Optimal Combination Of Process Parameters Research Articles

Multi-Objective Optimization of the Forming Process Parameters of Disc Forgings Based on Grey Correlation Analysis and the Response Surface Method

This paper introduces a multi-objective optimization problem (MPO) for the forming process parameters of disc forgings using grey relational analysis (GRA) and the response surface methodology (RSM). Firstly, an experimental design based on the Box–Behnken design (BBD) principle was established, and simulations were performed in Deform to obtain response data. Secondly, GRA was used to transform the MPO into a grey relational degree (GRD) problem, and the entropic weight method was integrated to ascertain the influence weights of each variable on GRD. Then, a quadratic polynomial prediction model based on the RSM was constructed, and its accuracy was ensured through model validation. Finally, the optimal process parameter combination was determined through the particle swarm optimization algorithm, which included a friction coefficient of 0.3, an initial temperature of 1250 °C, and a downward pressing speed of 7.5 mm/s. The results of the experimental investigation indicate that optimized process parameters significantly reduce the forming load, equivalent stress, and damage value, effectively enhancing the overall quality of forged parts.

Best practices for machine learning strategies aimed at process parameter development in powder bed fusion additive manufacturing

AbstractThe process parameters used for building a part utilizing the powder-bed fusion (PBF) additive manufacturing (AM) system have a direct influence on the quality—and therefore performance—of the final object. These parameters are commonly chosen based on experience or, in many cases, iteratively through experimentation. Discovering the optimal set of parameters via trial and error can be time-consuming and costly, as it often requires examining numerous permutations and combinations of parameters which commonly have complex interactions. However, machine learning (ML) methods can recommend suitable processing windows using models trained on data. They achieve this by efficiently identifying the optimal parameters through analyzing and recognizing patterns in data described by a multi-dimensional parameter space. We reviewed ML-based forward and inverse models that have been proposed to unlock the process–structure–property–performance relationships in both directions and assessed them in relation to data (quality, quantity, and diversity), ML method (mismatches and neglect of history), and model evaluation. To address the common shortcomings inherent in the published works, we propose strategies that embrace best practices. We point out the need for consistency in the reporting of details relevant to ML models and advocate for the development of relevant international standards. Significantly, our recommendations can be adopted for ML applications outside of AM where an optimum combination of process parameters (or other inputs) must be found with only a limited amount of training data.

Experimental investigation on in-situ laser-assisted machining of single crystal silicon based on response surface methodology

Abstract In this paper, Response Surface Methodology (RSM) was utilized as a robust and convenient predictive tool to establish the correlation between process parameters in in-situ laser-assisted machining and the surface roughness of single-crystal silicon. Based on the optimally designed diamond tool, a novel temperature field analysis method and RSM. The contribution rate of each process parameter on surface roughness was laser power > rotation speed > cutting depth > feed rate. The optimal process parameter combination is: rotation speed as 1001r/min, feed rate as 4.9μm/r, cutting depth as 7.55μm, and laser power as 28.81 W. Experimental validation of these optimal parameters compared surface roughness values obtained experimentally with those predicted. The surface roughness model showed a maximum relative error of 5.2%, with an average error of 4.8% across three experimental sets. These errors are within acceptable limits, indicating an alignment between predicted and experimental results.

Microstructure phase analysis and process optimization of ZL205A metal mold casting

In this paper, we take an elastomeric transmission structure part made of ZL205A alloy as the research object, and use the metal type gravity casting technology for casting trial production, analyze the quality and mechanical properties of the molded parts of ZL205A alloy parts under different process parameters, optimize the casting process parameters, get the optimal combination of process parameters, and obtain large ZL205A castings with less internal quality defects and high mechanical properties.

Simulation analysis and test of the spinning process of 304 stainless steel cylindrical parts

Abstract Taking the difference in wall thickness as the evaluation index, the simulation analysis of the spinning process of 304 stainless steel cylindrical parts is carried out. The influence of the roller nose radius, feed speed, and percent thickness reduction on the stress and strain of the roller is studied, and the optimum combination of process parameters is obtained. Through the spinning test, the stainless steel cylindrical parts with an outside diameter of 300.8 mm are successfully spun. After spinning forming and solution treatment, the tensile strength of stainless steel reaches 557~566 MPa, the yield strength reaches 220~232 MPa, and the elongation reaches 41.3%~42.5%.

