Metal 3D printing at 100 times the speed and a twentieth of the cost A new machine will print metal parts at a tenth of the cost of today's manufacturing systems, potentially launching a revolution in small part production, its creators claim. Desktop Metal, based in Boston, USA, has opened up pre-orders for its Studio System which uses inkjet-like technology, rather than laser-based techniques …
shilltastic.
Perhaps sending a reviewer than understood current 3d printer technology would have been an idea ?
Statasys have been doing 'inkjet style' 3d printing for years - nothing new there.
The innovation here is purely to do with that 'ink' now supporting metal. If that works it could be an interesting advance over DMLS for sure - but if that's all it is, then they'd be no barrier to the big boys like stratasys adapting there polyjets to print it too surely ?
From my understanding (based on looking at earlier sneak peaks and some tech data) the secret sauce is not in binding metal powder together with an inkjet (others have been doing that for years) but more so in the impressively high density for this process, with a fancy microwave assisted sintering oven for, and some methods to compensate for shrinkage. Also their conocept system pairs a single printer with multiple ovens to maintain throughput, so not quite an all in one solution (things may have changed since I last came across this system). Even though they are touting it as capable of making production ready parts I would expect them to be notably less strong than machined parts made using the ame metals but still useful for many practical purposes.
Boffins discover source of female pleasure, still puzzled about how it works.
I have to concur that the technology described in the article is already available now, from a variety of vendors. I have sent files to Shapeways and received quality parts for prototype knife fittings.
Take a look at their store where human creativity runs wild.
Their metal is a stainless bronze combo, which lacks ductility, but the parts are stunningly accurate.
The parts are expensive, but about 10 times less than Stratysys and most of the other vendors, who generally offer more robust metal properties.
The strength of this process is to create small complex parts that are unique, and that are of high value to the user. Anyone can create and profit from these services, many of which have software that will tell you if they can make your part and what it would cost in various materials, and how much you make when someone orders one of your creations.
To see my idea for 20x the speed, 1/10 the cost 3d metal fabrication go here:
http://www.instructables.com/id/Grapple-Wave-Technology/
I can believe that it is possible to print metal in new ways that allow for reduced cost or higher speed.
I also know that quite good yield stress (arond 1000 MPa) can be obtained in 3d printed steel. However, I have not heard of these printed materials also having good fatigue resistance (please correct me if I am wrong). Until fatigue is solved, I fear that 3d printed parts will remain difficult to use in high performance machines like the jet-engine mentioned in the article.
Also; part warping appears to be a problem with (some) free-standing thin walled structures
Still, I would love to be able to engineed high resolution 3d printed parts that are strong, fatigue resistant and liquid/gas-tight.
This, lets see metal lab reports on samples showing the yield points and microscopic grain structure of the dendrites before we get too excited with how its going to sweep traditional machining away. How those dendrites form as the metal recrystalizes makes a massive difference to the final material properties.
Some parts do not require known qualities in these aspects and will be suited to it, but there's also the risk of people applying the "if all you have is a hammer, every problem is a nail" ethos expecting it to replace every other tool in the workshop, and there are things which rely on these material properties to not fail in service.
Another feather to your cap to have around, to complement a rapid proto shop, if it lives up to the hype.
There are things you can do with certain metals to improve the printed strength. there's a process called HIP that is used for Titanium (and a 3d printed part that undergoes HIP is flight certified by the FAA). HIP basically means chucking the part in an autoclave at high temperature and pressure to force out any cavities and harden the surface to restrict crack growth.
Unfortunately, the process doesnt work for Aluminium at this Point in time (at the necessary Temp you melt the Aluminium). But there are other processes being investigated, but nothing is certified yet. Sorry dont know about Steel as we just dont use it in Aerospace or at least in my Company...
This, lets see metal lab reports on samples showing the yield points and microscopic grain structure of the dendrites before we get too excited with how its going to sweep traditional machining away. How those dendrites form as the metal recrystalizes makes a massive difference to the final material properties.
Most metal things aren't made from metal because they need to be immensely strong or tough, it's because they need to be a bit stronger, tougher, stiffer or just nicer looking than plastic. My toilet roll holder isn't loaded to anything approaching its ultimate tensile strength.
