Minimum lubrication in Al-Si drilling

Minimum lubrication in Al-Si drilling

D. U. Braga^I; A. E. Diniz^II; G. W. A. Miranda^III; N. L. Coppini^IV

^IDepartamento de Engenharia Mecânica Universidade Federal de São João Del Rei. E-mail: [e-mail] ^IIDepto de Engenharia de Fabricação, Faculdade de Engenharia Mecânica, UNICAMP. E-mail: [e-mail] ^IIIDepartamento de Engenharia Mecânica - Universidade de Taubaté. E-mail: [e-mail] ^IVDepto de Engenharia de Fabricação, Faculdade de Engenharia Mecânica, UNICAMP. E-mail: [e-mail]

ABSTRACT

The high percentage of the refrigeration costs (17%) in the operational production costs, the ecological subjects, the lawlegal demands related to the preservation of the environment and the human health justify the recent researches about the restriction towards the use of abundant cutting fluid in machining processes. However, it is important to point out that the use of a minimal mist lubrication (mixture of air and oil) has been possible in production processes for machining, due to the technological development of tool materials and also machine-tools. The objective of this work is to test the minimum quantity of lubricant (MQL) technique (10 ml/h of oil) mixed in a flow of air in the drilling process of aluminum silicon alloy (SAE 323) with a solid carbide drill (K10). The input variables were drill diameter (and consequentelyconsequently, depth of cut) and cutting speed. The results showed that when high depth of cut and cutting speed are used, the operation using abundant fluid is not possible because the drill breaks after a few holes. Moreover, the quality of the holes obtained using MQL is not worse than that obtained for with smaller cutting speed and depth of cut.

Keywords: Drilling, machining, aluminum-silicon alloy, minimal lubrication

Introduction

Lately a lot has been researched aiming to restrict the use of cutting fluid in the machining processes. The factors that justify such a procedure include operational cost production, ecological subjects, legal demands related to the preservation of the environment preservation, human health preservation, etc. (Heisel et al., 1998; Kalhöfer,1997; Klocke and Eisenblätter, 1997).

Two techniques have been proved successful to minimize the use of cutting fluids in machining processes. The first one is the dry cutting or cutting with no fluid (Granger, 1994; Lugscheider, 1997). To make it possible without causing a large decrease of tool life and loss of workpiece quality, it is mandatory to have suitable tool materials and cutting conditions. Hard coatings of tool materials, including the diamond coating, have been used to accomplish this purpose (Machado and Wallbank, 1997; Chiesa et al., 1995; Coelho et al.,1995). The optimization of cutting conditions to make them more suitable for dry cutting is made through the increase of feed and decrease of cutting speed. With this, roughly the same amount of heat is generated, but the area of the tool to receive this heat is bigger what makes the temperature lower and the amount of chip removed per minute constant (without increasing cutting time). This action may damage the workpiece surface finish due to the increase of the feed and, therefore, it is also necessary to increase the tool radius in order to keep the surface roughness at the same level (Klocke and Eisenblatter, 1997).

In those processes where dry cutting is either not possible or not economic, a second technique can be tried to reach the goal of minimizing the amount of cutting fluid in machining process: Minimum Quantity of Lubricant (MQL). This name is given to the process of pulverizing a very small amount of oil (less than 30 ml/h) in a flow of compressed air. Some good results have been obtained with this technique (Tonshoff, 1994). Lugscheider et al. (1997) used this technique in reaming process of gray cast iron - GG25 and aluminium alloy - AlSi12 with coated carbide tools and concluded that it caused a reduction of tool wear when compared with the completely dry process and, consequently, an improvement in the surface quality of the holes. Machado and Wallbank (1997) also used this technique in turning process of medium carbon steel and concluded that, in some cases, air or a mixture of air and water or air and soluble oil has been shown to be better than the overhead flooding application of soluble oil.

The drilling of aluminum-silicon alloys is one of those processes where dry cutting is impossible (Derflinger et al., 1999) due to the high ductility of the workpiece material. Without cooling and lubrication, the chip sticks to the tool and breaks it in a very short cutting time. Therefore, in this process a good alternative is the use of the MQL technique (Braga et. al., 1999; Braga et. al., 2002).

