|
|
Thread Tools | Display Modes |
#21
|
|||
|
|||
aluminum machining? A q that is technical but not bicyclic
Ryan Cousineau wrote:
I have a bit of an odd question with regard to laptop manufacture and aluminum machining. It's essentially one of curiosity, but I think a few people here may have some insight. If you don't know, Apple just introduced some revised laptop models. Great new features, bold design, whatever. The feature that made me sit bolt-upright and say "no way!" was the case structure. The old MacBook design used plastic molded cases. I think there might be a metal inner frame, too. The "Pro" model used aluminum cases, but I'm virtually certain they were stamped, and I think they relied on an inner metal frame as well (depending on the year, these frames were made out of aluminum or magnesium, I think, and presumably cast, but possibly stamped). This year's models are all sporting machined aluminum cases, and apparently have no inner frame to speak of (they call it "unibody" a la modern car designs). It appears the major components bolt more or less directly to the body. Web page here, including video of what may be actual manufacturing techniques and pictures of the bare case: http://www.apple.com/macbook/design.html It's a clever idea, and I can see benefits. But just watch that production video! How in the heck can they be doing CNC machining in a production-line fashion and be cost-competitive? I would have thought the cost and time (and manufacturing capacity?) would have been prohibitive. Anyone here have an insight as to how they're doing this without spending way more money than when they were molding cases out of plastic? I think these laptops are probably being made in the PRC, though there's a small chance this is a Taiwanese product. I mean, Apple's not crazy, so I assume this will all work, at least in theory, but when did machining costs and capacities get so cheap that they could conceive of milling out several million complex case designs every year? like hard drive cases, i expect they're cast to shape, then milled for finish. machining from a solid block makes no sense. |
Ads |
#22
|
|||
|
|||
aluminum machining? A q that is technical but not bicyclic
On Wed, 22 Oct 2008 17:00:29 -0700 (PDT), Frank Krygowski
wrote: On Oct 22, 3:44*am, Ryan Cousineau wrote: If you don't know, Apple just introduced some revised laptop models. Great new features, bold design, whatever. The feature that made me sit bolt-upright and say "no way!" was the case structure. This year's models are all sporting machined aluminum cases, and apparently have no inner frame to speak of (they call it "unibody" a la modern car designs). It appears the major components bolt more or less directly to the body. It's a clever idea, and I can see benefits. But just watch that production video! How in the heck can they be doing CNC machining in a production-line fashion and be cost-competitive? I would have thought the cost and time (and manufacturing capacity?) would have been prohibitive. Anyone here have an insight as to how they're doing this without spending way more money than when they were molding cases out of plastic? I think these laptops are probably being made in the PRC, though there's a small chance this is a Taiwanese product. I'm surprised the chassis isn't die cast. I'd expect that would be cheaper, and it's tolerances would be fine for that application. CNC is great if you're doing low production things like boutique derailleurs. It's not so economical for large production runs. And think of the volume of scrap (chips) that they're dealing with! It looks like over 50% of the raw material leaves as scrap. I could be wrong, but I've got a sneaking suspicion the production line involves something like pre-stamping or casting the aluminum into near-final shape, then machining to the final shape. Might be a "cheap" way to cover up for cheap casting or stamping... As for the video, well, you ever hear of advertising? Pat Email address works as is. |
#23
|
|||
|
|||
aluminum machining? A q that is technical but not bicyclic
In article
, Chalo wrote: Ryan Cousineau wrote: [laptops with machined-aluminum cases. This year's models are all sporting machined aluminum cases, and apparently have no inner frame to speak of (they call it "unibody" a la modern car designs). It appears the major components bolt more or less directly to the body. http://www.apple.com/macbook/design.html I don't know about that. Certainly it's a lot more expensive than making cases out of plastic, but it probably offsets thermal management features that would add cost and would otherwise be necessary. If you remember nice-looking CNC machined brake boosters from Tektro and others selling for very reasonable prices even after distributor and retail markup, you might be able to extrapolate the fundamentally similar plate construction of the Apple laptop costing a similar amount per area of plate, removed volume of material, etc. That's true. But I recall those parts being pretty expensive considering what they were, and being pretty simple machining jobs (I have a classic purple-ano brake booster on my MTB) from pretty thin plate, and where the nominal area would have been pretty close to the true area of those U-shaped plates, which was awfully small. And production volumes were almost certainly tiny compared to Apple's quantities. So from a cost perspective, it