Fossil Washer

We wanted to automate the fossil washing process to minimize the amount of time scientists waste. We therefore wanted to go from a hand held sieve on a stand to a fully automated process. The process of washing fossils is very delicate and time consuming since the sieves used are very fine and the fossils can easily break. Therefore we wanted to make a product that handles fossils with care and saves time for scientists.

1. 1 x Funnel
2. 1 x sieve
3. 1 x Solenoid
4. 1 x servo motor
5. 1 x Bunsen Burner Tripod
6. 1 x Turbidity holder
7. 1 x Servo Motor Holder
8. 1 x Solenoid Holder
9. 1 x Ball Valve
10. 1 x Battery Pack (6 x AA batteries)
11. 1 x 12v Power Adapter
12. 2 x Turbidity sensors
13. 2 x Plastic Tubes
14. 12 x screws + bolts
15. 1 x callipers
16. 11 x wires
17. 1x arduino uno
18. 1 x breadboard
19. 1 x 3D printed bridge mount
20. 2 x metal straps
21. 1x tupperware

Since washing fossils takes a long time, the idea is to automate the process. The criteria we needed to fulfill were as follows. We needed something that attaches to the pipe from the tap. Some kind of lid for the 20 cm diameter sieve to support the

device/mechanism. The water needs to move around over the sample in the sieve, it could spin around for example. It could have many small holes to make sure it sprays over a lot of the sieve. And finally, we need something to detect when the washing is complete. It could be to detect when the water leaving the sieves is clear.

So we needed to make a product that runs automatically, washes fossils carefully and notify scientists when done.

The idea is to have controlled waterflow in and out of the sieve using a solenoid. By using a solenoid, we can fill up the funnel and sieve as needed to break up and wash the fossils. A servomotor spins light plastic arms slowly to break up clay chunks and carefully wash the fossils. The motor is supposed to simulate a hand, softly running through the water. A turbidity sensor at the top bracket detects when water has reached the maximum height and turns off the solenoid. The water goes through a funnel with a ball valve and a turbidity sensor attached at the bottom. This allows the water to slowly run out, and get the cleanliness measured. When the water is sufficiently clear the whole machine stops and awaits human review for final washing. The overall idea is to minimize the amount of manual work

After many iterations, we decided on buying a funnel online. We mounted a plastic tube and used metal straps to waterproof it. In the end there is a ball valve screwed on with a holder for a turbidity sensor to measure the cleanliness of the water.

On top of the sieve we mount a platform with three mounts screwed onto it. The mounts are specifically designed to hold a turbidity sensor, a servo motor and a solenoid. The solenoid connects to the faucet via a metal tube. The turbidity sensor measures the water level which tells the solenoid to let water through or to hold it back. The servomotor is connected to a bar with plastic tubes. The servomotor turns them to wash the fossils.

We’ve primarily used Autodesk Fusion and FreeCad for the 3D designs.

For slicing we’ve used the open source tool OrcaSlicer and used the template for the “Bambu Lab X1 Carbon 0.4 nozzle”, alongside the plate type “PEI Textured”. For quite a lot of the earlier iterations we’ve also used an “Elegoo Centauri Carbon 0.4 nozzle”.

When printing difficult objects with specific geometry, we’ve relied on the Tree support type, as it is easier to peel off, than the conventional “Normal” support type. Everything has been exported to g-code, and the slice information has been studied to check whether the print will turn out fine, before printing.

We tested most of the flat designs (such as the tech bridge and some of the adapters) by laser cutting them, assembling and iterating, before 3D printing. Laser cutting was done in-lab and the settings for the machine used power=100, speed=80. We only cut in 3mm plywood for our quick iterations, but we’ve also thought about using acrylic for the bridge.

We made one arduino file containing the code for both turbidity sensors, the servo motor and the solenoid. This file was uploaded to one arduino board wired to a breadboard.

After printing we remove the supports from the prints. Then we realised that we didn't have any screws that fit our holes in the 3D prints. Therefore we had to widen the holes using a screwdriver and bruteforce. The 3D printed turbidity holder that screws onto the ball well needed small fixing. The part that holds the sensor was cut halfway toward the holder and filed down a bit to ensure that the sensor was far enough in the tube to measure correctly.

The electronics are the main part of the project, that basically makes everything work together. We programmed an Arduino Uno, to handle all the different components, making the process more seamless. The turbidity sensors, servo motor and solenoid, are controlled through the Arduino, and the different values that the Arduino detects from them, changes their behavior. The top sensor tells the rest when the funnel is almost full, and the bottom sensor tells the whole circuit when to stop. Most of the components are powered by a battery pack with 6 x AA batteries. The solenoid is the only component that is powered through a 12v power adapter, since it would drain the batteries too quickly.

There are plastic tubings around wires to waterproof as much as possible, arduino and breadboard should be kept in tuppaware box.

To assemble we first need to place the funnel in the tripod, the sieve in the funnel and our bridge on top.

