Improvement of prediction model for work roll thermal contour in hot strip mill

Though high accuracy of the thermal contour was obtained while adopting finite element method,it could not meet the real-time requirement.One-dimensional finite difference method could realize real-time control in the case of neglecting both the circumferential and radical heat exchange,but the over-simplified modes resulted in poor accuracy.And two-dimensional full explicit difference was also limited in practical application since its time step was restricted to keep the model stable.Consequently,a new method of alternating direction finite difference was introduced and discussed in the model's stability,calculating speed and precision.Specific work roll after one real rolling unit was researched,The result shows that error of temperature on work roll surface between measured and calculated values is within 5 ℃.The influence of rolling rhythm and strip width on thermal crown was also studied.The conclusion is verified theoretically and practically that it can maintain absolutely stable and meet the online requirement.

Thermal Behaviors of Work Roll in Finishing Trains of Hot Rolling

In order to analyze the thermal field and thermal contour of work roll in finishing trains of hot rolling in rolling process, a quasi two-dimension implicit finite difference model is developed. To improve the calculating speed and precision, some special solutions are introduced, including the development of this model, the simplification of boundary conditions and the computation of heat transfer coefficients. The results show that these solutions of thermal behaviors of work roll are very much efficient and the model can be used as an on-line profile and fatness control model of large industrial mills.

Determination of Optimal Work Roll Profile for Cold Strip Mill

The model of predicting distribution of strip thickness at the exit and tension stress is established by using weighted residual method and variational method with consideration of rolled metal lateral flow.A New method of calculating loaded roll gap in relation to target tension stress is proposed.The minimum square summation of loaded roll gap deviation is used as an objective function for the first time to optimize initial work roll profile.

Application of simulated annealing algorithm to improve work roll wear model in plate mills

Employing Simulated Annealing Algorithm (SAA) and many measured data, acalculation model of work roll wear was built in the 2 800 mm 4-high mill of Wuhan Iron and Steel(Group) Co. (WISCO). The model was a semi-theory practical formula. Its pattern and magnitude werestill hardly defined with classical optimization methods. But the problem could be resolved by SAA.It was pretty high precision to predict the values for the wear profiles of work roll in a rollingunit. After one-year application, the results show that the model is feasible in engineering, and itcan be applied to predict the wear profiles of work roll in other mills.

Improved model to solve influence coefficients of work roll deflection

According to the concept of virtual bending force,a rational explanation for SHOHET's model was presented. Considering the deformation characters of the work rolls in four-high mill,the deformation model of the work roll was regarded as a cantilever beam and new influence coefficients were deduced.The effect of the bending force was taken into account independently. Therefore,the contribution to work roll deflection caused by rolling load,rolling pressure between rolls and bending force can be got from the new formulas.To validate the accuracy of the formulas,the results obtained from the new formulas were compared with those from SHOHET's formulas.It is found that they highly coincide,which illustrates that the formulas are reliable.

Study on Vertical Vibration Characteristics of the 2-DOF Strip Rolling Mill Model with a Single Weak Defect on the Work Roll Bearing Outer Raceway

Considering the influence caused by a early single pit defect on the outer raceway of the work roll bearing,a 2-DOF plate strip rolling mill vertical vibration model with a single point weak fault on the outer raceway was established.With the practical parameters of the roughing mill of the 1780 hot continuous rolling mill,the vertical vibration characteristics of the rolling mill work roll with different rotating speed and different single pit defect area on the bearing outer raceway are analyzed by numerical simulation.It is found that with the change of the rotation speed of the work roll,different nonlinear vibration behaviors occurred,such as superharmonic resonance,main resonance,com-bined resonance and sub-harmonic resonance.Especially the subharmonic resonance of the work roll is more harmful than the main resonance when the work roll speed is twice the rotation speed corresponding to the first and second natural frequency of the rolling mill.This work provides a theoretical basis for further clarifying the effect caused by a early defect of the work roll bearing on the mill vibration.

