Load Cases Research Articles

Feedforward neural network-assisted parameter identification and tuning for uniaxial superelastic shape memory alloy models under dynamic loads

In this study, we introduce a machine learning-based method to predict the modeling parameters of superelastic shape memory alloys (SMAs). Our goal is to simultaneously determine and fine-tune all internal and material-related parameters, including thermodynamic ones, for a specific constitutive model using only cyclic tensile tests. We employ feedforward neural networks (FNNs) for their versatile structure. First, we sample the searched parameters within a predefined parameter space using the Latin hypercube sampling method. Then, using the constitutive model with the sampled parameters and representative strain loading, we generate the corresponding stress responses and finally train the FNN. To address the ill-posed nature of this inverse parameter identification problem and ensure a unique parameter set, during training, we use a dual network architecture with an additional FNN-based surrogate of the constitutive model. We also utilize transfer learning to accelerate the training process through knowledge transfer and handle multiple load cases simultaneously, ensuring consistent parameter identification across different scenarios. We validate the method by comparing the numerical results with the experimental data and demonstrate the importance of accurately identified parameter sets by numerical investigations on a SMA-retrofitted frame structure.

Differences for Cooling Capacity of a Cable Tunnel System: Install One or Three Additional Cable Troughs with Cable Circuit for 100/70 Heat Loads – A Numerical Study

The purpose of this report was to compare the numerical analysis for different cooling efficiency of a cable tunnel system when installing one additional cable trough or three cable troughs for cable circuit 100% or 70% heat load. Cable tunnel system is widely known for it is use for high voltage transmission in an underground. Due to the cables generating large quantities of heat, water pipes and air ventilation were used in the tunnel to cool down the transmission cable. To investigate the cooling efficiency of a tunnel, research for benchmark case was carried out. A Computational Fluid Dynamic (CFD) benchmark analysis about 2D natural convection inside the square cavity was investigated to further understand the air movement inside the square medium. Ansys Fluent was carried out for benchmark cases of heat transfer natural convection inside the square cavity and to check whether Ansys Fluent can be used for any CFD and heat transfer purposes by validating the results with previous studies. The benchmark case used a 2D square with different active temperature side walls and adiabatic walls for top and bottom side. From the benchmark simulations with different Rayleigh Number, a number representing buoyancy-driven of a fluid from 103 to 105 affected air differently and the heat transfer natural convection inside the square. At different Rayleigh Number air forms a recirculation passing the active walls and adiabatic walls. After investigated the benchmark, the results were used to validate with previous studies. To validate the results was by using Nusselt Number, ratio of convective to conductive. The benchmark case results were validated and Ansys Fluent proven can be used for CFD and heat transfer purposes. Two 3D cable tunnel were modelled with 100-meter length after concluded the benchmark. The two tunnel models were based on the number of cable troughs installed. The first tunnel was with one cable trough and two cable circuits on the bottom, as for the second tunnel was with one bottom cable circuit and three cable troughs. Cable tunnel simulations were based on cable circuit for 100% heat load and 70% heat load with each heat load case was given different water flow rate of 4.25 L/s and 6.5 L/s. A total of eight scenario cases were carried out to determine which provided a better cooling efficiency choice. The analysis for cable tunnel cases on how the copper cables cooled down through heat transfer of natural convection inside the cable trough, which the heat then was removed from the cable circuit by air and continue to the outside of the tunnel. During the heat transfer processes, conductivity was present as not all heat were removed by air and water, but continue to through the walls of the tunnel and to the ground soil. Based on the simulations, a cable tunnel with three cable troughs was proven more cooling efficiency compared tunnel with one cable trough. Some scenarios did not meet the criteria, due to insufficient heat relocation by air and water. The cooling efficiency was determined by which has the better balance heat relocation by air, water, and ground, also considering the comfort and safety of human inside the tunnel.

