Machinability of Titanium

Metal Matrix Composites

Roland Bejjani*, Bin Shi**, Helmi Attia**, Marek Balazinski*, Hossam Kishawy***

*École Polytechnique de Montréal, Québec, Canada H3T 1J4

**Aerospace Manufacturing Technology Centre (AMTC), Institute for Aerospace

Research, National Research Council Canada, Montreal, Québec, Canada

***University of Ontario Institute of Technology, Ottawa, Ontario, Canada

[e-mail]

Abstract: Titanium Metal Matrix Composite (Ti-MMC) is a new class of material,
which is widely used in several industrial applications. Although the ceramic particles
in the titanium matrix improve its wear resistance, they also cause high abrasive wear of
the cutting tool during machining, representing a serious challenge for this material to
be widely used by industry. Since abrasion is the main mode of tool failure when cutting
MMC, tool wear progression can be problematic, as the tool life will be relatively short.
In this paper, machining tests were conducted to assess the machinability aspect of Ti-
MMC, using Polycrystalline Diamond (PCD) tools, with a focus on the effect of the
cutting parameters on tool wear, surface roughness, cutting forces, and chip formation
mechanism; segmentation and morphology. The aim of the present work is to define the
optimal cutting conditions and to provide practical recommendations for finish turning
this class of material.
Keywords: Titanium Metal Matrix Composites, Machinability, Chip microstructure

1. INTRODUCTION

Metal matrix composites (MMC) are a new class of material that consist of a non-
metallic phase distributed in a metallic matrix with superior properties to those of the
constituents. The main advantages of MMC over other alloys are their higher wear and
fatigue resistance, higher rigidity, higher deformation resistance, lower thermal
expansion and better thermal shock resistance [Muthukrishnan, N. et al., 2008]. The
principal use of these materials is for lightweight load-bearing applications on account
of their enhanced mechanical and physical properties [Bell, J.A.E. et al., 1997]. Most of
these composites consist of an aluminum matrix reinforced with hard ceramic particles,
such as silicon carbide, alumina or boron carbide [Songméné, V., et al., 1999]. Titanium
MMCs have recently gained new grounds in industrial applications due to their higher
tensile strength at room and elevated temperatures, better creep strength, good fracture
toughness, excellent wear resistance, and other benefits that are typical of titanium
alloys, such as lightweight and corrosion resistance [Abkowitz, S.M., 2009].

The machinability of MMCs is generally poor due to the abrasive action of the hard

ceramic particles that significantly reduce the tool life during machining, resulting in
inconsistent part quality, increased manufacturing cost and reduced productivity. The

process of abrasion while cutting MMC is somewhat different from that of
homogeneous metal cutting. In fact, due to the high abrasive action of the hard particles
within the material itself, both two-body and three-body wear mechanisms have
generally been observed. In the two-body wear, the hard reinforcements penetrate into
the cutting tool surface and scrape out the material, while the trapped hard
reinforcements may roll and slide on the tool wear surface in the three-body wear
[Kannan et al., 2006]. Many researchers found that flank wear is the dominant form of
tool failure during cutting MMCs. It was reported that the progression of tool wear is
extremely fast and leads to very short tool life, sometimes of the order of only few
seconds at high cutting speeds [Songméné et al., 1999, Zhu et al., 2004, Tomac et al.,
1992]. Tomac and Tonnessen [Tomac et al., 1992] compared the performance of
Polycrystalline Diamond (PCD) inserts to that of TiN, TiC/N, and Al

2

O

3

coated tools.

PCD tools were shown to provide good overall performance due to their high hardness
and extended tool life. However, because of their high cost, cemented carbides and
ceramics were also used to machine these materials [Tomac et al., 1992].

Critical literature review indicated that limited machinability data are available for

Ti-MMC. To bridge this gap, this research work aimed at providing the needed
knowledge about the machinability of Ti-6Al-4V alloy MMC. The machining tests were
limited to finishing operations, where the MMC material is made to near net shape,
using PCD tool inserts.