Jet *engine* parts were not mentioned in the article - only a private jet. Private jets often have customised (suited for one-off or low volume production of parts) interiors. Whilst it is desirable for a bespoke drinks cabinet on a jet to be lightweight, no one dies if the hinge fails.
The fuselage and engines come from the established manufacturers.
'Depends on where the cart is located. Loose items in a passenger cabin during turbulence (or worse) are not good.
Embraer almost lost a KC-390 prototype this year due to stuff getting loose in the cargo bay!
https://theaviationist.com/2017/11/09/first-prototype-of-brazilian-kc-390-military-cargo-aircraft-almost-crashed-during-stall-tests-last-month/
An actual example I've seen used in the past was a cabin part that supported a tank(?) of some sort.
Similar enough in detail that I wonder if it was what they are referencing in the article. The end result was lighter, stronger and had fewer parts.
"Jet *engine* parts were not mentioned in the article"
No, but several combustor parts are being made using 3D techniques now. By all accounts they save several hundred hours of work per set and are much lighter than their predecessors (which were made up of dozens of parts) whilst being more reliable and longer lived.
GE's LEAP engines feature them and I wouldn't be at all surprised to see them built to fit into older engines, replacing hideously complex manually machined/assembled bits.
I figured powdered metal (the existing method) would be too weak for a proper gun barrel.
But _THIS_ sounds like it's a LOT more promising!
Now - I'd like to (be able to afford to) 3D print a few more items while I'm at it, like a jet engine [of my own design], a liquid-fueled rocket, and replacement car parts [as needed] for "legacy" cars.
This could become _VERY_ interesting!
Now - I'd like to (be able to afford to) 3D print a few more items while I'm at it, like a jet engine [of my own design], a liquid-fueled rocket, and replacement car parts [as needed] for "legacy" cars.This could become _VERY_ interesting!
Yup. Some parts can be bloody hard to find, hard to make, and stupidly expensive to have made for you. Though for me atm it's more engine parts (especially things like camshafts - not sure this would work very well for them).
While I fear for the traditional machine shops should the price of these things reach "household" levels, I also am drooling beyond flood-the-earth levels at the thoughts of what I could do with one of these toystools!
(Who'd you piss off to get the dv? :) Have a counterbalance from me)
--> imagine a stream of drool coming down the left side of that icon - now you know what I'm like atm!
(El Reg, for icon row #775, can we please have a "Drool!" icon?)
As an engineer, I'd be very interested to know the material properties of the printed metal parts. (i.e. the tensile strength, shear strength, compressive strength, and the typical failure strains.)
It's a lot like 3D printing of plastics; you can 3D print in ABS, but the inherent material alignments and small defects in the 3D printed part mean that it's likely to be weaker than an injection moulded ABS part.
Saying that, this should mean a lot less material wastage than using a CNC machine to make a prototype part from a solid block.
The 3-D metal printing has a potential to build objects by bonding complex shaped layers, like honey combs or hollow spirals, which would allow to create a light and strong design that is capable to withstand directional forces working on it during normal use.
It is hard to imagine that the strength of these objects would match forged steel objects like a heavy duty crankshaft.
@ naive: Dont think of 3d parts as replacements for forged/machined components. think of them as replacements for cast parts.
You can replace a machined component with 3d printed, but it's the same effect (from strength, etc) as if you replaced it with a cast part. but usually cheaper then cast and with sometimes fewer restrictions...
It can replace a machined part if it's not intended for specific extreme loads. One just needs to take the physics into consideration.
As the article notes, additive manufacturing (compared to subtractive such as a machined part) would definitely be an improvement for one-off/short-run parts so long as there aren't specific loads demanded of it.
And as others have noted, additive manufacturing can allow for structures that can't be achieved with machining, although I would think it's still not safe to create a closed-cell structure with this technique (this would trap byproducts inside them). Still, it opens options.
As an engineer, I'd be very interested to know the material properties of the printed metal parts. (i.e. the tensile strength, shear strength, compressive strength, and the typical failure strains.)
My employer is just wrapping up some long-winded evaluations of 3D printed aluminum, titanium, and stainless steel, each of specific alloys from a specific 3D printing machine. (Each alloy, printer, and process technique gets somewhat different results.)
