ii ISSN 1411-r284
Effects of Cutting Conditions and Chip Formation in Titanium Alloy Machining
Y. Burhanuddin, C.H. Che Haron', J.A. Ghani, G.A. Ibrahim, S. Junaidi Deparment of Mechanical and Matcrial Engineering
Universiti Kebangsaan Malaysia 43600 Bangi, Selangor, Malaysia
ondition and cutting tool. The approach required large oefeeseredrabsArrycab*hs.etstrrire,as'epcath_ora.arn-vTt"eeeo.o-vsdilaee1tralii,aolft"bfetlth.eeires-saJtsett*-hfaffreeo-"ctrirtome--noaeefchcmo;u;ceftrottit;hmnh;og;bedivn.maaTrto"iihoa;sinb;st;iei'oa-msfip;ocp;np;uotrtordtitonaalgcnht ttwDleoom.iitawthhpmoessuorpeanreitcduceorhdae(aPstrre,CnadcaaDtcce)hrar.eiirsnblUatiiincndtgiscev,oe,catlCyiotteaBnldnodNiwuic/timPaomCrnbhsoBiaddNsweublhuateoisnleeodnolsCPfgeBoaenrlNlyeae/srCPastluirCclyyiitBstayttNuba'mllDleiaenuraeedet mount of cost and cannot consider the combined effect of employed at high sfeed machining (Che Haron ;oumtttoibnigdneecvdoeneldofrpieticottnoosofloclnuifterteinsmgpoovndasere!ia.wbTthheeisch(pcutraupkotetsieinnogtfo.thoaics*coroeu,sinlet]aIt"rhcjihel,, 2}Al,Ezugwu eta12003). a^-M,tranceheinxapbeirliimtyendtaatlasotufditeitsa'nFiuomr siasviunsguathlley taimcqeuiraendd nd deprh of cur) by using design of experimenl d;i;;;; !o-1 experimenters used statistical design of :e:l:p!e-rim'hen"t' Statistical Cesigr of experiment is the process rctorial iesign nrethod. The first-order of tool life model are f cutting variables are investigated by the ,ppU""tio. o; of plapning the experiments so that the appropriate data l|eonrocrnatNetii.trTidhee (cCuBttNin)gatessctsutatirneg ctoonodlucwthedenustuinrgniCnguobfic should be collected which may be analysed by statistical methods resulting in valid and objective conclusions itanium in dry cutting condition. . (Montgomery 1991). Statistical design of experiment will Key.rord-:.oot tiie, fractionat factoriar, cBN, dry urning' titanium 6At-4v :;*:":lij:;j."" ""t"""ity and reduces the total number The objective of the study is to establislr the tool iife
is necessary to employ theoretical models making it
mociel cf CBN cutting when t.rrning Titaniurn 6Al-4V. It l- tNrRoDUc_rroN feasible to do predictions in function of operation conditions
such as spindle speed, feed rate,
cutting
depth,
Titanium
and
titanium alloy products is used mainly tool geometry and so on- Ttre study will investigate the n the aerospace industry due to the exceptional strength- *.*"p.ogr"rsion, the failue mode of cutting toollnd the o-weiglrt ratio, elevated temperafure perlormance and sisridcant factors that affect the CBN tool life. ;orrosion resistance. The applications of titanium are ltostly in jet engine and airfiame components where those rre subjeited to temperatures up to 600"C and for other II RESEARCH METHoDoLocY )ritical structurai parts. The usage is rvidespread in both hmmercial and military aircrafts. Titanium alloys fall A. DESTGN oF EX1ENMENTS uloys used today. i[P$eloe:uaLyplp.9.hc:aob"-mebJtepar*tia-pTshgearsHoseums:p.o,.Tr"Teh*iett,ahfmnafionuisImt5ilc0"o6:pmAlelmT-r4codV:nn',lIyt"mou1esfme"adb:lealirl;ltoJ-iybta,:fs:nta[hi.rui:se]m':lrTvvF:aaarrJii,aaohbtrieloieanslp.odrTeefhssseiepgnrneetsess4u*tul"tfdewye*i4"ltladpdkereowpsdtidhuinceoetloyfthcianuecctteooaxunopndletrlCiiftmheBeeaNnsstismcaoiunrnrevrtsaeopnnioevtsnoinsuaegss. ; I ttanium alloys has some certain characteristics that
several factors where it is
neceJsary
to study the combined
|mih on its machinability. Some of these are given as
effect of these factors on
responses.
