Artificial Neural Network Model for Predicting Stable and Unstable Regions in Cu-Zn Alloys

Processing maps are developed using the Dynamic Materials Model (DMM) and instability criterion, which help in choosing optimum process parameters for hot-working of materials. Certain high-level expertise is required to interpret and extract the information on instability regimes to be avoided during processing. In recent years, Artificial Neural Network (ANN) models have been developed to predict flow stress by using the input vector; namely, temperature, strain rate and strain. In this study, using the available Cu-Zn alloy data, ANN model has been developed to classify the hot-working process parameters, such as temperature, strain rate and flow stress for instability regime, directly from the corrected flow stress data without applying the DMM. This model uses 10 compositions of Cu-Zn system, ranging from 3% Zn to 51% Zn. The developed ANN model has been able to learn the nonlinear classifier, which separates unstable region from the stable region in the Cu-Zn alloy system with zinc content less than 40%.     

Artificial Neural Network Model for Predicting Stable and Unstable Regions in Cu–Zn Alloys

Processing maps are developed using the Dynamic Materials Model (DMM) and instability criterion, which help in choosing optimum process parameters for hot-working of materials. Certain high-level expertise is required to interpret and extract the information on instability regimes to be avoided during processing. In recent years, Artificial Neural Network (ANN) models have been developed to predict flow stress by using the input vector; namely, temperature, strain rate and strain. In this study, using the available Cu–Zn alloy data, ANN model has been developed to classify the hot-working process parameters, such as temperature, strain rate and flow stress for instability regime, directly from the corrected flow stress data without applying the DMM. This model uses 10 compositions of Cu–Zn system, ranging from 3% Zn to 51% Zn. The developed ANN model has been able to learn the nonlinear classifier, which separates unstable region from the stable region in the Cu–Zn alloy system with zinc content less than 40%.     

Cr-Mo-B系机械工程用钢高温流变行为及热加工图

在变形温度为850~1150℃、应变速率为0.1~10s -1 的条件下,对Cr-Mo-B系机械工程用钢进行高温热压缩实验。基于真应力-应变曲线,建立输入参数为温度、变形速率、应变和输出参数为流变应力的人工神经网络(ANN)模型。结果表明:神经网络模型的预测精度高,其预测流变应力的均方根误差为1.3858。根据动态材料模型理论(DMM),构建并分析材料在真应变为0.5和0.7时的热加工图,确定了最佳热变形工艺参数:当真应变ε=0.5时,变形温度为1050~1150℃、应变速率为0.1~0.4s -1 区域的功率耗散因子η≥37.20%;当真应变ε=0.7时,变形温度为1000~1150℃、应变速率为0.1~0.6s -1 区域的功率耗散因子η≥35.80%。     

The workability evaluation of wrought AZ80 magnesium alloy in hot compression

Workability, an important parameter in metal forming process, can be evaluated by means of processing maps on the basis of dynamic materials model (DMM), constructed from experimentally generated flow stress variation with respect to strain, strain rate and temperature. To obtain the processing maps of wrought AZ80 magnesium alloy, hot compression tests were performed over a range of temperatures 523–673 K and strain rates 0.01–10 s⁻¹. As the true strain is 0.25, 0.45, 0.65, 0.85 respectively, the response of strain-rate sensitivity (m-value), power dissipation efficiency (η-value) and instability parameter (ξ-value) to temperature and strain rate were evaluated. By the superimposition of the power dissipation and the instability maps, the stable, metastable and unstable regions were clarified clearly. In further, in the stable area the regions having the highest efficiency of power dissipation were identified and recommended. The optimal working parameters identified by the processing maps contribute to designing the hot forming process of AZ80 magnesium alloy without any defect.     

Modeling constitutive relationship of Ti40 alloy using artificial neural network

Constitutive relationship equation reflects the highly non-linear relationship of flow stress as function of strain, strain rate and temperature. It is a necessary mathematical model that describes basic information of materials deformation and finite element simulation. In this paper, based on the experimental data obtained from Gleeble-1500 Thermal Simulator, the constitutive relationship model for Ti40 alloy has been developed using back propagation (BP) neural network. The predicted flow stress values were compared with the experimental values. It was found that the absolute relative error between predicted and experimental data is less than 8.0%, which shows that predicted flow stress by artificial neural network (ANN) model is in good agreement with experimental results. Moreover, the ANN model could describe the whole deforming process better, indicating that the present model can provide a convenient and effective way to establish the constitutive relationship for Ti40 alloy.     