When assembling the electronics, you’ll start out with just the arduino and a breadboard.

To start, plug the battery pack into the arduino. The red wire goes into VIN and the black wire goes into GND. After this we’ll connect the power and the ground to the breadboard. You can do this by plugging the red wire from the 5V on the arduino, into the powerrail with a red plus on the breadboard. The black wire goes from GND on the arduino, to the powerrail with the blue minus on the bread board.

Now we want to connect the bottom turbidity sensor. The yellow wire goes to A0 on the arduino, the green wire goes to the powerrail with the red plus, and the blue wire goes to the powerrail with the blue minus.

Now we’ll connect the servo motor and the top turbidity sensor. The top turbidity sensor is connected like the bottom one, but the yellow wire connects to A1 instead. To connect the servo motor, the white wire goes to digital pin 7 on the arduino, and the green and blue wires are connected like the others alike on the powerrails.

To finish it off, we connect the solenoid. You’ll have to connect the white wire to digital pin 8, the orange wire to the positive power rail and the black wire to the ground power rail.

Finally plug the 12v Power adapter in the wall, and connect it to the solenoid adapter.

V1: Our initial version included a 3D printed stative, a case for the sieve and a lid that covers it where a servomotor spins a bar with PVC strips to stir the water. There was way too much material waste in this iteration which leads to version 2.

V2: This version we cut the stand away since we use the scientists tripod, we made the sieve case into a funnel with two "arms" up the side to hold the servomotor parts. The arms holding the bridge was the biggest bottleneck for this iteration. We had to tweak this a bit to make it easier to print which leads to version 3.

V3: In this version we extended the bridge to incorporate a water level switch to stop the water input from the solenoid. A turbidity sensor to measure turbidity of the water via a small pipe that leads to an automatic drain gate. And a better way to incorporate the bridge arms. We printed the funnel but unfortunately there were a lot of issues with it. Which is why we wanted to try a separate branch of the whole design process where we bought most of the main materials which leads to version 4.

V4: In this version we bought a funnel, and used metal straps and rubber pibes to attatch a ballvalve at the bottom. We designed a bridge where we want to attach servomotor, turbidity sensor and solenoid, and just let it stand freely on top of the sieve. The main idea with this version was that the whole system was running in cycles, such that water was inserted into the sieve until it hit the threshold, which was determined by the bridge turbidity sensor. And use a servo motor on the ballvalve to output 50% of the water which would get measured by the ballvalve turbidity sensor.

V4.1: This final version implements the ideas from V4 plus three customized holder parts screwed on for the turbidity sensor to meassure water level, servomotor and solenoid inlet pibe. It also has a 3D printed turbidity sensor holder that screws on to the ball well to meassure cleanliness of the water. And here it runs continuously instead of in cycles, this was easier to implement and is our final iteration.

V4.2: Added three clamps with plastic tubing to stabilize the sieve in the funnel while still ensuring seamless removal of the sieve.

We tested by rapid iteration. This means we tested each part individually and then made iterations based on performance. We tested solenoid by mounting it to a water faucet and testing if it successfully can stop waterflow, electronics by applying code to an arduino board and then testing if the turbidity sensors, servomotor and solenoid functioned as intended. The 3D printed parts were tested by laser cutting 2D versions to ensure correct fit to parts such as the turbidity sensor.

To further improve our product, making a case for the “bridge” to waterproof it is crucial, other waterproofing improvements could be: waterproofing spray, shrinkwrap, natural waterflow ways in case of overflowing.

Making the design more sturdy and as one instead of multiple moving parts, or keep it modular for ease of changing parts such as waterlevel sensor, solenoid and servomotor.

Furthermore we think adding a light shaking mechanism to improve washing time would be useful.

Making the funnel custom made to fit the sieve specifically, is also an important improvement, which was our initial plan in version 1-3. This will enable us to remove parts such as plastic tubes and metal straps since we can then design it to accommodate the ballvalve directly instead of pivoting to DIY.

Also swapping out the turbidity sensor on the bridge into a proper water level instead.

The current design uses a breadboard, further improvement could be moving to a shield for the arduino uno.

The turbidity sensor that measures cleanliness could be improved by giving it a better measuring opportunity rather than the water just running through it fast. Like running the whole washing program in cycles instead of a continuous outflow.

Calibrate and optimize our whole system to save water and material. Use better signal filtering for our turbidity sensors. Buttons and diodes (or a small screen) for a more clean user experience.

Casper: Communication with CaseOwners, Electronics, Arduino code, Lasercutting, 3D print sketching, Brainstorming and Reiterating, Purchasing of materials, Writing of assignments, Soldering, Instructable.

Alf: Sketch-Design for laser cutting and printing. 3D-printing at home and on multiple printers at the lab. Brainstorming and Reiterating. Purchasing of Materials. Writing of Assignments. Assembly. Project Management. Instructable.

William: Lasercutting, 3D prints and Sketching, Brainstorming and Reiterating, support in purchasing of materials, Instructable