Empirical Study of the Power Efficiency of Various Machining Processes

Available online at www.sciencedirect.com

ScienceDirect Procedia CIRP 14 (2014) 558 – 563

6th CIRP International Conference on High Performance Cutting, HPC2014

Empirical study of the power efficiency of various machining processes Hae-Sung Yoona, Jang-Yeob Leea, Min-Soo Kima, Eun-Seob Kima, Sung-Hoon Ahna,b,* a

School of Mechanical and Aerospace Engineering, Seoul National University, Seoul, 151-744, Korea b Institute of Advanced Machinery and Design, Seoul National University, Seoul, 151-744, Korea

* Corresponding author. Tel.: +82-2-880-7110; fax: +82-2-888-9073. E-mail address: [email protected]

Abstract Machining is one of the most widely used subtractive processes, and many studies have reported methods to improve the quality, productivity, and effectiveness of the process. Moreover, in recent years, energy efficiency has become an increasingly important consideration, largely as a result of new legislation and standardization in response to environmental concerns. However, the analysis of the energy efficiency of machine tools is not straightforward because of the complexity of the components and process. This study compared the power efficiency of various machining processes at different scales. Here, power efficiency is defined as the ratio of the process power to the total power consumption, and it was calculated using experimental results from conventional milling, micro-scale drilling, and brushing. The calculated power efficiency is compared for the processes reported here, as well as with selected published data. We found that the power efficiency varied regardless of machining scales or specific energy consumed, and also can vary widely in terms of the peripheral devices used. Moreover, for the case of laser-assistant machining, the present power efficiency metric decreased as the cutting load decreased. Therefore, it is need to consider the effect of the surrounding environment and effectiveness of machining, and suggestions were shown for the concept of novel power efficiency.

© 2014 2014The Published byPublished Elsevier by B.V. Open access under CC BY-NC-ND license. © Authors. Elsevier B.V. Committee theCIRP 6th CIRP International Conference Selection and peer-review under responsibility of the International Scientific Selection and peer-review under responsibility of the International Scientific Committee of theof6th International Conference on High on High Performance Performance Cutting. Cutting Keywords:Energy saving; Machining process; Power efficiency; Process scale

1. Introduction Mechanical machining is one of the most commonly used subtractive processes, in which a volume of material is removed using tool tip edges. Machining can deliver excellent geometric accuracy with good economic efficiency [1] and methods to improve the quality, productivity, and effectiveness of the process have been reported. In recent years, energy efficiency has been a growing concern, and directives and legislation have recently appeared in response to environmental concerns. The European Commission adopted the Energy Efficiency Directive 2012/27/EU [2] to realize energy efficiency targets, and the Cooperative Effort on Process Emissions in Manufacturing (CO2PE!) [3] was initiated to analyze the environmental impact of manufacturing processes. Manufacturing industries comprise a significant portion of the total global energy consumption, and potential for an 18–

26% reduction in the total energy consumption has been identified [4]. The energy consumption in manufacturing processes should be monitored carefully, and the efficiency of each component should be standardized to improve decision making. Duflou et al. proposed a systematic approach for energy saving at a system level [5]. They classified manufacturing systems into levels from the single device/unit level to the global supply chain level. Yoon et al. reported an energy saving strategy in terms of improved activity level, which could be implemented on single device level [6]. However, in order to implement energy saving technologies, it is necessary to monitor and model the energy consumption of each process. For machining, the energy consumption of machine tools has been classified in terms of the operating states of the machine [7] or components [8], with each component modeled individually. As an extension, the power consumption and efficiency should be defined and standardized. Since different types of

2212-8271 © 2014 Published by Elsevier B.V. Open access under CC BY-NC-ND license. Selection and peer-review under responsibility of the International Scientific Committee of the 6th CIRP International Conference on High Performance Cutting doi:10.1016/j.procir.2014.03.043

559

Hae-Sung Yoon et al. / Procedia CIRP 14 (2014) 558 – 563

machine tool at different scales consume different amounts of power, the efficiency of the process should be considered for effective optimization. Many reports define power efficiency as the ratio of the necessary process power to the total power consumption [9]. However, it is not straightforward to standardize machine tool efficiency using this definition, because the proportion of the process power can vary with the scale of the machining process, type of machine tool, and even the type of auxiliary devices used. This study provides an overview of the power efficiency of machining processes at various scales, and compares the efficiency with other processes and reported data. Power efficiency was calculated using experimental results from conventional milling, micro-scale drilling, and brushing. It was also calculated based on reports on conventional milling processes and laser-assisted turning. The main goal of this work is to provide an overview of the variation in the power efficiency at different scales of machining, allowing a direct comparison, and to file up the characteristics of the power efficiency at various machining processes. Due to the complexity of machine tools, power efficiency should be defined carefully, considering the characteristics and performance of each part of the process.

