Fast Reconstruction Method of the Stress Field for the Steam Turbine Rotor Based on Deep Fully Convolutional Network

2021 ◽  
Author(s):  
Ding Guo ◽  
Tianyuan Liu ◽  
Di Zhang ◽  
Yonghui Xie
Keyword(s):  
Finite Element ◽  
Stress Distribution ◽  
Stress Field ◽  
Temperature Measurement ◽  
Turbine Rotor ◽  
Convolutional Network ◽  
Fully Convolutional Network ◽  
Start Up ◽  
Fast Reconstruction ◽  
Reconstruction Model

Abstract Since it is difficult to directly measure the transient stress of a steam turbine rotor in operation, a rotor stress field reconstruction model based on deep fully convolutional network for the start-up process is proposed. The stress distribution in the rotor can be directly predicted based on the temperature of a few measurement points. First, the finite element model is used to accurately simulate the temperature and stress field of the rotor start-up process, generating training data for the deep learning method. Next, data of only 15 temperature measurement points are arranged to predict the stress distribution in critical area of the rotor surface, with the accuracy (R2-score) reaching 0.997. The time cost of the trained neural network model at a single case is 1.42s in CPUs and 0.11s in GPUs, shortened by 97.3% and 99.8% with comparison to finite element analysis, respectively. In addition, the influence of the number of temperature measurement points and the training size are discussed, verifying the stability of the model. With the advantages of fast calculation, high accuracy and strong stability, the fast reconstruction model can effectively realize the stress prediction during start-up processes, resulting in the possibility of real-time diagnosis of rotor strength in operation.

Fast Reconstruction Method of the Stress Field for the Steam Turbine Rotor Based On Deep Fully Convolutional Network

2021 ◽  
Author(s):  
Ding Guo ◽  
Tianyuan Liu ◽  
Di Zhang ◽  
Yonghui Xie
Keyword(s):  
Finite Element ◽  
Stress Distribution ◽  
Stress Field ◽  
Temperature Measurement ◽  
Turbine Rotor ◽  
Convolutional Network ◽  
Fully Convolutional Network ◽  
Start Up ◽  
Fast Reconstruction ◽  
Reconstruction Model

Abstract Since it is difficult to directly measure the transient stress of a steam turbine rotor in operation, a rotor stress field reconstruction model based on deep fully convolutional network for the start-up process is proposed. The stress distribution in the rotor can be directly predicted based on the temperature of a few measurement points. First, the finite element model is used to accurately simulate the temperature and stress field of the rotor start-up process, generating training data for the deep learning method. Next, data of only 15 temperature measurement points are arranged to predict the stress distribution in critical area of the rotor surface, with the accuracy (R2-score) reaching 0.997. The time cost of the trained neural network model at a single case is 1.42s in CPUs and 0.11s in GPUs, shortened by 97.3% and 99.8% with comparison to finite element analysis, respectively. In addition, the influence of the number of temperature measurement points and the training size are discussed, verifying the stability of the model. With the advantages of fast calculation, high accuracy and strong stability, the fast reconstruction model can effectively realize the stress prediction during start-up processes, resulting in the possibility of real-time diagnosis of rotor strength in operation.

Investigations on Residual Stress Distribution for 2A02 Aluminum Forgings

2013 ◽  
Vol 853 ◽  
pp. 135-142
Author(s):  
Jiang Cao ◽  
Chun Fu Li ◽  
Yan Wang ◽  
Xing Sun ◽  
Shu Yun Wang ◽  
...  
Keyword(s):  
Residual Stress ◽  
Finite Element ◽  
Stress Distribution ◽  
Stress Field ◽  
Residual Stress Distribution ◽  
Indentation Method ◽  
Absolute Value ◽  
Element Simulation ◽  
Quenching Process ◽  
Micro Indentation

High strength aluminum alloys have been widely used in aviation manufacturing due to their favorable combination of intensity, stress corrosion resistance and toughness. However, the research and control of residual stress distribution in aluminum components have become a key issue to be solved during the heat treatment and subsequent processes. By means of the analysis of micro-indentation method and ANSYS finite element method, the residual stress distribution in 2A02 aluminum components after water quenching were systematically investigated, mainly considering two factors of the symmetry of structure and the variation of surface constraint. This study may give great help to the technology of relieving forgings residual stress of two alloys.The results of micro-indentation method show that the absolute value of the residual stress within the sample tends to decrease as the condition of constraint increase at the location of the same thickness; the absolute value of the surface residual stress also tends to decrease as the thickness of the sample increase with the same constraint conditions. The tested results by micro-indentation method are in consistent with the results of finite element simulation to a great extent.The results of finite element simulation are as follows: for these two aluminum alloy, the stress field distribution during the process of quenching is mainly influenced by the thickness of the samples. In general, at the initial stage of the quenching process, the stress state at the components surface are controlled by tensile stress in the direction of both thickness and width, while the residual stress within the samples is dominated by compressive stress; at the end of the quenching process, the stress field distribution just turn to the opposite. These results are in great agreement with the corresponding results of the indentation method.

