准静态条件下金属材料的临界断裂准则研究

本文针对9种金属材料完成了具有不同约束程度的10类试样的延性断裂试验, 获得了发生拉、压、扭和裂尖断裂等破坏形式构型试样的载荷-位移试验关系; 基于圆棒漏斗试样拉伸试验所得直至破坏的载荷-位移曲线, 采用有限元辅助试验(finite-element-analysis aided testing, FAT)方法得到了9种材料直至破坏的全程等效应力-应变曲线, 以此作为材料本构关系通过有限元分析获得了各类试样直至临界破坏的载荷-位移关系模拟. 载荷-位移关系模拟结果与试验结果有较好的一致性, 表明用于解决试样颈缩问题的FAT方法所获得的全程材料本构关系针对各向同性材料具有真实性和普适性. 对应9种材料、10类试样的36 个载荷-位移临界断裂点, 通过有限元分析获得了对应的材料临界断裂应力、应变与临界应力三轴度, 研究表明, 第一主应力在延性变形过程中为主控断裂的主导参量; 通过研究光滑、缺口、裂纹等构型试样的断裂状态, 提出了$-1$至3范围的应力三轴度下由第一主应力主控的统一塑性临界断裂准则.      

准静态条件下金属材料的临界断裂准则研究

本文针对9种金属材料完成了具有不同约束程度的10类试样的延性断裂试验,获得了发生拉、压、扭和裂尖断裂等破坏形式构型试样的载荷-位移试验关系;基于圆棒漏斗试样拉伸试验所得直至破坏的载荷-位移曲线,采用有限元辅助试验(finite-element-analysis aided testing, FAT)方法得到了9种材料直至破坏的全程等效应力-应变曲线,以此作为材料本构关系通过有限元分析获得了各类试样直至临界破坏的载荷-位移关系模拟.载荷-位移关系模拟结果与试验结果有较好的一致性,表明用于解决试样颈缩问题的FAT方法所获得的全程材料本构关系针对各向同性材料具有真实性和普适性.对应9种材料、10类试样的36个载荷-位移临界断裂点,通过有限元分析获得了对应的材料临界断裂应力、应变与临界应力三轴度,研究表明,第一主应力在延性变形过程中为主控断裂的主导参量;通过研究光滑、缺口、裂纹等构型试样的断裂状态,提出了-1至3范围的应力三轴度下由第一主应力主控的统一塑性临界断裂准则.     

Cr15Mn9Cu2Ni1N不锈钢的高温拉伸变形特性分析

利用热/力模拟试验机,对Cr15Mn9Cu2Ni1N不锈钢进行了950℃~1200℃高温范围内的拉伸试验;采用有限元方法对试样的均匀变形过程进行了分析。高温拉伸过程中,试样在达到最大应力后并不立即颈缩,而是还要经历一段宏观均匀变形后才颈缩。分析结果表明:最大应力之后,试样端部区域等效应力降低,横截面积收缩量减小,而中心区域横截面积收缩量增大,形成了潜在颈缩区;在应变速率敏感性的作用下,潜在颈缩区的变形抗力随应变速率的增大而增加,使变形不能在该区域集中,而转向其它位置,保持了试样的宏观均匀变形,且颈缩未在最大应力后立即发生;在高温拉伸条件下,材料应变速率敏感性的增大是颈缩延迟发生的主要原因,随着变形温度升高,应变速率敏感性增大,也使得试样颈缩前的均匀变形量增大。     

0Cr18Ni9不锈钢材料断裂韧度的实验研究和数值分析

从实验研究和数值分析两方面研究了0Cr18Ni9不锈钢材料的断裂韧度。首先,通过单轴拉伸测试得到0Cr18Ni9不锈钢材料的应力-应变曲线,并验证了该应力-应变曲线符合Ramberg-Osgood弹塑性本构模型,利用全局优化算法反演得到其弹塑性本构参数。然后,结合声发射技术和扫描电镜断口观察,确定了0Cr18Ni9不锈钢紧凑拉伸试样的起裂载荷和I型裂纹实际的起裂断裂韧度。进一步,根据数字图像相关方法得到的全场变形计算了J积分,J积分随时间逐渐增大;最后,采用有限元法计算了起裂时沿裂纹前缘各处的J积分分布,发现J积分呈现中间较大、表面较小的分布特征,而且有限元计算的J积分平均值与从光学位移场数据计算的J积分接近。     

