激光穿透异种金属的熔池的动态行为影响了异种金属的焊接质量,本文采用仿真软件Fluent,通过流体体积法(VOF)对异种金属熔池形貌演化及匙孔瞬态行为进行模拟。阐明了匙孔失稳现象与温度场分布、熔池形态及流速分布的关系。研究以304不锈...激光穿透异种金属的熔池的动态行为影响了异种金属的焊接质量,本文采用仿真软件Fluent,通过流体体积法(VOF)对异种金属熔池形貌演化及匙孔瞬态行为进行模拟。阐明了匙孔失稳现象与温度场分布、熔池形态及流速分布的关系。研究以304不锈钢与镍为研究对象,在求解动量、能量、质量守恒方程的基础之上,同时考虑匙孔、反冲压力及熔池的耦合作用下建立三维数值模型。研究发现:在匙孔未能穿透Fe/Ni异种金属工件时,匙孔壁面的反冲压力与表面张力难以保持动态平衡,上部匙孔壁面凸起向匙孔下方倾斜,造成匙孔闭合坍塌现象;当匙孔穿透Fe/Ni异种金属工件瞬间,部分激光束经下部通孔逃逸导致反冲压力降低,又因熔池表面张力的存在引起部分液态金属向匙孔中心位置流动,二者共同作用导致通孔中间部位的失稳坍塌。The dynamic behavior of laser penetrating the molten pool of dissimilar metals affects the welding quality of dissimilar metals. In this paper, the simulation software Fluent is used to simulate the morphology evolution of molten pool of dissimilar metals and the transient behavior of key holes through volume of fluid (VOF). The relationship between keyhole instability and temperature field distribution, molten pool shape and flow rate distribution is elucidated. Study with 304 stainless steel and nickel as the research object, the momentum, energy and mass conservation equation, on the basis of considering keyhole, recoil pressure and molten pool under the coupled action of three dimensional numerical model is established. The research found that: when the keyhole fails to penetrate the Fe/Ni dissimilar metal workpiece, it is difficult for the backpressure and surface tension on the keyhole wall to maintain dynamic balance, and the upper part of the keyhole wall bulges and inclines downwards, causing the keyhole to close and collapse;when the keyhole penetrates the Fe/Ni dissimilar metal workpiece instantly, part of the laser beam escapes through the lower through-hole, resulting in a decrease in backpressure. Meanwhile, due to the existence of surface tension of the molten pool, some liquid metal flows to the center of the keyhole. These two factors jointly lead to the instability and collapse of the middle part of the through-hole.展开更多
针对神经网络训练时容易陷入局部最优的问题,提出一种基于工业入侵检测模型的帝王蝶优化算法。利用帝王蝶算法(monarch butterfl y optimization,MBO)和狮群进化算法(lion pride optimization,LPO)的全局寻优和局部寻优提高网络参数的...针对神经网络训练时容易陷入局部最优的问题,提出一种基于工业入侵检测模型的帝王蝶优化算法。利用帝王蝶算法(monarch butterfl y optimization,MBO)和狮群进化算法(lion pride optimization,LPO)的全局寻优和局部寻优提高网络参数的求解质量,从而提高入侵检测结果的准确率。实验在工业控制系统标准数据集进行评估,结果表明模型符合工业入侵检测标准,与其它流行算法相比较,优化后的工业入侵检测技术具有较高的检测精度和较快的收敛速度。展开更多
文摘激光穿透异种金属的熔池的动态行为影响了异种金属的焊接质量,本文采用仿真软件Fluent,通过流体体积法(VOF)对异种金属熔池形貌演化及匙孔瞬态行为进行模拟。阐明了匙孔失稳现象与温度场分布、熔池形态及流速分布的关系。研究以304不锈钢与镍为研究对象,在求解动量、能量、质量守恒方程的基础之上,同时考虑匙孔、反冲压力及熔池的耦合作用下建立三维数值模型。研究发现:在匙孔未能穿透Fe/Ni异种金属工件时,匙孔壁面的反冲压力与表面张力难以保持动态平衡,上部匙孔壁面凸起向匙孔下方倾斜,造成匙孔闭合坍塌现象;当匙孔穿透Fe/Ni异种金属工件瞬间,部分激光束经下部通孔逃逸导致反冲压力降低,又因熔池表面张力的存在引起部分液态金属向匙孔中心位置流动,二者共同作用导致通孔中间部位的失稳坍塌。The dynamic behavior of laser penetrating the molten pool of dissimilar metals affects the welding quality of dissimilar metals. In this paper, the simulation software Fluent is used to simulate the morphology evolution of molten pool of dissimilar metals and the transient behavior of key holes through volume of fluid (VOF). The relationship between keyhole instability and temperature field distribution, molten pool shape and flow rate distribution is elucidated. Study with 304 stainless steel and nickel as the research object, the momentum, energy and mass conservation equation, on the basis of considering keyhole, recoil pressure and molten pool under the coupled action of three dimensional numerical model is established. The research found that: when the keyhole fails to penetrate the Fe/Ni dissimilar metal workpiece, it is difficult for the backpressure and surface tension on the keyhole wall to maintain dynamic balance, and the upper part of the keyhole wall bulges and inclines downwards, causing the keyhole to close and collapse;when the keyhole penetrates the Fe/Ni dissimilar metal workpiece instantly, part of the laser beam escapes through the lower through-hole, resulting in a decrease in backpressure. Meanwhile, due to the existence of surface tension of the molten pool, some liquid metal flows to the center of the keyhole. These two factors jointly lead to the instability and collapse of the middle part of the through-hole.
文摘针对神经网络训练时容易陷入局部最优的问题,提出一种基于工业入侵检测模型的帝王蝶优化算法。利用帝王蝶算法(monarch butterfl y optimization,MBO)和狮群进化算法(lion pride optimization,LPO)的全局寻优和局部寻优提高网络参数的求解质量,从而提高入侵检测结果的准确率。实验在工业控制系统标准数据集进行评估,结果表明模型符合工业入侵检测标准,与其它流行算法相比较,优化后的工业入侵检测技术具有较高的检测精度和较快的收敛速度。
