- Email: [email protected]
Tribological aspects of the choice of metalworking fluid in cutting processes
Journal of Materials Processing Technology 124 (2002) 25–31
Tribological aspects of the choice of metalworking fluid in cutting processes Radoslav Rakic´a,*, Zlata Rakic´b a
NIS-Naftagas promet, Bulevar Oslobodjenja 69, 21000 Novi Sad, Yugoslavia b Vojvodjanska banka a.d., Miroslava Antic´a 2, 21000 Novi Sad, Yugoslavia Received 1 May 2000
Abstract The need for surface quality, shape and dimensional precision, or the use of materials that are difficult to cut, mean that in most cases metalworking fluids will continue to be important in the future. The recommendation of metalworking fluids depends on the special application. For applications where a metalworking fluid with better lubricating properties is needed, a non-water-miscible fluid is to be recommended (ISO-L-MHx). In other cases with high cutting velocities, a water-miscible fluid is often preferred due to its better cooling properties (ISO-L-MAx). The aim of this work is to explore the tribological aspects of the choice of metalworking fluid on cutting processes. Experimental investigations of the influence of the choice of metalworking fluid on cutting processes have been carried out at the metalworking factory. These investigations were carried out on 36 milling machines over four periods of time, each being 2000 working hours. The lifetime of milling machine elements up to failure due to the influence of metalworking fluids mostly shows large deviations. Using probability and statistical methods, it is possible to determine the influence of the choice of metalworking fluid on tribological processes in relation to the lifetime and reliability of the operation of the milling machines. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Tribological aspects; Metalworking fluid; Milling machine
1. Introduction Tribology includes the group of technologies required to understand and control the complex mechanical, metallurgical, chemical, physical and other phenomena that occur when two bodies are in sliding contact. The metalworking tribosystem can be extremely complex. In addition to the many tribological requirements of these manufacturing processes, developers of metalworking lubricants must optimize the solution properties and the environmental, safety and health characteristics of these lubricants [1]. Friction between the tool and workpiece depends on a multitude of factors such as process parameters, cutting tool geometry and tool material, acting forces, pressure between tool and work interface, heat generation during the process, temperature of contact zone and the cutting fluid applied [2]. Cutting processes are primarily governed by extremely complex and interdependent physical–chemical–mechanical, in other word tribological, phenomena in the contact zone of the cutting tool and material causing the tool to wear, the material to be removed from the surface of the blank part, thus generating the required surface geometrical *
Corresponding author. Fax: þ381-21-481-4505.
configuration, accuracy and surface quality [3]. Much is written in the technical literature of tribology about the lubrication mechanisms of various metalworking processes or the efficacy and efficiency of various metalworking additives. Schey [4] discusses the purposes and attributes of metalworking lubricants in some detail. Unfortunately, little information appears about the ways in which this technical information can be integrated into a lubricant development program or even about the way a new lubricant development effort should be conducted. The cutting process is present in all kinds of metal cutting operations such as turning, drilling, milling, grinding, etc., by the penetration of the wedge-shape portion of the cutting tool or grinding wheel into the workpiece material. During the cutting process a certain amount of workpiece material transforms into chips (particles), which slide across the front surfaces of the cutting wedge, leaving the cut area. The nature of the tribological process is very complex because of the high temperatures and pressures appearing on the contact surface. Transfer of particle mass, within the contact-making process, is carried out from the tool to the workpiece material and into the coolant, from the workpiece to the tool and into the coolant, and from the coolant to the tool and the workpiece. The progress made by tribology during the
0924-0136/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 4 - 0 1 3 6 ( 0 1 ) 0 1 1 5 0 - 5
26
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
past 25 years is indeed impressive, both in technological achievement, including improved understanding of the role of lubricants and of the atmospheric environment in cutting and abrasive machining processes [5]. The effectiveness of coolants has been studied in several projects and more detailed explanations can be found in [6–11]. Any metalworking fluid must satisfy certain fundamental requirements: removal of heat, lubrication and transport of metal removed. The coolant systems for metalworking are typical examples of a tribomechanical system, where several wear mechanisms are present simultaneously. The metal particles in the coolant system have the appearance of tiny chips, grains, coils of fine wire and so on. The particulate composition varies significantly depending on the cutting process, and on the operating conditions as well as the coolant characteristics and coolant systems. During the cutting process, the metalworking fluid is also exposed to changes due to chemical reactions between its substances, the tool surfaces and the workpieces; and due to the presence of particles and high temperatures on contact surfaces. The changes can impair the tribological properties and increase the risks associated with tribological processes on machine tool elements.
2. Coolants for metalworking The main functions of the coolants for metalworking are as follows: (i) to reduce frictional heating; (ii) to divert the generated heat from the workpiece and tool by an adequate flow of coolant; and (iii) to remove the chips produced in the process initially from the cutting tool. The recommendation of metalworking coolants depends on the special application. For applications where a metalworking fluid with better lubricating properties is needed, a non-water-miscible product is to be recommended. In other cases with high cutting velocities, a water-miscible product is often preferred due to its better cooling properties. From the maintenance point of view, non-water-miscible cutting oils cause relatively few problems. They undergo a relatively slow deterioration and the only necessary maintenance is filtering. Water-based coolants are liable to undergo a number of changes during use with respect to their composition and properties, which can impair the functional properties and increase the risks associated with abrasion, erosion and corrosion of tribomechanical system elements. The main influential factors in the coolant system are: particles, temperature, fluid composition and flow. The presence of particles has a negative influence on surface quality, machine life, tool life, coolant life and on health aspect. The temperature rise in coolants can be quite considerable, if the machine is of the type where the coolant sump forms an integral part of the machine framework. An increased temperature as compared to the normal 18–208C will result in a very considerable increase in bacterial growth and problem with tolerances.
