Elsevier

Measurement

Volume 82, March 2016, Pages 55-63
Measurement

Investigation on the effect of cutting fluid pressure on surface quality measurement in high speed thread milling of brass alloy (C3600) and aluminium alloy (5083)

https://doi.org/10.1016/j.measurement.2015.12.016Get rights and content

Abstract

The quality of a machined finish plays a major role in the performance of milling operations, good surface quality can significantly improve fatigue strength, corrosion resistance, or creep behaviour as well as surface friction. In this study, the effect of cutting parameters and cutting fluid pressure on the quality measurement of the surface of the crest for threads milled during high speed milling operations has been scrutinised. Cutting fluid pressure, feed rate and spindle speed were the input parameters whilst minimising surface roughness on the crest of the thread was the target. The experimental study was designed using the Taguchi L32 array. Analysing and modelling the effective parameters were carried out using both a multi-layer perceptron (MLP) and radial basis function (RBF) artificial neural networks (ANNs). These were shown to be highly adept for such tasks. In this paper, the analysis of surface roughness at the crest of the thread in high speed thread milling using a high accuracy optical profile-meter is an original contribution to the literature. The experimental results demonstrated that the surface quality in the crest of the thread was improved by increasing cutting speed, feed rate ranging 0.41–0.45 m/min and cutting fluid pressure ranging 2–3.5 bars. These outcomes characterised the ANN as a promising application for surface profile modelling in precision machining.

Introduction

High speed machining (HSM) is one of the most promising material removal processes and has been receiving growing attention in recent years. HSM is a popular manufacturing technique, due to its faster speed and reduced cost, as well as the ability to achieve very high quality surface finishes [1]. HSM operations can change the chemical composition and mechanical properties of materials being machined, causing complex wear mechanisms which have attracted the attention of a number of researchers.

Although different aspects of HSM such as chatter, force, vibration, tool damage, tool wear and microstructural damage [2], [3], [4], [5], [6] have been explored for improving the machinability of materials, surface roughness as a function of operation quality is currently receiving considerable attention [7], [8], [9], [10].

HSM on Cr12MoV steel using a two edged cemented carbide ball end milling cutter demonstrated that edge and machining contact length increased as a function of increasing helix angle. Hence, smaller helix angles and negative rake angles improved the quality of the surface manufactured as well as longer tool life [11]. HSM studies on Ti−10V−2Fe−3Al illustrated that optimisation of cutting parameters lead to a minimum surface roughness of (Ra < 0.8 μm) and fatigue cracks concentrating at the intersecting edges of machined surfaces. The study also showed that the strongest factors affecting surface roughness were feed per tooth, cutting width, and cutting speed respectively. This is important since the development of fatigue cracks are highly dependent on the quality of the measured surface [12]. HSM investigations on the surface roughness of Ti–6Al–4V using polycrystalline diamond (PCD) tools verified that, at high temperatures, adhesion between the workpiece and tool caused tool wear which lead to increased surface roughness. Increasing cutting speed and decreasing feed rate produced a high quality surface, cutting force and decreased tool wear [13]. Honghua et al. [14] studied roughness, defects, micro-hardness and microstructure of the machined surface using PCD and polycrystalline cubic boron nitride (PCBN) tools when HSM of Ti–6.5Al–2Zr–1Mo–1V (TA15) alloy. Results of this study show that PCD tools have a greater longevity compared to PCBN, with a higher quality surface (Ra < 0.3 μm) and changes in microhardness is observed.

Aluminium is one of the so called “light metals” that is widely used in aerospace, automotive, electronics and biomedical applications. A number of HSM studies of this alloy, such as analysing force and cutting conditions, have been reported in the literature [15], [16]. Analysis of HSM on metal matrix composites, such as particle reinforced aluminium [17], indicates that cutting speed and feed rate had the highest influence on the surface quality of this process and tool flank-wear was generated. Optical imaging of the surface illustrated material build up at edges, highly influencing tool wear and surface roughness. The effect of spindle speed, feed-rate and machining time on the surface roughness of aluminium Alloy 1100 shows the feed-rate and machining time contribute significantly to the production of surface roughness. A lower feed-rate produced a superior surface finish, but it also produced a continuous chip and would be inefficient in terms of the time taken for the operation. Continuous chips can enwrap/twist/weld around the tool decreasing chip clearance and preventing cutting edges, cooling, and subsequently causing material build up at the cutting edge of the tool or workpiece leading to surface defects [18]. Brass and its alloys are widely employed in commercialised systems because of their mixture of properties such as non-magnetic nature, high corrosion resistance, and good machinability. To increase the machinability of brass, adding lead and silicon is suggested. In machining of brass, as with other metals, high cutting speed and low feed rate improves surface roughness and chip breakage [19], [20], [21], [22].

