Professional Documents
Culture Documents
Experimental and Theoretical Study of Workpiece Temperature When End Milling Hardened Steels Using (TiAl) N-Coated and PcBN-tipped Tools
Experimental and Theoretical Study of Workpiece Temperature When End Milling Hardened Steels Using (TiAl) N-Coated and PcBN-tipped Tools
Lincoln Cardoso Brandão a , Reginaldo Teixeira Coelho b,∗ , Alessandro Roger Rodrigues c
a Federal University of São João del Rei, UFSJ, Praça Frei Orlando, 170 Centro, Brazil
b School of Engineering at São Carlos, EESC/USP, Avenida Trabalhador São-carlense, 400 Centro, Brazil
c Faculty of Engineering at Ilha Solteira-FEIS/UNESP, Av. Brasil Centro 56, Brazil
a r t i c l e i n f o a b s t r a c t
Article history: The present work shows an experimental and theoretical study on heat flow when end
Received 4 April 2007 milling, at high-speed, hardened steels applied to moulds and dies. AISI H13 and AISI D2
Received in revised form steels were machined with two types of ball nose end mills: coated with (TiAl)N and tipped
19 July 2007 with PcBN. The workpiece geometry was designed to simulate tool–workpiece interaction
Accepted 26 July 2007 in real situations found in mould industries, in which complex surfaces and thin walls
are commonly machined. The compressed and cold air cooling systems were compared to
dry machining. Results indicated a relatively small temperature variation, with higher range
Keywords: when machining AISI D2 with PcBN-tipped end mill. All cooling systems used demonstrated
Temperature measurements good capacity to remove heat from the machined surface, especially the cold air. Compressed
Heat conduction and convection air was the most indicated to keep workpiece at relatively stable temperature. A theoreti-
High-speed cutting cal model was also proposed to estimate the energy transferred to the workpiece (Q) and
Molds and dies the average convection coefficient (h̄) for the cooling systems used. The model used a FEM
simulation and a steepest decent method to find the best values for both variables.
© 2007 Elsevier B.V. All rights reserved.
∗
Corresponding author.
E-mail addresses: lincoln@sc.usp.br (L.C. Brandão), rtcoelho@sc.usp.br (R.T. Coelho), roger@dem.feis.unesp.br (A.R. Rodrigues).
0924-0136/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmatprotec.2007.07.049
j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 1 9 9 ( 2 0 0 8 ) 234–244 235
Table 2 – Geometry of the solid carbide ball nose for Fr2 and Fr3
Code Dc (mm) R (mm) Lc (mm) X1 (mm) X2 (mm) d1 (mm) 0 (◦ ) ˛0 (◦ ) zn (adm) Coating ISO code
Fig. 2 – Details of the ball noses Fr2 (F1827R.Z16.Z2.08 WXK15) and Fr3 (F1827R.Z16.Z2.08 WCB80). (a) Geometrical details:
for dimensions see Table 2; (b) (TiAl)N-coated Fr2; (c) PcBN-tipped Fr3.
Fig. 3 – Workpiece with its fixing holes: (a) general view of the dynamometer with workpiece on the top and (b) top view of
dynamometer without the wokpiece (details of the thermocouples and their fixing system).
j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 1 9 9 ( 2 0 0 8 ) 234–244 237
In order to find the best estimated values for Q and h̄, which
minimize the average error function, Eravg , the steepest decent
method was employed. It consists basically, in estimate a first
point Pi for the function and move to point Pi+1 in the direction
of the local downhill gradient −Eravg (Pi ) (Flannery et al., 1990).
This method was adopted due to its simplicity and it leads
to a local minimum, which seemed good enough since the
energy, for example, has its limit, between zero and the cut-
ting power. The convection coefficient has also some known
limits (Incropera and Dewitt, 1990; Özisik, 1980; Bejan, 1996).
To find the gradients, in both directions (Q and h̄) steps were
established and four new points were simulated around the
first guessed one. Each of their error function was calculated
and the steepest gradient followed until the error valley was
Fig. 8 – Detail of the load steps in the elements. found. At this point the best combination of Q and h̄ had been
found for each experiment set up. The step for the energy was
selected as 0.5 J and for the convection coefficient depends on
the final temperature distribution is obtained and the FEM the cooling system. The whole minimum search procedure
curves of temperature–time are plotted for those elements at was implemented using program developed using MatLabTM
the thermocouple positions, given by software. The stop criterion was less than 0.1 for Eravg , or 5
steps in any direction.
i i
TFEM = FEM (t) (1)
Table 4 – Maximum and minimum temperatures peaks for the AISI H13 and AISI D2 when dry cutting
Steel Cooling system Tool code Temperature peaks (◦ C) Range (◦ C)
T0 T1 T2 T3 T4 TMIN TMAX
AISI H13 Dry Fr1 33.2 32.1 33.2 34.6 36.5 32.1 36.5
Fr2 38.0 40.0 41.0 43.4 48.9 38.0 48.9
Fr3 29.7 31.6 31.5 32.3 33.0 29.7 33.0
AISI D2 Dry Fr1 34.7 35.3 35.9 37.5 43.5 34.7 43.5
Fr2 39.4 36.2 37.6 39.8 48.5 36.2 48.5
Fr3 37.1 37.2 36.2 36.5 42.3 36.2 42.3
j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 1 9 9 ( 2 0 0 8 ) 234–244 239
time, which left much less time for heat propagation into the
workpiece. That may be closer to an adiabatic cutting.
