沥青混合料劈裂试验的细观 有限元模拟与测试外文翻译资料
2022-09-01 18:00:10
沥青混合料劈裂试验的细观
有限元模拟与测试
摘 要:为分析沥青混合料的细观尺度破坏机理,采用数字图像技术将AC-25沥青混合料离散为集料、联结料、孔隙三相复合材料,基于ANSYS软件建立马歇尔试件的二维细观有限元模型,输入各相介质的力学参数,虚拟细观劈裂破坏试验。基于非接触式光学方法,测试AC-25马歇尔试件劈裂试验中各相介质应变,并与虚拟细观有限元分析结果进行对比。结果表明虚拟细观分析的各相介质参数影响较大,细观光学应变测试结果大于虚拟细观分析结果,但虚拟细观应变响应与光学测试结果规律相似,两种分析中沥青混合料破坏的薄弱环节均为联结料与界面。
关键词:道路工程,沥青混合料,马歇尔劈裂试验,细观有限元分析,细观光学试验,应变响应
0 引言
劈裂试验是进行沥青混合料抗裂性能评定、破损机理研究的主要途径,但仅从宏观尺度出发,将沥青混凝土简化为连续均匀的单一介质,很难全面合理的解释沥青混合料破坏机理。而从细观角度出发,沥青混合料通常可以分为集料、沥青联结料和孔隙三相介质,其不同介质的几何物理性质对劈裂破坏的影响值得深入探讨。
国内外已有一些学者开展了沥青混合料破损机理的细观尺度分析。于庆磊[1]等将数字图像处理技术与原有的RFPA2D系统相结合,通过数字图像处理技术和统计方法表征混合料内部骨料的非均匀性,从而模拟混合料在单轴荷载作用下破坏过程的变形非线性、应力重分布等现象和复杂应力状态下的裂纹扩展过程。邱延峻[2]等运用RFPA 真实破裂过程分析系统对马歇尔劈裂试验进行了全过程的数值模拟,发现了裂纹状态随均质度减小趋向离散,均质度越高,试件破坏越趋向于脆性破裂。皮育晖[3]等利用有限元软件ABAQUS,对沥青混合料劈裂试验进行数值模拟分析,研究了试验中应力、应变分布及其变化规律,讨论了网格划分密度、划分形式对模拟结果的影响。裴建中[4]等利用离散元方法生成沥青混合料马歇尔试件,数值模拟了不同粒径及不同加载速率对劈裂强度的影响和试件中裂缝形成的动态过程。Romanowicz Marek[5]从细观有限元模型出发,运用数值均匀化技术,简化模拟层压板结构模型,通过建立适当的本构方程,模拟两种不同的内部纤维造成基体开裂的失效机制。但总的说来,对沥青混合料破坏机理的分析侧重于数值模拟,缺乏实验分析和验证。
本文结合数字图像处理技术和有限元软件ANSYS,建立AC-25马歇尔试件的二维细观有限元模型,虚拟分析劈裂试验过程中的应变分布及其变化规律,同时应用非接触式光学方法测试细观尺度应变。通过对比分析有限元计算与试验结果,为沥青混合料细观破坏机理研究提供参考。
1沥青混合料马歇尔试件细观有限元分析
1.1 二维细观有限元模型
本文采用国内外已有的数字图像处理技术建立马歇尔试件的细观二维有限元模型,由于沥青混合料AC-25的原始截面较为复杂,如图1,因此需要结合Win Topo Professional和CAD等对试件截面模型进行转化处理,进而导入ANSYS软件进行建模分析。在细观尺度中,AC-25界面是由集料、沥青联结料和气孔组成的多相复合材料。有限元模型及细观组成部分如图2所示。
图1 AC-25试件原始界面图像
图2 AC-25细观有限元模型及细观组成成分
AC-25细观模型尺寸大小和宏观马歇尔试件相同,整体为一直径为101.6mm的圆形,上下压条宽度为13mm,厚度为4mm,内表面与试件同弧度,上下压条尺寸相同。AC-25细观模型采用PLANE42单元来定义单元类型,共有40350个单元,33549个节点,单元划分结果如图3所示。假定AC-25的集料和沥青联结料是线弹性体,物理性能不随外界条件变化而变化,集料与沥青联结料的性能参数如表1所示,相邻边界的单元按照共享节点处理。本模拟在上压条上边面中点上施加位移荷载,下压条刚体的参考点上设置对称边界条件,并限制底部的水平和垂直位移。
图3 AC-25劈裂试验细观模型有限元单元划分
表1 AC-25细观有限元模型材料参数
材料 |
弹性模量(MPa) |
泊松比 |
极限拉伸强度 [mu;ε] |
|
集料 |
55.5e3 |
0.16 |
97 |
|
沥青联结料 |
500 |
0.24 |
600 |
1.2 虚拟劈裂试验
在细观模型上通过分级加载来模拟劈裂试验,本模拟施加位移荷载且逐步加载,每次逐步增加0.05mm,按照0,-0.05mm,-0.10mm,-0.15mm,-0.20mm,-0.25mm位移荷载依次进行静态加载。通过计算,当施加位移荷载为-0.15mm时,AC-25细观界面集料和联结料的各向最大应变云图如图4和图5所示。
图4集料、和(位移荷载-0.15mm)
图5 联结料、和(位移荷载-0.15mm)
通过在细观模型上进行分级加载计算,集料及沥青联结料各向应变状态的变化趋势曲线如图6、7所示。可以看出,AC-25细观界面集料和沥青联结料的X向应变、Y向应变和第一主应变均根据逐步位移荷载呈线性趋势增长。集料的第一主应变与X向应变相差很小,第一主应变略大于X向应变,相差大约8%左右。沥青联结料的第一主应变大于X向应变,两者相差约为30%左右。因此,建议将第一主应变作为AC-25细观界面联结料的破坏应变参考指标。
图6 细观分析集料各向应变状态曲线
图7 细观分析联结料的各向应变状态曲线
2 劈裂试验的非光学接触测试
2.1 原理及公式
(1)试验仪器及准备
为了与有限元模拟的结果对比分析,观察AC-25试件的破坏路径,以及测定其试件即将破裂时和完全破裂时的细观界面应变状态,进行AC-25细观光学劈裂试验。AC-25试件是根据马歇尔试件制备的,直径为101.6mm,高为63.5mm,再利用切割机床切割试件剖面,形成AC-25细观界面,尺寸为直径101.6mm,高40mm。本试验基于双目立体视觉原理,采用三维数字图像相关方法,对马歇尔试件表面的三维形貌和荷载作用下的三维变形进行测量。图像测量系统如图8所示。
图8 图像测量系统
(2)材料非均质性描述[1]
根据以往的研究经验,通常运用统计数学的方法来描述材料的非均匀性。假设集料和联结料细观单元的弹性模量和强度均服从Weibull分布,
(1)
这里为细观单元力学性质参数,如弹性模量、强度等;为细观单元力学性质的统计分布密度;m为分布函数的形状函数,将其定义为材料的均匀性系数,它反映了材料的均质性; 为细观单元力学性质的平均值。从上式可以看出,随着材料均匀性系数m的增加,细观单元的力学性质将会收敛于一定的范围内,材料性质将趋于均匀。整个式子体现了材料内部介质细观力学性质非均匀性的分布情况。
(3)细观统计损伤模型
材料损伤力学研究表明,当材料发生变形破坏时,细观损伤的出现早于宏观裂纹,并且这些细观损伤将会影响材料与结构的强度的寿命。从这个角度出发,可以建立损伤模型来模拟损伤过程
(2)
式中:为视压平均应力;为有效应力;E为原状材料的弹性模量;D为损伤变量,当D=0时,对应无损伤的状态;0lt;Dlt;1时,对应不同程度的损伤;D=1时,对应完全损伤的状态。
考虑基元强度为Weibull分布时材料受单轴力作用下应力应变关系,材料数值模拟的本构方程为(为初始应变,为应变)
(3)
2.2 测试结果
根据PMLAB DIC-3D仪器观察,AC-25界面在加载到-0.07mm时,其界面上下两端产生很微小裂纹,但此时人为观察不到其界面的微小裂纹。此时AC-25界面上下两段裂纹处的X、Y和第一主应变如表2所示,界面上下两端处X向应变、Y向应变、第一主应变云图如。
表2界面应变状态(位移-0.07mm)
