The Mechanism of Chip Formation with Hard Turning Steel M.C Shaw (11, A.Vyas Arizona State University, Tempe, Arizona, USA Received on January 6,1998 Abstract Steels having a hardness of HRC 60 or greater are presently being finished by t r m n g instead of grinding This is usually done using a polycrystalline cubic boron nitride insert having a rather large nose radius on a very rigid machine at a relatively high cutting speed In order t o understand this process it is important that the sequence of events occurring in the formation of the unique type of chip involved be correctly identified Experimental evidence t o this end is presented and discussed in fundamental terms Keywords: Cutting, Chip formation, Hard turning Introduction In a previous paper (Shaw and Vyas,l993) the mechanics of chip formation involved when hard (brittle) materials are cut was considered Since that time a great deal more experimental work has been performed The basic picture presented there still holds, the only exception being the nature of the non-etching white layer observed when a hard high carbon steel is turned a t a relatively high speed This white layer has now been identified as a mixture of untempered martensite and ; iron by electron diffraction using a transmission electron microscope on a carefully thinned specimen The 1993 paper should therefore be considered an introduction t o this one Nomenclature In (Shaw and Vyas ,1993) the various chip types that have been identified were itemized as f 01lows: Steady state tvpxs: Concentrated shear zone, Pie shaped shear zone, Extended shear zone due t o a blunt tool-tip _Cy_fiLtypes: Discontinuous, Wavy, Sawtooth, Built up edge (BUE) In addition, noncyclic changes in chip thickness are sometimes obtained, particularly when pure materials are cut a t very low speed In the literature the term segmental chip is often used t o describe all of the cyclic types, particularly the wavy and saw-tooth types This is unfortunate since these two types of chips are distinctly different For example, the cycle frequency for a wavy chip is typically about 100 Hz while that for a saw-tooth chip is t o orders of magnitude greater Also, wavy chips not have sharp points while saw-tooth chips (Fig.1) Annals of the ClRP Vol 47/1/1998 Cycle Frequency The mean cycle frequency for either of these chip types may be readily determined by dividing the speed of the chip (vchip) by the mean spacing of points of maximum chip thickness (p,) = (vchip/pc : (1 Origin O f The Saw Tooth Chip The saw tooth chip was first identified about the same time as the wavy chip (wavy chip :Bickel 1954; saw tooth chip, Shaw et al, 1954) and there is a considerable body of literature pertaining to the mechanism of formation and the characteristics of each The saw tooth chip was found while studying the machining characteristics of a new structural material (titanium) having unusually low values of thermal conductivity and volume specific heat (Fig.2) Also, the concept of adiabatic shear introduced by Zener (1948) in connection with the mechanics of ballistic impact was a relatively new popular concept Unfortunately, it was suggested (Shaw et al 1954) that the saw-tooth chip observed when turning a titanium alloy might be due to periodic adiabatic shear This misconception has persisted to the present Low Speed Turning of a Very Brittle Material Nakayama (1972) found that saw-tooth chips were produced when highly cold worked (brittle) 40160 brass was cut under orthogonal conditions at very low speed He observed shear cracks forming periodically at the free surface which ran down the shear plane toward Ihe tool tip This divided the chip into blocks that slid past each other as the chip moved up the face of the tool (Fig.3) Quick-Stop Tests In the case of hard steel turned at a practical speed, chips are found to show the block-like structure of Fig near the free surface plus bands of concentrated shear extending downward from the cracks defining the edges of the blocks These concentrated shear bands curve downward and gradually become parallel to tool face as the chip moves up the tool face (Fig 4) In order 77 to obtain a betler idea of the sequence of events responsible for chips like Fig 4, a series of quick stop tests was performed on Ti-6AI-4V cut under orthogonal conditions