Analysis of numerical models of an integral bridge resting on an elastic half-space

Abstract The paper presents three methods of the numerical modeling of a 60 m long integral bridge structure resting on an elastic half-space. For the analysis, three bridge models were built using Abaqus FEA software. Models A and C represent complex three-dimensional numerical models consisting of the bridge structure and the soil layer beneath it. The soil layer on which the bridge is resting was modeled as a homogeneous, isotropic, continuous, and elastic semi-infinite body elastic half-space. Model B represents a simple three-dimensional numerical model consisting of just the bridge structure. The stiffness of the soil layer beneath the structure in model B was modeled with spring constants derived for shallow footing foundation based on the theory for an elastic half-space. This model represents an engineering approach to the design of an integral bridge. In all models, the bridge deck is monolithically connected with abutment walls and intermediate piers. The bridge is made of cast-in-situ reinforced concrete. All material constants used in the analysis are presented in the table. Self-weight, uniformly distributed load, and thermal longitudinal expansion of the bridge deck were applied to the bridge models. Due to the nonlinear boundary condition used in the supports of the bridge model A, such as contact and friction, the superposition principle cannot be used in the calculations of this model. For this reason, all the loads involved in all bridge models were combined into a single load case and the large displacement formulation was used in the static analysis. The self-weight of the soil layer beneath the structure was omitted in the analysis. The author is focused on the method of modeling an integral bridge structure resting on elastic soil. For the purpose of this paper, only two piers from each model were selected, from which the internal forces and displacements were compared. Based on the analysis, it was concluded that it is possible to design an integral bridge by building its simplified numerical model, once the conditions given in the conclusions are met.

ON THE SEAKEEPING RESPONSE VALUATION IN RANDOM SEA STATE CONDITIONS OF A LARGE LNG LIQUEFIED NATURAL GAS CARRIER

This seakeeping study refers to a large LNG liquified natural gas carrier with 289.3 m length, slender mono-hull, and fine shapes at both ends. From the loading cases of the large LNG carrier, two relevant are selected, full cargo and heavy ballast. The top speed of the large LNG carrier is 22 knots. To ensure the basis of the seakeeping response valuation in a random sea state, the navigation unrestricted conditions are modeled by the averaged ITTC wave power density spectrum. This study applies a linear strip theory hydrodynamic formulation for unrestricted water depth, delivering directly in the frequency domain the oscillations’ response amplitude operators. For the roll oscillations, this study includes the LNG carrier gravity center vertical position influence evaluation and the equivalent roll dampers effect. The short-term dynamic response for the random waves is evaluated by seakeeping criteria, delivering the safety navigation capabilities of the large LNG carrier.

Optimization design of stamping die for large-size three-dimensional curved thick plate

Stamping is a highly efficient and precise way to fabricate metal plates with complex shapes, commonly used in practical engineering. Lowering the weight of die components is essential to reduce manufacturing and operational costs. Despite using a single topology optimization process, the previous study experienced difficulty in gaining the optimal structural configurations and their specific dimensions while considering structural performance and die weight. Therefore, a three-phase optimization procedure is proposed, namely model preparation, conceptual design, and detailed design. Firstly, using the proposed node-to-node load mapping strategy, the history of the load distribution across the die face is transferred to multiple loading cases in static analysis models. Then, the solid isotropic microstructure with penalization (SIMP) method based topology optimization is employed to achieve better structural configurations. Finally, a multiple-surrogate-models-based size optimization approach is implemented to obtain the specific dimensions. Applications on stamping die for large-size three-dimensional curved thick plate were conducted. The results showed that topology optimization could deliver more efficient structural configurations, which reduced the weight of the upper die, restrike punch, and lower die by 22.6%, 48.7%, and 39.9%, respectively, while still meeting most of the performance requirements. Furthermore, after size optimization, an increase of the restrike punch weight by 22% was obtained to meet the performance constraints, while the upper and lower dies had a further reduction in weight of 19.1% and 8% without violating the performance constraints. Therefore, the proposed optimization procedure is verified to be effective in reducing the die weights while meeting all performance requirements.

