Application of composite materials in sports optics

J uly 2020, Vol. 20, No. 2 M ANUFACTURING T ECHNOLOGY ISSN 1213–248 9 200 indexed on: h p://www.scopus.com DOI: 10.21062/m .2020.038 © 2020 Manu ac u ing Technology. All igh s ese ed. h p://www.jou nalm .com Applica ion o Composi e Ma e ials in Spo s Op ics Radim Kupčák, Jan Zouha Facul y o Mechanical Enginee ing, B no Uni e si y o Technology, Technická 2896/2 616 69, B no, Czech Re- public, E-mail: 170299@ u b .cz, [email p o ec ed] CFRP (Ca bon Fibe Rein o ced Polyme s) a e o en used when designing pa s ha need o be s i , ligh and he mally s able. These bene i s a e a big mo i a ion o use CFRP in many applica ions, one o hem could be spo s op ics. Howe e , op ical de ices equi e p ecise dimensions wi h igh ole ances o he op ical assembly o wo k co ec ly. In o de o de e mine i CFRP could be a sui able ma e ial o choice o spo s op ics a simpli ied body o binocula s was designed. The ubula body was manu ac u- ed by p ep eg lay-up in o a 3D p in ed mold, ollowed by cu ing in an au ocla e. A e he p o o ype was manu ac u ed 3D measu emen s o he ube using 3D scanne GOM ATOS we e made. As expec ed, sh inkage o he mold and he epoxy esin in he ma ix o CFRP caused mino de o ma ions. Howe e , i he shape o he cu ed pa emains unchanged du ing condi ions simila o he gene al use o binocu- la s, hen he ini ial de o ma ions happening du ing manu ac u ing could be accoun ed o when desig- ning he pa . Keywo ds: Ca bon Fibe , Spo s Op ic, P ep eg, Composi e Ma e ials, 3D scan In oduc ion Composi e ma e ials, speci ically CFRP (Ca bon Fibe Rein o ced Polyme s) ha e ecen ly become mo e and mo e accessible as a s uc u al ma e ial. They ha e become so wide-sp ead because o new echnologies ha make manu ac u ing and use o CFRP cheape and mo e con enien . These new ech- nologies o en o e high le els o au oma ion and sho p oduc ion imes. This allows o use o compo- si e ma e ials in mass p oduc ion indus y segmen s such as lowe -end ehicles p oduc ion, which has no been he case ea lie . Tha can lead o signi ican weigh educ ions, which could compensa e o he ba e y weigh in elec ic ehicles, making hei o e all weigh mo e easonable. [1] On he o he hand, he e a e many composi e ma- nu ac u ing echnologies based on hand lay-up. Al ho- ugh hey migh seem ou -da ed, hey a e well sui ed o manu ac u ing o la ge -scale pa s o small-se ies p oduc ion in gene al. CFRP is used in many b anches o enginee ing inc- luding consume goods and spo s equipmen wi h he main bene i s being weigh sa ings and igidi y. I may come as a su p ise, ha in spo s op ics composi e ma e ials ha e no ound a use ye . Bu compa ed o i ea ms o example, (whe e ca bon ibe is some i- mes used o make ails ocks o handgua ds), spo s op ics de ices equi e di e en kinds o p ecision – in some cases ole ances in mic ons. Ca bon ibe composi e has densi y o a ound 1,7 kg/m3, which is simila alue o magnesium alloys and lowe compa ed o aluminum alloys (2,7 kg/m3 o 6061). Howe e , CFRP has much highe s eng h. [2 ] A compa ison o composi e ma e ials and me als is no s aigh o wa d because o many a iables in de- sign o composi es. Because o ha i is no possible o p edic po en ial weigh -sa ings o CFRP binocu- la s. I would be a ec ed by he numbe o laye s, hei di ec ion and ype o ibe s and ma ix used. These pa- ame e s would ha e o be ca e ully chosen and hen es ed on a p o o ype piece. This pape has been made in coope a ion wi h Czech manu