Figure gives two representative results Figure 5a shows the situation at the beginning of a cycle while Fig 5b is about half way through a saw tooth cycle These photomicrographs reveal the following: A complete crack (.con:inuous across the width of the chip) extending about half way down a straight shear plane toward the tool tip, followed by a region that does not appear to be completely cracked but weakened by microcracks (Fig 5a) A band of concentrated shear going all the way to the tool face in a straight line (Fig.Sa),followed by bands that begin to curve toward the tool face more and more as the chip moves up the tool face (Fig 5b) Movement of blocks of material that gradually proceed outward due to sliding along the fully cracked surfaces together with extension of bands of concentrated shear in the microcracked region (Figs 5a and b) Thinning of the microcracked region as the chip moves up the face of the tool (compare distance D1T with D2T in Fig 5b) A gradual approach to the final shape of the chip as it moves up the tool face requiring several cycles before the chip leaves contact with the tool No evidence of adiabatic shear is found along the fully cracked surfaces such as C2 D, in Fig 5b Fig.1 Chips commonly referred t o as segmental a) waW chip b) saw-tooth chip Fig Ti -140A chips a) Cutting speed (V) = Feed (f) = 0.0104 ipr (0.26 mm/r); Rake angle a = +so b) V = l O O fpm (30.5m/min; f = 0.0052 ipr (0.13 fDm (45.7 m/min.); Discussion In the discussion that follows the completely cracked region where a continuous crack extends across the chip width is designated GC (gross cracked region) while that corresponding to the region where cracks are discontinuous across the chip width is designated MC (microcracked region) The significance of the thinning of the MC region as the chip moves up the tool face is that this gives rise to a cutting ratio (r) greater than one This is usually the case when hard steel is turned with a negative rake tool Important consequences of r>l is that the speed of the chip (VC) will be greater than the cutting speed (V) and mm/r); cr - +So (after Shaw e t at, 1954) the shear angle will be greater than 45O The significance of the gradual approach to the final chip shape involving several cycles of chip-tool contact is that any slight variation in the cracking pitch (p,) will not be reflected in a fluctuation of the shear angle This causes the pitch of the “teeth” of the chip to be remarkably constant removing the effect of any slight variation in stress concentration in the original surface The fact there is no evidence of adiabatic shear on the GC surfaces (such as C2D1 in Fig 5b) suggests that the root cause of saw-tooth chip formation is periodic cracking and not adiabatic shear The only adiabatic shear involved in fig 5b is in the MC region which begins to develop only after a GC region forms Therefore, any attempt to predict the onset of saw-tooth chip formation due to an increase in cutting speed or feed (Fig 2) will involve fracture mechanics and not heat transfer 78 Fig Saw-tooth chip when turning highly cold worked brass a t very low speed with negative rake tool (after Nakayama, 1972) Hard Turning of Steel Steel in the hardened state is being finished today under conditions that produce surface finishes comparable with those in fine grinding (Ra=0.2to 0.4 pm) This is possible due to the availability of ceramic and cubic boron nitride tools of improved quality and machine tools of greater rigidity To produce surfaces of the desired finish at a reasonably high feed rate it is necessary to use tools having a relatively large nose radius Figure shows a typical turning arrangement where a nose radius (Iof )3 mm (0.118 in.) is making a cut at a feed rate (f) of 125 pm/r (0.005 in/r) , The depth of Cut (d) will be much less than r, so that all cutting is on the nose radius The scallop left behind on the finished surface will give a theoretical peak- to- valley roughness (Rt) of f2Br (independentof the depth of cut) T o a good approximation the theoretical arithmatic average roughness (Ra) wit1 be f2 /32 r) b r the example of Fig 6a: Ra= (125~lO-~)~/1(32)(0.003)] = 163pm (6.52 1.in) For dry turning with a sharp tool and a rigid system the actual surface roughness including vibration and other non- geometrical