Functionally Graded Impact Attenuator Using Bonded Construction

ABSTRACTAutomotive collisions are one of the major causes of death in the European Union, especially in childhood and adolescence. However, improvements in vehicle safety cannot be achieved only by increasing the size and dimensions of the structures, as this demands more materials and more costly manufacturing processes, which leads to an increased resource expenditure and lowered sustainability. Moreover, the increase of the structure size raises the vehicle's weight, leading to added fuel consumption, which cannot be accepted considering modern environmental regulations. To solve these issues, the application of bonded joints using the combination of several adherends materials, such as steel and aluminium alloys and fibber‐reinforced polymers with crash‐resistant adhesives presents itself as a novel solution, allowing to attain enhanced joint strength, energy absorption and weight reduction. The present works introduce a novel concept of an impact attenuator using bonded, geometrically optimized using the concept of functionally graded adherends to maximize energy absorption by ensuring that a load‐bearing path is kept during impact, converting the impact energy into the plastic deformation of panels with variable mechanical properties, tailored to withstand specific load cases, fulfilling a gap in the literature regarding impact absorption devices that combines graded materials and bonded construction. The results obtained present an increase in the energy absorption values of the graded impact attenuators above 200% when compared to the homogenous materials.

Argillaceous Soft Rock in-situ Test Program in Tunneling

AbstractThis study undertakes a comprehensive examination of the characteristics of the argillaceous, hard soil/soft rock (HSSR) lithology of the Unterangerberg formation in Tyrol, Austria, focusing on its swelling properties and anisotropic material behavior. The objective is to achieve an extensive material characterization to be able to calibrate material models accurately. This is to be achieved by means of an in-situ test campaign within the Angath adit tunnel construction site and accompanying laboratory tests. The geotechnical monitoring concept in a designated test gallery includes a chain inclinometer, extensometers, geodetic targets, shotcrete strain meters, photogrammetric observations, and a long-term irrigation test field, with a section of the invert left exposed intentionally. The in-situ tests yield valuable insights into the intricate behavior of the HSSR lithology, offering a comprehensive description of material variability, and recommendations for characterization. Results indicate that despite the high swellable clay mineral content, significant swelling occurred only to a small extent during the observation period to date. The study concludes that the in-situ behavior of such formations significantly contributes to a better understanding of their characteristics, leading to a substantial reduction in critical load cases for future planning phases.

Large Deflections, Loss of Stability and Over-Critical Behavior of Sloping Panels and Arches of Variable Thickness on an Elastic Base

The problem of large deflections of a flat arch in its plane (or an infinitely long panel) loaded with transverse loads is considered. A variational approach has been applied to solve it. The resolving nonlinear equations are reduced to finding the deflection and longitudinal force, which is considered constant along the length of the arch due to its flatness. An approximate solution method is proposed by decomposing the displacements into a Fourier series. The peculiarity of the approach used is that in order to pass the limit points, it is not necessary to use special algorithms such as the continuation method by parameter. It also allows you to trace the process of supercritical deformation of the arch. The proposed approach allows us to consider problems for arches of variable thicknesses, located on elastic supports, on an elastic base with a variable bed coefficient and various loads. Therefore, it is also convenient in problems of finding, for example, the optimal thickness distribution under restrictions on critical loads, on rigidity stiffness and on maximum compression or tension stresses. The results of numerical calculations are presented. The convergence is studied of the solution depending on the number of terms of the series, into which the desired deflection decomposes. A good agreement with the known analytical results is obtained earlier by solving the equilibrium equations of the arch element and panel in the case of simple types of loading. At the same time, even in more complex cases of loading and supercritical bending of an arch without an elastic base, the results are obtained when holding three, four and five members of the Fourier series, the maximum deflections and critical loads differed by no more than 2.5%. On the basis of numerical experiments, the features of the arch behavior caused by the rearrangement of geometry during loading are revealed. The processes of stability loss of the arch and its supercritical behavior are investigated. The effect of symmetric deformation in the case of kinematic loading is found, namely, it is revealed that at a certain value of the concentrated force, an infinite number of equilibrium forms are possible. Arches on an elastic base, as well as with variable thickness are considered (the results are presented in the form of load–displacement diagrams and in the form of pictures of deformed arch shapes). An interesting effect has been revealed based on numerical experiments. It turned out that the increase in thickness when moving away from the supports does not change the nature of the load–displacement diagram, i.e., with some load, a slap occurs. On the contrary, a decrease in thickness when moving away from the supports leads to a flattening of the diagram, and after a certain thickness value in the center, its further decrease leads to the fact that the arch does not occur.

