I-DCGAN and TOPSIS-IFP:A simulation generation model for radiographic flaw detection images in light alloy castings and an algorithm for quality evaluation of generated images

The intelligent detection technology driven by X-ray images and deep learning represents the forefront of advanced techniques and development trends in flaw detection and automated evaluation of light alloy castings.However,the efficacy of deep learning models hinges upon a substantial abundance of flaw samples.The existing research on X-ray image augmentation for flaw detection suffers from shortcomings such as poor diversity of flaw samples and low reliability of quality evaluation.To this end,a novel approach was put forward,which involves the creation of the Interpolation-Deep Convolutional Generative Adversarial Network(I-DCGAN)for flaw detection image generation and a comprehensive evaluation algorithm named TOPSIS-IFP.I-DCGAN enables the generation of high-resolution,diverse simulated images with multiple appearances,achieving an improvement in sample diversity and quality while maintaining a relatively lower computational complexity.TOPSIS-IFP facilitates multi-dimensional quality evaluation,including aspects such as diversity,authenticity,image distribution difference,and image distortion degree.The results indicate that the X-ray radiographic images of magnesium and aluminum alloy castings achieve optimal performance when trained up to the 800th and 600th epochs,respectively.The TOPSIS-IFP value reaches 78.7%and 73.8%similarity to the ideal solution,respectively.Compared to single index evaluation,the TOPSIS-IFP algorithm achieves higher-quality simulated images at the optimal training epoch.This approach successfully mitigates the issue of unreliable quality associated with single index evaluation.The image generation and comprehensive quality evaluation method developed in this paper provides a novel approach for image augmentation in flaw recognition,holding significant importance for enhancing the robustness of subsequent flaw recognition networks.

Optimal design of feeding system in steel casting by constrained optimization algorithms based on InteCAST

The traditional foundry industry has developed rapidly in recently years due to advancements in computer technology. Modifying and designing the feeding system has become more convenient with the help of the casting software, Inte CAST. A common method of designing a feeding system is to first design the initial systems, run simulations with casting software, analyze the feedback, and then redesign. In this work, genetic, fruit fly, and interior point optimizer(IPOPT) algorithms were introduced to guide the optimal riser design for the feeding system. The results calculated by the three optimal algorithms indicate that the riser volume has a weak relationship with the modulus constraint; while it has a close relationship with the volume constraint. Based on the convergence rate, the fruit fly algorithm was obviously faster than the genetic algorithm. The optimized riser was also applied during casting, and was simulated using Inte CAST. The numerical simulation results reveal that with the same riser volume, the riser optimized by the genetic and fruit fly algorithms has a similar improvement on casting shrinkage. The IPOPT algorithm has the advantage of causing the smallest shrinkage porosities, compared to those of the genetic and fruit fly algorithms, which were almost the same.

Digital management technology and its application to investment casting enterprises

With the advent of Industry 4.0, more and more investment casting enterprises are implementing production manufacturing systems, especially in the last two years. This paper summarizes three new common requirements of the digital management aspect in precision casting enterprises, and puts forward three corresponding techniques. They are: the production process tracking card technology based on the main-sub card mode; the workshop site production process processing technology based on the barcode; and the equipment data integration technology. Then, this paper discusses in detail the principle, application and effect of these technologies; to provide the reference for enterprises to move towards digital casting and intelligent casting.

In situ monitoring methods for selective laser melting additive manufacturing process based on images-A review

Selective laser melting(SLM)has been widely used in the fields of aviation,aerospace and die manufacturing due to its ability to produce metal components with arbitrarily complex shapes.However,the instability of SLM process often leads to quality fluctuation of the formed component,which hinders the further development and application of SLM.In situ quality control during SLM process is an effective solution to the quality fluctuation of formed components.However,the basic premise of feedback control during SLM process is the rapid and accurate diagnosis of the quality.Therefore,an in situ monitoring method of SLM process,which provides quality diagnosis information for feedback control,became one of the research hotspots in this field in recent years.In this paper,the research progress of in situ monitoring during SLM process based on images is reviewed.Firstly,the significance of in situ monitoring during SLM process is analyzed.Then,the image information source of SLM process,the image acquisition systems for different detection objects(the molten pool region,the scanned layer and the powder spread layer)and the methods of the image information analysis,detection and recognition are reviewed and analyzed.Through review and analysis,it is found that the existing image analysis and detection methods during SLM process are mainly based on traditional image processing methods combined with traditional machine learning models.Finally,the main development direction of in situ monitoring during SLM process is proposed by combining with the frontier technology of image-based computer vision.

