Well Windlass

Brainstorm

Another challenging brainstorm process! A windlass is not something that I see all the time in my life, but I definitely have an idea of what a windlass should look like in my mind. However, to my surprise, we came up with quite a few ideas that covered a whole foam core!

Similar with the bottle opener brainstorm, I realized that our structures are different combinations of variations of certain parts of the structure, i.e. we can summarize these different ideas by parts. There are four different categories: bottom supports, side supports, supporting rods, and handles.

The bottom parts are in shapes of a circle, a rectangle, two stripes across, and two feet. 

The side supports vary from simple walls, triangular structures, and boxes. The side support is very crucial to the stability of the model.

The supporting rods started with one simple rod, but then was increased to a combination of multiple rods, since the Delrin rod provided is not stiff enough. The most important part is that we want multiple rods to distribute pressure.
The handle also had a couple different structures, including both Delrin sheets and Delrin rods. We want to ensure the ease of usage but also strength of the handle, since it will take on a lot of pressure from the spinning.

The two key focuses of the windlass is stability and easy usage, but we also have to be aware of the limit on the material that we were allowed to use. As a result, for the bottom, we decided on a rectangular bottom which will hold the wall very firmly with multiple slots for press fit. For side supports, we initially chose rectangular side walls with a rectangular top, which is much easier to make than the other more complicated structures. For rods, we decided to go with three rods held apart by Delrin disks for its obvious strength. For the handle, we chose to add a rectangle to the Delrin disk holding the rods at the end to make sure that it would not break off as easily when turned.

Drawing of parts of our first iteration

However, this idea did not last long when we started thinking about the measurement--it uses way too much material. We very soon realized that certain parts of the structure do not contribute much to the stability of the structure or does not contribute enough compared to how much material it used. We first changed the bottom, since most of the part resting on the table with many notches for press fit does not contribute much to stability. Thus we changed it to the two-stripe design. The top also uses a lot of material. We decided to make it into a cross, which in fact provides more stability than a rectangle, since it is a triangular shape in the direction perpendicular to the gap (we believed that if our structure is not stable, it is most likely that it will shake in this direction. If structure shakes in the direction parallel to the gap, it would slide along the table, as opposed to collapsing.)

Drawing of parts of our second iteration

We were again restrained by the amount of materials that we have and decided that we must reduce the wall as well. The wall then became a trapezoid and thus changing the top into one stripe across, saving more material. The trapezoid wall, in fact, not only reduces the amount of material that we use, but also provide better support in the direction perpendicular to the gap, since it is harder to bend a triangle-like shape than a rectangle. The structure also becomes much easier to measure and to make, both as a foam model and in SolidWork.

Drawing of parts of our final iteration. The circular disks can be moved around and we can thus fit many more of them on our sheet of Delrin. There are also two holes on the wall for the Delrin rods to turn.

SolidWork drawing of the final version parts, before testing.

Engineering Analysis

SolidWork model, without securing circular disks

Foam Core model without handle

The overall stability of the structure is explained above during the brainstorming process. The problem of structure shaking in the direction perpendicular to the gap is solved by triangle-like walls. The problem of two walls falling to each other due to the potential bend of the winding structure is avoided by the top beam. In fact, there are are three pieces in the model that may bend under the pressure of the water bottle: the two bottom beams and the rods supporting and winding the rope. The risk of the bottom beams bending is avoided by putting stress on the walls which transfer the pressure to the table. The pressure on the winding rods are distributed to three rods so that two rods will always be bearing the pressure as the winding structure spins. It also ensures that no rod takes the full pressure during the whole time when the bottle is lifted, reducing the risk of bending and breaking.

The structure also saves energy and time for winding up the water bottle by using an axle-wheel structure. The axle-wheel structure is in fact also a form of leverage, following the formula:
F*R=f*r, where, in this case, F is the force applied by hand, R is the radius from finger tip to the center of the circular plate, f is the force applied by the water bottle, and r is the radius of the circle formed by the three rods.
Clearly, R is greater than r, then F is smaller than f. The three separated rod also allows the bottle to rise faster than it would have on just one rod. However, if we make r bigger, which allows the bottle to rise faster, we inevitable will increase F. Here we have to do a trade-off.

The string will be attached to one rod then wrapped around the three rods. This ensures that the rope will wind around the winding structure of three rods, as opposed to having the rods spin freely in the hoop made by the rope.