位移荷载(mm) |
(mu;ɛ) |
(mu;ɛ) |
(mu;ɛ) |
-0.07 |
495 |
-2973 |
583 |
图9 X向应变、Y向应变、第一主应变云图(位移-0.07mm)
AC-25试件加载到-0.19mm时,其界面产生较大的裂纹,如图10,此时可以观察到宏观裂缝(图11)。AC-25中间裂纹的X、Y和第一主应变见表3,界面中间裂纹X向应变、Y向应变和第一主应变如图12。
表3 宏观裂缝时应变状态(位移-0.19mm)
位移荷载(mm) |
(mu;ɛ) |
(mu;ɛ) |
(mu;ɛ) |
-0.19 |
1918 |
-10682 |
2574 |
图10 加载到-0.19mm时计算裂纹 图11 加载到-0.19mm时原始裂纹
图12 X向应变、Y向应变、第一主应变云图(位移-0.19mm)
3分析与讨论
根据AC-25细观光学实验,AC-25界面在加载到-0.07mm时界面开始出现很微小裂纹,在加载到-0.19mm时界面完全开裂。为了进行有限元模拟与试验结果对比分析,下面分别以加载到-0.07mm(裂纹产生时)和-0.19mm(显著开裂时)时界面的应变状态进行对比分析。
3.1.界面裂缝产生时应变状态对比分析
AC-25界面加载到-0.07mm时,试验界面有微小裂纹产生,此时有限元模拟细观界面和试验上下两端裂纹处的应变状态如表4,对比如图13。
表4 -0.07mm时的有限元模拟与试验应变
AC-25界面 |
(mu;ɛ) |
(mu;ɛ) |
(mu;ɛ) |
|
细观集料 lt;剩余内容已隐藏,支付完成后下载完整资料 Experimental Testing and Finite-Element Modeling of Indirect Tensile Test Based on the Mesoscopic Asphalt Mixture 【Abstract】 In order to analyze mesoscopic damage mechanism of asphalt mixture, AC-25 is distributed to aggregates, mastics and voids by digital image technique. Two-dimensional finite element model is established for the mesoscopic Marshall Specimen based on software ANSYS, and inputting mechanical parameters of each phase medium to simulate indirect tensile (IDT) test of mesoscopic damage. Contactless optical method is used to test the strain of each phase medium of AC-25 Marshall Specimen in IDT tests, and the result is compared with what is computed by FEM model. The result indicates that parameters of each phase medium influence in a large scale in mesoscopic simulation, and strains tested by optical method is higher than that of the simulation, but the strain response of simulation and optical test is of similar laws. Specifically, binder and interface are weaknesses in the damage of asphalt mixture by two kinds of analysis. 【Key Words】 road engineering; asphalt mixture; Marshall indirect tensile test; mesoscopic finite element analysis; mesoscopic optical experiment; strain response 0 Introduction Indirect tensile (IDT) test is widely adopted to evaluate the crack resistance and research the damage mechanism of asphalt mixture. Asphalt concrete is simplified to be isotropic and homogeneous from a macro scale, which is difficult to reveal its damage mechanism. From a meso scale, asphalt mixture is divided into aggregates, asphalt binder and voids. Meanwhile, the effect on IDT tests of geometry and physical properties of each medium is also worthy to be discussed. There have been some scholars to carry out the analysis of mesoscopic damage mechanism of asphalt mixture. Yu Qing-lei applied digital image processing technique and original RFPA2D system, using digital image processing technology and statistical method to characterize the inhomogeneous property of internal aggregates. Then he simulated the process of crack propagation under complex stress state and some phenomenon under uniaxial loading such as nonlinear deformation, stress redistribution and so on. Qiu Yan-jun adopted RFPA realistic failure process analysis system to carry out the numerical simulation of whole process for Marshall IDT tests, and found that cracks tend to disperse with the decrease of homogeneous degree and the higher the degree of homogeneity is, the more the damage of