Improved genetic algorithm based on Shapley value for a virtual machine scheduling model in cloud computing

IntroductionIn cloud computing, a common idea to reduce operation costs and improve service quality is to study task scheduling algorithms.MethodsTo better allocate virtual machine resources, a virtual machine resource scheduling algorithm, Shapley value method–genetic algorithm (SVM-GA) is proposed. This algorithm uses the SVM to obtain the contribution values of each component of the virtual machine, refine the topological network, and achieve the optimal solution of scheduling by the genetic algorithm.Results and DiscussionCloudSim simulation results indicate that SVM-GA has the lowest total task completion time when compared with existing intelligent optimization algorithms (such as the max–min algorithm, logistic regression algorithm, and differential evolution algorithm) with the same number of tasks, and the total task time is 25, 55, 81, 112, 145, and 175 s for 200, 400, 600, 800, 1,000, and 1,200 tasks, respectively. As the number of evolutionary generations increases, the ability of SVM-GA to reach the optimal solution of the model increases. In the simulated light load case, the SVM-GA migration time and Q10 migration count optimal solutions are slightly inferior to those of the logistic regression algorithm (3.02 s > 2.38 s; 1,129 times >999 times), but the migration energy consumption and service level agreement violation rate optimal solutions are superior. The SVM-GAA’s performance in the heavy load case is similar to that in the light load case. The experiments show the feasibility of the algorithm proposed in the study.

Analysis of Stress and Strain in Sandwich Structures Using an Equivalent Finite Element Model

The study aims to build an equivalent 2D model as an alternative to the 3D model of sandwich panel structures. This model enables for reducing model building time and calculation time in the design calculation of this sandwich structure. The research object in this study is corrugated core cardboard. First, the isotropic plasticity equivalent (IPE) model for the paper material is implemented in the Abaqus software, using the VUMAT user subroutine. Subsequently, the homogenization method is proposed as an equivalent elastic-plastic finite element model. This model is implemented in Abaqus using the UGENS subroutine. Finally, numerical simulations of different load cases between the 3D model and the equivalent 2D model are performed to confirm the accuracy of the proposed model. The comparison results indicate that the equivalent model ensures exceptional accuracy compared to the 3D model but significantly reduces model building time and CPU time.

Development of a Mechanical Vehicle Battery Module Simulation Model Combined with Short Circuit Detection

In recent years, electric vehicles (EVs) have gained significant traction within the automotive industry, driven by the societal push towards climate neutrality. These vehicles predominantly utilize lithium-ion batteries (LIBs) for storing electric traction energy, posing new challenges in crash safety. This paper presents the development of a mechanically validated LIB module simulation model specifically for crash applications, augmented with virtual short circuit detection. A pouch cell simulation model is created and validated using mechanical test data from two distinct out-of-plane load cases. Additionally, a method for virtual short circuit prediction is devised and successfully demonstrated. The model is then extended to the battery module level. Full-scale mechanical testing of the battery modules is performed, and the simulation data are compared with the empirical data, demonstrating the model’s validity in the out-of-plane direction. Key metrics such as force-displacement characteristics, force, deformation, and displacement during short circuit events are accurately replicated. It is the first mechanically valid model of a LIB pouch cell module incorporating short circuit prediction with hot spot location, that can be used in full vehicle crash simulations for EVs. The upscaling to full vehicle simulation is enabled by a macro-mechanical simulation approach which creates a computationally efficient model.