2. EXPERIMENTAL SET UP AND PROCEDURE

The principal machining parameters that control machinability are intrinsic parameters
(cutting speed, feed rate, depth of cut, type of tool), and extrinsic parameters (particulate
size, volume fraction, type of reinforcement) [Basavarajappa, S. et al., 2006]. In the
current experimental work, tests were carried out under dry conditions, in alignment
with the direction taken by industry to move towards environmentally benign
manufacturing. The selection of the cutting tool material was based on previous data for
machining Aluminum-MMC and tool manufacturers’ recommendations. Machining
tests were conducted on a 6-axis Boehringer NG 200, CNC turning center. Cutting
forces were measured using a three-component Kistler dynamometer, type 9121. The
roughness of the generated surface was examined after each machining test using Form
Talysurf Series S4C. High speed camera, model AOS X-Motion, was used to observe
the chip formation process under low and high cutting speeds. Chips were collected and
examined after metallographic preparation.

The tool wear progression has been established using an Olympus SZ-X12

stereoscopic microscope. A specially designed and developed quick stop device (QSD)
was used to examine the chip formation process and the chip segmentation
phenomenon.

The turning tests were carried out using 2.5” diameter cylindrical workpieces made

of CermeTi

®

MMCs, which consist of a non-metallic phase TiC (10-12% by weight)

distributed in a matrix of Ti-6Al-4V titanium alloy. This material was manufactured by
Dynamet Technology Inc. Table 1 shows some typical properties of TiMMC. The tool
insert used in this investigation was 0.5” round Polycrystalline Diamond (PCD)
supplied by KY-Diamond.

The Taguchi’s Parameter Design technique was utilized to provide an effective tool

for establishing the optimum conditions of the turning process of Ti-MMC, with a
relatively small number of experimental runs. This approach considers multiple input
parameters for a given response and uses the non-linearity of a response parameter to
decrease the sensitivity of the quality characteristic to variability. In optimizing the
surface roughness, ‘the smaller the better’ criterion was used and the signal-to-noise
ratio (S/N) was plotted for each control variable [Ross, Phillip J., 1988]. The S/N ratio
represents the magnitude of the mean of a process compared to its variation. The larger
the S/N, the more robust is the measured variable against noise. An L8: 4*2-Taguchi
array was used with 4 speeds and 2 feeds. The speed ranged from 60 to 230 m/min, the
feed was set at 0.1 and 0.2 mm/rev, and the depth of cut was kept constant at 0.15 mm.
The total length of cut was 40 mm, for each test.

Table 1; Properties of Cermet Ti-C; Ti-6Al-4V, 10-12% TiC.

Density

Yield

Strength

Tensile

Strength

Elastic

Modulus

Shear

Modulus

Thermal

Conductivity

Specific

Heat

4.5 g/cc

1014

MPa

1082

MPa

135 GPa

51.7 MPa

5.8 W/m/ºK

610 J/kg/ºK

Figure 1; The experimental setup for finished turning of Ti-MMC material using

PCD insert.

3. RESULTS AND DISCUSSION

3.1. Surface

roughness

Titanium alloys and TiMMC are generally used for parts that require high mechanical
performance. Therefore, surface roughness must be well controlled during the
machining process. In theory, the two main factors that affect surface roughness are the
tool geometry (nose radius) and feed. This can be shown by the following formula
[Shaw M.C., 2005]:

Ra=31.25

f

2

/r

Where Ra is the arithmetic average surface roughness (in

μm), f is the feed (in

mm/rev), and r is the nose radius (in mm). However, in real life situation and for
TiMMC, the cutting speed has also an effect on Ra. Figure 2(a) shows the effect of the
cutting speed (v) and the feed (f) on the surface roughness signal-to-noise ratio (S/N). It
should be noted that the higher the S/N ratio, the better the surface quality is (or the
lower the Ra value). Figure 2(a) shows that for speeds below 100 m/min and above 180
m/min, the slope of the S/N ratio is much steeper than that of the feed. This indicates
that the speed has a much higher effect on Ra, with a peak at 180 m/min. As expected,
Ra is better at the lower feed. Similar observation on the dominant effect of the cutting
speed on surface roughness has been reported for aluminium MMC [Kilickap, E. et al.,
2005]. Figure 2(b) shows that as the speed increase up to 100 m/min, the surface
roughness is reduced, mainly due to the increase in the temperature at the tool-chip
interface. This causes the softening of the workpiece material and the reduction in the
cutting forces. At cutting speeds higher than 180 m/min, a reversed effect was observed;
high speeds produce rougher surface. This can be attributed to the accelerated tool wear,
which results in the deterioration of the part’s surface quality. Another factor that
contributes to this poor surface quality is the excessive increase in the cutting
temperature, which causes the sparking of small particles that separate from the
machined surface, as will be shown in Figure 9. In addition, at higher cutting speeds, the
chips show more acute segmentations, which cause fluctuation in the cutting force
[Yang X. et al., 1999]. It was also observed that at a cutting speed of 230 m/min, the
same Ra was obtained for both feeds. This suggests that at higher cutting speeds, a

Figure 2; Surface roughness for PCD tool. (a); S/N ratio effect. (b); Speed effect.

(a)

(b)

Figure 3; Surface roughness profile. Cutting parameters:

speed =170m/min and feed= 0.2mm/rev.

different cutting mechanism is encountered. Furthermore, it is evident that an
interaction exists between the two variables, v and f, since the Ra lines are not parallel.
It is also worth reporting that at v=230 m/min and f=0.1 mm/rev, the chips caught fire
and the test could not be continued to the full cutting length. A sample of surface profile
of the machined surface is presented in Figure 3 for v= 170 m/min and f=0.2 mm/rev.

3.2.

Cutting

force

Figure 4 shows the effect of the cutting speed and feed on the cutting force F

c

. As

expected, the cutting force is reduced with the increase of the cutting speed up to 180
m/min. Above 180 m/min, the force level is stabilized and remains unchanged. Figure
4(b) shows that while the cutting force F

c

stabilizes quickly after the start of the turning

process, the thrust force F

t

reaches a steady value after much longer length of cut.

Figure 4; (a) Mean cutting force for PCD tool. (b) Forces profile. F

t

: thrust force, F

f

:

feed force, F

c

: cutting force (at v=180 m/min and f= 0.10 mm/rev).

(a)

(b)

F

t

F

f

F

c

3.3.

Tool

wear

TiMMC has all the machinability issues of Titanium alloys added to the problems
associated with MMC, which include hard abrasive particles. A literature review on the
machinability of Ti-6Al-4V indicated the following [Yang X. et al., 1999, Machado
A.R. et al., 1990]: The combination of the low thermal conductivity and a very thin
chip results in very high cutting temperature that is concentrated in a small area on the
tool/ material interface. This high temperature, added to the strong chemical reactivity
of titanium, results in high affinity to almost all carbide tools. Furthermore, titanium
alloys have generally low modulus of elasticity which can cause deflection and rubbing
action. The cyclic chip formation, i.e., segmentation, results in wide variation of the
cutting forces. In the current investigation, a newly developed PCD tool was used to
avoid the diffusion problem. Abrasive wear on the flank surface was found to be the
dominant tool wear mode. This is typical to MMC materials, which contain abrasive
particles.

The tool wear tests were carried out near the optimum values of speed and feed

established from the above tests; v=180 m/min and f=0.1 mm/rev, respectively. For safe
operation and to ensure that the chips do not catch fire, a speed of v=170 m/min and
feed f=0.2 mm/rev were chosen. The tests were terminated once a value of 300 µm was
reached for VB

bMax

. Figure 5 shows the wear progression pattern for the selected cutting

condition, as a function of the cutting length (or time). At v=170 m/min and f=0.2
mm/rev, 10 seconds represents approximately 10 mm of axial cutting length. It can be
seen from Figure 5 that the wear pattern consists of three regions that reflect the
transition from a highly-loaded tribological system (with an initial high wear rate due to
the high stresses on the sharp tool edge) to a lightly-loaded system (with a reduced and
uniform wear rate) and back to a highly-loaded system (with accelerated wear due to the
tool temperature rise and reduced hardness) [Shaw M.C., 2005]. With these near-
optimum cutting conditions, the end of tool life was reached at axial length of cut of
approximately 140 mm. The abrasive wear of the PCD tool insert is shown in Figure 6.

