nullScientific ResearchScientific Research郁明阳
浙江大学
聚变理论与模拟中心
*2013.10.16 BNUnull
什么是物理?
物事的道理
老郁*null
What?
Identify the problem
Why & How?
Understand the problemQuestions come from curiosity*null
The most important criterion for
research in basic science
Curiosity and Interest
Secondary criteria:
background knowledge and methods, personal relations, …, hard work.*nullInterest
can occur by itself, or develop when you are already working on a problem.*Method
There is no ‘correct’ or ‘best’ method!Method
There is no ‘correct’ or ‘best’ method!Most used and most effective:
TRIAL AND ERROR自古成功在尝试* “Trial and error”, or “guessing” “Trial and error”, or “guessing”The most natural process of learning is by trial and error.
It is also much more fun.
All babies learn this way!
Foreign languages are most effectively learnt this way!*null
LEARN HOW TO LEARN!
In the university, we learn how to learn.
What is gained should be much more!
举一而反三、闻一而知十 from guessing.
*School verses ResearchSchool verses Research In schools, carefully-prepared material are taught. Questions have clear answers.
In research, looking for questions,
i.e., new problems.
We don’t know what we are looking for!
Re-search = search and search again!*Learn broadly
Everything is worth learning
“博”士?Learn broadly
Everything is worth learning
“博”士?*Just learning is not sufficientJust learning is not sufficientUnderstanding is more important
But understanding usually comes
only after many unsuccessful trials!One gains much more knowledge
from the trials than from the books
and lectures! One also remembers them!*nullMost great scientific discoveries are by accident, (i.e., by luck!)
But not by 守株待兔
Luck is preparedness meeting opportunity
Napoleon
*What are physics problems?What are physics problems?*Physics also includesPhysics also includesTraffic flow
Medical theory
Molecular and atomic biology
Behavior of financial markets
Weather
Swarm effect of insects and birds
Social behavior
Psychology
... many more!*How to do research?
1.
Find a problem of your interestHow to do research?
1.
Find a problem of your interest*2.
Identify the
time and space scales
of the problem.2.
Identify the
time and space scales
of the problem.*
Identifying the time and space scales:
Different branches of physics are separated by
Identifying the time and space scales:
Different branches of physics are separated byVery large cosmos
Large planets
Normal daily life
Small micro, nano
Very small atoms, particlesDifferent problems are also separated by their scales. For each problem, again identify its time and space scales!*Finding a research problemFinding a research problemFrom your curiosity and interest. Sometimes guided by the professors, colleagues, seminars, papers, newspapers, and magazines, …
Nowadays it is easy: read the popular science magazines such as Physics Today, Physics World, etc. Information can also be found in Google, Wikipedia, etc. (But beware of errors!)*For a new problem
or research area,
very oftenFor a new problem
or research area,
very often
the first 80% of it is solved with
20% of the total human effort.
The remaining 20% of it requires
80% of the total effort.Working on “NOVEL” problems has a clear advantage!*How to read a paperHow to read a paperFind out what is new in the paper
Usually there is only one new idea, method, and/or result. Find it and describe it in a few words or lines.
Don’t go into the details!
Unless you need to use them, e.g., the experimental method, the derivation, etc. Most good papers are short. *nullTherefore:
Don’t closely follow the derivation of formulas in the papers unless you are really interested in or need them.
Try to derive the formulas independently if you are interested in them, or
if you suspect that they are wrong, or if there is a new trick.
If the paper is poorly written, don’t waste time on it!
Many published papers are trivial or wrong. Don’t blindly believe in them.
*nullAbout science research
Every problem is worth looking into!
Your selection should depend on your curiosity and interest.
This may be difficult in a company research group. However, the best companies and institutes do not have such restrictions.*null No problem is “completely solved”!
There is always something new to be discovered!
Therefore, always follow your interest and curiosity. *nullTry different approaches
Do not follow any specific method. Try different approaches.
Follow your intuition (直觉)!
*nullDon’t be afraid of asking questions!Before you ask a question, first try to answer it yourself and find out why you cannot answer it.
Then you should carefully formulate the question which you want to ask.In learning, you should have “thick face skin (厚脸皮)”*Don’t be afraid of making mistakesDon’t be afraid of making mistakes Don’t worry about wasting time on unimportant or unclear problems.
Most new results are not “clearly correct”.