Optimization of process parameters for Trifolium pratense L. seed granulation coating using GA-BP neural network

Regarding the issue of low granulation qualification rates during the granulation coating of red clover seeds, this study theoretically analyzed the force conditions of seeds and powder particles under the action of liquid to obtain the main factors affecting seed coating quality. During the seed granulation coating process, an intermittent powder supply method combined with continuous liquid supply was utilized to control the ratio of powder to liquid. Using the granulation qualification rate as the evaluation index, single-factor experiments were conducted to investigate the effects of coating pan fill ratio, single powder supply amount, powder supply interval, and liquid supply amount on the quality of red clover seed granulation coating. Based on the results of the single-factor experiments, orthogonal experiments were conducted, revealing that the interaction of factors would influence the experimental results. To further optimize the quality of seed granulation coating, the mechanisms of powder and liquid in the adhesion process on granulation coating were explored. Orthogonal experiments were conducted on the process parameters of the granulation coating machine, and the GA-BP model was employed for optimization and solution. The optimal process parameter combination obtained was a coating pan fill ratio of 33.78 %, a single powder supply amount of 5.17 g, a powder supply interval of 7.7 s, and a liquid supply amount of 0.42 mL/s. Under this optimal parameter combination, granulation coating experiments with red clover seeds were performed, and the seed granulation coating quality was relatively high, with a granulation qualification rate of 97.7 %. The research results can provide a reference for optimization experiments on coating irregular seeds.

Influence of spatial position of sound source on dual-frequency ultrasound-assisted preparation of magnesium alloy

In this work, a dual-frequency ultrasonic field was employed to improve the quality and mechanical properties of magnesium alloy ZW61 ingot. The effect of the spatial position of the sound source on the cavitation volume, microstructure size, and morphology was discussed in detail combined with numerical simulations and experiments. The mechanical properties were analyzed elaborately and the relationship between yield strength (YS), ultimate tensile strength (UTS), and average grain size (d) was established. The results showed that the dual-frequency ultrasonic field (DUF) had a larger cavitation range and superior grain refinement ability than the single-frequency ultrasonic field (SUF) at the same energy consumption, with the grain refining from 578 μm (untreated) and 395–433 μm (SUF) to 55.2 μm. The grain refining effect of DUF weakened with increasing sound source spacing, with d increasing from 55.2 μm to 163 μm. The radiating angle had little effect on the grain size. Too shallow or too deep inserting depth of the ultrasonic rods could seriously weaken the grain refinement. In this case, 30 mm source spacing, 60° radiating angle, and 30 mm inserting depth were the optimal process parameter combinations, and the ingot strength was increased from YS of 83 MPa and UTS of 151 MPa to 138 MPa and 241 MPa, respectively. YS and d for all conditions conformed to the Hall-Petch relationship, with σ0 and k values of 71.3 MPa and 503.9 MPa μm−1/2. The relationship between UTS and d could be expressed as UTS = 126.7 + 907.0 d−1/2.

Optimization of multiple responses in the laser cladding process parameters

ABSTRACT An orthogonal test method was used to optimize parameters for Ni60A-25% tungsten carbide (WC) laser cladding layers. Meanwhile, the tensile properties of the optimized laser cladding layer are explored. A multi-factor analysis of variance (ANOVA) was conducted to examine the effect of factors (laser power-A, scanning speed-B, powdering rate-C, and spot diameter-D) and interactions on responding variable. The results showed that the interaction of scanning speed and powdering rate with spot diameter has a more pronounced impact on the dilution rate compared to a single factor. Based on the ANOVA of the dilution rate, the optimal combination of process parameters was A3C2B1D3. Furthermore, the optimal combination of process parameters was C3B2D3A3 for the effective area. The results showed that the yield strength of the specimens was 512.3 MPa and 509.7 MPa, and the tensile strength was 646.3 MPa and 605.7 MPa for process parameters I and II, respectively.

