ME to Experimental or Theoretical Astrophysics?

[astroperson99]

Hi guys, I have a few questions that's been bugging me for a while and finally decided to post it in hopes to get some feedback.

Tech background: Mechanical engineer BS, will be getting MASc. in ME as well. Although my degree is ME, my focus in undergrad had been on CNC machines/CAD and my MASc. focus was on robotics and vision systems. So I feel that I personally have some decent knowledge on several fields in engineering.

Concern: I always aspired to become what I would call an astrophysicist from childhood through to the end of high school. However, reality gotten the best of my choice and I decided to make the logical choice of engineering (don't get me wrong, I don't mind it and I enjoy robotics quite a bit). Like many undergrads stressed during university, thoughts of jumping to astrophysics kept on creeping on me but after some consulting I decided it was best to finish what I started, so I completed engineering and even did a masters. But this annoying feeling keeps coming back.

So after some thought I decided to try astrophysics out and get it over with. I think I just need to see if I am really cut out for it, although it will waste some time but I only have one life. I've severed some relationships as a result of this decision.

After reading other posts of people in somewhat similar situations, I am not naive enough to believe I have what it takes to become a theoretical astrophysicist. I do plan on taking some courses for one semester to "start catching up" and then take my chances on the GREs this year. I want to take some courses first and consult with professors in the field to get a better understanding on what I'm getting into.

Question: Many sources have also told me that my chances for experimental astrophysics are much better and there are much better job prospects. Building telescopes/satellites sound pretty fun but it still doesn't address this nagging feeling of unfinished business with my childhood/teenage years. I was wondering if it is a better choice to aim for experimental physics, and use that as a stepping stone to get into theoretical (or not?) physics? Or should I just go straight to theoretical astrophysics and only try it out if I get into (1% chance, but heck it's still a chance!) a really good university?

The reason I am asking this is because I am your stereotypical theoretical physicist wannabe that wants to develop theories to make a warp drive/time machine and then construct/test it (star trek ref here). I am not foolish enough to believe that I can actually do those things, but I am deeply passionate about helping humanity with a means of easier/practical/cheap space travel (which was why I chose mech eng in the first place). After studying mech eng I feel that what really is needed now is a focus on unlocking the universe's deeper secrets (wormhole travel, etc) to obtain that dream, which is why theoretical physics is nudging me. This dream will probably not be achieved in my lifetime, but I still intend to try to help out.

I appreciate any feedback/insight or just a good smack in the head to knock some sense into me. I've been playing my life fairly cautiously, but now I want to take a risk. Is it a stupid move?

Well this turned into a really long post, so I apologize and thank you for reading this far!

 

[Pengwuino]

If you want to go into theoretical physics, you'll probably have to take a LOT of undergraduate physics courses and math courses. Since you're still in school, start taking some. Talk to the physics undergraduate adviser at your university and see what you can come up with. You're in a good position that you're still enrolled and are free to take courses still. If after taking a few courses and talking to the professors or TAs in the department you don't like it, well you still have the ability to make great contributions as a ME. If you like it, then doubly good. Mind you if you become a theoretical astrophysicist, the ME stuff will probably go by the wayside pretty quickly in my opinion...

 

[twofish-quant]

After reading other posts of people in somewhat similar situations, I am not naive enough to believe I have what it takes to become a theoretical astrophysicist.

You may. Most people think Stephen Hawking when they think theoretical astrophysicist, but that gives a very skewed view of what the field looks like.

I want to take some courses first and consult with professors in the field to get a better understanding on what I'm getting into.

Also do get some undergraduate research and start programming. Learn C++.

Many sources have also told me that my chances for experimental astrophysics are much better and there are much better job prospects.

I don't see this as being true. What happens is that most theoretical astrophysics is heavily computational, so people end up being computer programmers of one sort or
another.

I was wondering if it is a better choice to aim for experimental physics, and use that as a stepping stone to get into theoretical (or not?) physics?

I think it's better just to get some background in both and then figure out which one you like better. In some situations, it's not so much the field, but the people that you meet. Also having a mechanical engineering background is *VERY* useful in experimental work, since much of what you will be doing is designing and building instruments.

The reason I am asking this is because I am your stereotypical theoretical physicist wannabe that wants to develop theories to make a warp drive/time machine and then construct/test it (star trek ref here).

Sure. And then after a year of graduate school, you'll figure out why that doesn't work, so you'll end up writing computer code to simulate magnetic fields or dust flows. One thing about theoretical astrophysics is that it's more like engineering than people like Hawking make it seem.

