CONTENTS

WELDING PRODUCTION CHAIR OF NTUU «KPI» IS 80The 80th anniversary of the Chair of Welding Production ofNTUU «Kiev Polytechnic Institute» .................................................. 2

Kuznetsov V.D. and Stepanov D.V. Structure andproperties of weld metal modified by nanooxides .......................... 10

Kvasnitsky V.V., Ermolaev G.V. and Matvienko M.V. Effectof cooling mode after diffusion welding and brazing onresidual stresses in graphite—copper edge joints ........................... 17

Slivinsky A.A., Zhdanov L.A. and Korotenko V.V.Thermal-physical peculiarities of gas-shielded pulse-arcwelding using non-consumable electrode (Review) ....................... 24

Grinyuk A.A., Korzhik V.N., Shevchenko V.E., Babich A.A.,Peleshenko S.I., Chajka V.G., Tishchenko A.F. andKovbasenko G.V. Main tendencies in development ofplasma-arc welding of aluminium alloys ........................................ 31

Pashchenko V.N. Application of N—O—C—H gas systems forsynthesis of strengthening components in plasma coatings ........... 42

INDUSTRIALChervyakov N.O. Evaluation of thermal stressed state inwelded joint of alloy Inconel 690 ................................................... 48

Atroshenko M.G., Poleshchuk M.A., Shevtsov A.V., PuzrinA.L., Mishchenko D.D., Serebryanik I.P. and Borodin A.I.Physical and mechanical properties of transition zone ofbimetal produced by autonomous vacuum brazing ofcopper on steel ........................................................................... 52

Moltasov A.V., Gushchin K.V., Klochkov I.N., Tkach P.N.and Tarasenko A.I. Determination of force caused byheating of ring-type products in flash-butt welding ........................ 57

Olejnik O.I. Influence of metal shrinkage in longitudinalwelds of sleeves on contact pressure in main gas pipelinerepair .......................................................................................... 61

NEWSNew welding wire manufacturer in Ukraine .................................... 64

English translation of the monthly «Avtomaticheskaya Svarka» (Automatic Welding) journal published in Russian since 1948

EDITORIAL BOARD

Editor-in-Chief B.E. PatonScientists of PWI, Kiev

S.I. Kuchuk-Yatsenko (vice-chief ed.),V.N. Lipodaev (vice-chief ed.),

Yu.S. Borisov, G.M. Grigorenko,A.T. Zelnichenko, V.V. Knysh,

I.V. Krivtsun, Yu.N. Lankin,L.M. Lobanov, V.D. Poznyakov,I.A. Ryabtsev, V.F. Khorunov,

K.A. YushchenkoScientists of Ukrainian UniversitiesM.N. Brykov, ZNTSU, ZaporozhieV.V. Dmitrik, NTU «KhPI», Kharkov

V.F. Kvasnitsky, NUS, NikolaevV.D. Kuznetsov, NTUU «KPl», Kiev

Foreign ScientistsN.P. Alyoshin

N.E. Bauman MSTU, Moscow, RussiaGuan Qiao

Beijing Aeronautical Institute, ChinaA.S. Zubchenko

DB «Gidropress», Podolsk, RussiaM. Zinigrad

College of Judea & Samaria, Ariel, IsraelV.I. Lysak

Volgograd STU, RussiaYa. Pilarczyk

Welding Institute, Gliwice, PolandU. Reisgen

Welding and Joining Institute,Aachen, Germany

O.I. SteklovWelding Society, Moscow, Russia

G.A. TurichinSt. Petersburg SPU, Russia

FoundersE.O. Paton Electric Welding Institute, NASU

International Association «Welding»

PublisherInternational Association «Welding»

TranslatorsA.A. Fomin, O.S. Kurochko,

I.N. KutianovaEditor

N.A. DmitrievaElectron galley

D.I. Sereda, T.Yu. SnegiryovaAddress

E.O. Paton Electric Welding Institute,International Association «Welding»

11, Bozhenko Str., 03680, Kiev, UkraineTel.: (38044) 200 60 16, 200 82 77Fax: (38044) 200 82 77, 200 81 45

E-mail: journal@paton.kiev.uawww.patonpublishinghouse.com

State Registration CertificateKV 4790 of 09.01.2001

ISSN 0957-798X

Subscriptions$348, 12 issues per year,

air postage and packaging included.Back issues available.

All rights reserved.This publication and each of the articles contained

herein are protected by copyright.Permission to reproduce material contained in this

journal must be obtained in writing from thePublisher.