Generally speaking, a given 3D printed alloy will fall into the same vicinity of strength, elongation, hardness, etc. as the same alloy produced by other means (billet, bar, casting, etc.) If you 3D printed Ti-6Al-4V powder, it can have about the same properties as billet Ti-6Al-4V, give or take a bit.
However, there tends to be more scatter in properties in 3D printed parts. Even after trying treatments like hot isostatic pressing and tempering/stress relief/aging/annealing, etc., you get more scatter. The major reasons for the scatter are:
1) There's always more porosity in the final part than a conventional stock metal because you're welding together powder, and
2) You're basically working with a big piece of weld metal
Prior to heat treatment, there's also some variation in strength by direction. A specimen is stronger along the direction of deposition than across layers. My experience is that there's much less anisotropy in 3D printed metals (a few tens of MPa difference) than in 3D printed plastics (factor of 2 tensile strength difference between longitudinal and transverse directions), but that's speaking to specific 3D printing techniques. You can't fix the 3D printed plastics, but heat treating, especially HIPing can help metals.
There's also surface finish to consider. None of our evaluation parts went straight from 3D printer to Instron tensile tester or Charpy impact tester. The exterior was so rough that it was basically a stack of weld beads. So it went from 3D printer to CNC machine tool to finish the test specimen and then into the test rigs. Using the raw, printed metal would've had different properties before the polishing and prettying.
Basically, treat a 3D printed metal part with the care and concern you'd give a weldment or thick casting of the same material. It's more likely to have flaws, brittleness, and reduced elongation compared to properly worked billet material, but none of that means it's unusable and uncorrectable. With a good testing program, you can get A- and B-Basis mechanical properties for 3D printed metals that your designers can depend on.
And results are improving steadily - the 3D printed metals today are nothing like the porous, brittle crap from the 1990s.
@cray74, truly excellent post.
Ive been using the porous, brittle crap from the 1990's and have actually found it of some use: its good enough for form and fit prototypes. Sometimes we can use it to build mold sets for injection molding. But the real niche is that it makes it easy to build tools and jigs that facilitate conventional production, inspection, and test. Granted the photopolymer printers do this for me a lot more, but sometimes a metal jig that can take some heat and abuse, yet simplifies some other task ("robot" handling PCB in vacuum chamber, reflow oven, etc)
Of course Im benefittting from a sunk cost from years ago. All I have to do is justify the continued tax on depreciated capital equipment. Justifying a few hundred grand on a low end printer to build production gear can be ... difficult.
The final printed part is then first placed in a "debind" fluid that breaks down the wax and most of the plastic before being placed into a furnace where the rest of the binding agent (which has a boiling point of just below the metal) is burnt off, leaving just the metal.
Presumably the binding agent's boiling point is just below the metal's melting point, not its boiling point...
Keep in mind on the 4/4+ (and their m variants) you'll need to drop in a maintenance kit every ~110,000 pages or thereabouts at roughly ~$120 USD. that'll keep it running like the tank that it is. (that breaks down to roughly a dollar per brick of 500 pages of paper)
Accepting all the caveats noted above, this could be a serious game changer in so many areas. Traditional bulk manufacturing may be cheaper for millions of units, but by removing the setup costs, the unit costs for smaller runs become bearable. And of course, in the same way that print-on-demand means that books no longer need to be 'out of print', using this technique means that parts will never be 'out of stock' or 'no longer available'. If there is a method for converting old technical drawings to drive one of these things, then you can make parts for vintage cars, aircraft, trains etc - anyone fancy getting XH558 flying again?
I wonder if it could print replacement cabinet ministers?
C'mon - there's a lot of other bits they never use too - not to mention the bits they seem to use to excess and would be far better without.
Hitech custom eunuchs. Go for it I say.
Not as satisfying as the traditional method though, progress always spoils someone's fun.
Don't think so... my CRC materials handbook lacks an entry for bovine excrement. That means that I cannot immediately create your replacement politician. First I need to do a research on the material properties of BS first... and my manager will accuse me of "wasting money on materials research bullshit" and lay me off. Damnit! What if HE is ministerial material? Thats a thought. Want him? I need a drink.
Not quite sure how the lack of an "expensive laser" for direct sintering during printing (as can be currently done and is being used by NASA for some engine parts) is going to reduce part cost by a factor of 20 when their own marketing example is only a factor of 14 ...
