针对任意化学成分的热质量恒星实时线驱风质量损失估算器(LIME)
An on-the-fly line-driven-wind iterative mass-loss estimator (LIME) for hot, massive stars of arbitrary chemical compositions
摘要 Abstract
热质量恒星的质量损失率对于天体物理学的诸多应用至关重要。我们提出了一种快速、高效且易于使用的实时质量损失计算器,用于具有给定恒星参数和任意化学成分的热质量恒星线驱风质量损失计算,并通过万维网向公众开放使用(https://lime.ster.kuleuven.be/)。线驱动力基于激发和电离平衡计算,利用包含超过400万条光谱线的大规模原子数据库实时计算得出。随后,依据考虑有限恒星盘和气体声速影响的线驱风理论推导出质量损失率。对于给定的一组恒星参数和化学成分,我们在风临界点处预测质量损失率以及三个线驱动力参数Qbar、Q0和alpha。将我们的预测结果与来自XshootU合作项目(Vink等,2023年)的大量近期状态-of-the-art、同质化经验质量损失率进行比较表明,本文提供的超级简单计算器在平均意义上甚至优于Vink等(2001年)、Bjorklund等(2023年)和Krticka等(2024年)的质量损失公式(这些公式均基于更复杂但更为繁琐且灵活性较低的方法对限制模型网格的拟合)。除了速度和简便性之外,我们质量损失计算器的优势在于避免了将拟合公式应用于针对更受限参数范围计算的基础模型网格所导致的不确定性。特别是,这里可以轻松修改单个化学丰度,并探索其对预测质量损失率的影响。这使得直接应用于表面化学成分显著改变的恒星成为可能。
Mass-loss rates from hot, massive stars are important for a range of astrophysical applications. We present a fast, efficient, and trivial-to-use real-time mass-loss calculator for line-driven winds from hot, massive stars with given stellar parameters and arbitrary chemical compositions, and make it available for public use via the world wide web (https://lime.ster.kuleuven.be/). The line-force is computed on-the-fly from excitation and ionisation balance calculations using a large atomic data base consisting of more than 4 million spectral lines. Mass-loss rates are then derived from line-driven wind theory including effects of a finite stellar disk and gas sound speed. For a given set of stellar parameters and chemical composition, we obtain predictions for the mass-loss rate and the three line-force parameters Qbar, Q0, and alpha at the wind critical point. Comparison of our predictions to a large sample of recent state-of-the-art, homogeneously derived empirical mass-loss rates obtained from the XshootU collaboration project (Vink et al. 2023) demonstrates that the super-simple calculator provided in this paper on average performs even better than the mass-loss recipes by Vink et al. (2001), Bjorklund et al. (2023), and Krticka et al. (2024) (which are all fits to restricted model grids based on more sophisticated, but also more intricate and much less flexible, methods). In addition to its speed and simplicity, a strength of our mass-loss calculator is that it avoids uncertainties related to applying fit-formulae to underlying model-grids calculated for more restricted parameter ranges. In particular, individual chemical abundances can here be trivially modified and their effects upon predicted mass-loss rates readily explored. This thus allows also for direct applications toward stars that are significantly chemically modified at the surface.