Burr analysis in microgrooving

Transcription

1 Int J dv Mnuf Technol (1) 5: OI 1.17/s ORIGINL RTICLE Burr nlysis in microgrooving uy Le & Jong-Min Lee & Su-Jin Kim & ong-yoon Lee & Seok-Woo Lee Received: 3 July 9 /ccepted: 4 Jnury 1 /Published online: 4 Mrch 1 # Springer-Verlg London Limited 1 bstrct Microburr formtion ffects bdly on product's qulity, especilly the finishing surfce. Besides, deburring techniques on microcutting re lmost impossible or high cost. Understnding burr formtion phenomen nd minimizing burr size in correltion with cutting conditions nd mteril properties, in this cse, re more pproprite. For the purpose of exploring the burr phenomen in micrometl cutting, two cses of burr formtion in grooving microptterns re introduced in the pper. The burr hppens long the cutting direction side burr of prism pttern, nd the other hppens t the exit edge of the pttern in the cutting direction exit burr of pyrmid pttern. Besides, the exit brek off which occurs during the exit burr formtion is lso studied. The nlyticl solutions for predicting the burr nd brek off size in ech cse re lso proposed nd compred with experiments. Keywords Burr formtion. Microgrooving. Micropttern. Slip line theory. Hrdness 1 Introduction. Le : J.-M. Lee : S.-J. Kim (*) eprtment of Mechnicl Engineering, Gyeongsng Ntionl University, Gjw, Jinju, South Kore e-mil: sujinkim@gnu.c.kr.-y. Lee : S.-W. Lee Kore Institute of Industril Technology, nsn, Seoul, South Kore The motivtion for micromnufcturing rises from the trnsltion of the knowledge obtined from the mcromchining domin to microdomin. However, there re chllenges nd limittions on micromchining, nd simple scling might not be used to model the phenomen of micromchining opertions effectively [1]. High-ccurcy miniturized components re incresingly in demnd for vrious industries, such s erospce, biomedicl, electronics, environmentl, communictions, nd utomotive. This minituriztion will provide microsystems tht promise to enhnce helth cre, qulity of life, nd economic growth in such pplictions s microchnnels for lb on chips, shpe memory lloy stent, fluid grphite chnnels for fuel cell pplictions, subminiture ctutors nd sensors, nd medicl devices [ 5]. One of the most populr micromnufcturing methods is microcutting, especilly metl cutting, nd microptterns which re pplied in most of the pplictions listed bove re fbricted by microgrooving. Burr formtion in microgrooving usully ruins the finishing pttern surfce. eburring techniques on those surfces re lmost impossible or too specific, complicted, nd high cost [6]. Understnding burr formtion phenomen nd minimizing burr size in correltion with cutting conditions nd mteril properties, in this cse, re more pproprite [7]. But most of previous successful reserches re bout burr formtion in generl cutting, so microburr problems still need to be solved in order to serve the requirement of the industry. For the purpose of exploring the burr phenomen in micrometl cutting, two cses of burr formtion in turning microptterns re introduced in the pper. The burr hppens long the cutting direction side burr of prism pttern, nd the one hppens t the exit edge of the pttern in the cutting direction exit burr of pyrmid pttern.