effects will be within a factor of two of the above value Figure 7a) shorn the chip of Fig oriented along the negative rake tool face as a free body, and just below, the tool is shown in the process of making a cut This is a snapshot of a saw-tooth an instant after crack formation, where the element just formed has slid outward a small distance DC, Fb.4 Case carburized steel chip (HRC=62) V=338fpm (103 m/min); f=0.01 l i p r (0.28 m m h ; d=O.Ol l i n (0.28 mm); nose radius=0.125 in (3.1 mm); (x=-7O According t o Nakayama ( 974) the equal and oppositely directed forces R and R' should be parallel t o CD Forces R and R' are shown resolved parallel and perpendicular t o the shear plane (Fs and NS respectively) In this instance the tool face friction force (F) is very small while there is a very significant zone showing bands of concentrated shear The gross cracked region of the chip (GC) extends from C, t o D, while the microcracked region (MC) extends fom D, t o T The hodograph for the GC region is given in Figure 7b) The cutting ratio (r) for this chip may be found by dividing the undeformed chip thickness (t) by the mean chip thickness (Tc) However the composite surface CID1 + C2D2 + %D3 + etc was found t o corrrespond t o the equivalent length of uncut surface on the work This was demonstrated by coating the original surface of the work with soot and then producing a replica of the surface of the chip by pressing a soft plastic material into the back of the chip Valleys on the chip become peaks on the replica with slopes CD coated with soot Microscopic meeasurments on the replica revealed that mean length CD corresponds t o the mean distance between cracks on the work (p) Therefore a convenient method of finding r is t o divide the mean tooth pitch (pc) by the mean value of C2D2 (=p) Thus, r = pc/p = Vc/V (2) There is a small tooth-to-tooth variation of pc and p for the chip of fig When all combinations of pc and p are cosidered the mean value of pc/p is found t o be: pc/p = 1.59 t Fig Quick-stop photomicrographs of Ti-6AI-4V chips a) shortly after formation of gross crack (GC) at free surface showing extent of GC and MC b) about halfway between cyclic cracks V=l72fpm ( Z m h i n ) ; t=O.O07in (0.01 mm), b= 0.100 in (2.54mm); rx=-7'; tool material, WC 15 ( 59.t 10%) = VC/V 79 I \: ‘I I I I I r Fig Cutting geometry for hard turning with tool having relatively large nose radius The cutting ratio for the saw-tooth chip of Fig was obtained by the conventional method involving measurement of chip length and weight (see for exmple, Shaw 1984) and found to agree with the above value For the chip of Fig the’mean value of the cracking frequency was found t o be 18 kHz bv use of eq.1 This approahes the upper limit of the audio frequncy range and has been verified by dynamic masurement during sawtooth chip formatiion The very inhomogeneous strain in the MC region of Fig.4 will give high temperatures in the bands of concentrated shear along the microcracked extensions of the cJross cracks, for high cutting speeds It is only these bands that involve adiabatic shear For ferrous alloys the temperature in these concentrated shear bands may exceed the transformation temperature where ferrite (fi iron) changes t o austenite ( y iron) Evidence of this transformation is found in the nonetching white bands in Fig and other similar photomicrographs of saw-tooth chips of ferrous alloys machined a t high speed As previously mentioned these white areas have been identified as untempered martensite + y iron However, before rapid cooling and during chip foimation, these bands will be y iron a t high temperature which is a relatively soft material that offers little resistance t o plastic deformation, as would be the case for a molten metal It is the presence of high temperature y iron along the tool face that gives such low values of tool face coefficient of friction (about 0.05 in Fig 7) The distance one segment slides relative t o i t s neighbor during one cycle ( C2 in Fig 7a) will depend upon the distance between cracks on the work (PI When p p p (r>l), this is a result of the compressive stress on the material in the MC zone 80 being sufficient to cause elongation of the MC region of the chip Material in the GC region is carried along with the MC material, resulting in r for the entire chip being greater than one The drawing below the photomicrograph of Fig.7a) is consistent with the quick-stop photomicrographs of Fig, The hodograph in Fig.7b only holds for the GC region since the direction of the shear bands in the MC region are continuously changing direction as the chip moves up the tool face The shear angle may be obtained from Fig 5b where: r = V,/V = sin +/cos (C ++) (3) + Knowing r and a, the value of I.