IMR OpenIR
焊接残余应力及其爆炸处理的数值模拟
其他题名Numerical simulation of welding residual stress and its explosive treatment
管建军
学位类型博士
导师张劲松
2008-05-29
学位授予单位中国科学院金属研究所
学位授予地点金属研究所
学位专业材料加工工程
关键词温度场/应力场 焊接残余应力 爆炸消除应力处理 数值模拟
摘要焊接结构中普遍存在较高的焊接残余应力,其对焊接结构的安全可靠运行有着重要影响。因此研究焊接应力产生的规律及其控制或消除方法一直是焊接领域的研究热点。本文以数值模拟方法为主,对焊接残余应力产生有关的几个问题及爆炸消除应力工艺进行了研究。 本文首先对建立焊接温度场、应力应变场数值计算模型所涉及的理论基础、假设前提、相关概念,以及应用Ansys软件模拟所采用的方法进行了介绍和分析。在介绍爆轰波的C-J模型的基础上,对凝聚态炸药爆轰产物状态方程系数的拟合确定过程进行了推导和说明。介绍了钢板、空气的本构模型和状态方程,应用LS-DYNA软件模拟所采用的算法、沙漏问题的处理方式及时间步长的控制方法等。 为进一步认识爆炸处理对构件性能安全的影响,本文首次应用示波冲击试验机研究了单、双面爆炸处理后16Mn钢冲击性能的变化,从裂纹萌生功和扩展功两方面对冲击总功的变化进行了探讨,发现单面爆炸能改善韧性而双面爆炸却降低韧性。 控制薄板的焊接应力变形是一个难点。本文介绍并研究了一种新的焊接方法-跟踪加热焊接法,用随焊移动盖板限制瞬态应力产生的变形时,该方法焊接的薄板的焊接残余应力接近为零,翘曲变形也非常小。与常规焊相比,可以认为是无应力无变形焊接法。 为分析焊件横截面的应力、应变分布,常应用“平截面”假设。本文所做的数值模拟试验表明该假设与事实不符,并指出数值模拟后处理中根据初始坐标提取结果数据时,存在隐性的“平截面”假设前提。忽视该前提,可能会得到错误的结论。在本文的研究条件下,由于同一横截面上节点的相对纵向位移变化和前后相邻横截面上节点的物理量差异变化不同步,认为根据初始坐标提取任意时刻的结果可以用来表征该截面的状况。 本文提出存在加热和冷却过程时应该相对于自由热膨胀和自由热收缩来定义金属所承受的压缩应变和拉伸应变,而不能用相对于初始位置的实际应变的正负变化来表征。只有当任意两个相邻时刻金属的实际应变增量小于热应变的增量时金属受压缩,而当实际应变增量大于热应变的增量时金属受拉伸。本文认为焊件上各点应变的演化不仅仅由自身热应变决定,而且受比邻节点的影响和制约。焊接接头区域金属所承受的最大压缩应变的幅值超过其承受的最大热应变,完全冷却后最大拉伸应变也超过最大热应变。存在热作用时金属所承受的最大压缩应变、最大拉伸应变的代数和与残余总应变相等。发现超过母材屈服强度的残余拉应力区的宽度和最大实际压缩塑性应变超过0.2%区域的宽度相当,该区域残余实际压缩塑性应变约为0.2%。 本文根据板厚设计了8mm、16mm、32mm三种厚度的16Mn钢板的焊接工艺,应用Ansys软件模拟了试件的温度、应力的演变过程,得到了焊件的残余应力分布,为爆炸消除焊接应力工艺模拟打下了基础。根据埋弧焊的特点,采用双椭球热源来模拟其热输入,液固转变的热晗采用等效比热计算,以模拟金属熔化和凝固的吸热和放热过程。采用“生死”单元技术模拟焊缝单元的填充过程,使焊件的应力分布和变形与实际情况更为吻合。数值模拟计算结果表明所设计的焊接工艺参数是合理的,焊件的残余应力分布符合规律。 针对上述三种不同板厚的16Mn钢板,研究了爆炸工艺参数(主要是爆压)对焊接应力消除效果的影响。数值模拟试验表明:在爆轰波波阵面前方焊缝内部产生了先导膨胀波,由于其拉力较高,纵向残余拉应力得到彻底的释放,并且得到一定程度的拉伸,在爆轰波掠过前表现为压应力,炸药正下方的金属的爆炸处理状态处于“硬性爆炸条件”,而与炸药相隔较远的原始压应力区在爆炸处理后,应力值接近于0,其状态属于“中性爆炸条件”,更远的区域当属“软性爆炸区”;在“硬性爆炸区”若炸药爆压过高,则会在爆轰波掠过时产生过度压缩,使其应力反而向拉应力转化,同时爆压过高,也不利于结构的安全;爆压适当时,除表层金属外,内部单元的纵向应力都可为压应力;8mm和16mm钢板适合采用单面爆炸处理工艺,其理想的炸药的爆压分别为0.6GPa和1.5GPa,32mm厚钢板适合采用双面爆炸处理工艺,采用2GPa炸药同时爆炸工艺时效果很好,而2GPa炸药依次爆炸工艺的效果也很理想。
其他摘要There is higher welding residual stress in welded structure. It will endanger the safety of the structure. So it is a hot item to research the rule of the evolvement of welding stress and how to control or eliminate it. In this article, the research on some problem about welding residual stress and explosion treatment process was introduced. Those works were based on numerical simulation method mostly. Firstly, the theory, hypothesis, specific concept involved in the model of welding temperature field, stress strain field and the method how to carry out the simulation employed in Ansys software were introduced. Based on the C-J model of detonation wave, it was explained to how to determine the parameters of state equation of detonation product of condensed explosive. The constitutive and state equation of 16Mn steel and Air were interpreted also. The algorithm, solution to hourglass problem and control of time step employed by LS-DYNA software were commented, too. To further comprehend the effect of explosion treatment on structure safety, the impact properties of 16Mn steel, treated by single-side and double-side explosion treatment, were researched based on the initiation energy and expansion energy of the crack revealed by scillometric impact tester. It was inferred that single-side explosion treatment could improve impact toughness but double-side explosion treatment could deteriorate it. It is difficult to control the deformation of thin sheet when welded. A new welding method, welding process with trailing heating (WPTH), was introduced in this article. With a rigid cover plate to confine the deformation induced by transient stress, the residual stress of the thin sheet welded by WPTH was close to zero and the buckle deformation could be ignored. In contrast with normal welding process, WPTH could be considered as a stress-free and non-deformation welding method. The hypothesis of “planar section” is usually employed to analyze the distribution of stress and strain of any section of weldment. The result of simulation in this article indicated that the hypothesis didn’t accord with the fact. There was a “hidden” hypothesis of “planar section” when data of section by initial coordinate was drawn in the post treatment of simulation. A wrong conclusion about the section would be put forward if the “hidden” hypothesis was ignored. It was