The meaning of
Iollows: tow thermal conductivity, chemical reactivity
factorial design is that each complete trial or
replicalion wth
the
cuning lools materials al tool operating
of all the
pissible
combinations of the levels of the factors
is
investigated.
*f1. ' Cu.r"rpo,rding author: 'Iel.:(60)3-892 I 65 l6; E-rnail:chase@vlsi.eng.ukm.my '_ \. tr-*: m' &. ,'n ,r,', QIR Proceeding. 6-7 Sept 2006 F IMM-24 . Page I o16 l:_ii L' F!'l ki; H F, E E. E E :: responses (tool life, cutting forces) and machining The proposed relationship befween the machining independent variables can be represented by the following: T=CUlqfy dx) (l) ccuuttti(nmgms)pereedsspe(cmti/vmeliyn, )anfedeCd,r/a,tems,(mnmN/erecvon)staanndtsd.eEpq$.s(ol)f where 7 is the tool life in minutes, y, I and ,y' are the can be *ritten in the following logarithmic form: lnI= lnC+l ln Y +m lnf +n lnd (2) The fir$-order model can be expressed as: y = byxA + Drx, + brx, + brx, (3) In the present study, the parameters of Equation (3) wiil sInca4lwe,,hxhe6re:: yL1I iq(sd,utahmnedmbmy;ev,aa[sriauanrbedldeh)t,oaxor;el:tlhifIeenmYtoo,dxeazl =ploagIrnaamrfi,tehxtme,risc:. be estimated by using S.computer packige. To develop used. A design consisting of 8 experimeng lvas the first-order model, 2o-' parnal factorial aesigrr witt be conducted. LE VEL DE S IGNATIoN oF HiJRINT PRoCES S vARIAB LEs b;vel \r (n/min) F (r;:m/rev) D (mm) -l (Lor.4 180 0.05 0.1 I (tlieh) 300 0.25 t.00 CBN content Low Hish Illork piece material, I+{achine and cafiing inserts 4V rvas used for the tests. The machining tests were A 150 mm diameter x 300 mm long bar of Titanium 6.4.l- carried out on a Cincinnati Avenger 200T CNC lathe- A MCLNR 2020K09 tool holder was used to provide an g5o used were Kennametal grade KD050 and KDOgl cutting edge angle 4nd - 5" rake angles. The cutting tools designated CNGA 12040851020, in order to investigate the influence of CBN content. All of the experiments were conducted in dry condition.
Depending on the cutting
condition
and wear rate, machining was stopped at
varioui wear
of
the insert- Flank wear was considered as the interval of time varying from 5 t"" to I rnin to record
the criteria
of
tool failure and the wear was
mea-"ured using
a
Mitutoyo
toolmaker's microscope.
The machining i* rssN t4l, ,rrff stoppcd when an average flank rvear r\as Brealer than 0.30 I mm or fracturing happened. I III. REsuLrs AND DrscussloNs I A. . Starisrical Analysis I
Table 2 shows the experimental conditions together with the measured tool life values. The
results were transformed to half-normal probability plot as in Figure l- The figure shows the factors
(depth of cut, cutting speed, feed rate) and
the
interaction
(cuning speed-cutting depth), which may have an effect to cutting tool life. Then,
analysis of variance (ANOVA) was applied to
calculate
the main effects of cutting speed
(If,
feed rate (f), depth of cut (d),
and CBN content
together with their
two
-level interaction effects on
tool life.
