Cutting speed and feed based model for evaluating parallelism deviations in horizontal dry turning of aluminium alloys.

Salguero, Jorge ; Sanchez, Jose Miguel ; Sebastian, Miguel Angel 等

1. INTRODUCTION

Aluminium alloys are widely used in aerospace industry due to their excellent ratio weight/cost/mechanical properties. Thus, airship structural elements are mainly built by using aluminium alloys. These alloys need to be processed by machining operations because of assembly necessities, and a high quality finishing level are required for each element of an airship, particularly for the machined aluminium alloy pieces (Carrilero et al., 2002; Nouari et al., 2003). Thus, dimensional and geometrical design specifications must be rigorously followed. Commonly, dimensional precision has been studied in the cutting process of aluminium alloys. In the same way microgeometrical deviations have been also studied, mainly based on roughness analysis. However, macrogeometrical deviations, such as parallelism, are not ever taken into account in those studies.

On the other hand, cleaner technologies must be applied in order to reach higher levels of manufacturing sustainability. So, the employment of hazardous, toxic and high environmental impact cutting fluids must be minimised. In this way, cutting using minimum quantity of lubricant (MQL) or, better, in absence of it have been promoted as environmentally friendly alternatives to classical machining processes.

In this work, parallelism deviations (PD) of horizontally dry turned UNS A97050 (Al-Zn) and UNS A92024 (Al-Cu) cylindrical bars have been studied for different cutting speeds (v) and feeds (f) applied. The obtained results have allowed establishing v and f based exponential parametric model for the parallelism deviation as a function of those cutting parameters for both alloys.

2. EXPERIMENTAL

The workpieces used in the experiments were cylindrical bars (150 to 200 mm long with diameters between 80 and 120 mm) of both alloys. These samples were horizontally dry turned in a CNC Lathe. Cutting speeds from 43 up to 170 m/min, and feeds from 0.05 up to 0.3 mm/rev were applied (Tab.1), with a cutting depth maintained at 1 mm. The tools employed were TiN covered WC-Co turning inserts.

Parallelism Deviation (PD) was measured with an experimental device attached to the CNC Lathe, consisting in a comparer clock magnetically placed in the tool revolver, in order to draw four horizontal lines separated 90[degrees]. The points recording for the PD measurement was achieved in these lines. PD points were acquired each 5 mm in each line. Points were mathematically treated in order to obtain average values for the four lines.

Finally, PD value was defined as the difference between the maximum and minimum values of the recorded points. Parallelism deviation limits is considered defined by two lines which are parallel to rotation axis and which contains those extreme points.

3. RESULTS AND DISCUSSION

When a bar of aluminium alloy is dry-turned using TiN tools, a near to pure aluminium coating is developed in the rake face of the insert (Built-Up Layer, BUL). On the other hand, adhered alloy material can be distinguished in tool edge (Built-Up Edge, BUE) (Sanchez et al., 2005). Compositional characteristics of BUL and BUE are different because their formation mechanism are also distinct. BUL and BUE development influences the surface finish quality of the turned workpiece (Carrilero et al., 2002; Sanchez et al., 2005).

Figure 1 plots the evolution of the average values of PD as a function of cutting speed for the different feeds applied for each alloy.

As it can be observed, there is a similar trend in the PD(v) for each feed applied. In effect, a soft decrease of PD with the cutting speed can be appreciated. According to that concluded in previous works, it can be caused by the evolution of the changes in the tool geometry as a consequence of the Built-Up Edge and Built-Up Layer formation (Marcos et al., 2005; Sebastian et al., 2002).

[FIGURE 1 OMITTED]

As it can be observed in Figure 2, higher values of PD are obtained for lowest speeds and highest feeds. This can be explained in terms of the chip arrangement, that is also related to the surface finish of the machined samples. In the cutting speeds and feeds range applied, a high variety of chips geometry and length has been obtained, showing that more continuous and larger chip is formed when cutting speed increases and feed decreases (Rubio et al., 2006). However no convergence can be observed but, moreover, a change in the tendencies can be appreciated starting form a cutting speed of 85 m/min. This is in good agreement with the influence of the BUL and BUE formation in the finish quality observed in previous works (Carrilero et al., 2002; Sebastian et al., 2002).

Classical parametric models for establishing relationships between finishing quality variables (Ra, Rz) and cutting parameters (v,f,p) are usually potential models (Sebastian et al., 2002; Chan et al, 2001; Abouelatta and Madl, 2001; Sanchez et al., 2006). However, in this case, very bad results were obtained applying potential models. Thus, alternative models were tested. The best behaviour was reached for an exponential model such as:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

For determining the values of the coefficients, a linear regression procedure has been followed over the linear form of equation (1), written in a logarithmic scale.

ln (PD) = ln C + B = ln C + [2.summation over (i=1)] [2.summation over (j=1)] [K.sub.ij] x [a.sup.i] x [v.sup.yj] (2)

Multilinear regressions for each alloy reported the values included in Table 2.