After studying mech eng I feel that what really is needed now is a focus on unlocking the universe's deeper secrets (wormhole travel, etc) to obtain that dream, which is why theoretical physics is nudging me.

You'll find a lot of deep secrets pretty close at hand. Let me give you an example of a problem that is more typical of what most theorists do than quantum gravity. Look at a fireplace. You have all sorts of energy and complex transfers going on here. Try modelling it. You'll find it to be a *hard* problem. There are lots of deep secrets in how to model a fireplace that we don't understand, and a lot of that is applicable to the early universe and galaxy formation.

Once you figure out how to model turbulence in gaseous flows, then you can end up building a better rocket engine, or gas engine.

Personally, I find problems like "what happened before the big bang" to be uninteresting. They are uninteresting because we don't have much data, we don't have much in the way of theory constraints, so that you can make up anything. Now model a chimney (or a galactic nebula that behaves like a chimney). It's more interesting to me since you *can't* make stuff up.

I appreciate any feedback/insight or just a good smack in the head to knock some sense into me. I've been playing my life fairly cautiously, but now I want to take a risk. Is it a stupid move?

Just get involved in some research and see if you like it. If you do great, if you don't then do something else. :-) :-)

There are huge risks involved in not trying anything new.

 

[astroperson99]

Thank you guys very much for the replies. I am glad that I've posted this as I am gaining better insight into my situation.

Pengwuino: I plan on shaving off some of my savings and take some physics courses for one semester. On the side I will talk to some professors/graduates to get a better understanding on what it is actually like being a theoretical physicist.

Interestingly, much of my background involved a lot of programming (robotics/CNC machines and controls/vision systems). I can confidently say that I have a strong background in programming, knowing languages such as Matlab and c/c++. I heard from others that fortran is also important for simulations, as they are more optimized for for-loops.

This leads to my next question. From what I gathered from my previous discussions, research and your replies, it feels like theoretical astrophysicist graduate studies are:

1) Develop/Improve a model in a specific cosmological process (star formation, big bang model, etc...)
2) Simulate and wait (often times, wait really long for it to complete)
3) Examine results and compare with observed findings (telescope/spectrometers/etc)
4) Publish paper on findings, repeat step 1)

Am I wrong to think this? It is a similar process for any graduate work really, but for some strange reason I thought some theoretical physicists can get away with just doing step 1) and skipping to step 4), if given enough proofs. Then someone else can do the simulations and examine results.

As I am writing this, I am beginning to feel that I have a very skewed idea on what it is like to be a theoretical astrophysicist (as twofish-quant pointed out). I grew up (like many) admiring the big name physicists (Einstein, Faraday, Newton, Kepler (philosopher), to name a few), and I always thought that theoretical physicists are people who would have those Eureka moments, jot down a bunch of complicated equations and then revolutionize our way of thinking. The reason why they're so great, for me, is that those equations are then used in many technological advances, improving our way of life and even got us closer to the stars. For example, general relativity is used in GPS navigation systems. For this reason I imagined theoretical physicists to be more like people who examine relationships of physical phenomena and the implications from their findings eventually leads to practical applications. If this is the case then to me this is a worth while career to pursue.

In your opinion, would it be a better idea to learn the theory and come up with my own ideas and then build a test apparatus to test them, while having a job in engineering? I love tinkering/building with things and wouldn't mind the building process. However I think that physics has gone so far, that if I wanted to test anything to improve our understanding of physical phenomena it will have to be very expensive, like the LHC for example. But that is just an unrelated concern that I have...

Maybe I should seriously rethink my perception in theoretical physics before wasting other people's time and patience. I still intend to take courses and ask around more, but the more feedback the merrier. =P

Once again, thanks so much for the feedback, I am getting a better understanding of my situation and I am much much much less confused.

 

[nlsherrill]

Hi guys, I have a few questions that's been bugging me for a while and finally decided to post it in hopes to get some feedback.

Tech background: Mechanical engineer BS, will be getting MASc. in ME as well. Although my degree is ME, my focus in undergrad had been on CNC machines/CAD and my MASc. focus was on robotics and vision systems. So I feel that I personally have some decent knowledge on several fields in engineering.

Concern: I always aspired to become what I would call an astrophysicist from childhood through to the end of high school. However, reality gotten the best of my choice and I decided to make the logical choice of engineering (don't get me wrong, I don't mind it and I enjoy robotics quite a bit). Like many undergrads stressed during university, thoughts of jumping to astrophysics kept on creeping on me but after some consulting I decided it was best to finish what I started, so I completed engineering and even did a masters. But this annoying feeling keeps coming back.