November 2015No.11

Published since 2000

E.O. Paton Electric Welding Institute of the National Academy of Sciences of Ukraine

International Scientific-Technical and Production Journal

© PWI, International Association «Welding», 2015

THE 80th ANNIVERSARY OF THE CHAIROF WELDING PRODUCTION

OF NTUU «KIEV POLYTECHNIC INSTITUTE»

In 1935, on the initiative of Evgeny O. Paton,the outstanding scientist and engineer, the aca-demician of the UkrSSR, the Chair of WeldingProduction was founded at Mechanical Facultyof Kiev Polytechnic Institute, and preparationof mechanical engineers on specialty «Equipmentand Technology of Welding Production» was or-ganized. The need in education of the specialistsof such a profile was caused by the requirementsof intensive development of the domestic indus-try, transport, construction and many otherbranches of the national economy. By transfer-ring the students from other specialties the firstgroups of students were formed for educating thewelding engineers at the second and third yearsof study. The first graduation of engineers ofwelding specialty (17 persons) took place in1938, and within 3 years 104 welding specialistswere educated and trained.

Many research workers of the Electric Weld-ing Institute of the Academy of Sciences of theUkrSSR, organized in 1934, took an active partin the educational process at the Chair in thoseyears, in particular, V.I. Dyatlov, V.V. Shever-

nitsky, A.M. Sidorenko, F.E. Sorokovsky. Forteaching the profile disciplines the staff membersfrom other chairs of the Institute were also in-vited. Among them the Ass. Profs I.P. Trochunand N.V. Pines, the Assistant G.K. Blavdzevich,and M.M. Bort, the Chief metallurgist of theaircraft plant. The required number of rooms foreducational process and research works was al-located for the Chair. At that stage the formationof new engineering specialty «Equipment andTechnology of Welding Production» took place.For the first time the methodological grounds foreducation and training of welding engineers weredeveloped, including an integrated system of edu-cation based on close relation between the edu-cational institution and production.

From the beginning of its foundation the Chairwas headed by the Academician E.O. Paton.However, his high duties as the Director of Elec-tric Welding Institute impeded his work in KPI,and in 1938 he was forced to leave the institution,further he still continued to render a comprehen-sive assistance to the Chair, permanently em-bodying the originally laid idea of affinity andcooperation of both organizations.

Evgeny Paton was an outstanding scientistand teacher, and that fact had a rather beneficialeffect on the quality of preparation of the weldingspecialists. In those years as the basis of educatingthe welding engineers E.O. Paton laid the im-portant methodological principles consisting inneed of combining the theoretical and practicaltraining of students, subordinating the contentof the training specialists to the practical prob-lems of welding production, high requirementsto the work of students over the educational ma-terial, in extensive use of modern achievementsof science and technology of welding for educa-tion. The scientific and pedagogical staff of theChair carefully preserves these principles evenat the present time.

In 1938, the Ass. Prof. V.L. Ulasik becamethe head of the Chair. The outbreak of the warin 1941 interrupted education of welding special-ists. KPI was evacuated to Tashkent. Manyteachers, staff, postgraduate and undergraduatestudents were mobilized, some of them moved tothe plants producing defensive products. Due to

Evgy O. Paton, the founder of the Chair, and his sonsVladimir (left) and Boris (right) near new universal auto-matic welding machine TS-17 (1949)

2 11/2015

the absence of teaching personnel the educationof welding production engineers in Tashkent wasnot carried out.

In 1944, after return of KPI from evacuationto Kiev, the Chair of Welding Production imme-diately resumed its activity. From July to No-vember 1944 the Cair was headed by Prof. G.I.Pogodin-Alekseev, and then from the end of 1944to 1947 by the Ass. Prof. I.P. Trochun, and fora short period in 1947, by the Ass. Prof. M.N.Gapchenko and the Assistant M.M. Bort. Therewas a strong lack of qualified scientific and peda-gogical staff. Nevertheless in short terms the staffmembers and the students of the Chair managedto restore the classrooms and laboratories de-stroyed during the war. The Chair was equippedwith the necessary welding equipment. As a re-sult, in fact, the new educational and laboratoryfacilities were created with total area of about700 m2. The first post-war graduation of me-chanical engineers on specialty «Equipment andTechnology of Welding Production» was held in1947. It became possible due to the great contri-bution of Prof. K.K. Khrenov, who headed theChair from 1947 to 1957. After K.K. Khrenovthe Chair was headed by Prof. I.P. Trochun(1957—1967) and Prof. V.I. Dyatlov (1967—1969). Since 1965, at the Chair the educationand training of electromechanical engineers onspecialty «Electrothermal Equipment» with the

specialization in electric welding installationswas opened.

