Well Windlass

Our project this week was to create a well windlass that is able to support and lift a 1 liter bottle of water. We were given 500cm² of Delrin and 50cm of a Delrin rod to work with.

We created multiple designs of the windlass, as shown below. The basis behind all of our designs was a set of two towers or legs that support the rod in the middle. The rod would be attached to a wheel and handle for cranking up the bottle.

The first three sketches involves stacking small squares of Delrin into towers, while the bottom sketch involves two triangles, each being made of two rectangular strips of Delrin. We thought that the triangular shape of the structures would give us a more stable base.

We proceeded to make a foam model of the second sketch, as we believed the stacking would provide for a stronger structure than the two triangles made of rectangular strips. We cut out many small squares of the foam to mimic the stacking of the pieces that we planned to do with the Delrin. However, after learning of the various methods of fastening and attaching Delrin pieces together, we realized that our model would be very difficult to make.

We still wanted a triangular structure, so we came up with this sketch:

This design consists of two triangles, with holes in the top center for the insertion of the rod, and pegs on the bottom. The pegs fit into notches on two rectangular pieces of Delrin that act as the base for the triangles. Between the two triangles is a wheel made of four separate circular discs, two small and two large. A handle is attached to the rod on one side of the triangles.

A foam core model was created:

The foam core model also incorporated four hinges that are attached to the base plates in order to prevent the structure from sliding around on the table.

We then proceeded to create the design on Solidworks.

From the foam model, we decided to add holes to the sides of the triangles so that the remaining Delrin rod could be inserted to give more support. We also made twelve tight fit bushings for securing the wheel and rods. We made the circular discs for the wheel to be a tight fit with the rod, and the holes on the triangles to be loose fit. The pegs and notches of the triangles and base plates were also tight fits.

With the pegs, notches and bushings came the difficulty of testing the tight/loose fits. From last week's Fastening and Attaching assignment, we knew that what the laser cutter produced would most likely be larger in dimension that what is specified on Solidworks. We tested to see what radius of the holes of the triangles, wheels, handle and bushings would give us the desired fit. The same was done for the tabs and notches for the base.

The tabs and notches only took one test, but the circles took two or three test runs.

When we finally had the appropriate dimensions, we laser cut the Delrin!

The laser cutter did not completely cut out some of our pieces, so we had to manually force some of the pieces out of the Delrin sheet. This resulted in us using sandpaper to smooth out the jagged edges of the pieces. The bushings and pegs and notches needed to be sanded down a lot to be able to form a tight fit because of the imperfections left by forcing them out of the Delrin sheet.

The band saw was used to cut the Delrin rod into multiple pieces. We put together the entire structure and attempted to lift up the bottle. We realized that obviously it was necessary to secure the wheel and handle to the rod, as the tight-fit holes alone could not take on the heavy weight of the bottle.

The next step was to do something so that the four parts of the wheel rotated as one. We had planned to piano wire the four circular discs together, and then piano wire the entire wheel to the rod. However, we realized that inserting a small piece of the rod into the four discs would be an easier way of making sure they rotated together. Consequently, we re-cut the wheel pieces on the laser cutter, this time adding a tight fit hole between the center and edge of the discs.

Now, we were certain that the parts of the wheel would not rotate at different times. We put the wheel on the rod, secured both sides of it with tight bushings, and press-fit-piano-wired the smaller two discs to the rod. We first tested the piano-wiring with a small piece of the rod and an extra bushing. Once that worked well enough, we piano-wired the actual wheel.

A bit of piano wire was left outside of the wheel so that the rope used to lift up the bottle could be hooked on to the wheel. Using the piano wire to attach the pieces together was a learning experience. At first, the piano wire would not stay in the drilled hole because it was not thick enough. However, after finding the appropriate sized piano wire, the process was quite easy.

And, the final step was to piano wire the handle to the rod. Our first attempt at this resulted in a cracked handle, so we printed out another handle and piano wired it again. This time, it did not crack.

We tested the windlass again, and it worked! We were able to lift the bottle 10cm above the plane of the table.

Our original design had included braces on the base plates so that the plates would not slide around on the table. They were meant to be piano-wired to the plates, but it was found that they were not necessary.

Also, we found that the rods inserted in the sides of the triangles were not necessary either, as the device was sturdy enough without them. However, it was a good precaution to have them.

We chose to make a triangular-shaped structure so that the weight of the bottle would be widely distributed at the base. A rectangular-shaped structure would have the risk of tipping over when cranked. The rectangular plates holding the triangles further contributes to the sturdiness of the base.

According to the physics of beam bending, the longer a beam is, the more easily it bends. As a result, we were careful to not have the two triangles be separated by too much distance so that the rod would not be too long. The properties of the material are out of our control, as Delrin was the only available material to work with, so the only element that we could control was the length of the rod.

One improvement that I would make to our windlass is to heat stake a small Delrin rod, or cylindrical piece of Delrin, into the handle in order to create a crank. This would make the device more easy to use. In addition, since the Delrin is not an extremely heavy material, cranking up the bottle, even with an improved handle, may result in the side of the structure opposite of the handle moving around. If given the right materials, I would weigh down that side of the windlass to make the cranking more stable. We are limited by time and access to materials in this course, unfortunately, to make such improvements.

Material used:
Triangles: 18x3x2  111 cm³
Base plates: 10x7x2 = 140 cm³
Wheel: 2x(pi)x1.8x1.8 = 20.4 cm³
            2x(pi)x2.3x2.3 = 33.2 cm³
Handle: (2x4)+(2x2) = 12 cm³
Total: 316.6 cm³


Overall, I would say that our windlass is functional but could be made more convenient to use. It has a good base and is therefore quite sturdy. The piano wire required a lot of careful positioning of the drill and the pieces, but was not difficult to use. The Solidworks process was a challenge, as testing the bushings and pegs/notches took a considerable amount of time. This project showed me the challenges and complications of designing and manufacturing products. There is a lot of testing involved before the actual product can even be attempted to be produced. That is why it is extremely important to allow for enough time to test the appropriate pieces and to re-do certain things when necessary. Also, it is important to have many ideas before starting to build. If we had built the current design in the first place, rather than creating the foam model of the stacking-squares design, we could have saved quite a bit of time.