2 57 Int J dv Mnuf Technol (1) 5: Side burr in grooving microprism pttern.1 Cutting model u eformed chip The cutting model is built from mm in dimeter nd 1, mm in length roll mold in turning microprism pttern. The ide is tking smll squre piece of the roll surfce nd ssuming it is similr to grooving microprism pttern on plte mold. This will mke the first study bout side burr hppening long the pttern more generl s shown in Fig. 1. Cutting mteril is the copper coted outside the roll. Becuse the coting technology cnnot ensure tht everyplce on the roll surfce could hve the sme coting qulity, Vickers hrdness tests re required to check the mteril properties of the surfce. Then combining with the work of Hirno et l. [8], n pproximte reltion between Vickers hrdness nd yield stress could be climed. h ¼ :4 :7S y ð1þ where h (kilogrms per squre millimeter) is Vickers hrdness nd S y (newton per squre meter) is yield stress of coted copper. In microgrooving, ccording to this study, t every cutting depth, there obviously exists criticl thickness t cr t which cutting is impossible. It cuses plowing on the pttern s side surfce long with the deformed chip formtion upon it. In cse of the remined thickness of workpiece reches to t cr vlue, the top prt of the tringle begins to deform plsticlly long the feed direction. This concept comes from the comprison nd evlution between burr in generl cutting [9] nd micrometl cutting theory [1]. To determine the criticl thickness t cr, cutting of the infinite plte is considered. s shown in Fig., there is uniform stress distribution long its unit width. This scheme C k k d c τ s ø 45 t 45 cr k B Fig. etermintion of the criticl thickness t cr σ s Tool is obtined from the norml cross section of the tool edge, where u-xis is sme to y-xis nd v-xis is norml to pttern side, in Fig. 1. Nkym nd ri nd Toropov nd Ko s experiments [7, 9] hve shown tht burr formtion in the feed direction during turning opertion is cused by the stresses in sher plne. ue to these stresses, tension is expected in the plte when the thickness is close to t cr. Slip lines B nd C re the boundries of the tensile re. Since the exit surfce is free from externl stresses, the slip lines re inclined to the exit surfce t 45, following theory of plsticity [1, 11]. Besides, this theory implies tht the norml nd sher stresses on these lines re equl to the vlue of plsticity k which is determined from Silv s criterion [1] s k ¼ p sy ffiffi ðþ 3 where σ y cn be obtined from Eq. 1. The force blnce of element BC with respect to the u-xis gives ðt s cos f þ sin fþ d c sin f 4kt cr kd c ¼ ð3þ with d c is undeformed chip thickness nd ø is sher ngle which cn be estimted using Merchnt s theory τ s f ¼ p 4 þ g ð4þ Groove sequence x y z Tool v In which α is rke ngle, g is friction ngle which cn be obtined by pplying Merchnt s circle. g ¼ rctn F v þ F u tn ð5þ F u þ F v tn with F u nd F v cn be estimted from F y nd F z which re lredy found in previous reserch [13]. Then from Eq., t cr cn be clculted t cr ¼ d cðt s cot f þ s s kþ ð6þ 4k where d c ¼ d i sin, d i being reltive cutting depth. In this cse, criticl thickness is lso clled burr thickness. u w ϕ Fig. 1 Prism model

3 Int J dv Mnuf Technol (1) 5: Fig. 3 Continuous side burr formtion Next depth Overlp depth d i tn H Feed direction H + d c d tn c H + tn H t cr d i d c Burr Shpe d i+1 d i+1 i+1. Continuous side burr formtion Burr hppens t ech cutting depth of the grooving schedule, ccording to the cutting condition. So burr should be counted t ech cutting step to form the finl burr size, but ccording to Fig. 3, fter some depths, burr counting should be strted from the overlpped depth t which burr is clled initil burr. The overlpped depth is considered when the groove width is equl or greter thn the designed pitch. p tn ð7þ where is totl depth t step n; p nd re pitch nd ngle of pttern, respectively. Initil burr might develop continuously to the finl depth to form finl burr s Fig. 3. This model is pplied for the cse in which burr is fully formed without frcture. Strting from the overlpped cutting depth, the micro plte mteril of criticl thickness is shifted perpendiculr to the tool edge to form the initil burr. H ¼ t cr ð8þ tn t the step next to overlpped depth, the initil burr is rotted round the tip of the tringle due to the rdius effect of the cutting edge in microcutting, nd the mteril t the bottom of the initil burr is gin shifted perpendiculr to the tool edge. Those two phenomen hppened t once to form the continuous burr. So fr from the next steps, burr is formed continuously s described till the lst step. H m ¼ H þ Xm d i sin tn sin ð9þ i¼nþ1 which is finl burr height..3 iscontinuous side burr formtion Considering gin bout the grooving step fter the initil burr formed, for generl mteril, there will be two cses of burr formtion. First is continuous burr formtion s described bove. Second is discontinuous burr formtion which will be further discussed in this section. s observing Fig. 3, discontinuous in this cse mens fter the overlpped depth burr of the previous depth will be cut off t the next depth. ccording to Nkym nd ri s experiments on brss in generl cutting [7], when sher strin is greter thn three (g>3), the deformtion is ccompnied by side flow which cuses the side burr. This is the condition bout plstic deformtion bility of mteril. In the cse of this study, even when the plstic deformtion tht needs for side burr to develop continu- d c tn 1 st direction d m Fig. 4 Side burr formtion t finl depth Fig. 5 Pyrmid model

