means that for something where you're making millions of units, it's hard to see it competing with other processes which have much faster cycle times at the expense of higher tooling costs. But it may offer big advantages in allowing Apple to design parts that they couldn't make using die casting or machining techniques (deep internal undercuts for example, or large profiled surfaces), so I imagine that is why they're are giving it attention.

Re:No laser scintering of metal?

[tlhIngan]

First, the general rules behind CNC and FDM 3D printing are basically identical - the same G-code controls both in the end. It's just in one, you're programming paths for the cutter head to remove, while the other you're programming paths (slices) for the additive head to lay down material. Other than the actual machine head, the base mechanical unit is identical. The paths themselves are just pure software routines.

As for injection molding, that isn't going away - once you get the design laid down, injection molding is the fastest way to produce something - you squirt the material in, a few seconds later a part (or more) pops out. Sure it requires a few finishing steps, but that's something a CNC machine can rip through fairly quickly. It's just startup costs are fairly high. But once that's taken care of each replicated piece is trivially cheap to manufacture.

Now, Apple has been using traditional CNC for their products - they start with a billet of aluminum and then use a CNC machine to carve out the entire body of the device. This is what makes their products stronger - but there is a lot of waste (though it's aluminum, so highly recyclable).

Of course, if they could 3D print the cases out at a reasonable speed as they could carve them on a CNC, it would allow for the ability to make structures you can't make with a traditional CNC.

Re:

[avandesande]

They aren't the same, 3d printing requires requires building the entire volume while machining is just he surface. It will always be slower.

Re:

[guruevi]

Not entirely true, in one you are adding material, which with supporting structures you can stack just about anything, like miniature Lego blocks. With the other you are removing materials, which your tool, arm, position etc all need to get to the area you are removing things from. This is one of the many problems with CNC is that although there are systems now that have n-axis repositioning of the workpiece, you still have to get a drill bit or something in the area, which in many cases is blocked by a str

Re:

[omnichad]

It's slow, but when your product is essentially a ring with a giant hole in it, CNC machining is going to be mostly waste. You then waste a ton of energy re-melting the scrap to go through the process again.

Slow may not matter if you can parallelize it better.

Cha-Ching

[Anonymous Coward]

Just what I've always wanted. Print-quality parts with Apple prices attached!

Re:

[The-Ixian]

Exactly what I was thinking when I read "cut down on costs [for Apple, not for you]"

Right to repair?

[ghoul]

This is why Apple seems to have finally dropped lobbying efforts to stop Right to Repair laws. They are just going to print the whole damn thing as one unit. You cant repair it if you cant even open it as its just one unit. Any issue just throw and get a new one.

Re: Right to repair?

[ArmoredDragon]

3d printing inevitability involves assembly when you're combining significantly different materials.

Re:

[Anonymous Coward]

Yes, but don't assume that assembly has to come after printing. You can also have assembly in the middle of the 3D printing process where the prints pause at particular layers, parts are inserted into cavities, and the print continues with additional layers - therefore encapsulating whatever it was that you just inserted.

Re: Right to repair?

[ArmoredDragon]

Metal in particular means starting with powder, melting it with a laser of at least a kilowatt or more of power in the shape you want at each layer. You can't just drop a PCB in the middle of that, it won't work. I know this because we've got a lot of these printers where I work, and trust me, there's no "print head". It's not an inkjet and it doesn't use cartridges.

Re:

[omnichad]

For now. You just have multiple printheads in the same unit or push through an assembly line of machines. 3D printing of PCBs will likely make plenty of sense eventually because they are already made in several layers as it is.

Re:

[guruevi]

It would be trivial though to cut and do a 3D scan of the piece and reprint the enclosure after you've done the repair. This lack of imagination makes the whole "right to repair movement" basically seem like a bunch of lazy bastards that want the government to force companies to do their bidding.

Don't like it, don't buy it, and if you are a true hobbyist, then don't bitch that something is hard. If Apple can do it, so can you, there is nothing physically preventing you from making chips, making enclosures,

3D Printing Does Not Scale

[necro81]

I work in a field that relies heavily on manufacturing. I use multiple forms of 3D printing (FDM like your home printer, SLS of plastic and metal powders, STL of vats of good, Objet printing of UV-cure liquids, HP's Multi-Jet Fusion tech) as a regular part of that work. It is so very helpful for prototyping and limited production runs. I know it's being used for really sophisticated aerospace and rocket components and assemblies. But that's largely on the scale of hundreds to thousands of units per year.

3D-printing is *not* a technology that scales well to the 10^5-10^7-unit scale that Apple needs for consumer products. For that, molding will certainly win out. If Apple is concerned about all that CNC milling (also a technique that is hard to scale efficiently), then I'd suggest they look into Metal Injection Molding [wikipedia.org] first, which will produce a near net-shape part much more efficiently than printing.

Re:

[monkeyxpress]

I agree with all this, but you could have said the same thing about the MacBook air when it first came out. If the designers were dead set on the unibody design, an experienced manufacturing engineer would have recommended diecasting the blanks and then post machined the details that you couldn't cast. You'd basically end up with the same thing but with way less machining required, potentially less surface finishing, and remove the need to recycle huge quantities of swarf.

But Apple made the machining techni

Re:

[necro81]

diecasting the blanks and then post machined the details that you couldn't cast

Be careful there: diecasting materials don't have the same properties as billet materials. I'll take a stab and say the Air's body was machined from 6061 aluminum or very similar (maybe 7075). It's got good properties without being terribly expensive, it's widely available, and it machines well with decent accuracy. You can't really diecast 6061 aluminum and get a part with the same properties. Diecasting alloys (e.g., A380