Looking for recommendations : Large-ish format, commercial printing

We are working on a project that is something new for us, and are interested in any ideas.

The basic footprint of the product we are developing is custom cast, decorative brass pieces. Our plan is to create digital 3D models, print them out, use those prints to create flexible molds, use those flexible molds to create wax masters, then investment cast from those.

These pieces are round with a convex shape, roughly 8" in diameter, and roughly 1" tall. You can think of them as something like 2.5-D, since the front side is all that matters.

Our primary requirement is very high surface quality, and very high detail retention. These are essentially bas-relief art pieces with a high level of detail, but there will also be large convex areas that are not part of the decoration that need to be smooth, as they will be polished to a mirror finish after casting.

We are in the proof of concept stage, but do have an eye toward eventual production. We are targeting 100 units per month, so we will need to consider throughput, cost of consumables, etc.

Budget for the proof of concept machine is $6k to $8k.

Any ideas/thoughts/recommendations?

Thanks in advance!

With the one inch reference you've provided, I would suggest that you consider to use a CNC router of sufficient size. Your requirements of a smooth finish would mean substantial post-processing (sanding) of the 3D printed part, while many CNC routers have the ability to perform very smooth surfacing action as the final pass of the cutting bit.

Additionally, you may find it useful to perform the cutting process with a CNC router to create your wax master directly.

Your budget would allow for a suitable sized CNC router with quite a bit left over. You can get an OpenBuilds OX kit that will do 12" x 18" for less than a quarter of your proposed budget. Larger models are more expensive and more capable, of course. Fully assembled CNC routers will be more expensive, of course, but you miss the fun of assembly, depending on your point of view.

Appreciate the suggestion, Fred, but I don't think that will work. Our designs will have lots of "under cuts" that a CNC couldn't reach.

Best regards.

That's certainly a valid consideration. Undercuts are usually avoided in molding. Do you have a plan of attack to deal with that aspect of your project?

Back to the 3D printing part. If you do not require engineering level precision, a model printed in ABS plastic can be acetone-smoothed to give a very very smooth surface. Using 100 micron layers (0.100 mm) will reduce the irregularities on the surface as well. If you take a set of contours over 8 inches that have 200 or 300 micron layers, it will look like topographic maps but in real dimensions, not just visually. Of course, a 100 micron model will take three times as long, but I don't think that's a factor if surface is your focus.

Another aspect to consider in your search for a printer, beyond ABS printing, will be a heated bed (required for ABS) and a complete enclosure to prevent warping. ABS really likes to warp over large flattish models. Heated temperature controlled enclosures reduce (and maybe eliminate) that warping. If you use a glass bed, I'd recommend Wolfbite for ABS as it grips like crazy, yet releases at a touch when the bed cools.

Just to make things more interesting, there is also 3D printable casting wax. You can avoid the ABS warping problem, and maybe have easier post-processing with wax 3D printing. A heat gun, a soldering iron and you're good to go. Minimal requirements with respect to temperatures and warping for that material. I've attempted it, but ran into unrelated problems that had nothing to do with the material and have since forgotten the project.

So many options for your objective and your budget is quite generous.

If you're really going for high levels of detail, I doubt that you'll be happy with the results of FDM printing. By its nature, it will have fine grooves over the whole surface where the individual layers of melted plastic adhere to one another. I'd suggest holding off on your purchase for now, and send some samples out to be printed on various different types of machine. Go through your proposed workflow and see how well each of them works for you. If one printer does produce results superior to all the rest, you'll know which one to buy. Include a resin printer in your trials, like the Form2. These, while they have their problems, will produce better surface detail without the pronounced lamination lines, and don't need the acetone smoothing referred to above, which will melt away fine details.

The suggestion of CNC routing was a good one; the parts you're talking about seem like things that would be well-suited to that process, which can produce very clean surfaces. (This sort of part is not what's usually referred to as 2.5D, by the way - that's something that's a series of flat planes at different heights. I think the term you're looking for is "reliefs".)

You objected to the router idea because of undercuts, but realize that in most 3D printing processes those undercuts won't come out cleanly because they'll be filled with support structures that will need to be removed by hand and cleaned up. If there's not too many of these undercut areas, it would probably be just as easy to carve them in by hand after doing most of the work with a 3-axis router. If not, use a 5-axis router, which can carve them cleanly. Or use a SLS printer that doesn't require support structures. Neither of these machines are going to fall within your budget, though.

@awerby, I was hesitant to suggest SLA or SLS printing for something that could be eight inches square, due to the inordinate cost.

I like your idea of sending test prints out to model building services, though.

I suspect decorative brass isn't going to have too many fine details that would be lost in acetone smoothing, although acetone smoothing does indeed remove the tiny stuff. I printed a Yeti-clone mug handle with PLA, then put elbow grease into the equation, using micromesh all the way to a mirror finish (on some parts) but that's possible because it was mostly a cylinder. Any surface changes at all would be impossible to sand so smoothly.

I'm hopeful the OP will continue to post his progress and especially any results that are put into place.

You're going a very long way round.

You can simply either print with pla and use that for investment casting directy, or print in wax specifically designed for the purpose.

Clean that u and smooth it out and make the mould from that.

Otherwise you're going through three stages before you get the final metal oject. And at each stage you will lose detail.

Printed at a high resolution that would give you a much cleaner final model.

Also if you printed with this stuff: http://www.polymaker.com/home/polycast-info/
You can smooth it either with their gadget. Or smooth it yourself using (I think, isopropyl alcohol)

serious budget !
In which case there are a few more options.
worth looking at the form 2 - I know they make a castable resin.
not a massive print volume - around 6x6x7 inches
so probably not large enough.

Have a look at this one: http://www.solidator.com/news.html
As far as I can tell it's in your price range and also fits the specs and has castable resin. C
an only find one reviw that mentions cost, says $5'000-10'000
I really HATE companies that refuse to say what their kit costs and will only give you a quote. You damn well nobody gets the same quote.
Wouldn't normally recommend people like that - but there are very few sla machines around with that kind of build volume.

For what you want - I don't think fdm will suit.
The solidator looks like the best bet from what i can see around at the moment.
http://www.solidator.com/3D-Printer.html

Quote:

---

Originally Posted by curious aardvark

You're going a very long way round.

---

Funny you should say that, because we are starting to come to the same conclusion.

The idea was that the printing itself would be the "long pole in the tent", and in the cases where we need to make multiple copies of a given piece, or spoilage downstream forces a re-work, we wanted to be able to re-make it without having to re-print.

The more we learn, the faster that assumption is going away.

Thanks to all for the feedback. We'll keep this updated as progress continues.