Equally, I'd have to be convinced that a part that potentially needed several hours cooking in a furnace would be 100x quicker to produce than a laser sintered part ...
Be interesting if it can be done but looks a bit like a rehash of 10 year old technology to me.
Not quite sure how the lack of an "expensive laser" for direct sintering during printing (as can be currently done and is being used by NASA for some engine parts) is going to reduce part cost by a factor of 20 when their own marketing example is only a factor of 14 ...
The costs savings on not needing a shark + tank for the laser are external to the company, hence the difference.
A lot of the chatter about the strength comes from the FDM printing model.
GE have successfully printed and run a small jet engine whilst NASA is using a bunch of 3d printed parts in Orion and have also successfully test fired a small rocket motor.
Given the kind of engineering heavyweights those 2 will have behind that kind of public work, I am gonna guess that this stuff will be plenty strong engulf for the bulk of engineering work, the real question comes down to the economics of it.
I can see this a big step forward for engineering SMB's by allowing the cost of prototyping to come down and even if you cant print every part of the machine you are building you could print the bulk and outsource the hand full of high stress pars that need to be stronger.
I will admit my own bias i a have one of the aforementioned FDM machines and have found it to be one of the most useful tools in my box.
I will admit my own bias i a have one of the aforementioned FDM machines and have found it to be one of the most useful tools in my box.
That either means you have one small enough to fit in a box, or your box is rather large :)
I must admit I'd love to have access to one locally. I'm presently toying with the decision if I should set up a small hobby machine shop as I'm approaching retirement as I've always wanted to do a bit more mechanical engineering, but I reckon a metal capable 3D printer will probably be a tad beyond my budget. A small CNC I'll probably manage, though.
I wonder if this will ever get to the point that there are no factories as such: a local facility prints the object, doing away with shipping cost, factory labor, etc. In essence, your Maytag replacement part is not manufactured by Maytag - they own the design, which you pay to download and make.
How do you even plan for that type of disruption? Does the cheap labor of an emerging economy become meaningless when factories and laborers aren't needed?
" Does the cheap labor of an emerging economy become meaningless when factories and laborers aren't needed?"
More or less. Once you have the technology for on-demand production it makes the most sense to produce close to the point of consumption.
On the other hand you're saving a lot of people from quite shitty jobs and perhaps giving them opportunities that will allow them to leapfrog industrial economies.
"[...] by sending electronic files – all over the world and so avoid hefty import tariffs, [...]
Tariffs are usually imposed to protect local industries and skills. If the files were considered a threat to those then they would be banned or a high tax imposed on unit production.
What it could do is undercut offshore production if their low labour costs and lack of environmental controls were no longer major factors - thus reducing imports.
There are so many factors in the use of metal that I'll wait before making any judgement. If this works it will certainly ADD to the methods at our disposal to produce parts, but I would not be so quick to claim it's going to make other approaches redundant.
I accept that their marketing department probably has to to generate this sort of enthusiasm to sell the idea, but that doesn't mean that idea is realistic. Mechanical engineering is a bit conservative for a reason: they like stuff that works, so this new approach will only gain traction if that is demonstrated in the real world outside the marketing brochures.
If it delivers in the real world, excellent. Until then, meh. :).
How long is it?
How hot is it?
How much does it shrink the object by? Is it predictable? Is it repeatable? Is it the same in all axes?
Large(ish) high power furnaces are not cheap and take a while to heat up.
Potential game changer, maybe, if it hits its speed and cost targets.
Many of the somewhat carping comments here seem to have either not finished reading the article—which does *not* pretend that every form of metal-object manufacturing will be obsolete next year—or are focussed on high-strength, high-durability use cases which are, for the moment, excluded. Yes, it probably will be a while before we're making trustworthy HP turbine blades using 3D printing, but in the meantime I think I go along with another commenter who suggested that this new facility would be a really useful extra capability to supplement all the other stuff we do currently.
I was particularly struck by the point about reducing the number of moving parts in certain assemblies. Let's think of that as "separately manufactured and then tediously linked together" parts, and it becomes easier to see why a 100-piece assembly might be functionally replicated as a 10-piece one. The article itself makes the point that a functioning hinge can be printed as an integral item, instead of making the base plates, pins, then slotting them together etc, and when you consider the mechanical complexity of some bits of machinery ... it's not difficult to see why imagination and ingenuity can let you replace an item that once required 100 different bits to be made and assembled by something requiring only 10. I'd be surprised if young engineers will not find ways to massively improve on traditional designs, producing equivalent functionality cheaper, lighter and better. Not every metal engineered part has to support a London bus.
