Particle-laden flow is of importance in cloud formation, plankton in the ocean, paper-making industry and so on. In these fields, the shapes of particles are usually non-spherical, which have obvious effects on particle collisions, accumulations, and clustering. According to earlier research, the alignments of inertialess spheroidal particles are associated with local velocity gradient tensor, i.e. eigenvectors of strain-rate tensor and vorticity. The preferential alignments cause that oblate spheroids tumble more quickly than of prolate spheroids in homogeneous isotropic turbulence. Besides, the angles between two nearby prolate spheroids are anomalously large in the center of turbulent channel flow [Zhao et.al., Phys. Rev. Fluids (2019)]. In this study, we simulated spheroidal particles in homogeneous isotropic turbulence at a Taylor Reynolds number Reλ= 97.7 with 2523mesh grids by using a pseudo-spectral method (see figure 1). We use the Lagrangian method to track particles trajectories and calculate the rotational motion of particles. The tumbling rates of particles with different shapes are in a good agreement with the results by Parsa et al. (2012). We will present the results of angular structure functions of spheroids, defined by