宽板连轧锥形工作辊轴向窜动调控及工业应用

为了提高连轧锥形工作辊轴向窜动控制精度,对材料加工时发生横向流动的过程建立有限元模型,并构建了机架轧后屈服模型,并将模型置于现场环境中开展验证研究。研究结果表明:受到轧后屈服作用时,处于边降区(15~115)mm之间的断面误差可以控制到3μm范围内,与带钢边部的最大偏差在(0~15)mm之间。在机架间和轧制区内都出现了横向流动的情况,导致锥形辊发生了边降调控性能下降。随着窜辊量增大后,第一、二机架都出现功效下降的结果,第三机架则持续增大至最高功效。下游机架窜辊位都位于上游机架控制特征点范围内。这里调控方式形成了显著边部效果,可以观察到内部偏差。该研究对保证板材的轧制质量具有很好的实际指导价值。

基于脉冲涡流监测与Wiener过程的再制造工作辊实时剩余寿命预测方法

再制造工作辊在热轧工作过程中熔覆层易发生复杂的退化行为,而其性能退化会影响轧制产品的质量。为实时监测轧辊退化情况并预测其剩余寿命,构建一套在线脉冲涡流监测方法和系统,采用变分模态分解-希尔伯特变换(VMD-Hilbert)方法提取0~5000 Hz频率边际谱能量和的增量作为轧辊退化特征。基于此,构造考虑个体差异的状态退化空间模型,结合最大期望(EM)算法和Kalman平滑滤波算法联合对模型中未知参数进行自适应参数估计,并利用层次分析法融合监测的历史数据与实时数据,以实现对再制造工作辊实时剩余寿命预测。实验结果表明,所提方法能够准确预测再制造工作辊剩余寿命,具有一定的应用前景。

基于卷积神经网络的多工况多传感滚动轴承实时监控方法

针对工业环境中广泛在多工况下多滚动轴承实时状态监测的需求和部署环境受限的挑战,提出一种基于卷积神经网络(Convolutional Neural Network,CNN)的面向多传感器滚动轴承运行状态监控方法。该方法将两个不同工况下的一维时间序列数据集以均方根(Root Mean Square,RMS)指标标注,并通过将一维时间序列多传感器数据重构为二维空间张量的形式输入卷积神经网络训练。最后利用层融合和16比特量化优化,将网络部署到FPGA上,用以解决CNN的计算开销。实验结果表明,在结合了两种不同工况的数据集下,网络测试推理准确度依然高达99.24%,比多层感知机实现高10.48%,比多层感知机结合支持向量机的实现高2.91%,该算法对于新加入的数据集也有较强的鲁棒性,经过重训练,新加入的数据集准确率可以达到99.17%。基于FPGA部署优化的网络的峰值能效为76.217GPOS/W,为CPU实现的33.09倍,GPU实现的5.39倍。其中,16比特精度部署的网络测试精度相较32比特精度实现仅降低0.001%。

四工作辊轧机轧制304不锈钢板件截面塑性变形分析

针对四工作辊轧机轧制304不锈钢板件存在变形量大、尺寸不稳定的问题,利用Solidworks软件建立四工作辊轧制系统有限元模型,采用Deform软件对304不锈钢板件在不同轧辊直径、轧辊转速和轧辊压下量下进行轧制仿真,分析其对轧件截面高度和等效应力分布的影响,根据轧制仿真分析结果设计制造了四工作辊轧机,并对不同压下量下仿真和生产的轧件截面高度进行对比分析。结果表明,轧件截面塑性变形可分为相互作用Ⅰ区、过渡Ⅱ区和变形Ⅲ区,等效应力、等效应变和截面高度在Ⅰ区最大,在Ⅲ区最小且变化平稳,Ⅱ区值位于两者之间并呈U型分布。

超声滚压对DC53钢表面性能的影响

为了改善DC53冷作模具钢的表面性能,使用扫描电子显微镜、激光光谱共聚焦显微镜、显微硬度计、球-盘摩擦磨损试验机等设备分析了超声表面滚压加工工艺(USRP)的静压力、滚压道次、振幅、进给量对DC53冷作模具钢表面粗糙度、显微硬度、摩擦因数和磨损的影响规律。结果表明:与未处理DC53模具钢相比,USRP处理后DC53冷作模具钢表面产生了塑性变形,表面粗糙度由5.570μm最低可降为1.067μm,且在DC53冷作模具钢表面产生了晶粒细化现象,引入了梯度硬化层,表面硬度由基体的720 HV最大可提升为818.6 HV,硬化层深度最大可达300μm。摩擦因数降低,磨痕宽度减小,耐磨性能提高。通过极差分析法获得了对粗糙度、显微硬度、摩擦因数影响最大的工艺参数依次为进给量、滚压道次、静压力。通过灰色关联度分析法获得了工艺参数对表面性能的影响次序为静压力>进给量>滚压道次>振幅,最佳工艺参数组合为静压力400 N,滚压道次4次,振幅12μm,进给量0.05 mm。

