GLOBULAR-CLUSTERS IN THE MAGELLANIC CLOUDS .2. IR-ARRAY PHOTOMETRY FOR 12 GLOBULAR-CLUSTERS AND CONTRIBUTIONS TO THE INTEGRATED CLUSTER LIGHT

We report JHK results of observations of 12 globular clusters in the Large Magellanic Cloud (LMC), and present colour-magnitude diagrams down to K=16 (corresponding to M(K) similar to-2.6) for similar to 450 stars in these clusters. We merge our data with BV photometry for 11 LMC clusters, previously published in Paper I of this series, and use the merged data to study the evolution of integrated magnitudes and colours of simple stellar populations (SSPs), which are samples of coeval and chemically homogeneous stars. In particular, we examine the effect of phase transitions (ph-ts), which signal the appearance of the RGB or AGE in SSPs of increasing age. We find that the AGE contributes similar to 60 per cent of the integrated cluster light at K, while the contribution from the bright RGB stars (i.e., K-0<14.3, log L/L.similar to 2.66) is correlated with the s-parameter (Elson and Fall) ranging from similar to 0 per cent for s=0 up to similar to 20 per cent for s>35. The age at which the RGB ph-t actually takes place (i.e., the calibration of s with age) depends on the details of stellar evolutionary models. In 'classical' models (those without overshooting), the RGB ph-t occurs at similar to(6+/-2)x10(8) yr and lasts for 2.9x10(8) yr. In models with overshooting, the occurrence of the RGB ph-t is later [at similar to(1.5 +/- 0.3)x10(9) yr] and the duration is longer (4.3x10(8) yr). While the age and duration of the RGB ph-t depend on the treatment of mixing, both classical and overshooting models yield the same fractional contribution of RGB stars to the integrated cluster light before and after the RGB ph-t, in agreement with the Fuel Consumption Theorem (Renzini and Buzzoni). We report extensive experiments which show that the variations of the integrated colours of the LMC clusters from s=31 to 43 are controlled by the complex interplay of various factors, different from colour to colour and frequently dominated by the stochastic noise induced by a few very bright objects. The overall picture that emerges is consistent with the early conclusions drawn by Persson et al. and Frogel et al. that the J-K colour is mostly driven by the AGE stars, that V-K is substantially controlled by AGE and RGB stars (AGE stars being slightly more important), and that B-V is partially influenced by the whole population of red stars brighter than the bulk of the RGB clump, but is also quite strongly dependent on the progressive fading and reddening of the turn-off stars due to age increase.