So after some thought I decided to try astrophysics out and get it over with. I think I just need to see if I am really cut out for it, although it will waste some time but I only have one life. I've severed some relationships as a result of this decision.

After reading other posts of people in somewhat similar situations, I am not naive enough to believe I have what it takes to become a theoretical astrophysicist. I do plan on taking some courses for one semester to "start catching up" and then take my chances on the GREs this year. I want to take some courses first and consult with professors in the field to get a better understanding on what I'm getting into.

Question: Many sources have also told me that my chances for experimental astrophysics are much better and there are much better job prospects. Building telescopes/satellites sound pretty fun but it still doesn't address this nagging feeling of unfinished business with my childhood/teenage years. I was wondering if it is a better choice to aim for experimental physics, and use that as a stepping stone to get into theoretical (or not?) physics? Or should I just go straight to theoretical astrophysics and only try it out if I get into (1% chance, but heck it's still a chance!) a really good university?

The reason I am asking this is because I am your stereotypical theoretical physicist wannabe that wants to develop theories to make a warp drive/time machine and then construct/test it (star trek ref here). I am not foolish enough to believe that I can actually do those things, but I am deeply passionate about helping humanity with a means of easier/practical/cheap space travel (which was why I chose mech eng in the first place). After studying mech eng I feel that what really is needed now is a focus on unlocking the universe's deeper secrets (wormhole travel, etc) to obtain that dream, which is why theoretical physics is nudging me. This dream will probably not be achieved in my lifetime, but I still intend to try to help out.

I appreciate any feedback/insight or just a good smack in the head to knock some sense into me. I've been playing my life fairly cautiously, but now I want to take a risk. Is it a stupid move?

Well this turned into a really long post, so I apologize and thank you for reading this far!

I think you should definitely go for it. Although I am not one with much experience(2nd year physics/math major), I am sure if you can get a masters in ME you can do astrophysics. Astro is really a big area, and like twofish said, its highly computational. Most of the astrophysicists at my university are computational and model accretion disks, gamma ray bursts, and supernovas. I doubt you would have to learn THAT much extra math, but I do think astrophysics will be more math intensive than ME. If its been a passion of yours for this long, then I absolutely think you should do it. Like you said, you only live once!

Also, I would not be so intimidated by the field. Sure, its really hard compared to most things, but I would never underestimate your ability. I thought coming in as a freshman in physics I would be going up against "genius's." I have given it my all and have so far made an A in all my physics and math courses. I don't feel like I have a natural inclination for this area either, I just put a looooot of time into it.

Do it.

 

[astroperson99]

I think you should definitely go for it. Although I am not one with much experience(2nd year physics/math major), I am sure if you can get a masters in ME you can do astrophysics. Astro is really a big area, and like twofish said, its highly computational. Most of the astrophysicists at my university are computational and model accretion disks, gamma ray bursts, and supernovas. I doubt you would have to learn THAT much extra math, but I do think astrophysics will be more math intensive than ME. If its been a passion of yours for this long, then I absolutely think you should do it. Like you said, you only live once!

Also, I would not be so intimidated by the field. Sure, its really hard compared to most things, but I would never underestimate your ability. I thought coming in as a freshman in physics I would be going up against "genius's." I have given it my all and have so far made an A in all my physics and math courses. I don't feel like I have a natural inclination for this area either, I just put a looooot of time into it.

Do it.

Thank you for the encouragement! From experience, effort always pays off in the long run, so don't get troubled if you don't get the results you want even after a good amount of effort. =)

I enjoy math and proofs, but engineering has changed my mindset on the use of it in general. It's a lot less about proofs and much more on using equations to apply to an application.

 

[twofish-quant]

I heard from others that fortran is also important for simulations, as they are more optimized for for-loops.

C++ is superior to FORTRAN for this. There is something called template meta-programming which allows C++ compilers to optimize inner loops. All of the new simulation programming is done in C++.

1) Develop/Improve a model in a specific cosmological process (star formation, big bang model, etc...)
2) Simulate and wait (often times, wait really long for it to complete)

It's more like take code. Run a test. Watch the test work badly. Debug. Run a test. Watch the test fail. Debug. Repeat.

3) Examine results and compare with observed findings (telescope/spectrometers/etc)

Not really. It's more like "I ran a simulation with these assumptions and got these results. Discuss amongst yourself what that means." Often you run the simulation before anyone has any observational results, and your simulation tells the people building the instruments what to look for.

