I am digitizing hundreds of tapes in a small archive using a highly tuned Sony APR-5003V recorder for playback, and have to be extremely attentive to the behavior of the tape on the take-up reel. Sony supplied "Reel Platter Shims" made of thin rubber (0.75mm, 0.03") with every recorder, so that engineers can compensate for the difference in flange thickness between aluminum NAB reels and plastic reels. The goal, obviously, is to facilitate the development of a tape pack that is centered between the reel flanges. Since plastic is less strong than aluminum, the flanges are made thicker on 3–7" plastic reels, so these typically do not require the use of the shims. However, when playing tapes from the 1950s and 60s, I find that I need to study the way the tape pack is developing at the beginning of a transfer, as some old tapes contain a variety of stresses and may show a "coning" behavior that tends to create an asymmetrical tape pack (I have seen tapes cone so badly that they actually broke off an upper flange on a high-quality 3M 7" plastic reel!). Other tapes "pull" up or down as they build a tape pack. In every case, I squat down and study the spacing of the building tape pack on the take-up reel in order to decide whether to add, or remove a shim, and restart the playback. Note that all this variability is taking place on a mastering grade recorder that has been perfectly maintained, and on which the reel turntables themselves are precisely as they were set up by the factory. If there is a consistent problem with tape scraping a reel flange, there may be an issue with the tape path, and the relative position of components of the head assembly or tape guides in relation to the deck's reference plane.