Sensitivity analysis-based quality optimization strategy for multi-correlation parameters in spinning process

To address the problem of quality optimization of multi-correlation parameters in the spinning process, this paper proposes a new method based on a sparrow search algorithm (SSA). Firstly, a generalized regression neural network (GRNN) is used to investigate the impact of the spinning process parameters on yarn quality, and quality forward modeling in the spinning process is established. And based on the coupling and correlation characteristics of spinning process parameters, sensitivity analysis is used to analyze the influence of each spinning process parameter on yarn quality, the correlation spinning process parameters for further analysis. Then a model of quality optimization with spinning process parameters is established, and SSA is used to solve the model of quality optimization with multi-correlation parameters in the spinning process. Finally, the effectiveness of the proposed method was validated through an instance. The results show that the optimal spinning process parameters combination generation of [32.159 5.2 0.8 14.8 24.540 8588.677 21.708] occurs in a configuration with a fitness value of 0.0003. The proposed sensitivity analysis-based quality optimization strategy reveals good performances in terms of both convergence speed and optimization accuracy, which will provide guidance for improving yarn quality.

Optimization of laser-based glass frit bonding for optoelectronic packaging using response surface methodology

Laser-based glass frit bonding presents considerable advantages, including fast bonding speed and minimal thermal impact, showing significant application prospects in the packaging of temperature-sensitive optoelectronic devices. In this study, the individual effects of critical process parameters on bonding performance have been investigated through single-factor experiments. Response surface methodology is employed to design experiments systematically for the laser-based glass frit bonding process, focusing on exploring the interactive effects among these process parameters. The Box-Behnken design method is implemented for experimental setup, and a regression model has been developed to analyze the resulting data. Analysis of variance, 3D response surface plots, and contour plots are conducted to comprehensively evaluate the interaction effects of parameters on the bonding seam width and strength. The mathematical model is validated and an optimal combination of process parameters is identified, which significantly enhances the bonding quality. The optimized process parameters are also applied for the laser bonding of square-shaped seam to validate the feasibility of the optimized results and to promote the engineering application of glass laser-based glass frit bonding technology in microelectronic component packaging. Inspection of the bonded specimens shows excellent formation quality of the square-shaped bonding seams, confirming the methodology’s effectiveness in advancing optoelectronic packaging technology.

RSM-BASED STATISTICAL APPROACH TO ENHANCE THE COMPRESSIVE PERFORMANCE OF ABS P400 PARTS FABRICATED VIA FDM TECHNOLOGY

This research investigates the performance of ABS P400 parts under compressive loading. These ABS P400 parts are fabricated by fused deposition modeling (FDM) process using omnidirectional printing. The effects of three critical parameters, raster angle (A), air gap (B), and raster width (C), each at three different levels, were analyzed on three different performance measures, i.e. total deformation up to failure ([Formula: see text]) measured in %, maximum energy ([Formula: see text]) in MJ/m3 and break energy ([Formula: see text]) in MJ/m3. Results of experimentations were analyzed using analysis of variance (ANOVA), perturbation curve and 3D surface plots. The research established the anisotropic nature, durability and less brittleness of the FDM fabricated ABSP400 part, which leads to lower strength and also influence the energy absorbing behavior while deformation under compression. Complex relationships were exhibited between the chosen parameters and studied measures. It was also observed that [Formula: see text] is noticeable due to less brittleness of ABS P400 however, [Formula: see text] and [Formula: see text] vary in a contradictory fashion with changes in process parameters. The models were validated using normality plots. Multi-objective optimization was done using the desirability approach to determine the optimal combination of FDM process parameters for optimum responses. The desirability result showed that the optimized input parameter is [Formula: see text][Formula: see text]mm, and [Formula: see text][Formula: see text]mm which yield maximized optimum responses as [Formula: see text]%, [Formula: see text][Formula: see text]MJ/m3, and [Formula: see text][Formula: see text]MJ/m3. The work not only reflects the effect of three FDM parameters on the total deformation up to failure ([Formula: see text] in %) but, in addition, energy absorption behavior of FDM parts under compression is also investigated. The study on the behavior of energy at ultimate stress or ultimate load and breaking stress or breaking load were rarely investigated in earlier research which reflects the novelty of our research.

