The current codes and standards concerning wind loads on roof-mounted solar panels are discussed and summarized. Wind pressures on flat- and slope-roof-mounted solar arrays obtained from wind tunnel tests are compared with the recommended design values in ASCE 7-16 and JIS C 8955: 2017. Different parameters, including building side ratio, aspect ratio and parapet height, are examined. Results show that the largest wind pressures on flat-roof-mounted solar panels of all zones in ASCE 7-16 tend to be 10% to 26% smaller than the experimental results when normalized tributary area A _n is larger than 10 ³ . Uplift wind forces on flat-roof-mounted solar panels in downstream regions obtained from experiments can be larger than the recommended values in JIS C 8955: 2017 for adverse wind, but downward force coefficients are basically smaller than those in JIS C 8955: 2017 for fair wind. 40% to 60% increase on the pressure equalization factor for slope-roof-mounted solar panels is suggested for the potential refinement of ASCE 7-16 based on this study. Meanwhile, proposed pressures of slope-roof-mounted solar panels in JIS C 8955: 2017 might be too conservative according of experimental results.

Non-Gaussian wind field simulation plays a crucial role in studying wind-induced responses of wind turbine structures exposed to non-Gaussian wind excitations. To this end, the first, third, and fourth authors of this paper recently proposed an eigenvalue-based iterative translation approximation method (EITAM). Although theoretically superior to the conventional Cholesky decomposition-based iterative translation approximation method (CDITAM) in certain respects, the effectiveness of this method in non-Gaussian wind field simulations involving numerous simulation points has yet to be systematically investigated. In this paper, non-Gaussian wind field simulation via EITAM is presented for wind turbine structures. The CDITAM and EITAM are compared in detail through simulation cases with both weak and strong non-Gaussianity. Results showed that, compared to CDITAM, EITAM achieved smaller iteration errors, higher per-iteration efficiency, and greater robustness against non-convergence. In addition, the simulation accuracy achieved by EITAM was also slightly higher than that of CDITAM, which is consistent with the iteration results. Although the comparison result of iteration accuracy and total iteration efficiency of these two methods may vary with different cases, the overall performance of EITAM can still be considered superior to CDITAM due its conceptually simpler and higher convergence robustness. Therefore, the proposed method may hold significant potential for application in fluid–structure interaction analyses of wind turbines under non-Gaussian wind fields.