Finding ways to relate the weight, size (generally mouth diameter) and nominal frequency of bells is useful, because it allows values to be estimated what data is not available. Some work on this was done by André Lehr, as reported in his 1986 paper. Other models have been developed and refined since by bellfounders and others. The large collection of bells in my database (getting towards 10,000 at a recent count) allows investigation of patterns across bells of many different dates, founders and tuning styles. I have done two pieces of work so far:
An investigation into models estimating the weight of bells
An investigation into models of scale of thickness (very relevant to issues of strike pitch).