The composition of a coolant changes during use as a consequence of vapourisation of water and the removal of various materials in the filtering process. Depending on the relationship between these factors, these changes will thereby result in either a concentration or a dilution of the fluid. With respect to the significant vapourisation which takes place, it is essential to monitor the quality of the water which is added to maintain the correct dilution. The salt content in tap water causes accumulation of salts as more water is added. The main undesirable consequence of a high salt content on the process is excessive corrosion of tribomechanical system elements. The other influential factors are: free oils and micro-organisms. Free oil is a combination of ‘‘tramp oil’’ from the lubrication and drive systems of the process machines, oil washed down from guideways, rust-inhibiting oil from workpieces and oil resulting from emulsion breakdown. The presence of free oil on the coolant surface results in a number of undesirable consequences for both the effectiveness of the process and for health aspects. Large numbers of microorganisms (bacteria and fungi) shorten the coolants useful to life to a considerable degree and causes a number of technical problems. A large concentration of micro-organisms causes an accumulation of metabolic organisms in the fluid which can result in skin ailments. The recommendation of metalworking fluids depends on the special application. The present experimental investigations were carried out on milling machines and the milling conditions were determined using water-based coolant. International Standard ISO 6743/7 establishes the detailed classification of family M (metalworking) which belongs to the class L (lubricants, industrial oils and related products) [12]. This detailed classification of family ‘‘M’’ has been established by defining the categories of products required for the main applications of this family. Each category is designated by a symbol consisting of a group of three letters, which together constitute a code. The first letter of the category (M) identifies the family of the product considered but any following letters taken separately have no significance on their own. Classifications of aqueous fluids for metalworking (family M) are shown in Table 1, where: A — code letter; B — particular application; C — more specific application; D — product type and/or end use requirements; E — symbol ISO-L.
3. Coolant systems for a machine tool There are decentralized and centralized systems. In the first case, each machine has its own coolant system which consists of a tank, a pump, distribution plumbing and a swarf tray, as shown in Fig. 1. The main characteristics of decentralized systems are: (i) the system is difficult for cleaning, especially when the tank is an integral part of the machine frame; (ii) cleaning consists only of a course filtering and sedimentation, in
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
27
Table 1 Classification of aqueous fluids for metalworking (family M) A
B
C
D
E
Metal removal by cutting or abrasion and metal forming by punching deep drawing, ironing, power spinning, wire drawing, forging — hot and cold, extrusion, stamping, rolling — hot and cold
Operations primarily needing cooling
Concentrates giving, when blended with water, milky emulsions having anticorrosion properties
MAA
Concentrates of MAA type having friction-reducing properties Concentrates of MAA type having extreme pressure (EP) properties Concentrates of MAB type having extreme pressure (EP) properties Concentrates giving, when blended with water, translucent emulsions (micro-emulsion) having anti-corrosion properties Concentrates of MAE type having friction-reducing and/or extreme pressure (EP) properties Concentrates giving, when blended with water, transparent solution having anti-corrosion properties Concentrates of MAG type having friction reducing and/or extreme pressure (EP) properties Greases and pastes applied blended with water
MAB MAC MAD MAE
which only larger chips are retained; (iii) there often occurs an undesirable temperature increase of the coolant due to heat transfer from the machine; (iv) during periods when the coolants are not used, anaerobic bacterial growth will likely occur, which is extremely corrosive towards the majority of metals. Centralized systems are those in which the coolant is returned to a central unit in which it undergoes a recondi-
MAF MAG MAH MAI
tioning of some form or another before being recirculated to the individual machine tools, as is shown in Fig. 2. The main characteristics of centralized systems are: (i) it becomes economical to carry out an effective conditioning of the coolant by means of filtering, cooling and so on; (ii) monitoring and control are limited to a single point of both measurement and adjustment of concentration and flow, (iii) it becomes possible to automatically control the coolant concentration and flow.
4. Experimental investigations
Fig. 1. Decentralized (individual) coolant system.
Experimental investigations of the effect of the selection of the metalworking fluids on tribological processes of tribomechanical system elements have been carried out at a metalworking factory. The investigations were carried out on 36 milling machines in four periods of time (I, II, III, IV), each of 2000 working hours. In periods I and III of the investigation, decentralized systems were used as coolant systems on these milling machines. Maintenance consists only of a coarse individual filtering and sedimentation, in which only large chips are retained. In periods II and IV of the investigation, a partially centralized system was used as the coolant system. In the partially centralized system, an effective conditioning of the metalworking lubricant was carried out by means of: (i) fine filtering, i.e. removal of particulates; (ii) removal of tramp oil; (iii) temperature control; and (iv) monitoring with measurement and adjustment of concentration and flow. Removal of particulates, i.e. fine filtering, was carried out by passing the metalworking lubricant through a magnetic field whereby the particles were retained and then removed by scraping. Removal of tramp oil, i.e. free oil, was carried out by means of sedimentation followed by a removal of the oil by adhesion (using a tramp oil skimmer).
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
28
Fig. 2. Centralized coolant system where n is the number of machine tools.
Table 2 Milling conditions Machine tool Cutting tool Cutting conditions Speed, v Feed rate, s Depth, d Material of workpiece Coolant
Concentration Flow, Qc
Universal milling machines Facing (milling) cutter (YU) 28–50 m/min 0.13–0.40 mm/tooth 1–3 mm Steel (YU) Micro-emulsion (ISO-L-MAE) (periods I and II) Synthetic concentrate (ISO-L-MAG) (periods III and IV) 5% 10–20 l/min
Table 3 Typical characteristics of the commercial coolants for milling process
Fig. 3. Tribomechanical system for the milling process.