ac u e o op ical de ices (including spo s op ics), Meop a - op ics, L d. In Meop a ca bon ibe has been used as s uc u al ma e ial in many ap- plica ions, bu spo s op ics ha e no been one o hem ye . The pape ies o ackle he ques ion, whe- a e CFRP could be used a all o de ices o spo s op ics and analyses esul s o chosen manu ac u ing echnology. Binocula s ha e been chosen as a model de ice o he ca bon ibe o be used on. A simpli ied p o o ype was designed o de e mine i i would be possible o manu ac u e binocula s-shaped body ou o ca bon ibe a all. I lacks a ew key ea u es and componen s o binocula s – o example, op ical p isms o ocusing mechanism. Bu i wo ks as an exhibi ion piece and p oo o concep in e ms o manu ac u ing. The main pa o he p o o ype – ca bon ibe ube was 3D scanned wi h GOM ATOS – op ical 3D scan- ne . The p o o ype mold used o make he pa was scanned as well. The esul s we e compa ed wi h he CAD model and he de ia ions we e e alua ed. J uly 2020, Vol. 20, No. 2 M ANUFACTURING T ECHNOLOGY ISSN 1213–248 9 indexed on: h p://www.scopus.com 201 Design o he binocula s p o o ype In he beginning, i was necessa y o examine he possibili ies o manu ac u ing echnologies. Winding is a echnology ha was conside ed a i s . Because o i s na u e, i is mos ly sui able o pa s wi h ci cula symme y. Al hough i is possible o manu ac u e mo e complica ed shapes by winding, i o en b ings many p oblems. [3] The main ad an age o winding is p ecise con ol o ibe o ien a ion. I is c ucial o pa s wi h high s i ness and s eng h in one di ec ion. Howe e , o binocula s, all-a ound du abili y and e- sis ance o impac s, ib a ions and o he dynamic lo- ads, is mo e impo an . Žmindák [4] analyses he e ec o ibe o ien a ion on mechanical p ope ies o composi e ma e ials du ing dynamic loads. In he end, p ep eg lay-up was selec ed as he ma- nu ac u ing echnology o choice. I is mo e e sa ile in e ms o a ainable shapes and because an au ocla e was a ailable i was mo e cos -e ec i e. Ano he big a gumen o using p ep eg lay-up is he inse wi h p o usion, Fig. 1 and Fig. 7. I would be e y p oble- ma ic, i possible a all, o wind ca bon ibe s a ound he p o usion. Au ocla e cu ing allows o co-cu ing he pa wi h he inse , which is e y con enien and qui e a unique ea u e o his design. [5] The binocula s p o o ype (Fig. 1) consis s o wo CFRP ubes wi h 6061 aluminum alloy inse s. These inse s a e glued in he ubes and hey p o ide in e - ace o he objec i e and eyepiece o be moun ed. In he middle o he ubes, he e is he co-cu ed polyca - bona e inse (3D p in ed) wi h a cube-shaped p o u- sion. The inside s uc u e o he ube can be seen in Fig. 2. The p o usion ac s as an in e ace o he hinge pa s o be moun ed and glued on o he ubes. Fig. 1 Binocula s p o o ype Fig. 2 S uc u e o he binocula s p o o ype Manu ac u ing o he CFRP ube using p o o ype mold A p o o ype wo-piece nega i e mold om poly- ca bona e (PC) was designed o manu ac u e he ube [6]. The shape o he mold was compensa ed o ac- coun o he mal expansion o he mold ha happens a 120° C when he epoxy esin is cu ing. The olume ic he mal expansion o mula was used: ∆𝑉𝑉  𝛽∆𝑡 (1) 𝛽3𝛼 (2) Whe e: ∆V … Change in olume [m3], V0 … Ini ial olume [m3], β … Coe icien o olume ic expansion [K-1], ∆ … Change in empe a u e [K], α … Coe icien o linea expansion [K-1]. The coe icien α o PC is 68,4 x 10-6 K-1 [7]. The mold was 3D p in ed ou o polyca bona e using p o- essional-g ade 3D p in e S a asys Fo us 360 L (Fig. 3). The