How can an isolated system be defined and applied in different contexts?

[minhty96]

I have encountered 2 answers for this question.

1. In the natural sciences an isolated system is a physical system without any external exchange – neither matter nor energy can enter or exit, but can only move around inside. Truly isolated systems cannot exist in nature, other than possibly the universe itself, and they are thus hypothetical concepts only. It obeys, in particular, to the first of the conservation laws: its total energy - mass stays constant.

and

2. An isolated system is a system that is now acted upon by external forces; hence every force in the system has an equal and opposite action-reaction pair force within the system. So every impulse is balanced by an equal and opposite impulse, leading to conservation of momentum within a system.

Whats the best way to define it?

 

[Simon Bridge]

An isolated system is an idealization where there is no interaction between inside and outside - i.e. #1.

If every outside influence has an equal and opposite outside influence, at every point, then that is the same thing as saying there are no outside influences. However, there is a difference between two opposite forces that cancel and two opposite forces that pull you apart.

In practice we use this as a model for systems where the properties of interest are not influenced from outside the system, or the influence is very small.

i.e. a pendulum as often constructed in classrooms is not an isolated system - it's not even in an inertial reference frame - yet it compares well with a model worked out assuming it is isolated, on the scale of accuracy that is available to school stopwatch, rulers, and balances.

[edit]BTW: it is quite reasonable to call a system conditionally "isolated" - i.e. with respect to some particular influence. You can even leave the statement of condition to context.