Controlling the concentration is carried out by measuring with a refractometer and blending by means of a battery of ejector mixers. Fig. 3 shows the tribomechanical system for the milling process, which consists of the workpiece (element 1), the cutting tool (element 2), and the coolant (element 3).
Typical characteristics
Micro-emulsion
Synthetic concentrate
Appearance of concentrate Evaporation of residium Appearance and stability of 5% solution pH value of 5% solution Corrosion of 5% solution
Clear Oily Opalescent, stable
Clear Oily Transparent, stable
8.8–9.2 Negative
8.5–9.0 2%; negative
The milling conditions are given in Table 2. A medium carbon steel was selected as one of the workpiece materials because it is one of the most common engineering metals, being commonly used to manufacture machine parts and tools.
Table 4 Relevant parameters of the test procedure Symptoms of failure
Relevant parameters Indication
Maximum values before failure
Contamination with particles ðparticle size ¼ 503000 mmÞ Corrosion
Particle concentration, Kp Corrosion due to decreasing concentration Concentration, K Coolant temperature, t Tolerance, t
Kp > Kp max ; Kp max ¼ 30 wt:%
Increase of coolant temperature Decrease of machining accuracy Other symptoms Free oil Lime deposits on elements Sticking
The free oil content must be kept as low as possible Decreasing salt concentration (water hardness) Sticking due to increasing concentration ðKmax ¼ 10%Þ
K < Kmin ; Kmin ¼ 3ð2Þ% t > tmax ; tmax ¼ 35 C t > tmax of milling operation
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
29
The milling conditions were determined using the commercial coolants for the milling process. Typical characteristics of the commercial coolants for the milling process are summarized in Table 3. The relevant parameters of the test procedure in the experimental milling process are shown in Table 4.
5. Results and discussion of the experimental investigation The lifetime of the tribomechanical system elements up to failures due to the influence of coolant mostly shows large deviations. By the aid of probability and statistical methods, it is possible to determine the influence of coolants for metalworking on the tribological processes in relationship to the lifetime and reliability of operation of the tribomechanical system elements. Fig. 4 shows the main types of failure symptoms of tribomechanical system elements, which are: (1) contamination with particles, (2) corrosion, (3) increase of temperature, (4) decrease of machining accuracy due to the tribological processes, (5) other symptoms (composition, flow and so on), where p is the age percent of failures. Analysis of the results showed that the most common single symptom of element failures was contamination with particles, i.e. about 40%, with contamination leading progressively to decrease of machining accuracy. There were also a number of failures initiated by corrosion and increase of temperature and some of these could have been prevented by more attention to the maintenance of the coolants. The proportion of failures initiated by the other causes (composition and flow) is relatively small but not insignificant, especially when the high costs of milling machine failures are taken into account. The authors have discussed three major measures of the reliability effectiveness of tribomechanical system elements: (i) the failure rate — l; (ii) average life — T c ; (iii) the reliability curve — R(T).
Fig. 5. Analysis of the average life of critical elements T c .
The average life is the mean time between or to failure of critical tribomechanical system elements. Critical tribomechanical system elements are those which are the first and mostly breakdown, damage and premature failure happened. The results of the average life T c of critical elements of milling machine as a function of types of metalworking fluid and types of coolant system in periods I–IV of investigations are shown in Fig. 5. As is evident from Fig. 5, the average life T c (h) of critical elements after the use of decentralized coolant systems is 7.5% (7.7%) lower (average 1230 h/ISO-L-MAG, i.e. 1200 h/ ISO-L-MAE) than that of critical elements (average 1330 h/ ISO-L-MAG, i.e. 1300 h/ISO-L-MAE) after the use of a partially centralized coolant system. Fig. 6 shows the curves of reliability of critical tribomechanical system elements versus time on these milling machines as a function of the tribological properties of metalworking fluid and the coolant system for metalworking in periods I–IVof the investigations. On the basis of inspection (testing hypotheses), it can be concluded that the curves of reliability are approximately distributed experimentally. The most familiar way to test for Poisson/exponential character is to use a w2 goodness-of-fit test on the data presented in Figs. 4–6, against a theoretical distribution, with the parameter replaced by its maximumlikelihood estimator. The resulting statistic is then w2k ¼
n X ðfi ft Þ2 i
i¼1
Fig. 4. Analysis of the symptoms of element failures.
fti
where fi is the number observed in the ith frequency class, fti the number expected in the ith frequency class if the hypothesized distribution were correct, and k the number of degrees of freedom, always equal to the total number of classes less one and then minus one for each parameter estimated. The w2-test for the data of Figs. 4 and 5 is shown in Table 5 and the w2-test for data of Fig. 6 is shown in Table 6.
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
30
Fig. 6. Curves of reliability, where R(T) is reliability and T the time (h). Table 5 w2-test for the data of Figs. 4 and 5 (ISO-L-MAE/individual) Number
Actual
Theoretical
Difference
(Difference)2
(Difference)2/theoretical
0 1 2 3 4 5 6
13 19 13 8 5 1 1
10.96 18.64 15.86 8.97 3.82 1.29 0.46
2.04 0.36 2.86 0.97 1.18 0.29 0.54
4.16 0.13 8.18 0.94 1.39 0.08 0.29
0.3795 0.0069 0.5157 0.1048 0.3638 0.0620 0.6304
S
60
60.00
If n is the number of failures for 2000 h, then pn ¼
e
1:667
1:667 n!
n
Poisson distributed:
fti ð> TÞ ¼ 60 eT=T c
and if the degrees of freedom k ¼ 7 2 ¼ 5 and the confidence interval is 95%, then w20:05 ¼ 11:0705;
w2 ¼ 2:0631
w2 < w20:05
Based on the statistical goodness-of-fit test to these data, it can be concluded that the failures are approximately
where T c ffi 1200 h
The degrees of freedom are: k ¼ 7 2 ¼ 5 and confidence interval is 95%, so that w20:05 ¼ 11:0705;
w2 < w20:05
Based on the statistical goodness-of-fit test to these data, it can be concluded that the reliability curves are approximately exponentially distributed.