基于BP神经网络与小冲杆试验确定在役管道钢弹塑性性能方法研究

为了能够在不停输油气工况下获得在役管道材料的弹塑性力学性能,提出了一种人工智能BP (backpropagation)神经网络、小冲杆试验与有限元模拟相结合,通过确定材料真应力-应变曲线从而获得材料弹塑性力学性能的方法.首先,通过系统改变Hollomon公式中的参数K, n值,获得457组具有不同弹塑性力学性能的假想材料本构关系,其次,将得到的本构关系代入经试验验证的含有Gurson-Tvergaard-Needleman(GTN)损伤参数的小冲杆试验二维轴对称有限元模型,通过有限元计算得到了与真应力-应变曲线一一对应的457条不同假想材料的载荷-位移曲线,最终将两组数据作为数据库输入BP神经网络进行训练,建立了同种材料小冲杆试验载荷-位移曲线与真应力-应变曲线之间的关联关系.通过此关联关系,可利用试验得到的小冲杆载荷-位移曲线获取在役管道钢的真应力-应变曲线,从而确定其弹塑性力学性能.通过对比BP神经网络得到的X80管道钢真应力-应变曲线与单轴拉伸试验的结果以及引用现有文献中不同材料的试验数据对此关系进行验证,证明了该方法的准确性与广泛适用性.     

基于BP神经网络与小冲杆试验确定在役管道钢弹塑性性能方法研究

为了能够在不停输油气工况下获得在役管道材料的弹塑性力学性能, 提出了一种人工智能BP (back-propagation)神经网络、小冲杆试验与有限元模拟相结合,通过确定材料真应力-应变曲线从而获得材料弹塑性力学性能的方法. 首先,通过系统改变Hollomon公式中的参数$K$, $n$值,获得457组具有不同弹塑性力学性能的假想材料本构关系, 其次,将得到的本构关系代入经试验验证的含有Gurson-Tvergaard-Needleman(GTN)损伤参数的小冲杆试验二维轴对称有限元模型,通过有限元计算得到了与真应力-应变曲线一一对应的457条不同假想材料的载荷-位移曲线,最终将两组数据作为数据库输入BP神经网络进行训练,建立了同种材料小冲杆试验载荷-位移曲线与真应力-应变曲线之间的关联关系.通过此关联关系,可利用试验得到的小冲杆载荷-位移曲线获取在役管道钢的真应力-应变曲线,从而确定其弹塑性力学性能.通过对比BP神经网络得到的X80管道钢真应力-应变曲线与单轴拉伸试验的结果以及引用现有文献中不同材料的试验数据对此关系进行验证,证明了该方法的准确性与广泛适用性.      

半结晶聚合物损伤演化的实验表征与数值模拟

损伤本构模型对研究材料的断裂失效行为有重要意义, 但聚合物材料损伤演化的定量表征实验研究相对匮乏. 通过4种高密度聚乙烯(high density polythylene, HDPE)缺口圆棒试样的单轴拉伸实验获得了各类试样的载荷-位移曲线和真应力-应变曲线, 采用实验和有限元模拟相结合的方法确定了HDPE材料不同应力状态下的本构关系, 并建立了缺口半径与应力三轴度之间的关系;采用两阶段实验法定量描述了4种HDPE试样单轴拉伸过程中的弹性模量变化, 并建立了基于弹性模量衰减的损伤演化方程, 结合中断实验和扫描电子显微镜分析了应力状态对HDPE材料微观结构演化的影响. 结果表明缺口半径越小, 应力三轴度越大, 损伤起始越早、演化越快; 微观表现为: 高应力三轴度促进孔洞的萌生和发展, 但抑制纤维状结构的产生;基于实验和有限元模拟获得的断裂应变、应力三轴度、损伤演化方程等信息提出了一种适用于聚合物的损伤模型参数确定方法, 最后将本文获得的本构关系和损伤模型用于HDPE平板的冲压成形模拟, 模拟结果与实验结果吻合良好.      

