The weblab as a project had been in 'development' for a number of years but finally overcame its inertia after a donation from Dr. Bahari, an alumnus of the School of Chemical Engineering. The 'mark 1' version has been built and is now being tested in Physics before being moved to Chemical Engineering.
There were several challenges that had to be overcome before the weblab experiment could come online. As well as designing and building the physical apparatus, it was necessary to incorporate data acquisition hardware, write simulations, and build this website that allows communication between the apparatus and the people who wish to use it.
The chaotic waterwheel as a concept has been around for a long time, but this version was designed in-house by Dr. Phil Robbins and Iain Kings of the School of Chemical Engineering, and Steve Brookes of the School of Physics and Astronomy.
After a good deal of discussion about the configuration that would be able to exhibit the chaotic behaviour required, but that would also provide the most interesting visual experience for the remote user, we decided that the wheel would:
Have 8 transparent buckets distributed evenly around a single aluminium ring in a configuration similar to that of a ferris wheel.
Be supported solely at the rear so nothing would obstruct the view of a front-facing webcam. This required a 3-bearing support system to keep the wheel sturdy whilst minimisng friction in the absence of an axle, which would have been more desirable but visually obstructive.
Have water delivered via a single water nozzle at the top of the wheel
Once the deign had been finalised after discussions with Steve Brooks, the wheel was manufactured by the staff of the Physics workshop and it is currently undergoing testing to ensure that chaotic motion regimes can be achieved.
The video below shows the wheel running at a gentle rotation in one direction. There is still a little too much friction at the bearings and it takes more mass in the topmost bucket to start the rotation than would be desired. However, simple rotation has been achieved. This motion relies on the fact that there is always more mass on one side of the wheel than the other to drive the continued rotation. Ideally, there will be a rotation speed at which the buckets empty before they reach their lowermost position.
The video below shows the wheel running in a periodic motion regime. This bifurcation is the result of increasing the input flowrate of water, with the interplay between the inflow and the outflow of the water constantly changing the horizontal centre of mass (COM) of the wheel. However, this changing of the COM is still predictable, hence the lack of a chaotic regime.
We believe that increasing the input water flowrate further would allow a transition to the chaotic regime, but a number of limiting factors (mostly involving all of the water beginning to flood the floor of the physics workshop) have prevented us from testing this as yet.
We are now working on reducing the friction on the apparatus and optimising the drainage from each bucket to best achieve each regime of motion, as well as building the enclosure in which the apparatus will sit so that we can avoid flooding issues!
Data Acquisition Hardware
Data acquisition hardware was provided by National Instruments. We are currently using a CompacRio chassis with digital I/O modules for the flow control and angular velocity measurement data. The hardware is operated by a LabView Graphical User Interface (GUI).
A simulation of the motion of the waterwheel was developed using the Matlab computer program produced by MathWorks. It was decided quite early on that users of our website would be able to both observe the current state of the apparatus and run simulations of the apparatus when they are not running the experiment themselves.
It was also decided that simulations of different levels of complexity would be provided so that users of different academic levels would be able to use these tools multiple times as they progress through their education. As such there will be simulations aimed at the following ages and levels of expertise:
< 16 years (school)
16 - 18 years (college)
Postgraduate students, academics and industrial professionals