Abstract: An infill topology optimization method is proposed for the lightweight design of porous structures under thermo-mechanical loading. Traditional optimization methods often face challenges in thermoelastic structural topology optimization, such as design-dependent loads, complex sensitivity analysis, high computational costs, and the presence of gray areas in the designs, which hinder manufacturing. To address these issues, this paper proposes an infill topology optimization method that combines continuous and discrete variables, applied to solve the design problems of porous structures under thermo-mechanical loading. By introducing local volume constraints to replace global volume constraints, the proposed method avoids excessive uniformity in the generated porous structures. The local volume constraints not only facilitate the formation of substructures and guide the optimization process toward more uniform material distribution but also provide physical significance to the design variables in continuous variable optimization through the connection between continuous and discrete variable methods, preventing a significant increase in the objective function value when the design results are overly uniform. Furthermore, an improved sensitivity filtering strategy is developed, resulting in topology designs with better objective function values. Numerical examples demonstrate the effectiveness of method and the effects of filter radius and temperature difference is investigated. It is shown that the present method can transform continuous variable optimization results under coarse mesh into discrete variable porous structure designs under fine mesh, significantly enhancing the manufacturability of the designs.