Optimization of Agricultural Tractor Engine Hood Forming Quality Based on Neural Network Genetic Algorithm Function

To improve the forming quality of agricultural tractor engine hood and effectively solve the problems of wrinkling and cracking during the drawing forming process, DYNAFORM 5.9 finite element software was used for numerical simulation of the part. Based on a comprehensive scoring method of orthogonal experiments and grey relational analysis, the critical process parameters affecting the “maximum thickening rate” and “maximum thinning rate” were determined. Latin hypercube sampling was used to randomly sample the key process parameters, and the sampled samples served as the data basis for the optimization stage of the neural network genetic algorithm function; an optimization model of the neural network genetic algorithm function was established, with the key process parameters, and the “maximum thickening rate” and “maximum thinning rate” as inputs and outputs respectively, to construct a nonlinear mapping relationship and optimize the process parameters. The study showed that the blank holder force and die clearance had the greatest impact on the “maximum thickening rate” and “maximum thinning rate”, identifying them as key process parameters; the optimal process parameter combination was the blank holder force of 462.71 kN, the die clearance of 1.09 mm, and corresponding “maximum thickening rate” and “maximum thinning rate” of 3.97% and 25.96%, respectively. Overall, this research provides a practical optimization strategy to solve the quality issues of the engine hood, offering a theoretical basis for its actual production and processing with significant practical application value.

An Experimental Study in Laser-Assisted Machining of AerMet100 Steel.

To solve the problems of poor surface quality and low tool life in conventional machining (CM) of AerMet100 steel, an experimental study was conducted in laser-assisted machining (LAM) of AerMet100 steel. The effects of laser power, cutting speed, feed rate, and depth of cut on the surface roughness of AerMet100 steel were studied based on a single-factor experiment. The degree of influence of each factor on the surface roughness was evaluated by analyses of variance and range in the orthogonal experiment, and the combination of process parameters for the optimal surface roughness was obtained. The order of influence was as follows: laser power > cutting speed > depth of cut > feed rate; the optimal combination of process parameters was laser power 200 W, cutting speed 56.5 m/min, feed rate 0.018 mm/rev, and depth of cut 0.3 mm. Compared to CM, the surface morphology of the workpiece under the optimization of LAM was relatively smooth and flat, the surface roughness Ra was 0.402 μm, which was reduced by 62.11%, the flank wear was reduced from 208.69 μm to 52.17 μm, there were no tipping or notches, and the tool life was significantly improved. The study shows that the LAM of AerMet100 steel has obvious advantages in improving surface quality and reducing tool wear.

Parametric investigation of wire-EDM parameters for micro channel as a foundation of high-quality micro needle array fabrication

Micro channels are widely used in microfluidics, and medical science for the extraction of DNA, electrophoresis, and blood protein analysis. In the current work, the wire electrical discharge machining (WEDM) process was used to fabricate micro channels on the stainless steel 316 (SS316) substrate. To check the feasibility of fabricating high-quality micro channels and thereafter micro-scale structures with the WEDM process the effect of different process parameters such as pulse-off time, pulse-on time, servo gap reference voltage and wire feed rate were investigated. The quality attributes taken were namely surface roughness of the channel base, overcut, micro channel depth, and material removal rate (MRR). The pulse-on time significantly affects surface roughness and MRR. The servo gap reference voltage is the key process parameter affecting the overcut and micro channel depth. It was observed from the parametric study that 15 µs pulse-on time, 35 µs pulse-off time, 17 v servo gap reference voltage, and 8 m/min wire feed rate are the optimum levels for getting high MRR, low surface roughness, less overcut and high micro channel depth. Additionally, the topographical, morphological, surface, and sub-surface characterization were also carried out to assess the surface integrity and extent of thermal damage caused by the WEDM process. With the optimum combination of process parameters a 6 × 6 micro needle array was successfully fabricated on the same substrate using micro channels as a basis of the fabrication process for the first time using the WEDM process.