2. Literature reviews The power consumption of machine processes needs to be decomposed into that of individual elements and analyzed. Balogun and Mativenga [7] divided the machine states into basic, ready, and cutting states, and constructed a mathematical model using the power element and related period of each state, i.e., Et

Pb (t b  t r  t c )  Pr t r  Pair t air  ( Pr  Pcool  kv)t c (1)

where E t is the total energy consumption of the machine tool, and Pb , Pr , Pair , and Pcool are the power consumed at the basic, ready, air-cutting, and cooling states, respectively. The terms t b , t r , t air , and t c are the time consumed at the basic, ready, air-cutting, and cutting states, respectively; k is the specific cutting energy (in units of kJ/cm3), and v is the material removal rate (in cm3/s). Here, k and v are cuttingrelated components and represent the energy consumed in material removal. Yoon et al. [8] constructed the following energy model in terms of machine tool components: Et

E BASIC  E STAGE  E SPINDLE  E MACHINING

(3)

where E BASIC is the basic energy consumption element in the stand-by state, i.e., the constant energy used in stand-by mode, E STAGE and E SPINDLE are the energy consumed by stage and spindle, respectively, i.e., the momentum energy used for the process, and E MACHINING is the energy consumed in material removal. i.e., the cutting load during the process. Similarly, many authors have attempted to decompose the energy consumption elements. They divided the energy

consumption into in the basic setting-up or stand-by energy and the material removal energy. Machine tools consume more energy during cutting due to the cutting force. From the literature [9], the power efficiency of the cutting process, K , is simply defined as the material removal power as a portion of the total power consumption, i.e.,

K

(4)

PMACHINING / PTOTAL

For the energy efficiency, individual state or task period needs to be involved. However, they could varied with respect to the details of the process, hence the power efficiency, which is the representative efficiency values during the process, is considered in this study. Regarding the efficiency of process Gotowski et al. [10] analyzed the process with thermodynamic framework, using a concept of exergy with entropy and entalphy of material. However the present manuscript has a focus on effectiveness from the perspective of a machine tool, rather than a process itself. Hence a simple definition of power efficiency was adopted as mentioned above, to compare power utilization ratio of different machines at various scales. He et al. [11] constructed an energy estimation model based on numerical control code, and analyzed the power distribution of milling operations. Wang et al. [12] compared the process efficiency of various task levels, from machine tools to the workshop. Draganescu et al. [13] calculated the machine tool efficiency of various milling conditions, and reported that a maximum efficiency could be determined based on a response surface in terms of process parameters. In equation (4) PMACHINING can be determined from the cutting force, torque, feed, and rotational speed; however, many additional factors contribute to PTOTAL . Figure 1 shows the power distribution of a machine tool when performing aircutting [7]. The core auxiliaries accounted for over 60% of the total power consumption, and the spindle and stage each consumed 10%. Assuming that PBASIC does not change, the efficiency is determined by the absolute amount of cutting power, and the total power consumption can be larger at higher efficiencies. This study reviews the proportion PMACHINING comprises in various machining process and considers the effectiveness of various processes. Jog z-axis 5%

Jog x-axis Jog y-axis 4% 4%

Startup 1%

Home location 5% Tool clamp 7%

Spindle 8%

Core Auxiliaries 57%

Synchro magazine 9% Fig. 1. Power distribution of a machine tool performing air-cutting (redrawn from Balogun and Mativenga [7] with permission from Elsevier)