Application of neural networks to predict the elevated temperature flow behavior of a low alloy steel

In order to study the workability and establish the optimum hot forming processing parameters for 42CrMo steel, the compressive deformation behavior of 42CrMo steel was investigated at the temperatures from 850 °C to 1150 °C and strain rates from 0.01 s⁻¹ to 50 s⁻¹ on Gleeble-1500 thermo-simulation machine. Based on these experimental results, an artificial neural network (ANN) model is developed to predict the constitutive flow behaviors of 42CrMo steel during hot deformation. The inputs of the neural network are deformation temperature, log strain rate and strain whereas flow stress is the output. A three layer feed forward network with 12 neurons in a single hidden layer and back propagation (BP) learning algorithm has been employed. The effect of deformation temperature, strain rate and strain on the flow behavior of 42CrMo steel has been investigated by comparing the experimental and predicted results using the developed ANN model. A very good correlation between experimental and predicted result has been obtained, and the predicted results are consistent with what is expected from fundamental theory of hot compression deformation, which indicates that the excellent capability of the developed ANN model to predict the flow stress level, the strain hardening and flow softening stages is well evidenced.     

HMn62-3-3合金的热变形行为及热加工图

通过Gleeble-1500热模拟机在500~600℃、应变速率0.01~10s~(-1)条件下的近等温热模拟压缩试验,建立合金本构方程和热加工图。结果表明:HMn62-3-3合金在热变形过程中发生动态再结晶行为,其峰值应力随变形温度的升高或应变速率的降低而降低;采用Arrhenius方程能够较好地拟合HMn62-3-3合金的流变行为,其热变形激活能为201.525kJ·mol~(-1);根据DMM模型,计算并建立了HMn62-3-3材料的热加工图,由此确定热变形过程中的最佳工艺参数为变形温度610~640℃,应变速率为2~10s~(-1)。     

电子束冷床熔炼TC4钛合金的热变形行为

使用Gleeble-3800热模拟实验机进行一系列热模拟压缩实验,研究了电子束冷床熔炼TC4钛合金在变形温度为850℃~1100℃、应变速率为0.01 s^-1~10 s^-1条件下的热变形行为。根据真应力-真应变曲线分析变形参数对流变应力的影响,分别建立电子束冷床熔炼TC4钛合金在(α+β)两相区和β单相区的Arrhenius本构模型,绘制了基于动态材料模型的热加工图。结果表明：流变应力随着温度的提高和应变速率的增大而降低;(α+β)两相区的热变形激活能Q=746.334 kJ/mol,β单相区的热变形激活能Q=177.841 kJ/mol;用相关系数法和相对平均误差分析了模型的误差,相关系数R²=0.995,相对平均误差AARE=5.04%。这些结果表明,所建立的模型较为准确,可准确预测其热变形流变应力;合金的最佳加工区域为：变形温度1000~1100℃、应变速率0.01~0.1 s^-1。     

基于人工神经网络的Mg-Zn-Zr-Ce合金热压缩变形研究

为研究含稀土元素铈的镁合金中高温流变行为,利用热模拟试验机对Mg-6Zn-0.5Zr-1.5Ce合金在变形温度523～673 K、应变速率0.001～1 s-1范围内进行热压缩实验.基于真应力真应变实验数据构建了单隐层前馈误差反向传播人工神经网络模型,利用该模型对ZK60-1.5Ce合金的流变应力行为进行预测,并分析了变形温度、应变速率与真应变对流变应力的影响.研究表明:Ce添加可显著细化晶粒;该镁合金的流变应力随变形温度降低和应变速率升高而增加;其流变应力行为可用双曲正弦函数进行描述,依据峰值应力拟合求得该合金的表观激活能为161.13 kJ/mol;变形温度和应变速率对流变应力的影响高于真应变.所建立的人工神经网络模型可以很好地描述该镁合金的流变应力,其预测值与实验数值吻合良好.     

Prediction of metadynamic softening in a multi-pass hot deformed low alloy steel using artificial neural network

The metadynamic softening behaviors in 42CrMo steel were investigated by isothermal interrupted hot compression tests. Based on the experimental results, an efficient artificial neural network (ANN) model was developed to predict the flow stress and metadynamic softening fractions. The effects of deformation parameters on metadynamic softening behaviors in the hot deformed 42CrMo steel have been investigated by the experimental and predicted results from the developed ANN model. Results show that the effects of deformation parameters, such as strain rate and deformation temperature, on the softening fractions of metadynamic recrystallization are significant. However, the strain (beyond the peak strain) has little influence. A very good correlation between experimental and predicted results indicates that the excellent capability of the developed ANN model to predict the flow stress level and metadynamic softening, the metadynamic recrystallization behaviors were well evidenced.     

