A geometrical model for surface roughness prediction when face milling Al 7075-T7351 with square insert tools

Surface quality is important in engineering and a vital aspect of it is surface roughness, since it plays an important role in wear resistance, ductility, tensile, and fatigue strength for machined parts. This paper reports on a research study on the development of a geometrical model for surface roughness prediction when face milling with square inserts. The model is based on a geometrical analysis of the recreation of the tool trail left on the machined surface. The model has been validated with experimental data obtained for high speed milling of aluminum alloy (Al 7075-T7351) when using a wide range of cutting speed, feed per tooth, axial depth of cut and different values of tool nose radius (0.8 mm and 2.5 mm), using the Taguchi method as the design of experiments. The experimental roughness was obtained by measuring the surface roughness of the milled surfaces with a non-contact profilometer. The developed model can be used for any combination of material workpiece and tool, when tool flank wear is not considered and is suitable for using any tool diameter with any number of teeth and tool nose radius. The results show that the developed model achieved an excellent performance with almost 98% accuracy in terms of predicting the surface roughness when compared to the experimental data.     

Kernel based regression and genetic algorithms for estimating cutting conditions of surface roughness in end milling machining process

We observe a surface roughness in end milling machining process which is influenced by machine parameters, namely radial rake angle, speed and feed rate cutting condition. In this machining, we need to minimize and to obtain as low as possible the surface roughness by determining the optimum values of the three parameters. In previous works, some researchers used a response surface methodology (RSM) and a soft-computing approach, which was based on ordinary linear regression and genetic algorithms (GAs), to estimate the minimum surface roughness and its corresponding values of the parameters. However, the construction of the ordinary regression models was conducted without considering the existence of multicollinearity which can lead to inappropriate prediction. Beside that it is known the relation between the surface roughness and the three parameters is nonlinear, which implies that a linear regression model can be inappropriate model to approximate it. In this paper, we present a technique developed using hybridization of kernel principal component analysis (KPCA) based nonlinear regression and GAs to estimate the optimum values of the three parameters such that the estimated surface roughness is as low as possible. We use KPCA based regression to construct a nonlinear regression and to avoid the effect of multicollinearity in its prediction model. We show that the proposed technique gives more accurate prediction model than the ordinary linear regression’s approach. Comparing with the experiment data and RSM, our technique reduces the minimum surface roughness by about 45.3% and 54.2%, respectively.     

熔融沉积制造表面粗糙度参数显著性及预测模型研究

为了确定影响熔融沉积制造(FDM)打印件表面粗糙度的显著性因素,设计了基于 温度、打印速度和层厚的9 组正交实验。通过探针式粗糙度仪测量打印件表面粗糙度,并进行 了信噪比计算和波动分析,确定了影响表面粗糙度的显著性因素。利用田口法、多元回归方程 和指数方程对表面粗糙度进行预测,确定FDM 打印件最小表面粗糙度的参数组合。分析结果 表明：层厚对于表面粗糙度的影响程度最大,温度次之,打印速度最小;为了验证其有效性和 适用性,针对不同打印模型和FDM 打印机进行了验证性实验。实验结果表明：在预测模型方 面,多元回归方程的预测结果优于指数方程和田口法。并且,上述结论对不同打印模型和FDM 打印机具有较为宽泛的适用性。     

A Bayesian network model for surface roughness prediction in the machining process

The literature reports many scientific works on the use of artificial intelligence techniques such as neural networks or fuzzy logic to predict surface roughness. This article aims at introducing Bayesian network-based classifiers to predict surface roughness (Ra) in high-speed machining. These models are appropriate as prediction techniques because the non-linearity of the machining process demands robust and reliable algorithms to deal with all the invisible trends present when a work piece is machining. The experimental test obtained from a high-speed milling contouring process analysed the indicator of goodness using the Naïve Bayes and the Tree-Augmented Network algorithms. Up to 81.2% accuracy was achieved in the Ra classification results. Therefore, we envisage that Bayesian network-based classifiers may become a powerful and flexible tool in high-speed machining.     

