David Bailey - Opportunities for Students
If you are looking for summer research opportunities, your first stop should be our Summer travel, employment and research opportunities page. If you are interesting in working on any of the following projects, please contact me at firstname.lastname@example.org to find out if I have any openings.
An important prerequisite is that you enjoy doing hands-on experiments and connecting physics to the real world through data. For these projects, undergraduate students will be enrolled in either a Physics Supervised Study or Research course, an Advanced Physics Lab course, or an Engineering Science thesis. Students may have a better experience and be more productive if they are not working alone, so if you and a friend are both interested in a project, feel free to apply together.
If you are a high school student, please look at my High School Students section below.
Several possible projects build on work published in Not Normal: the uncertainties of scientific measurements, David C. Bailey, Royal Society Open Science 4 (2017) 160600.
- When a published scientific result later turns out to be wrong, we usually don't know why, except in rare cases such as "faster than light neutrinos" or "evidence for primoridal cosmic inflation". This makes it hard to improve our microscopic understanding of the observed heavy tails in scientific consistency. It would be helpful to build up a larger database of examples of systematic effects that caused, or almost caused, publication of inaccurate measurements. In particular, it would be worth looking in detail at historical measurements of important quantities such as Newton's gravitational constant or the speed of light
- It also would be interesting to look at data where it is possible to partially disentangle the variation due to statistical and systematic errors. Measurements by the same lab should share most systematic errors, so when multiple labs make measurements of the same several quantities, lower bounds on the systematic errors can be estimated from the correlations between the various measurements.
- It has long been known that the relative standard deviation of independent analytical chemistry measurements follows a quasi-universal " Horwitz curve". The reason for this behaviour is not well understood, but it is nevertheless sometimes used to judge fitness-for-purpose. It would be interesting to explore how universal the Horwitz relation actually is, and whether it is related to power-law uncertainty tails.
Energy Conservation is a fundamental law of physics that is believed to be true except when space-time itself is not constant, e.g. the redshift of photons caused by the expanding universe. Except for such general relativistic effects, however, any apparent energy non-conservation is always because energy is leaking in or out invisibly.
Physics knowledge is, however, always qualified by experimental or observational uncertainty. For example, the Review of Particle Properties summarizes the experimental constraints on everything from extra dimensions to magnetic monopoles to electric charge non-conservation. It is therefore surprising that it is not easy to find explicit laboratory constraints on energy non-conservation.
Searches for missing energy constrain theories with (almost) invisible particles, e.g. neutrinos, axions, or neutralinos, or theories with extra spatial dimensions into which the energy might leak. Searches for the appearance of expected energy constrain new physics such as dark matter (e.g. WIMPs), violations of conservation laws (e.g. electron decay), or universal scalar fields (e.g. continuous spontaneous localization).
Although it is highly unlikely, there is a tiny chance that energy simply might not always be conserved. Many popular "theories of everything" have extra spatial dimensions, so why not extra time dimensions? It turns out that it is appears essentially impossible to construct a sensible theory with more than one normal time dimension because energy conservation and causality get tossed out the window. Something like a gauge symmetry must be imposed to eliminate these problems (e.g. Two-Time Physics), but what would happen it this symmetry were broken, even by a tiny amount?
Researching limits on energy conservation requires compiling results from a wide variety of experiments and observations, and then organizing these results in a coherent manner. It would be interesting to set limits on:
- Classical mechanical systems, e.g. astrophysical orbits or rotations speeding up or slowing down.
- Spontaneous excitation of matter, e.g. nuclei, atoms, molecules
- Spontaneous creation of matter, e.g. photons, electron-positron pairs, other leptons, hadrons, ….
- Catalyzed, e.g. popping the vacuum and destroying the universe, or less catastrophically observing excess energy production at the LHC.
There is no easy way to parameterize energy non-conservation to compare the relative strength of all these different limits, so each is independently interesting.
Undergraduate Lab Projects
Possible undergraduate projects depend on my current interests and what equipment we have available. Many examples are listed as Advanced Physics Lab Special Projects, and other possibilities include:
Current Experiment Development
- Our new Earth's Field NMR/MRI experiment is strongly affected by spatial and temporal variations in the local magnetic field. We are working on active real-time compensation using fluxgate magnetometer feedback, and also plan thick passive shielding against noise.
- We would like to further develop Python analysis and modelling tools for our experiments. The desired tools include multi-peak fitting for experiments such as gamma ray and Mössbauer spectroscopy, and improved modelling of experiments such as Conductivity in less than 3 Dimensions.
- New experimental possibilities would open up if we had improved video tracking and measurement of objects for both existing and new experiments. We have used ImageJ and MatLab, but are currently moving to Python, using PIL and OpenCV.
All computational projects are designed to help other undergraduate students, so clarity, simplicity, and good documentation are more important than achieving maximum computational efficiency and speed.
Blue Sky Projects
I always have a few ideas kicking around the back of my head, but which are either low-priority, poorly defined, or not yet practical:
- Is it possible to observe special relativistic effects in a first year experiment?
Experiments for first year students must be inexpensive, easy to duplicate, and doable in just a few hours. An easy way to observe special relativity is to simultaneously measure two of energy, momentum, or velocity for electrons at momenta comparable to their mass (i.e. mec). This is quite doable for under $10k, but can it be done for less than a few $k? One possibility is Compton scattering of Cs-137 photons in a small cloud chamber with a magnetic field, imaged by a a webcam. Just observing cosmic ray tracks in a cloud chamber and doing linear track identification and fitting would be a useful first step, since this would allow students to observe the intensity, angular distribution and East-West asymmetry of cosmic rays. (The flux of muons at sea level is about 1/cm2/minute.) We have a large new cloud chamber that can be used for these studies.
I am interested in what works and what doesn't for student learning, and how do we tell.
- I often say that the purpose of physics lab courses is to teach all of physics not taught in lecture courses. I have a particular interest in advanced physics lab and practical courses, both at the University of Toronto and elsewhere. How can we assess our success in Advanced Lab courses?
I am sorry, but I do not normally have projects for high school students. Students are often too keen to work too early in someone else's lab, when they might learn more in their own bedroom/kitchen/basement/backyard/school. Just pick a physical system that interests you, and try to understand it experimentally or theoretically. You should have fun, and along the way you are likely to pick up useful skills.
- For example, if you are inclined towards theory and literature searching, the kinds of questions answered at what-if.xkcd.com can be lots of fun. Learning something like Sage Math (http://www.sagemath.org/) might help figure things out.
- If you like simulation of physical systems, vPython has lots of possibilities, or maybe Glowscript.
- If you like - as I do - experiments, then there are lots of possibilities. Here are a few examples of what I mean:
- How much corn starch do you need to add to water before a rock (or a marble) dropped onto the water will bounce? (This is the subject of a 15 April 2013 paper in the Journal of Colloid and Interface Science; or look atthis blog post for a related discussion.)
- If you drop a string into a shoe box, what is the probability it will form a knot when you try to pull it out of the box. (I have left out many details. I don't know the answer, although it is closely related to a somewhat well known paper. Just google "Spontaneous knotting of an agitated string". That is a very interesting paper - done by a student - but I have always found boxes of string or cables knot even if they aren't shaken.)
- If you drop a sheet of paper, what is the probability it will end up a certain distance from its drop point? The landing distribution probably depends on the shape, size, thickness, and stiffness of the paper, the orientation and height at which it is dropped, and the friction and roughness of the surface it is landing on. There has been quite a bit of research done related to this (e.g. falling maple seeds or dropped paper or tumbling cards or ....), but I couldn't easily find any examples addressing this particular question.
Last updated on 26 July 2017