Considering the use of the MQL in machining, the vapor, the mist and the oil smoke of oil can be considered undesirable sub-products, characterizing an increase in pollution by suspension in the air (Heisel and Lutz, 1998). In Germany, the maximum polluter pollution concentration in the air under the mist form is limited in to 5mg/m³ and for vapor oil the limit is 20 mg/m³ (Heisel(A) et al., 1998). Considering the patterns of German automotive industries, for in 1992, the volume of discarded soluble oil discard used in metal-mechanical transformation processes, represented 60% of the total consumption of lubricants. This represents a significant cost which varies from 7,.5% to 17% of the production costs for per piece, superior even to the costs of tools (Kalhöfer, 1997,Heisel Heisel and Lutz, 1998). Even if the costs of tools, in some cases, has have to be increased for the use of minimum lubrication, due to the increase of tool wear increase, nevertheless the total cost of production can be smaller when compared to the conventional process using abundant fluid. Other advantages of the MQL are connected to the maintenance of chips, reduction of reprocess costs, cleaning and conditioning of oil (Granger, 1994).

The objective of this work is to test the MQL technique in the drilling process of aluminum silicon alloy with a solid carbide drill in two different cutting conditions.

Experimental Procedures

Drilling experiments Materials

The material used in the samples of drilling tests was Aluminum-Silicon SAE 323 alloy, with approximately 7% of Silicon. The holes were made in two types of workpiece. The first was a sheet of the alloy, which received most of the holes. After several holes had been done made in this sheet (an the average, 33 holes when the drill diameter was D =10 mm and 17 holes when D = 20 mm) with the objective of wearing the tool, a hole was made in the second type of workpiece (this procedure was repeated, up to the end of the experiment). This type of workpiece had smaller dimensions and was fixed to the dynamometer. During the drilling of this workpiece, feed force and electrical consumed power were measured. After the cutting, diameter and roughness of the holes were measured.

Two kinds of cooling/lubrication systems were used during the experiments. The first one was the mixture of air and oil (Minimum Quantity of Lubricant – MQL), where 10 ml/h of mineral oil (MACRON A – Shell) was pulverised in an air flow of 4.5 bars of pressure. The second kind of system was a flood of soluble oil (4% of oil concentration in water) with a flow rate of 2.4 m³/h.

The tools used were solid K10 carbide drills, without coating (Titex model - A1263), according to DIN 338 Standard, with 10 mm and 20 mm diameters, and the holes drilled had lengths of 34 mm and 60 mm respectively, in order to have similar length/diameter ratios (L/D).

The experiments were carried out in a rigid CNC machining centre with 22 kW of power and maximum rotation of 12,000 rpm.

The minimum mist lubrication equipment used in the tests was the Magic Cut "Mist Coolant Equipment" model OS-21-AT-40, manufactured by Fuso Seiki Co. LTD which allows a fine adjustment of the oil flow, atomizing it in a compressed air flow with constant pressure.

The flank wear measurements were made in an optic microscope with 50 times of amplification.

The surface roughness of the holes was measured in a portable surface roughness equipment. Three measurements of average roughness (Ra) were taken in each hole, at the middle of the hole length and at 120⁰ from each other. The value used in the analysis for each hole is an the average of these three measurements. The diameter and roundness deviation of the holes were measured in an electronic equipment (column type) using a LVDT sensor. Feed force (F_f) of the cut was obtained during the machining of the holes in the workpiece fixed to a KISTLER, 9272 type dynamometer, connected to an A/D board and a PC computer. The sampling rate was 2.56 KHz and 0.4 seconds were monitored each time the sampling was carried out.

The cutting power was obtained indirectly by the monitoring of current and electrical tension through a home made equipment, containing a hall sensor to measure current and a transformer to decrease the value of electrical tension, up to a value possible to be input in the electronic board. In order to obtain the cutting power, an electronic board was built to multiply these two signals. This final signal was input in and A/D board connected to a PC computer with a sampling rate of 50 Hz.This signal was sampled when the drill was cutting the first third part of the hole length.

