Pre-meeting
After modifying some of our old designs and creating new designs for our chain idea of moving the solenoid actuator up and down, we headed down to the MnT lab to print and test our new designs. We also brought our fan to the MnT lab so that we can carry out actual testing of the fan.
In the MnT Lab
Fortunately for us, there were not many groups using the 3D printers in the morning, so we were able to quickly print our different designs on multiple printers.
Chain
For our chain idea of moving the solenoid actuator up and down, we scaled down a 3D chain from Thingiverse to make it smaller so that we can have finer control over the position of the solenoid actuator. We initially tried out printing at 0.5 scale, however, we realised that because of the smaller size, the supports that were supposed to be removed were stuck to the main body, hence we tried 0.75 scale next.
![](https://blogs.ntu.edu.sg/ps9888-2020-g07/files/2020/07/0.5-and-0.75-chains-side-by-side-300x169.jpg)
Right: Chain printed at 0.5 scale
Unfortunately, even at 0.75 scale, our chain was not able to work well. The main issue was still the supports being stuck to the main body, which made it almost impossible to remove without affecting the actual design. Our chain of 20 links ended up breaking at multiple points when we tried to remove the supports.
![](https://blogs.ntu.edu.sg/ps9888-2020-g07/files/2020/07/0.75-chain-300x169.jpg)
On a brighter note, we found that the links that didn’t break were still able to turn and twist as intended. We peeled a few links open to expose the inner and outer cylinders and observed that even though the supports fused together, most of the cylinders which allowed for the turning to occur did not fuse together and were largely separated. Some had a few points where they fused, but it was not severe and was thus still functional. as can be seen in the photo below, the inner cylinder did not fuse such that it touched the outer circle, as intended. However, ultimately we decided if we are to proceed with this design, we will print it at the intended size on Thingiverse.
![](https://blogs.ntu.edu.sg/ps9888-2020-g07/files/2020/07/closeup-of-0.75-chain-300x169.jpg)
Chain to solenoid bar-holder attachment
We did test our attachment method of having the bar holder for our solenoid actuator attached directly to the chain link.
![](https://blogs.ntu.edu.sg/ps9888-2020-g07/files/2020/07/bar-holder-chain-169x300.jpg)
The bar holder has a chain link at the right side of the photo above. This link would then be attached to the rest of the chain in the pulley system, allowing the bar to move up and down when we use a motor to turn the sprocket. Theoretically, we still believe this would work if we proceed with the idea of using chains. However, we are likely to be changing to a different motor (stepper motor and screw shaft) or using a timing belt that we attach the bar holder to. The image above also depicts how the bar holder would hold the bar, thus allowing the bar to slide in and out so the buttons can once again be exposed. This is to counter the previously raised issue that our original design meant that the buttons could no longer be pushed manually as the original design had permanent bars that block the physical buttons. The sliding mechanism would thus be a suitable option that we will continue to proceed with for now.
New direction for vertical motion
It would be difficult to control precisely the location of the bar holder with a chain and DC motor setup, which is a bigger concern than we initially expected. Tony suggested that we use a T6 linear CNC slide stage actuator, powered by a stepper motor, to achieve vertical motion, similar to how a 3D printer achieves vertical motion. This was similar to one of our previous ideas, using a screw shaft, which was rejected due to it being too slow.
![](https://blogs.ntu.edu.sg/ps9888-2020-g07/files/2020/07/t6-stepper-motor-225x300.jpg)
Another suggestion, by Hanyang, was to use a timing belt. Our concern was that the timing belt and the load might slip, since we’re transporting our load against gravity. We rotated a Creality Ender 3 3D printer by 90 degrees to see if the extruder slipped, and even though it didn’t slip, it’s still better to play safe and use a stepper motor with a screw shaft to achieve our vertical motion.
Mounting our things onto the fan body
Our idea of using 2 screws and 2 rectangles with slits to fit the screw in, and tightening, could not work, because the M4 screw that we planned to use had a maximum length of 60mm, whereas our fan body required a screw that is at least 90mm long. We will be changing the design to a rectangle with one of the long sides missing so that we can use 2 shorter screws on the 2 short sides of the rectangle to tighten it. There’s also the back up option of using cable ties (suggested by Tony).
Multiple linear actuators branch
For the other half of our group that is continuing with the design that involves multiple linear actuators, the new design that tried to solve the problem of the small retraction of the solenoid actuators fell short.
![](https://blogs.ntu.edu.sg/ps9888-2020-g07/files/2020/07/multiple-actuators-design-169x300.jpg)
We will be further refining this.
Post-meeting
For the vertical motion branch of our group, we will be looking into how we’re going to mount the T6 linear stage actuator + stepper motor onto our fan body, and how we will control the position using the stepper motor driver and a Raspberry Pi.
For the static branch, we will be further refining the design of the bar that pushes the button.