In 1969, to the position of the Head of theChair (as a combining job) Prof. B.S. Kasatkin,the Head of Department of the Electric WeldingInstitute, was invited, who headed the Chair un-til June 1972. This event contributed to the sig-nificant widening and strengthening of compre-hensive and creative cooperation between theChair and the E.O. Paton Electric Welding In-stitute, especially in the field of scientific re-search. The contractual research works and train-ing of scientific personnel were essentiallyboosted through postgraduate courses and com-petition. At the same time, the difficulties asso-ciated with the limited educational, laboratoryand scientific base of the Chair were noticeablemore significantly.

In the early 1970s, the Government ofUkraine took the decision to found the Educa-tional Center for the joint training and improvingthe qualification of welding engineers on the fa-cilities of KPI and PWI. The new educationaland laboratory building (now the 23rd buildingof the National Technical University of Ukraine«KPI») of 6000 m2 area was built. In September,1977 the Chair moved to the new educationaland laboratory facilities, equipped with the mod-ern equipment and devices. For educational workthe PWI leading scientists G.I. Leskov, A.I.

Unveiling of monument Evgeny Paton at NTUU «KPI»

11/2015 3

Chvertko, A.G. Potapievsky, V.R. Ryabov, V.N.Zamkov, V.E. Moravsky, V.I. Makhnenko, B.A.Movchan, A.A. Rossoshinsky et al. were invitedto the Chair on the terms of a combining job.

Since 1972 to 1974, the Chair was headed byProf. M.N. Gapchenko, and since 1974 it washeaded by Prof. I.R. Patskevitch.

From 1962 to 1978, under the supervision ofV.P. Chernysh the works on creation of methodsof magnetic control of solidification of weld poolmetal (V.P. Chernysh, V.V. Syrovatka, I.V.Malinkin, V.D. Kuznetsov et al.) were carriedout at the Chair. The obtained results confirmedthe possibility of active influence on the processof primary solidification by electromagnetic stir-ring of weld pool and showed that stirring leadsto refinement of weld metal structure, and changein depth and shape of penetration.

In 1978, from the Chair of Welding Productionthe Chair of Welding Equipment was detachedunder the directorship of Prof. V.P. Chernysh,where the works on the control of solidification ofweld pool metal by the magnetic influence werecontinued, further developed in the works ofProf. R.N. Ryzhov, Dr. V.A. Pakharenko et al.

From May, 1989 to April 27, 2015, the Chair ofWelding Production was headed by Prof. V.M.Prokhorenko (the graduate from the Chair in 1962).

In 1991, from the Chair of Welding Productionthe Chair of Restoration of Machine Parts wasdetached, which by that time was headed by theAss. Prof. V.M. Dukhno, and from October, 1993it was headed by Prof. V.N. Korzh. It was chargedwith preparation of specialists on specialty «Equip-ment and Technology for Improving Wear Resis-tance of Machine Parts». At present time the Chairis called the Chair of Surface Engineering and isheaded by Prof. V.D. Kuznetsov.

On the basis of the Chair of Welding Produc-tion in 1975 the Welding Faculty resumed itswork, whose dean until 2002 was Prof. A.M.Slivinsky, the graduate of the Chair of WeldingProduction in 1960. Since 2002 until now theWelding Faculty is headed by Prof. S.K.Fomichev, who headed the Chair of ElectricWelding Equipment since 2001.

Today the Faculty consists of three specialchairs: Welding Production, Electric WeldingEquipment and Surface Engineering. The Fac-ulty prepares bachelors in the direction «Weld-ing», as well as specialists on «Equipment andTechnology of Welding Production», «WeldingMachines», «Equipment and Technology of In-creasing Wear Resistance and Restoration of Ma-chine Parts», and masters in the field of weldingscience and technology on the basis of bachelorpreparation. Currently, about 400 students arestudying at the Faculty. Besides the education

of students of the Faculty the scientists carry outa wide range of scientific research on the mostrelevant and advanced directions.

On the territorial facility of building of theWelding Faculty, the Welding Training Centerwas founded in 1977, where the UNIDO workshopand courses for improving the qualification of do-mestic specialists in the field of welding and non-destructive testing of welded joints and trainingcourses for welders-workers were opened. TheChair of Welding Production takes also an activepart in the work of the Educational Center. At themoment this direction is actively realized in theprogram of double diploma – International Weld-ing Engineer (IWE), International Welding Tech-nologist (IWT) and International Welding Inspec-tor (IWI) (according to standard DSTU ISO14731:2008). This diploma is realized on the basisof the Welding Faculty and the teachers of thechairs of the Welding Faculty form the workinggroup on training specialists on the basis of labo-ratories of the Chair.

