OK, so I haven't remembered the words, but that was the gist of the exchange. The fact is, despite good intentions and probably a belief that we should share our skills and know-how more widely, there is a strong element of keeping-it-in-house when it comes to research. We all want to land those big funding grants, attract more PhD students, and so on, and to do that we need to be the best people in the country or world to do our particular job - so why tell others your trade secrets?
Well, the PhD student isn't completely misguided in her wish. It does happen. On Tuesday, for example, I visited BrainResource - a company with their main office in Sydney. One of their core businesses is BRAINnet: collecting together a huge database ofelectroencephalograms - from healthy people, from people with various conditions (epilepsy, depression etc) - in various situations - e.g. resting, sleeping, carrying out particular tasks - for the purposes of facilitating research. This data is made available (in certain forms) to others. The idea is that the database helps bring in research grants - for both the company and for other institutions. Overall, the winner will be the governments and health services that fund these projects, as they'll get a better outcome, as well as the BrainResource employees who get employed to do a fascinating job.
It's an interesting model of doing business, but it appears to be working.
]]>
I quickly got the impression that the way many (most?) tourists see the Blue Mountains is on a coach tour. The coach pulls up in the vast carpark at Echo Point, the tourists pile out with their cameras, take lots of photos of 'The Three Sisters', grab an overpriced icecream, then pile back into the coach to go to their next viewpoint, wherever that is. Certainly Echo Point has a feel of an international airport about it.
However, escape the crowds, and it becomes rather more pleasant. The way I did that was to head to the bottom of the cliffs. There's three ways of doing this. 1. Take the 'railway' (I assume this is a funicular - I didn't get a look at it) or the cable car down. 2. Walk. 3. Jump. I chose option two. That involved heading down the 'Giant Staircase' - about nine hundred steps in total down 250 metres or so of nearly sheer sandstone cliff.
One of the surprising experiences going down this number of steps is just how hard going it is. Near the bottom, my legs were burning - rather like having done a REV class in the gym. What's happening is that I'm exercising muscles that don't usually get a good workout, and it hurts. In physics terms, I need to provide a force against gravity to slow me down as I descent. The force of gravity alone would cause an acceleration downwards; in order to descend at a constant speed I need a force that balances this (Newton's first law), and that comes from under-used muscle groups within my legs. After several minutes of this, it hurts.
Incidentally, in energy terms this force is not supplying energy to you, unlike the reverse case where it takes energy to climb the steps back up. That's because the force is in the opposite direction to your movement - therefore it doesn't do work on you (transfer energy to you). In fact, it is doing the opposite - your legs are absorbing the potential energy lost as you head downwards. No wonder they are burning at the bottom.
Once at the bottom, it was a lovely walk through the bush (though I did get a little worried about what zero- or eight-legged nasties might be lurking in the undergrowth). I went back up a different route - past many lilttle waterfalls. I lost count at 1200 steps. And at 1000 m altitude, or thereabouts at the top, it's quite an effort for a near-sea-level-dweller like myself.
]]>
The gap in knowledge here are at least three-fold. First, how does the electric field interact with neurons and what does it get them to do? There's experimental data on how they respond, but the mechanisms of this are unclear. And then, when they do respond in the way they do, how does that lead to a change in neural connections? And finally, what is it about the neural connectivity that causes improvements in symptons of depression or Parkinson's?
These questions are a mix of many different disciplines: molecular biology, neuroscience, mathematics, and, from my perspective, physics, and probably a good deal more. Making progress on this question is going to require all of these skills. That's why a cross-discipline collaborative approach is so important to making advances in science. Being on study leave gives me a good opportunity to explore some of these possibilities. Yesterday's visit was certainly very enlightening in this regard.
]]>Virga can rapidly cool the air around it, causing downdrafts in the atmosphere which then heat as the pressure increases nearer the ground. Personally, I didn't notice anything particularly bizarre, but then again I was indoors for most of the day.
There's a nice video of virga on youtube.
]]>It was very reminiscent of a day a few years ago when a new PhD student in our research group turned up at Waikato and needed to get keys, out-of-hours-access card, library card, computer account, etc etc. I spent half the day taking her around the campus. Now at Waikato, this problem should now be much diminished by our student centre in the library building. A lot of the administrative things new students need to do are now located in one place. That's logic for you. Does it work? I'll have to track down a few new students when I get back and interrogate them.
