<div class="textcontainer">
<p class="margin"> </p>
<h3>Week 5: 3D Design</h3>
<div class="flexrow">
<img src="./Head2.jpeg" alt="" width = 600>
</div>
<p class="caption"></p>
<p class="margin"> </p>
</p>
I found my favorite pair of sunglasses on the ground at Harvard-Yale. They make me feel powerful. The bands are heavy and geometric; they scream "excess".
</p>I wear them everywhere: on the road, at school, at home. I think I've even fallen asleep in these. There's just one problem; they kill my peripheral vision and I have to crank my head like an owl to make sure I don't plow into someone when I change lanes.
</p>
<div class="flexrow">
<img src="./process.jpeg" alt="" width = 500>
</div>
<p class="caption"></p>
<p class="margin"> </p>
</p>
So I embarked on a long arduous journey in CAD. The mission? Recreate the glasses from scratch in Fusion360 and add "windows" to the bands, reminiscent of the McLaren Senna's door cards, to improve peripheral visibility. Ultimately, I would use the 3D model to cast the sunglasses in resin.
<p class="margin"> </p>
<div class="flexrow">
<img src="./senna.jpeg" alt="" width = 500>
</div>
<p class="caption"></p>
<p class="margin"> </p>
</p>
<p class="margin"> </p>
<div style="display: flex; flex-direction: column; align-items: center; justify-content: center;" class="container-fluid">
<p class="margin"></p>
<div style="display: flex; flex-direction: column; align-items: center; justify-content: center;" class="container-fluid">
<h3>Here's what I would tell my past self:</h3>
<p class="margin"></p>
</p>
Sunglasses have two main parts: band and frame. Both bend along 3 axes. this means simple extrusion and cutting won't do. Find another solution.
</p>
Your first plan, scanning the glasses and then "just cutting a hole" will be an utter failure, although it makes a good shark fin!
<p class="margin"></p>
<div class="flexrow">
<img src="3DScan.jpeg" width = 400 alt="a piece of paper with some diagrams drawn in black marker">
<img src="Fail.jpeg" width = 400 alt="a screenshot of possible keypad layouts">
</div>
<p class="caption">Technology is incredible.</p>
</p>
Printing tiny plastic hinges is unreliable. Harvest the hinges from an old pair of glasses and adapt the fitment in your model.
It is crucial to import canvases on both x and y axes to model from an image.
<div class="flexrow">
<img src="./Band.jpeg" alt="tinkercad simulation with code and wiring diagram" width = 400>
</div>
<p class="caption"></p>
<p class="margin"> </p>
</p>
Offset planes are your best friend. First, use these to create 6 sketches along the y axis to estimate the geometry of the frame from the top view.
</p>
Next, after sketches and intense visualization in your head, draw lines from which to create offest planes at the odd angles where your frames and bands are cut.
</p>
You won't be able to recreate the geometry perfectly. I'm still not sure how some of the angles work on the real glasses.
</p>
</p>
<div class="flexrow">
<img src="./Band1.5.jpeg" alt="" width = 400>
</div>
<p class="caption"></p>
<p class="margin"> </p>
</p>
The craftsmanship is incredible. Some of the traingles are not flat planes as they seem, but bent surfaces disguized as planes to make the bands look like a series of flat triangles.
</p>
Your first attempt is the wrong size.
</p>
</p>
<div class="flexrow">
<img src="./CAD.gif" alt="tinkercad simulation with code and wiring diagram" width = 500>
</div>
<p class="caption"></p>
<p class="margin"> </p>
</p>
So far, you've been making offset planes by drawing perpendicular lines, drawing a rectangle, and extruding that rectangle. This is much less efficient than the offset plane function. Stop doing that (I will continue to do this anyway later on).
</p>
Your second attempt is a little bit better, but you forgot to bend the frames. Thankfully, a heat gun will probably be fine.
<p class="margin"> </p>
When you go back to edit your first attempt, you will have created so many sketches and bodies that you'll have to start from scratch to make a small edit.
</p>
Even after hours of modeling, you still won't be completely sure how you're going to fit lenses in these glasses.