At the end of 1994, on the initiative of A.M.Slivinsky, the Dean of the Welding Faculty, andthe Ukrainian Welding Society according to thegeneral resolution of the National Academy ofSciences of Ukraine and the Ministry of Educa-tion of Ukraine, the Ukrainian Committee forWelder Certification (UCWC) was founded. Itstechnical director was V.T. Kotik, the Ass. Prof.of the Chair of Welding Production. During theperiod of the UCWC activity, more than 700experts were trained and certified, and more than230 commissions on certification of welders wereopened in Ukraine. The database on certifyingcommission and certified welders (annually cer-tified about 10,000 persons), created by theUCWC, provide a rapid control of the certifica-tion processes as well as obtaining the informa-tion on the status and dynamics of developmentof welding production in Ukraine. Thus, the sys-tem of welders certification in Ukraine demon-strated its necessity, relevance and effectiveness,providing the high quality of welding works.

Since the moment of its foundation the re-search work gained the extensive developmentat the Chair. The postgraduate courses were or-ganized and successfully operate. The postgradu-ate students and many staff members of the Chairdefended their theses for candidate of technicalsciences degree. The teachers of the Chair pub-lished a number of books, which received a widerecognition. At the Chair the steadily developingresearch directions were formed. In particular,the staff members of the Chair under the generalsupervision of Prof. K.K. Khrenov carried out anumber of works related to the investigation ofwelding arc (V.E. Moravsky, G.B. Serdyuk,

4 11/2015

G.V. Vasiliev, L.A. Byalotsky). In 1949, K.K.Khrenov published fundamental monograph«Electric welding arc», which became the firstsummarized work on the subject. These workson study of arc discharge were continued on anew qualitative level with the use of electronicoscillography of arc in order to study the tran-sient processes (L.A. Zhdanov, V.L. Kovalenko).The results of these investigations are reflectedin the thesis for Cand. of Tech. Sci. Degree ofV.L. Kovalenko (2013). As a result, a new com-plex criterion for evaluation of stability of exist-ence of arc discharge was proposed, which in-cludes power and technological characteristics ofthe arc, the features of existence of alternatingcurrent arc were explained.

The other demanded direction of research ac-tivity of the Chair was the creation of new origi-nal slag systems for ceramic fluxes and develop-ment of their compositions intended for weldingand surfacing, as well as the technology of manu-facture of ceramic fluxes (D.M. Kushnerev, I.M.Zhdanov, M.P. Grebelnik et al.). The results ofinvestigations on the development and applica-tion of ceramic fluxes are described in monographof K.K. Khrenov and D.M. Kushnerev «Ceramicfluxes» published in 1954. The works were con-tinued in 1970—2015 by A.M. Slivinsky, V.N.Kopersak, V.I. Prokhorov, V.T. Kotik, O.A.Gaevsky, L.A. Zhdanov and N.M. Strelenko.During that time the unique ceramic flux forwelding the metal after oxy-fuel cutting (A.M.Slivinsky, V.T. Kotik), fused flux AN-44 for

welding of low-carbon steels of increasedstrength (A.M. Slivinsky, V.M. Prokhorov, B.N.Kopersak), fused flux AN-69 for surfacing (A.M.Slivinsky, L.A. Zhdanov), agglomerated fluxANK-45 (jointly with the PWI Department) andflux ANK-73 for surfacing (L.A. Zhdanov, N.M.Strelenko) were developed. The results of thework are reflected in numerous publications andprotected by the copyright certificates and pat-ents of Ukraine.

The development of original technologies forsurfacing of cutting tool, worn-out parts of ma-chines and mechanisms are also an important di-rection of work of the Chair of Welding Produc-tion. The works of V.D. Kuznetsov, N.A. Gor-penyuk, Yu.A. Yuzvenko, M.S. Samotryasov andB.N. Gorpenyuk should be noted.

The important works on theoretical problemsof welding were performed under the supervisionof V.I. Dyatlov, who defended the thesis for Dr.of Tech. Sci. Degree on these directions in 1963.They include original developments in the fieldof theory of freely expanding and contractedwelding arc, calculations of conditions of auto-matic submerged arc welding, electrode metaltransfer and a number of other investigations.The theoretical models of V.I. Dyatlov on arcprocesses, electrode metal transfer, metallurgicalinteraction in submerged arc welding did notloose their relevance even today. Prof. V.I. Dyat-lov dealt not only with fusion welding, but alsowith solid-state welding. At the Chair under hissupervision the thesis for Cand. of Tech. Sci.