一种海洋自升式修井平台船型分析

针对渤海海域油气开发中自升式修井平台的船型设计,进行系统分析。从桩腿数量维度出发,考虑波流载荷、风载荷和平台重力的共同作用,通过有限元数值模拟软件,建立自升式修井平台的仿真模型,计算不同方向角下平台的波流载荷、桩腿位移和抗倾力矩等评价指标,综合对比分析三桩腿和四桩腿两种不同船型的设计特点,以揭示各自在平台方案、预压载能力和稳定性等方面的特点。研究结果发现,四桩腿平台在舱内设备布置、预压载能力和抗倾稳性方面具有明显优势,特别是在单桩失效情况下,四桩腿平台的稳定性和安全性更高。研究结论对自升式修井平台的设计优化具有重要的指导意义,为进一步提升平台的安全性和高效性提供理论依据。

四辊轧机工作辊辊颈端部铸接修复部位寿命评估

轧辊作为带钢生产轧机的核心部件,其消耗对生产成本有重要影响。随着钢铁企业竞争加剧导致的降本增效需求逐步提高,轧辊再制造循环利用得到逐步推广。大型轧机大尺寸轧辊工作层消耗后改造用于小轧机为其中主要方式之一;对于局部尺寸尤其是辊肩尺寸不足的轧辊通常采用铸接方式加长以满足进一步使用的需求;铸接再制造后轧辊的使用寿命成为保障安全生产的关键。针对上述问题,在轧辊铸接部位取样开展了疲劳性能测试,同时选用扭转疲劳模型在证明计算精度的基础上,结合基于现场生产数据建立的载荷谱,对工作辊修复后的辊颈进行了寿命预测,预期服役寿命满足现场对修复后轧辊的性能需求,可为现场轧辊安全使用维护制度建立提供理论支撑。

基于流形特征域适配的滚动轴承故障诊断

针对变工况下存在两域特征分布复杂,数据在原始空间进行分布对齐时,特征扭曲和发散难以消除等问题,提出一种基于流形特征域适配(manifold feature domain adaptation, MFDA)的滚动轴承故障诊断方法。该方法通过无监督的方式生成一个与目标域具有相似分布的中间域,构建与源域、中间域和目标域相关的公共子空间,并利用局部生成差异度量保留数据在子空间中的流形局部几何结构,以避免数据对齐时出现扭曲和发散;同时利用最大均值差异度量对齐中间域和目标域,以最小化两域间的分布差异,保证数据间局部与全局结构的相关性。最后,利用学习到的特征,以最小二乘法实现滚动轴承的跨域故障识别。在三组滚动轴承数据集上进行试验验证,与其他智能识别算法相比,该方法能有效避免特征扭曲和发散,且具有优良的泛化性能。

Solidworks在辊锻制坯中的应用

为适应三维辊锻模的设计要求,提出了基于三维中间毛坯模型的辊锻及模具设计方法,即通过构建辊锻件中间毛坯模型来进行变形量的计算和分配、确定型槽截面尺寸和纵向尺寸,并用Solidworks进行成功实现,提高了辊锻模具的设计质量和效率。

Effect of ceramic work rolls on surface roughness of rolled SUS304 ultra-thin strips