was a smaller, simpler part where the only real alternatives (before cheap carbon fibre) were heavier, cruddier castings, or forgings (which were probably uneconomic at likely production volumes). But the data point is notable. These computer cases are competing with injection-molded plastic or stamped aluminum, which are both cheap and fast in the volumes we're talking about. Make no mistake: I take it as virtually proven by the existence of these cases that the production is viable, it just shocked me that it was so, and wondered what it was I didn't know about these production processes. That said, I prototyped a series of avionics component boxes once that were machined from 9" x 9" x 2.25" blocks of aluminum. They had lots of little ribs and small machine threads down in their bottoms, and their wall thicknesses were reduced to 1/16". After I had vetted and modified the original design for manufacturability, each one of these boxes still cost on the order of $500 from an efficient production shop with more suitable machinery than I had available to do the prototyping. If they had only been 3/4" thick, though, they would have been fast/easy/cheap by comparison. These are the sort of numbers that I would have guessed at (Well, I'm not surprised at the prototyping cost, but the CNC-production cost surprises me a little. It is higher than I would have guessed, even for a complex box of that size). Okay, this laptop part is a third of the thickness, but would it have cost you less than a third the cost to make these laptop cases using the same production process? For high-quality, high volume CNC milling, I'd look to Taiwan first. I'm guessing that their equation is partly a matter of being willing to take the high road design-wise and escalate costs a little bit in the process, and the one-piece machined case allowing them to trim away other features and components that would have added up to some comparable cost figure anyway. Looking carefully at the photos and commentary in this disassembly... http://www.ifixit.com/Guide/First-Lo...acBook-Unibody http://www.ifixit.com/Guide/First-Lo...y/Page-7#s3006 ....I find your theory compelling. It looks like the part-count is WAY down, and it looks like the case is being used as a heat sink. -- Ryan Cousineau http://www.wiredcola.com/ "In other newsgroups, they killfile trolls." "In rec.bicycles.racing, we coach them." |
#24
|
|||
|
|||
aluminum machining? A q that is technical but not bicyclic
Ryan Cousineau wrote:
That said, I prototyped a series of avionics component boxes once that were machined from 9" x 9" x 2.25" blocks of aluminum. *They had lots of little ribs and small machine threads down in their bottoms, and their wall thicknesses were reduced to 1/16". *After I had vetted and modified the original design for manufacturability, each one of these boxes still cost on the order of $500 from an efficient production shop with more suitable machinery than I had available to do the prototyping. *If they had only been 3/4" thick, though, they would have been fast/easy/cheap by comparison. These are the sort of numbers that I would have guessed at (Well, I'm not surprised at the prototyping cost, but the CNC-production cost surprises me a little. It is higher than I would have guessed, even for a complex box of that size). Okay, this laptop part is a third of the thickness, but would it have cost you less than a third the cost to make these laptop cases using the same production process? Oh heavens, yes. The problem with machining those avionics boxes was twofold: They required the use of expensive long skinny (thus relatively flexible, thus vibration-prone) tools, and the remaining wall thickness was thin enough to "sing" when the tool passed across it as well as thin enough to flex away from the tool. Neither of these things would have been a noteworthy issue with a total plate thickness of only 3/4". A shorter, stiffer tool would have allowed dramatically faster removal rates and better finish, and a shallower cavity would have cleared chips better. Aluminum plates are more expensive now, no doubt. But at Apple's scale of production, most of the aluminum is probably carried right back to where it came from, to be turned into more aluminum bars for machining more computers. That's what Boeing does after machining away roughly 95% of the mass of the aluminum plates that come into their machine shops. (Jet airliners may have minimum skin thicknesses of about 1.0 mm, but those skins have been machined from plates several inches thick, with ribs and reinforcements and fastener points carved out of solid metal.) Chalo |
#25
|
|||
|
|||
aluminum machining? A q that is technical but not bicyclic
On Wed, 22 Oct 2008 21:58:37 -0500, Patrick Lamb