Now, at Sydney this morning, in order to get a key for my office in the Physics building I had to walk close to a kilometre, through the campus, across a main road, to the security office, which for some reason is located right on the periphery of the campus. Apparently, all keys are issued centrally - departments don't issue keys for their own buildings. Then I had to walk back again. An easy experience? Hardly. At least getting the computer access could be done through the physics department. One positive thing from the trip, however, was that I've now located the swimming pool, which is large and indoor and cheap and available to short-term visitors like me and open early morning. Yay.
]]>
Well, I thought, that depends on how pedantic you want to be about it. I'll be travelling to Sydney very soon as part of my study leave, and that is going to slow down my aging by a few nanoseconds. That's simply a consequence of special relativity. The important factor here is denoted by the Greek letter gamma by physicists, and is the reciprocal of the square root of [1 - (v^2/c^2)]. In this expression, v is your velocity, and c is the velocity of light, or 299,792,458 metres per second. (The ^2 means 'square'). It tells you how much time slows when you are travelling at this speed.
What is gamma for a commercial jet aircraft? At 900 km/h, or 250 m/s, it comes to 1.00000000000035 . That means, when I'm on board the plane, for every 1 second of aging I do, people on the ground will age 1.00000000000035 seconds, that is and extra 3.5 times 10 to the power of minus 13 seconds. Over the course of a three hour flight that comes about an extra 4 times 10 to the power of minus 9 seconds, or 4 ns. Not enough so you'd notice, but it's measurable with atomic clocks.
One of the problems with teaching special relativity is that its effects nearly all lie outside the realm of everyday experience. They are only apparent when something travels very fast. However, they are extremely important in physics. That's the motivation for this piece of software I was told about a couple of weeks ago from the Australian National University in Canberra. The software, which you can download for free, lets you fly a spaceship around a city-scape at close to light speed, and observe some of the effects that are occurring, or observe different clocks at speed to see the time dilation effect. It's well worth a play, but to get the most out of it you should follow through the student instructions which come with it. The research article that comes with it is worth a read too.
]]>
Anyway, our new, glorious library and student centre gets better every time I go in. Not only has a new cafe opened up inside, but there are now screens telling you just how eco-friendly the building is being right now, by displaying data on the building's power consumption, solar power generation, water consumption and water capture, etc.
So, in the last month (which I assume means April) the building has used 182 792 kWh of power (that certainly makes my electricity bill look tiddly!) but has generated a cool 1 847 531 kWh from its solar panels. Now, I know April has been unusually sunny this year (shame the sunshine couldn't have come in summer when it was supposed to) but there is something clearly wrong with this figure.
One metre squared of area, under full sun, gets about 1 kW of power on it. That means in about an hour it captures 1 kWh of energy. I don't know how much of the building is covered in solar panel or other capture device, but I reckon the footprint of the building is about square with a side of 40 or 50 metres, so let's say about 2000 m2 in roof area. So, if that were covered in solar panel, under full sun it would capture about 2000 kWh in one hour. In April there are 720 hours, so that gives us 1 440 000 kWh of energy.
But I've assumed that the panels are illuminated 24 hours a day! That's clearly rubbish. So halve that, since half the time it's night. In April, the sun isn't anywhere near the zenith, so let's halve that again. We are down to about 400 000 kWh. Then there are cloudy days (albeit not too many this month) which will take it down again, and still a large factor to apply because the power conversion from light to electricity or hot water isn't 100% efficient. I reckon we might be down to a more reasonable estimate of 100 000 kWh.
So what does the 1 847 531 kWh represent? It's either a mistake, or I'm misinterpreting the display.
]]>
That I think is something worth thinking about. Nearly all the experiments that students do in class are things that have been done many, many times before. There's good reasons for that - they are (mostly) reliable, the equipment is available at sensible cost, the experiment can easily be discussed in terms of underlying theory, it's not dependent on the expertise of a single teacher that will be lost if the teacher leaves, it saves planning new experiments from year-to-year, and so on.
But it also means that students can look-up what the answers should be. That's got a couple of clear problems: 1. The student loses their ability to critically judge the credibility of their own work and 2. there is the temptation to 'cook' the results so you can get a better mark. The first of these is a real skill that a scientist needs to develop - if you can't look critically at what you've done and decide whether you trust the results then your science is on shaky ground. In real science, there is no answer to look up in a book - so how would you know if you've made a mistake?