Staff members of the Chair of Welding Production (September, 2015)

11/2015 5

Degree on diffusion welding in the glow dis-charge of magnetic and nonmagnetic steels wasmade by D.I. Kotelnikov, who later defendedthe thesis for Dr. of Tech. Sci. Degree on thesame subject.

Under the supervision of M.N. Gapchenko atthe Chair a number of works was carried out aimedat the study of weldability, brittle fracture of met-als and modernization of technology of weldingsteels and alloys. The results of investigations aresummarized in monograph of M.N. Gapchenko«Brittle fractures of welded joints and structures»,as well as in his thesis for Dr. of Tech. Sci. Degreedefended in 1969. Later the investigations of tech-nological strength and weldability for nickel al-loys, stainless and high-strength steels were con-tinued by A.A. Slivinsky. The results of these in-vestigations in 1998—2015 were published in nu-merous articles, including foreign ones, and theywere also included into educational book «Weld-ability of structural materials».

The works in the field of gas-flame treatmentof metals, carried out under the supervision ofM.M. Bort and A.D. Kotvitsky, gained a greatdevelopment in 1950—1960. Especially the oxy-gen cutting of metal of large thicknesses at lowpressure, pack cutting of metal, development ofthe whole series of installations for cutting theround profile, hot crops of steel casting, hot roll-ing, etc. should be noted. The fundamentallynew types of cutters were created, which founda wide application in industry. The developmentof plasma welding, surfacing and cutting of ma-terials was headed by M.N. Gapchenko. The sig-nificant results were obtained under the supervi-sion of V.N. Korzh in the field of creation of equip-ment and technologies for welding using hydro-gen-oxygen flame. The results of these investiga-tions formed the basis of the thesis for Dr. of Tech.Sci. Degree of V.N. Korzh defended in 1991.

The works in the field of welding stresses anddeformations were developed under the supervi-sion of I.M. Zhdanov and I.P. Trochun, whodeveloped the simple and descriptive methods forengineering calculations of welding stresses anddeformations in metal structures. The main re-sults of these investigations are presented inmonograph of I.P. Trochun «Inner forces anddeformations in welding». The investigations ofthermal deformation processes in welding (theworks of I.M. Zhdanov, I.M. Chertov, E.A. Kor-shenko, V.M. Prokhorenko, A.S. Karpenko,V.N. Korzh, B.V. Medko, A.K. Gonchar et al.)were intensively developed in the 1960s, carriedout under the supervision of I.M. Zhdanov. Themain directions of research are the study of regu-

larities of formation of deformations and stressesin welding process, regularities of brittle fracture,creation of new experimental methods of investi-gations and devices, development of methods forreducing residual stresses and deformations.

The results of investigations of influence ofstress-strain state of welded structures on brittlefracture of welded joints are reflected in the the-sis for Dr. of Tech. Sci. Degree of V.M. Prokho-renko «Methods for calculating stress intensityfactors and crack opening in welded joints withaccount for residual stresses», the defence ofwhich was held in 1989. The further investiga-tions of stressed state of welded structures werecarried out in the direction of development ofnew engineering methods of calculation based onthe modern conceptions about the kinetics of de-formations in welding. The result of these studieswas the thesis for Cand. of Tech. Sci. Degree ofD.V. Prokhorenko. The research results de-scribed above were partially included into edu-cational books «Stresses and deformations inwelding» by B.S. Kasatkin, V.M. Prokhorenko,I.M. Chertov, as well as «Stresses and deforma-tions in welded joints and structures» by V.M. Prok-horenko, D.V. Prokhorenko, published, respec-tively, in 1987 and 2009. In the recent years thestudy of stress-strain state in welding is successfullycontinued by the young generation of researchersD.V. Prokhorenko and A.A. Perepichaj using nu-merical methods of mathematical modeling basedon finite element method. In particular, they car-ried out modeling of stress-strain state of the mainpipeline at the site of repair of crack-like defect,study of thermomechanical processes in surfacingthe weld on the surface of a semi-infinite body,calculation of stresses and deformations for differ-ent technological schemes of welding of butt jointsof thin metal. One of the results of these studieswas the thesis for Cand. of Tech. Sci. Degree ofA.A. Perepichaj.

