Printing metal

[Don Y]

Anyone done this? I think it might be less hassle than having
small quantities (~60 prototypes -- too few to justify a casting)
machined.

[Bill Sloman]

Yes.

https://en.wikipedia.org/wiki/3D_printing_processes

Selective laser melting works with titanium alloys, chrome-cobalt alloys, stainless steel, and aluminium.

Electron beam melting works too.

It's probably going to be horribly expensive, but the techniques have been around for years now, and you should be able to find a potential subcontractor close enough to visit.

--
Bill Sloman, Sydney

[Tom Gardner]

3d printing is in the stage that micros were in the 70s:
wild evolution with lots of new possibilities being
explored and niches being filled.

There are several ways of printing metal. Bill mentioned SLS,
and that is used for specialist components (aerospace,
medical, etc).

One method available to amateurs is lost wax printing. I've
used Shapeways to print brass. They will also print gold,
but that is somewhat expensive.
https://www.shapeways.com/materials

[Clifford Heath]

I've seem info on an amateur producing very high-quality intricate
aluminium castings using a 3D printed plastic pattern as "lost wax".

It doesn't need to be wax. To be able to 3D print a pattern then use
normal casting processes is very powerful.

CH

[Don Y]

On 6/24/2021 12:37 AM, Tom Gardner wrote:
> On 23/06/21 22:44, Don Y wrote:
>> Anyone done this? I think it might be less hassle than having
>> small quantities (~60 prototypes -- too few to justify a casting)
>> machined.
>
> 3d printing is in the stage that micros were in the 70s:
> wild evolution with lots of new possibilities being
> explored and niches being filled.
>
> There are several ways of printing metal. Bill mentioned SLS,
> and that is used for specialist components (aerospace,
> medical, etc).

I was more interested in folks who've actually *done* it (or "had
it done") to get a feel for the issues that arose in the process.

[I hacked together some plastic prototypes of a different assembly
to test for fit and function -- willingly ignoring appearance. But,
found that they were inferior mechanically to injection molded
parts -- they sheared along the laminations when "snap-fitted"
together. I'm hoping a service bureau will produce more robust
parts but may need to change the orientation of the part to
get the laminations to "cooperate"]

E.g., printing "lack of detail" is often just wasted money (and
time in the printer). So, if the "detail" could be assembled ONTO
something "more traditionally manufactured", there's some low
hanging fruit (economies) to be gained, there.

Likewise, post-processing/cleanup is always an issue. So,
the effort saved printing has to also include the costs of making
a *usable* part (not just an outline of one).

By contrast, *machining* prototypes already has that sort
of cleanup baked into the base part.

[And, a casting beats the printed part on cost but also bears
the "post-casting" work]

Of course, each "production technology" has consequences to
the design so getting a handle on that early on can save
headaches, later!

> One method available to amateurs is lost wax printing. I've
> used Shapeways to print brass. They will also print gold,
> but that is somewhat expensive.
> https://www.shapeways.com/materials

Yes, that's how much dental work is/was produced (e.g., crowns).

[Tom Gardner]

On 24/06/21 09:23, Don Y wrote:
> On 6/24/2021 12:37 AM, Tom Gardner wrote:
>> On 23/06/21 22:44, Don Y wrote:
>>> Anyone done this?  I think it might be less hassle than having
>>> small quantities (~60 prototypes -- too few to justify a casting)
>>> machined.
>>
>> 3d printing is in the stage that micros were in the 70s:
>> wild evolution with lots of new possibilities being
>> explored and niches being filled.
>>
>> There are several ways of printing metal. Bill mentioned SLS,
>> and that is used for specialist components (aerospace,
>> medical, etc).
>
> I was more interested in folks who've actually *done* it (or "had
> it done") to get a feel for the issues that arose in the process.
>
> [I hacked together some plastic prototypes of a different assembly
> to test for fit and function -- willingly ignoring appearance.  But,
> found that they were inferior mechanically to injection molded
> parts -- they sheared along the laminations when "snap-fitted"
> together.  I'm hoping a service bureau will produce more robust
> parts but may need to change the orientation of the part to
> get the laminations to "cooperate"]

Shouldn't be a problem with nylon or SLA. Very definitely
a problem with domestic PLA/ABS printers.