4 57 Int J dv Mnuf Technol (1) 5: Fig. 6 Exit burr formtion () Initition (b) evelopment (c) Finl Formtion h C B dβ β ω β dx t C x γ β γ 1 B x ω β H Burr γf B ously is higher thn tht, it is cceptble to nlyze the discontinuous cse with this condition. In this cse, sher strin is estimted s lwys less thn three (g<3). It mens tht burr from the previous cutting depth is lwys removed t the next depth or the burr size of the scheduled grooving is burr size of the finl cutting depth H ¼ t cr tn which is shown in Fig Exit burr nd brek off in grooving micropyrmid pttern 3.1 Cutting model ð1þ Pyrmid pttern cn be formed in the sequence of first groove to crete prism pttern nd the second groove which is orthogonl with the first one. uring the second groove ech time when tool moves cross one prism pttern, there is exit burr or brek off t the edge s shown in Fig. 5. Inthiscse, rke ngle is zero so tht exit burr hppens in orthogonl cutting. Tking cross section in the pttern s Fig. 6 llows exploring the cutting process nd exit burr phenomenon. Exit burr mechnism cn be divided into three prts bsed on the observtion from the mchining tests on plsticine [14]. Initition: s the tool pproches the end of the workpiece, there is trnsition point t which the chip formtion stops nd plstic deformtion below the mchined surfce, long cutting direction, begins. Initition of burr formtion is chrcterized by the initil negtive sher ngle β nd initil distnce of tool tip ω. n interesting point is tht the initil negtive sher ngle, β, is lmost regrdless of the workpiece mteril nd the cutting conditions, when exit ngle is 9. This hs been verified by previous reserchers [14, 15]. evelopment: s the tool move forwrd fter initition from to 1, negtive sher plne lso rottes from B to B 1. Once the initil negtive sher plne is formed, the finl point of the negtive sher plne which crosses the exit surfce of the prism pttern t point B in Fig. 6 will ct s plstic hinge nd not trnslte during the burr development. Formtion: Finlly, the burr is formed with or without frcture becuse of incresing strin long the negtive sher plne s the tool pproching the end of workpiece. If frcture occurs long the negtive sher plne or through existing burr, it will remove or reduce the burr. 3. Exit burr formtion without brek off In the first stge of studying the exit burr of micropyrmid pttern, it is recommended to pply the model bout exit burr in generl cutting model of [15] following the cutting Fig. 7 Exit brek off model 1 R Brek off ε mx ε f dβ B β S ω

5 Int J dv Mnuf Technol (1) 5: ir chip suction b Silicon Smple (VK-97, Keyence) SEM (S-4, Hitchi) (VHX-6, Keyence) imond tool Fig. 8 Experiment setup. Turning copper mold with dimond tool nd chip suction. b Cpturing pttern with digitl microscope nd SEM schedule of the second groove. t ech depth, initil length is lso ssumed s burr thickness nd cn be obtined s 1 k s d c sin f þ sl sin f w ¼ ð11þ cos f k cos b þ S y 4 tn b with k is the yield sher strin obtined by Mise s criterion, k s is sher yield strength in sher zone, L is tool chip contct length, nd burr size cn be obtined geometriclly from Fig. 6 by B ¼ w tn b ð1þ H ¼ ðd c þ BÞsin g f ð13þ which B nd H re burr thickness nd burr height, respectively. 3.3 Exit brek off during burr formtion Beside the burr development, edge brek off during burr formtion gretly ffects the product qulity. If edge brekout occurs insted of burr formtion, crck long the negtive sher plne will be initited. It depends on the mteril behvior, prticulrly the frcture strin in this cse. frcture criterion is necessry for predicting the occurrence of frcture during burr formtion. While the criteri for initil yielding nd brittle frcture require only the current stte of stress, behvior of the ductile mteril is not clerly defined due to the extent of the plstic deformtion tht cn occur before frcture hppens: The behvior depends on the deformtion history in the ductile mteril. The behvior of most mterils is prcticlly.8 d.5µm d1µm dµm Burr Height ( µ m) Pttern ngle (deg) Fig. 9 Burr height in different pttern ngles with different finl depths Fig. 1 Ninety-degree prism pttern

