新型Ti-Al-Zr-Nb-Mo-Si钛合金热变形行为及基于BP神经网络模型的本构关系研究

通过热压缩模拟实验,研究了一种新型Ti-Al-Zr-Nb-Mo-Si高强度、高弹性模量钛合金在温度为950～1 150℃、应变速率为0.05～1 s~(-1)条件下的流变行为。真应力-真应变曲线表明,变形温度、应变速率对该合金的流变应力影响显著。基于实验数据,利用包含应变参量的双曲正弦型Arrhenius方程和BP人工神经网络模型分别构建了变形参数和流变应力的本构关系,并对两种模型进行了对比评价。结果表明,两种模型的平均相对误差值分别为11.21%和2.163%,整体上均可以较好地预测Ti-Al-Zr-Nb-Mo-Si钛合金热压缩流变应力;但相对传统Arrhenius方程,BP人工神经网络模型具有更高的精度和可靠性。     

Zirlo锆合金高温变形行为及本构关系EI北大核心CSCD

采用Gleeble-3800型热模拟试验机,对Zirlo合金进行等温恒应变速率压缩实验,研究其在变形温度550~700℃,应变速率0.01~10 s^(-1)范围内的热变形行为;并在Arrhenius型双曲正弦函数方程基础上引入应变量,构建了基于应变补偿的Arrhenius本构模型,同时构建了基于位错密度演化加工硬化模型和基于唯象型的软化模型的分段唯象型本构模型。结果表明:Zirlo合金的流变应力随着温度的降低和应变速率的提高而升高,低应变速率下流变应力呈现更高的温度敏感性,流变应力曲线在不同变形条件下分别呈现加工硬化、动态回复、动态再结晶特征。经过误差分析可知,基于应变补偿的Arrhenius本构模型大部分预测值的误差均在15%以内,具有较高的准确性,而分段唯象型本构模型相对平均绝对误差最大值不超过3%,具有97%以上的准确率,可以很好地预测合金的应力-应变曲线,具有良好的拓展性,并且可初步判断曲线类型,具有良好的实用性。     

Ti-6Al-2Zr-1Mo-1V合金热塑性流变行为的变参数Arrhenius模型表征

在Gleeble-1500热模拟试验机上进行多组热压缩试验,获得Ti-6Al-2Zr-1Mo-1V合金在温度1073～1323K、应变速率0.01～10s-1下的真应力-应变数据,基于此分析工艺参数对流变应力演变的影响规律,识别应力-应变曲线的动态再结晶型和动态回复型软化特征及其出现时机。引入不含应变影响的传统Arrhenius型本构方程,将其进一步拓展使用,在不同应变量条件下,通过多元线性回归计算与多项式拟合,求解获得Ti-6Al-2Zr-1Mo-1V合金形变表观激活能Q、材料常数n、α及结构因子A等参量对应变的响应规律,从而建立Ti-6Al-2Zr-1Mo-1V合金含应变、温度及应变速率影响的变参数流变应力本构方程。模型预测值与实验值对比显示最大相对误差为4.55%,最大平均误差为2.19%。     

38MnVS6非调质钢两种高温本构模型的对比

采用Gleeble-1500热模拟实验机对38MnVS6非调质钢在温度为950~1200℃、应变速率为0.01~5s-1进行等温、等应变速率热压缩实验。依据实验所获得的流动应力曲线,分别采用双曲正弦形式的Arrhenius方程和BP人工神经网络模型建立该种钢的高温本构模型。分别统计计算两种模型预测值与实验值之间的相关系数和平均相对误差。结果表明:神经网络模型具有更高的精度,更适合用于该种钢的高温流动应力预测。     

等轴细晶TC4钛合金应变补偿本构关系及热加工图的研究

本工作对等轴细晶TC4钛合金进行了热压缩实验,研究了变形温度为800～950℃、应变速率为0.01～10 s-1下TC4钛合金的变形行为,并建立相应的Arrhenius型本构方程和热加工图,再基于实验获得的真应力应变曲线对本构方程进行应变补偿修正.结果表明:合金的真应力值随温度升高、应变速率下降而减小;修正后本构方程真应力预测值与实验值相关系数R为0.985,相对误差ARRE为6.8％.结合热加工图和相应区域的电子背散射衍射(EBSD)分析可知:失稳区的温度为875～950℃,应变速率为0.3～10 s-1,组织特征表现为长条状晶粒;最适宜加工区的温度为800～875℃,应变速率为0.01～0.3 s-1,组织特征表现为等轴细晶.     