su ace o he mold was hand inished wi h a ine-g i sandpape . The alignmen o bo h pa s is secu ed wi h wo cen e ing pins ha i in holes, which we e inished wi h a eame . Gene al dimensions o he mold can be seen in Fig. 4. Fig. 3 Two-piece mold Be o e lay-up sealan and elease-agen we e ap- plied o hose su aces o he mold, which will be in con ac wi h p ep egs. Fo he lay-up, ou laye s o p ep egs om Del a P eg ha e been applied [8] – wo shee s o GG380 DT121 and wo shee s o GG245 DT121, Table 1. Laye s o ca bon ha e been layed up in bo h hal es o he mold o c ea e symme ic and balanced lay-up, Figu e 5. Whe e he wo pa s o he mold mee , he p ep egs we e o e lapped by 10 mm. Se e al o he lay-up s yles we e es ed in he p ocess. The pa was bagged, acuumed and cu ed in Ma oso au oma ic au ocla e (diame e o he chambe 1 m, leng h 2 m). The hea and p essu e cycles a e shown in Fig. 6. A e manu ac u ing 10 pieces o ubes, he 3D p in ed mold was a he end o i s li e. J uly 2020, Vol. 20, No. 2 M ANUFACTURING T ECHNOLOGY ISSN 1213–248 9 202 indexed on: h p://www.scopus.com Fig. 4 Mold dimension Tab. 1 Lay-up scheme Mold 245 g/m2 ±45° 380 g/m2 0°, 90° 380 g/m2 0°, 90° 245 g/m2 ±45° Peel Ply Sepa a ion laye B ea he clo h Vacuum bag Fig. 5 P ocess o lay-up A e cu ing, he ube needed o be inished. Faces o he ube we e milled and he emains o pa ing line we e sanded o . Milling composi e ma e ials equi es special ools wi h di e en geome y o ools ha a e used o machining me als. Because only 10 p o o ype ubes we e made, egula ools ha e been used. Be- cause o ha , ool wea was apid and he su ace i- nish wasn' sa is ac o y. The pa needed o be hand- inished. In p oduc ion on a g ea e scale, p ope ools and op imized cu ing speeds would ha e o be used. [9] Fig. 6 Au ocla e p essu e and empe a u e cycle Fig. 7 Finished ube wi h inse s glued in place J uly 2020, Vol. 20, No. 2 M ANUFACTURING T ECHNOLOGY ISSN 1213–248 9 indexed on: h p://www.scopus.com 203 The las s ep o he manu ac u ing p ocess was gluing he h eaded inse s (Fig. 7). Because he main pu pose o his p o o ype was being displayed as an exhibi ion piece, achie ing he concen ici y o he in- se s was no necessa y. Howe e , his would be a c u- cial pa ame e in eal, p ope ly wo king binocula s. Concen ici y could be secu ed wi h alignmen jig o by u ning he inse s a e gluing. 3.1 The p ocess o 3D scanning The mold and he ca bon ibe ube we e measu ed using an op ical 3D scanne GOM ATOS. The pa s we e sp ayed wi h an i-gla e sp ay. I lea es empo a y ma whi e inish. Sel -adhesi e e e ence poin s we e placed on he su ace o he pa . [10][11]. Fig. 8 3D scanning o he ube To ensu e e en scanning, he pa s we e placed on o a p og ammed u n able du ing scanning (Fig. 8 andFig. 9). Resul o he measu emen s was a polygo- nal mesh ha was expo ed as a STL ile. Fig. 9 3D scanning o he mold 3.2 E alua ion o da a The measu ed da a was e alua ed using so wa e: GOM Inspec and Mic oso Excel. In GOM Inspec a .STL ile was impo ed. The alignmen o he ac ual da a wi h he CAD model was made wi h Local Bes Fi on he conical pa o he ube. Acco ding o equi- emen s by Meop a o ge a comple e opog aphy o he mold, c oss-sec ions wi h in e als o 10 mm we e c ea ed ac oss he pa in GOM Inspec . In each c oss-sec ion measu emen s e e y 5 deg ees we e a- ken (Fig. 10). The analyzed mold in he GOM Inspec en i onmen can be seen in Fig. 11, esul alues discussed in chap e 4. Fig. 10 Measu ed c oss-sec ions and poin s o he mold J uly 