Surface morphology and roughness are important parameters of surface quality of cold-rolled ultra-thin strip steel.In order to study the effect of Si_(3)N_(4) ceramic work rolls on the surface roughness of SUS304 ultra-thin strip,ABAQUS finite element model calculation,3D laser profilometry,and other methods were adopted based on the physical characteristics of Si_(3)N_(4) ceramic roll and 9Cr2MoV steel roll,like anti-flattening performance and oil wettability.The surface morphology and roughness of SUS304 ultra-thin strip rolled by different material work rolls under oil lubrication conditions were measured.The results showed that under the rolling force of 10 kN,the maximum flattening amount of Si_(3)N_(4) ceramic roll was reduced by 35.14%compared to that of 9Cr2MoV steel roll.The reduction rate of SUS304 ultra-thin strip rolled by ceramic roll was 34.19%,while that rolled by steel roll was only 22%;Si_(3)N_(4) ceramic roll can further increase the number of alternating convex peaks and concave valleys in the rolled thin strip,reduce the profile steepness,reduce surface roughness,and enhance the planarization effect of surface micro-convex bodies,with improving surface quality and surface glossiness.Finally,the mechanism of Si_(3)N_(4) ceramic roll significantly improving the surface roughness of rolled SUS304 ultra-thin strip was proposed.

固态基板表面硬化层对液-固铸轧1070A/7075复合板组织与性能影响研究

采用钎焊金刚石碗磨片、钢丝刷和百叶片3种不同材料对7075铝合金基板进行打磨,制备了具有不同表面加工硬化层状态的7075铝合金基板,之后采用液-固铸轧的方式制备出1070A/7075铝合金复合板,探究了不同打磨方式下加工硬化层状态对1070A/7075铝合金复合板组织与性能的影响。研究结果表明:3种机械打磨方式下复合板硬化层厚度情况分别为百叶片(45μm)>钢丝刷(25μm)>钎焊金刚石碗磨片(15μm),固态基板经钎焊金刚石碗磨片打磨处理后的1070A/7075铝合金复合板元素扩散距离为7.5μm,结合强度为55.2 MPa。

复合多尺度注意熵在旋转机械多工况损伤识别中的应用

针对传统旋转机械损伤识别方法存在模型精度低和抗噪性差的问题,提出了一种基于复合多尺度注意熵(CMATE)和随机森林(RF)的旋转机械多工况损伤识别方法。首先,提出了一种新的测量时间序列复杂度的非线性动力学工具——复合多尺度注意熵;然后,利用CMATE提取旋转机械振动信号的损伤特征,其表征了旋转机械不同工况下的损伤特性;接着,将损伤特征输入至基于随机森林构造的多类别分类器中,进行了损伤识别;最后,采用滚动轴承-齿轮箱、齿轮箱和离心泵3种旋转机械数据集,并分别构造了9种工况和20种工况的多工况损伤数据集,对该损伤识别方法进行了实验研究。研究结果表明:该旋转机械损伤识别方法分别取得95%、97%和100%的识别准确率,在准确率和特征提取效率两方面优于其他的非线性动力学工具;并且在0 dB、1 dB、2 dB和3 dB这4种不同信噪比的噪声干扰下,依然取得了不错的损伤识别结果,证明了该模型具有可观的抗噪性;同时,该损伤识别方法能够稳定地识别旋转机械的不同负载和转速下的损伤,平均识别准确率分别达到了97.2%和93.5%,具有较好的实际应用潜力。

变工况及小样本情况下滚动轴承故障迁移学习方法综述

准确预测滚动轴承损伤类型及剩余寿命归于各类旋转机械运行的可靠性和安全性具有重要意义。迁移学习方法通过实验数据使网络模型学习到相关故障类型的知识,训练好的模型可以直接应用到实际工业生产问题当中。突破了传统深度学习方法所需大量有标签数据、模型使用有局限性、通用性差的局限。首先对不同类型的迁移学习方法在滚动轴承故障诊断中的应用进行分析,归纳总结面对不同情况下的迁移学习方法。其次,针对变工况、小样本及一些其他情况下的问题进行总结分析。最后,给出了滚动轴承的迁移学习发展趋势。

大采高超长工作面调斜开采技术研究与应用

根据梁北煤矿11071综采工作面瓦斯地质情况和采煤工艺条件,通过对实中心调斜和虚中心调斜方案优缺点的分析比较结果,选择采取围绕虚中心旋转调斜开采的施工方案,并通过精确计算得出调斜开采的技术参数指标,同时确定了调斜开采的工艺流程和安全技术保障措施。实践证明,该调斜开采方案简化了机巷运煤系统,解决了调斜期间虚中心控制、支架姿态控制、运输机上窜下滑控制、转载机和运输机搭接控制和超前应力控制等难题,实现了工作面在调斜开采期间不间断推进,给矿井带来了巨大的技术进步和经济效益。