Dec 31, 2014

Candy Dispenser part 2

Mason Jar model from Daniel Flores and GrabCAD

I'm planning on entering this candy dispenser into a mason-jar instructables contest (what timing!), but I completely forgot to take in-progress photos when I was building it. So instead, I made a CAD model and will submit renderings instead. Below are the resulting photos:



























Dec 12, 2014

Secret Santa Candy Dispenser


This year, I made a wooden candy dispenser for Secret Santa. My hall's particular version of this tradition disallows spending any money for materials, but luckily we have an abundance of scrap wood.

I got one of these canning jars with the sealing lids. Throw out the seal part, superglue the lid's rim into a hole, and now the mason jar can screw directly into the piece of wood.


I laid out gear profiles using Solidworks's toolbox. Solidworks gears aren't involute, but my wooden gears don't care about power transmission. Not caring about power transmission also allows me to position my non-beveled wooden spur gears at right angles. (don't do this for anything real!)

I attempted to lasercut my gears from 3/8" plywood, but only got through 1/4" before the teeth turned to charcoal. On my second attempt, I lasercut guide lines into the plywood then completed the job with a scroll saw. The thicker large gear has two shapes glued together.

Scrap 2x6 and 4x4 form the body of the candy dispenser. I drilled holes and superglued metal pins in place. Scavenged dowels cap the pins. 

If I do this project again in the future, I would cut the dispenser hole further away or cut an indent into the body. Currently the user has to hold their hands flush to the machine or candy spills.





Oct 19, 2014

Motorized Ripstik: Preliminary CAD

Though "Ripstik" is a proprietary name, so I will need to come up with a better one eventually.

Ripstiks are great toys. They have small enough turning radii to travel in my dorm's hallways, and can actually handle cracks well enough to go outside. Their method of propulsion means I don't have to shift my weight to place my feet on the ground while moving, and carving in tight circles is fun. Casterboards also don't pretend to be efficient modes of transportation, unlike skateboards, so their silliness is fine with me.

If you don't know how casterboards work, here's a decent video (I don't intend on doing tricks with this board, sorry).

So I want to make a motorized version. To attach this motor, I want to incorporate it into a custom-made caster. And if I'm making my own casters, I might as well make the rest too, right? Lots of fun machining things.

I have some beautiful blue shim steel sheet that would be perfect for the deck, and I have some plywood to stick underneath it for stiffness. The rest of the materials I would have to buy, so I'm applying to techfair for funding. If they sponsor me, my project would happen through winter and be presented in February.  
shim steel has such a beautiful iridescent blue
yay techfair! (please fund me)
Plywood and metal composite works really well, actually. I could use 1/8" sheet aluminum and plywood for the ripstik chassis, since I have access to a metal brake and wood clamps. The strange geometries required are made possible through bandsaw and waterjet. And Solidworks has a great sheet metal tool, "flatten", that converts shells between 2D and 3D.



Front chassis and its flattened version

The overall shapes of my board closely mimic the contours of a real ripstik. My end goal is to achieve something that could be mistaken for a regular ripstik from far away, but is clearly custom up close.

Both front and back chassis are mostly hollow to accomodate batteries. Real ripstiks are only 0.5" thick, which severely limits battery space. I intend to use A123 batteries, so my ripstik requires a bit more than 1.25" thickness. Probably no one will notice.

There's a hole in the rear to accommodate the torsion rod that connects the front and back chassis. This hole is smaller than required, since I am still unclear about the capabilities of the metal brake. A proper-sized hole might have too little material to bend without breaking, and I might have to drill the actual hole after folding.

I am also currently lacking the bracket mounting holes to attach the deck. Working on that soon.




Back chassis and its flattened version

The back chassis is a little different. Acceleration pedal, back lights, and other hardware are going there, and the rear is raised to provide clearance for the modified back caster.

Casters on ripstiks are at a slight angle (~25 degrees), but adding an angle to the chassis would be unnecessarily difficult. My plan is to incorporate the angled portion into the mounting block half of the caster.

I'm still debating whether to make these unibody casters or to bolt on separate wheel-holding-armature-things (these really need a name) to the rotational part of the caster. Stresses on the wheel would likely break the connection points, but separating parts would make machining easier and waste less material. Either way I would have to carefully plan out the geometry, thickness, and material for these casters. Anyway, here's some preliminary CAD:

generic caster frame
Yes, I know some of these angles are not actually mill-able. But it gets the idea across. Thrust bearings for the casters are idealized as beveled cylinders.