The ANOVA output and the calculated F ratios are shown in Table
3
for each significant effect. The
5 per cent level for testing the significance of the main
effecB
and
the
interaction
was used. Table 3 shows that
depth of cut is most significant,
and follcwed
by
cutling
speed and
cutting speed-depth of cut
interaction.
Feed rate is
not significant because its "Prob > F" values greater than 0.1. Even ifthe feed rate has a greater value than 0.1, this term wiii not be negiected in order to take into account the feed rate contribution to tool life. While the CBN ccntent does not have any effect to tool life absolutely. TABL,F 2 E)GERIMENTAL CONDINONS AND RESULTS Facton Run Cutting Feed rate Tool speed (mm/rev) Cutting CBN depth life content (sec) (m/min) (mm) 2 1 3C0 0.05 I.0 Low 40 3 180 0-05 0.1 Low 1740 180 4 0.2s 0.1 High 820 5 300 0.2s 0.1 Low 130 I80 0.05 6 L0 I .0 High 150 180 0.2s Lorv l0 8 7 300 0.05 300 0.25 0.1 High 420 t.0 Hish l0 9ft Int'l QIR proceed ing, 6-7 Sept 2006 tMM-24 Pqe2 of 6 rssN l41l-1284 Normal Plot of Residuals I 99 95 l7 ee0 n o il-l 9ooI800 lso o i /Ll 6Eo3z0o z. 1g 5 = ttl. ,l :TI n 181.25 362.5q i4:.15 rti 08 lE-fectl
Fig. 2. Normal plot of residuals for
tool life
data Fig.
I Half
normal % probability
plor ANovA
FoR
sELECTEo .IH3?inr MoDEL [pARrr.AL suM oF SQUARESj Source Model l c B AC Residual ConTotal Sum of Squares DFMean Sqr_rare F Prob > F Value 2,302,00A 4 575,600 7.27 0.0575 561,800 I 56r,800 '1.10 0.0761 238,0C0 I 238,000 3.0t 0.1813 1,05 t,000 I 1,05t,000 13.28 0.a356 45 1,300 I 451,300 5.?0 0.0969 237,s00 J 79t'50-00 2,540,000 7 The CBN content seems to be suitable for intemrpted cutting than continuous cutting as in this experiment-'I'he Rex-spqlauinasred90s.6ta5tiys'ticofinthdeicavtaersiabthilairty thinetofiorsltl-ioferd(eTr)m.oTdheel calculation also indicates that the model has an adequate ; signal to noise ratio. .
The normal probability
plots
of the residuals and the
, lrte
are shown in
Figwes 2 and3, respectively. A check on plots ofthe
residuals versus the predicted response
fortool 1
the
plots in Figures 2 and 3
revealed that-
ihe
'residuals generally fall on a straight line implying that the
erors
are
drstributed
normally. This implies that the models proposed are adequate and there is no reason to suspect any violation of the independence or constant variance
asswrption. It means that the proposed nrodel using partial factorial design is suitable for running the experiment. But, in order to obtain a more precise result, the second-order model is proposed. B. Tool wear and tool life Figure 4 shows the ftank wear progression curves of CBN tool at the different cutting conditions. The flank wear progressions were generally divided into three stages: the rapid initia!, the relative steady state and finally the abrupt ofwear.
The flank
vrear
rate was rapid at higher cutting speeds, feed rates
and depth of cuts.
There are several
fypes
of wear mechanisms that can influence the tool wear and subsequently the tool
iife.