</p>
</p>
<div class="flexrow">
<img src="./Draft1.jpeg" alt="" width = 400>
</div>
<p class="caption"></p>
</p>
You create a last-minute lip on the inside of the frames in hopes that this will allow the lenses to sit flush. Still unclear whether this strategy will pay off.
</p>
You will test your scrap lenses to see whether they can be shaped with a heat gun. They can... kind of.
</p>
You will learn to create lofts and intersect a front-face sketch with 5 successive sketches on offset layers on the y axis to account for the 3-axis geometry of the frame.
</p>
<p class="margin"></p>
<div style="display: flex; flex-direction: column; align-items: center; justify-content: center;" class="container-fluid">
<h3>And it gets better.</h3>
<p class="margin"></p>
After 6 hours modeling the frame, you'll remember that there are more cuts you forgot to make. You'll spend another 45 minutes trying to understand from which direction you need to extrude your triangles to cut the frame, and you'll remember that you already mirrored the body, so you have to do everything twice and mirror all new cuts across the midline.
</p>
This is also a lesson in organization, as if you had just created components from the start, you could just reverse in the timeline and work on one side at a time.
</p>
<div class="flexrow">
<img src="./Band2.jpeg" alt="tinkercad simulation with code and wiring diagram" width = 600>
</div>
<p class="caption"></p>
<p class="margin"> </p>
</p>
</p>
Your first 3 attempts printing the almost-final sample will crash and you'll spend 2 days in the lab trying to
</p>
Insecurity will build as you look to a nearby table and realize in the time it took you to inch your way through a YouTube tutorial on lofts in Fusion360, your friend who has never used CAD created a custom printed DSLR Sony A7 3-lens attachment harvested from instant cameras.
</p>
Ever since taking apart your favorite glasses, they haven't quite fit right, and you worry you ruined them for this project.
</p>
<p class="margin"></p>
<div style="display: flex; flex-direction: column; align-items: center; justify-content: center;" class="container-fluid">
<h3>But there's good news too!</h3>
<p class="margin"></p>
You become less insecure when your friend admits his design failed and he had to copy one online.
</p>
Despite committing to get to bed at 11:30pm, you spend another hour trying to create components so you can combine your frame and bands in a single STL. Meanwhile, it's started to pour rain and the Dhall closed
</p>
Fast forward a day, you'll go into the lab again and crash 3 prints over 2.5 hours before your design finally prints correctly.
</p>
Print supports screw up flat surfaces, so flip your designs in Prusa to ensure the fronts are facing upwards.
</p>
<h3>Hinges are a problem.</h3>
</p>
The most difficult part of making a pair of sunglasses is constructing the hinge mechanism. In total, the hinges took me around 12 hours over 2 days using 2 different methods.
</p>
<div class="flexrow">
<img src="./fitment.jpeg" alt="tinkercad simulation with code and wiring diagram" width = 500>
</div>
<p class="caption"></p>
<p class="margin"> </p>
First, I tried a simple 2-piece metal hinge harvested from an old pair of sunglasses. I thought it would be easiest to adapt to my pair while being strong enough to ge rotated hundreds of times without breaking.
</p>
Attaching the hinges to my pair was surprisingly difficult. Worst of all, the angle was wrong. I was also feeling guilty for not trying to create the hinges from scratch, so I put the metal hinges on hold and designed a new pair with preattached hinges.
</p>
</p>
A few failed print attempts aand 32 hours later, my new version was complete. Sadly, these were also terrible and broke immediately. I also realized that the mechanism was likely not going to survive the crude molding process I would soon embark on to turn the PLA models into resin.
</p>
I went back to the metal hinges and adjusted my strategy by conducting a quick printing study with hinge attachments of incremental size. The best fit was put back into my 3d model and I printed the final PLA bands and frame.
</p>
<p class="margin"> </p>
<h2> Final Results, for now: </h2>
<p class="margin"> </p>
<div class="flexrow">
<img src="./Final.jpeg" alt="tinkercad simulation with code and wiring diagram" width = 600>
</div>
<p class="caption"></p>
<div class="flexrow">
<a id="btn" href="CAD.zip" download>Download my CAD files from this week!
</a>
</div>
<p class="margin"> </p>
</div>