External thread milling and whirlwind operations involve complicated machining approaches due to their elaborate tool geometry, effects of this tridimensional tool trajectory have not yet been fully understood [23], [24], [25]. Different aspects of thread milling such as cutting force, cutting parameters, various cutting lubrications, cutting tool angle, wear and interference have been discussed by researchers using mathematical and numerical models [26], [27], [28]. However, there is still a lack of knowledge on the surface quality of threads due to difficulties in the measurement of these surfaces. Knowledge-based systems are highly useful for modelling, simulation, prediction and optimisation of industrial processes [29]. In HSM notably, milling and turning, the application of artificial intelligence for surface roughness modelling are common and have been the subject for some investigations [30], [31], [32], [33], [34], [35], [36], [37].

A number of studies on the surface quality, cutting condition, cutting force, lubrication and tools in HSM of Al and brass alloys are found in the literature. However, thus far, there are very few reports detailing the effect of cutting parameters on the quality of the surface of the threads. In this paper the effect of cutting fluid pressure on the predicted surface roughness of HSM thread operations for brass C3600 and Al 5083 have been outlined. In order to perform the necessary experiments a Taguchi L32 design of experiment (DOE) together with MLP and RBF ANNs have been used.

Section snippets

Taguchi approach

The Taguchi approach requires less experimental operations to obtain accurate results compared with other approaches Furthermore, processing of these outcomes can be carried out to improve the accuracy.

Cutting conditions were adapted from those recommended by the machine tool manufacturer (Datron). In this experiment, the Taguchi L32 orthogonal array DOE is used to decrease the number of experiments required. Based on this DOE two factors (cutting fluid pressure and cutting speed) were chosen

Selection of process parameters

In HSM the modality of operation, cutting conditions and the geometry of cutting tools allow operators to choose suitable cutting conditions. In thread milling, effective parameters that measure the quality of produced materials are tool material and geometry, cutting speed, feed rate, cutting fluid and workpiece material. The outlined factors are highly dependent on the machine tool and the nature of the operation.

Workpiece material cutting tools and machine tool specifications

In this study HSM thread milling operations were investigated in brass C3600

MLP and RBF neural networks modelling

To determine the capability of the respective ANN models, cutting fluid pressure, spindle speed and feed rate were chosen as inputs and roughness on the crest of thread was used as a target. In MLP the best structure for designing the network was obtained 3 × 2 × 1, a hyperbolic secant was chosen as the transfer function. Fig. 3 illustrates the correlation of the network output and targets for both training and test processes. The correlation of output and targets for train and test were obtained

Conclusions

In this paper the effects of cutting fluid pressure at the surface roughness, from HSM thread milling of brass C3600 and Al 5083 by using ANNs has been explored. Two different ANNs containing MLP and RBF were investigated and their ability to model the mentioned operation compared. The results allowed characterising MLP ANNs as a high accuracy application for modelling and predicting the nano-surface roughness on the crest of the thread.

Results are summarized as below:

  • MLP neural networks have

References (37)

Cited by (26)

  • Fatigue fracture surface of bronze RG7 under proportional cyclic torsion and bending

    2023, Measurement: Journal of the International Measurement Confederation
  • Modelling of laser powder bed fusion process and analysing the effective parameters on surface characteristics of Ti-6Al-4V

    2020, International Journal of Mechanical Sciences
    Citation Excerpt :

    This tool can draw logical relationships between inputs and outputs in multi-dimensional space and are used to predict, model and analyse the contribution of each input to the obtained results. Moreover, this tool can show the interaction of different parameters on the results and explain the linear and non-linear behavior of input parameters on outputs [47–49]. ANNs are based on three stages; train, test and recall/validation.

  • Variants of high speed machining

    2020, High-Speed Machining
  • Multi-hole simultaneous drilling of aluminium alloy: A preliminary study and evaluation against one-shot drilling process

    2020, Journal of Materials Research and Technology
    Citation Excerpt :

    Their results also showed that dry machining of Al5083 performs better at high Vc and produces lower cutting forces than in wet cutting. Khorasani et al. [24] have used different machining parameters and coolant pressure to minimize Ra of Al5083 using a high speed milling operation. They concluded that the best surface quality is possible when the cutting fluid ranges between 2.5–3.5 bars and f between 0.41–0.45 m/min.

  • Experimental investigations and empirical modeling for optimization of surface roughness and machining time parameters in micro end milling using Genetic Algorithm

    2018, Measurement: Journal of the International Measurement Confederation
    Citation Excerpt :

    Surface quality plays a significant role in the concert of micro and miniature parts. It improves corrosion resistance, fatigue strength, creep behavior, surface friction and minimizes stress concentration, Khorasani et al. [4]. Micro milling has the capability to fabricate a profile with desirable accuracy and precision surface finish for applications include micro dies, molds, channels, gears, propellers, fluidic devices, etc.

View all citing articles on Scopus
View full text