Multi-objective optimization of high-power fiber laser cutting process using data augmentation-based ANN-Adam model

During fiber laser cutting of sheet metal, excessive heat-affected zone (HAZ) and recast layer thicknesses (RLT) can adversely affect subsequent precision machining and welding processes on the parts. As a result, it is necessary to optimize the process parameters to control the HAZ and RLT. In this work, laser power, gas pressure, focal plane position (FPP), and cutting speed are selected as input parameters. The cutting experiments are conducted using response surface methodology, and the factors that significantly affect the cutting quality are identified by analysis of variance (ANOVA). To improve the generalization ability and robustness of the model, a data augmentation-based artificial neural network (ANN) model is developed to achieve accurate prediction of cut quality characteristics. Further, this study performs a multi-objective optimization of the cutting process parameters using the Adam optimizer. The research results show that the mean relative errors of the ANN model in predicting the HAZ and RLT are 2.44 % and 1.55 %, respectively. The determination coefficients R2 are 0.9542 and 0.9779, respectively. The optimal combinations of process parameters obtained through optimization are laser power 8.9 kW, gas pressure 0.8 bar, FPP 11.8 mm, and cutting speed 1.3 m/min. Experimental validation shows that the relative errors between the predicted and the experiment values of the cut quality characteristics are kept within 5 % under the optimal process parameters. This fully verifies the effectiveness of the optimization model in guiding the selection of laser-cutting process parameters and improving cutting quality.

Orthogonal experimental study of ultrasound-assisted fused deposition molding

To improve the quality of fused deposition molding (FDM), we introduced ultrasonic vibration into the molding process to enhance the strength of the tensile samples. We used the response surface method to optimize the FDM molding process. We carried out orthogonal experiments to investigate the effects of ultrasonic power and process parameters (extrusion temperature, hotbed temperature, and delamination thickness) on the mechanical properties of fused deposition acrylonitrile butadiene styrene products. We obtained the quadratic regression equations and response surfaces of the data through analysis of variance and significance analysis by software. Last, we received the optimum combination of process parameters for tensile strength as extrusion temperature 233.85 °C, hotbed temperature 93.38 °C, delamination thickness 0.1 mm, and ultrasonic power 43.09 W.

Parametric analysis, modeling and optimization of the process parameters in electric discharge machining of aluminium metal matrix composite

Optimizing electric discharge machining (EDM) for aluminum/SiCp metal matrix composites poses challenges due to intricate machine parameters and process complexity, impacting process economy and elevating product costs. The research aims to find the optimal combination of process parameters which include pulse on-time, pulse current, duty cycle (%), gap voltage, sensitivity and flushing pressure for EDM of Al/SiCp-MMC using a copper electrode for the selected response factors such as material erosion rate and surface roughness, Ra. The experiments were designed using the central composite design of response surface methodology and an advanced optimization technique known as Teaching–learning-based optimization (TLBO), is applied to find the optimal combination of process parameters to obtain maximum material erosion rate subject to the desired range of surface roughness (SR), Ra. The combination of the high pulse on-time (i.e. 150 μs) and high pulse current (i.e. 12A) results in high material removal rate with deep craters on the machined surface clearly visible in SEM images contrasting the minimized surface roughness at lower values of pulse on-time (50 μs) and the pulse current (4A). Pulse on - time (Ton) is found to be the most significant factor for material erosion rate and surface roughness with percentage contribution of 70.86 and 54.9 respectively for optimization of the response. The regression models were developed at 95% confidence level for material removal rate and surface roughness with R2 value of 0.93 and 0.95 respectively signifying high degree of accuracy in predicting the response. Confirmation tests conducted to check the adequacy of the established models revealed that the percentage error between the predicted and experimental responses is found to be within acceptable levels. Electron discharge machining of the aluminium metal matrix composite at the optimized conditions could provide economical aspect in the aerospace and automobile industry.