> E.g., printing "lack of detail" is often just wasted money (and
> time in the printer).  So, if the "detail" could be assembled ONTO
> something "more traditionally manufactured", there's some low
> hanging fruit (economies) to be gained, there.
>
> Likewise, post-processing/cleanup is always an issue.  So,
> the effort saved printing has to also include the costs of making
> a *usable* part (not just an outline of one).
>
> By contrast, *machining* prototypes already has that sort
> of cleanup baked into the base part.
>
> [And, a casting beats the printed part on cost but also bears
> the "post-casting" work]
>
> Of course, each "production technology" has consequences to
> the design so getting a handle on that early on can save
> headaches, later!
>
>> One method available to amateurs is lost wax printing. I've
>> used Shapeways to print brass. They will also print gold,
>> but that is somewhat expensive.
>> https://www.shapeways.com/materials
>
> Yes, that's how much dental work is/was produced (e.g., crowns).

The Shapeways pages have a good description of what's
achievable in their processes.

Finish detail and precision are good with wax/metal and
SLA. Nylon is better than PLA/ABS

[Clive Arthur]

This was done for a project I did the electronics for over 15 years ago.
We had a machine shop, but it was deemed cheaper to get these made
externally by laser sintering.

They were small helical impellers for a multi position flowmeter and
were about 2" long, 0.5" diameter in Titanium. Very good quality but I
don't remember who or where. I guess it's more common nowadays.

I think the product ended up using plastic impellers. Still not cheap as
these are machined from something high temperature, stable and resistant
to chemical attack. Don't know what.

--
Cheers
Clive

[Don Y]

On 6/24/2021 3:06 AM, Tom Gardner wrote:
> On 24/06/21 09:23, Don Y wrote:
>> On 6/24/2021 12:37 AM, Tom Gardner wrote:
>>> On 23/06/21 22:44, Don Y wrote:
>>>> Anyone done this? I think it might be less hassle than having
>>>> small quantities (~60 prototypes -- too few to justify a casting)
>>>> machined.
>>>
>>> 3d printing is in the stage that micros were in the 70s:
>>> wild evolution with lots of new possibilities being
>>> explored and niches being filled.
>>>
>>> There are several ways of printing metal. Bill mentioned SLS,
>>> and that is used for specialist components (aerospace,
>>> medical, etc).
>>
>> I was more interested in folks who've actually *done* it (or "had
>> it done") to get a feel for the issues that arose in the process.
>>
>> [I hacked together some plastic prototypes of a different assembly
>> to test for fit and function -- willingly ignoring appearance. But,
>> found that they were inferior mechanically to injection molded
>> parts -- they sheared along the laminations when "snap-fitted"
>> together. I'm hoping a service bureau will produce more robust
>> parts but may need to change the orientation of the part to
>> get the laminations to "cooperate"]
>
> Shouldn't be a problem with nylon or SLA. Very definitely
> a problem with domestic PLA/ABS printers.

I hadn't planned on actually trying to USE the few parts I
made. Rather, got curious to see what they would look like
"mated"... then discovered the material wasn't suitable
for those types of forces over very short distances.

<shrug> That's why you make prototypes!

>>> One method available to amateurs is lost wax printing. I've
>>> used Shapeways to print brass. They will also print gold,
>>> but that is somewhat expensive.
>>> https://www.shapeways.com/materials
>>
>> Yes, that's how much dental work is/was produced (e.g., crowns).
>
> The Shapeways pages have a good description of what's
> achievable in their processes.

I've "reviewed" several service bureaus so know what they
*claim* -- and, some of their sample pieces look very nice!
But, I haven't settled on a design until I'm reasonably
confident about how I will make the pieces (short- and
long-term)

[I'm not an ME so a lot of this is hit-and-miss on top of
common sense]

> Finish detail and precision are good with wax/metal and
> SLA. Nylon is better than PLA/ABS

Some surfaces I'll need to be able to press fit bearings,
etc. So, dimension as well as low Ra/z/q are significant
design constraints. (hence the potential appeal of
post-processing)

Time to check on the ice cream...