Table 6 w2-test for data of Fig. 6. (ISO-L-MAE/individual) T(h)
Actual
Theoretical
Difference
(Difference)2
(Difference)2/theoretical
0–500 501–1000 1001–1500 1501–2000 2001–2500 2501–3000 3001–3500
17 14 10 7 5 3 4
20.4 13.6 8.8 5.8 4.0 2.5 4.9
3.4 0.6 1.2 1.2 1.0 0.5 0.9
11.56 0.36 1.44 1.44 1.00 0.25 0.81
0.5667 0.0264 0.1636 0.2483 0.2500 0.1000 0.1653
S
60
60.0
w2 ¼ 1:5203
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
As is evident from Fig. 6, the least reliable are tribomechanical system elements in the Ist period of investigation, i.e. for micro-emulsion ISO-L-MAE from individual systems (the greatest inclination of the curve R(T)). They are the most reliable in the IVth period of investigation, i.e. for synthetic concentrate ISO-L-MAG from a partially centralized system (the least inclination of the curve R(T)). According to the milling conditions (Table 2) and coolant characteristics (Table 3), the consequent differences in life in relation to the tribological properties of coolants are: 1. The effect of metalworking fluid type. The metalworking fluid (synthetic concentrate) ISO-L-MAG obviously gave a longer life, compared with the metalworking fluid (micro-emulsion) ISO-L-MAE. 2. The effect of coolant system type. The partially centralized coolant system obviously gave a longer life compared with decentralized coolant systems in the milling process.
6. Conclusions The data were arranged according to different types of metalworking fluid of the same coolant system and different coolant systems of the same type of metalworking fluid to identify the influence of the metalworking fluid and type, and the coolant system type on the milling process. The following conclusions can be drawn from the results presented above: 1. The tribomechanical system failures were found to be affected by both the coolant systems and the type of metalworking fluids, under the present operating conditions. 2. The metalworking fluid ISO-L-MAG gives a longer tribomechanical system life compared with the case of ISO-L-MAE. 3. The partially centralized coolant system gives a longer tribomechanical system life compared with the case of decentralized coolant systems. The paper presents the following: 1. The analysis of the causes and the symptoms of failure of critical elements. 2. The analysis of the average life of critical elements as a function of the effect of the metalworking fluids.
31
3. The curve of the reliability of critical tribomechanical system elements of milling machines as a function of the effect of coolants for metalworking and the maintenance of the coolant system under the operating conditions of the present investigation. The results obtained have established that the right choice of coolants, and by monitoring and maintenance of the coolant systems for metalworking, the maximum reliability of the machine tools can be achieved and the tribomechanical system elements can be protected from damage and premature failure. References [1] M.E. Kipp, L.B. Riddle, A Guide to the Development of Advanced Metalworking Lubricants, in: Proceedings of the Eighth International Colloquium, Vol. 2, Ostfildern, Germany, 1992, pp. 18.1.1–18.1.14. [2] A. Kaldos, An Investigation into the Tribological and Environmental Aspects of Cutting Processes, World Tribology Congress, London, 1997, p. 901 (Abstracts of Papers). [3] H. Czichos, The role of tribology as science and technology — what are the essentials? Tribol. Int. 28 (1) (1995) 15–16. [4] J.A. Schey, Purposes and attributes of metalworking lubricants, Lubr. Eng. 24 (5) (1967) 193–198. [5] H.P. Jost, The first 25 years and beyond — achievements, shortcomings and future tasks, in: Proceedings of the Eighth International Colloquium, Vol. 12, Ostfildern, Germany, 1992, pp. 1.1.1–1.1.17. [6] R. Rakic´ , Reliability of coolant systems for metalworking, in: Proceedings of the Fifth International Conference on Flexible Technologies, MMA’94, Novi Sad, Yugoslavia, 1994, pp. 559–564. [7] R. Rakic´ , Z. Rakic´ , Influence of coolants for metalworking on erosion and corrosion of tribomechanical system elements, in: Book of Abstracts of Eighth International Conference on Erosion by Liquid and Solid Impact (ELSI VIII), No. 15, Cambridge, UK, 1994. [8] R. Rakic´ , The effects of maintenance of coolants for metalworking on tribological processes, in: Proceedings of the Symposium on YUNG 4P’95, Vol. P4, Vrnjacˇ ka Banja, Yugoslavia, 1995, pp. 157– 162 (in Serbian). [9] R. Rakic´ , The influence of coolants for metalworking on machine tool failures, in: Proceedings of the Second International Conference on Tribology, Balkantrib’96, Thessaloniki, Greece, 1996, pp. 827–833. [10] R. Rakic´ , Z. Rakic´ , The influence of coolants for metal working on tribological processes, in: Proceedings of the International Manufacturing Engineering Conference, Storrs, CT, 1996, pp. 540–542. [11] R. Rakic´ , Z. Rakic´ , The effect of the selection of the metalworking fluids on tribological processes, in: Proceedings of the 11th International Colloquium Industrial and Automotive Lubrication, Ostfildern, Germany, 1998, pp. 57–62. [12] Classification of lubricants, industrial oils and related products (Class L), Part 7, Family M (metalworking), ISO 6743, 1986.