Figure 5; PCD Tool wear progression.

Figure 6; Worn PCD tool after machining 160 mm axially with v=170 m/min ,

f=0.2mm/rev and 0.15 mm depth of cut.

3.4. Chip morphology and formation mechanism

Segmented chips that are typical for titanium alloys, were also observed during the
machining of TiMMC. This phenomenon of catastrophic or adiabatic shear is well
discussed in literature [Machado A.R. et al., 1990, Shaw M.C., 2005]. In the present
study, the chip morphology was analyzed for two different cutting speeds of 100 and
230 m/min. Some interesting observations and revealing information were established
from this analysis. Figure 7 shows that at the low speed, the chip crystalline structure
are more elongated and deformed, producing longer chips. At the higher cutting speed,
the shear strain in the primary shear zone is reduced, as demonstrated by the larger shear
angle and the smaller chip deformation. It was also observed that at lower speeds, more
particles are smashed, while they are just displaced or cut at high speeds.

A Quick Stop Device (QSD) was especially developed to investigate the chip

formation mechanism; morphology and segmentation. This also allows us to examine
the adhesion of the workpiece material to the tool face, the separation of hard particles
that result in three-body abrasion of the cutting tool, as well as, measuring the thickness
of the primary shear zone.

Figure 7; Chip morphology at (a) 100 m/min and (b) 230m/min

(a)

(b)

Figure 8; Chip formation using QS at cutting speed of 70m/min,

feed of 0.2mm/rev, and depth of cut of 0.15 mm.

Using this QSD, samples of the chip roots formed during the machining of Ti-

MMC were obtained and their microstructures were subsequently examined. The
photomicrograph in Figure 8 shows clearly the root of the chip, the primary shear zone,
as well as the chip segmentation. This phenomenon is familiar to Titanium machining
and is know as adiabatic shear banding. This effect can be explained by the high strain
rate and strain hardening effects combined with thermal softening in materials with low
thermal conductivity [Shaw M.C. et al., 1998].

In an attempt to further understand the different phenomenon associated with the

machining of Ti-MMC, a high speed camera was used to compare the process of chip
formation at low and high cutting speeds (120 and 230 m/min). The analysis showed
that at high cutting speeds, some particles are expelled at the tool-chip interface then
spark as they move away from the tool, as shown in Figure 10.

Figure 10; Chips catching fire recorded with High Speed Camera at v=230 m/min and

f=0.1 mm/rev. (The time elapsed between the two frames is 70 ms).

It was also established that at a cutting speed of 230 m/min, and feed of 0.1

mm/rev, the chips catch fire as soon as they are formed. However, when the feed is
increased to 0.2 mm/rev, this phenomenon disappeared, possibly due to the reduced
temperature in the chip, as the thermal energy is dissipated in larger volume.

4. CONCLUSIONS

The following conclusions can be drawn from this experimental investigation on
finished turning of Ti-MMC using Polycrystalline Diamond (PCD) tools. The
application of Taguchi’s Parameter Design technique showed that the cutting speed has
a much higher effect on surface roughness than feed. Additionally, there is a strong
interaction between these two variables. It was also established that the optimum cutting
conditions were v = 180 mm/min and f = 0.1 mm/rev, where the lowest cutting force
and the lowest surface roughness were obtained. At a higher cutting speed of 230
m/min, the feed showed no effect on the surface roughness, and the chip morphology
was clearly different. This suggests that there is a change in the cutting action and the
chip formation process at this cutting speed. This observation is further supported by the
fact that the chips catch fire as soon as they formed when a small feed of 0.1mm/rev
was used. When the feed increases to 0.2 mm/rev, no more fire was seen.

ACKNOWLEDGEMENT

The authors wish to acknowledge Dynamet Technology Inc., for providing the Ti-MMC
material used in this investigation. The support of the Aerospace Manufacturing
Technology Centre (AMTC), Institute for Aerospace Research (IAR), National
Research Council Canada (NRC), where the experimental work was carried out, is
indeed appreciated.

