2012A&A...542A..18L

The β Pictoris system with its debris disk and a massive giant planet orbiting at ≃9AU represents an ideal laboratory for studying giant planet formation and evolution as well as planet-disk interactions. β Pic b can also help in testing brightness-mass relations at young ages. Other planets, yet undetected, may of course be present in the system. We aim at directly constraining the mass of β Pic b and at searching for additional jovian planets on orbits closer than typically 2AU. We used high-precision Harps data collected over eight years since 2003 to measure and analyse β Pic radial velocities. We show that the true mass of β Pic b is less than 10, 12, 15.5, 20, and 25M_Jup if orbiting at 8, 9, 10, 11, and 12AU, respectively. This is the first direct constraint on the mass of an imaged planet. The upper mass found is well in the range predicted by brightness-mass relations provided by current ``hot start'' models. We also exclude the presence of giant planets more massive than 2.5M_Jup with periods less than 100 days (hot Jupiters), more massive than 9M_Jup for periods in the range 100-500 days. In the 500-1000 day range, the detection limit is in the brown dwarf domain. Beyond the intrinsic interest for β Pic, these results show the possibilities of precise RV measurements of early-type, rapidly rotating stars.