FeCrNiMn高熵合金本构方程的建立

在变形温度为900～1050℃、应变速率为0.001～1 s-1下对FeCrNiMn高熵合金进行热压缩试验.通过真应力-真应变曲线,分析流变应力与变形参数之间的关系.在Arrhenius型双曲正弦方程的基础上建立了FeCrNiMn高熵合金本构方程.误差分析表明,所建立的本构方程与实验值基本吻合,可以作为热变形工艺参数选择的依据.     

新型超高强度热冲压用钢的热变形行为及本构关系

利用Gleeble-1500D热模拟机对新型超高强度热冲压用钢22MnB5Nb进行等温单向拉伸实验,研究了其在变形温度为650~950℃,应变速率为0.1,1.0,10s~(-1)下的热变形行为,并采用3种本构分析方法,即基于传统拟合回归方法的Arrhenius型、考虑材料常数应变补偿的Arrhenius型和本工作新提出的基于Quasi-Newton BFGS算法的Arrhenius型本构方程来描述22MnB5Nb钢的热变形行为。结果表明:22MnB5Nb钢表现出典型的加工硬化和动态回复软化行为,变形温度与应变速率均对其流变应力有较大影响;3种方程均可以准确预测实验钢的峰值流变应力,其中,Quasi-Newton BFGS算法具有可一次性求解所有材料参数、求解步骤简单和预测精度最高(R=0.99578,Re=11.03MPa,E=2.48%)的特点,考虑材料常数应变补偿的Arrhenius型本构方程预测精度相对较低,但能直接预测不同变形条件下的流变应力曲线且可以较好地预测变形过程中的加工硬化效应、动态回复软化效应和应变速率强化效应。     

一种新型高淬透性Ni-Cr-Mo-B钢的热变形本构分析

基于Gleeble单道次热压缩实验并结合OM表征手段及回归方法,研究了一种新型高淬透性Ni-Cr-Mo-B特厚板钢在850～1 150℃和0. 01～10 s-1热变形参数下的热变形行为。结果表明,Ni-Cr-Mo-B钢在热变形过程中流变应力随变形温度的升高而减小,随应变速率的增加而增大。建立的典型应变补偿Arrhenius型本构方程可用于粗略预测Ni-Cr-Mo-B钢的流变行为。考虑到应变速率、变形热对变形过程的影响,提出修正应变补偿Arrhenius型本构方程,其统计学参量Rc=0. 994 68、AARE=3. 69%、RMSE=6. 44 MPa、NMBE=1. 25%,相对误差大部分(94. 87%)在±10%之内,表明该方程具有极高的流变应力预测精度。     

A100钢的热变形行为及加工图

目的 研究A100钢的热变形行为,确定热加工范围并优化工艺参数.方法 使用Gleeble-3800热模拟实验机,对A100钢进行应变为0.6,变形温度为1073～1473 K,应变速率为0.01～10 s–1的等温热压缩实验.利用A100钢的热压缩实验数据,建立在不同变形温度、不同应变速率下的真应力-真应变曲线.建立A100钢基于唯象的本构模型与基于物理的本构模型以及基于Murty失稳准则的热加工图.结果 当应变速率一定,温度升高或一定,应变速率下降时,A100钢的流变应力会减小,流变应力曲线上主要表现为动态再结晶的软化机制.结论 构建的基于唯象的本构方程可以对A100钢在应变为0.6时的流变应力进行较好的预测,基于物理的本构方程可以反映出A100钢的物理特性,通过构建的基于Murty失稳准则的加工图可以得到A100钢的加工范围是温度为1173～1223 K,应变速率为0.01～0.1 s–1和温度为1323～1373 K,应变速率为0.05～0.15 s–1时.     

3003铝合金热变形流变应力及动态再结晶模型

在变形温度为300～500℃,应变速率为0.01～10.0s~(-1)的条件下,通过Gleeble-1500热模拟试验机对3003铝合金进行高温等温压缩实验。结果表明,该合金在热变形过程中的峰值流变应力可用双曲正弦本构方程来描述,由本构方程计算获得模型的流变应力预测值和实测值的相对误差在±7%范围以内。根据热力学不可逆原理确定动态再结晶临界应变,建立动态再结晶开始时间与变形温度关系的RTT(Recrystallization Start Time)图,研究表明:动态再结晶开始时间随着应变速率的减小与变形温度的降低而增大,由流变应力曲线计算动态再结晶体积比例,其大小随变形温度的升高和应变速率的减小而增大,并获得3003铝合金动态再结晶体积分数数学模型。     