The second is an interesting point, too. I would say if students find that they are getting better marks by making up results to conform to what the textbook says there is something going wrong with the marking process. Assessment should always be done with a view to promote learning. Copying things out of a book is not learning. I'd rather have a student try and fail to produce valid results, but, at the end of the experiment, understand what they were doing wrongly and why, and be able to correct it next time, than just to produce a graph that looks nice. Any experimentalist knows that most of his or her lab time is taken trying to troubleshoot the experiment. So why do we set experimental labs for our students that teach them otherwise?
This isn't an easy problem to get around, especially when you have a huge class of students. I tackle it with the second and third year classes by marking a student's work with the student present (in fact, getting them to mark it themselves first, then tell me why they deserve the mark they think they do). That way we can talk through the difficult issues and they can see that I'm not just after graphs that look good. But in a first year class, where there are 100 or more students to shuttle through in a week, that's not practical.
The placement student was very happy to have the opportunity to tackle a real experiment. Perhaps the answer lies in bringing in our research more into our teaching - getting the students to do some real research things.
]]>
One thing that's clear is that the heat pump generates a lot of convection in the house. That, of course, is the point - that's how the heat transfers from the indoor unit to the room - but in our case the air currents can be pretty strong. There are two ways this happens - 'Free' convection and 'Forced' convection.
Forced convection just describes the air flow due to the fact that there is a fan inside the unit - it's blowing the air around. Given that air isn't created or destroyed, this means that the air has to circulate around the house. Free convection describes the currents caused by warm air being a lower density than cold air. The warm air will be buoyant, and rise, which means (by continuity again) that colder air has to sink. The two mechanisms are both in place, and they are both very evident. First of all, one can feel the air moving around (and see the light shade swing a little bit as the air blows past), and secondly upstairs becomes nice and toastie while downstairs is rather cooler.
So, the problem then becomes setting the unit so that we get the 'best' situation. One obvious problem is that the forced convection results in the airflow into the unit being rather cooler than the average room temperature. The pump measures the temperature of the incoming air, and then attempts to take this to the target temperature you set. So if you set the target temperature as 19 Celsius, it means that the air temperature of the majority of the area, especially upstairs, will get to rather higher than 19 C. That we can cope with, simply by setting the target temperature a bit lower.
We've also found that the angle of air flow leaving the pump makes a large difference to the distribution of heat around the house. If it's too flat, it just scoots across the downstairs ceiling, and then heads upwards when the ceiling runs out - we end up with an oven upstairs and a fridge downstairs. If it's angled too far down, it blasts into the dining room table - it keeps you dinner nice and warm but it feels like you are eating dinner in a Canterbury Nor'wester. Again, with some experimentation, we think we've got a reasonable setting.
So that's all good for now, but there is a nagging doubt that when it gets much colder still the convection currents will change again and we'll have to 're-optimize'.
]]>
How do you know for sure? Take a photo of it, silly. And yes, that means yet more light. With a single atom, you can't take a conventional photograph, because the wavelength of visible light is around 500 nanometres compared to the atom size of around a nanometre. It's impossible to focus that small. But what you can do is measure exactly how much light is being scattered by the atom, and from that know for sure that there is only one atom there.
I had a go at this experiment. Admittedly, it wasn't desperately taxing on my abilities as a physicist - in fact all I had to do was push a button on a computer window and in half a second the machine did its thing, completely silently, and produced a nice looking photo with a splurge of light in the centre. And that light was being scattered by just a single atom.
So, you CAN see single atoms, after all. I have a picture to prove it. (If you want to see one, have a look on the group's website here).
Postscript. I'm staying with my sister-in-law and her family. When I got back to their house this afternoon, I showed the picture to my eight year old niece and asked her if she could guess what it was. She gazed at it for a few seconds, then said "An atom?"
]]>
In this paper, we first do a bit of quantum stuff, then develop the solid-state. At this moment in time, towards the end of the quantum section, it's not really obvious to many of the electronic engineering students why they should learn any of this stuff. It's all very abstract, and not clear how they would every apply it in their careers. That's not a good place to be in as a teacher, because students who can't see the point don't pay attention, and eventually slink off and do something else with their careers. One of the roles of the teacher is to show them the point of what they are learning. (Incidentally, I think if you the teacher don't know the point, then you shouldn't be teaching it, because if you can't see it your students certainly won't.)
