Introduction: Light Ring on Underwater Robot -- Part 1 (Lathe)

Hi. For my engineering job, I had the opportunity to work on an underwater robot -- for science! (Picture 2)

One of my first tasks was to redesign the light ring on the front of the vehicle. It's dark under the ocean, and we need to be able to see.

The old light ring is a potted vessel. Basically, the previous team put lights in an aluminium housing and filled it with epoxy. (Picture 3).

Some problems with the old light ring:

- It would leak, short, and break.
- It is difficult to take apart and maintain.
- It would diffuse the lights and the lasers (yes we have lasers).

So the new light ring has to:

- Not leak!
- Be easy to maintain.
- Not diffuse the lasers.

Picture 4 shows a first prototype of the new light ring. It is an aluminum housing with an acrylic cover and some o-rings. In this photo, we had just dropped the light ring off the side of a boat to 400 meters (600 psi... think a few cars parked on a CD). We had applied different colored ink to each sealing surface to check for leaks. There was the movement of ink but no water, so it works as a proof-of-concept.

This document will detail the final light ring. Specifically, it covers the first half of fabricating the housing -- we are making the part on the left in the cover picture. These operations occur primarily on a metal lathe. Please make sure that you are properly trained on this tool, since it involves spinning jaws of death (seriously).

I am not a machinist, so take anything I say with a few teaspoons of salt. This instructable is probably best for a beginner on the lathe who has used the machine a couple of times before and could benefit from hearing the thoughts of another beginner on the lathe.

My workplace did not have the appropriate tooling, so I made this part at TechShop, a community maker space. See

Step 1: Materials and Tooling

A few things I used (Picture 1, left to right):

- A notebook (for notes)
- Calipers -- these go up to 8 inches
- My stock. I started with a 7-inch diameter piece of aluminum.
- Dial indicator
- Three cutting tools: a boring bar, turning tool, and grooving tool.

These photos are from the third piece that I made. On this piece, I started by cutting a 3-inch diameter hole in the middle of the stock using the waterjet. This saved me about an hour of work. Boring is, well, boring!

TechShop has rather nice (powerful and heavy) engine lathes (Picture 2). But, as a communal tool in a learning environment, it is a bit beat up. So making a precision part can be a bit challenging.

I started out by blocking off the area to discourage people from walking past. This particular lathe has a tendency to spray slippery coolant all over the floor.

The stock was about $20 from

Step 2: Workholding and Cut Plan

I started by mounting my part (Picture 1). We need two set-ups (one for the back features and one for the front features), and if you want control over the concentricity of these two sets of cuts, I'd recommend a 4-jaw chuck. I centered my part using the technique detailed in this 2-minute YouTube video.

My part drawing is attached (Picture 2). I circled a few important numbers for this operation.

We start with the back features. The cut plan is below. These notes might be confusing, so I'll include some pictures in the next steps.

1. Bore out the inside. (R 2.316)
2. Turn down the outside. (R 3.053)
3. Face off the back. (So previous two features are 0.53 deep)

Then we flip it around to the front and re-center. From there:

4. Face off the front. (So entire part is 0.70 thick and flange is 0.17 thick)
5. Cut the outer groove.
6. Cut the inner groove.

It will be more clear with the pictures in the following steps.

Step 3: Back Features

Picture 1 shows me slowly boring out the inside. I go through the whole part before I start leaving the flange. This is because I need to create a precision surface that I can later use for centering when I flip the part around. I don't actually have room on the outside to make the cut (it will be inside the jaws when I flip the part around, since the part is so short). So I make this surface on the inside.

Picture 2 shows me turning down the outside. For tooling, I used the same boring bar. I just flipped it upside-down and ran the lathe in reverse.

Picture 3 shows the part with the back faced to depth. I used the same boring bar in the same orientation, I just cut in a different direction. By not changing the tool set-up, it saves me a bit of time in set-up and zeroing. Just auto-feed slowly and you will have a nice surface finish.

I encountered a lot of difficulties with chip control when cutting the outside and inside diameters. The boring bar that I was using had a tendency to make terrible pigtails that would clog the tool, so I would constantly have to stop and rip the chips off with a pair of pliers. I tried dozens of combinations of different feeds, speeds, depths of cut, coolant control and even tool height. I spent hours calculating and getting things square and whatnot, and always got the same lousy chips.

Eventually, I said screw it, and cut at whatever angle, way off-center, with a slanted tool, whatever speed and feed, all sorts of ugly things. Lo and behold, the chips came off in perfect little 6's and 9's.

So to me, the lathe is a kind of a mystical creature.

What I really think, though, is that a good carbide-tipped tool with a fancy geometry would have given me the chips and speed that I desired. But, alas, I was stuck using the very plain and economical HSS communal tools in the cabinet.

I switched to the turning tool to put in the chamfers. It's a good thing to learn how to do -- other than deburring the part, the chamfers catch the light and make the part shine.

Step 4: Front Features

Time to start on the front face! I flipped the part over and re-centered it off of the bored-out inside diameter. The outer diameter would have been better to use, but in this case the chuck jaws were in the way.

I started by facing the part to depth (Picture 1).

Next, I cut the grooves (Picture 2).

Finally, I added the chamfers (Picture 3).

I want to write a little more about the grooves. In our first prototype (pictured in the intro), I had cut the grooves on a CNC mill. Long story short, it wasn't a particularly rigid mill, and I could not manage a good surface finish on the bottom face of the o-ring grooves. That surface finish is important, since it is a sealing surface.

The lathe at TechShop is a much more rigid tool, and it would also leave a circular lay (which would be good for the sealing action). Since I was using the lathe anyway, I decided to do as many cuts as possible on this tool. It works out since there are less people trying to share the lathes than there are people trying to share the CNC mill.

Picture 4 shows the grooving tool that I borrowed from the machine shop next door to my work. It cuts like a beaut. Notice the relief that allows it to complete face grooving operations without rubbing.

Picture 5 comes from Sandvik's website. We couldn't afford their face grooving tools, but I could still mimic their cutting procedure.

For more information on surface finish, Wikipedia has a pretty good article.

The Parker O-Ring Handbook is a definitive source of o-ring information.

Step 5: Part 1 Complete!

Here is an image of the finished part.

And there we are!

Some ideas on the time commitment:

- It took me 13 hours straight on this machine to make the first one.
- It took me 5 hours to make the third one.
- A better boring bar would have cut better chips and at least an hour off the time, since I wouldn't have to constantly stop and clear away tangles.

As a beginner, I was pleased with the results. A 32 finish was required on the bottom of the grooves, and I came away with about an 8.

That's it for now. Click here for Part 2, in which we take the part to the CNC mill to finish the features.