Optimization of steel casting feeding system based on BP neural network and genetic algorithm

The trial-and-error method is widely used for the current optimization of the steel casting feeding system, which is highly random, subjective and thus ineff icient. In the present work, both the theoretical and the experimental research on the modeling and optimization methods of the process are studied. An approximate alternative model is established based on the Back Propagation(BP) neural network and experimental design. The process parameters of the feeding system are taken as the input, the volumes of shrinkage cavities and porosities calculated by simulation are simultaneously taken as the output. Thus, a mathematical model is established by the BP neural network to combine the input variables with the output response. Then, this model is optimized by the nonlinear optimization function of the genetic algorithm. Finally, a feeding system optimization of a steel traveling wheel is conducted. No shrinkage cavities and porosities are induced through the optimization. Compared to the initial design scheme, the process yield is increased by 4.1% and the volume of the riser is decreased by 5.48×10~6 mm3.

Effects of hot isostatic pressing temperature on casting shrinkage densification and microstructure of Ti6Al4V alloy

The Ti6Al4V alloy castings were produced by the investment casting process, and the hot isostatic pressing(HIP) was used to remove shrinkage from castings. The processing pressure and holding time for HIP were 150 MPa and 20 min, respectively. Four different HIP temperatures were tested, including 750 ℃, 850 ℃, 920 ℃ and 950 ℃. To evaluate the effects of temperature on densification and microstructure of Ti6Al4V alloy treated by HIP, non-destructive testing and metallographic observation was performed. The experimental results show that the shrinkage was completely closed at 920 ℃ and 950 ℃. The densification of Ti6Al4V alloy increased as the HIP temperature increased below 920 ℃. The lamel ae were more uniform, the thickness of lamel ae was obviously broadened and the structure was coarsen. Besides, the Norton creep equation was used to simulate the effect of different temperatures on the densification of Ti6Al4V alloy during HIP. The simulation results were in good agreement with the experimental results. It was also found that 920 ℃ is a suitable temperature for HIP for Ti6Al4V alloy.

Additive manufacturing and foundry innovation

Additive manufacturing is expected to transform and upgrade the traditional foundry industry to realize the integrated manufacturing and rapid and low-cost development of high-performance components with complex shapes.The additive manufacturing technology commonly applied in casting mold preparation(fusible molds,sand molds/cores and ceramic cores)mainly includes selective laser sintering(SLS)and binder injection three-dimensional printing(3DP).In this work,the research status of SLS/3DP-casting processes on material preparation,equipment development,process optimization,simulation and application cases in aerospace,automotive and other fields were elaborated.Finally,the developing trends of the additive manufacturing technology in the future of foundry field are introduced,including multi-material sand molds(metal core included),ceramic core-shell integration and die-casting dies with conformal cooling runners.

Temperature field simulations during selective laser melting process based on fully threaded tree

Selective laser melting(SLM) is a promising technique for additive manufacturing. During SLM of metallic powder, the temperature field and thermal history are important to understand physical phenomena involved. The purpose of this study is to simulate the temperature field during the SLM process of a hollow cylinder shape part based on a fully threaded tree(FTT) technique, and to analyze the temperature variation with time in different regions of the part. A revised model for temperature field simulation in the SLM process was employed, which includes the transition of powder-to-dense sub-model and a moving volumetric Gaussian distribution heat source sub-model. The FTT technique is then adopted as an adaptive mesh strategy in the simulation. The simulation result shows that during the SLM process of cylinder part, the temperature of inner powder bed is obviously higher than external one. The temperature at the internal of the molten pool is also higher than external, which may lead to differences in microstructures and other properties between the two regions.

基于改进蚁群算法的加热炉温度控制研究

加热炉温度控制系统具有非线性、时变性、滞后性等缺陷,导致系统控制过程中响应速度慢、抗干扰能力差,传统的控制方法无法对其进行精确控制。将免疫算法引入蚁群算法中,根据免疫算法中的亲和力原理,增加蚁群的多样性,并对蚁群的初始信息素规则进行改进。利用改进的蚁群算法来调节PIT神经网络(PIDNN)的权值,提出一种新型PITNN控制方法,并采用计算机仿真软件对其进行实验。仿真实验结果表明,与传统的PITNN控制方法相比,当采用改进蚁群算法的PIDNN控制器对加热炉进行控制时,系统达到稳态所需时间缩短了约34%;当加入扰动后,系统恢复到稳态所需时间减少了约26%,振动幅度明显降低,加热炉控制系统的抗干扰能力增强。