Table 1 shows the cutting conditions and the number of holes machined in each experiment. The number of holes made in each experiment was chosen because it represents a feed length big enough to verify the influence of tool wear in all the measured parameters.

Each hole was made using the pecking technique, i.e., the tool was removed from the hole three times during drilling, to allow expelling of the chip. The experiments with D = 20 mm and v_c = 450 m/min is referred in this work as condition 1 and the experiments with D = 10 mm and v_c = 300 m/min as condition 2.

Measurement of the Cooling Capacity of the Fluids

The goal of these experiments was to find the cooling capacity of each of the fluids used in the experiments. They were carried out using an electric furnace. The maximum temperature measured in the middle of the aluminum-silicon workpiece (42x42x34mm) was 400° C, obtained after keeping it inside the furnace for a period of 8 minutes. After the heating, the workpieces were submitted to cooling conditions similar to those used in the experiments, i.e.: environment temperature (air), minimnum quantity of lubricant (air + oil) with oil flow of 10ml/h, 30ml/h and 66ml/h and flow of abundant cutting fluid with soluble oil with an oil/water ratio of 1:25.

Using an A/D board and a PC computer with a suitable software, it was possible to monitor the indirect signal of temperature (in milivolts) with a thermosensor type K (Cromel-Alumel) for two minutes (600 points sampled). This thermosensor was connected to the workpiece through a hole that allowed it to reach the center of the workpiece. The hose of fluid was 15 mm distant from the upper part of the workpiece.

Results and Discussion

In a previous work (Braga et al., 1999) the authors demonstrated that the use of MQL did not damage either the quality of the holes, or tool wear when compared to the use of abundant fluid, for the drilling of this aluminum-silicon alloy with v_c = 300 m/min and drill diameter of 10 mm. The intention of this work was to make a similar comparison with higher cutting speed (450 m/min). Due to the fact that the maximum rotation of the machine tool used is 12,000 RPM, it was impossible to use the same drill diameter and, therefore, it was increased to 20 mm. With this, not just the cutting speed (v_c) was changed, but also the depth of cut (a_p). As feed was kept constant, the volume of chip removed per minute increased 3 times, because a_p was increased 2 times and v_c 1.5 times. This intention could not be completely carried out because it was impossible to perform more than 4 holes with abundant fluid. In this condition, the chip stuck to the tip and to the body of the tool in such a way that, if the cut went on, the drill would be broken. To try to explain why the cut can be carried out with MQL but not with abundant cutting fluid, the cooling capacity of these fluids was measured and the results are shown on fig. 1. It can be seen in this figure that MQL has a much lower cooling capacity than that of the flow of abundant cutting fluid and not much higher than that of the environment temperature. Besides, it also shows that the increase of oil flow inside the air does not improve the cooling capacity of the fluid.

With these results, it is possible to explain why the drilling of aluminum-silicon alloy could be done with v_c = 300 m/min and drill diameter D = 10 mm in both cooling/lubrication systems (see Braga, 1999), but could not be done with v_c = 450 m/min and drill diameter D = 20 mm, when abundant fluid was used. To drill aluminum alloys with a carbide drill, the cooling of the tool is not so important, because temperature is not so high and it is lower than what the tool can withstand. Lubrication is important, because, without it, the chip sticks to the tool, jamming its channels, what can cause drill breakage. To perform the drilling with a high volume of chip per time, like when v_c = 450 m/min and drill diameter D = 20 mm were used, a high level of lubrication must be present. Abundant flow of soluble oil just has a good cooling capacity, as shown in fig. 1, but it does not have good lubricant capacity (Shell, 1986). Therefore, it is not possible to carry out the drilling process using this kind of cooling system. On the other hand, MQL does not have a good cooling capacity, which is not necessary in this process, but presents a very good lubrication capacity, since the oil inside the airflow is neat oil. Besides, it is not necessary to have a great amount of oil in the airflow to make the drilling process of this alloy feasible. Preliminary tests demonstrated that it is not possible to carry out the drilling without any kind of fluid nor MQL up to 6 ml/h of oil in the air flow, due to the sticking of the chip on the body of the tool. They also demonstrated that with 10 ml/h of oil, the drilling becomes possible and the increase of oil flow up to 60 ml/h does not increase the performance of the process. Therefore, to perform the drilling with v_c = 450 m/min and drill diameter D = 20 mm, it is necessary either to use MQL or to make a pre-drill to avoid the damaging influences of the drill center and to decrease the volume of chip removed per time in the process. The authors believe that, with less volume of chip removed per time (pre-drilled hole), it will be possible to carry out the drilling with v_c = 450 m/min and abundant flow of oil, since it was possible to do so when v_c = 300 m/min and D = 10 mm (therefore with a volume per minute 3 times smaller than with v_c = 450 m/min and Dd = 20 mm).

