Materials...

I was planning on making my WALL-E head/eyes mostly out of MDF with the critical parts being made out of aluminum. It looks like I was underestimating the total weight I would be dealing with. As it stands, each eye is estimated to weight about 9.5 pounds! Wow.. Even though each eye is about 8 inches wide and 16 inches long, 9.5 pounds is way too much.

The density of MDF varies but is 30-60 lb/ft^3 with ultralight MDF being close to 30 lb/ft^3. Aluminum on the other hand is 168 lb/ft^3. Some alternative materials are balsa wood which is about 8 lb/ft^3 and then there is precision board which starts at 4 lb/ft^3 and goes up past 40 lb/ft^3.

What I will probably have to do is start by first removing as much weight from the design as possible. Most of the pieces will end up being made from balsa with some of the larger pieces being made from precision board if I can't find large enough pieces of balsa. The more critical pieces will be made from MDF and I'll have very few pieces made from aluminum if need be.

The weight will be very important, especially for the head and arm pieces because they are extended out so far from the main body... =\

Final Eye Version (Almost)

As noted last time, I've made some changes to the eye socket assembly and this time I'm almost done with the rest of the eye itself. This time I redesigned the rear portion of the eye with the support beam in mind unlike last time where I put it in at the last minute and made everything inelegant. I've also update the pivot brackets and the way that they mount to the rear of the head to be simpler and just plain better.

Here you can see the 3 pivot brackets with a larger relief fillet than last time.

It's a little difficult to see with all the shadows but here's the inside portion

And here are some renders without most of the outer shell components.

Looks pretty cool, huh? :)

One problem that's been on my mind is that I might not have enough room to run all the wires through the rear connector. For each eye, there will be 4 servos, 1 stepper motor, 1 switch, and 1 accelerometer. I need to run all those cables through the rear connector. So if I were to run those separately, that would be 15 wires-ish (assuming the grounds are common). I might be able to combine some of the power lines if they ran on the same voltage but I don't know quite yet. Right now, I have a 1/2" ID line to play with but I was hoping for a 1/4" ID line to better fit the look.

Anyways, I was finally able to understand why my original head pivot (pitch, roll) didn't work! The original design called for the same two actuators with a center pivot to provide 2 D.o.F. but during the simulations, with both actuators locked, I still had an unrestrained D.o.F. That's why I changed it to the current design with the support spring and all that junk.

After taking another look, it was such a dumb mistake. The 2 D.o.F. design is pretty common as a simulator platform but I couldn't figure out why mine didn't work while all the ones I've seen online worked. Well, the problem was that I used a ball joint to support the weight while every was using U-joints. I totally forgot that with the shafts constrained, the U-joint doesn't allow for roll along the axis while the ball joint under the similar constraints will.... duh.

So, tomorrow I will be shopping online for a suitable U-joint (probably from SDP-SI) and continue on with the rest of the head joint. I'm just hoping that I will have enough clearance for the range of motion that I want in the head.

Alright, time to take a break!

Update: Okay, so here's the revised head pivot!

I had to reverse mount the servos due to clearance issues and I will have to create spacers to offset them even more to stick out more towards the rear of the pivot.

There are going to be some clearance issues with the linkage hitting the lower sides of the pivot.. you can already see it being pretty close. I'll either shave off part of the pivot or flip the servos such that the arms are on the outside edge. That'll also lead to some clearance issues with the eye's bottom cover but these issues can all be addressed without too much difficulty. I just have to find that sweet spot... The other issue I'm worried about are the rod ends. The website where I'll be purchasing them do not specify the maximum swivel angle and I have a feeling that I'm already pushing them beyond their limits, especially on the two located on the very bottom. I might be able to alleviate this issue a little bit by drilling the mounting holes at an angle downwards. Hmm...

And here are the final renders for tonight:

Final Revision Underway

I've started the final revision for my WALL-E. The final revision will include all the hardware callouts and will incorporate designs that I can fabricate with the tools that I have.  In addition, I will be going back and forth between parts and assemblies to see if I can re-use similar fasteners and other hardware to cut down on costs.

So far, here is the Eye Socket in it's (most likely) final form:

I went ahead and combined the individual element rings into a single piece because I think my CNC router should be able to fabricate it. I also replaced the individual ringed spacers with plastic spacers which should reduce cost and complexity. In addition, this design will give me the ability to adjust the spacing. I changed some hardware for the SPDT switch mounting bracket and added the remaining hardware callouts.

Back To Business

Now that school is over, I'm going to continue where I left off. Hopefully I will be to get some real progress done this semester. There are several things that are falling into place to make things easier for me:

1. I'll have access to the weld shop to make the body structure
2. I'll have access to CNC machines for the entire year (hopefully I can finish my assigned projects and then I'll be able to work on personal projects in my free time)
3. Found 2 friends with access to commercial 3D printers (this should be handy for some body panels and making molds)

So, I guess it's time to get started on the body structure. It'll probably take a couple of days to really nail it down.