基于改进粒子群算法的PIDNN解耦控制研究（英文）

对于具有非线性、强耦合、大迟滞特点的多输入多输出的非线性系统,传统的控制方法无法对其进行精确解耦,导致控制精度较低。提出一种基于免疫机制的改进粒子群算法,并用此方法对PID神经网络权值进行优化,形成新型PID神经网络控制器。利用两个PID神经网络控制器对双输入双输出耦合系统进行控制以减弱系统的耦合影响。通过仿真结果表明:相对于传统PID解耦控制,该算法具有更好的动态和静态特性。可为控制领域中的解耦问题提供一定的参考。

Pyrolysis characteristics and thermal kinetics of expanded polystyrene(EPS) and styrene-methyl methacrylate(St-MMA) copolymer in LFC process

The pyrolysis behaviors of foam patterns have critical influences on fluid morphology and defect formation in Lost Foam Casting(LFC). The pyrolysis behaviors of expanded polystyrene(EPS) and styrenemethyl methacrylate(St-MMA) foams were compared using synchronous thermal analysis(STA), which was performed under argon atmosphere at different heating rates(from 10 to 40 K·min^(^(-1))). The degradation heat was calculated by integrating DSC curves. Results show that the calculated degradation heat of St-MMA(605.28 J·g^(-1)) was significantly lower than that of EPS(706.71 J·g^(-1)). Furthermore, the non-isothermal iso-conversional method was used to determine the pyrolysis apparent activation energies of EPS and St-MMA, and results show that the activation energy of St-MMA(200.36 kJ·mol^(-1)) was apparently higher than that of EPS(167.92 kJ·mol^(-1)). These calculated results indicate that the weight loss rate of EPS is greater than St-MMA in the pyrolysis process. In addition, the apparent activation energies at various pyrolysis stages demonstrate that the pyrolysis reactions of EPS and St-MMA may involve physical and chemical changes in the decomposition layer of the LFC process.

Finite element simulation of real cavity closure in cast Ti6Al4V alloy during hot isostatic pressing

The healing behavior of shrinkage cavity inside the cast Ti6Al4V alloy during hot isostatic pressing(HIP)was investigated experimentally by interrupted hot isostatic pressing tests.The X-ray micro computed tomography was used to record the morphology changes before and after hot isostatic pressing.The two-dimensional geometry obtained by the microCT scan was used in simulation to study the evolution of the real shrinkage cavity during hot isostatic pressing.Shrinkage cavities,shrinkage porosity and small gas pores can be effectively eliminated under proper HIP conditions.The two-dimensional morphology in the simulation results agrees well with the experimental results.This study reveals that plastic deformation,creep and diffusion are the main mechanisms of cavity closure during hot isostatic pressing.In addition,the simplified elliptical pores with aspect ratios at different positions were used to replace the real pores to further study the factors affecting the position of dimples after HIP by simulation.It is found that the position of the dimples mainly depends on the aspect ratio of the elliptical pore and the distance between the pore surface and the external surface of the geometric model.

Effect of alloying elements W,Ti,Sn on microstructure and mechanical properties of gray iron 220

Experiments were carried out to observe the variation in microstructure and mechanical properties of gray cast iron by adding pearlite promoting alloying elements such as Ti,Sn and W.Results show that adding Sn,Ti,and W with different concentrations improve the microstructure,Brinell hardness and tensile strength of gray cast iron.With the increase of alloying element concentration,the average graphite length and graphite content increase linearly.At the same time,average cell size and the maximum graphite length also decrease linearly.Brinell hardness and tensile strength of gray cast iron also increase with an increase in alloying elements contents,and attain the maximum when Ti=0.561%,Sn=0.561%and W=0.945%.However,at higher concentrations of Ti=0.810%,Sn=0.631%and W=1.351%,the tensile strength decreases from 333 MPa to 297 MPa and the Brinell hardness decreases from 248 HB to 225 HB.The decrease in tensile strength and Brinell hardness at the higher concentration level is attributed to the formation of coarse and thick graphite flakes.