After trying without success to carry out the drilling operation with v_c = 450 m/min and drill diameter D = 20 mm with abundant cutting fluid, two experiments were carried out with the same cutting speed and tool diameter, but now using MQL with 10 ml/h of oil, instead of abundant fluid. These two experiments presented results very similar for cutting forces, tool wear and quality of the holes. Therefore, just the results of one of these experiments are going to be shown and be compared to the results obtained when v_c = 300 m/min, D =10 mm and MQL with 10 ml/h of oil was used.

Figure 2 shows cutting power consumed in the drilling process against feed length for the two conditions tested. As already said, condition 1 had 3 times greater volume of chips removed than condition 2 and, at least in the beginning of tool life, cutting power for condition 1 was around 3 times bigger than that obtained for condition 2. Cutting power (Pc) can be expressed by Eq. (1) (Diniz et al. 2000):

where:

Fc = cutting force

Ks = specific cutting force

ƒ= feed per revolution

f is constant and the product "a_p.v_c" is three times greater in condition 1 than in condition 2. The result obtained for cutting power shows that the variation of cutting speed and depth of cut did not influence Ks. It was expected that the increase of cutting speed could have some influence on the friction coefficient,which could decrease Ks, however this did not happen. Another thing that has to be pointed out is, as already cited in the "Experimental Procedures" item, that cutting power was monitored at the beginning of the drilling, before the drill had reached the third part of the holes. The authors believe that, if this measurement were made closer to the end of the drilling, the results would be different, because the cutting power for the holes with D = 20 mm would be influenced by the lack of lubrication caused by the higher length (L) of the holes (for D = 20 mm, L = 60 mm).

It also can be seen in this figure that the slopes of both curves are very similar. Comparing this behavior with the behavior of flank wear against feed length (fig. 3), it can be seen that they are also very similar, i.e., wear also increased linearly with feed length, with very similar slopes for both conditions. Therefore, it can be said that the increase of power followed the growth of flank wear.

It can be seen in fig. 3 that flank wear was greater for condition 1 than for condition 2. This result was expected, since the former condition had a greater cutting speed and, due to the greater drill diameter, the drill in condition 1 had a bigger contact length with the walls of the holes than in condition 2, for the same feed length. Nevertheless it is interesting to note that the slopes of the curves are similar, though the initial tool wear value for condition 1 is greater. It was expected that flank wear for condition 1 grew faster than for condition 2, due to the higher cutting speed and higher contact length between tool and workpiece, which did not happen. It may be connected to the greater length of the cutting edge, which may provide a better distribution of the higher amount of heat generated.

Figure 4 shows the results of feed force against feed length. In the beginning of the experiments (when wear is very small and can be neglected) feed force for condition 1 is about twice that found for condition 2. As the volume of chip removed per time was 3 times greater when condition 1 was used, it can be said that something happened which caused feed force not to increase in the same proportion. It may be connected to the higher amount of chip stuck to the tip of the tool when condition 2 was used, due to the lower cutting speed. However, feed force for condition 1 increased with feed length, while for condition 2, it remained almost constant. The amount of chip stuck on the tip of the tool remained almost constant for condition 2, while for condition 1 it increased as cutting time increased, due to the bigger length of the holes, which harmed the lubrication of the tool tip. This result also demonstrates that feed force is not very influenced by flank wear (which, as already said, presented a similar growth with feed length for both conditions), but it is strongly influenced by the conditions of the tool tip.