wrote: On Wed, 22 Oct 2008 17:00:29 -0700 (PDT), Frank Krygowski wrote: On Oct 22, 3:44*am, Ryan Cousineau wrote: If you don't know, Apple just introduced some revised laptop models. Great new features, bold design, whatever. The feature that made me sit bolt-upright and say "no way!" was the case structure. This year's models are all sporting machined aluminum cases, and apparently have no inner frame to speak of (they call it "unibody" a la modern car designs). It appears the major components bolt more or less directly to the body. It's a clever idea, and I can see benefits. But just watch that production video! How in the heck can they be doing CNC machining in a production-line fashion and be cost-competitive? I would have thought the cost and time (and manufacturing capacity?) would have been prohibitive. Anyone here have an insight as to how they're doing this without spending way more money than when they were molding cases out of plastic? I think these laptops are probably being made in the PRC, though there's a small chance this is a Taiwanese product. I'm surprised the chassis isn't die cast. I'd expect that would be cheaper, and it's tolerances would be fine for that application. CNC is great if you're doing low production things like boutique derailleurs. It's not so economical for large production runs. And think of the volume of scrap (chips) that they're dealing with! It looks like over 50% of the raw material leaves as scrap. I could be wrong, but I've got a sneaking suspicion the production line involves something like pre-stamping or casting the aluminum into near-final shape, then machining to the final shape. Might be a "cheap" way to cover up for cheap casting or stamping... As for the video, well, you ever hear of advertising? Pat Email address works as is. Dear Patrick, It seems unlikely that Apple would lie about machining the frame from a single rectangular block of solid aluminum and put together a video showing the steps in detail and making specific claims: http://www.apple.com/macbook/design.html#designvideo If anything, Apple is boasting about going to all this fuss, which is of little practical interest to a user--if no one told you that the laptop case was laboriously machined out of a solid chunk of aluminum, how would you tell without disassembling it? Somehow I doubt that the price will be substantially lower. The accompanying text is so absurd that it requires no comment: "Take the thumbscoop, for example. It’s the indentation that allows you to open the display. If the scoop is too deep, you put too much pressure on the display to open it. If it’s too shallow, you struggle to open the display. It may seem incidental, but if the thumbscoop is well designed, it makes the difference between a bad experience and a good one. The challenge of the thumbscoop was to create a crisply machined scoop that was still comfortable to use. The designers at Apple worked on hundreds of versions of the thumbscoop — even examining them under an electron microscope — to get it right." As an analogy, I'm willing to believe that someone has ceramic ball bearings in his bicycle headset, but he has to proclaim it--otherwise, it's not likely that anyone will notice that the ordinary grade 25 steel balls have been replaced. Cheers, Carl Fogel |
#27
|
|||
|
|||
aluminum machining? A q that is technical but not bicyclic
On Oct 22, 5:00*pm, Frank Krygowski wrote:
On Oct 22, 3:44*am, Ryan Cousineau wrote: I mean, Apple's not crazy, so I assume this will all work, at least in theory, but when did machining costs and capacities get so cheap that they could conceive of milling out several million complex case designs every year? I'm surprised the chassis isn't die cast. *I'd expect that would be cheaper, and it's tolerances would be fine for that application. I don't think you could cast all the detail into the chassis at the required precision, like the holes for the keys, and some of the support structures and ribs. Plus screw holes. So they would need a lot of machining after the fact. It might have been possible to cast a sort of bowl shaped blank and machine that, which would save having to machine away a bunch of metal. However, removing the large chunks of metal is probably not the rate limiting step, because they could use a relatively large tool and feed quickly. Machining all the small surfaces probably takes the most time. CNC is great if you're doing low production things like boutique derailleurs. *It's not so economical for large production runs. *And think of the volume of scrap (chips) that they're dealing with! *It looks like over 50% of the raw material leaves as scrap. As Chalo said, I imagine that the scrap is recaptured. They probably squeeze a little of the cutting oil out and send it off to be stamped into Dells. Aluminum billet isn't super cheap, but the material cost is a smallish part of the whole process. Where I bet they are saving money is on assembly. Conventional laptops have multiple layers of frame, shielding, outer body and so on that can get rather involved. Some of Apple's earlier laptops were notoriously complex in order to cram everything into the available space (and because they used to discourage user serviceability). I'm typing this on an antique iBook G3 that wishes for more disk space, but there are over 30 steps in getting to the hard drive. (Seriously. There are illustrated guides at ifixit.com. I've fixed cameras that looked easier.) Ben |
#28
|
|||
|
|||
aluminum machining? A q that is technical but not bicyclic
In article ,