argued that the relative longitudinal displacement of nodes on the same section didn’t change following to the change of physical variable of nodes on adjacent sections. So it was regarded as low risk that the status of any section at any time was represented by data drawn by initial coordinate of the section. This article put forward a problem how to give right judgment on tensile or compressive status of metal when heated or cooled. It was argued that the compressive and tensile strain would be defined according to free thermal expansion and contraction. The two strains couldn’t be remarked by sign of actual strain which is calculated according to node’s real displacement. When the increment of actual strain is lower than the increment of thermal strain between two following time points, it is considered that metal is compressed, otherwise stretched. The development of strain of points on the weldment is not only determined by itself thermal strain, but also is affected and confined by adjacent nodes. The absolute value of maximal compressive strain of metal located in welded joint exceeds to the maximal thermal strain and the maximal tensile strain after cooled is also greater than the maximal thermal strain. The sum of the maximal compressive strain and the maximal tensile strain is equal to residual strain. It was indicated in this article that the width of area in which residual stress is greater than yield strength at room temperature is same to that of area in which the maximal actual compressive strain exceeds to 0.2% and the residual actual compressive strain remains 0.2% at the same area. Different welding process parameters were designed according to the thickness, 8mm, 16mm, 32mm, of the 16Mn steel. The development of temperature, stress and strain was simulated by Ansys software. The residual stress distribution of the weldment was gained and founded the basis of simulation of explosion treatment. The double ellipsoid heat source model was selected to simulate the heat input of SMW. The heat absorption and release of liquid-solid phase transformation was calculated as effective specific heat. The technologic of “birth-death” element was employed to simulate the filling of weld element for more agreement of the simulation result on real world. The simulation result indicated that the parameters of welding process were reasonable and the residual stress distribution was perfect. A series of parameters of explosion treatment process were designed and their affection on elimination of welding residual stress was analyzed. The simulation tests showed that the expansion wave propagating on the weld before front wave of detonation wave induced the relief of longitudinal tensile stress because of its higher amplitude and that it stretched the weld elements so much that there was compressive stress in those elements before detonation wave passed them. The explosion treatment status of the metal under the explosive was regarded as “excessive” status. That of the metal with residual compressive stress far away from the explosive, where stress was close to zero after explosion treatment, belonged to moderate status. And that of the metal more far away was classed as “soft” status. It was argued that excessive compress would convert compressive stress into tensile stress and that it would deteriorate structure safety if the detonation pressure was too high. The longitudinal stress of inner weld elements could be changed into compressive stress. The single-side explosion treatment was suited to be employed on the 8mm-thickness and 16mm-thickness steel. The ideal detonation pressure was 0.6GPa and 1.5GPa for two steel respectively. The double-side explosion treatment could be designed for 32mm-thickness steel. The double-side explosion treatment was divided into two processes: coinstantaneous explosion and one-by-one explosion. The ideal detonation pressure for two processes was same, 2GPa.
页数141
语种中文
文献类型学位论文
条目标识符http://ir.imr.ac.cn/handle/321006/17018
专题中国科学院金属研究所
推荐引用方式
GB/T 7714
管建军. 焊接残余应力及其爆炸处理的数值模拟[D]. 金属研究所. 中国科学院金属研究所,2008.
条目包含的文件
条目无相关文件。
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[管建军]的文章
百度学术
百度学术中相似的文章
[管建军]的文章
必应学术
必应学术中相似的文章
[管建军]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。