As one can notice, the higher measured values with dry
cutting were around 48 ◦ C, which is relatively low compared to
values found for dry cutting elsewhere (Toh, 2005). However,
in the present work thermocouples were used embedded in
small holes at 2.0 mm below the milling surface. There is a
spreading of heat throughout the workpiece and a high gra-
dient before it is detected by the thermocouples. Those can
be the causes of comparatively low measured values, but at
that position there was a better indication of the whole body
temperature, which was used to estimate absorbed energy.
Fig. 10 shows a typical graph when cutting H13 using com-
pressed air as a cooling system.
In general, it can be observed a small progressive increase
on the temperature peaks but with lower values, compared to
dry cutting. That demonstrates a more efficient heat removal
system, when using compressed air. Table 5 shows all the
Fig. 9 – Temperature at 2.0 mm below surface when cutting
results for the tests with compressed air.
AISI H13 using Fr1 with vc = 200 m min−1 ,
Comparing the temperature peaks amongst the different
fz = 0.1 mm tooth−1 , ap = 0.20 mm, ae = 0.25 mm, dry cutting.
tools used, the same tendency observed in the dry cutting
conditions can be observed, i.e., Fr1 showed the lowest tem-
perature values, followed by Fr3 and Fr2. In general, values for
D2 were also slightly higher, following the hardness.
The cold air system caused a significant decrease on work-
piece temperature demonstrating an excellent capacity to
remove heat. Even so, there was a peak when cutting edge
ran over the thermocouple position and the central thermo-
couple, number 2, shown the lowest peak. Table 6 shows all
the results for the tests with cold air.
Fig. 11 shows a typical graph when cutting H13 using cold
air as a cooling system.
Comparisons amongst the tools and workpiece materials
showed the same tendency as those previously discussed.
Overall, the experiments generated higher peaks of temper-
ature for the dry cutting, followed by compressed air and cold
air. Additionally, the cold air system managed to low the work-
piece temperature in about 2 to 3 ◦ C. That behavior can result
in a dimensional effect as bad as the dry cutting, but in oppo-
site sense.
Fig. 10 – Temperature at 2.0 mm below surface when Fig. 12 shows a typical graph of temperature versus time for
cutting AISI H13 using Fr2 with vc = 200 m min−1 , dry cutting, obtained from the FEM simulation.
fz = 0.1 mm tooth−1 , ap = 0.20 mm, ae = 0.25 mm, compressed The original curve presents an oscillation, since the heat
air as cooling system. source (chip formation) runs up and down on the curved work-
Table 5 – Maximum and minimum temperatures peak for the AISI H13 and AISI D2 using compressed air
Steel Cooling system Tool code Temperature peaks (◦ C) Range (◦ C)
T0 T1 T2 T3 T4 TMIN TMAX
AISI H13 Compressed air Fr1 27.3 27.8 27.4 27.8 29.4 27.3 29.4
Fr2 30.4 31.5 31.3 32.5 32.9 30.4 32.9
Fr3 27.2 28.6 28.0 28.2 28.4 27.2 28.6
AISI D2 Compressed air Fr1 31.7 30.8 30.7 32.5 33.7 30.8 33.7
Fr2 31.6 31.8 30.6 32.4 33.6 30.1 33.6
Fr3 28.3 29.0 28.6 29.2 28.9 28.3 29.1
240 j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 1 9 9 ( 2 0 0 8 ) 234–244
Table 6 – A maximum and minimum temperature peaks for the AISI H13 and AISI D2 using cold air
Steel Cooling system Tool code Thermocouples temperature (◦ C) Temperature (◦ C)
T0 T1 T0 T1 T0 TMIN TMAX
AISI H13 Cold air Fr1 24.2 24.0 23.9 24.0 24.8 23.9 24.8
Fr2 27.7 27.4 27.4 28.2 30.8 27.4 30.8
Fr3 25.2 24.7 24.0 24.2 25.2 24.0 25.2
AISI D2 Cold air Fr1 26.7 24.8 24.5 24.0 24.5 24.0 26.7
Fr2 26.2 25.7 25.8 26.3 28.8 25.7 28.8
Fr3 25.9 28.9 29.1 29.7 29.8 25.9 29.8
Table 7 – Summary of the average quadratic error, Eravg , for all tools and workpiece materials cut in dry conditions
piece surface and also the elements had a finite mass and error. In some of these cases, lower values could be found. That
volume. That oscillation was also observed on the experimen- was necessary since the algorithm only finds local minimums
tal curve, but with lower intensity, since there is some time and another minimum could be in the neighborhood.
