Spinning to Wiggling: An Example of Rotary-to-Rectilinear Motion

Assignment: "Write a blog posting on a mechanism [to convert rotational motion to linear motion] you find compelling, explaining how it works and why you find it of interest."

My favorite model of motion was the bottom portion of "Model 042: Rotary into Rectilinear Motion", by Virginia Downward & William M. Clark, 1930.  I'm excited it seems like a woman contributed to this design, which is exciting for 1930.  It's a pretty neat way to change rotational motion into back-and-forth motion; the pattern is very regular, and I find the triangle design asethetically appealing.  Its motion was not immediately obvious to me--I did not anticipate the change of direction from the bottom arm of the inner piece pushing on the outer piece alternately with the topmost arm.

How it works:

Starting:

The outer piece is stationary.  The inside piece rotates couter-clockwise (blue curved arrow) so its top arm pushes the outside piece to the left (blue arrow).  The outer piece will quickly speed up as the middle piece continues to shove to the left, but the outer piece will slow down as soon as its bottom edge is hit by the lowest arm of the inner piece.

Going left:

The inside piece rotates couter-clockwise so its top arm pushes the outside piece to the left.  The outside piece was initially moving to the left (green arrow), but the combined forces of friction and the inside piece's rightward push soon slow, then momentarily stop the motion.

Turning around:

The inside piece has now exerted enough force to momentarily stop the previously-leftward motion of the outer piece; as the lowest arm of the inside piece continues to push, the outer piece will start to move to the right.

Going right:

The outer piece is now going right.  Its top edge has just been hit by the uppermost arm of the inner piece, which is starting to push the outer piece to the left; the outer piece will soon slow, then momentarily stop, then start to go left again (back in the position for "Going Left" above).

A movie and more information about the model:

http://kmoddl.library.cornell.edu/model.php?m=472&movie=show

Windlass Part III: SolidWorks Works!

As mentioned previously (see Windlass Part II), last weekend my partners and I spent a fair amount of time putting the windlass together in SolidWorks.  After some trials and tribulations, we changed a few dimensions to make sure all the pegs fit perfectly; then we measured our Delrin and realized we needed to change a few more dimensions.  We were excited to print out our pieces on the laser cutter!

We didn't assemble the full windlass, since that would have given us a disproportionately small amount of information for the amount of additional time we would need to invest in SolidWorks--here's at least one of each joint assembled:

Another copy of the above assembled structure will stand on the other side of the "well" gap (so the A frames are on the edges of the gap, with the horizontal piece that's currently sticking out mirrored on the other side to hold the two frames together).  The complete structure will have two Delrin rods stuck through the holes in the bearings in the top of the A frames (a tight fit to provide structural integrity), with two larger bushings on either side of each bearing to hold them inside the A-frame holes.  We will heat-stake together the pegs to provide a sturdy base.

Everything fits!  Huzzah!

Alas, our excitement was short-lived.  Unfortunately, 5 minutes before printing, the laser cutter started billowing smoke (well, not so much billowing, but smoking a fair amount).  Apparently the blower fan, which cooled the machine and blew away the noxious fumes, wasn't adequately communicating with the laser cutter.  We hope it will be fixed before the end of the semester, but for now Olin has graciously allowed the few groups who didn't manage to print their windlasses to use the engineering college's laser cutter.  The design has been sent off to Olin with one of our fearless TAs, and we eagerly await assembling the final Delrin product!

Windlass Part II: Another SolidWorks Adventure

Ashley joined Hannah and me at the SolidWorks stage; we worked in class on 2/15 and came in on Sunday to finish (see Windlass Part III).
We spent too many hours working on this part:

The dimensions had to be perfect, and the SmartDimensions were a little tricky sometimes, but we persevered and eventually made everything fit nicely.  The middle circle, when cut out, will be the bearing for the two Delrin rods (a snug fit around the rods to provide structural support); we'll make a few more bushings with slightly larger outer radii to keep the bearings in place and serve as a handle to easily turn the axle. (We'll make two of the front face pieces, one for each side; see the cardboard mock-up photos in Windlass Part I.)

