当前位置: 首页>博士论文>资源详情
刚体冲击作用下建筑夹层玻璃的动态响应和破坏行为研究
中文摘要

夹层玻璃在建筑幕墙和结构领域的应用愈加广泛,然而冲击荷载下玻璃易脆性特性易引发安全风险,相关的设计规范中关于夹层玻璃抗冲击设计内容极为匮乏,其冲击破坏过程和机理缺乏足够的基础性研究。本文以PVB和SGP胶层的建筑夹层玻璃为研究对象,采用落重冲击实验和组合有限元-离散元模型分析等手段,对建筑夹层玻璃冲击破坏响应和机理展开研究。首先,设计并加工了一套结合有高速摄影和动态采集系统的冲击实验系统;随后采用最小平均碎裂速度测试法(MMBV)对23组83块足尺PVB和SGP夹层玻璃板进行了实验研究,测定了每层玻璃碎裂对应的冲击能量阈值和冲击响应,并利用高速相机捕捉夹层玻璃起裂和裂纹高速扩展过程;对夹层玻璃在玻璃层数、玻璃类型和构型、玻璃尺寸、支撑条件、胶片厚度这六种设计参数影响下的抗冲击性能和破坏机理进行分析,确定了碎裂能、碎裂刚度、冲击力、等效动态刚度和能量耗散行为关键影响因素和规律;为开发夹层玻璃冲击破坏数值模型,首先针对四类常用的数值方法关于玻璃冲击破坏仿真能力的优劣性进行了系统的对比研究,确定了适用于夹层玻璃冲击破坏的仿真分析方法,即FEM/DEM法;其次基于FEM/DEM法开发离散裂纹模型、胶片超弹性模型和胶合面耦合模态脱层模型的数值分析模型,通过对其模拟结果和实验观测进行比较,验证了该分析模型的有效性并揭示其尚存的问题。本文通过实验和数值研究剖析了建筑夹层玻璃于刚体冲击下的破坏机理,结果可对土木工程应用中夹层玻璃抗冲击设计提供指导。本文主要结论如下: (1)夹层玻璃板冲击破坏下的裂纹/变形模态:双层夹层玻璃裂纹模态以内层的放射型裂纹和环向次级裂纹网格、外层的涟漪型裂纹和花瓣状裂纹为主要特征;三层夹层玻璃易出现不规则裂纹区,钢化水平越高则碎片尺寸越小;破坏后变形模态以十字形屈服线(同种玻璃)、一字型屈服线(外半钢化内钢化)和X型屈服线(外钢化内半钢化)为主。两块玻璃面板同时碎裂通常触发于内层,裂纹触发滞后时间小于1ms,涟漪型裂纹扩展速度高于呈现递减趋势的放射型裂纹。 (2)冲击头运动特征:玻璃类型、胶层类型对冲击头运动特征的影响可忽略,玻璃面板支撑条件对冲击头运动特征影响在玻璃尺寸较小范围内比较明显;冲击头最大位移值较为稳定,无玻璃碎裂时不超过4㎜,有任意一层玻璃碎裂时不超过6㎜,两层均碎时通常大于30㎜。 (3)各设计参数对夹层玻璃抗冲击性能及冲击响应影响规律如下:a)胶片厚度:夹层玻璃采用1.52 ㎜PVB胶层相较0.76 ㎜和3.04 ㎜厚度具有更优的抗冲击性能,胶层低于0.76 ㎜非线性临界速度降低且增加面板同时碎裂的风险;增加SGP厚度对抗冲击性能无显著作用,对能量耗散反而有负面影响;b)支撑条件:边缘夹固式支撑相对驳接点支式具有更好的抗冲击性能,其提升比率可高于1倍,且后破坏阶段有较好的抗往复冲击的“延性”;C)玻璃类型/构成:内半钢化-外钢化的构成可显著提高抗冲击性能和临界速度,外层采用半钢化玻璃会削弱抗冲击性能,玻璃面板的能量耗散特性受玻璃类型的影响可忽略。降低外层-内层厚度比可提高抗冲击性能和临界速度,而单纯增加内层玻璃厚度同样可以提升抗冲击性能,但其呈现出易在同一次撞击中两片玻璃同时碎裂的趋势,增加脆性破坏的概率。玻璃厚度对耗散冲击能能力影响可以忽略。玻璃类型和玻璃面板总厚度对冲击力峰值作用明显,且遵循钢化玻璃面板总厚度越高可承受冲击力更高的准则;d)玻璃层数/尺寸:增加玻璃面板层数无助于提升抗冲击性能,却使内层玻璃更易碎裂,中层玻璃对刚度作用相对外层和内层更显著。玻璃尺寸增加对抗冲击性能有较小的负面作用。 (4)揭示了往复冲击(冲击物质量13.5㎏)下的夹层玻璃刚度退化规律,其中双层夹层玻璃碎裂前刚度退化的临界速度都小于3.1m/s(PVB夹层)和2.5 m/s(SGP夹层),后破坏阶段临界速度通常小于3.5m/s(PVB夹层)和2.9m/s(SGP夹层);三层夹层玻璃碎裂前阶段均无刚度退化,后破坏阶段临界速度都小于3.1m/s(PVB夹层)和2.9 m/s(SGP夹层)。 (5)夹层玻璃冲击破坏主要出现最外层脱层和中间层脱层两种脱层行为,边缘夹固式支撑会导致往复冲击下产生连续脱层,而驳接点支式的夹层玻璃最外层脱层扩展区域则较为稳定(直径≤100 ㎜)。 (6)用DEM和FEM/DEM方法可以较好地模拟典型的裂纹模态及碎裂的演变过程;然而,DEM的参数校准方法无法有效匹配能量释放率,因此其模拟的材料能量耗散性能有较高的不可控性,该问题可能导致模型能量耗散及其他全局结果有显著偏差。 FEM/DEM对高速及斜向冲击下玻璃表面的损伤模态和尺寸特征及接触持时等结果模拟效果较好,但当接触区域单元尺寸设定过小时,高速冲击碎裂过程会将冲击能量过度耗散掉。XFEM在研究动态断裂问题时存在局限性,难以获得典型的裂纹模态。FEM方法可以模拟典型的裂纹模态和碎裂过程,但单元消除法的应用会低估整体冲击响应。 (7)采用本文开发的FEM/DEM模型计算的应力波传播特征和实验观测及解析结果吻合较好,并发现和冲击瞬时产生的压碎型裂纹和脱层将会严重延缓应力波的传播;本文所采用的改进Xu-Needleman模型可成功重现冲击点附近模态Ⅰ和模态Ⅱ主导的脱层行为。 关键词:夹层玻璃,冲击破坏,结构玻璃,刚体冲击,组合有限元-离散元法,脱层,抗冲击性能

英文摘要

As one type of safety glass, laminated glass (LG) is increasingly used in curtain wall and structural purpose. The brittle cracking characteristic of glass under impact renders great risk during service. However, impact resistance design models is not available in most design codes or standards so far, and the impact damage mechanism lacks adequate insightful information supported with sufficient experimental and numerical data. It is therefore of importance to conduct studies on the impact resistance of laminated glass. This work concerns on the architectural laminated glass using polyvinyl butyral (PVB) and SentryGlasⓇPlus (SGP) interlayers, and carries out the investigation into the impact damage mechanism and the numerical prediction model with both drop weight impact tests and combined finite-discrete element analysis. A drop weight impact test system encompassing high speed filming and dynamic impact response acquisition system is firstly devised and manufactured. An experimental investigation into the damage behavior of full sized PVB and SGP laminated glass panels is then carried out. A mean minimum breakage velocity (mmBV) test approach has been employed in testing 83 specimens of 23 groups, to determine the breakage energy that triggers glass breakage. The crack initiation and crack propagation process is captured using high speed photos. This study carefully examines the effects of six design variables, that are, quantity of glass panels, glass make up, glass type, panel size, support condition, interlayer