6 574 Int J dv Mnuf Technol (1) 5: Tble 1 Single grooving, depth = pitch, V 8, 1 Hv v c epth m/min 5µm 1µm 15µm Burr.3µm.48µm.57µm 3 m/min Burr.7µm.45µm.53µm 4 m/min Burr.5µm.5µm.55µm locted between two extreme cses: fully ductile mteril nd perfectly brittle mteril which frcture without plstic deformtion, like crystl. In ductile mteril such s copper which lrge plstic deformtion occurs before frcture, it is not convenient to decide the frcture condition by using stress stte of the workpiece. It is suggested to use strin insted of stress to pply with the frcture criterion. Even in coted copper, less ductile mteril which is ccompnied by less plstic deformtion before frcture, the ductile frcture criterion by McClintock is ssumed to be pplied. pplying McClintock s criterion, Ko nd ornfeld s work [15] shows tht with pure copper using strin hrdening index.54 gives frcture strin in exit burr formtion ε f =.11, nd ccording to the work of Lin [16], the frcture strin ε f of copper lloy is 1.. So its suitble for pplying frcture strin ε f =1. in this study. The strin t the initil point is zero nd increses lmost linerly s the negtive sher ngle increses. So by monitoring the mximum sher strin of the re under grooved surfce, frcture cn be detected. If the mximum strin is greter thn the frcture strin then frcture hppens. " mx " f ð14þ 4 Experiment nd results Roll mold ws coted with copper nd grooved on precise turning center, s shown in Fig. 8, with single crystl dimond (SC) tool V shpe 6, 8, nd 9 ; hrdness is within the rnge of 1 to 7 Hv. First groove formed prism pttern nd second groove formed pyrmid pttern. Test cses were lso done following tht sequence nd cptured results by digitl microscope (Keyence VHX-6 nd VK-97) nd SEM (Hitchi S-4). 4.1 Prism pttern Figure 9 shows the reltion of burr height nd pttern ngle in cses of finl cutting depths re.5, 1, nd μm. ccording to the prediction, burr height grows proportionlly to the ngle decrement nd lso the increment of depth, nd this stisfies the previous experimentl works of Nkym nd ri bout generl burr [7]. Test cses re tken with SC tool V shpe 6, 8, nd 9 ; cutting speed v c, 3, nd 4 m/min; single nd multiple grooving. Theoreticlly, there is no burr when the pttern ngle reches to 9 s test cse in Fig. 1, but in other experimentl conditions somehow show tht burr still hppens. Tble 1 shows the surfce result fter grooving with single depth equl pitch, SC tool V shpe ngle 8, nd workpiece hrdness 1 Hv. Tble shows the results fter grooving with two depths, cutting speed 3 m/min, nd workpiece hrdness 1 Hv. Observing the surfce of prism Tble Two-step grooving, 3 m/min, 1 Hv ngle 8º (pitch = 15µm) Burr epth 1+5µm.3µm The mximum sher strin, ε mx, here cn be pproximted s the strin t the tool tip in Fig. 7, nd it cn be obtined s Z b pffiffi s ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi! 6 " tooltip ¼ " mx ¼ 3 b þ 3 tn 4 db ð15þ b b 6º (pitch = 13.µm) Burr.75µm

7 Int J dv Mnuf Technol (1) 5: m/min 3m/min 4m/min predict Forwrd Burr & epth (9 pttern 5Hv).5 4% Side Burr ( μ m) % 4% epth ( μm) (Pitch) b d = 15μm, = 9 b ω ( μm) m Hv c d = 15μm, 5Hv Fig. 11 Comprison between prediction nd experiment on side burr of prism pttern. Single grooving, depth = pitch, V 8, 1 Hv. b Two-step grooving, 3 m/min, 1 Hv pttern from bove, wvy lines t the top of ptterns which re shped by side burr cn be seen clerly. In order to compre the results with the prediction, n orthogrphic projection of burr height in Eq. 1 is tken. side burr ¼H sin ð16þ ω ( μm) m (deg) Figure 11 gives the comprison between prediction nd result from Tble 1 to Tble of side burr. For single groove, mx error is 5% nd min is 4%. For two-step grooving, mx error is 1% nd min is 3.7%. 4. Pyrmid pttern Some predicted results for exit burr of pyrmid pttern shown in Fig. 1 conclude tht exit burr size grows proportionlly to the decrement of hrdness nd lso the increment of cutting depth. In the reltion with pttern Fig. 1 Predicted results on exit burr of pyrmid pttern. Exit burr nd cutting depth, V 9, 5 Hv. b Initil distnce nd Hv. c Initil distnce nd pttern ngle ngle, burr size increses when ngle increse, ginst the rule in side burr of prism pttern. The reson comes from the geometry of the cutting model, s ngle increses then cutting depth on the side wll of the pyrmid increses so tht burr size increses. Those predictions gree with the rules bout burr formtion in generl cutting with previous reserches [7, 14, 15].

