Young, active radio stars in the ABDoradus moving group

Context. Precise determination of stellar masses is necessary to test the validity of pre-main-sequence (PMS) stellar evolutionary models, whose predictions are in disagreement with measurements for masses below 1.2 M-circle dot. To improve such a test, and based on our previous studies, we selected the ABDoradus moving group (AB Dor-MG) as the best-suited association on which to apply radio-based high-precision astrometric techniques to study binary systems. Aims. We seek to determine precise estimates of the masses of a set of stars belonging to the ABDor-MG using radio and infrared observations. Methods. We observed in phase-reference mode with the Very Large Array (VLA) at 5 GHz and with the European VLBI Network (EVN) at 8.4 GHz the stars HD160934, EKDra, PWAnd, and LO Peg. We also observed some of these stars with the near-infrared CCD AstraLux camera at the Calar Alto observatory to complement the radio observations. Results. We determine model-independent dynamical masses of both components of the star HD160934, A and c, which are 0.70 +/- 0.07 M-circle dot and 0.45 +/- 0.04 M-circle dot, respectively. We revised the orbital parameters of EK Dra and we determine a sum of the masses of the system of 1.38 +/- 0.08 M-circle dot. We also explored the binarity of the stars LO Peg and PWAnd. Conclusions. We found observational evidence that PMS evolutionary models underpredict the mass of PMS stars by 10% 40%, as previously reported by other authors. We also inferred that the origin of the radio emission must be similar in all observed stars, that is, extreme magnetic activity of the stellar corona that triggers gyrosynchrotron emission from non-thermal, accelerated electrons