Reliability of cold-formed sections with web holes susceptible to failure by web crippling due to concentrated force

Cold-formed steel (CFS) are structural elements widely used in the construction industry and, in many situations, holes in the web is a design necessity, because it facilitates the passage of ducts, wiring and piping. This study evaluates the reliability of cold-formed sections with web holes subjected to concentrated force in sections without transverse stiffeners susceptible to web crippling failure. FOSM (Second Moment and First Order Method) and FORM (First Order Reliability Method) methods were used. A database was created with experimental results extracted from the literature, covering the four loading cases (end-one-flange loading – EOF, interior-one-flange loading – IOF, end-two-flange loading – ETF, interior-two-flange loading – ITF) presented in the standards used as reference, North American specification AISI S100 (2016), Brazilian standard NBR 14762 (2010) and European code EN1993-1-3 (2006). To obtain statistical data on the professional factor, one of the variables of the reliability problem, experimental results were compared with theoretical ones obtained by norm equations. Only AISI S100 (2016) standard has a criterion for considering web holes by considering a specified reduction factor for EOF and IOF. The results showed that the web holes significantly reduce the resistance of sections when subjected to the EOF loading case, while they reduce to a lesser extent in sections exposed to the IOF loading case. Analyzing the specimens subjected to the IOF loading, the results showed that another condition that influences the resistance is the fixation of the flanges to the supports. Very similar specimens groups reached the target reliability index when connected to the supports during the tests and did not reach when not connected.

Probabilistic Approach for Lateral Buckling Analysis in Virtual Anchor Spacing (VAS) Models of Rigid Flowlines Subjected to High Pressures and High Temperatures (HP/HT)

This study deals with the main topics related to the phenomenon of thermomechanical buckling of rigid flowlines laid on the seabed. It focuses on the latest versions of DNV-ST-F101 [1] and DNV-RP-F110 [2]. Additionally, the study presents a probabilistic approach for lateral buckling analyses in Virtual Anchor Spacing (VAS) of rigid flowlines exposed on the seabed, subjected to High Pressures and High Temperatures (HP/HT) in deep-water environments. The methodology includes determining the probability curves of the parameters with the greatest influence on the analysis: Pipe-Soil Interaction (PSI), friction between the pipe and sleeper (buckling mitigator), and geometric imperfections (lateral out-of-straightness). This is followed by a Design of Experiment (DoE) analysis to create a matrix composed of a representative set of load cases to be simulated using the Finite Element Method (FEM). These simulations are performed using Abaqus to obtain the probability distributions of the effective Critical Buckling Force (CBF) and Post-Buckling Force (PBF). The tolerable VAS is determined through FEM analyses applied to a dedicated load case matrix to identify the most critical failure criteria. The results are post-processed to generate response surfaces and failure probabilities for each of these criteria using the Monte Carlo Method. Finally, a case study is performed to illustrate the probabilistic approach, which requires more simulation time than a deterministic method but offers more detailed probability ranges for lateral buckling responses, potentially reducing a project's overall cost. It can be used to assess the reliability of lateral buckling responses during the detailed design phase of subsea pipelines susceptible to lateral buckling.

Finite Element Model Updating using Bayesian Optimization Algorithm

Structural engineering projects require reliable information about parameters of the designed systems that convey precise predictions about the behavior of the system under different load cases. These predictions are derived from numerical engineering models containing real structure geometry and parameters, resulting in responses using the well-established Finite Element Method (FEM). However, due to the presence of uncertainties and assumptions made during the construction of the model, the resulting response may not well represent the real structural behavior. Thus, the established approach involves supplying the numerical model with experimental data to reduce numerical-experimental error, improving the correlation between numerical and observed structural behavior of the system, in a process called Finite Element Model Updating (FEMU). Overall, engineers have developed many different approaches to solve this problem, mainly consisting of sampling or optimization algorithms, applying these methods in FEMU, damage detection, and optimization of sensor placement. The focus of this paper is on the implementation of the Bayesian Optimization Algorithm (BOA) to solve FEMU problems. BOA is a derivative-free optimization algorithm that aims to minimize the number of evaluations of the objective function by trading some computational resources for the construction and optimization of a cheaper auxiliary function, which evaluates the best point to evaluate next. This can be beneficial when heavy sampling of the optimization function can be very time-consuming or when limited evaluations are available, rendering sampling methods and some heuristic algorithms inefficient. To assess the behavior of the algorithm, many cases are analyzed from lower evaluation time to higher complexity cases and then compared with other methods, such as Bayesian Inference, Particle Swarm Optimization, and Genetic Algorithm. The first case is to optimize test functions, followed by some FEMU applications on an uncertain boundary condition beam and finally, a honeycomb plate, in which evaluation of the objective function is more expensive. Results show that BOA can attain good results in a short amount of time when applied to FEMU for low dimensionality problems compared with well-established methods. Overall, this paper reviews FEMU methods and proposes the implementation of a novel approach, demonstrating its effectiveness in solving model updating problems.