Before the analysis of the quality parameters of the holes, it is important to note that neither diameter nor any other quality parameter of the hole was influenced by tool wear. In other words, these parameters presented no tendency as feed length increased.

Diameter close to the entrance of the holes

Table 2 shows the average, maximum and minimum diameters measured in the first third part of the hole length. It can be seen in table 2 that both kinds of holes presented the same standard deviation of average diameter, which means that the greatest diameter (D = 20 mm) presented a better quality. In other words, both holes presented values of diameter inside IT 9 range of tolerance, but the capability of the processes were different, since for D = 10 mm the IT 9 interval is 36 µm, while for D = 20 mm, it is 52 mm. The capability of the process can be given by Eq. (2) (Novaski, 1994):

Therefore, for the process with D = 20 mm, CP = 52/36 = 1.44, while the capability of the process for D = 10 mm is CP = 36/36 = 1. The reason for this result may be connected to the greater rigidity of the drill. Even with the higher cutting force caused by the higher volume of chip removed per time, the bigger drill was more capable of providing a better result in terms of dimensional tolerance of the holes.

Roundness

Figures 5 and 6 show the roundness of the holes close to the entrance and close to the end of them. Some things must be pointed out in these figures:

a) roundness deviation did not change from the beginning to the end of the holes, when for D = 10 mm – this result can be attributed to the lower cutting forces and the shorter diameters. Even with the drill penetrating further into the hole, the forces were not able to deviate the drill more than in the entrance of the hole, and the roundness deviation was kept almost constant.

b) for D = 20 mm, the roundnees deviation was smaller close to the entrance than in the end of the holes – in the beginning of the drilling, because the drill was bigger, its rigidity did not allow the drill to deviate, even with the higher cutting forces generated by the higher volume of chip removed. Therefore, its roundness deviation was smaller than even that found for D = 10 mm. At the end of the holes the cutting forces increased more because, due to the high length of the holes, the oil was not able to reach the cutting zone to lubricate it. Therefore, the drill rigidity was no longer able to keep the drill rotating without more eccentricity and the deviation increased. These high values (for D = 20 mm) at the end of the holes may prevent the use of this process (at least with long length) when high quality of the holes is desired.

Average surface roughness (Ra)

The roughness of the hole wall was greater for condition 1 than for condition 2, as can be seen in fig. 7. This result may be related to the bigger depth of cut used in condition 1, which increased the cutting forces and damaged roughness. On the other hand, the drill was more rigid and able to decrease vibration caused by cutting forces. But the lack of lubrication in the longer holes, may have caused the cutting forces to increase more than the drill rigidity and made the roughness to increase. Roughness values were not just bigger when D = 20 mm, but also more scattered.

Taper

The average taper values and their dispersion were smaller for condition 1 (fig. 8). Moreover, in both conditions the average taper values were positive, i.e., the diameters in the entrance of the holes were bigger than at the end. These bad results found for the holes made with the smaller drill are due to its lower stiffness, which made the diameter in the beginning of the holes to increase. When the tool reached the end of the holes, the diameter decreased, due to the alignment of the tool caused by the hole wall. When D = 20 mm, the hole was not enlarged in the entrance due to the high tool rigidity and its diameter did not increase at the end of the holes, even with the higher forces caused by the lack of lubrication, due to the alignment of the tool caused by the hole wall.

Conclusions

It is important to point out that, among the parameters used to characterize hole quality, just roughness and roundness deviation at the end of the holes presented worse results when condition 1 was used. Therefore, it is possible to drill holes with D = 20 mm and L = 60 mm (L/D ratio = 3) using MQL without pre-drilling them. Moreover, even for roughness and roundness deviation, the results would be better if the holes were shorter.

