Evolution of protostars accreting mass at very high rates: Is Orion IRc2 a huge protostar?

The recent near-infrared spectroscopy of scattered light from Orion IRc2 suggests that the illuminating source at the K' band is either a protostar with a radius greater than or similar to 300 R-circle dot or a disk with an accretion rate approximate to 10(-2) M-circle dot yr(-1). To test the former interpretation, we present a simplified stellar model accreting mass at a very high rate, approximate to 10(-2) M-circle dot yr(-1). We find that the protostar is fully convective at almost all stages of the stellar mass M less than or similar to 15 M-circle dot, and thus a polytrope of index 1.5 is a good approximation of the stellar structure. The maximum radius less than or similar to 30 R-circle dot is attained at M similar to 7 M-circle dot. The shell deuterium burning, which would occur afterward cannot greatly blow up the protostar because the energy released by deuterium burning is small and also because the protostar is already shrinking rapidly. The only remaining possibility to make a huge protostar resides in the rotation of the surface layer almost at its breakup velocity. On the other hand, we find no difficulty in the alternative interpretation that the illuminating source is the accretion disk. In this case we predict that the 2.3 mu m CO absorption lines should be observed with a width similar to 50 km s(-1) due to the Keplerian motion in the disk. The accretion rate as high as 10(-2) M(circle dot)yr(-1) is compatible with the velocity dispersion in the Orion KL molecular cloud core. Because the luminosity of IRc2 is dominated by accretion, the protostellar mass is overestimated if the observed luminosity is regarded as intrinsic. Because the K'-band luminosity is emitted in the disk region far from the protostellar surface, the total accretion luminosity must be significantly higher than the observed K'-band luminosity.