The observed wear mechanisms were abrasion,
attrition,
adhesion,
diffirsion- dissolution,
chipping and
fracture. lnt
'l QIR Proceeding, 6-7 Sept 2006 IMM-
24
Page 3
of6
ISSN
Residuals vs. Predictecj machining- If the material welded to the tool dynamical stress, this eventually results in severe failure- Figures 5(b) and 6(b) show ttr. v) '" surface ofCBN tool due to severe failure. pJ Another w€ar mechzrnism which present u experiment is dissolution and diffusion of the n,,r E o 0.00 The dissolution-diffirsion mechanism .urr"A' 1'il1 .N E@l c P6(a.)Tsh:o*w 4the 3fo*rm"'elad:cTraft-e1rob:gviio"usllry.d rig., . ilai 0 -1.50 cfiugtutirFneig,gudpreaerpat7mhaeoltsfeorcssuhatotiwsditfrhfeeesruemlntsotsodtfeupcrutfhltutioenfngcctiuontog.lBflaaifsceetodornotosnr tool life, followed by cutting speed and feedrate. 652.50 Predicted Fig. 3. Residuals vs predicted responses plot low so that the wehr based on ptastic shear or diffixion mechanism. At the lower cutting ,p""d, t"*p"rahres are Figure 5(a) show clearly rubb;ng and ath-ition #ear does not occur. The flow of metal past the cutting edge is tdah-i.1ieusc^^eu^+i^ici^a+ocnrd-w-.i:lt.iLiro'tntants-,elat-oroglemrafyi:abgemieesnstcsonmtinauyoubs.eUntdoemr more irregular, laminar, a builrup edge may be formed test, the well-defined built-up edge has nct appeared- interminently from tool surfaceiTrent l99tj. nut in tnis processing, e-specially d,ry cutting, the lemperature bcaenlnavlesoenbethae-sctriobeodl toanthde cwheomrkicanlrraetaecritaiolsn. aDiduaridnhgestiohne Beside the abrasion and attritiorr, .rttirrg tool wear 4V at cutting speed 180 rrr,/min, feedrate 0.25 mm/rev (a) depth of cut 0.1 ,,. Fierl" 5: SEM micrograph of CBN tools while machirring Ti_6AI- mm, O)
depth of cut I mm the
tenCency for the titanium to weki to the tooi during mcreases rapidly. The increasing temperature wili induce
Flank wear (VB)
05t) 0.45 0.40 0.35
0.
30
0.
25
0.
20
0.
15 :
0.
to ::: i il r. la 0r+oloroott 200 400 600 8@ looo 1200 T@t ,ife (se) Fig. 4. Flank wear progression curves 1400 1600 1800 2ooo \ at cufting speed.lO! nVrnin (a) feedrate 0.25 mm/rev anO Oepth of'cut 0.1 {ig. 6. SEM-micrograph of CBN tools while machining Ti_6At4V mm, (b) feedrate
0.05 mm/
rey
and depth of cut
I
mm
;{ .* I 4
Int'l QIR Proceeding, 6-7 Sept 2006
tMM-24
Page 4
of6 200c 1900 1600 ? .r,{00 s 1200 J rooo : Eoo I ooo 400 200 0 rssN 1411_1284 (a) E 180025Hish El 300 0.05 High EI €00.05 Lm O3000251il eza (b) 120 !ro f,,0 .rl60 Fig. 8.Chip formed at cutting speed 180 mimin,
feed rate 0.
25 mrn
/rev and depth
ofcut
0.
1
mm (a)
top view (b) side view JO Fig. 7 Tool life a1 different depth of cutting (a) 0.1 mm (b) I mm . Figures 5(b) and 6(b) confinned the influence ofCepth cf cut to tool wear mechanism. In this case, higher depth oicut enlarged the possibility oftool to severe failure and shortens the tool life. It is suggested to use lower depth of cut Ln finish nrachining of titanium using CBN tcol. It seems that CBN content (grade) give small significance level of to tool life..Lower CBN content gave a reiatively good performance and it is parallel to the finding of Ezugwu et al (2003). C. Chip formation Figures 8 and 9 show the differences of two formed chips at the
different cutting speeds and
depth of cuts.
The depth of cut
will influence the width of chip, tlpe of flal.ies and peak shape of flakes. Chip formgd in turning
at a depth of cut
I
mm
has approximately 2.5 times wider than chip formed at
depth of cut 0.1 mm. The
bigger flake and the smaller flake were formed at a
depth of cut 0.1 mm. The
smaller flake was formed at the right side-
At the depth of cut of
I
mm, the
smaller flakes were not formed. The sharper peak shape was formed at the
depth of cut 0.1 mm
than at the
depth of cut
1
mm.
