3D Printing to solve a computer part problem?

[Altamira]

----- Second hand information. I still don't have a 3d printer but I've been watching the technology for years. - - - -

From the ROV community there are guys doing 3d printing components which are pressure tolerant to 100m. Just a heads up the typical 3d printing method results in micro air spaces in the layers and crushes/deforms even at shallow depths.

Seem possible with the right amount of time and patience. People are cloning all sorts of little parts.

From experience, provided that cap is not some odd material (dried chewing gum perhaps?) and won't bond, an apoxy repair might be the easy solution.

If you go ahead with this I'd love to see the results.
Cameron

@northernone. If you come across a name and number for anyone in the ROV community using pressure tolerant materials for 3D print, I would love to chat with them to see what they are using. One company that does 3D printing in a variety of materials recommended Rigid Polyurethane, but I would like to cross check that recommendation with folks that actually use pressure resistant materials. Thank you.

 

[Fibonacci]

Given the close tolerances and smooth surface finish required to ensure the o-ring will seal 100% reliably under pressure at depth I don't think 3D printing in either plastic or metal will be feasible.
A better approach IMO would be to 3D model in Solidworks then get the part CNC machined in acetal (Delrin) or preferably 316 stainless

All volume production processes involve tolerances, and the battery cover you measure could be nominal size if you are very lucky or smaller/larger than nominal. So it is far better to accurately measure key dimensions of several covers (ie the sealing surfaces) and tabulate the data in Excel to obtain Min/Max/Median/Std Deviation S.

The CNC process will add its own tolerances (generally +/- .005" for metal and +/- .010" for plastics, but capable of +/- .001" in critical areas) so it is vital your specified CAD dimensions are correct.
Some light reading on General Dimensions and Tolerancing (GD&T) here:
http://www.pages.drexel.edu/~rcc34/Files/Teaching/MEM201 L6-Tolerance_RC.pdf

If you get a group of interested owners together on SB and do a small run it will be much more cost-effective... or do a bigger run and sell some on eBay!
At work we use Xometry, they have a Solidworks Add-in and an online automated quote service (I have no interest in the company, nor gain any benefit but they are very helpful and have a good reputation...)
CNC Machining Services - Xometry
Some good process resource material and how-to guides here:
Resources - Xometry
CNC Guide: Best Design Practices for Custom Machined Parts
Auto quote facility here:
Instant Quoting for 3D Printing and CNC Machining

Good luck!

 

[Altamira]

@Fibonacci. Wow, this project sounds a lot more complicated, and it is obvious to me that I am going to have to hire a very savvy person to make this work. It amazes me to think that a 3D printing process in plastic or metal might be less likely to seal properly than the crappy plastic hatch cover that is part of the computer. My guess is that cover is made in China or similar country. I will show your post to the guy I am meeting with tomorrow to discuss options. I really appreciate the references that you cited. Thank you very much.

 

[Fibonacci]

@Fibonacci. It amazes me to think that a 3D printing process in plastic or metal might be less likely to seal properly than the crappy plastic hatch cover that is part of the computer. .

My pleasure... hope the project is successful!
The 3D additive printing process involves generating multiple layers of material in either plastic or metal, even a $100,000 machine at its finest settings will still have a slightly 'stepped' surface finish which would need post-processing to creative a smooth sealing surface.
Yes its highly likely the battery cover would have been injection moulded in China but they are perfectly capable of producing top-quality parts IF the production drawings are correctly specified AND off-tool samples are closely checked before signing off the run.
Depending on expected volume, soft tooling could be used with your 3D CAD model as a lower cost alternative to full OEM production tools.
You then have a different set of design parameters to worry about... part lines, draft angles so the part ejects cleanly from the tool, flow lines and weld points, injection gate location etc.

Yet another very low volume alternative (5-50 parts) is to take a silicon mould off the repaired original part and vacuum inject it with high density urethane resin. Part rigidity and dimensional stability around the o-ring sealing surface could again be an issue, and you would just be replicating the same design-flawed part.
I'd be looking at an improved 3D design with CNC...

 

[diversteve]

A better approach IMO would be to 3D model in Solidworks then get the part CNC machined in acetal (Delrin) or preferably 316 stainless

I agree. I wonder though if you could cut it on a waterjet - some high performance bike/car shops have one - they use them to make custom metal accents. Maybe someone would be willing to cut you one cheap. IDK if they can cut to fine enough tolerances though.

Delrin would probably work, I've got some Delrin bushings a machine shop made for me years ago - it would be smooth enough with an o-ring under them. I forgot about the stair-stepping when I mentioned using a 3d printer in my other post - it would be an issue.

I think a really high quality industrial 3-d printer can be fine enough but they're really expensive. You won't get that level on a Makerbot.Even a low-volume extrusion would be out of your price range as someone has to machine the mold. Once they have that the cost to produce the parts is relatively affordable - my dad used to design them part of the time - he's a mechanical engineer.

From what I've seen of his work, Solidworks is the right tool. He once made a bunch of small plastic fittings used in oxygen delivery systems in hospitals - similar resolution to what you need most likely. They were extruded in mass production.

It would help if we could see a photo of the part in question. I know what mine looks like (Suunto) but I'm guessing yours is different from how you've described it.

 

[Spinal33]

Done it before for other parts - don't see what all the cynicism is...

1. Get a cad program (google sketchup is basic, but free... depending on the size of the cover and how many edges there you'll probably need something better)
2. Open an account on 3dHubs.com - and upload your design
3. Get it printed in plastic (shouldn't cost more than a tenner) or CNC'ed out of metal (probably a better option).
4. Fit the oring and use... (you may need to file/sand first to get a smooth edge)

 

[dmaziuk]

I agree. From what mine looks like (cressi leonardo) the "proper" CNC machining sounds like a bit of an overkill. You may need to hand-finish the surface, but if the underside is flat that shouldn't be too hard. OTOH if the battery compartment is not sealed separately, I'd find a chamber to test it in before diving.

 

[Fibonacci]

Anyone have images of the actual cover?
OP mentioned the need for a limited production run... that will have major implications for the RP type most suitable.

 

[Ghetto Diver]

I agree that CNC is a bit overkill. No need to draw it up.
The hatch will need to removed somehow. Using a caliper, get dimensions, thread size etc. Cut on Delrin. This is what I would do if I could get away with it.
Likely would take an attempt or two to get it right. Even when using CNC, you will go through 3-4 pieces to get the specs right.

 

[Altamira]

I gave the damaged hatch cover to a 3D guy this morning and he is going to scan it, making a few design modifications in the process (adding a notched square on the top to enable opening with a wrench, pliers, coin, or screwdriver), and then send the scan to a manufacturing company for cost options with a variety of materials. If it reasonably cost effective to make 3 hatch covers, I'll try one to see if it seals properly and withstands underwater pressure. If that goes well, I will make the scan available with details on the manufacturer, and other SB divers can pursue options as they choose. Will keep progress posted on this thread.