$ that satisfies Eq may be found An energy balance for the specific energy for the chip of Fig may be readily perfomed but space limitations not permit this to be included here The Adiabatic Shear Theory According t o the adiabatic shear theory the root cause of saw-tooth chip formation is a catastrophic thermoplastic instability where the decrease in flow stress due t o thermal softening associated with an increase in strain more than offsets the associated strain hardening A number of papers suggest adiabatic shear as the origin of saw-tooth chip formation A few of these are Davies et al, 1996 and 1997; Komanduri et al, 1982; Koenig e t al, 1984; Recht, 1964 and 1985; Sheikh-Ahmed and Bailey, 1997; and ZhenBin and Komanduri, 1997 The most recently proposed model based on the adiabatic shear theory is given in Flg Comparison of the Two Theories The quick stop photomicrographs of Fig are useful in comparing the two theories First of all, a thermally initiated process should be expected t o have its origin where the temperature is a maximum which is a t the tool tip This is in agreement with the model of Fig but not with reality (Fig 5) The crack in Fig 5a clearly runs from the surface toward the tool tip, initially along a relatively straight shear plane A shear crack should be expected t o initiate near a point of maximum shear stess where the compressive stress is a minimum (i.e.at the free surface) Although it is normally assumed that stress along the tool face and shear plane are constant, this is not so The normal stress on the shear plane rises exponentially from zero a t the free surface to a maximum a t the tool tip Evidence for this and for a similar variation of stress on the tool face is given in Sarnpath _and Shaw 1983 AS higher normal stresses are encountered as a shear crack progresses downward from the free surface toward the tool tip, a continuous gross crack will be gradually converted into a Fig Chip of Fig.4 oriented to tool face a) Free body of chip b) Hodograph of GC region discontinuous microcrack This is seen to be the case in Fig.5a It should be noted that a titanium alloy was employed in obtaining Fig.5 which should favor the adiabatic shear theory due to its low values of thermal conductivity and specific heat Figure shows a recent model (Zhen-Bin and Komanduri, 1995) employed to explain saw-tooth chip formation in terms of an onset of adiabatic shear In this model an adiabatic shear band runs from the tool tip A in Fig.8a along a straight line t o the free surface As the chip moves forward the concentrated shear band just formed rolls down onto the tool face as block (1 ) glides outward along two adjacent shear bands While this model explains the sharp point it is not in agreement with Fig 5a in that the adiabatic shear band shown from ‘C to D in Fig 8c is not found experimentally in Figs or Even when surfaces CB in Fig.8~ are carefully protected with a hard material t o prevent alteration or loss during polishing, no evidence of adiabatic shear has been found on such surfaces with the electron microscope a t very high magnification While there is no apparent reason the concentrated shear bands bend down and approach the tool face in the model of Fig.8, the reason for this is evident in the shear crack theory The fact that work-piece hardness (brittleness) is so important relative to the onset of saw-tooth chip formation further supports the crack initiation theory 81 impel fections in the original surface However, this small variation is not reflected as a change in rake angle, since several cycles are involved in chip-tool contact as the final geometry of the chip evolves (Fig 5) The need for nomenclature t o distinguish between the several types of cyclic chips cannot be overemphasized because the basic mechanisms involved for each is entirely different (b) Fig Model used by Zhen-Bin and