A comparative study on modified Johnson Cook,modified Zerilli–Armstrong and Arrhenius-type constitutive models to predict the hot deformation behavior in 28CrMnMoV steel

The true stress-strain data from isothermal hot compression tests on Gleeble-3500 thermo mechanical simulator, in a wide range of temperatures (1173–1473 K) and strain rates (0.01–10 s⁻¹), were employed to establish the constitutive equations based on modified Johnson Cook, modified Zerilli–Armstrong, and strain-compensated Arrhenius-type models respectively to predict the high-temperature flow stress of 28CrMnMoV steel. Furthermore, a comparative study has been made on the capability of the three models to represent the elevated temperature flow behavior of this steel. Suitability of the three models were evaluated by comparing the accuracy of prediction of deformation behavior, correlation coefficient, average absolute relative error (AARE) and relative errors of prediction, the number of material constants, and the time needed to evaluate these constants. The results showed that the predicted values by the modified Johnson Cook and Zerilli–Armstrong models could agree well with the experimental values except under the strain rate of 0.01 s⁻¹. However, the strain-compensated Arrhenius-type model could track the deformation behavior more accurately throughout the entire temperature and strain rate range.     

A comparative study on the capability of Johnson–Cook and Arrhenius-type constitutive equations to describe the flow behavior of Mg–6Al–1Zn alloy

In the current study, the predictability of two phenomenological constitutive equations, Johnson–Cook (JC) and Arrhenius-type ones, for describing the flow behavior of a magnesium alloy (Mg–6Al–1Zn) under hot deformation conditions has been evaluated. Towards this end, a series of hot compression tests were performed over a temperature range of 250–450 °C, under strain rates of 0.001, 0.01 and 0.1 s⁻¹. Using the experimental results obtained through implementing the predetermined compression tests, the related parameters and material constants in the constitutive equations were calculated. In order to compare the performance of the models, the statistical parameters of correlation coefficient and absolute mean error were employed. The results imply that the predictability of the Arrhenius-type equations is much stronger in estimating the flow behavior compared to that of the JC model; although more constants are needed to be calculated when using the former equation. It is concluded that the JC model, in contrast to the Arrhenius-type equations, is not reliable for the materials possessing tangible softening stage in their stress–strain curves such as magnesium AZ series.     

Microstructural modeling and simulation for GCr15 steel during elevated temperature deformation

The microstructural evolution of GCr15 steel, one of the most commonly used bearing steels, was investigated and simulated by physical experiments and finite element method (FEM). Physical experiments were conducted on the Gleeble-3500 thermo-simulation system. Effects of initial grain size and plastic strain on the microstructural of the materials were investigated by setting different heating temperature, holding time and deformation degree, respectively. Based on the results of stress–strain curves and metallographic analysis, the constitutive equations for flow stress, austenite grain growth and dynamic recrystallization of GCr15 steel were formulated by linear regression method and genetic algorithm. In addition, the coupled thermo-mechanical finite element method integrated with the developed constitutive models was used to simulate the microstructural evolution of GCr15 steel during hot compression. Good agreement between the calculated and experimental results was obtained, which confirmed that the developed constitutive models can be successfully used to predict microstructural evolution during hot deformation process for GCr15 steel.     

Artificial neural network modeling to predict the high temperature flow behavior of an AZ81 magnesium alloy

In the present work, the capability of artificial neural network (ANN) has been evaluated to describe and to predict the high temperature flow behavior of a cast AZ81 magnesium alloy. Toward this end, a set of isothermal hot compression tests were carried out in temperature range of 250–400 °C and strain rates of 0.0001, 0.001 and 0.01 s⁻¹ up to a true strain of 0.6. The flow stress was primarily predicted by the hyperbolic laws in an Arrhenius-type of constitutive equation considering the effects of strain, strain rate and temperature. Then, a feed-forward back propagation artificial neural network with single hidden layer was established to investigate the flow behavior of the material. The neural network has been trained with an in-house database obtained from hot compression tests. The performance of the proposed models has been evaluated using a wide variety of statistical indices. The comparative assessment of the results indicates that the trained ANN model is more efficient and accurate in predicting the hot compressive behavior of cast AZ81 magnesium alloy than the constitutive equations.     