Also, you have the "I ran a simulation with these assumptions, and it didn't work, there is something wrong with these assumptions." or "I came up with this new technique which works great on this test problem. Please try this on your own simulations."

4) Publish paper on findings, repeat step 1)

Upload paper onto Los Alamos. No one reads journals to get the latest research, since it can take a few months to get something published at which point it's outdated. Also, the paper is a final product. There is a lot of proofreading before you publish.

Also publishing is only one way of communicating results. Talking to someone at a conference is useful, because they can tell you things that are useful, but aren't publishable. For example, if you think that person X has results that are non-sense and they don't really know what they are doing, this is the type of stuff you say over drinks in a hotel bar. Also, if you suspect that person X's results are non-sense but aren't sure, this is also stuff you say over drinks at a conference.

It is a similar process for any graduate work really, but for some strange reason I thought some theoretical physicists can get away with just doing step 1) and skipping to step 4), if given enough proofs. Then someone else can do the simulations and examine results.

Not really. There wouldn't be enough "meat" for a paper.

I grew up (like many) admiring the big name physicists (Einstein, Faraday, Newton, Kepler (philosopher), to name a few), and I always thought that theoretical physicists are people who would have those Eureka moments, jot down a bunch of complicated equations and then revolutionize our way of thinking.

Actually...

You have a lot of Eureka moments. It's just that most of them turn out to be false alarms. You have this wonderful moment in which you think you know what's going on. You then run some tests and then read some papers, and bummer, doesn't work. But what about?? Drat. But maybe? Nope. Umm what about? Darn.

Also, there is a lot of mini-Eureka moments. You figure out a better way of coding three lines that saves you 10% of run time, or you figure out why this function is unstable.

Looking at things from the end is misleading because we know what worked, so you don't see the 100 things that didn't. What's fun sometimes is to read some old papers and watch as people slowly and painfully realize something. Also part of research is listing what didn't work. If you spend three years and come up with a list of five clever things that don't work, that's progress.

The other thing is often the person that gets the big credit is the person that puts the last piece in. But he or she couldn't have done this without lots of people putting other pieces there.

For example, general relativity is used in GPS navigation systems. For this reason I imagined theoretical physicists to be more like people who examine relationships of physical phenomena and the implications from their findings eventually leads to practical applications. If this is the case then to me this is a worth while career to pursue.

Sure. But GR is isn't the only thing that results in major advances. Magnetic fields. You do realize that we really don't know how the Earth's magnetic field works. We have a model that is a good guess, but we really don't know.

Take a match and light it. There are tons of mysteries in what happens next. I light a match, and then put a thermometer two inches above that. Now I want you to write a set of equations that describes the temperature on the thermometer. You could spend *years* working on that problem, and people do.

Also a lot of it involves making things simple. GR is an incredibly rich and complex theory. It's too rich and complex. One thing that you need in order to do GPS is to come up with a simplification that works well enough so that you can compute it with minimal computing power.

In your opinion, would it be a better idea to learn the theory and come up with my own ideas and then build a test apparatus to test them, while having a job in engineering?

The problem is that science is a social activity. You have to surround yourself with people that are interested in some small topic (like burning matches) so that you know what people are thinking about. The problem with the idea of the "lone genius" is that you really can't get that much done without talking to other people.

Also people specialize. People find some small niche in which they are really good at, and talk to people that are in other niches. Some people are expert tool builders.

I love tinkering/building with things and wouldn't mind the building process. However I think that physics has gone so far, that if I wanted to test anything to improve our understanding of physical phenomena it will have to be very expensive, like the LHC for example. But that is just an unrelated concern that I have...

If you want to do particle physics, that's expensive. If you want to do something in combustion physics, that's less expensive. One reason I didn't to into particle physics, is that the experiments are so big and expensive that there isn't room for the "crazy artist."

Now, if you want to be the world's expert on using GPU's to compute cellular automata... Or if you could figure out a clever way of using cellular automata to figure out how much heat transfer you can get with the match experiment...

Maybe I should seriously rethink my perception in theoretical physics before wasting other people's time and patience.

You might find (as I did) that it's really more fun than you first thought. Part of the thing that I find interesting is that you see the world with different eyes. Think about dust. Have you ever thought about dust? Probably not, but just think about it for a moment. There are people that spend their entire careers thinking about dust. Once you realize that this happens, and you realize how interesting and useful thinking about dust is, that just changes how you see the world.