So, I spent a while (not for the first time) describing how the paper will develop, and how the paper then supports further learning in the next year. From the bottom up, it goes something like this.
Quantum mechanics applies to very small things. Of particular importance here are electrons, since these are what are doing the work in 'electronics' (note the name is no co-incidence). In order to understand how electrons interact with other matter, we need, at a lowest level, quantum mechanics. Now, the QM that we do is pretty simple stuff, really only applying to a single particle in simple situations. But we take our simple, one electron situations, then use what we have learned to help us understand more complicated single electron situations (the hydrogen atom). From there, we think about how multi-electron atoms will behave. Then we bring in some quantum statistics, and talk about how many atoms will behave together. Finally, this leads on to describing how the all-important semi-conductor (e.g. silicon or germanium and others) will behave.
Once we are at the semiconductor stage it is the purpose becomes more obvious to the students. We can use our semiconductor physics (band theory) to explain why diodes and transistors do the things they do, and, because we understand why, we can start talking about some of the effects that happen when you pile in zillions of transistors close together on a single piece of silicon (a 'chip', or integrated circuit). And integrated circuits drive just about everything these days, whether it's your iPhone, computer, TV or car. This means our engineers aren't just working with a black-box understanding of a diode or transistor - they have some idea of why it does what it does, which will really help in careers where understanding an electronic circuit is vital.
To get to that point, we need some quantum mechanics. Yes, it's abstract, and yes it's very weird, but it is very very useful, for engineers as well as physicists.
]]>Then, today was spent marking tests. This had some degree of extra interest because this was a test that the students could talk in. I was fascinated to see how they did. Overall, I think they were all helped by it. It was interesting to watch them on Thursday, form little groups, and there was way more talking than I'd ever got out of the class before in a lecture or tutorial. That's got to be a good thing. Most of the groupings got most of the stuff correct - figured it out between them in the end, though there were a few who got one concept in particular wrong. I wonder whether this was because all the students in the group didn't get it to start with, or maybe because there was one forceful personality in the group who convinced others that the incorrect method was the way to go. Without recording the conversations going on, it's hard to know for sure. That would be a nice study in itself .
I haven't had much feedback from the students yet about what they thought of this, but what I have had is more positive than negative.
I'm running another test on these lines with a different class after the Easter Break is finished, so will watch carefully what the students do in that one, too.
]]>
The UK Education Secretary wants universities to have a much greater role in driving A-levels. He has concern that they are not doing what they should; that is, preparing students for university. The exam boards are failing to do their job in ensuring a quality qualification, and this role should be taken away from them. The perennial issue of grade inflation also rears its head - that's where year after year the grades awarded to students get higher - some would say it's because the teachers are teaching better - others that the exams are simply getting easier.
The complaint about falling entry standards at university is common - if courses ran at the level of difficulty they did 20 years ago, with a similar amount of assumed knowledge, there would be a huge and unacceptable drop-out rate, therefore standards have to fall.
The problem that I have with these articles and the debate in general is that so much of it is based on ideology. For example, ask yourself the following questions.
1. Should universities be there for the 'elite', or should they be open to all? And who should pay for them?
2. Is the purpose of an exam at the age of 18 to prepare a student for university, or is it to prepare him or her for the workplace, or for life, or what?
3. Should exams identify the performance of students relative to their cohort, or should they identify the absolute performance of the students (i.e. would it be acceptable for every student in an exam to get a grade 'A' if they all were good students)? Are 'grades' the way to go anyway?
4. Who should have control of the examination process? Should it be the government (or their appointed agency), universities, professional bodies, employers, the teachers who teach the stuff?
I would suggest that the answer you'd give to these questions would to a great extent depend on your ideologies. Debating these are often pointless - no-one is likely to shift their position in a hurry. So the UK government says that A-levels are no longer up to the job of preparing students for university. But who says that this should be the job of the A-level? Ideology. Is grade inflation acceptable? Ideology again. Universities should drive school exams? More ideology.
