3D printing article

[Guest]

Below is a link to an article that just appeared in the Scientific American regarding 3D printing and how some new software can help in the design process. Thought it may be of use.

http://blogs.scientificamerican.com/observations/2012/09/20/3-d-printing-software-helps-designers-avoid-rookie-mistakes-video/?WT_mc_id=SA_CAT_TECH_20120925

Here's another article I just found that's just too strange/interesting to not pass along.

http://www.nbcnews.com/technology/gadgetbox/hermit-crabs-move-3-d-printed-guggenheim-shaped-shell-1B6103330

[DR]

Anybody here using 3D printing for anything useful? The only use I've had for it was a foundry used it to make the wax pattern from a CAD file for an aluminum investment casting. That was a practical, smart use of the technology.

Some uses I've seen are plain nonsense like the Jay Leno video making the rounds awhile back. He had in hand a broken part from a vehicle restoration. They scanned it to get a digital file, then 3D printed, then machined. In that case the 3D printing was totally useless except to demonstrate the process.

[Paul Alciatore]

My impression of 3D printing is mixed. You will not get +/-0.001" tolerance of finish. But you can get usable parts if they can tolerate the tolerance and finish that it will produce. Also, you do not get the full strength of the material used due to bonding limitations between layers. Again, it is not yet ready for high volume applications. It is just plain slow. On the plus side, you can set it up and then do something else so little of your time is needed. It can even be run overnight while you sleep. Another plus it many odd shapes can be made in a single operation. And it is probably more fun to watch than the washing machine.

So if you only need a few parts and they have wide tolerances and time is not a major problem, it can work for you. One trick is to make several identical small parts in a single run. You can make as many as will fit on the table. This can cut down on the total time for multiple runs.

Like any technique or shop tool, there are things it is good for and others that it is not. It's value is, in large part, due to the ingenuity of the user.

If you are interested, there was an article on this in the latest issue of Digital Machinist.

[BridgeportBill]

I posted on the other thread a few days ago. I do use a uPrint machine for the manufacturing of three parts for our generator. The stator, armature, and an impeller. The parts are then dipped in MEK and left to dry, this is believed to give a stronger part and more water tight. I am still curious as to the quality of mine compared to maker bot or other smaller machines.

[lazlo]

My impression of 3D printing is mixed. You will not get +/-0.001" tolerance of finish. But you can get usable parts if they can tolerate the tolerance and finish that it will produce.

In what material? ABS?

The huge gap between the pro machines and the Repraps et al is that the professional machines use laser-sintered metal powder, which is then heated at high temperatures to fuse into something that's reasonably structurally sound. But I don't think even the professional machines can make a valve for a car engine, for example.

It's really tough to separate the claims from reality. I saw a demo from IMTS where they had a shiny turbine impeller that they supposedly made from sintered powered stainless. But if you read the fine print, they post-processed the impeller with a combination of electro- and mechanical polishing, and it didn't seem like the part was actually usable in a turbine.

George mentioned that the shop he used to work at had a 3D printer that could make custom tooling. That seems more achievable, since you could sinter stainless and carbide, for example.

As always, Devil's in the details.

[loose nut]

Some uses I've seen are plain nonsense like the Jay Leno video making the rounds awhile back. He had in hand a broken part from a vehicle restoration. They scanned it to get a digital file, then 3D printed, then machined. In that case the 3D printing was totally useless except to demonstrate the process.

Actually, they used the process to make a copy of the missing (or damaged, I can't remember) part to use it as a pattern to cast a new part which is very useful.

[lazlo]

Actually, they used the process to make a copy of the missing (or damaged, I can't remember) part to use it as a pattern to cast a new part which is very useful.

If you watch the video again, you'll see that Stan is right: Leno is plugging the 3D printer. He prints the 3D part in ABS plastic to hold up on the show, but then mentions that the replacement was machined on a Fadal he has in the shop.

[justanengineer]

On the plus side, you can set it up and then do something else so little of your time is needed. It can even be run overnight while you sleep. Another plus it many odd shapes can be made in a single operation. And it is probably more fun to watch than the washing machine.

Personally, Id rather watch the washing machine as that gets done in under an hour. The quote above to me is a rather comical misnomer that regularly pops up. Suggesting that you dont have to monitor the machine is like suggesting you dont have to monitor a cnc. If youre running multiples of a part based on a known good setup/program you dont have to watch every minute, but youd better check on it regularly. If youre creating a single prototype, youd better watch it pretty closely and be prepared to interrupt/e-stop it. I used to run several RP machines among other duties for a former employer creating models for fluid testing (flow bench, wind and smoke tunnel, etc), mostly stereolithography in ABS, and can attest to even the professional machines being a PITA. While the programming is mostly foolproof (due to the machine handling most of it), the media itself is often a weak link in the process. I cant count the number of "bees nests" of plastic filament Ive cleaned out of these machines despite regular monitoring. Head nozzles clog up - bees nest. 1st layer of build material decides not to stick to the base/build plate - bees nest. Build material decides not to stick to support material or support material collapses - bees nest...etc etc etc - bees nest...