 

[astroperson99]

C++ is superior to FORTRAN for this. There is something called template meta-programming which allows C++ compilers to optimize inner loops. All of the new simulation programming is done in C++.

Thanks for the correction, I shall correct my colleague next time I see him.

It's more like take code. Run a test. Watch the test work badly. Debug. Run a test. Watch the test fail. Debug. Repeat.

Ah yes, how did I miss that inner loop of hair ripping madness. I guess the head-banging-on-table-repeatedly technique is universal amongst cs, engineers and physicists!

Not really. It's more like "I ran a simulation with these assumptions and got these results. Discuss amongst yourself what that means." Often you run the simulation before anyone has any observational results, and your simulation tells the people building the instruments what to look for.

Also, you have the "I ran a simulation with these assumptions, and it didn't work, there is something wrong with these assumptions." or "I came up with this new technique which works great on this test problem. Please try this on your own simulations."

Yup, sounds like my experience alright. Except, add "Hey, look at it working in this specially designed/controlled environment where I know it will work and look cool. Just look at those flashy pictures/videos! Give me an applause."

Upload paper onto Los Alamos. No one reads journals to get the latest research, since it can take a few months to get something published at which point it's outdated. Also, the paper is a final product. There is a lot of proofreading before you publish.

Also publishing is only one way of communicating results. Talking to someone at a conference is useful, because they can tell you things that are useful, but aren't publishable. For example, if you think that person X has results that are non-sense and they don't really know what they are doing, this is the type of stuff you say over drinks in a hotel bar. Also, if you suspect that person X's results are non-sense but aren't sure, this is also stuff you say over drinks at a conference.

I agree entirely. The best research is definitely socializing.

Actually...

You have a lot of Eureka moments. It's just that most of them turn out to be false alarms. You have this wonderful moment in which you think you know what's going on. You then run some tests and then read some papers, and bummer, doesn't work. But what about?? Drat. But maybe? Nope. Umm what about? Darn.

Also, there is a lot of mini-Eureka moments. You figure out a better way of coding three lines that saves you 10% of run time, or you figure out why this function is unstable.

Looking at things from the end is misleading because we know what worked, so you don't see the 100 things that didn't. What's fun sometimes is to read some old papers and watch as people slowly and painfully realize something. Also part of research is listing what didn't work. If you spend three years and come up with a list of five clever things that don't work, that's progress.

The other thing is often the person that gets the big credit is the person that puts the last piece in. But he or she couldn't have done this without lots of people putting other pieces there.

I guess I've cheapened the efforts of many others which allowed those who come to those conclusions. It's actually a really good way of looking at things, I should remember not to forget this philosophy again. Thanks for the insight. =P

You do realize that we really don't know how the Earth's magnetic field works.
...
Take a match and light it. There are tons of mysteries in what happens next. ... Also a lot of it involves making things simple...

I always thought they knew how magnetic fields work actually. But then again, I guess we are generally working our understanding from top down and not down up. Maybe there's still many small experiments we can do at home, which for me sounds pretty exciting. I gather that your message is "Small and seemingly insignificant are actually really significant." Or rather, everything is significant. And that there still much that we don't know about all sorts of things.

The problem is that science is a social activity. You have to surround yourself with people that are interested in some small topic (like burning matches) so that you know what people are thinking about. The problem with the idea of the "lone genius" is that you really can't get that much done without talking to other people.

Also people specialize. People find some small niche in which they are really good at, and talk to people that are in other niches. Some people are expert tool builders.

You are right on that one, I always had trouble finding people to talk to (in real life) who were also interested in astrophysics. Talking to people who's interested in the same niche really helps with motivation and learning.

Now, if you want to be the world's expert on using GPU's to compute cellular automata... Or if you could figure out a clever way of using cellular automata to figure out how much heat transfer you can get with the match experiment...

I always found cellular automata and fractals really fascinating. I guess you can say those can classify as cheap experiments.

You might find (as I did) that it's really more fun than you first thought. Part of the thing that I find interesting is that you see the world with different eyes. Think about dust. Have you ever thought about dust? Probably not, but just think about it for a moment. There are people that spend their entire careers thinking about dust. Once you realize that this happens, and you realize how interesting and useful thinking about dust is, that just changes how you see the world.

Actually, I do take breaks to go to the library just to read up on some space related books. Once I remember reading up on atmospheres of different planets, and when I got out of the library I found myself feeling like I'm in a sea of air. It was definitely a different feel altogether. It was very interesting indeed. =P

Thanks for the post, I definitely got much more wisdom out of it, as well as a very essential smack on the head about my way of looking at things as well.