front caster assembly with 76mm wheel

back caster assembly, missing sprocket and chain

Sep 17, 2014

cantilever loft

All photos taken post-completion
My current room is a bit smaller than the freshman double I had last year, so when I moved my glass desk here, I had trouble fitting a bed in the space. The conventional loft-bed setup calls for posts at each corner, but this setup would either conflict with a desk or have a post right in front of the door. Not ideal.
I had a few ideas for what I wanted in a loft:
  • 6'+ height: I didn't want to hit my head on it, nor did I want my tall-ish friends to hit their heads on it either.
  • minimal diagonal bracing: I wanted my loft out of the way and not be the only thing people see when they walk in my room. Diagonal bracing takes too much space from an aesthetic point of view.
  • open space: there will be no posts blocking my desk or the center of my room. All posts will be as close to the walls as possible.
  • stability: I should be able to hang from all points of my loft. I should be able to climb on and off my loft without creaking or other sad noises. 
East Campus rush provided all the lumber I needed (we use up a lot of wood) and even several 3/4" bolts, which proved useful (though overkill) in assembly. I made the bedbox using 2x8s and 2x4 slats; posts are 8' 4x4s; I had one long 2x6 crossbeam and some 2x4 braces.

Tools used: chopsaw, circ saw, drill w/ 3/4" auger bit.

Drawings from CAD: my working drawings are much more chicken-scratch

CAD model of the planned loft
2x8 lumber make great lofts. They are wide enough that you can screw a 2x4 beam to them, lay 2x4 slats on top, and still have room to insert a mattress. Plus, they're strong and unlikely to have been cut into sub-8' sections at the end of rush.

The bedbox was the first thing I built. Outside, I cut and dry-fitted all beams and slats, leaving room for posts, then drilled pilot holes for each screw to prevent splitting. Slats were spaced 12" apart. I then clamped 4x4 posts in place to drill bolt holes through both the bedbox beams and posts. This ensured that the bolt holes would line up. I stuck bolts through the posts and leaned them against the wall. Then friends helped lift the bedbox to 6'-1/4" as I secured the bolts.

I chose 6'-1/4" to be the height of the loft because I have a 6' tall bookcase. I gave myself a 1/4" error to ensure that the loft would rest on its legs and not the bookcase (1. dorms get upset when their institutional furniture gets locked in wooden structures, and 2. I have no idea how structural that bookcase is over time.)

1/2" plywood added after loft was completed. Here you can also see the 2x4 screwed into 2x8 for slats to rest on.
At this stage of construction, the loft was only supported on three legs. The bookshelf served as a temporary support while I attached the crossbeam and the 4th leg. After I learned an unhappy lesson about torque with falling wooden beams (luckily nothing broke!) I had more friends hold the 4x4 post steady while I screwed the crossbrace to the loft.

Once the loft was secure, it was a simple matter to climb up and screw in bracing. I did grudgingly add one diagonal brace, but it's not evident from the ground.

Crossbrace added
Sideview: One horizontal support further attaches the corner to the post.
A bookshelf nook underneath stores my mess!
(There's a few outlets that prevent shelves from sitting flush against the wall)

Though fun, using the bookcases and slats to climb on my loft wasn't going to be feasible with a mattress, so I quickly attached a ladder to a post. The 2x4 leg of the ladder is slightly offset from the edge of the bookcase so I still have grip room.

The one diagonal brace. I also wedged a piece of wood between the crosspiece and the wall to prevent lateral movement.

I'm really happy with the results. There are no ominous creaking noises when climbing/moving around my loft, and I don't notice it when I'm working at my desk. I've even hung from the middle of the crossbeam without seeing noticeable deflection. Running up and swinging from it, though, does produce sad creaking noises (not sure what my friend was thinking there)

This cantilevered loft design is the first of its kind in East Campus, so I was a little worried about how well it would hold up with use. In particular, I wanted to know how much force I could put on the cantilevered end before strain would break it or before it fell over (there's glass under that!) If it did well, maybe I could hang a swing from it in the future.

Well, I have a CAD model in Solidworks. Might as well put weight on it in FEA and look at the pretty colors, right? 

In Solidworks, I assumed my dimensional lumber to be red pine and Douglas fir and my bolts/screws to be generic steel alloy. (There a few varieties of pine common to Massachusetts, so I just picked one that had material info available.) I then put 300lbs of force on the farthest slat (right next to the diagonal brace) and ran the simulation for static displacement and strain. Below are the results.

The loft moves about the pivot leg
Crosspiece and pivoting post experience the most strain, as expected
Close up picture of strain around bolts

The simulation says that with 300lbs, the free corner of the loft deflects ~1cm (~0.4 in) This number seems reasonable, so I asked Ben and Wesley to stupid-check my results. They said "Oh, that's easy to do" and hanged from the corner.Their combined weight (almost 300lbs) displaced the corner around 0.25".  

So, the simulation is pretty close. Let's try 500 lbs.



More sad colors appear when 500lbs are placed on the last slat. It also appears that even at 500lbs, the loft moves as a whole and the screws/bolts don't fail. There's some scary displacement going on at the last post.

What happens when you put a literal ton of force on your loft? Solidworks complains and refuses to finish static analysis. (it outputs the message "Large Displacements Detected" before failing, so I assume the loft falls over.) The following is its output of "33% load", so with 667lbs.