While it is cool to design a hollow model of an intake manifold in a few hours that can be tested and tweaked, followed by creating an actual plastic intake manifold that can be tested on vehicle...crap happens all the time in these processes, theyre not exactly easy or cheap, and no single technology is likely to ever displace conventional machine tools.

[dp]

I've been in the McNeel offices a few times (They're the Rhino Cad people here in Seattle) and they have a number of prototype displays that were created with 3D printing using their software products. I don't know what kind of printers were used or how much post-printing time is required for final fit and finish, but there is some damn cool stuff there. Maybe one of the Rhino lurkers will chime in and describe the stuff and or post images.

[Dr Stan]

Actually I've use the professional machines made by Stratasys. They are FDM (fusion deposition modelers) that lay down layers of ABS resulting in parts that have about 75% of the strength of injected molded components. The weak point(s) are located where the layers join so if one plans accordingly you can have any force applied to the strongest area of the part.

I would not dismiss FDM as not a professional level process as it was used to make intake manifolds for the first gen Viper. They certainly were not expected to last 100K miles, but on the other hand stood up well so the engineers could prove the design before committing to metal. They were also used to produce many other prototypes and Boeing is investigating the possibility of using FDM to make short run production parts for military aircraft & submarines.

If we would have had this type of equipment available when I was on both the aircraft carrier and on the repair ship I'm certain it would have saved us many many hours of machining. There were instances when we had to make repair parts from brass or aluminum not because they needed the strength, but because we did not have the means to process plastic parts.

[mc_n_g]

I built a Reprap machine and I can get around 0.015" to 0.025" of accuracy for location and dimension. Much of this comes down to machine and extrusion speeds.
I am happy to get that considering it is constructed of drill rod and threaded rod and ran me about $325 total. Five stepper motors and building the electronics was the biggest $ chunk.
If I can prototype a part without wasting metal I will gladly do it. Not everything has to be 0.001" and made of metal.
I don't know why there is such a downplay of this. It is not meant to be a replacement for machined metal parts.
Maybe Leno is overselling a process that has been out there for a number of years but it is here to stay.
In some cases it is a better alternative to get something done cheap or see if a design change works. Why is that bad?
Bring on the flames....
Signed, a Blasphemer

[dp]

Maybe Leno is overselling a process that has been out there for a number of years but it is here to stay.
In some cases it is a better alternative to get something done cheap or see if a design change works. Why is that bad?
Bring on the flames....
Signed, a Blasphemer

I think Leno failed to include a step in what he was describing. They scanned a broken part and recreated it in plastic. They then used the Fadal to machine the plastic to fit using generated code from the scan. Very likely some code fiddling happened as it always does and once happy with the part they then used the Fadal to machine an iron part. The process is an example of data entry using a scanner rather than a human to draw the part in CAD.

The reason I think this is because that is how I've seen this scanner used in other videos. The whole early chat in the Leno video is about eliminating the CAD step.

[lazlo]

I don't know why there is such a downplay of this. It is not meant to be a replacement for machined metal parts.

No flames, but 3D Printing is subject to a lot of hyperbole. It's often pitched, as noted by Leno, as "a photocopier for replacement parts."

It will be someday, but that's probably at least another 20 years away. The hurdle isn't the printer mechanism, it's finding a material that can be extruded or sintered into a structurally sound part. The sintered stainless steel powder is a start, but there are a LOT of steps involved, and no one's talking about how durable the final parts are.

[dp]

No flames, but 3D Printing is subject to a lot of hyperbole. It's often pitched, as noted by Leno, as "a photocopier for replacement parts."

If you ignore Jay and look at what is being done in the industry you'll see that printing end parts are not as important as 3D prototyping. Leno showed, rather poorly, I think, 3 different technologies that are highly intercoupled: Scanning, 3D printing, CNC machining. His example was rather gratuitous given he had the original and could have used that directly to make a casting. While he could have machined the printed prototype to show the gcode works to fit, very likely he went straight to CNC milled gray bar iron but didn't show it. I wasn't there so can only speculate.

The show is for a wider audience than a bunch of gearhead machinists and would have been a turn-off if too many details were included. I guess I'm saying that video is not what you would use to show the value of 3D scanner/CNC machined prototypes. Youtube has a load examples of that and it seems the female body is a commonly scanned part

[lazlo]

His example was rather gratuitous given he had the original and could have used that directly to make a casting.

I think that's a killer app for 3D printing: making molds for casting. Gary Martin, of Martin Models, was apparently doing exactly that at the Techshop in Seattle, before it went bust.