A comprehensive modelling effort to analyse sputtering erosion and redeposition in the DIII-D/DiMES 70, 71 and 79 experiments using metal films (beryllium, vanadium, molybdenum, tungsten) on a carbon divertor probe has been performed. These materials were exposed at the outer strike point of an attached H mode plasma with peak at Te ∼75 eV. The analysis uses coupled impurity transport (REDEP, WBC) and related codes with inputs of measured plasma parameters. Thecode output was compared with measured erosion and redeposition profiles, and other data. The predicted redeposition profiles for beryllium, carbon, molybdenum and tungsten agree well with the data. Beryllium and carbon exhibit longer transport distances compared with those of high-Z metals - primarily due to longer mean free paths for sputtered atom ionization. Photon emission calculations for beryllium also compare well with the data, thereby tending to validate coupled models for plasma parameters, impurity transport and atomic data. For vanadium the comparison of the code with the data varies from poor to fair depending on the ionization model used. For most metal films, the absolute erosion is less than that predicted for the `pure' metal, a fact we attribute to the effect of a carbon overlay or mixture. For carbon the predicted peak net erosion rate (∼4 nm/s) is approximately 5 times less than the gross rate, and this is confirmed by the data. Carbon net erosion and core contamination result primarily from the quiescent (intra-ELM) period oblique incidence deuterium physical sputtering and self-sputtering. ELMperiod sputtering and chemical erosion appear to play a small role in net erosion in these plasmas.