In the 1980s, at the Chair the new scientificdirections appeared. Under the supervision ofI.R. Patskevitch the development of matters ofthe technology of welding cast iron and investi-gations of surface phenomena in welding wasstarted. The wettability and leaking of differentpairs of liquid metals in isothermal and non-iso-thermal conditions were studied. The effect ofexternal influences on the mentioned phenomenawas determined. The results of these works aredescribed in monograph of I.R. Patskevich, V.R.Ryabov and G.F. Deev «Surface phenomena inmetals during welding» (1991). Their workswere continued by V.P. Bojko, who created aunique experimental installation for investiga-

6 11/2015

tion of high-temperature processes of wetting andinterphase interactions at the gas—slag—metalboundary. According to the results of researchworks the numerous papers were published. Themethods of increasing the accuracy of produc-ing welded structures were intensively devel-oped under the supervision of I.M. Zhdanov.The works were continued by V.V. Lysak. Ac-cording to the results of the work the thesis forCand. of Tech. Sci. Degree was defended andan original method for welding of thin sheetmaterial was developed.

Many interesting results were obtained byV.V. Batyuk with the staff members of the Chair(B.A. Bobin, S.N. Minakov, I.N. Grisha et al.)in the direction of development of devices andtechnology of non-destructive testing of residualwelding stresses in different welded structures.

The investigations in the field of temperatureconditions of weld pool and electrode metal dropswere carried out by V.M. Dykhno and S.M. Get-manets, and later they formed the basis of thesisfor Cand. Tech. Sci. Degree and made a significantcontribution to the study of heat content of elec-trode metal drops and pool in argon arc welding.

Since the late 1980s, at the Chair under thesupervision of S.K. Fomichev a large complex ofworks was carried out to improve the corrosionresistance of welded structures, the materials ofthese investigations formed the basis for the the-sis for Dr. of Tech. Sci. Degree defended by himin 1994. Since 1987, at the Chair under the su-pervision of I.P. Belokur the matters of flawdetection and quality control of welded jointsare intensively developed. The results of researchworks in this direction are reflected in his thesisfor Dr. of Tech. Sci. Degree, defended in 1991,and in his numerous publications.

Since the 1990s, at the Chair the direction isactively developed associated with the develop-ment of CAD systems of technological processesof fusion welding using the computer technology.Under the supervision of I.F. Korinets and withthe participation of V.P. Bojko and Yu.I. Okhajthe mathematical models for melting of steel andtitanium base metals, solid and flux-cored wiresand heating of covered electrode were developed.As a result, the original methods of calculationof conditions of mechanized and automatic arcwelding in shielding gases and under flux weredeveloped.

Modern scientific and technical directionsof the Chair of Welding Production. The re-search directions of the Chair were formed in theclose creative relationships with the PWI, withthe participation of not only the leading scientists

of the Institute, but also Boris E. Paton himself.Prof. B.E. Paton, the Director of the PWI andthe President of the NAS of Ukraine, personallyassisted in solving the problems of material andtechnical base, training of scientific personnel ofthe Chair and solving the strategic tasks of de-velopment of welding production. Below the ba-sic problematic directions of work of the Chairare considered.

Technologies and metallurgical processes inelectric arc welding:

• creation of theoretical models for calculationof gas phase composition, effect of welding con-sumables on composition of weld metal, contentof gases and nonmetallic inclusions in it duringarc welding on the basis of physical thermody-namic modeling;

• investigation of metallurgical processes inwelding and development of new fused agglom-erated fluxes and flux-cored wires for weldingand surfacing;

• study of weld metal tendency to crack for-mation on the basis of technological samples;

• investigation of arc discharge in welding andits technical characteristics on the basis of com-plex factors of stability and transient processesusing synergic power sources;

• investigation and modeling of features ofpore formation in welding;

• modeling of thermal processes of electric arcwelding;

• investigation of influence of thermal defor-mational cycles of welding on phase compositionand structure of welded joint metal;

• technological features of welding usingmodulated current with synergic regulation ofarc;

• creation of mathematical models of base andelectrode metal melting in arc fusion weldingand optimization of welding processes in shield-ing gases on their basis.

Stresses and deformations in welding:• modeling and calculation of welding stresses,

deformations and displacements of elements ofwelded structures using finite element methodbasing on the modern computer technologies;

• study of influence of technological schemesof welding on residual displacements of longitudi-nal axis of welded structures and development ofoptimal technological sequence of their welding;

• modeling of stress-strain state of weldedstructures for beam and arc welding methods;

• determination of energy input of thermalstraightening of welded one-dimensional struc-tures using engineering calculation methods.

