Automatic Generation and Design Optimization of Microstructure of Building Materials Based on Simulated Annealing Algorithm

The microstructure of building materials has a significant impact on their performance. Higher performance and lower cost can be achieved by optimizing and automatically generating the microstructure of building materials. Simulated annealing algorithm is a stochastic search based optimization algorithm that has been widely used in design optimization problems in various fields. The traditional manual design methods have problems such as low efficiency and limited design space. Therefore, in this paper, torsional deformation and annealing treatment are used to modulate the microstructure and improve the mechanical properties of MP35N alloy, and its mechanical behavior is analyzed by micro-mechanism. Torsional deformation introduces a gradient dislocation density in the radial direction of the alloy, which refines the grains and introduces a large number of laminar dislocations at the surface with the highest strain. The introduction of the gradient structure resulted in a significant increase in the strength of the alloy after torsional deformation to similar to 767 MPa, with an appreciable homogeneous elongation ( similar to 40%); further annealing treatment did not change the gradient structure introduced by the torsional deformation (both in terms of the gradient dislocation density and grain size), but the annealing treatment resulted in the annihilation of a portion of dislocations in the alloy, and the transformation of layer dislocations at the surface region into nano-twin crystals. The formation of nanotwin crystals due to the annealing process and the "Suzuki effect" result in secondary hardening of the alloy. The secondary hardening resulting from the annealing process counteracts the annealing softening due to dislocation annihilation, and the secondary hardening provides a strength of about 76 MPa. Through the optimization algorithm, the mechanical properties of the material are improved and the design requirements are met at the same time. It is hoped that the exploration and practice of this study will provide useful ideas and methods for the automatic generation and design optimization of the microstructure of building materials.