Effect of Steam Turbine Rotor Notch Fillet Radius on Stress and Deformation

2013 ◽  
Vol 860-863 ◽  
pp. 1770-1781
Author(s):  
Dong Mei Ji ◽  
M. H. Herman Shen ◽  
Shi Hua Yang ◽  
Gang Xia
Keyword(s):  
Finite Element ◽  
Thermal Stress ◽  
Finite Element Model ◽  
Steam Turbine ◽  
Element Model ◽  
Turbine Rotor ◽  
Critical Value ◽  
Steam Turbine Rotor ◽  
Fillet Radius ◽  
Start Up

A thorough investigation on the effect of a 320MW steam turbine rotor notch fillet radius on thermal and mechanical stresses during start up is presented. The approach consists of a shape design and analysis procedure which incorporates a finite element model. The finite element model is used to characterize the radius of the rotor notch fillet for ensuring the designed thermal and mechanical stress state/pattern and associated deflection during start-up. The results indicate that the notch fillet radius r has significant impact on the total stress of the rotor, in particular on thermal stress. It is determined that the thermal stress is decreased as the notch fillet radius r increases to a critical value. However, the thermal stress becomes saturated as the radius is increased to values larger than the critical value. The results also indicate that the rotor notch fillet radius has little effect on the deflection of the rotor during start-up. This investigation could be very useful to designers for construction of the design guidelines for steam turbine rotors.

scholarly journals Voids Nucleation at Inclusions of Various Shapes in Front of the Crack in Plane Strain

2016 ◽  
Vol 61 (3) ◽  
pp. 1587-1592 ◽  
Author(s):  
A. Neimitz ◽  
U. Janus
Keyword(s):  
Finite Element Method ◽  
Finite Element ◽  
Stress Distribution ◽  
Stress Field ◽  
Plane Strain ◽  
Crack Front ◽  
The Finite Element Method ◽  
Element Method

Abstract An analysis is presented of the stress field in and around inclusions of various shapes. Results were obtained by the finite element method. Inclusions were located within elementary cells located at the centre of the specimen next to the crack front. The influence of an in-plane constraint on the stress distribution was tested.

Simulation of Single Crystal Nickel-Based Superalloys Creep on Finite Element Method

2012 ◽  
Vol 433-440 ◽  
pp. 2029-2033
Author(s):  
Shu Zhang ◽  
Lei Meng
Keyword(s):  
Mathematical Model ◽  
Finite Element Method ◽  
Finite Element ◽  
Numerical Calculation ◽  
Stress Distribution ◽  
Stress Field ◽  
Single Crystal ◽  
Microstructure Evolution ◽  
The Finite Element Method ◽  
Element Method

Based on finite element Method a dynamic mathematical model is established, and the simulation of stress distribution around the defects of single crystal nickel-based superalloysis also established with ANSYS. After the change of stress field with time is analyzed, the result is compared with that achieved through numerical calculation and experimental analysis. The comparison shows that the finite element method is effective to study the stress distribution and can provide basis for creep features and microstructure evolution.

scholarly journals Modeling and Thermal-Mechanical Coupling Analysis of Piston in Car Engines

2020 ◽  
Vol 45 (1) ◽  
pp. 83-92
Author(s):  
Feifei Zhao
Keyword(s):  
Finite Element ◽  
Temperature Field ◽  
Stress Distribution ◽  
Stress Field ◽  
Boundary Conditions ◽  
Thermal Load ◽  
Mechanical Load ◽  
Field Analysis ◽  
Coupling Analysis ◽  
Mechanical Coupling

In this paper, finite-element analysis (FEA) is carried out on the temperature field and stress field of automobile engine piston, as well as the thermal-mechanical load coupling stress field. Through the analysis, the authors grasped the thermal load and combined stress distribution of the piston, and thus optimized the piston design to improve its operational reliability. Specifically, a 1/4 solid model of the piston was constructed in the three-dimensional (3D) computer-aided design (CAD) software Pro/ENGINEER, and then converted into a finite-element model in Pro/Mechanica. Then, an alternating load was imposed on the piston model, and fatigue analysis was performed to identify the parts of the piston prone to fatigue failure, and judge whether the piston structure satisfies working requirements. Next, temperature field analysis was carried out on the piston model. The distribution of the steady-state temperature field as determined by applying temperatures and heat transfer coefficients as required by the boundary conditions of the third kind. Finally, the piston model was subject to thermal-mechanical coupling analysis. The stress and deformation distributions of the piston under the coupled stress field were ascertained under the boundary conditions of temperature field distribution and mechanical load. Through the above work, the authors obtained the basis for safety evaluation of piston, laying the foundation for further reducing the thermal load and optimizing the stress distribution of piston.

Computational Analysis of Stress Field and Contact Pressure for Rubber Seal Based on Non-Linear Finite Element

2012 ◽  
Vol 268-270 ◽  
pp. 1080-1083 ◽  
Author(s):  
Jian Bing Sang ◽  
Li Fang Sun ◽  
Su Fang Xing ◽  
Dong Ling Zhang
Keyword(s):  
Finite Element ◽  
Stress Distribution ◽  
Stress Field ◽  
Contact Stress ◽  
Von Mises Stress ◽  
Rubber Seal ◽  
Linear Finite Element ◽  
Maximum Contact Stress ◽  
Von Mises ◽  
Maximum Contact

This paper aims to research on the sealing capability of rubber seal by utlizing non-linear finite element analysis. After discussion on various types of strain energy functions of rubber like materials, material parameters of Mooney-Rivilin model are determined by curve fitting based on the stress strain curve from the uniaxial tensile experiment. Axis symmetric and isotropic finite element model is developed to analyze the stress field and contact pressue of O-ring seal and lip seal at different medium pressure. Von-mises stress distribution rule and contact stress distribution curve on contact surface are achieved. The results show that maximum Von Mises stress increase with the increases of oil pressure. The maximum contact stress appears on the middle contact zone and the maximum contact stress and contact width increases obviously with the increases of oil pressure. In the meanwhile, the maximum contact stress is greater than midum pressure which can prevent the leakage of midum and achieve the function of sealing. The research results will be a useful technique for predicting the properties of rubber seal and providing reference for engineering design.

Sign in / Sign up

Export Citation Format

Share Document