Wormster,
Your mention of thermite reminds me of what some conspiracy theorists claimed about the 9/11 atrocity, i.e. traces of combustion products proved that thermite demolition charges had been used to bring the buildings down, thus proving that the whole thing was indeed a conspiracy.
But what they forgot that the steel frame of the building almost certainly had surface rust, which, together with burning aluminium from the aircraft, would indeed produce traces of thermite combustion products. It would be very surprising if it were not so.
Now I doubt that you will get a sustained thermite reaction initiating with casual contact between aluminium and rusty steel. The thermite mixture they use for welding on the railways is actually quite hard to ignite. But as you are suggesting, there does seem to be the potential for creating sparks in freak conditions. I assume that it would be a localised thermite reaction, where the frictional heat between the two surfaces causes tiny hot spots, then the aluminium ignites locally, intensifying the spark. Between identical metals, there would also be hot spots, but no thermite reaction to boost the temperature.
However, steel on its own is notoriously good at causing sparks, in conjunction with other substances commonly found everywhere, such as quartz particles, which may be found in the floor and/or roof of the mine, but it does not participate in a violently exothermic reaction, merely mechanical abrasion, so the spark energy will be less.
It is interesting that where explosives are concerned, steel tools are banned, and brass or wood is used.
But aluminium and steel in contact in a damp environment will cause electrolytic corrosion (the voltage generated is insufficient to ignite anything), and aluminium has bad fatigue properties, so it would be a very poor choice for structural use in a mine.
I think that the ban on aluminium may be for multiple reasons, not just the possibility of causing an explosion by rubbing against steel. Thin foil, in particular, may be able to ignite if it shorted an electrical circuit which was otherwise within the intrinsic safety limits, but I don't know if it has ever actually happened.
Strangely enough, I was doing some electronic design for an intrinsically safe requirement some time ago, and being new to that, I went right through the relevant standards, BS, ISO etc, and found lots of things that we had to satisfy, but no mention of aluminium and steel at all. I was working to the highest levels of the standards, because it seemed prudent (it was for use in what is effectively an explosives magazine, although the possible vapour was nowhere near as easy to ignite as petrol or firedamp) and it was perfectly permissible for the case of our equipment to be aluminium while the structure it is mounted on is steel, and possibly rusty.
I am thinking that it would be interesting to bang bits of aluminium and rusty steel together in a dark room to see if any visible sparks are produced. But I don't anticipate much success. Of course, in an explosive atmosphere, it does only need the one in a million attempts....
But there may well be more to this than we yet know.
Alan