Thought experiments - a didactic tool? I am wondering whether thought experiments (as from a didactic perspective) are being discussed in the context of teaching physics by using another term or notion? - since I think thought experiments (just as "real" experiments) are somehow a didactic tool and I am unable to find decent documentation or literature considering thought experiments from a didactic perspective.
Thought experiments do appear in didactic literature but rarely! Am I missing something by using the wrong term for it?
 A: I take a thought-experiment to be the application of theory to a particular set-up that is easily visualised. So the logical consequences of the theory are worked out for that set–up.
An example would be applying the postulates of Special Relativity to flashes of light as logged by observers on the ground and in a moving train. Relativity of simultaneity and so on emerge in this special case.
The traditional way to arrive at the concept of entropy without microscopic analysis uses the idea of the cyclic engine, especially an ideal engine performing a Carnot cycle. We could say that the argument proceeds by means of thought experiments. Is the argument pedagogy? I would not class it as such, but as part of the fabric of thermodynamics. Against this view, perhaps, is the unease among some physicists with this time-honoured argument, and their desire to replace it with a more conventional sort of reasoning (e.g. the famous attempt of Caratheodory).
A thought experiment, then, is a powerful (and rigorous) device that may be put to didactic use. But a thought-experiment may well be  a stepping-stone towards a general formal argument not dependent on a particular set-up.
A: I am not aware of another term in the didactic of physics. But I would go as far as to say that the way most experiments are described in textbooks, they actually could be considered thought experiments - they are not carried out in reality and neglect all kinds of real physics. Turning this around, you could use most descriptions of "experiments" as "thought experiment". I would therefore expect that much of the didactic of examples applied verbatim to the didactic of thought experiments.
A: The phenomena of thought experiments include all experiments that are not actual real physically performed experiments. This is regardless of the use of written text, drawings, formulas or given values. More often than not, they deal with idealized conditions that could never be reproduced in practice. As such they form the bulk of didactic mass in physics, mainly because they are way cheaper than actual experiments.
Loosing connection to the fact that they still remain mere thought experiments is one of physics main hazards. This is primarily the result of these thought experiments being quite useful in practice, raising the notion they define practice, which they don't.
The expression "thought experiment" however, is commonly used for a much less defined thinking process, in which the persons participating are free to let go of laws and conventions in order to obtain understanding that is otherwise hard or even impossible to communicate. Certain parts of our physical reality simply not fitting the box we call physics.
Physics as a field of interest naturally tends towards the latter kind of 'thought experiments', whereas physics as a science seeks to define the conditions of thought experiments, preferringly allowing no freedom whatsoever.
The practice of teaching physics is a mixture of these two explanations of the term, in which the objective of the study determines the balance of their use.
The objective of thinking out of the box in physics however, will always remain to get whatever is found to fit inside the box.