Tribological aspects of the choice of metalworking fluid in cutting processes Radoslav Rakic´a,*, Zlata Rakic´b a
NIS-Naftagas promet, Bulevar Oslobodjenja 69, 21000 Novi Sad, Yugoslavia b Vojvodjanska banka a.d., Miroslava Antic´a 2, 21000 Novi Sad, Yugoslavia Received 1 May 2000
Abstract The need for surface quality, shape and dimensional precision, or the use of materials that are difficult to cut, mean that in most cases metalworking fluids will continue to be important in the future. The recommendation of metalworking fluids depends on the special application. For applications where a metalworking fluid with better lubricating properties is needed, a non-water-miscible fluid is to be recommended (ISO-L-MHx). In other cases with high cutting velocities, a water-miscible fluid is often preferred due to its better cooling properties (ISO-L-MAx). The aim of this work is to explore the tribological aspects of the choice of metalworking fluid on cutting processes. Experimental investigations of the influence of the choice of metalworking fluid on cutting processes have been carried out at the metalworking factory. These investigations were carried out on 36 milling machines over four periods of time, each being 2000 working hours. The lifetime of milling machine elements up to failure due to the influence of metalworking fluids mostly shows large deviations. Using probability and statistical methods, it is possible to determine the influence of the choice of metalworking fluid on tribological processes in relation to the lifetime and reliability of the operation of the milling machines. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Tribological aspects; Metalworking fluid; Milling machine
1. Introduction Tribology includes the group of technologies required to understand and control the complex mechanical, metallurgical, chemical, physical and other phenomena that occur when two bodies are in sliding contact. The metalworking tribosystem can be extremely complex. In addition to the many tribological requirements of these manufacturing processes, developers of metalworking lubricants must optimize the solution properties and the environmental, safety and health characteristics of these lubricants [1]. Friction between the tool and workpiece depends on a multitude of factors such as process parameters, cutting tool geometry and tool material, acting forces, pressure between tool and work interface, heat generation during the process, temperature of contact zone and the cutting fluid applied [2]. Cutting processes are primarily governed by extremely complex and interdependent physical–chemical–mechanical, in other word tribological, phenomena in the contact zone of the cutting tool and material causing the tool to wear, the material to be removed from the surface of the blank part, thus generating the required surface geometrical *
Corresponding author. Fax: þ381-21-481-4505.
configuration, accuracy and surface quality [3]. Much is written in the technical literature of tribology about the lubrication mechanisms of various metalworking processes or the efficacy and efficiency of various metalworking additives. Schey [4] discusses the purposes and attributes of metalworking lubricants in some detail. Unfortunately, little information appears about the ways in which this technical information can be integrated into a lubricant development program or even about the way a new lubricant development effort should be conducted. The cutting process is present in all kinds of metal cutting operations such as turning, drilling, milling, grinding, etc., by the penetration of the wedge-shape portion of the cutting tool or grinding wheel into the workpiece material. During the cutting process a certain amount of workpiece material transforms into chips (particles), which slide across the front surfaces of the cutting wedge, leaving the cut area. The nature of the tribological process is very complex because of the high temperatures and pressures appearing on the contact surface. Transfer of particle mass, within the contact-making process, is carried out from the tool to the workpiece material and into the coolant, from the workpiece to the tool and into the coolant, and from the coolant to the tool and the workpiece. The progress made by tribology during the
0924-0136/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 4 - 0 1 3 6 ( 0 1 ) 0 1 1 5 0 - 5
26
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
past 25 years is indeed impressive, both in technological achievement, including improved understanding of the role of lubricants and of the atmospheric environment in cutting and abrasive machining processes [5]. The effectiveness of coolants has been studied in several projects and more detailed explanations can be found in [6–11]. Any metalworking fluid must satisfy certain fundamental requirements: removal of heat, lubrication and transport of metal removed. The coolant systems for metalworking are typical examples of a tribomechanical system, where several wear mechanisms are present simultaneously. The metal particles in the coolant system have the appearance of tiny chips, grains, coils of fine wire and so on. The particulate composition varies significantly depending on the cutting process, and on the operating conditions as well as the coolant characteristics and coolant systems. During the cutting process, the metalworking fluid is also exposed to changes due to chemical reactions between its substances, the tool surfaces and the workpieces; and due to the presence of particles and high temperatures on contact surfaces. The changes can impair the tribological properties and increase the risks associated with tribological processes on machine tool elements.
2. Coolants for metalworking The main functions of the coolants for metalworking are as follows: (i) to reduce frictional heating; (ii) to divert the generated heat from the workpiece and tool by an adequate flow of coolant; and (iii) to remove the chips produced in the process initially from the cutting tool. The recommendation of metalworking coolants depends on the special application. For applications where a metalworking fluid with better lubricating properties is needed, a non-water-miscible product is to be recommended. In other cases with high cutting velocities, a water-miscible product is often preferred due to its better cooling properties. From the maintenance point of view, non-water-miscible cutting oils cause relatively few problems. They undergo a relatively slow deterioration and the only necessary maintenance is filtering. Water-based coolants are liable to undergo a number of changes during use with respect to their composition and properties, which can impair the functional properties and increase the risks associated with abrasion, erosion and corrosion of tribomechanical system elements. The main influential factors in the coolant system are: particles, temperature, fluid composition and flow. The presence of particles has a negative influence on surface quality, machine life, tool life, coolant life and on health aspect. The temperature rise in coolants can be quite considerable, if the machine is of the type where the coolant sump forms an integral part of the machine framework. An increased temperature as compared to the normal 18–208C will result in a very considerable increase in bacterial growth and problem with tolerances.