SolidWork and Testing

To put everything in SolidWork is in fact much easier for us this time, since the measurement is already settled beforehand and we are more familiar with the program. The biggest challenge came when we starting our testing process.

We decided to press fit the bottom beams to the side walls and heat stake the top. The winding structure will be joined by tight fitting rods through the plate, but we want to make sure that it can still slide when we apply enough force so that we can move each plate to the right place.

The press fitting, fortunately, did not take too many tries. We had to hammer the pegs into the notches, but it did no damage to the Delrin sheet and would not affect the strength and sturdiness of the structure. In fact, a good press fit ensures that the whole structure would wobble very minimally in the direction perpendicular to the gap.

Successful test piece for press fit 

The fitting between the circle plates and the rods took very little time as well, since we are not very strict about the tolerance of the fit. We simply want to make sure that the plates do not slide along the rods freely without much force applied to it. The plates will endure force caused by the bottle transferred through the rods. However, these forces will be perpendicular to the rods as opposed to parallel, thus having little effect on moving the circle plates around.

The heat staking, which we initially thought would not be as troublesome as press fitting, took five tries before it worked. At first, the peg of the initial design was too wide to fit under the heat staking machine. We then changed it into two pegs, but when we cut out the test piece, it became a press fit and the peg cannot go pass the notch and then be melted onto the top beam. The third try had a peg that was too long. The fourth try had a peg that was too short so there was not enough Delrin to form a heat stake. Eventually, the fifth try was successful. However, this was not the end! After we printed out the whole structure and when I was trying to heat stake the top, it did not heat stake as it did on the test piece! I consulted Professor Banzaert and learned that this may due to a lack of length in Delrin, the time of heat staking being too short, or the temperature being too low. I also learned that the best way to determine whether the staking is done is to look for curled up edge. Since the air tube was leaking, it was blowing more air to the heat staker than usual and thus decreasing the temperature. I solved this problem by moving away the air tube during the heating process and by checking the edge before I stop heating.

Failed test pieces. Top row: 1, 2 Bottom row: 3, 4

Successful test piece

Reflection

 

Our design ended up saving much more material than we expected.
The total area = side*2+bottom*2+top+circle*6+handle
(14+1)*15*.5*2+2*14*2+2.4*14+pi*1.3^2*6+3*1=349.5cm^2
Everything fits in a rectangle of 500cm^2 comfortably. In fact, we did not even need six plates. While the rods do twist when there were only four plates (two inside the frame and two outside), it was only when the center two plates are far from each other. When they are placed closer to each other, the twisting does not happen severely. We still added two more plates to the structure, because we had a lot of extra room in our sheet, and we wanted to reduce the twisting to an extent that it was not visible.

One thing I would have changed was to press fit a small piece of rod on the handle, so that when winding the machine, the operator can hold onto a bar parallel to themselves, which makes it easier to operate with one hand. However, our design did end up using all of the rod provided. Thus such a design may require more material. We could also press fit another piece of Delrin sheet.

In terms of saving materials, we had a satisfying balance between sturdiness and reduction of material use. However, I truly admire the windlass designed by Olivia and Brooke (need to double-check) which is an arc and uses a multi-rod winding structure but much shorter. Their design allows the winding structure to sit parallel to the gap as opposed to across it, significantly reducing the length of the whole structure, providing more material to distribute pressure on the winding structure.

To me, it is challenging and interesting to make something I already know, since I find it a good practice for me to think outside of the box. In the past two projects, I feel like my brain was still rather restricted by the stereotypes that I already had in the back of my mind. My windlass, though not the exact same as the what I normally think of a windlass, takes on a very similar structure. It is hard to notice the flaws or restrictions of things that have existed for so long and have been used so widely. However, I truly appreciate watching different projects made by different people, since it provides me with examples on how to break boundaries.

I also learned a lot by analyzing how my brain process information. It turned out that my brain is very good at comparing things and organizing differences very well. Such way to organizing information is indeed helpful, since it breaks down the process of creating and innovating. I do not have to pull out a random design out of thin air. Instead, when I run out of ideas, I can alter the ones I have parts by parts, and who knows what will come out eventually!

Last but not least, I would like to thank my hardworking and talented co-worker Julie Chase, who has devoted much of her time to our project even after she decided to drop the class. We will miss you and the knowledge you bring to us a lot!