I'll look for a detail. We haven't done one that needed a sliding surface in a while. Usually, we can accommodate the translation movements by deformation of the elastomeric pad.
We usually bond the PTFE sheet to the top of an elastomeric pad that sits on the concrete bearing seat. The girder will have a steel sole plate welded onto the bottom flange with the mirror-finished stainless steel sheet welded to the sole plate for the PTFE sheet to slide against.
We have some bearings that make use of a 'self-lubricating' bronze plate that the steel slides on, but the those haven't worked well over the long-term, due to the dirt and debris that blows into them. In an indoor application, I would imagine they could work well.
I don't know what issues may arise with just letting the steel slide on steel. I suppose it would depend on how often it moves. The only time we'd use something like that is for energy dissipation during an earthquake, where it would be anticipated to slide maybe once or twice during its service life. I'd imagine steel on steel would produce sounds that are loud and very annoying when they slide, so it this connection is in a building where it would be sliding due to thermal movements or wind loads, I can't imagine that would be acceptable.
