Análisis espectroscópico cuantitativo de estrellas de tipo o en la región 30 doradus de la gran nube de magallanes

Supervised by:
  1. Artemio Herrero Davó Director
  2. Sergio Simón Díaz Director

Defence university: Universidad de La Laguna

Fecha de defensa: 26 May 2014

  1. John Beckman Chair
  2. Basilio Ruiz Cobo Secretary
  3. Francisco Najarro de la Parra Committee member
  1. Astrofísica

Type: Thesis

Teseo: 363558 DIALNET


This thesis focuses on the quantitative spectroscopic study of the O dwarf population in 30 Doradus (Large Magellanic Cloud, LMC), aiming at clarifying some long-standing open questions about the formation, evolution and physical properties of O stars during their main-sequence phase, where they spend most of their lifetime. This work has been conceived within the VLT-FLAMES Tarantula Survey (VFTS), which has provided high quality optical spectra of hundreds of O stars in the H II region 30 Doradus, located within the LMC. The VFTS O dwarf population is an ideal laboratory to test our theoretical predictions and the e¿ects of metallicity in order to shed light on the main open questions about O dwarf stars. The main problems studied in this thesis are basically divided into four blocks: (1) O Vz stars, a subclass of O-type dwarf stars, have been suggested to be younger than normal O dwarfs, thus representing a direct link from the formation phase to the ¿rst stages in the evolution of massive stars, but this suggestion required further investigation. (2) The use of a calibration of the stellar parameters with spectral type for O stars is necessary to characterize their physical properties and evolutionary state when only photometry and low-resolution spectroscopy are available. (3) The determination of the stellar mass of an O star is a di¿cult task as shown by the mass discrepancy between evolutionary and spectroscopic masses reported by di¿erent authors in the last decades. (4) Several O dwarfs with relatively low luminosities and located within environments with di¿erent metallicity have shown thinner winds than predicted, a feature known as the weak wind problem. We have performed a quantitative spectroscopic analysis of a sample of 105 apparently single-star VFTS O dwarfs using the IACOB Grid-Based Automatic Tool (IACOB-GBAT), an accurate and fast method that compares, by means of a chi^2 algorithm, observed and synthetic H and He line pro¿les in an automated way. It is based on a pre-computed grid of almost 200000 stellar atmosphere models calculated with the fastwind code. Using the IACOB-GBAT, we have performed a series of tests aimed at (1) checking the behaviour of the automatic tool, and (2) investigating the e¿ects of several sources of error (i.e., nebular contamination, erroneous rotational velocities and wrong continuum normalization) on the determination of the stellar parameters. We have ¿nally derived mean values and uncertainties of stellar and wind parameters for our sample of stars, which have been used to study the problems described above. From the investigation of O Vz stars in 30 Doradus by comparing analytical results and fastwind predictions we can say that the Vz phenomenon can be explained by the natural combination of stellar parameters, mostly depending on e¿ective temperature and wind-strength. As a population, O Vz stars tend to be younger and have weaker winds than O V stars, but the Vz characteristic does not imply necessarily a young object with a thin wind. The large number of O Vz stars in 30 Doradus is a consequence of the low metal content of the host region. We have provided new spectral type calibrations for e¿ective temperatures and gravities of O dwarfs in the LMC, which are now the most complete calibrations, in terms of homogeneity and number of stars. In contrast to previous studies, we have not seen a clear e¿ect of metallicity on the temperature scale. A systematic di¿erence between evolutionary and spectroscopic masses has been derived for the O dwarfs in 30 Doradus, with a certain tendency to increase with higher temperatures and lower gravities. Finally, almost half of the O dwarfs in our sample present winds which cannot be properly characterized with optical diagnostic. These objects are potential candidates to have weak winds, but alternative diagnostics, such as those provided by the UV spectral range, are necessary.