While the interval offlakes depend on the cutting speed. Fig.9. Chip formed at cutting sped 300 n/miq feed rate 0.25 mnlirev and depth ofcut I mm (a) top view (b) side view N. CONCLUSIONS It
has been
shown
in this work,
machining of titanium
6Al-4V with
CBN cutting tool, the partial factorial design can be applied in designing the experiments. The desigrr is very helpful in the running of expensive cutting tool- material combinations. The work showed
that the depth of cut is the most significant factor
to tool life,
followed
b1' 9il' Inr
'l QIR Proceeding,6-7 Sept 2006 IMM-
24
Page
5 of6 E tri F. E. p- E. CBN tool has three stages of wear process: rapid initially, slowly w€ar progresi and abruptly failure. ihe dominantly wearing mechanisrn present rvas abrasion, attrition, adhesion, diffirsiondissolution, chipping and fracture. The cutting speed induced atbition wear. While cut. Diffirsion-dissolution was induced by increasing chipping and fracture was generated by higher depth of temperature during tuming operation. Beside the biggei flakes, some smaller flakes were formed when turning at lower depth of cut. The width and height of formed chip are depended on depth of cut while interval between peaks offlakes is depended on cutting speed. AcxNowrBocl,rgr.rr Ministry of Science, Technologl and Environment fof
The authors would like to thank the Malaysian
sponsoring this work under project
tRpA
03-02-02-
0A62- EAt22. REFERENcES lll Arbizu, I.P. &P€req C.J.L. 2003. Surface roughness prediction by factorial design of experiments in tuming processes. J_ Mat. Processing Tech . I 43-l 44: 390-396 ISSN I4ll_1281 L2l Bhaumik, S.K., Divakar, C. & Singh, A.K. 1995. Machining Ti. 6AI-4V alloy with a wBN-cBN composite tool. Mat. & Design. l6(a): 221-226. t3I Che-Haron, C.H.. 2001. Tool life and surface integrity in turning titanium alloy. J. Mat. hocrssing Tech . I 18:. 231-23'7 t41 Choudhury, I.A. El-Baradie, M.A. 1998. Tool-life prediction & model by design ofexperiments for tuming high strength sreel (290 Bfil$. J. Mat. Processing Tech-77:3t9-326. I5t D.C. Montgomer_v. l9l. Design and Analysis cf Experimen6. John Wiley and Sons: New York. t61 E.M. Trent. 1991. Metal Cutting- 3rd ed. Buttreworth-Heinemqn. Oxford. 171 Ezugwu, E.O. Bonney, J. & Yamane, Y. 2003. An overview of the machinability of aeroengine alloys. J. Mat. Processing Tech. 134: 233-253. t8l Lin, Z.C. & Chen, D.Y.. 1995. A study of cufting with a CBN tool. J- Materials Proc. Tech. 49:.149-164. tel Nabhani, F. 2001. Machining of aerospace ritanium alloys. Robotics & Comp. Intcg. Manuf. 17:.99-106 UOl Narutaki, N., Murakoshi, & Motonishi, S. 1983. Srudy on A Machining of Titanium Alloys. Annals of the CIRP.32: 65-69. I l] NoorCin, M.Y., Venkatcsh, V.C., Sharif, S. , Elting, S. & Abdullah, A- 2004- Application of response surface methodology in describing the performance of coated carbide tools when tuming AISI 1045 stecl, J. Mal Processing Tech. 1.15:46-58. |21 Zhou, J.M., Walter, H., Andersson, M. & Stahl, J.E. 2003. Effect of chamfer angle on wear of PCBN cutting tool. Int. J. Mach. Tools Manuf.43: 301-305. t: ; &tr I ts E. r. F. F. il- IE s s prnt
'l QIR Proceeding, 6-7 Sept 2006
TMM-24 s &
Page
6
of
6 ffii: H 6 &. m & fafi E re € E
