3D Printing Of Commercial Project Enclosures
idbruce
Posts: 6,197
I am just curious if anyone is 3D printing enclosures for commercial projects and how it is working out, as compared to either making molds and going into full production or subcontracting out the enclosures.
Comments
It depends on the quantity. 3D printing is really slow, so if you need many you are probably better off getting a mold.
3D printing enclosures is better suited to getting the design right, or for low quantities.
Bean
Thanks for responding.
Do you abs or pla for the prototypes? And how well does the plastic hold up to adding and removing screws?
ABS lasts longer, but it's a bit foamy inside and probably only half as strong as injection molded. And the finish quality is much lower.
Per Bean, it works best in custom, prototype and very low volume applications.
One cheap path on molds is to use aluminum. Simple designs won't take much to prepare and mill. Tool life is short to moderate, depending on injection material and pressure, etc...
But, machining a replacement remains cheap.
Often, a tool and a lot of parts can be had for less than a great tool. You need to understand your product life and changes to understand if a soft tool can make sense.
Aluminum can be done on modest mills too, which can make for other cost savings, like making the tool in a hacker space or something.
I know this approach is being used by several companies, some large, and it can be very cost effective on even hundreds of thousands of parts.
-Phil
Your response intrigues me. Care to elaborate on what your thoughts are?
@potatohead
Although injection molded plastic would be ideal for what I have in mind, that would be way out of league, because I am still struggling to get back on feet from my last endeavor. And I am not exactly sure about the end of your response, whether that was pertaining to injection molds or milling the entire enclosure from aluminum. I am sure that milling enclosures would be a nice alternative to injection molding, but once again, currently out of my league.
@Phil
Of course that would be the ideal solution for me, but I need something cosmetically appealing. For instance, I need something that a person would not be ashamed of leaving on the kitchen table or a woman would not have a problem tossing it into her purse, or a man storing it in the breast pocket of a suit.
///////////////////////////////////////////////////////////////////////////
Although not the most ideal or economical solution.............. And it might work, but it may not.
I don't know if you guys remember or even read the thread on the new PCB saw, but in that thread, I experimented with casting JB Weld. I was really quite impressed with the workability and machinability of the cast. I am fairly certain that a casted enclosure would be durable enough for most end users, however I am uncertain how well it would stand up in high heat conditions as Erco pointed out with the car example, plus it would require paint, and I am uncertain how well paint will stick to JB Weld.
Perhaps use a 3D printed mold as Beau suggested, and fill the mold with thinned JB Weld. One of the main problems would be that the mold would have to be resistant to laquer thinner and acetone, because these are the suggested thinning agents for JB Weld.
Great call, Beau. Mattel's first 3D printer many moons ago only printed wax for sculptings. Weak, but high resolution. We would print molds for gears and pour Ciba epoxy into the cavity. Worked great.
Erco nailed the application. ... and this is widely used more often than you might think.
WAX is used so that it can easily melt away and potentially be re-claimed and used again.
Like everything else there is a process involved where you build upon the result of a previous result.
Another benefit of using a WAX medium, is that you can create intricate "internal" features that you can't achieve with a CNC process.
@Bruce: What I mean is aluminum injection tooling has low machining costs, and where parts change, or other variables are in play, such as intermittent production runs, the investment in long life, hard tooling, doesn't always pay off well, or at all.
A "soft tool" easy to machine, etc... can pay off on the very first run of parts, and replacing tooling on demand rather than building one tool that can run for years and years is a great alternative.
Some scenarios:
1. Initial run unsuccessful, warranty failures, market changes, whatever, mean a revision would be more optimal and yield a better profit. Make new, cheap, soft tool, and continue. Innovation happens at moderate costs and risks compared to hard tools. This also matches well with the higher cost of 3D printing, in that understanding if there really is demand for many items can be had for incremental, modest dollar amounts.
2. Initial run successful. Sales aren't at expected rates. Subsequent runs may not be needed. Investment in hard tool would have ate up all the dollars resulting in a loss, or just no profit. Soft tool life actually matched market realities, and profit made. There is a gamble in this often overlooked.