8 576 Int J dv Mnuf Technol (1) 5: ε ω f = f = m ω By the wy, in this study, the min results from the experiments re bout the exit brek off which re lso tken plce during the exit burr formtion. Test cses re tken with SC tool V shpe ngle 9 in cse of single grooving. Cutting speed for second groove is seprted into two groups: high speed 3 m/min with second groove follows turning direction, nd low speed 1, 6, nd 5 m/min with second groove follows feed direction, s shown in Fig. 14. The reltion of exit brek off size, shpe, nd frcture strin is predicted s Fig. 13. Figure 14 shows the comprison between prediction nd experiment on exit brek off in grooving pyrmid pttern. The predicted result for this cse gives brek off size 5.97 μm t negtive sher ngle 45 s Fig. 14. The mesured results from experiments with cutting speed t 3 nd 1 m/min give the brek off size 7.5 nd 6 μm (Fig. 14b, c). Fig. 13 Reltion of exit brek off nd frcture strin ε f β reserches. Methods of predicting side burr in prism pttern nd exit brek off in pyrmid pttern were suggested. Grooving prism nd pyrmid pttern on roll mold with V shpe SC tool, mesuring finished surfce in nd 3, were setup to check with the predicted results. In prism pttern, new concept of continuous nd discontinuous burr development through scheduled grooving ws proposed. Eqution 6 shows the criticl thickness where the cutting is impossible, or it cn be clled side burr thickness. Equtions 9 nd 1 give the totl side burr height nd burr height t finl depth, respectively. The errors from the predicted nd experimentl results in cse of two-step grooving re less thn 1%. In pyrmid pttern, exit burr thickness nd height cn be found by Eqs. 1 nd 13, but more importntly, mteril brek off model ws pplied into grooving pyrmid pttern. Exit brek off is predicted by compring the mx sher strin ε mx long the negtive sher plne, which is found by Eq. 15, with the frcture strin ε f of coted copper, nd the errors of predicted exit brek off compred to experiments re less thn 14.7%. The gp could be cused by lcking of cutting velocity effect on the burr formtion model while there re severl cutting speeds in the experiments. The burr formtion process is quite complex in generl nd mcro size nd even more complicted in microcutting. This work ttempts to develop new reserch field bout microburrformtioninmicrogrooving which is widely used in nowdys industry, especilly light trnsmission industry. Further work might worth to be crried out for the requirement of higher qulity micropttern. 5 Conclusion Microburr mechnism in grooving prism nd pyrmid pttern ws investigted ccording to previous burr cknowledgement The uthors wish to cknowledge the Ministry of Knowledge Economy of South Kore for supporting this work. b c 1 st groove nd groove (1m/min) ε β nd groove (3m/min) 7.5µm 1 st groove 6µm β (deg) Fig. 14 Comprison between prediction nd experiment on exit brek off in pyrmid pttern. Prediction of β till brek off occur with ε f =1.. b Brek off size with v c =3 m/min. c Brek off size with v c =1 m/min

9 Int J dv Mnuf Technol (1) 5: References 1. Che TJ, Prk SS, Freiheit T (6) Investigtion of micro-cutting opertions. Int J Mch Tools Mnuf 46: Lng W (1999) Reflexions on the future of microsystems. Sens ctutors 7: Mdou MJ (ed) () Fundmentls of microfbriction: the science of minituriztion (nd edn). CRC, Boc Rton 4. Corbett J, McKeon P, Peggs GN (9) Nnotechnology: interntionl developments nd emerging products. CIRP nn 49: Weck M, Fischer S, Vos M (1997) Fbriction of micro components using ultr precision mchine tools. Nnotechnol 8: Schller T, Bohn L, Myer J, Schubert K (1999) Microstructure grooves with width of less thn 5 µm cut with ground hrd metl micro end mills. Precis Eng 3: Nkym K, ri M (1987) Burr formtion in metl cutting. CIRP nn 36: Hirno, Skne M, Hmd N (7) Reltionship between Vickers hrdness nd inelstic mteril constnts. J Soc Mter Sci 56(5): Toropov, Ko SL (5) model of burr formtion in the feed direction in turning. Int J Mch Tools Mnuf 46: stkhov VP (ed) (1999) Metl cutting mechnics. CRC, Boc Rton 11. Chkrbrty J (ed) (1987) Theory of plsticity. McGrw-Hill, New York. ISBN-1: Silv V (ed) Mechnics nd strength of mterils. Springer, Berlin 13. Lee JM, Le, Kim SJ, Lee SW, Che TJ (9) evelopment of micro pttern cutting simultion softwre. Trns KSME 33:18 3. doi:1.3795/ksme Ko SL, ornfeld (1991) study on burr formtion mechnism. J Eng Mter Technol 113: Ko SL, ornfeld (1996) nlysis of frcture in burr formtion t the exit stge of metl cutting. J Mter Process Technol 58: Lin ZC, Lin YY (1) Three-dimensionl elstic plstic finite element nlysis for orthogonl cutting with discontinuous chip of 6-4 brss. Theor ppl Frct Mech 35: