Can one use thermal expansion to "passively" realign solar panels (whether they are used for photo voltaic production or direct heat production)?
BACKGROUND
The efficiency of Solar panels depends on their alignment with the sun's position in the sky. Keeping this alignment is costly, whether done by hand or via an electronic device. Could thermal expansion be used to transform heat in movement and automatically rotate a solar panel when slightly misaligned, and hence progressively adjusting the reflectors' position during a day of uninterrupted sun exposure?
I would like to study theoretically the potential of the thermal expansion of metal (e.g. Aluminum, Linear coefficient 23.1), Polymer (e.g. Polypropylene, 150) and liquids (Gasoline, 317) to generate enough movement to realign a set of reflectors beyond a "clicker" threshold (so that the solar panel does not move back once the material cools down), so that the material used for the thermal expansion can cool down and be used again at the next misalignment. And then, maybe, to validate such theoretical work using a small scale set of reflector illuminating a photovoltaic sensor, used to monitor the process of realignment with the sun during the day.
But the idea seems a bit too simple: if it worked, somebody would have done it already, and if nobody did it, it could be because one can easily demonstrate that the forces produced by thermal expansion are insufficient to generate sufficient movement. Yet searching for "solar panels realignment" did not yield any useful results, hence the question here, in order to avoid "reinventing the wheel"...