The second important conclusion of this work is related to the use of cutting fluid in this kind of process. When drilling aluminum silicon alloys it is not so important to have a fluid with cooling effect, because the temperature of the cutting zone is not so high as to damage the tool. What is important is the lubrication effect, which prevents the chip sticking on the tool and makes the cut feasible. Due to the fact that the abundant cutting fluid has a much better cooling effect than MQL, although its lubrication action is not as good as MQL, it was not possible to carry out the process with it when a high volume of chip per time had to be removed. MQL, however, provide good lubrication for the process and allowed it to be carried out.

Paper accepted October, 2002. Technical Editor: José Roberto de F. Arruda

• Braga, D. U., Diniz, A. E., Miranda, G. & Coppini, N. L., 1999, "Minimal Lubrication in the Drilling Process of Aluminum-Silicon Alloy - A356 (SAE 323)", Proceedings of the 3^rd International Machining and Grinding Conference (SME), Vol. 1, Cincinnati, USA, pp. 683-697.
• Braga, D. U., Diniz, A. E., Miranda, G. & Coppini, N. L., 2002, "Using a minimum quantity of lubricant (MQL) and a diamond coated tool in the drilling of aluminum-silicon alloys", Journal of Materials Processing Technology, Vol 122, pp. 127-138.
• Chiesa, F., Fuoco, R., Santarini, M., 1995, "Fundição de ligas de alumínio usadas em componentes para a indústria automobilística". Máquinas e Metais, Vol.357, pp. 126 137.
• Coelho, R. T., Yamada, S., Aspinwall, D. K., Wise, M. L. H., 1995, "The application of polycrystalline diamond (PCD) tool materials when drilling and reaming aluminium based alloys including MMC", International Journal of Machine Tools and Manufacture, Vol. 35, No. 5, pp. 761-774.
• Derflinger, V., Brändle, H. and Zimmermann, H., 1996, "New hard/lubricant coating for dry machining", Surface Coating Technology, Vol. 113, pp. 286-292.
• Diniz, A. E., Marcondes, F. C. & Coppini, N. L., 2000, "Tecnologia da Usinagem dos Materiais", Artliber Editora, São Paulo, Brazil, 242 p.
• Granger, C., 1994, "Dry machining's double benefit", Machinery and Production Engineering, Vol.152, No. 3873, pp.14-20.
• Heisel, U., Lutz, M., Spath, D., Wassmer, R., Walter, U., 1998, "A técnica da quantidade mínima de fluido e sua aplicação nos processos de corte", Máquinas e Metais, Vol. 385, pp.22-38.
• Heisel, U., Lutz, M., 1998, "Pesquisa de fluido de refrigeração e de lubrificação", Máquinas e Metais, Vol. 388, pp.40-49.
• Kalhöfer, E., 1997, "Dry machining - principles and applications", Proceedings of the "2° Seminário International de Alta Tecnologia", Vol. 1, Santa Barbara d'Oeste, Brazil, pp. 53-62
• Klocke, F., Eisenblätter, G., 1997, "Presented at the Opening Session of Dry Cutting", Annals of the CIRP, Vol.46, No.2, pp.519-526.
• Lugscheider, E., Knotek, O, Barimani, C., Leyendecker, T., Lemmer, O., Wenke, R., 1997, "Investigations on hard coated reamers in different lubricant free cutting operations", Surface and coatings technology, Vol. 90, pp. 172-177.
• Machado, A. R., Wallbank, J., 1997, "The effect of extremely low lubricant volumes in machining", Wear, Vol. 210, No.12, pp. 76 82.
• Novaski, O., 1994, "Introdução à Engenharia de Fabricação Mecânica", Editora Edgard Blücher Ltda., São Paulo, Brazil, 119 p.
• Shell, S. A., 1986, "Fluidos de Corte: Curso de Lubrificação Industrial", São Paulo, Brazil.
• Tönshoff, H. K., Spintig, W., 1994, "Machining of holes developments in drilling technology", Annals of the CIRP, Vol. 43, No. 2, pp. 551 561.

Publication Dates

• Publication in this collection
  18 Mar 2004
• Date of issue
  Mar 2003

History

• Received
  Oct 2002