560

Hae-Sung Yoon et al. / Procedia CIRP 14 (2014) 558 – 563

3. Power efficiency of various machining processes We compared the power efficiency of conventional milling, micro-scale drilling, and brushing processes. Typically, conventional milling is carried out using an end-mill with multi-tool edges, and material is removed to form the desired geometry. Micro-scale drilling is performed using micro-bits with typical diameters of several hundred micrometers, and the material removal rates are significantly lower than with conventional milling. Brushing uses a cup brush with many wires to sweep the workpiece to achieve the desired surface requirements, and the material removal is almost negligible. These three processes involve material removal rates on different scales; hence, the power consumption characteristics were compared. 3.1. Conventional milling with a 3-mm-diameter tool The power consumption of a three-axis machining center (F400, Hyundai WIA Corp. ltd., Korea) was measured using a power meter (PAC3200, Siemens, Germany). An end-mill tool with diameter of 3 mm (R216.24-03050ACC05P, Sandvik Coromant, Sweden) was used, and C45 steel was used as the workpiece. The machining parameters were as follows: rotation speed 12,000 rpm, reed rate 12.9 mm/s, and axial depth of cut 1 mm. Machining was performed using a fullwidth 3-mm cut. The power consumption of the machining center was decomposed using Equation (3) by measuring each element individually. PBASIC was 2,132 W, including 602 W for fluid pumping, and PMACHINING was calculated from the power profile during the cutting process. Figure 2 shows the power profile measured while cutting. The overall power consumption of the machine tool increased slightly while cutting, due to the cutting force and torque. Only constant power regions were considered as the cutting state, and the regions of the curve where the power consumption changed were neglected for simplicity. PMACHINING was defined as the difference between the power in the cutting and non-cutting states. 3000

PMACHINING, 7.6% PSTAGE, 1%

PSPINDLE, 22.6%

PBASIC, 68.9%

Fig. 3. Power distribution of the conventional milling process.

Figure 3 shows the power distribution of the conventional milling process. PBASIC was approximately 69% of the total power consumption, and the power for material removal, PMACHINING , occupied for only 7.6% of the total. Hence, the power efficiency of the conventional milling process was 7.6% in this case. 3.2. Micro drilling For micro-scale drilling, PCB (Printed Circuit Board) drilling process was targeted. As a preliminary research, power consumption model and manufacturing cost model were constructed with respect to various parameters [8]. For micro-scale drilling, a printed circuit board (PCB) drilling process was examined. As a preliminary study, an energy consumption and manufacturing cost model was constructed [8]. A three-axis stage (Justek, Korea) with a high-speed spindle (D1733, Westwind Air Bearings, UK) was used to construct a machining system, and a power meter (SKPM200, Seon Kwang System, Korea) was used to measure the power consumption of the system. A micro-drill with diameter of 400 μm was used to drill three stacks of epoxy-resin copper-clad laminate (CCL) board with aluminum entry and wood backing boards. The machining parameters were as follows: rotation speed 90,000 rpm, feed rate 35 mm/s, and drilling depth 6 mm. The power consumption was modeled using Equation (3), and the machining power was defined as the mean power during drilling. In this case, the change of power consumption due to drilling was negligible.

Power consumption (W)

2800

PMACHINING, 1%

PMACHINING, Power for material removal 2600

PSTAGE, 24%

Air-cutting power = PBASIC + PSPINDLE + PSTAGE

2400

PBASIC, 60% PSPINDLE, 15%

2200

2000

5

6

7

8

9

10

11

12

13

T(s) Fig. 2. Power profile of a conventional milling machine tool during cutting.

Fig. 4. Power distribution of the micro-scale drilling process (redrawn from Yoon et al. [8] with permission from Elsevier)

561

Hae-Sung Yoon et al. / Procedia CIRP 14 (2014) 558 – 563

Figure 4 shows the power distribution of the micro-scale drilling process [8]. Some results were published as the preliminary study, and PMACHINING was negligible in microscale machining. 3.3. Brushing PCB manufacturing also involves brushing processes, primarily for de-burring after drilling. In the experiment, a brushing system was constructed. A hand grinder (GWS7100E, BOSCH, Germany) was attached to the stage used for drilling, and brushing was carried out. A stainless cup brush (Inwoo Tech, Korea) with a diameter of 10 cm was used. The cup brush consists of many steel wires used to remove microscale burrs from the surface. The workpiece material was the same as in Section 3.2; however, the entry and backing boards were not used. The machining parameters were as follows: rotation speed 5,400 rpm and feed rate 50 mm/s. Contact between the brush and work material was measured using visual observation of the surface of the workpiece, and the brush was positioned with depth of cut of 0.7 mm. Since the brush wire is elastic, this ‘virtual’ depth of cut affects the cutting force and hence the material removal power. Figure 5 shows the power profile during the brushing process using the hand grinder. Since the round hand grinder passed over rectangular work material, the power increased linearly and then decreased with the motion of the grinder. 600

Power consumption (W)

500

PMACHINING, Power for material removal

400

Air-cutting power = PBASIC + PGRINDER + PSTAGE

300 200

100 0

1

3

4

5 6 Time (sec)

7

8

9

Fig. 5. Power profile of brushing process with hand grinder. Here, the term PGRINDER was used rather than of PSPINDLE.