11/2015 7

Diffusion welding and brazing of metals, al-loys and composite materials:

• mathematical modeling of thermal deforma-tion processes during diffusion welding and braz-ing;

• development of technologies of diffusionwelding and brazing with the controlled stress-strain state;

• investigation of influence of surface modifi-cation by highly-concentrated energy flows onproperties of diffusion-welded and brazed joints;

• creation of new materials for producing dif-fusion-welded and brazed joints.

The active development of the direction is pro-vided owing to the support of the PWI manage-ment and staff members, in particular, K.A. Yush-chenko and I.V. Krivtsun, the Academicians of theNASU, V.F. Khorunov, the Corr.-Member of theNASU, and the staff members of the National Ship-building University (Nikolaev). The great contri-bution to the performance of works on control ofstress-strain state in the process of producing jointsin the solid state was made by V.I. Makhnenko,the Academician of the NASU, and to producingof metal-ceramic joints by O.K. Nazarenko, theCorr.-Member of the NASU.

Many of the graduates of the Chair becameprominent workers of science and welding pro-duction. Among the graduates from the Chairthere are many candidates and doctors of tech-nical sciences. In this article it is difficult tomention the names of all the prominent graduatesof the Chair. We only note that many of thembecame the academicians of the NASU and man-agers of large enterprises and organizations.

In the period of 2005—2015, a significantbreakthrough in raising the scientific potentialof the Chair and Welding Faculty in general wasdone. This became possible due to the implemen-tation of a target program of preparation of Can-didates and Doctors of sciences, signed byProf. B.E. Paton and M.Z. Zgurovsky, the Aca-demician of the NASU, the Rector of NTUU«KPI», and is realized at their active support.In the frames of this program at the WeldingFaculty 5 Doctors and 10 Candidates of TechnicalSciences were prepared.

Recently, the Chair has a fruitful cooperationwith well-known companies Fronius Ukraine andBinzel Ukraine GmbH, as well as the PWI PilotPlant of Welding Equipment and Pilot Plant ofWelding Materials, where our graduates areworking. These companies are equipping thelaboratories of the Chair with innovative equip-ment, provide information stands and weldingconsumables, carry out lectures-presentations on

the latest achievements in the field of weldingproduction. The students visit these enterpriseswith a great interest.

The staff members of the Chair take an activepart in the development of international coopera-tion. In particular, Prof. V.V. Kvasnitsky,Ass. Profs L.A. Zhdanov and A.A. Slivinsky, andthe Head of the laboratory A.A. Grinyuk take anactive part in the international projects of the Chi-nese-Ukrainian E.O. Paton Welding Institute. Thejoint research and educational programs with theBelarusian State University, Otto-von-GuerickeUniversity Magdeburg (Germany), Federal Uni-versity of Uberlandia (Brazil) and organizationsof other countries, are actively carried out.

During 80 years the Chair of Welding Produc-tion prepared thousands of highly-skilled special-ists, who played an important role in developmentof welding science and production. Our graduatessuccessfully work not only in Ukraine but also inmany other countries like Germany, USA, Canada,Russia, Australia, New Zealand and Israel.

Students of the Chair in its history. Manyglorious pages to the history of the Chair werewritten down by its students. They worked ac-tively in the student construction brigades inTyumen, Sakhalin and other regions of the formerSoviet Union and also in Czech Republic.

In 1985—1987, the specialized student bri-gades of the Chair participated in the large-scalescientific and technical experiment on the construc-tion of pipelines of polyethylene pipes for gasifi-cation and water supply of towns of Novoodessadistrict of Nikolaev (Ukraine) region. This was agood example of fruitful cooperation of the Chairwith the PWI, when the students under the super-vision of teachers participated in the implementa-tion of developments of the scientists.

The students of the Chair participate con-stantly in the Ukrainian student competitions onwelding and occupy the first and prize places inthe individual and team standings. The masterand postgraduate students deliver papers at theannual scientific conferences. In 2010, our teamoccupied a prize place at the All-Ukrainian en-gineering competitions and represented Ukraineat the European competitions. The Chair wasalways famous by its activists, who participatedin organization of the Faculty Days – editionof the multi-issue student newspaper, organiza-tion of broadcasting in our building, creation ofthe student library of several thousand books onwelding. The students of the Chair contribute tothe victories of sport teams of the Faculty infootball, basketball, shaping and achieve highresults in individual sports like track-and-field

8 11/2015

athletics, wrestling, boxing, gymnastics, weight-lifting, sport archery. The best students of theChair as to the results of study, research andsocial works receive scholarships of EvgenyO. Paton and Boris E. Paton, as well as schol-arships of the Rector of NTUU «KPI» and themayor of Kiev.

