PUPS 8 & 9 | Most Critical Module & Complete Machine
To see all the calculations and work below as well as the problem and description head to my Transforming Table page.
Problem
Small apartments need furniture to be multifunctional and flexible. Who has space for a coffee table and dinner table and work table? What if the same table can transform between the three?
The MCM most critical module of this transforming table is the leg linkage and how it connects to the motor, via wires. In the early stages of this design I mostly focused on designing the linkage to optimize mechanical advantage and determine geometries and joints that would meet the desired height requirements for the transformation. Now, I focused on designing the connection between the leg and the motor, including the motor mount, the wire guide, and wire attachments on both ends.
Final Solid Model
Solid Model Updates
-
Created spline to match involute curves and spur gear pattern for motor connection (some iteration to best match the unknown spur gear attached to the motor)
-
Designed round wire guide spindle for height adjustment
-
Designed adjustable wire attachment mechanisms on both ends to ensure legs were adjustable to be at the same height
-
Created bi-stable feet for the table legs to reduce friction during transition, but also to provide flat contact surfaces with the floor while static
Bill of Materials
Manufactured Parts
Machines used: Mill | Lathe | Router | Waterjet | 3D printer | Chop saw, band saw, cold saw | Drill press | Sand blaster |
With limited time and resources in a term and on a student budget I worked to design parts that were functional, but simple to manufacture with materials that I could find around the lab. This meant that often in my design process I designed an optimal part then found the stock I would use, and then redesigned the part based on the stock available and the machines accessible.
One example are the long leg wood mounts. Originally I designed them to be waterjet out of a thick piece of aluminum stock, but the waterjet was down in MakerWorks so I found u-channel that wasn't a perfect fit but was good enough. However, this did lead to pin joints that had more washers and spaces than originally anticipated, with more wobble and compliance.
Below is chart of all the parts I manufactured for this desk and a brief description of the processes I used to make them. Following the chart are CAD images of each part.
2 Bearing carriage
3 Table Top
5 Wood leg mount
4 Pulley mount
6 Leg connector
7 Bistable foot cap
9 Wire spindle
8 Short leg
10 Wire alignment
11 Motor mount
Assembly
AXIS
Assembling the carriage with the bearings and rails went smoothly, but the pin joint and leg assembly was more tricky to insert the washers and spacers into the small gap and align the pin. This challenge made me think about I would preload this joint connection if I needed to? I'm glad that I ended up designing the pin hole such that it was a snug sliding fit, and that it was a through hole so that I could push the pin out and re-insert. I ended up being able to assemble a full leg plus the wire and duct taped the wire to the center of the table, which held 15 lbs. on the tip of the leg. After assembling my first leg I realized I would constantly be re-aligning the bearings whenever they fell off the rails, so I quickly made some stopper blocks to prevent the carriage from sliding off the rail ends. This really made life a lot better for the rest of the build and test!
After building my MCM it was clear to me that the legs did not provide adequate lateral stability. Even though my overall design with the legs placed in a kinematic geometry provides some lateral stability from a side force, I wanted to evolve my linkage design to be more stable. Unfortunately, with the time constraints I wasn't able to implement my ideas before racing towards full assembly.
Safety review - yes I would operate and let my loved one do it too. Manufacturing review - yes, I'll just be able to fully build and assemble before the final showcase. Some good teamwork as shop buddies and great reviews from Maha were essential to developing my manufacturing plans and talking through the details.
FULL TABLE
Pre-drilling pilot holes in the bottom of the table made for a smooth as butter assembly of all the table mounted parts, the rails, pulley mounts, and long leg mounts. Pre-drilling the center of the table also allowed me to line up the holes for mounting the motor so that it would pull equally on all three wires.
The most difficult part of assembling the table precisely was adjusting and securing the wires so that the legs were at equal heights, for a level table top. I finally decided to put all the carriages against their stopper blocks at the lowest position, measure the height of the legs and secure the wires. I included an adjustable mechanism on the pulley block to provide secondary adjustment once the wires were secured to the wire spindle.
TOLERANCES
Knowing that my table would be easier to manufacture if I could design it to allow for looser tolerances I designed many pieces to be adjustable or not require tight tolerances. The most important toleranced part is the carriage with tolerances discussed in PUPS 7, but when Protolabs machined the part it came back with holes 0.3mm under spec, at least better than 0.3 mm to big! So I reamed the holes to size for the pin joints.
The table top pilot holes were located with a tolerance of +/- 0.5 mm which led to accurate positioning of the elements attached to the table. Furthermore, in the leg attachments the holes for the pin joints were sized +/- 0.1 mm, and mostly met that spec with one measuring 0.5mm off.
Important to the measuring and assembling process was deburring all the machined edges of my parts. This is best practice for accurate measurement, assembly, and safety for the people handling the parts post machining.
MEASURED PERFORMANCE
To better understand the error of my linear axis and leg linkage I connected a laser pointer to the tip of my leg and pointed it at a wall about 20 feet away. I marked the laser center on a piece of paper, and then moved the leg up and back down into it's lowest position and made a new mark. I repeated this 11 times. I then chose a spot in the lower left corner of the data as a reference point and measured the x and y coordinates of each data point, recorded and graphed below. From this data I was able to calculate the average, standard deviation, and angular error between trials. I will also conduct this analysis for the full table, also with loading on the desktop.
Looking at the table above, the first thing to notice is that the standard deviation value is a zero, if you look at this calculation in the "one spreadsheet to rule them all" under a tab titled error measure you'll see that the values in the deviation column happen to sum exactly to zero. If I exclude the first or the last data data point, the x-deviation is on the order of 0.02 - 0.03 in. The y-deviation is also very small with a number with magnitude E-17. Looking at the x and y values for the laser points and the distribution on the graph it shows that there is a greater range, correlating to more error, in the x-direction than the y direction, which makes sense physically interacting with a leg assembly, as there is more lateral instability that vertical. Even so, there is only 0.217 inches of max error in the x-direction, and only 0.141 inches in the y-direction. This is less than my expected error of 0.25" in the y-direction projected in my error budget.
This data shows that the leg movement (unloaded) is repeatable to at least a 0.1 inch resolution. The nominal height of the table at it's low point is 22.25 inches. I updated the predictive model with this data to keep the error budget current.
Based on the results of this testing and closing the design loop with the MCM, I recognized the need to evolve the design of the bearings and actuator. I plan to add press fit nylon bushings in the short aluminum legs, which will reduce the vertical slop (when not pre-loaded) and will reduce friction and wear linkage.
2. Update error budget and FRDPARRC table and if needed iterate on the design. (1 pt)
4. Check the error budget one last time. (1 pt)
5. Make part toleranced part drawings, thinking carefully how everything will fit together. (2 pts)