PMACHINING, 23.8%

The maximum of the power curve was considered to be the machining power and the difference between the brushing power and power when not machining was considered in the calculation of the power used for material removal. Figure 6 shows the power distribution of the brushing process. This process had a very low material removal rate; however, PMACHINING accounted for 24% of the total power consumption. Since PMACHINING in conventional machining consists of the material shear power and friction power, it follows that this process involves significantly more friction than the material shear power. 3.4. Power efficiency at different scales In addition to these experimental results, more cases were analyzed through preceding researches. In this section power efficiency of various scales of machining processes were compared with each other. Researchers performed power modelling research on conventional milling or turning processes in general. In addition to the experimental results reported here, we analyzed previous reports on machining power consumption. This section compares the power efficiency of machining processes on various scales, including simulation studies of energy consumption in conventional milling or turning processes. He et al. [11] investigated the power consumption of a vertical milling center (PL700, Chengdu Precise CNC Machine Tool, China). The machining conditions were as follows: rotation speed 2,000 rpm, feed rate 25 mm/s, axial depth of cut 0.2 mm, and width of cut 10 mm. The workpiece was C45 steel. The cutting power was 100 W, accounting for 7–8% of the total power consumption (excluding the z-axis feed motor). The authors also analyzed the power consumption while machining a sample workpiece, and found that the cutting power was 7% of the total power consumption. Li et al. [13] analyzed the power consumption of a machining center (BMC-20LR, Hurco CNC, USA). The machining conditions were as follows: rotation speed 2,500 rpm, feed rate 4.16 mm/s, depth of cut 0.3 mm, and width of cut 15 mm. The diameter of the tool was 24 mm, and the workpiece material was C45 steel. PMACHINING, 12% PSTAGE, 1%

PBASIC, 32.6%

PSPINDLE, 41% PBASIC, 46%

PGRINDER, 5% PSPINDLE, 38.6%

Fig. 6. Power distribution of the brushing process.

Fig. 7. Power distribution of milling process with a width of cut is 15 mm (data from Li et al. [13] with permission from Elsevier)

Hae-Sung Yoon et al. / Procedia CIRP 14 (2014) 558 – 563

In their study, the cutting power accounted for 111 W, or approximately 12% of the total power consumption, as shown in Figure 7. The authors constructed a power consumption prediction model, and performed machining with various process parameters. The portion of the cutting power changed from 9% to 18% of the total when the material removal rate changed from 18.75 to 30 mm3/s. Draganescu et al. [9] analyzed the power consumption of an FV-32 machine tool. The machining parameters were as follows: rotation speed 750 rpm, feed rate 32 mm/s, axial depth of cut 6.5 mm, and width of cut 65.6 mm. The workpiece material was aluminum alloy ATSI10Mg, and the diameter of the tool was 250 mm. The calculated cutting power accounted for 86.4% of the total power consumption.

Energy efficiency (%)

562

20 18 16 14 12 10 8 6 4 2 0 0

10

20

30

40

Material removal rate (mm3/s)

Fig. 9. Power efficiency in terms of the material removal rate (data from Li et al. [13]) Table 1. Comparison of the energy consumption of conventional turning and laser-assisted turning (data from Zhao et al. [15] with permission from Elsevier) Process

Electricity consumption (kWh)

Time (min)

Specific energy consumption (J/mm3)