Al-Cu-Mg-Ag合金的热变形加工图及其分析

为确定Al-Cu-Mg-Ag合金的热加工工艺制度提供理论依据以及便捷的途径,基于动态材料模型(DMM)理论和Ziegler失稳判据,采用Al-Cu-Mg-Ag合金的热变形实验数据,建立了热变形加工图,并利用加工图理论分析了该合金在热变形过程中的变形行为.结果表明:Al-Cu-Mg-Ag合金热变形时有2个失稳区域,一是变...     

Artificial neural network approach to predict the flow stress in the isothermal compression of as-cast TC21 titanium alloy

Isothermal compression of as-cast TC21 titanium alloy at the deformation temperatures ranging from 1000 to 1150 °C with an interval of 50 °C, the strain rates ranging from 0.01 to 10.0 s^?1 and the height reduction of 60% was conducted on a Gleeble-3500 thermo-mechanical simulator. Based on the experimental results, an artificial neural network (ANN) model with a back-propagation learning algorithm was developed to predict the flow stress in isothermal compression of as-cast TC21 titanium alloy. In the present ANN model, the strain, strain rate and deformation temperature were taken as inputs, and the flow stress as output. According to the predicted and experimental results, the maximum error and average error between the predicted flow stress and the experimental data were 4.60% and 1.58%, respectively. Comparison of the predicted results of flow stress based on the ANN model and those using the regression method, it was found that the relative error based on the ANN model varied from ?1.41% to 4.60% and that was in the range from ?13.38% to 10.33% using the regression method, and the average absolute relative error were 1.58% and 5.14% corresponding to the ANN model and regression method, respectively. These results have sufficiently indicated that the ANN model is more accurate and efficient in terms of predicting the flow stress of as-cast TC21 titanium alloy.     

Hot working behavior of a WE54 magnesium alloy

The plastic deformation and recrystallization behavior of the commercial magnesium alloys WE54 was analyzed using the strain rates 0.01, 0.1, 1, and 5 s^?1 in the temperature range from 400 to 550 °C. The dependence of the flow stress on the temperature and the strain rate was modeled using the Garofalo hyperbolic sine equation. Thereby, the activation energy for plastic deformation of 224 kJ mol^?1 was determined considering the flow stress at a strain of 0.5. The analysis revealed a stress exponent of 3.2. Furthermore, processing maps were generated by plotting the efficiency of power dissipation and the instability parameter considering different instability criteria as a function of the temperature and the strain rate. Depending on these parameters the extent of the recrystallization and the localization of the nucleation varied, significantly. At 400 °C, the recrystallization is very limited and was observed at grain boundaries (GB), shear bands (SB), and twin boundaries (TW). Increasing temperatures result in an increased recrystallized fraction, while lower strain rates promote grain boundary nucleation and reduce the amount of SBN and TW. The prediction of the processing map was verified by large scale extrusion trials, which proof that the evaluation of hot compression data can provide an effective tool to establish viable processing parameters.     

Numerical simulation of finish hard turning for AISI H13 die steel

A coupled thermo-mechanical model of plane-strain orthogonal turning of hardened steel was presented. In general, the flow stress models used in computer simulation of machining processes are a function of effective strain, effective strain rate and temperature developed during the cutting process. However, these models do not adequately describe the material behavior in hard machining, where the workpiece material is machined in its hardened condition. This hardness modifies the strength and work hardening characteristics of the material being cut. So, the flow stress of the work-material was taken with literature [H. Yan, J. Hua, R. Shivpuri, Development of flow stress model for hard machining of AISI H13 work tool steel. The Fourth International Conference on Physical and Numerical Simulation of Materials Processing, Shanghui in China, 2004, p. 5] in order to take into account the effect of the large strain, strain-rate, temperature and initial workpiece hardness. Then a series of numerical simulations had been done to investigate the effect of machining parameters on the machinability of hardened steel AISI H13 in finish turning process. The results obtained are helpful for optimizing process parameters and improving the design of cutting inserts in finish turning of hardened steel AISI H13.     