The composition of a coolant changes during use as a consequence of vapourisation of water and the removal of various materials in the filtering process. Depending on the relationship between these factors, these changes will thereby result in either a concentration or a dilution of the fluid. With respect to the significant vapourisation which takes place, it is essential to monitor the quality of the water which is added to maintain the correct dilution. The salt content in tap water causes accumulation of salts as more water is added. The main undesirable consequence of a high salt content on the process is excessive corrosion of tribomechanical system elements. The other influential factors are: free oils and micro-organisms. Free oil is a combination of ‘‘tramp oil’’ from the lubrication and drive systems of the process machines, oil washed down from guideways, rust-inhibiting oil from workpieces and oil resulting from emulsion breakdown. The presence of free oil on the coolant surface results in a number of undesirable consequences for both the effectiveness of the process and for health aspects. Large numbers of microorganisms (bacteria and fungi) shorten the coolants useful to life to a considerable degree and causes a number of technical problems. A large concentration of micro-organisms causes an accumulation of metabolic organisms in the fluid which can result in skin ailments. The recommendation of metalworking fluids depends on the special application. The present experimental investigations were carried out on milling machines and the milling conditions were determined using water-based coolant. International Standard ISO 6743/7 establishes the detailed classification of family M (metalworking) which belongs to the class L (lubricants, industrial oils and related products) [12]. This detailed classification of family ‘‘M’’ has been established by defining the categories of products required for the main applications of this family. Each category is designated by a symbol consisting of a group of three letters, which together constitute a code. The first letter of the category (M) identifies the family of the product considered but any following letters taken separately have no significance on their own. Classifications of aqueous fluids for metalworking (family M) are shown in Table 1, where: A — code letter; B — particular application; C — more specific application; D — product type and/or end use requirements; E — symbol ISO-L.
3. Coolant systems for a machine tool There are decentralized and centralized systems. In the first case, each machine has its own coolant system which consists of a tank, a pump, distribution plumbing and a swarf tray, as shown in Fig. 1. The main characteristics of decentralized systems are: (i) the system is difficult for cleaning, especially when the tank is an integral part of the machine frame; (ii) cleaning consists only of a course filtering and sedimentation, in
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
27
Table 1 Classification of aqueous fluids for metalworking (family M) A
B
C
D
E
Metal removal by cutting or abrasion and metal forming by punching deep drawing, ironing, power spinning, wire drawing, forging — hot and cold, extrusion, stamping, rolling — hot and cold
Operations primarily needing cooling
Concentrates giving, when blended with water, milky emulsions having anticorrosion properties
MAA
Concentrates of MAA type having friction-reducing properties Concentrates of MAA type having extreme pressure (EP) properties Concentrates of MAB type having extreme pressure (EP) properties Concentrates giving, when blended with water, translucent emulsions (micro-emulsion) having anti-corrosion properties Concentrates of MAE type having friction-reducing and/or extreme pressure (EP) properties Concentrates giving, when blended with water, transparent solution having anti-corrosion properties Concentrates of MAG type having friction reducing and/or extreme pressure (EP) properties Greases and pastes applied blended with water
MAB MAC MAD MAE
which only larger chips are retained; (iii) there often occurs an undesirable temperature increase of the coolant due to heat transfer from the machine; (iv) during periods when the coolants are not used, anaerobic bacterial growth will likely occur, which is extremely corrosive towards the majority of metals. Centralized systems are those in which the coolant is returned to a central unit in which it undergoes a recondi-
MAF MAG MAH MAI
tioning of some form or another before being recirculated to the individual machine tools, as is shown in Fig. 2. The main characteristics of centralized systems are: (i) it becomes economical to carry out an effective conditioning of the coolant by means of filtering, cooling and so on; (ii) monitoring and control are limited to a single point of both measurement and adjustment of concentration and flow, (iii) it becomes possible to automatically control the coolant concentration and flow.
4. Experimental investigations
Fig. 1. Decentralized (individual) coolant system.
Experimental investigations of the effect of the selection of the metalworking fluids on tribological processes of tribomechanical system elements have been carried out at a metalworking factory. The investigations were carried out on 36 milling machines in four periods of time (I, II, III, IV), each of 2000 working hours. In periods I and III of the investigation, decentralized systems were used as coolant systems on these milling machines. Maintenance consists only of a coarse individual filtering and sedimentation, in which only large chips are retained. In periods II and IV of the investigation, a partially centralized system was used as the coolant system. In the partially centralized system, an effective conditioning of the metalworking lubricant was carried out by means of: (i) fine filtering, i.e. removal of particulates; (ii) removal of tramp oil; (iii) temperature control; and (iv) monitoring with measurement and adjustment of concentration and flow. Removal of particulates, i.e. fine filtering, was carried out by passing the metalworking lubricant through a magnetic field whereby the particles were retained and then removed by scraping. Removal of tramp oil, i.e. free oil, was carried out by means of sedimentation followed by a removal of the oil by adhesion (using a tramp oil skimmer).
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
28
Fig. 2. Centralized coolant system where n is the number of machine tools.
Table 2 Milling conditions Machine tool Cutting tool Cutting conditions Speed, v Feed rate, s Depth, d Material of workpiece Coolant
Concentration Flow, Qc
Universal milling machines Facing (milling) cutter (YU) 28–50 m/min 0.13–0.40 mm/tooth 1–3 mm Steel (YU) Micro-emulsion (ISO-L-MAE) (periods I and II) Synthetic concentrate (ISO-L-MAG) (periods III and IV) 5% 10–20 l/min
Table 3 Typical characteristics of the commercial coolants for milling process
Fig. 3. Tribomechanical system for the milling process.