Boundary element method for buckling analysis of elastically connected double-beam system

Studies of Double-beam systems have received significant attention from researchers due to their wide applications in civil and mechanical engineering. Commonly, finite element method or exact solutions have been employed to solve double-beam systems, with few others considering buckling of axially loaded double-beam systems with classical boundary conditions. This paper presents a novel formulation of the Boundary Element Method (BEM) to determine the critical buckling load of the double-beam system elastically connected by a Winkler elastic layer with generalized boundary conditions. Based on the Euler-Bernoulli beam theory, the double-beam system is composed of two identical beams. This paper provides a detailed discussion of each step involved in the BEM, including the fundamental solution, boundary integral equations, and algebraic system. Examples considering different boundary conditions, material properties, and load cases are done and compared to corresponding analytical solution. The results show excellent agreement between the BEM approach and the analytical solution, confirming the accuracy and effectiveness of the technique.

Fatigue and Rutting Life Design of Layered Flexible Pavements Via Advanced MultiSmart3D

AbstractThis paper introduces a graphical user interface (GUI) built on the foundation of MultiSmart3D_Dynamics (or simply MultiSmart3D) code. Through this GUI, we calculate the pavement design parameters associated with fatigue cracking and rutting deformation. Besides the time-harmonic loading case, the moving loading case is further analyzed based on a simple relation between the moving velocity and frequency. Several numerical examples pertaining to static and dynamic (time-harmonic and moving) responses are studied for two typical pavement structures in Taiwan. The efficiency and accuracy of MultiSmart3D GUI are validated by comparing the present results with those from experimental loading frequency and real-world frequency. Our results reveal that the thickness of the asphalt concrete and base layers have a significant impact on fatigue and rutting lives. In addition, we have found that incorporating dynamic load into the structure may offer potential benefits in terms of cost reduction for optimizing the lifespan of the pavement.

Wear Behavior of 2024 Aluminum Alloy Reinforced with SiC Particles

The wear behavior of Aluminum Alloy (AA) 2024 reinforced with SiC particles were investigated at room conditions. Stir casting method was used to prepare the matrix composite of AA 2024 alloy reinforced with different weight percentages of SiC (3, 6, 9, 12 wt%). The wear behavior of both AA 2024 and AA 2024/SiC composite was studied using the reciprocating wear test. Loads of 2.5, 5, 7.5, 10, and 12.5 N were applied with sliding speeds of .0.55, 0.72, 0.88, 1.04 and 1.2 m/sec in dry sliding contact conditions. The microstructure and phase distribution were investigated using Optical Microscope (OM) and Scanning Electron Microscopy (SEM). The results demonstrate the presence of fine dendritic structure with semi-homogeneously distributed SiC particles in the alloy matrix. The micro-hardness increased with increasing SiC percentage. The highest value of hardness of 79.3 HV was found at 12 wt% SiC, from the initial 44 HV of the base alloy. The wear tests showed that the wear rate increased with increasing applied normal contact load at a contact sliding speed of 0.88 m/sec. The wear rate of the as cast material was 18.59 ×10-8 g/cm. Adding the SiC particles decreased the wear rate to 13.78, 9.76, 7.98, and 5.81 ×10-8 g/cm for 3, 6, 9, 12 wt% SiC addition at 12.5 N applied load, respectively. SEM examination of worn surfaces showed that severe and abrasive wear was exhibited at higher loads in the dry case while mild and adhesive wear were observed at lower loads in the lubricated condition.

