On the interpretation of the Sandage period-shift effect among globular-cluster RR Lyrae variables
A new grid of canonical zero-age horizontal branch (ZAHB) models is constructed to study the Sandage period-shift effect, and a number of possible explanations for the discrepancy between the observed and predicted period shifts are explored. Various factors which determine the period shift in ZAHB models are discussed in detail, showing that possible uncertainties in the input physics used for computing the preceding red-giant evolution are unlikely to resolve the discrepancy. Some test calculations for the ZAHB phase are presented which suggest than an increase in the metal opacity at temperatures around a million K might substantially increase the predicted period shift. It is found that a nonsolar CNO to Fe ratio is not a plausible explanation of the period-shift discrepancy, and the effects of evolution off the ZAHB in the period-effective temperature diagram are discussed. It is shown that evolution can increase the predicted period shift but not to the extent required by the Sandage effect. The possibility that rotation might explain the discrepancy is considered and rejected.