Power efficiency as used here represents the proportion of the effective power used for material removal. Figure 8 shows the proportion of process power to the total power consumption in various machining processes mentioned in this research. In conventional milling, processes with large material removal rates typically have higher power efficiencies, as shown in Figure 9. The data in Figure 9 were compiled from several milling results performed with various process parameters [13]. This trend can be explained by considering that higher material removal rates correspond to a larger cutting force and hence a higher material removal power. These results also agree with the trend of the electricity requirements in various manufacturing processes. Gutowski et al. [14] analyzed electricity requirements in various manufacturing processes, and revealed that the electricity requirements for the processing unit mass of material have a negative correlation with the process rate in log scale. This trend is valid in individual process; higher process rate condition deliver the higher proportion of process power, and the the lower electricity requirements. Nevertheless, the relationship between PMACHINING and PTOTAL should be considered carefully. For the brushing process, PMACHINING comprised 23.8%; however, most of the machining power was due to friction between the brush wires and workpiece. In the micro-scale drilling process, PMACHINING comprised only 1%; however material was removed more effectively than with the brushing process.

Conventional turning

0.21 kWh for 315,243 mm3 of material removed

15.5

2.50 J/mm3

Laserassisted turning

0.242 kWh 528,000 mm3 of material removed

10.8

1.65 J/mm3

Proportion of cutting power (%)

4. Discussion

25 20

15 10 5 0 Brushing Micro-scale Milling a drilling

Milling b [11]

Fig. 8. Proportion of various machining processes

Milling c [13]

Intuitively, this efficiency should not be applied to the power optimization of the process, because higher efficiency does not necessarily represent the minimum power consumption. Assuming that PBASIC is kept to constant in a single machine tool operation, a higher PMACHINING is required for greater power efficiency. In advanced machining techniques, the proportion of PMACHINING is lower. For example, laser-assisted machining increases the temperature of the area of the workpiece to be processed via laser irradiation, which helps to reduce the stress of the workpiece material and to improve the tool life and process rate. However, PBASIC increases and PMACHINING decreases. Zhao et al. [15] compared the efficiency of laser assisted turning and conventional turning, originally studied by Skvarenina and Shin [16]. They used a 1.5-kW CO2 laser and a turret lathe, and performed laser-assisted turning. Table 1 shows the results of the comparison [15]; with laser assistance, the turning consumed more power and hence more energy; however, from the perspective of the specific energy consumption in Joules per unit volume, the laser-assisted turning was more effective than the conventional turning process. The definition of power efficiency must be considered carefully and require re-definition to reflect the effectiveness of cutting. The definition given above does not reflect the effective material removal rate of the process, or the effect of the cutting force. In the case of assisted machining, the power efficiency becomes even worse, although the specific energy consumption actually decreased. The specific energy consumption itself might be considered as the standard; however, the performance of the process must be considered simultaneously. Helu et al. [17] reported that rough cut turning of titanium alloy consumed 10–30 kJ/cm3, while finish cutting consumed 60–700 kJ/cm3.

Hae-Sung Yoon et al. / Procedia CIRP 14 (2014) 558 – 563

Therefore, the rough cutting condition is much more energy efficient; however, finish cutting cannot be omitted from the overall process. Moreover, the proportion of basic power consumption should be considered in order to evaluate the effectiveness of the machine tool hardware components. Therefore, a novel and standardized definition of power efficiency is necessary for the machine tools. The present definition of the power efficiency could be a possible metric in individual machine, however it could not deliver a valid information such as material removal effectiveness, or the effect of assistance. Another plausible metric is to distinguish the hardware efficiency and the process efficiency. The hardware efficiency refers the power conversion efficiency from electrical to mechanical power, and the process efficiency refers the effectiveness of the material removal. The process efficiency could be defined as the function of cutting geometry, cutting load, and material embodied energy. For the effective process planning of machining process, standardized power efficiency will be essential information with the specific energy consumption values.

5. Conclusion As the importance of energy labeling and standardization has increased, a standard for power efficiency should be constructed. We discussed the power efficiency of various machining processes, which is defined as the ratio of material removal power to the overall power consumption. Conventional milling with a diameter of several mm, micro-scale drilling with diameter of several hundred micrometers, and brushing processes were considered, as well as published data. The power efficiency of conventional milling was approximately 7.6%, while those of micro-scale drilling and brushing were 1% and 23.8%, respectively. However, from the perspective of the amount of material removed, the conventional milling process was the most efficient. The definition of power efficiency did not reflect the effect of these processes well, as higher power efficiency simply reflects a higher material removal power, which, in the case of the brushing process, was dissipated via friction. Moreover, with assisted machining, the power efficiency of the process can even decrease due to the power consumption of the assisting devices. In this case, the reduced cutting force that results from the assistance leads to reduced power efficiency. Proper consideration of the process is required for effective representation of the power efficiency of a given machine tool process. The specific energy consumption has been considered as an energy indicator; however, the actual effectiveness of the basic power consumption should be considered, as well as the overall process performance.