REFERENCES

[Rohati, P., 1991] Rohathi, P.; “Cast Aluminium Matrix Composites for Automotive

Applications” J. Organomet 1991, Chem., pp. 10-15.

[Bell, J.A.E. et al., 1997] Bell, J.A.E.; Stephenson, T.F.; Waner A.E.M.; Songmene, V.;

“Physical Properties of Graphitic Silicon Carbide Aluminium Metal Matrix
Composite”, SAE Paper No. 970788, 1997.

[Songméné, V., et al., 1999] Songméné, V.; Balazinski, M.; “Machinability of

Graphitic Metal Matrix Composits as a Function of Reinfoncing Particles” CIRP
Annals Vol. 48/1. 1999, pp 77-80.

[Kannan, S., et al., 2006] Kannan, S.; Kishawy, H.; Balazinski, M.; “Analysis of

Abrasive Wear During Machining Metal Matrix Composites” in Jahazi M.,

Elboudjaini M. and Patnaik, P., Aerospace Materials and Manufacturing
Processes; proc. of the 3rd Int. Symp. of CIM, COM 2006, 1-4 Oct., pp.227-236.

[Kilickap, E., et al., 2005] Kilickap, E.; Cakir, O.; Aksoy, M.; Inan, A.; “Study of tool

wear and surface roughness in machining of homogenised SiC-p reinforced
aluminium metal matrix composite”; Journal of Materials Processing Technology,
Volumes 164-165, 2005, pp. 862-867.

[Machado A.R. et al., 1990] Machado A.R.; Wallbank J.; “Machining of Titanium and

its alloys-a review”; Proceedings of the institute of mechanical engineers Part-B:
Journal of Engineering Manufacture 204, pp53-60, 1990.

[Yang X. et al., 1999] Yang Xiaoping; Liu C. Richard; “Machining Titanium and its

Alloys”; Machining Science and Technology, 3:1, 107-139, 1999.

[Muthukrishnan, N. et al., 2008]. Muthukrishnan, N.; Murugan M.; Prahlada Rao K. ;

“Machinability issues in turning of Al-SiC metal matrix composites”. Int. J. Adv.
Manuf. Technol. DOI 10.1007/s00170-007-1220-2008.

[Abkowitz, S.M., 2009] Abkowitz, S.M.; “Titanium Alloy and MMC Materials via

Powder Metal Technology”, Presentation to P&W Canada, Montreal, May2 2009.

[Basavarajappa, S. et al., 2006] S. Basavarajappa; G. Chandramohan; K. V.

Narasimha Rao; R. Adhakrishanan; V. Krish, ”Turning of particulate metal matrix
composites - review and discussion”, Proceedings of the Institution of Mechanical
Engineers; Jul 2006; 220, B7; Wilson Applied Science & Technology Abstracts .
pp. 1189.

[Tomac et al., 1992] Tomac, N; Tonnessen, K; “Machinability of particulate aluminium

matrix composites”, Annals of the CIRP, Vol. 41/1, pp.55-58, 1992.

[Ross Philip J., 1988] Ross, Phillip J. ; “Taguchi techniques for quality engineering” ;

McGraw-Hill, 1988.

[Kishawy et al., 2005] Kishawy, H. A.; Kannan, S.; Balazinski, M.; “Analytical

modeling of tool wear progression during turning particulate reinforced metal
matrix composites”, Annals of the CIRP, Vol: 54/1, Turkey, 2005.

[Zhu et al., 2004] Zhu, Y. and Kishawy, H. A; “Influence of alumina particles on the

mechanics of machining metal matrix composites”, International Journal of
Machine Tools and Manufacture, Vol. 45/4-5, pp. 389-398, 2005.

[Tomac et al., 1992] Tomac, N; and Tonnessen, K; “Machinability of particulate

aluminium matrix composites”, Annals of the CIRP, Vol. 41/1, pp.55-58, 1992.

[Shaw M.C. et al., 1998] Shaw M.C.; Vyas A.; “The Mechanism of Chip Formation

with Hard Turning Steel”, Annals of the ClRP Vol. 47/1/1998.

[Shaw M.C., 2005] Shaw M.C. ; “Metal Cutting Principles” Second edition, Oxford

University Press, 2005.