Controlling the concentration is carried out by measuring with a refractometer and blending by means of a battery of ejector mixers. Fig. 3 shows the tribomechanical system for the milling process, which consists of the workpiece (element 1), the cutting tool (element 2), and the coolant (element 3).
Typical characteristics
Micro-emulsion
Synthetic concentrate
Appearance of concentrate Evaporation of residium Appearance and stability of 5% solution pH value of 5% solution Corrosion of 5% solution
Clear Oily Opalescent, stable
Clear Oily Transparent, stable
8.8–9.2 Negative
8.5–9.0 2%; negative
The milling conditions are given in Table 2. A medium carbon steel was selected as one of the workpiece materials because it is one of the most common engineering metals, being commonly used to manufacture machine parts and tools.
Table 4 Relevant parameters of the test procedure Symptoms of failure
Relevant parameters Indication
Maximum values before failure
Contamination with particles ðparticle size ¼ 503000 mmÞ Corrosion
Particle concentration, Kp Corrosion due to decreasing concentration Concentration, K Coolant temperature, t Tolerance, t
Kp > Kp max ; Kp max ¼ 30 wt:%
Increase of coolant temperature Decrease of machining accuracy Other symptoms Free oil Lime deposits on elements Sticking
The free oil content must be kept as low as possible Decreasing salt concentration (water hardness) Sticking due to increasing concentration ðKmax ¼ 10%Þ
K < Kmin ; Kmin ¼ 3ð2Þ% t > tmax ; tmax ¼ 35 C t > tmax of milling operation
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
29
The milling conditions were determined using the commercial coolants for the milling process. Typical characteristics of the commercial coolants for the milling process are summarized in Table 3. The relevant parameters of the test procedure in the experimental milling process are shown in Table 4.
5. Results and discussion of the experimental investigation The lifetime of the tribomechanical system elements up to failures due to the influence of coolant mostly shows large deviations. By the aid of probability and statistical methods, it is possible to determine the influence of coolants for metalworking on the tribological processes in relationship to the lifetime and reliability of operation of the tribomechanical system elements. Fig. 4 shows the main types of failure symptoms of tribomechanical system elements, which are: (1) contamination with particles, (2) corrosion, (3) increase of temperature, (4) decrease of machining accuracy due to the tribological processes, (5) other symptoms (composition, flow and so on), where p is the age percent of failures. Analysis of the results showed that the most common single symptom of element failures was contamination with particles, i.e. about 40%, with contamination leading progressively to decrease of machining accuracy. There were also a number of failures initiated by corrosion and increase of temperature and some of these could have been prevented by more attention to the maintenance of the coolants. The proportion of failures initiated by the other causes (composition and flow) is relatively small but not insignificant, especially when the high costs of milling machine failures are taken into account. The authors have discussed three major measures of the reliability effectiveness of tribomechanical system elements: (i) the failure rate — l; (ii) average life — T c ; (iii) the reliability curve — R(T).
Fig. 5. Analysis of the average life of critical elements T c .
The average life is the mean time between or to failure of critical tribomechanical system elements. Critical tribomechanical system elements are those which are the first and mostly breakdown, damage and premature failure happened. The results of the average life T c of critical elements of milling machine as a function of types of metalworking fluid and types of coolant system in periods I–IV of investigations are shown in Fig. 5. As is evident from Fig. 5, the average life T c (h) of critical elements after the use of decentralized coolant systems is 7.5% (7.7%) lower (average 1230 h/ISO-L-MAG, i.e. 1200 h/ ISO-L-MAE) than that of critical elements (average 1330 h/ ISO-L-MAG, i.e. 1300 h/ISO-L-MAE) after the use of a partially centralized coolant system. Fig. 6 shows the curves of reliability of critical tribomechanical system elements versus time on these milling machines as a function of the tribological properties of metalworking fluid and the coolant system for metalworking in periods I–IVof the investigations. On the basis of inspection (testing hypotheses), it can be concluded that the curves of reliability are approximately distributed experimentally. The most familiar way to test for Poisson/exponential character is to use a w2 goodness-of-fit test on the data presented in Figs. 4–6, against a theoretical distribution, with the parameter replaced by its maximumlikelihood estimator. The resulting statistic is then w2k ¼
n X ðfi ft Þ2 i
i¼1
Fig. 4. Analysis of the symptoms of element failures.
fti
where fi is the number observed in the ith frequency class, fti the number expected in the ith frequency class if the hypothesized distribution were correct, and k the number of degrees of freedom, always equal to the total number of classes less one and then minus one for each parameter estimated. The w2-test for the data of Figs. 4 and 5 is shown in Table 5 and the w2-test for data of Fig. 6 is shown in Table 6.
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
30
Fig. 6. Curves of reliability, where R(T) is reliability and T the time (h). Table 5 w2-test for the data of Figs. 4 and 5 (ISO-L-MAE/individual) Number
Actual
Theoretical
Difference
(Difference)2
(Difference)2/theoretical
0 1 2 3 4 5 6
13 19 13 8 5 1 1
10.96 18.64 15.86 8.97 3.82 1.29 0.46
2.04 0.36 2.86 0.97 1.18 0.29 0.54
4.16 0.13 8.18 0.94 1.39 0.08 0.29
0.3795 0.0069 0.5157 0.1048 0.3638 0.0620 0.6304
S
60
60.00
If n is the number of failures for 2000 h, then pn ¼
e
1:667
1:667 n!
n
Poisson distributed:
fti ð> TÞ ¼ 60 eT=T c
and if the degrees of freedom k ¼ 7 2 ¼ 5 and the confidence interval is 95%, then w20:05 ¼ 11:0705;
w2 ¼ 2:0631
w2 < w20:05
Based on the statistical goodness-of-fit test to these data, it can be concluded that the failures are approximately
where T c ffi 1200 h
The degrees of freedom are: k ¼ 7 2 ¼ 5 and confidence interval is 95%, so that w20:05 ¼ 11:0705;
w2 < w20:05
Based on the statistical goodness-of-fit test to these data, it can be concluded that the reliability curves are approximately exponentially distributed.
