What does an experimental particle physicist do?

[Mr rabbit]

Hi,

I'm about to finish my degree on Physics (this will be my last year). I have plans to do a PhD with a professor who works too at the LHCb experiment (CP violation), so I'd research on this topics. But I have doubts on the specifical work of an experimental particle physicist... only analyze data? Maybe I'm little lost on this issues, but it's the feeling that causes me.

For example: if you're researching on superconductivity you prepare the samples, control and keep and eye on the experiment, you take data, analyze it... you do many things. But in the sector of particle physics you only receive data and analyze it... I'm right? In a first look, although you work on topics that you're interested, this seems a little boring.

 

[BvU]

Ask the prof for an introduction to someone who can show you around over there. Particle physics is the cutting edge of human knowledge and yes, it takes a lot of sitting behind a screen. But there's also design, building, maintenance, etc. of detectors, of accelerators, electronics, superconducting, and so on. There's hardly a corner of science where particle physicists don't stumble around !

 

[Gigaz]

Few physics PhDs nowadays build their experiment from scratch. When you're working on superconductivity, chances are that you analyse other peoples samples at a commercial device which you don't understand nearly as well as you would like to. I think the best topics for a diverse environment are those which deal with nanotechnology close to application. When your system consists of like 10 different interacting materials, it doesn't really matter to have the best measurement device, because you cannot explain all the details anyways.

 

[Vanadium 50]

this seems a little boring.

That's fair. And it can be boring. Testing a thousand photomultipliers is not exciting. Monitoring 2560 low voltage power supplies is not exciting.

My question to you is if you find this boring, why get a doctorate in it?

 

[jtbell]

Testing a thousand photomultipliers is not exciting. Monitoring 2560 low voltage power supplies is not exciting.

Neither is monitoring the pressure and flow gauges on the gas-handling system of a Čerenkov detector, but I once spent a trip to Fermilab doing just that.

 

[Mr rabbit]

Thank you for the answers.

I like particle physics, and I like experimental physics. But for me, until now, the data analysis was the last part; before, you have to take good data, try to minimize the errors, etc. Now is the only part. You don't have to see the experiment, or find errors on the devices, or take good data.

But maybe I have a diffuse idea of what it really is.

 

[mfb]

It depends on the topic.

Looking for CP violation is purely physics data analysis, of course, but typically PhD students are also involved in other tasks. Simulations of current and future detector conditions, event and object reconstruction, detector calibration, monitoring the detector conditions, doing shifts in the control room, sometimes even physically assembling detector or infrastructure components, testing hardware for future upgrades, ...

LHCb plans a major upgrade in 2019-2020. If you start a PhD in Europe (after a MSc, starting with research) that would be the time frame of your work, if you start a PhD in the US (after a BSc, starting with coursework) you'll be a bit late for that, but then commissioning of the new detector components will be a big task.

 

[Mr rabbit]

Thank you for your answer. You motivated me to move on.

I mean, PhD is for 4-5 years. If you have to be 4 years analyzing data there is no problem. But when you already have a fixed position as a researcher and your job is always to analyze data ... I don't want to analyze data for the rest of my life.
But I like tasks like simulation and others that you mention too, so maybe it can be a good job

 

[mfb]

Well, what exactly do you call "data analysis"? Everything that is done on computers is some sort of analysis of data. If you go by this broad category, then most of the work is data analysis. Some of it has nothing to do with particle physics, e. g. software to monitor the temperature, humidity or whatever else in some detector part. These tasks are often not the most popular ones (they rarely lead to publications; in the best case everything works and everyone takes it for granted, in the worst case it does not work and it is your fault), but they have to get done as well.

I don't have a proper statistics, but as a very rough estimate, 50% the time is spent on data analysis for specific publications (e. g. "let's measure the cross section for this process"), 50% is spent on other tasks.

 

[Mr rabbit]

Well, for me the data analysis is to treat the data statistically. For example, you go to the lab and do an experiment on the photoelectric effect, take voltage and current data for some wavelengths and now you need to process the data to get conclusions. You can get the stopping potential by two or three methods and you can get Planck's constant, compare methods, measure the work function, etc. This is what I mean by data analysis. I don't know how this works in particle physics, but it can't be very different.

When you do a simulation, for example, you are not doing data analysis, you are actually generating data (like doing an experiment).

 

[mfb]

When you do a simulation, for example, you are not doing data analysis, you are actually generating data (like doing an experiment).

Well...
You install some software, figure out how to configure it, plug in what you want to simulate, run it, repeat the last three steps several times until it actually does what you want[/size]... and then you analyze the simulation result. If it is some detector part: How many tracks are found under these conditions, or how good is the energy resolution, or whatever. If it is a simulation for some physics analysis, you do a simplified version of this analysis to estimate how well some property can be measured in the future.