Modelling of surface roughness in inclined milling operations with circle-segment end mills

There is currently a lack of knowledge in the manufacturing of high complexity aerospace components. Impellers or blade-integrated disks (blisks) are expensive, and manufacturers tend to prefer reliability over productivity. Thus, manufacturing times are higher than they should be. These challenging parts need to be machined using new advanced tools for several reasons, such as requirement of 1) special and complex tool paths, 2) smoother cutting forces, and 3) good accessibility. Circle-segment or oval-form cutters have recently demonstrated their usefulness and adaptability in the machining of profile and free-form surface operations, and are becoming a solution for a wide range of applications and materials. However, machinists who use them know very little about such tools. In fact, there has been a lack of real-world modelling applications. This paper proposes for the first time a geometrical model that allows the prediction of the surface topography in flank-milling operations using circle-segment end mills. This time-domain model includes the most important mechanical and kinematical parameters during cutting: the tool geometry, feed rate, radial immersion, and tool runout. Tool orientation angles commonly used in 5-axis operations are also included. The developed model was positively verified against experimentally measured values in a milled wall made of aluminium Al7075T. This knowledge-based tool is useful for manufacturing companies and suppliers interested in optimizing and controlling their production parameters.     

An intelligent approach for the prediction of surface roughness in ball-end machining of polypropylene

Manufacturing paradigms over the last 150 years have changed from craft production, to mass production and now to mass customisation. One further extension of mass customisation is personalised manufacture, which is the concept of providing bespoke products to the individual consumer. As a result this has brought about the need for a greater degree of sophistication in manufacturing practises and the technologies employed. This bespoke form of manufacture of consumer goods is now being pursued on CNC machining centres as opposed to the alternative of highly expensive rapid prototyping methods. The problem with this form of manufacture is that the products are generally free formed objects which require sophisticated setups and machining. Ball-end machining is a method used to create cusp-type geometry, which is employed on CNC machines to create sculptured surfaces. The objective of this research is to provide a predictive model using a design of experiments strategy to obtain optimised machining parameters for a specific surface roughness in ball-end machining of polypropylene. This paper reports on new manufacturing knowledge to machine polypropylene using ball-end tooling in order to generate personalised sculptured surface products.     

Simulation and optimization of machining parameters in drilling of titanium alloys

The present study attempts to analyse the effect of various drilling parameters such as spindle speed, feed rate and drill bit diameter on performance characteristics such as thrust force, torque and circularity at entry and exit of the holes in drilling of titanium alloy using coated drill bit. A three dimensional machining model based on Lagrangian approach is developed using DEFORM-3D software. The performance characteristics obtained through simulation model is compared with experimental results. The simulation model closely agree with the experimental results as percentage relative error of 4.93, 9.01, 6.04 and 3.0 is observed for thrust, torque circularity at entry and circularity at exit respectively. The experimental data is used to develop valid empirical models to relate performance characteristics with drilling parameters using non-linear regression analysis. The empirical model helps to predict various performance characteristics without resorting to rigorous analysis through the numerical model. The values of various performance characteristics predicted from empirical models are compared with experimental results and the percentage relative error within 10% is observed. Finally, an improved version of latest evolutionary approach known as Harmony Search (HS) algorithm has been proposed to obtain favorable machining conditions through optimization of each performance characteristic. The optimal value of circularity at entry is obtained as 0.985 (approaching towards ideal value of one) when spindle speed, feed rate and drill bit diameter are set at 530.86 (≈531) RPM, 44.8 (≈45) mm/min and 7 mm, respectively. Similarly, optimal value of circularity at exit reaches 0.979 with same spindle speed and drill bit diameter but feed rate of 50 mm/min.     

Artificial neural network models for the prediction of surface roughness in electrical discharge machining

In the present paper Artificial Neural Networks (ANNs) models are proposed for the prediction of surface roughness in Electrical Discharge Machining (EDM). For this purpose two well-known programs, namely Matlab^® with associated toolboxes, as well as Netlab^®, were emplo- yed. Training of the models was performed with data from an extensive series of EDM experiments on steel grades; the proposed models use the pulse current, the pulse duration, and the processed material as input parameters. The reported results indicate that the proposed ANNs models can satisfactorily predict the surface roughness in EDM. Moreover, they can be considered as valuable tools for the process planning for EDMachining.     