thickness, on the impact resistance and damage mechanism of laminated glass. The key factors that influence the breakage energy, breakage stiffness, impact force, equivalent dynamic stiffness and energy dissipation behavior are then identified. In order to overcome the problems in modelling the impact failure of laminated glass, a comparative review on the available numerical approaches (finite element method, FEM; extended finite element method, XFEM; discrete element method, DEM; combined finite-discrete element method, FEM/DEM) is carried out to illustrate their fundamental principles, modelling techniques and applications by using several example cases. An example of monolithic glass beam under impact is examined to identify the weakness and advantages of each approach. The most feasible approach, i.e. FEM/DEM, is determined and further examined and their results were compared with the experimental data for modelling the high speed and oblique impact tests on glass. By addressing the key problems of FEM/DEM in modelling the delamination of glass-interlayer interface, this work develops a model encompassing the formulation of the discrete crack model (DCM) for glass, the Mooney Rivlin model to represent the hyperelasticity of PVB interlayer, and the adapted Xu and Needleman model to describe the irreversible combined damage-plasticity behavior of interface. The comparison between the simulation and experimental results for several glass make-ups validates the applicability of the proposed FEM/DEM model and identifies the shortcoming and resulting errors. This work investigates the damage mechanism of architectural laminated glass under hard body impact through experimental and numerical studies, it can provide insightful information and suggestions for impact resistance design of laminated glass. The concluding remarks are collarated and listed as follows: (1)Crack/deformation pattern: The crack pattern of laminated glass is featured with the radial cracks and circumferential crack network in inner glass panel, the rippled cracks and petal shaped cracks in outer glass panel. An abnormal crack pattern is commonly seen in triple layered LG. Smaller fragments coincide with higher strengthening level. The deformation patterns are featured with cross-type yield line (in LG made of same glass types), single lateral yield line (in HSG-FTG) and x-type yield line (FTG-HSG). If glass panels experiences simulatation cracking in all constituent panels, the crack commonly initiates in the inner panel, and the lagging time of the following cracks is less than 0.16 ms (PVB LG) or 0.96 ms (SGP LG). The propagation speed of rippled crack is higher than that of radial crack. (2)Impactor motion: The effects of glass type and interlayer type on the impactor motion are negligible, support condition only shows obvrious influence when glass panel is larger. The peak impact displacement is close in each LG type. It is less than 4 ㎜ for intact glass, is less than 6 ㎜ at single glass panel breakage, and is larger than 30 ㎜ when both panels break. (3)The effects of design variables investigated are summarized as follows: a) Interlayer thickness: The use of 1.52 ㎜ PVB interlayer in LG can obtain better impact resistance than 0.76 or 3.04 ㎜ PVB. The interlayer thinner than 0.76 ㎜ will reduce the critical impact velocity that triggers the stiffness degradation and increases the risk of both panels breaking at the same impact. Increasing the SGP thickness presents no improvement of impact resistance and even has negative effect on dissipating impact energy.b) Support condition: Edge clamping can provide better impact resistance than bolted connection. The corresponding increase ratio may be larger than 100%. It also provides better ductility in the