[Don Y]

On 6/24/2021 3:42 AM, Clive Arthur wrote:
> On 23/06/2021 22:44, Don Y wrote:
>> Anyone done this? I think it might be less hassle than having
>> small quantities (~60 prototypes -- too few to justify a casting)
>> machined.
>
> This was done for a project I did the electronics for over 15 years ago. We
> had a machine shop, but it was deemed cheaper to get these made externally by
> laser sintering.

Hmmm... interesting. Most places that I've worked (or with which I've worked)
that have had any sort of "additional facilities" (shake-n-bake, wave,
pick-n-place, PCB fab, machine shop, etc.) have pushed to use THOSE facilities,
wherever possible. Odd that your firm would have taken the opposite approach
(presumably, you had the skillsets in-house as well as the kit?).

But, I wasn't always aware of burdened rates for these sorts of facilities
so couldn't attest to their ACTUAL "economy".

> They were small helical impellers for a multi position flowmeter and were about
> 2" long, 0.5" diameter in Titanium. Very good quality but I don't remember who
> or where. I guess it's more common nowadays.

So, they were machinable parts (else doing it yourself wouldn't have been an
option). Did anything "special" have to be done to accomodate the outsourcing?
Did you just turn over CAD files and a purchase order?

> I think the product ended up using plastic impellers. Still not cheap as these
> are machined from something high temperature, stable and resistant to chemical
> attack. Don't know what.

Yeah, I got a tour of a ball bearing factory when I was a kid. At one point,
a display of products they'd made "locally" -- of course, the exotic stuff
is all anyone really has an interest in). I asked the CEO about a *glass*
bearing -- which seemed odd (until you realize most bearings don't carry any
static load).

He sent an underling off to find the engineer who'd designed the bearing
to explain it's purpose and design criteria to me. IIRC, it was used *in*
a corrosive stream for the shaft of a large pump (the balls were an inch in
diameter with an outer of ~6 inches)

Amusing to see what you can do when you feel like spending money!

(also saw some stainless steel bearings with ~36" inners -- in a pile
marked "scrap"... THAT had to be painful!)

[Piotr Wyderski]

Doable, but wouldn't an off-the-shelf CNC service be cheaper?

Best regards, Piotr

[Clive Arthur]

On 24/06/2021 13:29, Don Y wrote:
> On 6/24/2021 3:42 AM, Clive Arthur wrote:
>> On 23/06/2021 22:44, Don Y wrote:
>>> Anyone done this?  I think it might be less hassle than having
>>> small quantities (~60 prototypes -- too few to justify a casting)
>>> machined.
>>
>> This was done for a project I did the electronics for over 15 years
>> ago.  We had a machine shop, but it was deemed cheaper to get these
>> made externally by laser sintering.
>
> Hmmm... interesting.  Most places that I've worked (or with which I've
> worked)
> that have had any sort of "additional facilities" (shake-n-bake, wave,
> pick-n-place, PCB fab, machine shop, etc.) have pushed to use THOSE
> facilities,
> wherever possible.  Odd that your firm would have taken the opposite
> approach
> (presumably, you had the skillsets in-house as well as the kit?).

It may just have been workload, they were always busy.

>
> But, I wasn't always aware of burdened rates for these sorts of facilities
> so couldn't attest to their ACTUAL "economy".
>
>> They were small helical impellers for a multi position flowmeter and
>> were about 2" long, 0.5" diameter in Titanium.  Very good quality but
>> I don't remember who or where.  I guess it's more common nowadays.
>
> So, they were machinable parts (else doing it yourself wouldn't have
> been an
> option).  Did anything "special" have to be done to accomodate the
> outsourcing?
> Did you just turn over CAD files and a purchase order?

It was the mechanical engineer who did that, but I think so. [There were
about ten each of mechanical and electronics engineers. Easy to tell
who was which - they all wore shirts, ties and polished shoes, we all
wore jeans or shorts, t-shirts and trainers or sandals.]

--
Cheers
Clive

[Spehro Pefhany]

Last time I priced it, it was EXTREMELY expensive per cubic cm, and
the parts went though a sintering and another process to fill the
pores with a second metal.

I used the same outfit to CNC machine some cup-shaped parts out of
aluminum and it was relatively reasonable.