Acknowledgements This work was supported by the Brain Korea 21 plus project at Seoul National University, Korea Institute for

563

Advancement of Technology (KIAT) funded by the Ministry of Knowledge Economy (No. 2010-TD-700203-001), the National Research Foundation of Korea (NRF) grant funded by the Ministry of Education, Science and Technology (No. NRF-2010-0029227), and SNU-Hyundai NGV cooperative research projects funded by Hyundai WIA corporation.

References [1] Yoon H-S, Wu R, Lee T-M, Ahn S-H. Geometric optimization of micro drills using Taguchi methods and response surface methodology. International Journal of Precision Engineering and Manufacturing 2011;12(5):871-5. [2] European Commission. Energy Efficiency Directive [homepage on the Internet]. c2012 [Updated 2012 October; cited 2013 Nov 19] Available from http://ec.europa.eu/energy/efficiency/eed/eed_en.htm [3] College International Pour la Recherche en Productique (CIRP). Cooperative Effort on Process Emissions in Manufacturing [homepage on the Internet]. c2009 [cited 2013 Nov 19] Available from http://www.co2pe.org/ [4] International Energy Agency. Worldwide Trends in Energy Use and Efficiency. Paris: IEA publications; 2008. [5] Duflou JR, Sutherland JW, Dornfeld D, Herrmann C, Jeswiet J, Kara S, Hauschild M, Kellens K. Towards energy and resource efficient manufacturing: A processes and systems approach. CIRP Annals – Manufacturing Technology 2012;61:587-609. [6] Yoon H-S, Lee J-Y, Kim M-S, Kim E-S, Ahn S-H. Effect of scale of machining hardware on energy consumption. International Symposium on Green Manufacturing and Application (ISGMA 2013); 2013 June 25-29, Ohau, Hawaii, USA. [7] Balogun VA, Mativenga PT. Modeling of direct energy requirements in mechanical machining processes. Journal of Cleaner Production 2013;41:179-86. [8] Yoon H-S, Moon J-S, Pham M-Q, Lee G-B, Ahn S-H. Control of machining parameters for energy and cost savings in micro-scale drilling of PCBs. Journal of Cleaner Production 2013;54:41-8. [9] Draganescu F, Gheorghe M, Doicin CV. Models of machine tool efficiency and specific consumed energy. Journal of Materials Processing Technology. 2003;141:9-15. [10] Gutowski TG, Branham MS, Dahmus JB, Jones AJ, Thiriez A, Sekulic DP. Thermodynamic Analysis of Resources Used in Manufaturing Processes, Environmental Science & Technology 2009;43(5):1584-90. [11] He Y, Liu F, Wu T, Zhong F, Peng B. Analysis and estimation of energy consumption for numerical control machining. Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture 2012;226:255-66. [12] Wang Q, Liu F, Li C. An integrated method for assessing the energy efficiency of machining workshop. Journal of Cleaner Production 2013;52:122-33. [13] Li L, Yan J, Xing Z. Energy requirements evaluation of milling machines based on thermal equilibrium and empirical modelling. Journal of Cleaner Production 2013;52:113-21. [14] Gutowski T, Dahmus J, Thiriez A, Branham M, Jones A. A Thermodynamic Characterization of Manufacturing Processes. Proceedings of the 2007 IEEE International Symposium on Electronics and the Environment; 2007 May 7-10, Orlando, FL, USA. [15] Zhao F, Bernstein WZ, Naik G, Cheng GJ. Environmental assessment of laser assisted manufacturing: case studies on laser shock peening and laser assisted turning. Journal of Cleaner Production 2010;18:1311-19. [16] Skvarenina S, Shin YC. Laser-assisted machining of compacted graphite iron. International Journal of Machine Tools and Manufacture 2006;46:717. [17] Helu M, Behmann B, Meier H, Dornfeld D, Lanza G, Schulze V. Impact of green machining strategies on achieved surface quality. CIRP Annals Manufacturing Technology 2012;61:55-8.