The final front face (x2):

The final base piece (x4):

The front face's three pegs will fit in the three holes on the base piece; the middle hole is for the lengthwise support (heat staked with the T), and the hole farthest right is for the trapezoid support.
The final lengthwise support (x2):

We built this part at least four times:

I tried to do magic with similar triangles, but things quickly got complicated because we really had to work with similar trapezoids.  In the end, we used the same angles and modified the lengths based on what we measured in the Assembly phase of SolidWorks.
The final support for the trapezoid faces (x2):

The peg to heat-stake together the lengthwise supports (x4):

We carefully measured the holes to match the heat-stake machine, 5.45 mm by the 4.56 mm Delrin thickness of our claimed sheet.
After much thinking and pondering and a little bit of head-banging when VirtualBox gave us the Blue Screen of Death, we moved on to Assemble the pieces in SolidWorks (and iterated back from there a bit).

Windlass Part I: The Beginnings

The task: raise a 1-liter bottle of soda 10 cm above a 12-cm gap, using Delrin, our ingenuity, and a laser cutter. (Hopefully a laser cutter that doesn't billow smoke.)
The first designs:

The semi-final design and cardboard "looks-like" model:

The Delrin rods would form an axle spanning the two holes; they would be held together with various specialized bushings (E1 and E2 in the sketches above) to decrease deflection and function as bearings in the holes.
Although we initially chose 3 Delrin rods for the axle of the windlass, the cardboard model showed that 50 cm of rod would not be enough to easily turn the crank.  We were anxious about compromising the structural integrity of the model by having fewer rods--we tested the deflection of one, two, and three rods when pulling up a bottle and then switched to a two-rod design, which didn't seem to bend too much.
After further consideration, we decided to add another support across the triangular base, parallel to the trapezoid.  We used the cardboard model to test out two designs for the heat stakes that would hold together the trapezoid face and the wide base pieces, and discovered that pegs perpendicular to the trapezoid face would work best (as shown in the right-most sketch above).
With those few kinks worked out, we got the go-ahead from Professor Banzaert and dived into another SolidWorks adventure!

Reflections on a Bottle Opener: Part II

Unfortunately, when I came to the engineering lab to continue the bottle opener design on Friday, I discovered that my partner had dropped the class.  I had just begun to attempt to draw the bottle opener in SolidWorks, and Dana joined me as my partner for its completion.  Together we struggled to enforce the correct relations between the partial ellipses we used for the finger holds--while we knew how to put most things together, attempting to delete the guide ellipse we had drawn proved more difficult than expected.
Things went awry:

With the help of our many TAs we finally managed to complete the sketch and print out the bottle opener with the laser cutter.
The final sketch:

The final Delrin model:

 Success!

In the end, the bottle opener worked spectacularly well.  Our only minor issue was a slight discrepancy between the Delrin part and the foam core model; estimation with the ruler on the laser cutter screen proved less accurate than we had hoped.  However, the part was still comfortable to hold and easy to use with right or left hand, and but for an almost-imperceptible chipping of part of the center ellipse the bottle-opener remained unscathed after popping the cap off at least 5 bottles.  Given our opener's success, we would change only the sizing, by using Smart Dimensions in SolidWorks to specify the exact diameter desired.  All in all, our bottle opener was an exciting accomplishment to begin a semester of cool projects!

Reflections on a Bottle Opener: Part I

Although I outlined the design process in previous posts, I did not cohesively summarize all the trials and tribulations that finally led to success.  Here's Part I of the reflection.
The first step of designing the bottle opener was brainstorming with my partner Laura to come up with some possible opening mechanisms.  After about 15 minutes, we came up with all of these ideas:

We started out with a half-moon shape with a handle on one end, and considered using teeth to grip the underside of the cap.  A simple wrench design seemed too flimsy, so we thought about making the handle much wider, or having a handle on both sides to minimize the stress on the part.  This eventually evolved into a design that was basically gripping the bottle cap with one's whole hand wrapped around the part, which would have indents oriented nicely for each finger.  We used a Pugh chart (shown at the bottom of the page in the picture) to compare our four favorite design, based on user-friendliness, manufacturing difficulty (how difficult it would be to draw on the computer), stress (how breakable it seemed), and aesthetics.  The hand-shaped design was our favorite, although we knew it would be difficult to draw the appropriate curves.
We toyed around with how this part would actually open the bottle.  We did have the crazy idea of designing something to simply punch through the metal on the top of the bottle, but we expected that Delrin would have a tough time with that, and so confined ourselves to the simpler task of merely popping off the cap.  With the hand-shaped design, we initially planned to have a circular opening, and wedge one half of it beneath the bottle cap and press down on the other half on top of the cap, to pry the cap off.  We thought about putting the whole part below the bottle cap and trying to use the glass part of the bottle as a fulcrum, but then we measured the angles and realized this would be much more difficult, so we decided to just put the hole halfway over the bottlecap.
We then went to make our foamcore "looks-like" model.  I traced my hand onto the foam and cut it out with an Exacto knife (which was more difficult than expected, but manageable).  At this point, Laura came up with another very similar design, which was slightly smaller and symmetrical (the design on the left in the picture below).  We suddenly realized that we weren't sure how large to make the middle hole, so we measured the bottle cap and experimented with a few different sizes.  We finally settled on an ellipse slightly larger than the cap, where pulled up on the part to rotate around the major axis to open the bottle.  We liked both designs and decided to start putting both into SolidWorks.

Class 2/6/13: Joining things!

After formally testing the bottle openers (everyone's worked, especially mine!), we split into four groups to rotate around stations to learn how to join together pieces of plastic.  We used piano wire to make a rotating or fixed hinge, melted plastic at 400 degrees to fuse it together, played with calipers to measure the diameters of Delrin rods and bushings to see which would fit, and finally made a plate with notches of various sizes and semi-matching pegs to see how accurate our measurements needed to be to make a peg fit into a hole.  There are pros and cons to each joining method:

Piano wire:

Pros:

• Can be a slip fit for a hinge
• Can be tight fit that doesn't wiggle
• The drill press is really cool

Cons:

• Must be able to drill all the way through the piece--limited length of the hinge

Heat press:

Pros:

• Very sturdy seal
• Aesthetically pleasing

Cons:

• Bubble on the back (may cause instability of final structure if trying to stand on that plane)
• Can only join small pegs (not much fits inside the heat tip)
• Difficult to join things at a non-90-degree angle
• The cooling air is really loud

Pegs:

Pros:

• Can be any size
• Measure exactly with SolidWorks

Cons:

• Must fit exactly; difficult when trying to account for the laser cutter
• Only 90-degree fitting (unless file by hand)

Delrin rods/bushings:

Pros:

• Screw-in bushings are very sturdy
• Loose bushings allow for rotation, and can serve as a sort of washer or spacer between larger pieces of Delrin
• Tight bushings may be used to attach rods to larger pieces of Delrin, like nuts to hold on an axle

Cons:

• Loose bushings tend to make joints wobbly and less sturdy
• Loose or tight bushings must be sized perfectly; difficult when trying to account for the laser cutter
• Tight bushings are difficult to put on or remove

Bottle opener finished!

After some help with the amazing TA Essie, the SolidWorks difficulties were solved!  We had more luck with another of the various versions of the "trim entities" sketching tool, and managed to finally remove the last few remnants of the guiding ellipse by greatly magnifying the view to make sure the correct tiny portions of ellipse were deleted.  Then I printed the bottle opener on the laser cutter!  I used 3/16" Delrin for the part, as it was thick enough to be sturdy but thin enough to fit easily under the bottle cap.  I ran to open the first bottle I could find immediately after the laser cutter finished.

Ahhhh...sweet success!
Unfortunately, we had some difficulties with inaccurate laser cutter sizing, so we were only able to get an approximate size; while the bottle opener works quite well to remove a cap, the part is a tiny bit too large for the perfect grip.  While it is by no means uncomfortable, and for someone with larger hands this size would be ideal, the part does not exactly fit my hand in the way of my foamcore model.  But despite this minor issue, the bottle opener works fantastically, easily opening bottle after bottle!