The parts had a spherical shape on the bottom and flattened areas on
the top to fit at the bottom of a 4K dewar (plus slots and threaded
holes and that sort of thing). So not the sort of thing Joe Blow
machinist could crank out on a manual lathe very easily.

--
Best regards,
Spehro Pefhany

[Don Y]

Not all geometries are machinable. So, I have to have some idea
as to how I want the final parts made before I commit to
their design!

[Don Y]

I don't want to have to "hold any hands" as that increases the cost
in far less desirable ways!

> [There were about
> ten each of mechanical and electronics engineers. Easy to tell who was which -
> they all wore shirts, ties and polished shoes, we all wore jeans or shorts,
> t-shirts and trainers or sandals.]

Yeah, each of my employers would gently -- but firmly -- remind me to wear
a tie. I never did (unless I was representing the company to guests
or clients). "You didn't specify a dress code as a condition of employment..."

One, otherwise attractive, job offer I turned down and, when the HR person
asked "why", I flatly told her "everyone in your office was wearing a
tie..." and let her figure out the rest! :> Given the salary offered,
I suspect she found that odd!

My normal attire, for the past 30+ years, has been jeans and T-shirt.
The past 6 or 7 it has been jeans and THE SAME T-shirt (I have 7 identical
T-shirts from which I simply select the topmost). Every 2 years I
replace the 7 with 7 more of an alternate "style" -- then switch back
to the previous style two years after *that*.

It makes "deciding what to wear" a complete no-brainer! (socks are all
the same color, etc.) :>

[The two different T-shirt styles were gifted to me by a friend (22 of one
style and 18 of another) as a joke -- because she complained that I
ALWAYS wore a black T-shirt (both of these styles are black, but with
something printed on the front -- her idea of "spicing up" my appearance).]

I just can't see wasting time trying to decide WHAT to wear (or, hunting
for a "matching sock"!)

[Anthony William Sloman]

On Thursday, June 24, 2021 at 7:44:43 AM UTC+10, Don Y wrote:

You might be living in the wrong country - Boeing seems to have had to buy it's printed metal parts from Australia.

https://indaily.com.au/news/business/2021/07/01/3d-metal-printer-takes-off-with-boeing-deal/

Apparently Boeing is getting mandrels printed in invar-36.

--
Bill Sloman, Sydney

[DecadentLinux...@decadence.org]

Don Y <blocked...@foo.invalid> wrote in news:sb0a06$h3q$1@dont-
email.me:

> Anyone done this? I think it might be less hassle than having
> small quantities (~60 prototypes -- too few to justify a casting)
> machined.
>

Only costs about $12k to find out.

https://markforged.com/3d-printers/metal-x

[Robert Baer]

How about metal vapor deposition for films as thick or thin as you want

[Anthony William Sloman]

> How about metal vapor deposition for films as thick or thin as you want.

You'd better not want any mechanical strength - a thick vapour-deposited metal film is one you can't see through - and the metal goes down as crystal islands.

Once you've got a more or less continuous conductive film, you can build it up faster by electroplating. Electroforming is a known technique, but you may have to anneal what you get fairly close to melting temperature to get much mechanical strength. Fast-pulse laser heating can let you do this on selected areas if the laser can get at them.

--
Bill Sloman, Sydney

[DecadentLinux...@decadence.org]

Robert Baer <rober...@localnet.com> wrote in
news:16a5329db549bbf9$1$107610$6add...@news.thecubenet.com:

$10k and you are in!

I have a laser engraver, and a 3D printer. I want to work my way
up to a metal set-up so I can do proto stuff for folks. It even does
Inconel!

<https://www.youtube.com/results?
search_query=markforged+metal+printer+demo>

[DecadentLinux...@decadence.org]

Anthony William Sloman <bill....@ieee.org> wrote in
news:8991b087-8371-4ac6...@googlegroups.com:

3D metal printing is happening. I'm a gonna get one soon.
The finished parts are quite hard and strong.

GE 3D prints Turbine blades.

<https://www.ge.com/news/reports/future-manufacturing-take-look-
inside-factory-3d-printing-jet-engine-parts>

https://www.youtube.com/results?
search_query=markforged+metal+printer+demo