198
C. Tang (唐晨宇) and Y. Wang (王延颋)
5.1.1 Computational Physics: A Bridge Connecting Theories
and Experiments
Since the beginning of the usage of computers in science, computational physics has
become an important branch of physics study. Before the appearance of computational physics, approximations usually had to be taken when theoretical studies were
conducted to investigate physical problems. Such approximations often came with
inevitable disadvantages that most of the time the consequence of making approximations could not be estimated ahead of time. On the other hand, experimental results are
usually too comprehensive to establish the causality among various factors. Therefore, it was a common practice that a large gap between theory and experiment existed
when investigating a certain physical system.
The emergence of computational physics provides contemporary researchers a
brand new method in predicting the properties of many systems in ways that have
never been done before. It fills in the gap between theory and experiment by generating accurate (in the sense that the underlying model is good) numerical results
while keeping input variables controllable. Current researches have been following
a certain route in terms of using computational method. The theoretical works often
offer a model to specific systems, which renders certain directions for computations
to follow, whereas the results of computational works feed theories with solutions that
may not be easily found without numerical calculations. The computations further
play a significant role in advising experimentalists in conducting related experiments, whose work in return marks parameters for computational physics, which are
essential in narrowing the gap between simulated systems and the real ones.
There are certain typical functions of computational physics used under most
circumstances among studies. One of them is to numerically solve analytical equations. In some cases, it triggers much difficulty to solve equations with analytical
approach. It is thus unwise and sometimes impossible to solve most sets of equations,
no matter differential or integral ones, without numerical calculations. Bringing in
computational methods as an effective way to provide sets of numerical solutions are
of much necessity. Another major function resides in the simulation of many-body
problems to obtain more realistic results. The many-body systems usually require sets
of equations with too many degrees of freedom for physicists and mathematicians to
draw analytical solutions. Although their approximations, often using perturbation
or series expansion methods, have been remarkably successful in depicting simple
systems, they lead to unacceptable deviations in most classical or quantum manybody systems. In avoiding such deviations, it is wise to use computational methods
that offer much more precise results with iterative algorithms in due systems.
Moreover, there are certain systems that mere experiments may be insufficient
to describe. Some of the experiments should be conducted under extreme conditions that is of overwhelming difficulty in reality. Computational physics therefore
becomes crucially needed because manipulating simulated systems renders much
more freedom than doing so under real laboratorial environments. Such freedom
C. Tang (唐晨宇) and Y. Wang (王延颋)
5.1.1 Computational Physics: A Bridge Connecting Theories
and Experiments
Since the beginning of the usage of computers in science, computational physics has
become an important branch of physics study. Before the appearance of computational physics, approximations usually had to be taken when theoretical studies were
conducted to investigate physical problems. Such approximations often came with
inevitable disadvantages that most of the time the consequence of making approximations could not be estimated ahead of time. On the other hand, experimental results are
usually too comprehensive to establish the causality among various factors. Therefore, it was a common practice that a large gap between theory and experiment existed
when investigating a certain physical system.
The emergence of computational physics provides contemporary researchers a
brand new method in predicting the properties of many systems in ways that have
never been done before. It fills in the gap between theory and experiment by generating accurate (in the sense that the underlying model is good) numerical results
while keeping input variables controllable. Current researches have been following
a certain route in terms of using computational method. The theoretical works often
offer a model to specific systems, which renders certain directions for computations
to follow, whereas the results of computational works feed theories with solutions that
may not be easily found without numerical calculations. The computations further
play a significant role in advising experimentalists in conducting related experiments, whose work in return marks parameters for computational physics, which are
essential in narrowing the gap between simulated systems and the real ones.
There are certain typical functions of computational physics used under most
circumstances among studies. One of them is to numerically solve analytical equations. In some cases, it triggers much difficulty to solve equations with analytical
approach. It is thus unwise and sometimes impossible to solve most sets of equations,
no matter differential or integral ones, without numerical calculations. Bringing in
computational methods as an effective way to provide sets of numerical solutions are
of much necessity. Another major function resides in the simulation of many-body
problems to obtain more realistic results. The many-body systems usually require sets
of equations with too many degrees of freedom for physicists and mathematicians to
draw analytical solutions. Although their approximations, often using perturbation
or series expansion methods, have been remarkably successful in depicting simple
systems, they lead to unacceptable deviations in most classical or quantum manybody systems. In avoiding such deviations, it is wise to use computational methods
that offer much more precise results with iterative algorithms in due systems.
Moreover, there are certain systems that mere experiments may be insufficient
to describe. Some of the experiments should be conducted under extreme conditions that is of overwhelming difficulty in reality. Computational physics therefore
becomes crucially needed because manipulating simulated systems renders much
more freedom than doing so under real laboratorial environments. Such freedom