2124铝合金的热压缩变形和加工图

采用热模拟实验研究2124铝合金在应变速率为0.01~10s-1、变形温度为340~500℃条件下的流变应力行为。结果表明:2124铝合金热变形过程中的流变应力可用双曲正弦本构关系来描述,平均激活能为170.13kJ/mol。根据动态材料模型,计算并分析2124铝合金的加工图。利用加工图确定热变形的流变失稳区,并且获得了实验参数范围内的热变形过程的最佳工艺参数,其热加工温度为450℃左右,应变速率为0.01~0.1s-1。     

Effective parameters modeling in compression of an austenitic stainless steel using artificial neural network

In this study, the prediction of flow stress in 304 stainless steel using artificial neural networks (ANN) has been investigated. Experimental data earlier deduced—by [S. Venugopal et al., Optimization of cold and warm workability in 304 stainless steel using instability maps, Metall. Trans. A 27A (1996) 126–199]—were collected to obtain training and test data. Temperature, strain-rate and strain were used as input layer, while the output was flow stress. The back propagation learning algorithm with three different variants and logistic sigmoid transfer function were used in the network. The results of this investigation shows that the R² values for the test and training data set are about 0.9791 and 0.9871, respectively, and the smallest mean absolute error is 14.235. With these results, we believe that the ANN can be used for prediction of flow stress as an accurate method in 304 stainless steel.     

Processing maps of Ti662 unreinforced and reinforced with TiC particles according to dynamic models

Ti–6Al–6V–2Sn produced by powder metallurgy by Dynamet unreinforced (CermeTi^®-C-662) and reinforced with 12 vol.% of TiC particles (CermeTi^®-C-12-662), and ingot Ti662 are deformed at high temperatures. The processing maps of these materials are derived using the dynamic material model (DMM) developed by Prasad et al., and the modified DMM developed by Murty and Rao. Although both models result in similar power dissipation values, the instability zones predicted by them are quite different. The processing maps predicted by the modified DMM can be correlated to the deformation behaviour of these materials, with respect to the shape of their flow curves and to their microstructure after deformation. The concentration of stresses produced during compression is released by cracking at the triple junction of grain boundaries in the CermeTi^®-C-662, whereas in the CermeTi^®-C-12-662 by fracture or debonding of the reinforcing particles.     

Neural network modelling of flow stress in Ti–6Al–4V alloy with equiaxed and Widmanstätten microstructures

Abstract

In the present study, artificial neural networks (ANNs) were used to model flow stress in Ti–6Al–4V alloy with equiaxed and Widmanstätten microstructures as initial microstructures. Continuous compression tests were performed on a Gleeble 3500 thermomechanical simulator over a wide range of temperatures (700–1100°C) with strain rates of 0˙001–100 s^–1 and true strains of 0˙1–0˙6. These tests have been focused on obtaining flow stress data under varying conditions of strain, strain rate, temperature, and initial microstructure to train ANN model. The feed forward neural network consisted of two hidden layers with a sigmoid activation function and backpropagation training algorithm was used. The architecture of the network includes four input parameters: strain rate ?, Temperature T, true strain ? and initial microstructure and one output parameter: the flow stress. The initial microstructure was considered qualitatively. The ANN model was successfully trained across (α+β) to β phase regimes and across different deformation domains for both of the microstructures. Results show that the ANN model can correctly reproduce the flow stress in the sampled data and it can predict well with the nonsampled data. A graphical user interface was designed for easy use of the model.     

TA5钛合金的应变补偿物理本构模型和加工图

使用Gleeble-3800热模拟试验机对TA5钛合金进行等温恒应变速率压缩,研究其在变形温度为850~1050℃、应变速率为0.001~10 s^-1和最大变形量为60%条件下的高温热变形行为;建立了引入物理参量的应变补偿本构模型,并根据DMM模型得到了加工图。结果表明：TA5钛合金为正应变速率敏感性和负变形温度相关性材料;考虑物理参量的应变补偿本构模型具有较高的预测精度,其相关系数R为0.99,平均相对误差AARE为8.95%。分析加工图和观察微观组织,发现失稳区域(850~990℃,0.05~10 s^-1)的主要变形机制为局部流动;稳定区域(870~990℃,0.005~0.05 s^-1)的主要变形机制为动态回复和动态再结晶。TA5钛合金的最佳热加工工艺参数范围为870~990℃和0.005~0.05 s^-1。     

ZK60镁合金加工图的构建及失稳分析

采用Gleeble-1500热模拟试验机对ZK60镁合金在变形温度为150～400℃,应变速率为0.001～10 s-1条件下的热变形行为进行研究,利用双曲正弦关系式描述了该合金在热变形过程中的稳态流变应力;根据合金动态模型,计算并分析了该合金的加工图.研究表明:利用加工图可确定出该合金热变形的流变失稳区,导致变形失稳的原因主要是孪生和局部流变;试验条件下热变形的最佳工艺参数为变形温度350℃,应变速率0.001 s-1,在该条件下合金发生完全再结晶,具有较好的塑性.