基于改进PSO神经网络的板形板厚解耦控制研究

针对板形板厚综合系统具备强耦合、非线性、大时滞等特性,传统的控制方法无法对其完成精确解耦,导致控制精度较低。提出一种基于免疫机制的改进粒子群算法,同时借助此算法来优化处理PID神经网络(PIDNN),形成新型PIDNN控制器。利用两个PIDNN解耦控制器对板形板厚综合系统进行控制以降低系统耦合影响。通过仿真结果可以看出,在动态性能与静态性能上,此算法较以往PIDNN解耦控制均存在明显优势。可为控制领域中的解耦问题提供一定的参考。

基于混沌PSONN-PID神经网络的热连轧活套系统自适应解耦控制（英文）

To improve the performance of PSO （particle swarm optimization） optimization algorithm, a new algo-rithm - CPSO （chaotic particle swarm optimization）was proposed. The algorithm chaotic search mechanismwas introduced to the particle swarm algorithm to increase the diversity of particle. In order to improve the dimin-ishing policy flaws, the algorithm also adopts the methods of increasing particle interaction strategy and the first-increased-then-decreased inertia weight factor model to set inertia weight factor. The improved algorithm and PIDsingle neuron are combined, and which is used in hot rolling looper decoupling control system. Simulation resultsshow that the algorithm can overcome the defects of PSO in prematureness and being easy to fall into local opti-mum. This research puts forward a new and effective way to solve the high tension coupling problem in looper sys-tem.

热连轧活套系统解耦控制（英文）

热连轧机中的活套系统是多变量强耦合、非线性、多约束的复杂对象,其解耦控制问题一直是控制界关注的热点。采用PID神经元网络解耦控制方法来消除活套高度和轧件张力之间的耦合,PID神经元网络连接权值由粒子群算法进行学习优化。仿真结果表明所建模型和所提出控制方法的有效性。新的方法可在大范围内克服系统的非线性和强耦合问题,具有较强的鲁棒性。

Global heat transfer model and dynamic ray tracing algorithm for complex multiple turbine blades of Nibased superalloys in directional solidification process

High-quality solidification microstructure during directional solidification relies on precise temperature gradient control, so accurate calculation of the temperature field is critical. In this study, a 3D transient global heat transfer model of directional solidification by the Bridgman method based on the finite difference method is developed. The radiation heat in this model is calculated by the discrete transfer method, and a modified method of external surface area for irregular geometric models is proposed to reduce the zigzag shape caused by finite difference grids. Considering the radiative heat transfer between any surface elements of all materials in the directional solidification furnace, a dynamic ray tracing algorithm is developed to simulate the entire process of directional solidification. Then, the simulated results are compared with the theoretical results and experimental results, respectively. Finally, based on the present model and method, the simulation program developed is applied to the directional solidification of actual castings. The simulated results are in good agreement with the experimental results, which indicate that the model and method developed in this study is effective and practical.

Multi-scale coupling simulation in directional solidification of superalloy based on cellular automaton-finite difference method

Casting microstructure evolution is difficult to describe quantitatively by only a separate simulation of dendrite scale or grain scale, and the numerical simulation of these two scales is difficult to render compatible. A three-dimensional cellular automaton model couplling both dendritic scale and grain scale is developed to simulate the microstructure evolution of the nickel-based single crystal superalloy DD406. Besides, a macro–mesoscopic/microscopic coupling solution algorithm is proposed to improve computational efficiency. The simulation results of dendrite growth and grain growth of the alloy are obtained and compared with the results given in previous reports. The results show that the primary dendritic arm spacing and secondary dendritic arm spacing of the dendritic growth are consistent with the theoretical and experimental results. The mesoscopic grain simulation can be used to obtain results similar to those of microscopic dendrites simulation. It is indicated that the developed model is feasible and effective.

Phase field study on crystal orientation effects in eutectoid phase transformation

In this present work, a multi-phase field model was used to simulate the eutectoid transformation process, and on the basis of the nucleation model that was previously proposed by our research team, anisotropic and orientation relationship models were introduced to study the growth mechanism of the pearlite lamellae with anisotropy. It was found that the growth direction of the pearlite lamellae is related to its orientation and spacing. In the process of lamellar growth, deflection growth of pearlite will appear along with the adjustment of lamellar spacing, and the deflection angle is equal to the orientation difference between the austenite and the pearlite. Comparison between experimental and numerical results indicates a good consistency in pearlite morphology.

Life-threatening hemorrhage after liver radiofrequency ablation successfully controlled by transarterial embolization

A 59-year-old man underwent liver radiofrequency ablation under laparotomy for recurrent hepatic carcinoma located in the caudate lobe,however,near-fatal bleeding occurred 1 wk after the operation.The intraoperative ultrasound study during laparotomy revealed left hepatic artery pseudoaneurysm.Suture and packing with ribbon gauze was used to obtain hemostasis.A secondary hemorrhage followed 11 h later and hepatic angiography was performed immediately.Bleeding from the pseudoaneurysm in a branch of the left hepatic artery was found and the artery branch was embolized with coils.Other than slight bile leakage,post-embolization continued satisfactorily.Bleeding did not reoccur.The follow up visit 1 mo later found the pseudoaneurysm disappearing and no tumor recurrence.