The article examines approaches to the optimization of structural calculations, which require a comprehensive analysis of deformation processes under the action of operational loads. The finite element method is usually used to solve such problems. Modeling of structures taking into account stress concentrators requires the use of large-sized calculation grids. In real operating conditions, most structural elements are under the influence of mechanical and thermal loads. The package of application programs «MIRELA+» is intended for solving problems of thermoelasticity of structures made of weakly compressible elastomers and composites with an elastomeric matrix. One of the stages of solving problems of thermoelasticity is the formation of stiffness matrices taking into account weak compressibility and thermal conductivity of finite elements. When forming the stiffness and thermal conductivity matrices, it is necessary to perform integration over the area of the finite element. For elements of the same type, calculations are used according to the same procedure, and the number of such calculations determines the time to solve the problem. In the traditional approach, these calculations are performed sequentially. In the case of large grid sizes, the number and time of calculations increase, which requires optimization of calculations using parallel calculation algorithms. Using the OpenMP library, a parallel algorithm for forming the matrix of solving equations was built. The calculation of terms according to the quadrature scheme is included in the parallel area of the algorithm. Filling of the stiffness matrix of the finite element takes place in the sequential region. The execution time of parallel calculations is determined by the slowest subtask. Using the moment scheme of finite elements, displacement and deformation are approximated by the same polynomials, which simplifies the calculation of integrals. Parallel programming algorithms for constructing solving equations of the “MIRELA+” program package have been developed. Solutions for calculation grids of different sizes have been obtained. The effect of parallelization on the calculation time was studied.