I see posts lifting off the ground
That's a bright blue.
Solidworks says the entire thing will tip over (sorry desk), but also says the stress on the pine crossbeam exceeds red pine's fiber stress limit (not shown). So, 667 lbs at the corner is the limit before the loft tips over or breaks.

(no, I don't have a physical ton to test this out empirically)



Sep 1, 2014

cooking with powertools

What do you do when you live on a hall that loves to build ridiculous things? You propose a ridiculous building event. Putz Hall Rush 2014 featured an event called "Cooking with Powertools", where the freshmen made edible things in silly motorized/electrical ways. 


Earlier in the year Putz acquired an apple lathe, a device that can core, slice, and peel apples using only a hand-crank. By hand, it takes 2-3 minutes to turn an apple on the lathe. A motorized version would be a grand improvement.

An easy way to do this would be to remove the hand-crank and rotate the shaft in a drill chuck. But after examination, the only removable part of the apple lathe was a single screw at the end of the hand-crank. I didn't want to ruin a functional kitchen device for the sole purpose of silly hall events, so I made do with just modifying the screw.


I designed a plate that allowed a drill to be held in-line with the shaft and turn the hand-crank. The hexagonal hole in the center mates with a drill driver bit, and the screw-hole is a loose fit to avoid misalignment and overconstraint.

The plate is circular mainly for simplicity. I wanted to have Pi Tau Zeta visible on a hall-rush device, and the curved cutouts were placed for aesthetics and to reduce weight. I also wanted to minimize the risk of injury by avoiding using a spinning arm that could whack unsuspecting fingers. 

An unexpected bonus with this design was that, when not in use, the plate hangs from the hand-crank screw and Pi Tau Zeta is always right-side up.  



The improved apple lathe


(it doesn't work with tomatoes)


I've always been a fan of wood lathes, and too many kids these days have never seen one. I brought highschool shop class into the kitchen by constructing a carrot lathe and teaching the new freshmen to turn carrots. An old single-speed drill was clamped to the table with a 5/8" spade bit in the chuck. A philips driver bit jammed into a 2x4 block comprised the endstop. The endstop could be moved to three different positions to accommodate carrots of all sizes.

First, I carefully measured the drill dimensions to make sure the spade bit and endstop would be aligned. A 3/16" pilot hole and a hammer drove the driver bit into the endstop. A 1/4" throughhole supported the spade bit. 

The rest of the lathe box was simple 1x3 structure. I started with the endstop against the back wall, screwed in a 1x3 wall, then moved the endstop forward. Repeat. Afterwards, I secured the drill to the table with a small box of 1x3 and clamp. The lathe box was left unsecured for tolerancing. 

I finished the carrot lathe in the two hours right before hall rush, and was pleasantly surprised with the results. This was the first time my creations had been perfectly aligned on the first try. I considered clamping down the lathe box as well to preserve this perfect alignment, but leaving it unclamped turned out better during the course of the night.   


The carrot lathe was a two-person machine. The operator pressed the drill trigger to turn on the lathe and placed pressure on the back of the endstop to ensure carrot stability. The freshman used a chisel to carve the rotating carrot with varying degrees of success.

If I made this again, I would add two things: a stabilizing bar for the chisel and a splash guard. Many students had trouble smoothly moving the chisel along the carrot, so the addition of a 1x1 bar or metal rod would help. I would also add an acrylic screen to the back of the lathe as a splash guard, since carrot shavings got absolutely everywhere in the kitchen. In addition, a splash guard would also protect the operators hands from possibly getting nicked by a stray chisel.



I was also involved in Spirograph Pancakes, inspired from Nathan Shields's Saipancakes. (He made a cool build-process video as well.) But I wanted to add a level of silliness and fun by making the batter-mixing an interactive activity as well. I took 1x3 blocks, carved out a mortise with some chisels for a 1-1/8" spade bit, then clamped the blocks together using four 1-1/4" drywall screws (sanitized with isopropyl, of course.) I rounded out the corners to better fit the mixing attachment to our bowl, taking care to avoid unbalancing the center of mass. Someone pointed out to me that real mixers have holes in them to force batter to flow, so I added those. Stick the attachment in a drill chuck and done! The stand mixer was complete with a quick frame of 2x4 and one 2x8 with a 1-7/8" hole to accomodate the drill (big enough to fit; small enough to minimize splash.)

Lauren Herring designed spirograph templates that we lasercut from 1/4" acrylic. The pancakes turned out beautifully, but using acrylic over a hot griddle for 6 hours ended up not being the best decision. Acrylic warps over high heat, so next time we will use plywood.

Wesley's event "Resistor Sausages" was the most noticeable. "Would you like your hotdogs cooked in series or in parallel?" was the best tagline ever invented, and the smell of burnt electrode permeated the entire length of hall. His sausages were extremely tasty, though.      


Lauren and Wesley being awesome with their food events


Cooking with Powertools: would do again every year.