Simulation-based model for the optimization of machining parameters in a metal-cutting operation

To achieve a certain measurable performance in cutting machines, the machine parameters need to be optimized. Several constraints determine the possible values that these parameters can take. Although parameters are usually assumed to be deterministic, in practice, it is common to find variations on the characteristics of the products or the processes. Modeling machining parameters as stochastic factors provides a more realistic representation of cutting operations. Moreover, multiple operational objectives are of interest, in many real situations, these multiple objectives are conflicting. Consequently, the problem of setting the parameters becomes a trade-off situation. This paper presents a novel Simulation-based Multi-Objective Optimization (SimMOpt) solution procedure. The procedure is divided into two phases: (1) finding initial solutions and, (2) using a simulated annealing-based method for finding neighboring solutions. In the first phase, non-linear goal programming is used for finding high quality initial solutions. The second phase incorporates Pareto Archive Evolution Strategy (PAES) and hypotheses testing for searching near-optimal solutions for a set of stochastic parameters (i.e., cutting speed, feed rate, and depth of cut) in metal cutting operations. Three objectives are optimized (i.e., operation time, operation cost, and quality of the product). The results from implementing this procedure are analyzed and compared to a baseline methodology based on the Multi-Objective Simulated Annealing (MOSA) algorithm. The analysis demonstrates that the proposed method outperforms the Genetic Algorithm (GA), which was the benchmark algorithm, in terms of the solution quality of all the objectives. More importantly, the solutions from using the SimMOpt procedure outperform those obtained from using an enhanced MOSA-based approach (i.e., 4.71% improvement in the hypervolume approximation).     

A new hybrid differential evolution algorithm for the selection of optimal machining parameters in milling operations

This paper presents a novel hybrid optimization approach based on differential evolution algorithm and receptor editing property of immune system. The purpose of the present research is to develop a new optimization approach to solve optimization problems in the manufacturing industry. The proposed hybrid approach is applied to a case study for milling operations to show its effectiveness in machining operations. The results of the hybrid approach for the case study are compared with those of hybrid particle swarm algorithm, ant colony algorithm, immune algorithm, hybrid immune algorithm, genetic algorithm, feasible direction method and handbook recommendation.     

Prediction of surface roughness in ball-end milling process by utilizing dynamic cutting force ratio

The aim of this research is to propose the practical model to predict the in-process surface roughness during the ball-end milling process by utilizing the dynamic cutting force ratio. The proposed model is developed based on the experimentally obtained results by employing the exponential function with five factors of the spindle speed, the feed rate, the tool diameter, the depth of cut, and the dynamic cutting force ratio. The experimentally obtained results showed that the frequency of the dynamic cutting force corresponds with the frequency of the surface roughness profile in the frequency domain. Hence, the dimensionless dynamic cutting force ratio is proposed regardless of the cutting conditions to predict the in-process surface roughness by taking the ratio of the area of the dynamic cutting force in X axis to that in Z axis. The multiple regression analysis is adopted to calculate the regression coefficients at 95 % confident level. The experimentally obtained model has been verified by using the new cutting conditions. It is understood that the developed surface roughness model can be used to predict the in-process surface roughness with the high accuracy of 92.82 % for the average surface roughness and 91.54 % for the surface roughness.     

Prediction of surface roughness in CNC face milling using neural networks and Taguchi''s design of experiments

In this paper, a neural network modeling approach is presented for the prediction of surface roughness (R_a) in CNC face milling. The data used for the training and checking of the networks’ performance derived from experiments conducted on a CNC milling machine according to the principles of Taguchi design of experiments (DoE) method. The factors considered in the experiment were the depth of cut, the feed rate per tooth, the cutting speed, the engagement and wear of the cutting tool, the use of cutting fluid and the three components of the cutting force. Using feedforward artificial neural networks (ANNs) trained with the Levenberg–Marquardt algorithm, the most influential of the factors were determined, again using DoE principles, and a 5×3×1 ANN based on them was able to predict the surface roughness with a mean squared error equal to 1.86% and to be consistent throughout the entire range of values.     

Analysis and optimization of surface roughness in the ball burnishing process using response surface methodology and desirabilty function

In the present study, an optimization strategy based on desirability function approach (DFA) together with response surface methodology (RSM) has been used to optimize ball burnishing process of 7178 aluminium alloy. A quadratic regression model was developed to predict surface roughness using RSM with rotatable central composite design (CCD). In the development of predictive models, burnishing force, number of passes, feed rate and burnishing speed were considered as model variables. The results indicated that burnishing force and number of passes were the significant factors on the surface roughness. The predicted surface roughness values and the subsequent verification experiments under the optimal conditions were confirmed the validity of the predicted model. The absolute average error between the experimental and predicted values at the optimal combination of parameter settings for surface roughness was calculated as 2.82%.     