Class 2/1/13

Much to my dismay, EXTD 160 no longer fit in my lab partner Laura's schedule, so I was sadly alone for the first part of class.  Happily, Dana joined on Friday, so together we ventured forth to continue the bottle opener saga!
Dana approved of my design, so we spent most of this class trying to put it into SolidWorks.
We wanted the part to look like this:

Unfortunately, it spent much of its time looking like this:

We weren't really sure what happened, but basically my first forays into SolidWorks were...less than successful.  Our wonderful TA Essie (one of many amazing supplemental instructors) showed us how to make curves much more easily with the fillet tool, and patiently fiddled with things when they continuously failed to work.  For some reason, when we tried to trim away parts of the original large ellipse base we had used as a guideline for the part, more ellipses would pop up, and the curved spaces for the fingers would completely change.  We think the main problem was SolidWorks incorrectly defining the relationships between all the different ellipses, though the repeated appearance of extra ellipses remains a mystery.  After spending almost the entire class fussing with SolidWorks to get the curves the right sizes and orientations and attempting in vain to remove the guide ellipse, we discussed our problems with Professor Turback, who had many interesting ideas about what might be wrong but was unfortunately unable to help much.  Class ended with slightly dampened spirits, as we envied the many groups that had already printed out their designs on the laser cutter and were happily opening and re-opening their soda bottles.  But Dana and I found a time to meet with Essie outside of class to work with the program some more, and we were hopeful that we would soon enjoy a similar success.

Class 1/30/13

We had a brief overview of engineering, then tried to build the sturdiest cantilevers with one piece of paper and a binder clip.  My group was quite successful with our tightly-rolled design, but we only tested one other idea, so our brainstorming session was less effective.  After we briefly touched on the physics behind cantilevers, we discussed the general process of design:

• Identifying the problem
• Developing concepts
• System-level design
• Detail design
• Testing and refinement

We emphasized the iterative nature of this process, which seemed reminiscent of the scientific method.  Another way to think about this:

• Research--it's important to know some physics so you have some idea of how things could work (and what won't ever work)
• Brainstorming--15 minutes of chucking out creativity as fast as possible, without discouting any ideas yet; don't get stuck on the first good idea, but push through the "crazy idea" trough to get to the really brilliant breakthroughs
• Selection--throw out the utterly impractical ideas
• Analysis--make a Pugh chart to compare the best ideas; combine the best features
• Prototyping--make "looks-like" or "works-like" models, or even actual full-size things that are just not manufactured on a huge scale yet
• Experimentation--check to see if things work!
• Evaluation--note what needs to be changed, go back to the beginning and start over

Finally we divided into pairs to try to design a bottle opener (lab partner: Laura Liu).  Spurred on by the promise of real South American Fanta, we excitedly began scribbling down many ideas in our initial brainstorming.  We used the Pugh chart to whittle down our ideas into two good designs, then cut out "looks-like" models with foam board.  Next class, we will draw our designs in the CAD software SolidWorks and then use the laser cutter (!!!) to cut out the final product from 1/8, 3/16, or 1/4" Delrin (my group chose the 3/16" Delrin as it will be thick enough to be sturdy but thin enough to fit underneath the bottle cap).
Although we initially discounted the similarities between a bottle-opener and a cantilever, we came back to the physics to try to figure out which design would be most effective with the Pugh chart.  Deflection is directly proportional to the force and the cube of the length, and inversely proportional to Young's modulus and the area moment of inertia of the cantilever.  Length seemed the most important term here, since deflection is proportional to its cube, so we tried to make the lever length small to ensure a sturdy bottle-opener.  On the other hand, we know from physics that the amount of force needed to apply a given torque must be greater for a shorter lever--so a shorter handle on the bottle-opener will be more difficult for the user.  We discussed the benefits and drawbacks of lengthening the handle, and eventually decided on two very similar designs with no real handle at all:

Above: lots of different designs, with a Pugh chart comparing our four favorites.
Below: the two designs Laura and I finally picked (very different from our first 7 or so ideas)--traces of our foamboard "looks-like" models.

Laura and I both enjoyed our designs; while Laura's was more symmetrical and aesthetically pleasing, my opener was ergonomically designed to fit just between the fingers.  Either design could be used equally by right- or left-handed people (mine would flip upside-down).  We fiddled with the center hole for a while--we briefly toyed with the idea of giving the bottle opener teeth, but discarded that after realizing how easily they might break off.  We eventually decided on a smooth ellipse whose minor axis was the diameter of the bottle cap, with a slightly bigger major axis--we would rotate the bottle opener around the major axis to pull off the cap.
We were excited to draw our design on SolidWorks and print it on the laser cutter soon!