System reliability analysis of cold-formed steel structures for serviceability limit states

The adoption of numerical analyses in structural design practice has ensured greater control over the actual behavior of structures. Design-by-analysis approaches deviate from traditional methods by directly considering the load-bearing capacity of the entire system rather than individual element checks. These approaches can even incorporate inherent nonlinearities into the analysis, such as second-order effects, material inelasticity, geometric imperfections, and semi-rigid connections. These advanced analysis models may produce results closer to the actual behavior of the structure in service, enabling the safe utilization of lighter structural systems. In this context, the evaluation of serviceability limit states becomes relevant, particularly for cold-formed steel structures, which are typically assembled using slender elements. This paper presents the application of a reliability-based procedure to assess the compliance of cold-formed steel frames with serviceability limit state criteria for excessive lateral displacement. Considering two groups of cold-formed steel structures based on their functions: single-story building frames and industrial storage rack frames, the lateral drift limits are extracted from the AISI S100, ASCE/SEI 7, ANSI MH 16.1, and Brazilian NBR 14762 and NBR 15524-2 standards, along with the target reliability indices for the corresponding limit state. The subsequent step involves evaluating system reliability, considering typical load cases for these structures (dead and wind loads for single-story frames and dead and product loads for racks) selected from the mentioned specifications, accounting for the uncertainties in loads and in the material and geometric parameters of the steel members. For this purpose, a MATLAB reliability framework is presented, coupling the First Order Reliability Method (FORM) to the MASTAN2 finite-element package, in order to apply the structural analysis results in each evaluation of the limit state functions. Conclusions regarding reliability indices and the potential adequacy of resistance factors are presented.

Hybrid data-driven and physics-informed regularized learning of cyclic plasticity with neural networks

Abstract An extendable, efficient and explainable Machine Learning approach is proposed to represent cyclic plasticity and replace conventional material models based on the Radial Return Mapping algorithm. High accuracy and stability by means of a limited amount of training data is achieved by implementing physics-informed regularizations and the back stress information. The off-loading of the neural network (NN) is applied to the maximal extent. The proposed model architecture is simpler and more efficient compared to existing solutions from the literature using approximately only half the amount of NN parameters, while representing a complete three-dimensional material model. The validation of the approach is carried out by means of results obtained with the Armstrong–Frederick kinematic hardening model. The mean squared error is assumed as the loss function which stipulates several restrictions: deviatoric character of internal variables, compliance with the flow rule, the differentiation of elastic and plastic steps and the associativity of the flow rule. The latter, however, has a minor impact on the accuracy, which implies the generalizability of the model for a broad spectrum of evolution laws for internal variables. Numerical tests simulating several load cases are presented in detail. The validation shows cyclic stability and deviations in normal directions of less than 2% at peak values which is comparable to the order of measurement inaccuracies.

Research on Stiffness Analysis and Technology of the Heavy Spidle Top

Background: The spindle top is an important component used to withstand the shaft workpiece on machine tools so that the spindle can meet high efficiency and high precision requirements. However, the selection principles under various load conditions are not stipulated in use. In addition, material selection, manufacturing, heat treatment technology, etc., are of practical significance for the production of high hardness, high wear resistance, and high precision spindle tops. Objectives: The spindle top type, material selection principles, heat treatment, cold working, and other manufacturing processes are given. Provide a reference for highperformance and top-notch design and manufacturing. Methods: The model of the spindle top will be created in UG software, then using ANSYS finite element analysis software to analyze stiffness of spindle top whose height-to-diameter ratios are 1:4 and 1:7 types in a variety of different load cases. The design and manufacturing process of the spindle top is analyzed and expounded from the selection and performance comparison of metal materials, heat treatment of different materials, cold manufacturing technology, and other aspects. Results: The deformation laws of different types of spindle tops are obtained. According to the deformation regular, find the selection principle of height to diameter ratio of spindle top. The defects that are easy to occur in the technology are obtained and the preventive and solution measures are put forward. Conclusion: According to the deformation regular, find the selection principle of height to diameter ratio of spindle top. The material selection, heat treatment technology, and other technical research on the spindle top provide the necessary basis for the production of the spindle top.