From these values, the PD(v,f) parametric model can be constructed by substituting them in equation (1) for each alloy:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] ( 3)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)

In equations (3) and (4) B coefficients are:

[B.sub.7050] = -0.22 x a x [v.sup.0.60] + 1.01 x [a.sup.2] x [v.sup.0.60] - 0.01 x [a.sup.2] x [v.sup.1.20] (5)

[B.sub.2024] = 14.38 x a x [v.sup.-0.08] - 21.26 x [v.sup.-0.16] - 119.88 x [a.sup.2] x [v.sup.-0.08] + 174.13 x [a.sup.2] x [v.sup.-0.16] (6)

Simulated PD(f,v) values showed a good approximation (higher that 95%) to those experimentally measured for v and f values included into the range studied.

[FIGURE 2 OMITTED]

4. CONCLUSIONS

Parallelism deviation (PD) has been studied as a function of cutting speed for each feed applied as well as a function of feed for each cutting speed applied.

Results obtained have shown that higher values of PD are obtained, in both alloys, for lowest speeds and highest feeds.

These results can be related with both chip arrangement and adhered material forms in either tool edge or tool rake face.

Marginal PD(v) and PD(f) relationships allow considering the establishment of a PD(v,f) parametric model.

The best adjustment has been reached for an exponential model as that included in equation (1), whose results showed a good correlation between experimental and simulated values of the parallelism deviation.

5. REFERENCES

Abouelatta, O.B. & Madl, J. (2001). Surface roughness prediction based on cutting parameters and tool vibrations in turning operations. J. Mater. Proc. Tech., 118, 269-277, 0924-0136

Carrilero, M.S.; Bienvenido, R.; Sanchez, J.M.; Alvarez, M.; Gonzalez, A. & Marcos, M. (2002). A SEM and EDS insight into the BUL and BUE differences in the turning processes of AA2024 Al-Cu alloy. Int. J. Mach. Tools & Manuf., 42, 215-220, 0890-6955.

Chan, K.C.; Cheung, C.F.; Ramesh, M.V.; Lee, W.B. & To, S. (2001). A theoretical and experimental investigation of surface generation in diamond turning of an Al6061/SiCp metal matrix composite. Int. J. Mech. Sci., 43, 2047-2068, 0020-7403

Marcos, M.; Sebastian-Perez, M.A.; Contreras-Samper, J.P.; Sanchez-Carrilero, M.; Sanchez-Lopez, M. & Sanchez-Sola, J.M. (2005). Study of roundness on cylindrical bars turned of aluminium-copper alloys UNS A92024. J. Mater. Proc. Tech, 162-163, 644-648, 0924-0136

Nouari, M.; List ,G.; Girot, F. & Coupard, D. (2003). Experimental analysis and optimisation of tool wear in dry machining of aluminium alloys. Wear, 255, 7-12, 1359-1368, 0043-1648

Rubio, E.M.; Camacho, A.M.; Sanchez-Sola, J.M. & Marcos, M. (2005). Surface roughness of AA7050 alloy turned bars. Analysis of the influence of the length of machining. J. Mater. Proc. Tech, 162-163, 682-689, 0924-0136

Rubio, E.; Camacho, A.M.; Sanchez-Sola, J. M. & Marcos, M. (2006). Chip arrangement in the dry cutting of aluminium alloys. J. Achiev. Mat. Manuf. Eng., 16, 1-2, 164-170. 1734-8412

Sanchez, J.M.; Rubio, E.; Alvarez, M.; Sebastian, M.A. & Marcos, M. (2005). Microstructural characterisation of material adhered over cutting tool in the dry machining of aerospace aluminium alloys. J. Mater. Proc. Tech, 162-163, 911-918, 0924-0136

Sanchez, J.M.; Sebastian, M.A.; Rubio, E.M.; Sanchez Carrilero, M.; Sevilla, L. & Marcos, M. (2006). A Parametric Model for the Straightness Deviation in the Cutting Processes of Aluminum Alloys. Mater. Sci. Forum, 526, 31-36, 0255-5476

Sebastian, M.A.; Sanchez, J.M.; Carrilero, M.S.; Gonzalez, J.M. & Marcos, M. (2002). Parametric model for predicting surface finish of machined AA2024 alloy bars. Int. J. Manuf Sci. & Prod, 4, 181-188, 0793-6648

SALGUERO, J[orge]; SANCHEZ, J[ose] M[iguel]; SEBASTIAN, M[iguel] A[ngel]; SANCHEZ--CARRILERO, M[anuel] & MARCOS, M[ariano] *

* Supervisor, Mentor

Tab. 1. Cutting conditions

v (m/min) 43 65 85 125 170
f (mm/rev) 0.05 0.10 0.2 0.3

Tab. 2. Multilinear regressions values for each alloy.

 C y [K.sub.11] [K.sub.12] K21 K22

UNS A92024 48.40 -0.08 14.38 -21.26 -119.9 174.13
UNS A97050 62.10 0.60 -0.22 0.00 1.01 -0.01