Fact: you can understand more from your errors. *Old referee “report form” of
Physical Review LettersOld referee “report form” of
Physical Review Letters… please judge the manuscript regarding its
1) Impact on the specific field: very low ( ) ( ) ( ) ( ) (X) very high
2) Impact on physics research: very narrow ( ) ( ) ( ) ( ) (X) very broad
3) Degree of innovation: very low ( ) ( ) ( ) ( ) (X) very high
4) Validity: not valid ( ) ( ) ( ) ( ) (X) valid
5) Readability for a nonspecialist: not accessible ( ) ( ) ( ) ( ) (X) very accessible*New referee report form of PRLNew referee report form of PRL… Letters published in PRL must meet a high standard of importance and interest.
a) Please judge the importance of the paper to its specific field. Not important O O O O Very important
b) Please judge the broad interest of the paper, apart from its importance to its specific field, to a wide spectrum of physicists. Not interesting O O O O Very interesting
c) Please judge the validity of the paper. Probably not valid O O O O Probably valid*What Nature editors seekWhat Nature editors seekHigh degree of novelty or innovation
高度新奇或创新的主
Interesting to a broad range of readers
可引起广泛的读者感兴趣
Significant step forward
有重要的进步
Breakthrough in performance
有突破的
现
High impact in the field
有重大的影响
Important advance in scientific understanding that provides new directions for research
在科学了解上有重要的进展或提供新的研究方向
Data persuasively supports conclusions
有说服性的数据来支持结论Rachel Peichin Won
Chief Editor
Nature Photonics*nullFor both PRL and Nature:
correctness and usefulness
are not crucial !
Most new discoveries are of uncertain nature!*nullAlways listen to critics, and discuss with others.
Never believe that you are definitely correct!
Acknowledge your errors! Don’t be afraid of making mistakes (continued)
But…
*nullLook out for new results…
If your result differs from others or looks strange, it may be wrong, but it may also be new!
Check it carefully. Find out what is “wrong” before giving up!
Remember:
Most great discoveries are accidental!*nullDon’t be sad/angry if your work is wrong:
You have gained much knowledge and experience that will be useful in the future.
Psychology:
Never think that you have wasted time! Otherwise, the gained knowledge and experience will quickly fade!*nullAt the beginning, most of your trials may be unsuccessful.
But with each unsuccessful trial, you gain more experience.
After many trials, your guesses will improve. That is why many PhD research work were actually done in the final 6 months.
If you merely solve a well-posed problem given by your advisor, it is only a homework exercise!*nullThanks!*Some interesting physics problemsSome interesting physics problems*Discovery of a new materialDiscovery of a new materialA few years ago, a professor and his students at Harvard University
put a silicon chip in a gas
shoot it with ultrashort laser pulses
a careless student burnt the chip
Instead of replacing the chip,
they analyzed the burnt chip.
Discovered the blackest material ever known!
The new material has the best heat and light absorption properties. Besides of basic physics interest, it has very important applications in communications and sensor technology, etc.*Charged dust particles in a plasma
in the International Space Station (extremely small gravity)Charged dust particles in a plasma
in the International Space Station (extremely small gravity)Dust crystal structure with a void at center*nullSnow (H2O) flakesWhat causes the many (> 100) different forms of the snow-flake crystal structures? Still an unsolved problem!
The form depends sensitively on the temperature, and also on how the temperature changes.*nullvertically vibrating dishA simple experimentOSCILLONS*nullWave patterns formed by tiny balls in dishAlready noticed in the 17th century!*nullOscillons
in colloidal fluidsclay mix*nullPossible explanationBut most probably wrong!*nullFreak wave*Water spikeWater spikeNormal toiletGerman toiletA channel of lower pressure is created by the dropping!*Sand dunes interactionSand dunes interactionA sand dune can pass through another without much change!
Solitons?*Comet Hale-Bopp has 2 tailsComet Hale-Bopp has 2 tailsThe white one is the tail (aligned to the comet’s trajectory) consisting of melted ice particles, the blue one is photoionized gas (plasma) deflected by the solar-wind magnetic field.*Traffic and fluid flowTraffic and fluid flowRoad interaction浙大大门浙大大门*Taffic jam can also happen without accident or blockage!
Taffic jam can also happen without accident or blockage!
Wider roads are not useful for smoother flow, just like water pipes.*Physics is useful if you can apply it!*Physics is useful if you can apply it!nullThanks!*