specimen tend to be brittle fracture. By using the finite element software ABAQUS, Pi Yu-hui carried out numerical simulation analysis for IDT tests of asphalt mixture, then studied the stress and strain distribution in the test and its change law, and discussed the effect on the result of simulation of the mesh density and form. Pei Jian-zhong applied the discrete element method to generate the Marshall Specimen of asphalt mixture, and simulated the effect on IDT tests of different particle sizes and different loading rate and the dynamic process of crack formation. Starting from mesoscopic finite element model, Romanowicz Marek simulated model of laminate structure in a simple way by using numerical uniformity technology, and simulated two different failure mechanism of matrix cracking created by internal fiber by proper constitutive equation. In sum, the analysis of damage mechanism of asphalt mixture focused on numerical simulation and lack experimental analysis and verification. This paper combined digital image processing technology with FEM software ANSYS, then established two-dimension finite element model of mesoscopic Marshall Specimen AC-25, and analyzed strain distribution and its variation law in IDT test. Meanwhile, mesoscopic strain is tested by contactless optical method. By comparison, providing reference for the study of the mesoscopic damage mechanism of asphalt mixture. 1 Finite element analysis of mesoscopic Marshall Specimen of asphalt mixture 1.1 Two-dimension mesoscopic finite element model This paper established two-dimension mesoscopic finite element model of Marshall Specimen by using the existing digital image processing technology at home and abroad. Because the original section of AC-25 is complicated, as presented in Fig.1, it was recommended to convert the cross-sectional model by utilizing Win Topo Professional and CAD, then introduced into ANSYS software to analysis. From meso scale, AC-25 is a multi-phase composite consisting of aggregates, asphalt mastics and voids. The finite element model and mesoscopic component are shown in Fig.2.
Fig.1 Original image of AC-25 Fig.2 Mesoscopic FEM model and component specimenrsquo;s interface Mesoscopic model size of AC-25 is the same as the macro Marshall Specimen, which is a circle as a whole, and the radian of the inner surface and specimen are identical. The circlersquo;s diameter is 101.6mm, and there are identical battens whose width are 13mm and thickness are 4mm. PLANE42 element was employed to define element type, and there were 40350 elements and 33549 nodes, as presented in Fig.3. Assumed that aggregates and asphalt mastics are linear elastic, whose physical properties canrsquo;t change with external condition and they are shown in Table 1. Meanwhile, adjacent elements were considered as sharing the nodes. This simulation applied displacement load to the center of upper surface in the upper batten and symmetric boundary conditions were set in the reference point of rigid body in the lower batten, then horizontal and vertical displacement were limited. Fig.3 Mesh of mesoscopic FEM model based on IDT test Table 1. Material Parameter of Mesoscopic FEM Model AC-25
|