3. Initial run successful, sales are brisk. Invest profits into hard tool, using sales from soft tool to fund it. You get a free hard tool at very moderate risk due to the market proving out being done on the cheap with a soft tool.
4. Derivatives make sense. Roll soft tool design into variations. Each is a risk, and may not see the higher volume the primary product does, but why leave money on the table? Serve more prospects.
Wax???? Gee thanks guys.... Now you really got me thinking that I am building the wrong kind of 3D printer. I have read about people printing with wax, but I never thought about the printed object being used for casting, although I have read about PLA being used for investment casting. DOH
Considering my interest in casting, this is definitely worth some more thought and research.
@potatohead
Thanks for the explanation, because now it all makes sense to me, and I can definitely see where "soft tooling" would be advantagous for unproven products.
These may work in place of wax. I don't know, but it's something to consider.
http://store.makerbot.com/dissolvable-filament.html
Bean
Thanks for the heads up about this filament. It is always nice to know what kind of techniques and materials are available. I appreciate it.
Cast however you can, cut parts free, and finish.
For small, high detail parts, make a lot of them, and most will fail in casting, but some will be fine. Work with the good ones and think hard about manufacturing processes that will perform better.
If it is a couple inches, soft tools can be done on modest, or even desktop mills. What you do is make something 5hay fits in a tool holder base for injection machines. Life won't be optimal, but you may find it produces many parts.
For smaller things, inject wax, then cast. Wax injection can be done for small things with old school, air driven portable wax injectors. You may find a tool for wax does not cost much, and you can actually use hard plastic for the tool as injection pressures can be low.
Back in the day we used to mill the cases themselves out of big blocks of aluminium and ship them.
No, seriously.
Someone on the forum suggested these guys for enclosures:
http://www.polycase.com/
I think they have a lot of nice looking enclosures. I think the prices are reasonable.
If using extruded aluminum you can make a custom PCB as the front panel. Dave of EEVblogs has a video on this topic.
I am still in the thought process, but my intention is to use as many prexisting products as possible. At the bare minimum, it has to be large enough to house a large LCD (4 X 20) and most likely a board such as the Spinneret, with ample room for for two pushbuttons and three LEDs. Something similar to an older caller ID for phones.
Thanks for the link.
LCD - 4 rows of 20 characters
FWIW, a run of 100 or so in sheet metal might be inexpensive as well. Just a thought.
I've printed stuff at similar sizes. Material costs can be a few bucks, and print times a few hours.
Sheet metal may be a few bucks per part, with a non recurring engineering charge on the first run.
If I were you, I would look hard at surplus or some standard enclosure to hit that same few bucks per part cost, but be able to buy on demand and let the sales fund better processes.
I suppose it would definitely be worth my while to at least search for a suitable surplus or stock enclosure. What I need and perhaps what we all may need eventually, is a nice long list of links for surplus and stock enclosures. In my opinion, the enclosures are almost as important as the electronics, because if it is not aesthetically pleasing, most people probably wouldn't purchase a product, unless of course, they really need the functionality of the product.
So in my belief, 3D printers are same bubble, that going to burst like it was dotcom craze.
The "dotcom bubble" was all about the internet, web sites, web services and what you can do with them. If you have not noticed the internet, web sites, web services and what you can do with them are still with us, more and bigger than ever.
The "bubble" part was all about investors losing a ton of money. Ah well.
Similarly I don't see 3D printers going away. They may be a bit of a fad or craze today but they have valid uses now as prototyping tools and one off production of odd things. They will evolve and improve.
Investors may once again lose a ton of money. Ah well.
I see too many 3D printers around me, bought inder "wow so cool" mood, but now idling. Printing is slow, accuracy is low, supplies are expensive, head needs special care....
All good and valid points, yet they still have their uses for things like prototyping, one (or a few) off manufacturing, They may never compete with current methods for mass production but they will improve in speed, reliability, and ease of use over time.
Some key patents have expired, and there is a lot of effort being put toward the $5k commercial price point. As those come online, we will see better software and farms of machines used much like farms of milling machines are.
Looking at Maker bot class machines and extrapolating yields a crappy picture. Instead, look at the really great commercial machines and scale down and there is a much better story.