Finite element analysis and experimental study of microstructure evolution of nitrogen alloyed ultralow carbon stainless steel during hot deformation

Abstract

Hot compression experiments of a nitrogen alloyed ultralow carbon stainless steel were performed in the temperature range of 1223–1423 K, at strain rates of 0.001–1 s^?1, and with deformation amounts of 30–70% on a Gleeble-3500 thermal-simulator. Based on the results from thermo-physical simulation experiments and metallographic analyses, a physically-based constitutive model and a dynamic recrystallisation (DRX) model of the studied steel were derived, and the developed models were further embedded into a finite element method (FEM) software. The microstructure evolution of the studied steel under various hot deformation conditions was simulated by FEM, and the effects of deformation amount, strain rate and temperature on the microstructure evolution were clarified. The results obtained from the finite element analysis were verified by the experiments. The finding confirms that the thermal-mechanical FEM coupled with the developed constitutive model and DRX model can be used to accurately predict the microstructure evolution of the studied steel during hot deformation.     

基于动态再结晶37CrS4特种钢的流变应力预测模型

使用Gleeble-1500D热模拟实验机对37CrS4特种钢进行单道次热压缩实验,研究了37CrS4钢在950~1100℃和0.01 s^-1~10 s^-1条件下的热压缩流变应力行为。结果表明：这种钢的真应力应变曲线出现了明显的高温塑性变形动态再结晶行为;热变形后的微观组织为典型的板条状马氏体,发生动态再结晶行为的临界应变值与峰值应变比值为0.77162,拟合相关性R²=0.9576;其软化机制为动态回复与动态再结晶的共同作用。引入Zener-Hollomon参数(Z参数)建立再结晶动力学模型,得到了37CrS4特种钢基于动态回复、动态再结晶的分段式流变应力本构模型。本构模型的平均相关性R²=0.9756,分段式本构模型的预测应力与实验应力具有较高的一致性,能较为准确的预测37CrS4高温塑性变形时流变应力的变化。     

Constitutive modelling for high temperature behavior of 1Cr12Ni3Mo2VNbN martensitic steel

The deformation behavior of 1Cr12Ni3Mo2VNbN martensitic steel in the temperature range of 1253 and 1453 K and the strain rate range of 0.01 and 10 s⁻¹ are investigated by isothermal compression tests on a Gleeble 1500 thermal-mechanics simulator. Most of the stress-strain curves exhibit a single peak stress, after which the stress gradually decreases until a steady state stress occurs, indicating a typical dynamic recrystallization (DRX) behavior of the steel under the deformation conditions. The experimental data are employed to develop constitutive equations on the basis of the Arrhenius-type equation. In the constitutive equations, the effect of the strain on the deformation behavior is incorporated and the effects of the deformation temperature and strain rate are represented by the Zener-Holloman parameter. The flow stress predicted by the constitutive equations shows good agreement with the experimental stress, which validates the efficiency of the constitutive equations in describing the deformation behavior of the material.     

高强钢TRB高温下热流变特性

为了准确仿真高强钢板热冲压成形过程,获得高强钢高温下的材料本构关系模型,利用Gleeble3500热模拟试验机在不同温度和应变速率下对不同厚度的高强钢B1500HS钢板进行了单向拉伸试验,获得各种工艺条件下的应力-应变曲线,并基于变形抗力数学模型,引入板材厚度参数,通过最小二乘法进行数据拟合获得高强钢TRB高温下的材料本构关系.利用试验结果对本构关系模型进行的拟合验证表明,拟合程度较好,说明建立的材料本构关系能很好地描述高强钢TRB在高温下的应力-应变关系.     

具有层片状α相组织的TB8钛合金热变形行为及本构方程

主要研究具有层片状α相组织的TB8钛合金在α+β双相区的热变形行为。结果表明,在应变速率为1s-1时,变形温度为650℃的流变曲线展现出连续的流变软化,当温度高于650℃时,流变曲线呈现出不连续屈服现象。不连续屈服现象随变形温度的增加和应变速率的降低而消失。当应变速率为0.001s-1时,750℃和800℃的流变曲线呈现出典型的动态再结晶特征。峰值应力σp,温度T和应变速率ε·三者之间的关系已通过Arrhenius-type本构方程进行表征,建立了材料常数α,A,n和Q值与真应变之间的关系模型,并分析了应变对α,A,n和Q值的影响。α值随真应变的增加而增加,而A,n和Q的值随真应变的增加而逐渐降低。实验应力值和预测应力值之间的相关系数和平均相对误差参数分别为0.945和9.08%。这表明本工作建立的应变补偿的热变形本构方程能够很好地预测具有层片状α相组织的TB8钛合金在α+β双相区热变形过程中的流变应力。