Komanduri for thermal analysis based on adiabatic shear theory (after Zhen- Bin and Komanduri, 1995) of saw-tooth chip formation further supports the crack initiation theory Concluding Remarks Considerable experimental evidence supports the concept that the root source for sawtooth chips is cyclic cracks that initiate a t the free surface of the work and proceed downward along a shear plane toward the tool tip and not adiabatic shear If the material is sufficiently brittle these cracks may be continuous across the width of the chip (called gross cracks, GC) essentially all the way t o the tool tip (Fig 3) For less brittle materials higher cutting speeds are required for saw-tooth chips t o form and continuous cracks will become bands of discontinuous microcracks as high crack arresting normal stresses are encountered close t o the tool tip There are then two regions as the chip proceeds up the tool face - the material between gross cracks sliding outward, and deformation in the MC region confined primarily t o concentrated shear bands that gradually bend downward and run along the tool face (Figs and 7) If the chip speed is high enough when hard turning steel, the temperature may reach a value high enough t o cause a transformation to austenite This will offer little resistance t o deformation acting as though it were molten metal After rapid cooling the transformed shear band material becomes a very hard non-etching white layer consisting of untempered martensite and retained austenite Tool face friction is found t o be unusually low when a white layer exists along the tool face side of a polished and etched chip The point spacing in a saw-tooth chip is remarkably constant but does vary slightly due t o 82 References Bickel, E., (1954) Hochfrequenten Zeitlupenaufnamen (Spandbindung), Anna!s-of-CIRR 2, 90 Davies, M.A., Chou,Y., and Evans, C.J., (1996) On Chip Morphology, Tool Wear, and Cutting Mechanics in Finish Hard Turning, Annalspf CIRP , 45/1, 77-82 Davies, M.A., Burns, T.J., and Evans,C.J., (1997) On the Dynamics of Chip Formation in Machining Hard Metals, Annals of CIRP 46/1 1-6 Elbestawi, M.A., Srivastava, A.K., and ElWardany, T.I., (1 996) A Model for Chip Formation During Machining of Hardened Steel, Annals Of CIRP -_45/1 71-76 Komanduri, R., Schroeder, J.A., Hazra, J., von Turkovich, B.F., and Flom, D.G., (1982) On the Catastrophic Shear Instability in Hlgh Speed Machining of an AlSl 1040 Steel, Trans ASME (J Ena for Ind.) 104, 121-131 Koenig, W.A., Komanduri, R., Toenshoff, H.K., and Ackeshott, G., (1984) Machining of Hard Metals, ~ Annals of- C I R P , 3/ 7Nakayama, K., (1 972) Private Communication Nakayama K., (1 974) The Formation of Saw Tooth Chips, Internat Conf o n P r o d Ena Tokwo_, p.572-577 Recht, R.F., (1 964) Catastrophic Thermoolastic Shear, LApRlied Mechanics, _3_9,189-193 Recht, R.F., (1985) A Dynamic Analysis of High Speed Machining , J Ena for Ind 101, 309-315 Sampath, W.S., and Shaw, M.C (1983), Fracture On The Shear Plane in Continuous lla& Cutting, Proc Amer Metal Workina Research Co& SME, Dearborn, pp 281 -285 Shaw, M.C., Dirke, S.O., Smith, P.A., Cook, N.H., Loewen, E.G., Yang, C.T., (1954) Machining Titanium, MIT RePort t o U.S Air Force Shaw, M.C., Metal Cuttina Principles Clarendon Press, Oxford, (1984) Shaw, M.C and Vyas, A (1993) Chip Formation in the Machining of Hard Steel, Annals of CIRP 4211, 29-33 Sheikh-Ahmad, J., and Bailey, J.A Flow Instability in the Orthogonal Maching of C.P Titanium, J of Mfa Sc and Ena 1 307-318 Zener, C., (1918) The Micromechanism of Fracture , Fracture of Metals, ASM ~ p 3-31 Zhen-Bin, H and Komanduri, R (1995) On a Thermomechanical Model of Shear Instability in Machining, Annals of CIRP 4411 69-73 ... coating the original surface of the work with soot and then producing a replica of the surface of the chip by pressing a soft plastic material into the back of the chip Valleys on the chip become... Thinning of the microcracked region as the chip moves up the face of the tool (compare distance D1T with D2T in Fig 5b) A gradual approach to the final shape of the chip as it moves up the tool... fluctuation of the shear angle This causes the pitch of the “teeth” of the chip to be remarkably constant removing the effect of any slight variation in stress concentration in the original surface The