In passing, mention of Nasa immediately made me think of Item One in the packing list for the next manned Moon or Mars-shot: a couple of top-end multi-material 3D printers, a tonne of various ingots, and the data needed to repair or remanufacture every critical part of the spaceship, suits, habitats—to be used by astronauts/colonists who've been trained to design and make new stuff as needed, too.
Plus, of course, the data and materials needed to make additional 3D printers ;-)
The error could have originated in engineering. Or at least in the drawing office.
I remember the design spec for a piece of avionics that incorporated a "backwoods power detector." It was for use in the Multi Roll Combat Aircraft (obviously capable interestingly ineffective evasive manoeuvres).
You might be able to print in an office-like environment, but that's not the whole process. You have to process the part in a sintering furnace. That's not office-friendly, nor is it fast.
My impression is that sintering cycle is effectively a day. Five minutes to print a (simple) part in a machine the size of an old copier sounds appealing. Sending it out to the long line of ovens in the basement.. not so much.
I was once on a DTI project to try and identify the most important potential manufacturing process of the future. The front runner was to develop something that made metal parts at net size.
Over 20 years later we are still only slowly getting there (like ceramics, this process involves some shrinkage at the firing stage).
"like ceramics, this process involves some shrinkage at the firing stage"
Ceramic firing shrinkage is unpredictable and not necessarily uniform in a piece.
IIRC Charles Dickens describes the rejects from the process of making Parian figures eg heads much too big/small for the body. Admittedly those figures were assembled from components that each may have had a different moisture content.
One of the comments above send my mind down a rabbithole. How feasible would it be, with metal 3D printing, to set the device up in a vacuum, and print (say) a sphere with internal bracing against collapse? Then, could you flood the chamber with air and keep a balloon full of vacuum which would float?
And then what use would it be? I guess you could add buoyancy to all sorts of things, but would *lots* of buoyancy be better than just using a lightweight gas?
I think you could put just the sintering stage in a vacuum if you decompressed slowly - I have a feeling the print is pretty porous until it's been heated up. You may have just described how it would work on a space station - like another commenter it was just about the first thing I thought of when I read this.
(that said, it's a logical conclusion for people who have grown up with science fiction :) ).
Take a weighing balance. At equal distances from the fulcrum place an inflated balloon on one side, on the other place two uninflated balloons. If the inflated balloon is heavier than the two uninflated balloons you might just be in with a chance. (Issues: the air inside the inflated balloon will be slightly denser than atmosphere because of the elasticity of the balloon material. )
I would like to know how does compare the mechanical resistance of a device made with this 3D printer with a device made from a block of metal worked with a CNC. I'm not convinced by the "fusion" process described in the article, I doubt it can bind layers together strong enough to make something mechanically similar to a "full" piece without impacting the geometrical characteristics.
Sintered gears have been in mass production for decades. They have significantly better accuracy and balance than cast parts.
They are generally made with finely powdered metal molded at room temperature using very high pressure. The high pressure allows using a metallic binder (e.g. copper), which deforms to provide near-100% density. Even so, the unidirectional nature of compaction result in directionality to the part strength, especially off-compaction-axis impact resistance.
My guess is that the 3D printed process, with no compaction, results in significantly inferior parts compared to existing pressure molding techniques.
The freedom you have in geometry can save you mass for the same structural performance, so if material tests told you that your printed part was only 98% as strong as a machined part of same geometry, you can redesign your printed part and likely still be ahead in terms of strength/weight ratio.
Any critical part - be it printed, cast, sintered or machined - will be tested before deployment anyway. And modern CAD systems allow the results of such physical testing to be stored for reference in the design of future parts.
@Evertime; good points but the devil is literally in the details and why a professional engineer won't typically answer a 'is conventional or additive manufacturing better?' question with anything but a 'it depends' answer. One can harden AM printed gears. Surface peening is an option, and is electroplating, exotic lubes, and so forth. Nearly all the materials post processing used for conventional parts are applicable to AM. Of course one can fail to design the gear well and all the king's manufacturing men cannot make a go of it. Or truly excellent mechanical design can make a go of a product even with crap materials and processes- the key being knowing just what quality you have. Or I can overthink my gears to the point where they outlast my otherwise dodgy product and I go out of business having spent too much capital on my lovely gears.