response delay from the thermocouple. In the FEM simulation Considering that the errors were an average quadratic for
the oscillation was much more evident because thermocou- all the temperature points, each in highlighted value in Table 8
ple time response was not modeled. Therefore, the original would represent the squared temperature error for a single
graphs were filtered using a median filter function of LabView temperature measurement. Therefore, the theoretical curves
program with 50 points, similarly to that applied on the exper- can be considered quite precise in predicting the energy and
imental curves. After the filtering, the graphs obtained from convection coefficient for end milling H13. Table 8 shows the
FEM were confronted with the experimental ones and submit-
ted to the minimum error search, according to Eq. (4), aiming
at the best values for Q and h̄. Table 7 shows a summary of the Table 8 – Estimated values for Q and h̄ when dry
results of average error, highlighting the minimum found. machining
The shadowed error values were the initial guess and then
Workpieces AISI H13 AISI D2
the steepest descent method took the way down to the low-
est value, keeping constant the selected steps. In all those Tools Fr1 Fr2 Fr3 Fr1 Fr2 Fr3
cases the error criterion was met before reaching the maxi- Q (J) 2.5 4.5 3.0 3.5 4.0 3.5
mum number of steps. When the criterion was met at the first h̄ (W m−2 ◦ C−1 ) 750 500 1000 500 500 500
selected point, the program was run again to search for a lower
j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 1 9 9 ( 2 0 0 8 ) 234–244 241
Table 9 – Summary of the average quadratic error, Eravg , for all tools and workpiece materials cut with compressed air
Fig. 11 – Temperature at 2.0 mm below surface when Fig. 12 – Typical curve of temperature versus time obtained
cutting AISI H13 using Fr3 with vc = 400 m min−1 , from FEM simulation, when dry cutting.
fz = 0.1 mm tooth−1 , ap = 0.20 mm, ae = 0.25 mm, cold air as
cooling system.
Table 11 – Summary of the average quadratic error, Eravg , for all tools and workpiece materials cut using cold air
j o u r n a l o f m a t e r i a l s p r o c e s s i n g t e c h n o l o g y 1 9 9 ( 2 0 0 8 ) 234–244 243
Table 12 – Estimated values for Q and h̄ when machining using cold air
Workpieces AISI H13 AISI D2
4. Conclusions
When end milling hardened AISI H13 and AISI D2 steels using
solid carbide (TiAl)N-coated and PcBN-tipped tools, the follow-
ing conclusions can be drawn:
references Incropera, F.P., Dewitt, D.P., 1990. Fundamentals of Heat and Mass
Transfer, 3rd ed. Wiley, New York, 919 p.
Lazoglu, I., Altintas, Y., 2002. Prediction of tool and chip
temperature in continuous and interrupted machining. Int. J.
Abukhshim, N.A., Mativenga, P.T., Sheikh, M.A., 2006. Heat Mach. Tools Manufacture 42, 1011–1022.
generation and temperature prediction in metal cutting: a Ming, C., 2003. Experimental research on the dynamic
review and implications for high speed machining. Int. J. characteristics of the cutting temperature in the process of
Mach. Tools Manufacture 46 (n7–8), 782–800. high-speed cutting. J. Mater. Process. Technol. 6675, 1–4.
Bejan, A., 1996. In: Edgard Blücher (Ed.), Transferência de calor, Özisik, N.M., 1980. Heat Conduction, 2nd ed. John Wiley & Sons,
2nd ed. John Wiley Sons Inc., p. 477. pp. 572.
Dewes, R.C., Ng, E., Chua, K.S., Newton, P.G., Aspiwall, D.K., 1999. Paul, S., Dhar, N.R., Chattopadhyay, A.B., 2001. Beneficial effects
Temperature measurement when high speed machining of cryogenic cooling over dry and wet machining on tool wear
hardened mould/die steel. J. Mater. Process. Technol. 92/93, and surface finish in turning AISI 1060 steel. J. Mater. Process.
293–301. Technol. 116, 44–48.
Flannery, W.H., Teukolsky, B.P.S.A., Vetterling, W.T., 1990. Toh, C.K., 2005. Comparison of chip surface temperature between
Numerical Recipes in Pascal—The Art of Scientific Computing. up and down milling orientations in high speed rough milling
Cambridge University Press, Cambridge, pp. 735. of hardened steel. J. Mater. Process. Technol. n167, 110–118.
Fröhlingsdorf, W., Unger, H., 1999. Numerical Investigations of Yvonnet, J., Umbrello, D., Chinesta, F., Micari, F., 2006. A simple
the compressible flow and the energy separation in the inverse procedure to determine heat flux on the tool in
Ranque-Hilsch vortex tube. Int. J. Heat Mass Transfer. orthogonal cutting. Int. J. Mach. Tools Manufacture 46 (n7–8),
Pergamon 42, 415–422. 820–827.