wrote: On Wed, 22 Oct 2008 21:58:37 -0500, Patrick Lamb wrote: On Wed, 22 Oct 2008 17:00:29 -0700 (PDT), Frank Krygowski wrote: On Oct 22, 3:44*am, Ryan Cousineau wrote: If you don't know, Apple just introduced some revised laptop models. Great new features, bold design, whatever. The feature that made me sit bolt-upright and say "no way!" was the case structure. This year's models are all sporting machined aluminum cases, and apparently have no inner frame to speak of (they call it "unibody" a la modern car designs). It appears the major components bolt more or less directly to the body. http://www.apple.com/macbook/design.html#designvideo The accompanying text is so absurd that it requires no comment: "Take the thumbscoop, for example. It’s the indentation that allows you to open the display. If the scoop is too deep, you put too much pressure on the display to open it. If it’s too shallow, you struggle to open the display. It may seem incidental, but if the thumbscoop is well designed, it makes the difference between a bad experience and a good one. The challenge of the thumbscoop was to create a crisply machined scoop that was still comfortable to use. The designers at Apple worked on hundreds of versions of the thumbscoop — even examining them under an electron microscope — to get it right." It pains me to admit this, but I totally get what he's talking about with respect to the thumbscoop design, and (electron microscope notwithstanding) I'm quite sure they prototyped multiple scoop designs and spent a while flipping them with their thumbs. It sounds like nonsense, but I get to put my hands on a lot of computer equipment (including laptops) every year. In general, Apple's physical designs are way more thoughtful than those of any other company in this industry, and spending time fussing over the thumbscoop is a vignette that suggests how they got there. To an extent, yes, Apple wants to sell the idea of their own greatness, for marketing reasons. But holy moly is their stuff nice, both on the hardware and software side. As an analogy, I'm willing to believe that someone has ceramic ball bearings in his bicycle headset, but he has to proclaim it--otherwise, it's not likely that anyone will notice that the ordinary grade 25 steel balls have been replaced. People sometimes wonder why I work in IT support a mostly-PC environment but have a Mac as my home laptop. I enjoy my work, and it pays well, but I don't want to do my job at home. -- Ryan Cousineau http://www.wiredcola.com/ "In other newsgroups, they killfile trolls." "In rec.bicycles.racing, we coach them." |
#29
|
|||
|
|||
aluminum machining? A q that is technical but not bicyclic
On 2008-10-23, Ryan Cousineau wrote:
In article , wrote: [...] The accompanying text is so absurd that it requires no comment: "Take the thumbscoop, for example. It’s the indentation that allows you to open the display. If the scoop is too deep, you put too much pressure on the display to open it. If it’s too shallow, you struggle to open the display. It may seem incidental, but if the thumbscoop is well designed, it makes the difference between a bad experience and a good one. The challenge of the thumbscoop was to create a crisply machined scoop that was still comfortable to use. The designers at Apple worked on hundreds of versions of the thumbscoop — even examining them under an electron microscope — to get it right." It pains me to admit this, but I totally get what he's talking about with respect to the thumbscoop design, and (electron microscope notwithstanding) I'm quite sure they prototyped multiple scoop designs and spent a while flipping them with their thumbs. It sounds like nonsense, but I get to put my hands on a lot of computer equipment (including laptops) every year. In general, Apple's physical designs are way more thoughtful than those of any other company in this industry, and spending time fussing over the thumbscoop is a vignette that suggests how they got there. I can also believe they did spend a lot of time on the thumbscoop. They are very rigorous about putting all the effort and expense into the bits the user sees and feels and saving money somewhere else. They must save several millions, perhaps even billions, of dollars by helping themselves to FreeBSD-licensed software (and why not) and by only supporting their own hardware. To an extent, yes, Apple wants to sell the idea of their own greatness, for marketing reasons. But holy moly is their stuff nice, both on the hardware and software side. Most of the software they get for free. They put their effort into the GUI and the bits that make it nice to use. It's a good strategy, especially for a consumer product. |
#30
|
|||
|
|||
aluminum machining? A q that is technical but not bicyclic
On Wed, 22 Oct 2008 19:34:28 -0700, jim beam wrote:
Tom Ace wrote: On Oct 22, 12:44 am, Ryan Cousineau wrote: I mean, Apple's not crazy, so I assume this will all work, at least in theory, but when did machining costs and capacities get so cheap that they could conceive of milling out several million complex case designs every year? Hard disk drives have had intricately machined aluminum cases for years. Tom Ace they're /finished/ with machining operations, but the bulk of the shaping is done by casting. finish is much cheaper/simpler than forming. Wrong again. Apple says they start with a solid block. |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Machining a Square-Taper Crank | pkittle | Unicycling | 1 | February 18th 08 05:28 AM |
Machining a Square-Taper Crank | pkittle | Unicycling | 10 | February 16th 08 07:06 PM |
bike component tolerance & other machining info? | Nate Knutson | Techniques | 8 | October 16th 06 09:42 PM |
Machining an integrated HS to accept threadless | D'ohBoy | Techniques | 2 | August 4th 06 06:29 AM |
Machining axles/sleeves and some questions. | total uni | Unicycling | 19 | January 5th 05 04:45 AM |