So many tradeoffs... This is why engineering is an art, really - making the best product humanly possible is fun. But making an economically competitive, safe, "good enough" product that users like is both fun and bloody difficult!
"sending electronic files – all over the world and so avoid hefty import tariffs"
Just the same as any other CNC machine could be used to make parts locally from files sent across the world. Claims like that make you wonder how much of the rest is bullshit.
Interesting.
The devil is in the details, but I'll paraphrase something from the software world - a 10x increase in speed is a feature, because it allows you to use a system in different ways and different purposes than previously, even if the final result remains the same.
A 20-fold drop in price would indeed make a huge difference in 3D printing and manufacturing. If their idea has legs and if it isn't just another bit of investor-bait, then it will advance the state of the art. Even if they're not perfect, others will follow and improve on it. If they have big, established backers that are savvy to this type of tech, might be something to it*.
That's some pretty big IFs, but still interesting to see how quickly this field is moving nowadays.
* or.... Thanatos?
I'm out of a job
No more f***ing about with stupid robots/cnc machines or the attendent personel problems caused by low quality operators.... and as a bonus, when the demented designers upstairs come up with even more stupidily designed crap , they'll have the honour of running the machine that makes it so they can explain to the board why the part fails in service instead of blaming us lot in the machine room.
Boris
<<muttering darkly about the idiot today who insisted on sharp recessed corners on a part subject to high shock and vibration loads.....
If you could print with an alloy that melts at 240C, and a thermally conductive alloy that sinters at 240C, and an insulator that sinters at 240C, you'd have 3D PCBs. Place the components in it as it's built-up then bake it. It sounds hard but it can't be worse than designing the sandwich in the iPhone X. More advanced designs could even accommodate liquid cooling ducts.
The Borg collective is calling. We must go.
Already exists.
It's called Low Temperature Cofired Ceramic.
It's been around about 30 years. the layers can be CNC punched or laser cut and are in 2,4.5,6.5,10 and 20 mil (0.001inch units) thicknesses. Various ink pastes can be used to make conductors, resistors, inductors or capacitors inside the boards by screen printing. Cavities for chips can also be made, protected by a carbon paste that burns out during the bonding process.
The tech is also good for making MEMS. It's also a hell of a lot cheaper than making custom chips.
A further note on Low Temperature Cofired Ceramic.
There are also tapes that incorporate both piezo electric precursors and ferromagnetic particles. A new technique developed (and patented) by Sandia labs allows Full Through Thickness Features (as opposed to ink paste layers of around 1/2-1 mil (12-25.4 micrometres) by CNC punching from the back supporting plastic film to the front, screen printing without a special mask) and acting as a "liftoff" lithography system, peeling off the excess with the backing tape.
PZ and FM allow you to make in situ actuators for MEMS, or (in principle) speakers and microphones, along with (in principle) memory devices, all good to very high (by Silicon) operating temps.
There are (sort of) semiconductor pastes (ZnO and TiO) but the joker is there don't seem to be any viable ways to dope them to change their conductivity. If there were you could make big (and therefore slow, c30 micrometres) transistors that could operate comfortably at 5-600c+ down a borehole.
Indeed, we have a friend in Edinburgh who has been without a working boiler so no heating, no HW since earlier this week and has to wait for Tuesday next week for a part to come, almost certainly some electronic part.
Being able to produce those cheaply and easily locally would indeed be a boon.
These modern boilers break down a lot, because of the complexity which has to be included to make them efficient. We have an ancient Baxi back boiler and it has just developed its first fault since we moved in nearly 20 years ago. Pilot light kept going out and was hard to relight. BG engineer had the necessary part in his van, a new thermocouple was required, It shuts off the gas supply to the pilot light if it goes out (to stop the house slowly filing with gas) but hours was shutting the gas off while the pilot light was on.
We will replace this ancient machine soon, after the costs of the youngest getting married in NZ next year are done. It just needs to survive one more winter. Moved up into the attic. But I'm bracing myself for more callouts and more problems.
"These modern boilers break down a lot, because of the complexity which has to be included to make them efficient."
