Background
I was looking for some fun, clean, April Fools pranks when I found a printable joke about the worlds smallest sand castle, "carefully crafted by microscopic nanobots." What you really do is glue a tiny pinch of sand to the paper.This might fool some gullible, elementary aged children, but in my line of work we can easily view objects on the micro and nanoscale. We have powerful optical microscopes, scanning electron microscopes, confocal microscopes, laser interferometry, etc. So I got thinking about what it would take to really build the world's smallest sand castle.
The Idea
I think it would be a fun project to actually build the world's smallest sand castle (if it hasn't been done already). To accomplish this I would build it from real grains of sand by stacking them one at a time like giant building blocks. Okay, they are actually tiny building blocks, but think about buildings that are made from huge granite blocks. I would treat each small grain as if it were a huge building block. That gives you a sense of scale of what I want to build.There are many different types of sand, but the typical size distribution goes from 0.06mm to 2.0mm. To make building easier I would try to find a high purity silica sand with grains filtered towards the smaller end of the spectrum, around 0.10mm. I think the biggest problem with building on this small scale is that the building blocks won't be square. Unlike salt crystals, silica crystals take a much more random form and will be rounded on all sides. This means I'll need to be creative to get the grains to stick together. The most obvious answer would be to glue them in place.
The last, and I think most difficult / expensive, challenge to overcome is how to actually perform the building. How do you move the sand grains into place and glue them there?! Really steady hands and lots of patience? I think not. I've done lots of work lately playing with servos, and motor controllers, but those just don't have micro-scale accuracy. After a little research I think I'll use micromanipulators. These are mechanical devices used mostly by biologists to work with individual cells under a microscope. If I can get my hands on one or two of these things then I might have a shot at a world record.
Progress
02/03/2015 -Has it Been Done Before?
Using my Google-fu I checked for any prior attempts at my idea. The current claim to the world's smallest sand castle comes from Vik Muniz. He has taken an individual grain of sand and used a laser to etch a picture of a castle. I think this is awesome, and a clever twist on the idea of a sand castle. However, it does not qualify as a 3D structure build from multiple grains of sand.I also found a few photos of small castles built with more traditional techniques by J.W. Gruber. He makes a tightly packed block of sand then carves out the castle. Probably the smallest freestanding examples with extensive detail.
Here is a gentleman who does micro-sculpting by hand between his heart beats. He could probably do it without using any extra equipment. His pace is a little slow, but that's exactly the attribute that makes him successful: Willard Wigan TED Talk
Looking For a Micromanipulator
- New on Amazon: $1,500
- Used Industrial: $150 - $500
- 3D Printed Low Res: $100
03/31/2019 - I still want to do this one. It just seems unique and fun. I've had the 3D-printed micromanipulator and VR goggles on my desk for at least a year.
I've been stuck on the optics. I'm trying to maximize both the free working distance and the depth of focus around the optimal magnification level for medium sand grains (about 0.1 mm). If you've ever used a microscope, you may have noticed that everything looks flat or out of focus. This is because when you zoom in to high magnifications, only a thin section of material is in focus.
With my design, I've been trying to overcome this flatness and give a sense of depth and dimension by using stereo 3D microscopy. This has been done successfully by Gary Greenberg, but he looks like an academic gone commercial. I can't afford his equipment, and he doesn't publish the design since he's selling them for over $40,000 on Amazon. From his websites it looks like he's relying on Z-stacking to get full-depth 3D images. In other words, the camera automatically takes images at different focal depths, then a computer combines all the parts that are in-focus to create a neat looking picture. So it's not REALLY real-time, live 3D viewing.
That depth of focus is going to really make this project a challenge. My problem is that I don't have the money to just play around with expensive optics, and I don't know enough optics theory to design my system on paper to know exactly what to buy and how that translates to a workable system. I've gotten a rough idea by studying scopes for soldering electronic components, and I found DIY stereo microscope that I might try.
Or maybe I should build the end-effector first and try to find a lab with a standard stereomicroscope and see if a standard mag like 65 or 90 is sufficient for what I need. Then I'd feel okay dropping the $250-$2500 for a standard lab unit.