2020, Vol. 20, No. 2 MANUFACTURING TECHNOLOGY ISSN 1213–248 9 204 indexed on: h p://www.scopus.com Fig. 11 Di e ence epo mold wi h CAD model Fo he ube, he same me hod was applied wi h one di e ence - he measu emen s we e aken a ound he ull ci cum e ence o he c oss-sec ions (Fig. 12). The c oss-sec ions o he pa and he mold align wi h each o he . The ube in he GOM Inspec en i on- men can be seen in Fig. 13. Fig. 12 Measu ed c oss-sec ions and poin s o he ube J uly 2020, Vol. 20, No. 2 M ANUFACTURING T ECHNOLOGY ISSN 1213–248 9 indexed on: h p://www.scopus.com 205 Fig. 13 Di e ence epo ca bon ibe ube wi h CAD model Resul s In he ollowing chap e , he esul s a e analysed and discussed. 4.1 Mold Only a small sample o measu ed poin s is eco - ded in able 2. Fig. shows he chosen coo dina e sys- em in e ms o he angle alue o he poin s in each c oss-sec ion. Poin s wi h ± 90° angle we e excluded om he sample because o he cham e on he edge o he mold. The a i hme ic mean c oss-sec ion is calcula ed om all measu ed c oss-sec ions, no only he ones eco ded in able 2. Fig. 14 Dimensioning o measu ed poin s on he mold Tab. 2 De ia ions o poin s o he mold Angle [ ° ] De ia ion [ mm ] o a ious c oss-sec ion 25 mm 65 mm 95 mm 125 mm 155 mm A i hme ic mean c oss-sec ion -85 -0.20 -0.31 -0.28 -0.31 -0.38 -0.294 -80 -0.16 -0.27 -0.21 -0.26 -0.33 -0.246 -75 -0.15 -0.21 -0.15 -0.20 -0.28 -0.198 -70 -0.13 -0.18 -0.10 -0.14 -0.24 -0.162 -65 -0.09 -0.15 -0.05 -0.10 -0.23 -0.130 . . . . . . . . . . . . . . . . . . . . . 55 -0.02 -0.07 -0.09 -0.03 -0.12 -0.076 60 -0.06 -0.11 -0.12 -0.10 -0.20 -0.124 65 -0.07 -0.16 -0.16 -0.19 -0.31 -0.178 70 -0.10 -0.20 -0.20 -0.26 -0.39 -0.226 75 -0.13 -0.22 -0.26 -0.31 -0.47 -0.269 80 -0.15 -0.25 -0.31 -0.39 -0.55 -0.317 85 -0.20 -0.28 -0.37 -0.43 -0.62 -0.370 J uly 2020, Vol. 20, No. 2 M ANUFACTURING T ECHNOLOGY ISSN 1213–248 9 206 indexed on: h p://www.scopus.com Fig. 15 plo s alues o de ia ions om able 2. La ge de ia ions o alues abo e ± 50 ° can be ob- se ed. I sugges s signi ican wa page o he mold in he a ea nea he pa ing line. The mold had endency o close in on i sel . Because he 3D scanning was ca ied ou a e manu ac u ing 10 CFRP pa s in he mold, i is no clea whe he he wa page happened di ec ly a e 3D p in ing o was a esul o hea cycles in he au ocla e. The wa page o “C-shaped” pa s is common due o s ess dis ibu ion in hese pa s. As a esul , mea- su ed cylind ici y on he wide end o he mold (Fig. 11) is 0.24 mm. Fig. 15 De ia ion o measu ed poin s in di e en sec ions As can be obse ed on he “A e age c oss- sec ion” - alues he ca i y is on a e age smalle han he nominal alue om CAD da a (0 on he y-axis in he plo ). This is due o olume compensa ion o he shape o he mold. which was ca ied ou based on he au ocla e hea cycle (Fig. 6). 4.2 Tube The ube was measu ed in a simila way o he mold (Fig. 16) – sample da a o da a is eco ded in a- ble 3. Fig. 16 Dimensioning o measu ed poin s on he ube Tab. 3 De ia ions o poin s o he ube Angle [ ° ] De ia ion [ mm ] o a ious c oss-sec ion 35 mm 65 mm 95 mm 125 mm 145 mm A i hme ic mean c oss-sec ion -85 0.07 -0.01 0.04 -0.04 -0.19 0.000 -80 0.08 0.03 0.11 0.02 -0.14 0.038 -75 0.08 0.07 0.16 0.08 -0.08 0.071 -70 0.09 0.10 0.2 0.12 -0.05 0.095 -65 0.12 0.12 0.25 0.16 -0.02 0.130 . . . . . . . . . . . . . . . . . . . . . 