Sculpture 2 Process: Philomel
It was important for me to base one of the sculptures on this cycle of coming and going which happens all the time above our heads. As our encounters with migrating birds are brief, it is difficult to see the disruption that is going on whilst we are not looking up.
Climate change is increasingly affecting bird migration patterns during their seasonal journeys between wintering and breeding grounds. The growing unpredictability of seasonal conditions is causing mismatches between birds’ arrival times and the availability of food, shelter, and suitable weather. As a result, many species are being forced to migrate and breed earlier, shift their traditional ranges, and face heightened risks during their migrations.
I will link here some interesting article on this pressing issue that informed my initial research:
-
Climate Change Is Pushing These Migratory Birds to the Brink
-
Climate Change and the UK bird by the British Trust for Ornithology
This video is also a great way into the topic that I found helpful:
My initial idea was to make a sculpture with wings. I wanted them to move in a graceful flapping motion but get progressively faster as someone approached it.
I collated images into a moodboard of inspiration as well as paying extra special attention to the birds in my neighbourhood looking for both calm and chaotic behaviours.
I started drawing out my ideas for the form the sculpture will take. I knew that the form would depend a lot on the mechanism I could create and this would also determine the scale so my initial designs were very loose. I knew I want the piece to be suspended above the viewer and for the wings to have 2 articulated extensions to the wing like birds I had watched. I was also open to the possibility of mulitple sets of wings to suggest a flock.
✈︎PROTOTYPING MOTION✈︎¶
For the mechanism I was looking into open source projects and tutorials on building a motor powered ornithopter with articulated wings.
These were the sources that inspired me and that had the motion closest to how I was visualising it in my head.
Again, I started by drawing out the mechanisms to understand how they worked and how I could put them together into a prototype that suits my needs:
I had been looking at the Michelle Vossen's documentation, in which she experiments with building an ornithopter during Fabacademy. I thought she took a nice approach to working out a reliable linkage system by laser cutting lengths of wood with lots of holes in so that she could try out different dimensions to find the best movement. Again as I don't have prior knowledge of mechanical engineering or the physics behind these mechanisms, trial and error seemed the best approach.
My goal was to make a system that is as adaptable as possible so that I could try out many options and refine which created the motion I wanted.
I quickly got to work making a file based on rough estimates and laser cutting the pieces from 6mm ply. I would use these lengths to create a mechanism with 2 gears as I had seen in other projects.
Next, for my gears I generated some basic gears using STL gears.com. Again I estimated what I might need ensuring that the diameter of the gear was wide to maximise my movement and that there weren't too many teeth so that it would be hard to print.
In Rhino I added various holes from which I could connect my wings at different distances from the Bore.
I also made some spacers of different sizes so I could test how far parts of the wing would need to be offset from the gears.
Here is the stl file for my gear, you can download it from Sketchfab below. And you can download the .dxf file to laser cut the test pieces here: 1
Iteration 1¶
The assembly process was very quick and dirty and largely intuitive. I took inspiration from the basic design in this video to make a frame from the pieces and gluing it to a solid base. I then connected the different lengths with a M3 screw and nut until the dimensions looked about right. I used some scrap metal rod I found in the work shop and bent it to create a length with more tension also.
ANALYSIS
-
I was pretty happy with the shape of the movement as it had the articulate wing and extention I wanted.
-
However, the bolts and nuts I used for hinges constantly came loose disabling the movement
-
My gears were very misaligned. With this prototype it was difficult to get the two wings moving at the same time and I had to constantly adjust things so it would keep working.
Going into prototype 2 I would:
- Use mini bearings and book screws instead of screws at linkage points and on gear bore.
- Use a gear simulator to work out the optimum distance between the gears.
- I needed stronger fixed points, these points were initially glued and this left too much flex in the mechanism so it was hard to tell if the dimensions were right or not.
- Intergrating connections for a motor.
Linkage Hardware¶
Before I could get back on to working on the file for iteration 2, I needed to work out the best linkage hardware to use.
After some research I decided that ball bearings would be most appropriate for the gears as this would reduce friction in my system. I would also test these on linkage points to see if it again reduced inertia. I used these ones as they were the smallest I could find: Mini Bearings.
I also decided to test Chicago Screws (or book screws/ furniture screws). Firstly because they are made with a smooth cylinder which creates less friction than a threaded screw, but also because the end locks into the cylinder with the thread inside, making it much more secure. I used these ones: Chicago Screws.
I tested each out to see how the movement was:
Due to the height of each bearing (4mm) I had to use multiple per connection to link each piece and put a ordinary screw through them. This was not ideal as it created a lot of friction and it was difficult to make the connection strong.
I decided only to use the bearings on the gears and connect the rest with Chicago screws as they gave a smooth and unimpedimented motion whilst holding the connection well:
Iteration 2¶
With this decided I made my in Rhino with the correct dimensions, resized the holes for the chicago screws and generated some more gears with the bore size the same as the mini bearings.
