Structural health monitoring (SHM) is a research focus involving a large category of techniques performing in-situ identification of structural damage, stress, external loads, vibration signatures, etc. Among various SHM techniques, those able to monitoring structural deformed shapes are considered as an important category. A novel method of deformed shape reconstruction for thin-walled beam structures was recently proposed by Xu et al. [1], which is capable of decoupling complex beam deformations subject to the combination of different loading cases, including tension/compression, bending and warping torsion, and also able to reconstruct the full-field displacement distributions. However, this method was demonstrated only under a relatively simple loading coupling cases, involving uni-axial bending and warping torsion. The effectiveness of the method under more complex loading cases needs to be thoroughly investigated. In this study, more complex deformations under the coupling between bi-axial bending and warping torsion was decoupled using the method. The set of equations for deformation decoupling was established, and the reconstruction algorithm for bending and torsion deformation were utilized. The effectiveness and accuracy of the method was examined using a thin-walled channel beam, relying on analysis results of finite element analysis (FEA). In the analysis, the influence of the positions of the measurement of surface strain distributions on the reconstruction accuracy was discussed. Moreover, different levels of measurement noise were added to the axial strain values based on numerical method, and the noise resistance ability of the deformation reconstruction method was investigated systematically. According to the FEA results, the effectiveness and precision of the method in complex deformation decoupling and reconstruction were demonstrated. Moreover, the immunity of the method to measurement noise was proven to be considerably strong.