55 0.11 0.08 0.10 0.25 0.26 0.154 60 0.10 0.07 0.06 0.21 0.20 0.127 65 0.10 0.05 0.02 0.16 0.12 0.087 70 0.10 0.04 -0.02 0.10 0.06 0.056 75 0.11 0.02 -0.07 0.05 0.00 0.032 80 0.11 -0.02 -0.06 0.00 -0.08 0.006 85 0.11 -0.04 0.00 -0.01 -0.12 0.002 In Fig. 17, da a om able 2 a e plo ed. As expec- ed, he shape o he lines closely esembles he su ace o he mold. This esul sugges s, ha he quali y o he ube would be highly imp o ed i me al machined mold had been used. Di ec compa ison o a e age c oss-sec ions o he 3D p in ed p o o ype mold and he ube is shown in Fig 18. Dha al and Pa el p esen a compa ison o pa manu ac u ed by in usion and p ep eg lay-up in hei pape [12]. In hei case, he p ecision o he in used pa was g ea e han he pa made om p ep egs. The e is no appa en sp ing-back o he pa due o i s ubula shape. Pape by Dha al and Gajja [13] deals wi h sp ing-back o “L-shaped” pa s. J uly 2020, Vol. 20, No. 2 M ANUFACTURING T ECHNOLOGY ISSN 1213–248 9 indexed on: h p://www.scopus.com 207 As he plo s show, he a e age c oss-sec ion o he ube has de ia ions consis en ly abo e he nominal a- lue o he CAD da a. This sugges s, ha he he mal compensa ion men ioned in Chap e 3 does no di- ec ly co espond wi h he eal si ua ion. The ac ha he de ia ion o he ube is g ea e , han he de ia ion o he mold caused di icul y when emo ing he pa om he mold. Fig. 17 De ia ion o measu ed poin s in di e en sec ions Fig. 18 Compa ison o de ia ions o a e age c oss-sec ions o he mold and ube The Following esea ch Mo e esea ch needs o ollow up in o de o examine he use o CFRP in he ield o spo s op ics in a comp ehensi e way. E en mo e impo an han he ini ial p ecision o manu ac u ing is he he mal s abili y o he pa i sel and he bonds wi h o he componen s. I is impo an ha adhesi e ha is used can wi hs and clima ic con- di ions a which binocula s a e es ed. Because o ha empe a u e s abili y es o adhesi e join s es s ha e been designed. Fig. 19 Samples on he alignmen jig Fig. 20 Adhesi e es sample Measu ed samples consis o wo me al cylinde s ha a e glued oge he on a jig o ensu e concen ici y Fig. 19 and Fig. 21. The gap ha is illed wi h glue is 3 mm wide. Because o ha , he long- e m concen ici y o he inne and ou e ube is de e mined by he s abi- li y o he glue. In he u u e, a ious sizes o he gap will be es ed. When he adhesi e is ully cu ed con- cen ici y o he ubes is measu ed on CMM Zeiss Con u a G2 o ge a e e ence Fig. 21. Fig. 21 Measu emen o he adhesi e es sample Tempe a u e es s ollow. The samples will be pla- ced in a clima ic chambe ( ange o empe a u es will be es ed). Concen ici y is measu ed a e e e y es . The i s se o es s has he samples s anding on hei aces. Tes s in he ho izon al posi ion will ollow o J uly 2020, Vol. 20, No. 2 M ANUFACTURING T ECHNOLOGY ISSN 1213–248 9 208 indexed on: h p://www.scopus.com ge esul s wi h di e en ypes o s ess on he glue. The es s a e s ill in p og ess and he esul s a e no eady ye . Ca e ul choice o manu ac u ing echnology o he ubula CFRP pa would be ano he poin o in- e es in u u e wo k. This p oblem needs o be ackled om mul iple poin s o iew. One o hem is o examine which echnology b ings pa s wi h he bes mechanical p ope ies ega ding s i ness and s eng h. Kulha ý deals wi h his p oblema ic in his pape s, in his wo k he compa es expe imen s wi h nu- me ical simula ions as well as compe ing di e en ma- nu ac u ing echnologies [14][15]. Ano he p oblem is he p ecision and s abili y o CFRP pa s. The e is no much esea ch ha has been made a his momen , he impac o manu ac u ing echnology on p ecision and s abili y needs o be quan i ied by expe imen s made wi