I also added an attachment for a motor to the gear and made them slightly bigger to increase the size of the movement a little more.
The new gear for the mechanism iteration 2 can be downloaded from sketchfab below and you can download the .dxf file for the pieces here: 2.
I glued the base pieces together again and used the same bent pieces of metal as I did before to assemble iteration two.
I was much happier with the movement I created in iteration 2. I loved the extension of the wings and the way the central and extended part of the wings moved differently from each other.
♒︎ELECTRONIC INTEGRATION♒︎¶
Choosing a motor¶
When I began thinking about how to intergrate a motor into this system I didn't know what kind of motor to use. I knew that a continuous servo would probably not have enough torque to move the linkage system so I was looking for different options.
I used videos like this to understand the options out there:
I made a pros and cons between a DC motor and a Stepper motor:
DC MOTORS PROS AND CONS
✓PROS✓
-
High Speed
-
Gear box can be added to decrease speed and increase torque
-
Great for continous smooth rotation
✗CONS✗
-
Low Torque
-
Need a motor driver
-
very high speeds
-
Doesn't know its position
-
Speed adjustment not possible without a gearbox
STEPPER MOTORS PROS AND CONS
✓PROS✓
-
Holds position when excessive torque (doesn't stall)
-
Can control position and speed very precisely
-
Microstepping control
-
High torque
✗CONS✗
-
Rotation not as smooth due to steps
-
Need a motor driver
-
Heavier
-
Harder to programme
From these considerations I decided to go with a Stepper motor. Because I knew I would need to programme variations in the speed based on the sensor.
I decided to use the Two Trees Nema 17 17HS4401 as it has a 42 N.cm/m holding torque which I felt would be sufficient to move my linkage system and was relatively small and light in comparison to the alternatives.
Choosing a Motor Driver¶
What is a motor driver?
A motor driver is like an amplifier which converts the low current signal into a high current signal. It goes between the motor and the microcontroller, getting power from an external power source and control signals from the microcontroller.
I looked at options such as the A4988, DRV8825, TB6600 or the L298N.
The DRV8825 looked like the best option for my situation as it was relatively cheap whilst still having a high enough max. current to use with my motor. It was also very well documented in use with the Nema 17
Image from mytectutor, DRV8825 Stepper Motor Driver With Arduino.
What is a Current Limit?¶
Setting the current limit prevents the current flowing through the stepper motors coils from exceeding the rated current limit of the motor.
Stepper motors like the NEMA 17 17HS4401 have a rated current (1.7A). If you push more current than that the motor coils heat up and excessive heat can permanently damaging or reducing the lifespan of the motor.
Motor drivers have current handling limits too. If the motor tries to draw more current than the driver can handle, the driver overheats.
By setting the current properly avoids damage and ensure smooth & accurate running fo the motor. Current affects torque, vibration, and position accuracy. If it’s too high you might get more torque, but at the cost of heat and noise. If too low, the motor might skip steps or stall.
I used these tutorials to work out how to set the current limit correctly on the DRV8825.
Working out the Vref¶
The V ref is a reference voltage, it tells the driver how much current to allow through the motor coils.
You work it out differently for different motor drivers but for the DRV8825 the equation is:
Vref = Current Limit (of stepper motor)/ 2 .
So for the Nema 17 17HS4401 the current limit is 1.7A, I reduced this to 1.68 A for safety margin.
Vref = 1.68 A / 2
Vref = 0.84V.
Setting the Current Limit¶
To set the Vref we use a multimeter to measure the reference voltage between the driver’s potentiometer screw and ground. The circuit is set up as shown below where 5V is supplied to the SLEEP and RESET pins. The DRV8825 should also be powered via VMOT. I used a 12V power supply.
Image from mytectutor, DRV8825 Stepper Motor Driver With Arduino.
Instructions
Use a small screwdriver and an alligator clip to connect the potentiometer on the stepper driver to the positive (red) probe of your multimeter.
Connect the negative (black) probe of the multimeter to GND on the driver or controller board.
Set your multimeter to DC voltage mode.
Gently place the tip of the screwdriver on the potentiometer (the small screw on the driver). VERY GENTLY, I BROKE SO MANY DRIVERS BY PUSHING TOO HARD!!
While touching the potentiometer:
Turn clockwise to increase the reference voltage.
Turn counterclockwise to decrease the reference voltage.
Watch the voltage reading on your multimeter and adjust until it matches your calculated Vref, in my case 0.84 V.
HERE ARE SOME IMAGES OF MY BREADBOARD FOR REFERENCE:
Stepper motor circuit¶
To get the stepper motor set up and spinning I used this tutorial:
){ width=600 }
I made some changes to their circuit as I was using the Xiao ESP32C3 instead of the Arduino and incorperating a sensor so my circuit was as so:
Fail
New linkage
Stepper working and interactio
ACRYLIC PROTOTYPE
STEPPER MOTOR/E ELECTRONIC INTERACTION
MATERAIL UPGRADE< DESIGN
FEATHERS>