Table 6 w2-test for data of Fig. 6. (ISO-L-MAE/individual) T(h)
Actual
Theoretical
Difference
(Difference)2
(Difference)2/theoretical
0–500 501–1000 1001–1500 1501–2000 2001–2500 2501–3000 3001–3500
17 14 10 7 5 3 4
20.4 13.6 8.8 5.8 4.0 2.5 4.9
3.4 0.6 1.2 1.2 1.0 0.5 0.9
11.56 0.36 1.44 1.44 1.00 0.25 0.81
0.5667 0.0264 0.1636 0.2483 0.2500 0.1000 0.1653
S
60
60.0
w2 ¼ 1:5203
R. Rakic´ , Z. Rakic´ / Journal of Materials Processing Technology 124 (2002) 25–31
As is evident from Fig. 6, the least reliable are tribomechanical system elements in the Ist period of investigation, i.e. for micro-emulsion ISO-L-MAE from individual systems (the greatest inclination of the curve R(T)). They are the most reliable in the IVth period of investigation, i.e. for synthetic concentrate ISO-L-MAG from a partially centralized system (the least inclination of the curve R(T)). According to the milling conditions (Table 2) and coolant characteristics (Table 3), the consequent differences in life in relation to the tribological properties of coolants are: 1. The effect of metalworking fluid type. The metalworking fluid (synthetic concentrate) ISO-L-MAG obviously gave a longer life, compared with the metalworking fluid (micro-emulsion) ISO-L-MAE. 2. The effect of coolant system type. The partially centralized coolant system obviously gave a longer life compared with decentralized coolant systems in the milling process.
6. Conclusions The data were arranged according to different types of metalworking fluid of the same coolant system and different coolant systems of the same type of metalworking fluid to identify the influence of the metalworking fluid and type, and the coolant system type on the milling process. The following conclusions can be drawn from the results presented above: 1. The tribomechanical system failures were found to be affected by both the coolant systems and the type of metalworking fluids, under the present operating conditions. 2. The metalworking fluid ISO-L-MAG gives a longer tribomechanical system life compared with the case of ISO-L-MAE. 3. The partially centralized coolant system gives a longer tribomechanical system life compared with the case of decentralized coolant systems. The paper presents the following: 1. The analysis of the causes and the symptoms of failure of critical elements. 2. The analysis of the average life of critical elements as a function of the effect of the metalworking fluids.
31
3. The curve of the reliability of critical tribomechanical system elements of milling machines as a function of the effect of coolants for metalworking and the maintenance of the coolant system under the operating conditions of the present investigation. The results obtained have established that the right choice of coolants, and by monitoring and maintenance of the coolant systems for metalworking, the maximum reliability of the machine tools can be achieved and the tribomechanical system elements can be protected from damage and premature failure. References [1] M.E. Kipp, L.B. Riddle, A Guide to the Development of Advanced Metalworking Lubricants, in: Proceedings of the Eighth International Colloquium, Vol. 2, Ostfildern, Germany, 1992, pp. 18.1.1–18.1.14. [2] A. Kaldos, An Investigation into the Tribological and Environmental Aspects of Cutting Processes, World Tribology Congress, London, 1997, p. 901 (Abstracts of Papers). [3] H. Czichos, The role of tribology as science and technology — what are the essentials? Tribol. Int. 28 (1) (1995) 15–16. [4] J.A. Schey, Purposes and attributes of metalworking lubricants, Lubr. Eng. 24 (5) (1967) 193–198. [5] H.P. Jost, The first 25 years and beyond — achievements, shortcomings and future tasks, in: Proceedings of the Eighth International Colloquium, Vol. 12, Ostfildern, Germany, 1992, pp. 1.1.1–1.1.17. [6] R. Rakic´ , Reliability of coolant systems for metalworking, in: Proceedings of the Fifth International Conference on Flexible Technologies, MMA’94, Novi Sad, Yugoslavia, 1994, pp. 559–564. [7] R. Rakic´ , Z. Rakic´ , Influence of coolants for metalworking on erosion and corrosion of tribomechanical system elements, in: Book of Abstracts of Eighth International Conference on Erosion by Liquid and Solid Impact (ELSI VIII), No. 15, Cambridge, UK, 1994. [8] R. Rakic´ , The effects of maintenance of coolants for metalworking on tribological processes, in: Proceedings of the Symposium on YUNG 4P’95, Vol. P4, Vrnjacˇ ka Banja, Yugoslavia, 1995, pp. 157– 162 (in Serbian). [9] R. Rakic´ , The influence of coolants for metalworking on machine tool failures, in: Proceedings of the Second International Conference on Tribology, Balkantrib’96, Thessaloniki, Greece, 1996, pp. 827–833. [10] R. Rakic´ , Z. Rakic´ , The influence of coolants for metal working on tribological processes, in: Proceedings of the International Manufacturing Engineering Conference, Storrs, CT, 1996, pp. 540–542. [11] R. Rakic´ , Z. Rakic´ , The effect of the selection of the metalworking fluids on tribological processes, in: Proceedings of the 11th International Colloquium Industrial and Automotive Lubrication, Ostfildern, Germany, 1998, pp. 57–62. [12] Classification of lubricants, industrial oils and related products (Class L), Part 7, Family M (metalworking), ISO 6743, 1986.