post breakage stage when subjected to repeated impacts.c) Glass type / glass make up: FTGHSG configuration can evidently improve impact resistance and critical velocity. Placing HSG in the outer side will weaken impact resistance. The effects of glass types on energy dissipation characteristics can be omitted. In addition, reducing the ratio of outer-inner panel thickness without increasing total thickness of LG can enhance impact resistance and critical velocity, while it can also improve impact resistance but tends to have brittle failure in the same impact and has no enhancement in breakage stiffness if only increasing inner panel thickness. The glass thickness has negligible influence on dissipating energy. One conclusion can be found that glass type and total glass thickness will significantly affect the amplitude of impact force, that is, if LG panel adopts more FTG panels or FTG panel with greater thickness, it can sustain greater impact force; d) Glass panel quantity / glass panel size: Increasing glass panels cannot improve impact resistance, it may even cause higher risk of inner glass breakage. It can be found that the middle glass panel contributes more in LG stiffness than both outer and inner glass panel. Increasing panel size will cause negative effects on impact resistance, and its effects on improving breakage stiffness depend on breakage sequence. (4) Critical velocity that triggers the stiffness degradation: for double layered LG, critical velocity is found to be less than 3.1 m/s (PVB) and 2.5 m/s (SGP) before the first breakage,3.1m/s (PVB) and 2.9 m/s (SGP) before the second breakage. For triple layered LG, no stiffness degradation can be seen before breakage. The critical velocity is less than 3.1 m/s (PVB) and 2.9 m/s (SGP) in the post breakage stage. The corresponding impactor mass is 13.5㎏. (5)Two types of delamination can be found, that are, outmost delamination and inter delamination. Edge clamping leads to continuous delamination under repeated impacts, while bolted connection will maintain constant size of delamination zone of which the diameter is less than 100 ㎜. (6)DEM and FEM/DEM approach can satisfactorily simulate typical crack pattern and fracture growth. However, DEM presents difficulty in calibrating realistic properties such as energy release rate. It may lead to a high degree of unpredictability in modelling energy dissipation features of materials, and further results in the remarkable deviation of global response. On the contrary, FEM/DEM can successfully produce the pattern and the featured size of glass surface damage, and predicts contact duration in the high-speed impact cases. However, FEM/DEM will overestimate the energy dissipation during the high-speed impact because of the small element size setting in the contact zone. XFEM shows limitations in modelling dynamic fracture and obtaining typical crack pattern. FEM with element erosion method can adequately model typical crack pattern, whereas the element erosion method will underestimate the global impact response. (7)The developed FEM/DEM model can adequately reproduce the characteristics of stress wave propagation that map well with experimental and analytical results. The results reveal that crushing cracks and delamination will cause dramatic delay of wave propagation. The delamination model proposed in this work can successfully produce the Mode I and Mode II dominate delamination behavior near impact point. This work also identifies the shortcoming and resulting errors of discrete crack model for glass when modelling thermally strengthened glass. KEY WORDS: laminated glass, impact failure, structural glass, hard body impact, combined finite-discrete element method (FEM/DEM), delamination, impact resistance

作者相关
主题相关
看过该书的人还在看哪些书