h a la ge sample o pa s. Conclusion The main pu pose o his pape was o design and manu ac u e a p o o ype sample using a CFRP manu- ac u ing echnology. I se ed as a p oo o concep in e ms o manu ac u ing a binocula s-like body using CFRP. The wo-piece p o o ype mold and he CFRP ube we e 3D scanned using op ical 3D scanne GOM ATOS. The measu ed da a we e e alua ed and com- pa ed o he nominal CAD model. As expec ed, he p ecision o he mold s ongly a ec s he inal p ecision o he manu ac u ed pa . The wo-piece nega i e p o o ype mold men ioned in his pape , which was made om polyca bona e, showed a wa page de o ma ion and as a esul , he ube pa was na owe nea he pa ing line. As ex- pec ed, he opog aphy o he su ace o he inal pa closely esembles he opog aphy o he su ace o he mold. The a e age de ia ion o he mold is nega i e – meaning ha he mold is smalle han expec ed. I is caused by ela i ely high coe icien o he mal ex- pansion o polyca bona e. Du ing he condi ions in au ocla e – 6 ba and 120 °C he ube expanded and a e cooling o he oom empe a u e sh inked. A e cu ing he pa was di icul o emo e om he mold as a esul o small cu e sh inkage o he epoxy esin. These impe ec ions a e accep able when conside- ed he ma e ial and echnology o he manu ac u ing o he mold. FDM is p ima ily an al e na i e echno- logy o p o o ype p oduc ion in he case o molds. Men ioned p oblems would be elimina ed i machined me al mold had been used. As men ioned in he chap e “The ollowing e- sea ch” he he mal s abili y o pa s and join s may be mo e c ucial, he e o e i will be he ocus o u u e es s. Acknowledgemen This esea ch wo k was suppo ed by he B no Uni e si y o Technology, Facul y o Mechanical Enginee ing, Spe-ci ic esea ch 2019, wi h he g an " Resea ch o pe spec i e p oduc ion ech- nologies ", FSI-S-19-6014. Re e ences LOSCH, D.; EKANAYAKE, S.; NIENHEYSEN, P.; BETHLEHEM- EICHLER, K.; SCHMITZ, R.; BIER, S.; ROSSMANN, J. (2017); An Indus y 4.0-based epai concep o s uc u al CFRP compo- nen s in he au omo i e sec o , 2nd IEEE In e - na ional Symposium on Robo ics and Manu ac u ing Au oma ion, Roma, A icle numbe 7847801 COMPOTECH [4.11.2019], h ps://compo- ech.com/ echnology/composi e-cha ac e is- ics/, Composi e cha ac e is ics, Compo ech, 2019 CAMPBELL, F. (2010) S uc u al composi e ma e- ials. Ma e ials Pa k, Ohio: ASM In e na ional. ISBN 16-150-3037-9. ŽMINDÁK, M.; PELAGIC, Z.; SOUKUP. (2015); Analysis o ibe o ien a ion in luence o dynamic p ope ies o composi e s uc u es, Manu ac u ing Technology, Vol 15, p. 490-494 CAMPBELL, F. C. (2004). Manu ac u ing p oces- ses o ad anced composi es. New Yo k: Else ie , 2004. ISBN 978-185-6174-152. HADAŠ, Z.; ZOUHAR, J.; SINGULE, V.; ONDRŮŠEK, Č. (2008) Design o Ene gy Ha es ing Gene a o Base on Rapid P o o y- ping Pa s. 13 h Powe Elec onics and Mo ion Con- ol Con e ence. 1. Poznan, POLAND: s. 1688- 1692. ISBN: 978-1-4244-1741- 4 STRATASYS [4.11.2019], h ps://www.s a- asys.com/-/media/ iles/ma e ial-spec-she- e s/mss_ dm_pc_1117a.pd , PC (Polyca bo- na e) ma e ial spec shee , STRATASYS, 2017 DELTA PREG, [4.11.2019], h p://www. anso.se/wp-con en /uplo- ads/2017/03/Ca alogA4_2017Feb.pd , Del a- P eg Ca alogue, Del a-P eg, 2017 BÍLEK, O.; ŽALUDEK, M.; ČOP, J. (2016). Cu ing ool pe o mance in end milling o glass ibe - ein o ced polyme composi es, Manu- ac u ing Technology, Vol 16, p. 12-16 ZOUHAR, J.; PÍŠA, Z.; SEDLÁK, J.; SEDLÁČEK, J. (2007) P oduc i e machining wi h e e se enginee inng me hods. In F ézo ání