形状记忆合金伪弹性行为模拟

基于对伪弹性形状记忆合金(SMA)典型应力-应变曲线的特征分析,在原Graesser本构模型中增加简洁多项式来描述SMA应力诱发马氏体相变完成后在变形马氏体相下继续加载阶段的变形特征;并引入应变幅值与混相下SMA弹性模量的关系来改进不同应变幅值下卸载时SMA的应力-应变关系,从而提出了一种新的SMA一维本构关系模拟其伪弹性力学行为。该模型对直径为0.5mm的NiTi合金丝的拉伸加载、卸载试验曲线的模拟结果表明:改进本构模型与原Graesser模型相比,其能够准确地模拟SMA在不同应变幅值下加载和卸载应力-应变关系。此外,通过研究SMA本构模型的物理关系,推导出了控制SMA滞回曲线特征的关键参数fT与相变临界应力、弹性常数之间的明确关系,可利用该关系直接确定参数fT,摆脱了只靠试算获取该参数的传统做法,其准确性得到了试验验证。     

正交各向异性韧性材料应力-应变关系

采用大变形弹塑性有限元方法分析了各向同性和正交各向异性韧性材料光滑圆棒拉伸试件的颈缩问题．首先给出了采用计算机模拟确定各向同性韧性材料真实应力-应变曲线的具体方法；对正交各向异性韧性材料的分析表明，颈缩截面呈椭圆形，其长短轴方向的等效塑性应变基本上均匀分布，与Bridgman假设一致；轴向拉伸载荷-位移曲线与其它两方向的各向异性参数关系不大．在此基础上，建议了一种确定正交各向异性韧性材料真实应力-应变曲线的方法．     

2D-C/SiC复合材料的单轴拉伸力学行为及其强度

通过单调拉伸和循环加卸载试验, 研究了平纹编织C/SiC复合材料的损伤演化过程及其应力-应变行为. 结果表明, 残余应变、卸载模量和外加应力的关系曲线与拉伸应力-应变曲线具有类似的形状. 基于剪滞理论和混合率建立了材料的损伤本构关系和强度模型, 分析计算表明, 残余应变主要由裂纹张开位移和裂纹间距决定, 而卸载模量主要由界面脱粘率决定; 材料的单轴拉伸行为主要由纵向纤维束决定, 横向纤维对材料的整体模量和强度贡献较小. 理论模拟结果与试验值吻合较好.      

Obtaining a Wide-Strain-Range True Stress–Strain Curve Using the Measurement-In-Neck-Section Method

Background

Measuring true stress–strain curve over a large-strain-range is essential to understand mechanical behavior and simulate non-linear plastic deformation. The digital image correlation (DIC) technique, a non-contact full-field optical measurement technique, is a promising candidate to obtain a long-range true stress–strain curve experimentally.

Objective

This paper proposes a method for measuring true stress–strain curves over a large-strain-range during tensile testing using DIC.

Methods

The wide-strain-range true stress–strain curves of dual-phase and low carbon steels were extracted on the transverse direction in the neck region. The axial strain on the neck section was estimated by averaging the inhomogeneous deformation on the cross-section of the tensile specimen. The true stress was calculated from the engineering stress and the cross-sectional area of the neck.

Results

The validity of the proposed method was assessed by comparing the experimental load–displacement responses during tensile testing with the finite element method (FEM) simulation results. The stress and strain on the neck section estimated using the FEM and DIC, respectively, were proven to satisfy the uniaxial condition and successfully obtained.

Conclusions

The experimental results agree well with the FEM results. The proposed concept can be applied to various deformation modes for accurately measuring long-range true stress–strain curves.

     

一种从离散模拟到连续介质弹性模拟的过渡方法

提出了一种从离散分子动力学模拟(MD)到连续介质弹性有限元计算分析(FEA)的过渡方法, 简称MD-FEA方法. 首先通过MD计算获得晶体材料原子的移动位置, 然后根据晶体结构的周期性特征构造连续介质假设下的有限单元变形模型, 进一步结合材料的力学行为本构关系获得应变和应力场. 为了检验MD-FEA方法的有效性, 将该方法应用于详细分析Al-Ni软硬组合两相材料纳米柱体的拉伸变形问题和基底材料为Al球形压头材料为金刚石的纳米压痕问题. 采用MD-FEA方法获得了上述两种问题的应力?应变场, 并将计算结果分别与传统MD方法中通过变形梯度计算的原子应变以及原子的位力应力进行了比较, 详细讨论了用MD-FEA方法计算的应力?应变场与传统MD原子应变和位力应力的区别, 并对MD-FEA方法的有效性及其相较于传统MD方法所具有的优势进行了探讨. 结论显示, MD-FEA方法与传统MD方法在应力?应变变化平缓的区域得到的结果接近, 但在变化剧烈的区域以及材料的表/界面区域, MD-FEA方法能够得到更加精确的结果. 同时, MD-FEA方法避免了传统MD方法中, 需要人为选取截断半径以及加权函数所导致的误差. 另外, 当应变较大时, MD-FEA方法计算的小应变与传统MD方法计算的格林应变存在一定差异, 因此, MD-FEA方法更适合应变较小的情形.      

A computational micromechanics constitutive model for the unloading behavior of paper

In this study, a computational micromechanics material model for the unloading behavior of paper and other nonwoven materials is presented. The asymptotic fiber and bond (AFB) model for paper elastic–plastic behavior [Sinha, S.K., Perkins, R.W., 1995. Micromechanics constitutive model for use in finite element analysis, In: Proceedings of the 1995, Joint ASME Applied Mechanics and Materials Summer Meeting, Los Angeles, CA, USA, Jun 28–30, 1995] has been extended to model the unloading process through a computational algorithm and implemented using the UMAT subroutine in ABAQUS finite element code. For every unloading increment, the material model assumes elastic unloading with a slope equal to the initial elastic modulus. The Jacobian matrix of the constitutive model is updated at every unloading increment by applying the incremental form of AFB model for a planar element with an elastic fiber and bond condition. A uniaxial tensile and a biaxial Mullen burst loading–unloading experiments were carried out for a paperboard sample and simulated using the model. The stress–strain curve and residual strain for the uniaxial loading were in good agreement with experimental results. The finite element model of the burst test with the AFB unloading material model predicted the general shape of the pressure versus deflection curve. However, the model over predicted the residual deflection by more than 50%. The loading portion of the pressure–deflection curve had a significant offset from experimental curves, and the nonlinearity in the unloading curve towards the end was not predicted. The discrepancies with experimental results are attributed to the burst test itself, model parameter estimation inadequacies, boundary conditions used in the FEA, and neglecting time-dependant effects. Nevertheless, the model can be useful in parametric studies relating microstructure to unloading behavior in structural problems.     

Sequential evaluation of continuous deformation field of semi-crystalline polymers during tensile deformation accompanied by neck propagation

The localized deformation field of high density polyethylene and polypropylene during a tensile test accompanied by neck propagation was quantitatively evaluated based on the network digital image correlation method. In the proposed method, the continuity of the deformation field around a point of interest was introduced for accurate evaluation of the displacement. The accuracy of the proposed method was verified through test images. Using the proposed method, the development of a non-uniform displacement field during tensile tests was evaluated from sequential digital images. The local strain rate was almost uniform until the nominal stress reached its maximum value. After the maximum stress was reached, non-uniform deformation developed at a part of the gauge region of the specimen. A decrease in nominal stress induced a reduction of the local strain rate at regions other than the necked zone. In this study, the cross section average local true stress, strain, and strain rate can be evaluated from the local displacement field. Thus, the relationship between these quantities was evaluated during the tensile tests. Using the proposed method, the local response under wide ranges of strain and strain rate can be evaluated from a few test conditions of tensile strain rate and a small range of tensile strain. Finally, the relationships between gradients of stress, strain, and strain rate under uniaxial tension are discussed. These non-local quantities deviated from those predicted by constitutive equations when the domain size used to evaluate the local quantities was large.     

Evaluation of the stress–strain curve of metallic materials by spherical indentation

A method for deducing the stress–strain uniaxial properties of metallic materials from instrumented spherical indentation is presented along with an experimental verification.An extensive finite element parametric analysis of the spherical indentation was performed in order to generate a database of load vs. depth of penetration curves for classes of materials selected in order to represent the metals commonly employed in structural applications. The stress–strain curves of the materials were represented with three parameters: the Young modulus for the elastic regime, the stress of proportionality limit and the strain-hardening coefficient for the elastic–plastic regime.The indentation curves simulated by the finite element analyses were fitted in order to obtain a continuous function which can produce accurate load vs. depth curves for any combination of the constitutive elastic–plastic parameters. On the basis of this continuous function, an optimization algorithm was then employed to deduce the material elastic–plastic parameters and the related stress–strain curve when the measured load vs. depth curve is available by an instrumented spherical indentation test.The proposed method was verified by comparing the predicted stress–strain curves with those directly measured for several metallic alloys having different mechanical properties.This result confirms the possibility to deduce the complete stress–strain curve of a metal alloy with good accuracy by a properly conducted instrumented spherical indentation test and a suitable interpretation technique of the measured quantities.     

Dynamic strain aging and related instabilities: experimental,theoretical and numerical aspects

Experimental data from uniaxial tensile tests on smooth and notched specimens of aluminium alloy 5083-H116 show that the material exhibits negative strain-rate sensitivity for strain rates within a certain range. The negative strain-rate dependence, which is attributed to dynamic strain aging, leads to serrated stress–strain curves, discontinuous plastic flow and propagating deformation bands during plastic straining (also denoted as the Portevin–Le Chatelier effect). Band analysis and linear perturbation analysis are performed using simple elastic-viscoplastic constitutive equations that include negative strain-rate sensitivity in a simplified manner. The negative strain-rate sensitivity allows for jumps in the plastic strain rate, which in turn permits the existence of localisation bands for the elastic-viscoplastic model. The simple elastic-viscoplastic constitutive model has been implemented in LS-DYNA, and non-linear finite element simulations of smooth and notched tensile test specimens are performed, allowing more detailed investigations into the effects of the negative strain-rate sensitivity on the material's behaviour.     

A 3-phase model for the numerical analysis of semi-crystalline polymer films in finite elastoplastic strains

In this paper, the mechanical behavior of semi-crystalline polymer films in finite elastoplastic strains is investigated. A 3-phase constitutive model has been specially developed in a previous paper and validated for various materials in both uniaxial and biaxial uniform hot drawing. In the present study, the numerical implementation of this 3-phase model in a finite element software is outlined in the perspective of using this model in more general non-uniform cases of complex geometries and/or loadings. In the present case, only polyethylene films at room temperature are considered. First, uniaxial tensile experimental tests are performed so as to calibrate the model parameters. Then, for validation purposes, two series of experimental tests are conducted on tensile specimens with central holes and double edge notched tensile (DENT) specimens. During these tests, digital image correlation is used to analyze the strain (or displacement) field history during loading. Finally, numerical computations are performed with the help of the finite element software including the 3-phase model previously implemented (cohesive elements are also needed for the simulation of the crack propagation in DENT specimens). In both cases, the comparison between the experimental and numerical force–displacement curves, together with the comparisons between the experimental and numerical strain fields at different times, give very satisfactory results.