Using kernel data in machine tools for the indirect evaluation of surface roughness in vertical milling operations

The goal of this research is to compare the capabilities of kernel data and external sensor data, captured with piezoelectric accelerometers, for the indirect evaluation of surface roughness in vertical milling operations. Experiments were conducted to obtain data for developing algorithmic models that will be utilized to predict surface roughness. Seventy-two samples were used to develop two neural networks; one based on accelerometer inputs and the other on kernel inputs, and to compare the performance of the data source when calculating the average surface roughness parameter (Ra). Results show that accelerometer data and numerical control kernel (NCK) data can be useful for the indirect evaluation of average surface roughness as shown by a high correlation between outputs and targets. The main conclusion of this work is that when evaluating the average surface roughness parameter, it is more interesting to use the data obtained directly from the NCK than from external accelerometers.     

Multi-objective optimization in wire-electro-discharge machining of TiC reinforced composite through Neuro-Genetic technique

This paper proposed a Neuro-Genetic technique to optimize the multi-response of wire electro-discharge machining (WEDM) process. The technique was developed through hybridization of a radial basis function network (RBFN) and non-dominated sorting genetic algorithm (NSGA-II). The machining was done on 5 vol% titanium carbide (TiC) reinforced austenitic manganese steel metal matrix composite (MMC). The proposed Neuro-Genetic technique was found to be potential in finding several optimal input machining conditions which can satisfy wide requirements of a process engineer and help in efficient utilization of WEDM in industry.     

Investigation of surface roughness effects on fluid flow in passive micromixer

Surface roughness effects are dominant at microscale. In this study, microchannels are fabricated on Silicon substrate. The roughness morphology is modeled for the fabricated structure using Weierstrass-Mandelbrot function for self-similar fractals. A two dimensional model of hexagonal passive micromixer is analyzed with surface roughness present on inner walls of channels using parallel Lattice Boltzmann method, implemented on sixteen node cluster. The results are compared by simulating this micromixer structure using Navier–Stokes equations. The experimental results on the fabricated micromixers are also presented. The effects of relative roughness, fractal dimension and Reynolds number are discussed on laminar flow in hexagonal passive micromixers. The study concludes the importance of modeling surface roughness effect for better mixing efficiency.     

Diffuse response surface model based on moving Latin hypercube patterns for reliability-based design optimization of ultrahigh strength steel NC milling parameters

We focus here on a Response Surface Methodology adapted to the Reliability-Based Design Optimization (RBDO). The Diffuse Approximation, a version of the Moving Least Squares (MLS) approximation, based on a progressive sampling pattern is used within a variant of the First Order Reliability Method (FORM). The proposed method uses simultaneously the points in the standard normal space (U-space) and the physical space (X-space). The two grids form a ??virtual design of experiments?? defined by two sets of points in both design spaces, which are evaluated only when needed in order to minimize the number of the ??exact?? thus supposed costly, function evaluations. At each new iteration, the pattern of points is updated with the points appropriately selected from the virtual design, in order to perform the approximation. As an original contribution, we introduce the concept of ??Advancing LHS?? which extends the idea of Latin Hypercube Sampling (LHS) for the maximal reuse of already computed points while adding at each step a minimal number of new neighboring points, necessary for the approximation in the vicinity of the current design. We propose panning, expanding and shrinking Latin patterns of sampling points and we analyze the influence of this specific kind of patterns on the quality of the approximation. Then we analyze the minimal number of data points required in order to get well-conditioned approximation systems. In the application part of this work, we investigate the case of optimizing the process parameters of numerically controlled (NC) milling of ultrahigh strength steel.     

On the evolution of welding residual stress after milling and cutting machining

Residual stresses develop during most manufacturing processes involving material deformation, heat treatment, machining or processing operations that transform the shape or change the properties of a material. They have a not negligible effect on the material strength, especially on fatigue. For this reason, it is important that some knowledge of the internal stress state can be deduced either from measurements or from modelling predictions. The object of this paper is forecasting the modification and the evolution that a residual stress field, originated by welding, suffers after chip-forming machining, such as milling and cutting. Numerical results have been critically compared to experimental measurements and show the potentiality but also the limitations of numerical techniques.     

Adaptive network-based fuzzy inference system for prediction of surface roughness in end milling process using hybrid Taguchi-genetic learning algorithm

In this paper, an adaptive network-based fuzzy inference system (ANFIS) with the genetic learning algorithm is used to predict the workpiece surface roughness for the end milling process. The hybrid Taguchi-genetic learning algorithm (HTGLA) is applied in the ANFIS to determine the most suitable membership functions and to simultaneously find the optimal premise and consequent parameters by directly minimizing the root-mean-squared-error performance criterion. Experimental results show that the HTGLA-based ANFIS approach outperforms the ANFIS methods given in the Matlab toolbox and reported recently in the literature in terms of prediction accuracy.