Research on cutting force characteristics of laser ultrasonically assisted turning of cemented carbide

Cemented carbide possesses excellent characteristics such as high strength and hardness, but its poor machinability has restricted its widespread application. Laser ultrasonically assisted turning (LUAT) can enhance the machinability of workpieces and is widely used in precision machining of difficult-to-machine materials. In this paper, LUAT is applied to the cutting of cemented carbide, and the characteristics of cutting force in the machining process are studied. A theoretical model of cutting force is established by analyzing the interaction between the cutting tool, chip and workpiece during the turning process. The maximum error was found to be 18.6% by comparing with the experimental data, which better predicted the changing trend of cutting force and verified the accuracy of the model. Furthermore, experiments on LUAT of cemented carbide are carried out using the Taguchi method. The influence of cutting tools and processing parameters on cutting force is analyzed, revealing that the amplitude has the most significant impact on cutting force. The optimal combination of processing parameters is also obtained. Finally, a comparison is made between different turning methods. Compared to conventional turning, laser assisted turning, and ultrasonically assisted turning, the total cutting force in LUAT is reduced by 53.44%, 45.62%, and 21.64%, respectively, resulting in improved surface quality of the workpiece.

Numerical Simulation of Lost-Foam Casting for Key Components of A356 Aluminum Alloy in New Energy Vehicles.

The intricate geometry and thin walls of the motor housing in new energy vehicles render it susceptible to casting defects during conventional casting processes. However, the lost-foam casting process holds a unique advantage in eliminating casting defects and ensuring the strength and air-tightness of thin-walled castings. In this paper, the lost-foam casting process of thin-walled A356 alloy motor housing was simulated using ProCAST software (2016.0). The results indicate that the filling process is stable and exhibits characteristics of diffusive filling. Solidification occurs gradually from thin to thick. Defect positions are accurately predicted. Through analysis of the defect volume range, the optimal process parameter combination is determined to be a pouring temperature of 700 °C, an interfacial heat transfer coefficient of 50, and a sand thermal conductivity coefficient of 0.5. Microscopic analysis of the motor housing fabricated using the process optimized through numerical simulations reveals the absence of defects such as shrinkage at critical locations.

Optimization of porosity behavior of hybrid reinforced titanium metal matrix composite through RSM, ANN, and GA for multi-objective parameters

Titanium matrix composites (TMCs) have high specific strength and stiffness, and high-temperature TMCs can reduce weight by up to 50% when compared with monolithic super alloys while preserving equal stiffness and strength in jet engine systems for propulsion. The purpose of this work examines the use of mathematical models and learning approaches to optimize response such as porosity and control variables in synthesized hybrid titanium metal matrix composites (HTMMCs) reinforced by B4C-SiC-MoS2-ZrO2. To further understand the impacts of process factors on porosity reduction, the study employs methodologies such as the response surface methodology (RSM), integrated artificial neural networks (ANN), and genetic algorithm (GA). The findings indicate that these strategies have the potential to contribute to the industry. The optimal combination of 7.5wt.% SiC, 7.5wt.% B4C, 7.5wt.% ZrO2, 4wt.% MoS2, and 73.5wt.% Ti compositions was determined utilizing process factors such as milling period (6h), compaction pressure (50MPa), compact duration (50min), sintering temperature (1200°C), and sintering time (2h) as compared to pure Ti grade 5. The mechanical properties of the optimum combination of reinforcement weight percentage and process parameters resulted in a minimum porosity of 0.118%, density of 4.36gcm3, and micro-hardness of 63.4HRC boosted by 1.76%, and compressive strength of 2500MPa increased by 2.6%. In addition, these HTMMCs had a minimal wear rate of 0.176mm3/Nm and a corrosion resistance rate of 2.15×10-4mmpy. The investigation result analysis discovered that the RSM and combined ANN-GA models considerably enhanced the forecasting of multidimensional interaction difficulties in composite material production that were highly statistically connected, with R2 values of 0.9552 and 0.97984. The ANN-GA model provided a 95% confidence range for porosity predictions, which increased the production use of titanium-based particle composites. Furthermore, HMMCs can be utilized in the automotive and aviation industries with enhanced corrosion and wear resistance.