No, they break down a lot because of shitty design and that has a lot to do with poor expectations of british consumers. These are the same cruddy units that die due to limescale buildups in the heat exchangers.
DECENT boilers don't suffer electronics breakdowns (hint: test the capacitors and make sure the designs take ESD protection into account). DECENT designs also use a secondary water loop so that fresh mains water is never exposed to flame in the primary heat exchanger and DECENT installations don't run the condensate line where it's going to be exposed to freezing temperatures in winter.
I have a nice 15 year old Bosch combi setup which has never failed - and the water around here is so limey that if you boil it in a bucket you can see flakes precipitating out. On the other hand people who installed cheap "potterton" and co crud have endless problems with the things due to rotten british quality control and poor design. TCO is about far more than just being cheap to buy.
Currently helping a mate try to find some parts for a rare old vehicle.
Something like this could be really good for one-off parts that are expensive to make and even more expensive to buy. Though no idea how it'd handle for stuff like cam shafts.
A pity about the price. I'd get one and whip up some stuff for him (and others) but.. Still, given a decent 3-d scanner expect non-OEM parts to hit the fleabay markets soon.
As to BMW... Shit product made from shit metal. After the fun and games another mate had with the mount for the drivers door handle breaking constantly (something my ancient POS has had the original of for >20 years without any hint of issues - and this is a common BMW problem!)... Well, if BMW can make terrible shit like that for something every other model has last the lifetime of the vehicle, it says a lot about the many many many many many many many other issues he had for the short time he had the cracar. BMW are the worst option for advertising your product, if they're known to use it they'll drive other prospective customers away!
[bmw hate brought to you by direct experience of their garbage]
Well the metal quality might be OK but the surface finish is the issue.
That said there are a number of "liquid polishing" techniques. One of which is called "liquid honing." It use a proprietary mix of plastic beads with embedded abrasive in to give a surface finish below 64 microinches (1.6 micrometres) and even to 0.4micrometres.
Well the metal quality might be OK but the surface finish is the issue.That said there are a number of "liquid polishing" techniques. One of which is called "liquid honing." It use a proprietary mix of plastic beads with embedded abrasive in to give a surface finish below 64 microinches (1.6 micrometres) and even to 0.4micrometres.
It's the metal quality that I'm more interested in. The polish on the lobes isn't a big issue, but whether or not the lobes or other parts will fail under load. Then again, I guess the pressures on a cam shaft aren't really that great. As I write this I'm wondering if I could make my own clay mold and get someone to fill it with an appropriate material for me. Getting the exact "as new" specs (especially the lobe shape) would be nice, but I doubt the maker is willing to let such details out.
Discussed a couple of paragraphs before the end.
"The metal fuses creating a density of between 96 and 99.8 per cent, according to Desktop Metal. "
So near full metal density.
If the standard spec metal was borderline you would probably have to switch metals. In terms of powder metallurgy I think the nearest is where the metal is compressed in a mold and has an electric current put through it.
If you are an engineer seriously interested in material properties of additive manufactured materials, I'd recommend getting in bed with NIST (link: https://www.nist.gov/document-3511) or its UK/EU equivalent.
I haven't worked with NIST on metal AM but know they've got access to just about every polymer printer under the sun and a spectacular set of materials data and at least an equivalent if less mature program on the metals side. For my polymer work the relationship has proven essential.
Absolutely have your prospective vendor print standard ASTM coupons and perform proper materials testing. If you cannot do so in house, materials test houses abound and can provide full traceability. Try different lots of powder. Try powder from the top and bottom of a cartridge. Try powder that is new and some that's been stored right at the margins of its storage envelope. For polymer printers, test parts printed at all corners of the materials' temp and humidity space. You get the idea.
The people I know who do DMLS in a production environment always include two test coupons (x-y and y-z axes) and do at least a basic yield measurement on every single one to catch any process drift. Fatigue strength, hardness, etc are occasionally tested per SPC.
The bottom line is that there is nothing inherent in AM that absolves you of the need for robust processes control. Indeed until many thousands of hours accrue in a given process one tends to keep a weather eye on outcomes.
Note that you generally still need some CNC capability to.clean up parts post-build, no matter what the salesman claims.
Mine's the one the with the 'trust but verify' button on the lapel.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