Anyone who wades into this mire is going to have to expect trouble on many sides, and it's no surprise the Education Secretary is getting it. A bit more open discussion might be a better approach, though I doubt this would change the general principle of "if you talk about school exams in any way, no matter what position you take, expect to greatly offend around 50% of your listeners".
]]>When I set an assignment, I often throw in a not-very-obvious-at-all situation for students to think about. The idea is to get them to think through what they have learned and how it applies in a real situation. In that way they can get to grips with how a piece of theory actually works and I can see whether their understanding is along the right tracks.
I good example of this came out of the second year quantum physics class. We had been talking about Heisenberg's uncertainty principle, (also see the nice movie here) and rather than set them a straightforward question along the lines of "if I measure the position of something to an uncertainty of xxx, what is the minimum uncertainty to which I can know the momentum?", which basically tests their ability to stuff numbers into a formula and use a calculator, I set this curlier problem:
"I'm driving along a straight road, and go from a 100 km/h speed zone into a 50 km/h zone. I don't slow down quickly enough and, unfortunately for me, there's a revenue-gathering traffic cop hiding behind a bush 20 metres into the 50 km/h zone. He clocks me doing 61 km/h. Being a physicist, I try to wriggle my way out with the following: 'By Heisenberg's uncertainty principle, it is impossible for you to know exactly where I am and my momentum at the same time. If you've clocked me at exactly 61 km/h, that means you have specified my momentum exactly, and therefore cannot say anything about where I was. So, I might still have been in the 100 km/h zone, in which case I wasn't speeding.' Is my argument valid?"
This got some interesting answers - nearly all said my argument had to be invalid, though some had difficulty in pinning down why. From my point of view, as the teacher, this question worked really well in flushing out exactly what the students made of the uncertainty principle, and identifying misconceptions. For example, several students incorrectly believed this was to do with relativity -that the principle only applies when things are moving close to the speed of light. Another popular thought was that the principle only applied to elementary particles, such as electrons. By seeing this incorrect thinking, I was able to talk about it in a subsequent lecture.
So this question, in fact, worked really well - helping not just the students but also me. Next time I teach this stuff I'll be wiser to what the students might make of it.
]]>
Not what I wanted to blog about. Halogen lights. You know the ones, the little light bulbs that are all the fashion. You fit lots of them snuggly into your ceiling to provide a nice even illumination of the room without obtrusive cables and fittings hanging from the ceiling.
They are nasty things, for three reasons. First, they ain't energy efficient. They might look hi-tec, but they rack up your electricity bill as fast as a few incandescent bulbs will - especially when you have lots of them.
Second, fitting snuggly into the ceiling means that you have further to reach to change them. Whereas I could do a hanging light fitting without even getting a chair (oh, the advantages of being tall) a chair is the minimum I need in our new house. For many of them, a step ladder is called for. Indeed, we have one light in the most stupidly located position imaginable. It is where the ceiling is the highest, on the upstairs landing right at the top of the stairs. To get to it, I need to put a step ladder at the top of the stairs and go to the top of the ladder and reach upwards. ACC would cringe at the prospect (and so do I). This light bulb is remaining unchanged.
And third, once you finally reach them, they are SO difficult to get in and out of the fitting. With a conventional bulb, you could grasp it and turn; but with this there is nowhere to grasp. To generate a torque on the bulb, which is what you need to do to turn it in the fitting, you need to use friction between your fingers and the bulb. But to stop your fingers sliding across the bulb, you need to press really hard. But also, because the bulb is small, you can generate very little torque. The torque, which is the total 'strength' of the turn, is given by the force you exert, times the radius of the circle that is being turned. A small radius means that you need apply a lot of force, which isn't easy to do through friction alone. You can experience a similar thing with opening jam jars. There are a variety of inventions to cope with stiff lids - mostly they work by making the lid effectively bigger - e.g. by gripping the lid and providing you with something larger to turn. That means the radius of the turning circle is bigger, so the force you need to exert to give the same torque is lower.
A couple of days ago I changed a couple of bulbs in a friend's house - he's unable to do them himself. One of them, in particular, took an age to get out of the socket. I left with skin missing from my fingertips, because of the friction involved. People with arthritis would have no chance.
I feel like I need to invent a device to grip these things and let you remove them more easily - unless of course someone's already invented it.
]]>