At different times many foreign students fromChina, Iran, Vietnam, Cuba, Poland, Hungary,Bulgaria and other countries studied at the Chair.Our Ukrainian students in their turn have anopportunity to study abroad. We have a doublediploma programs with Otto-von-Guericke Uni-versity Magdeburg and the Federal Universityof Uberlandia. More than a dozen of our studentshave already received the master diplomas ofUkrainian and German sample, as well as Ukrain-ian and Brazilian sample. In the recent years thestaff members of the Chair were essentially re-newed. It was joined by young and perspectivepersonnel. Over the last 10 years the staff mem-bers of the Chair defended 6 theses for Cand. ofTech. Sci. Degree, and at the present time mostof the teachers are the associate professors of theChair. The Chair is headed by Prof. V.V. Kvas-nitsky.

Participation of the Chair in development ofdefense industry of Ukraine. From the momentof foundation of the Chair its staff members underthe supervision of Evgeny O. Paton took an ac-tive part in the work on defense subjects. Thegraduates of the Chair worked in the defenseindustry and contribution of welders to creationof tank and other machinery in 1941—1945 is well-known. Such assistance has become urgent aswell in our days.

Since September, 2014, the initiative groupof the staff members and students of the WeldingFaculty under the supervision of Dr. A.A. Sliv-insky (students A. Suprun, E. Bilytsky, post-graduate student A. Bogach, staff memberS. Nestulya, Ass. Prof. L. Zhdanov et al.) is car-rying out works on manufacture and assembly ofprotective anti-cumulative screens for armoredvehicles of the ATO forces.

Due to cooperation with the Central ResearchInstitute of Armament and Military Equipmentof the Armed Forces of Ukraine into the designof screens and technology of their manufacturea number of improvements was introduced. Thescreens successfully passed ballistic tests at thesite of the Ministry of Defence of Ukraine. Theapplication for a patent was drawn up and theworks on adoption of the screens of the giventype for the armament are carried out. Today bythe efforts of students a single production in the

mechanical workshop of the Welding Facultywas developed into a serial production of screensat one of the enterprises of Kiev. At the moment,the Faculty carries out a complete design andtechnological support of the screens manufac-ture, a set of technical documentation on thescreens for the main models of light-armouredvehicles of the AFU – BTR-80 and BMP-2 wasdeveloped.

During combat missions of the AFU, underthe enemy fire, due to anti-cumulative screensmounted on armoured vehicles the lives of manymilitary men were saved. The staff members ofthe Chair received official thanks from the AFUcommand.

By the specialists of the Welding Faculty theprotective screens were manufactured and in-stalled on armored vehicles of the mobile unitsof special operation forces of Ukraine. The coop-eration with the Ukrainian defense enterpriseson working out the technological recommenda-tions on welding of armor steels of foreign pro-duction and vibration processing of welded armorstructures was arranged.

At the present time the further investigationson the processing of working elements of thescreens are carried out to improve the effective-ness of protection and guarantee the destructionof the warhead of anti-tank ammunition.

The staff members of the Chair of WeldingProduction of the National Technical Universityof Ukraine «Kiev Polytechnic Institute» cele-brate the 80th anniversary, being fully resolvedto conduct a high-level preparation of weldingspecialists for the independent state of Ukraine.

S.K. Fomichev, V.P. Bojko,V.V. Kvasnitsky, L.A. Zhdanov,

A.A. Slivinsky,V.L. Kovalenko, NTTU «KPI»

Assembly of anti-cumulative screens on armored personnelcarrier BTR-80 of the Armed Forces of Ukraine in the ATOzone by the Chair staff members

11/2015 9

STRUCTURE AND PROPERTIESOF WELD METAL MODIFIED BY NANOOXIDES

V.D. KUZNETSOV and D.V. STEPANOVNTTU «Kiev Polytechnic Institute»

37 Pobeda Ave., 03056, Kiev, Ukraine. E-mail: v.kuznetsov@kpi.ua

Given are the results of investigation of structure and properties of weld metals at introduction of nanooxidepowders into weld pool. It is shown that portion of nanooxides, being introduced in the weld pool, shouldnot exceed 0.5 % in a range of recommended modes of welding of low-alloy steels. It is determined thatnanooxides introduction results in formation of mainly non-metallic inclusions in nanosize range of 0.07—0.13 μm as well as in 0.3—0.8 μm range. At that increased content of carbon, titanium, oxygen, aluminumand manganese are observed in them in comparison to reference welds. It is shown that introduction of0.5 % titanium nanooxide reduces content of brittle constituents of the structure and content of acicularferrite rises to 40 %, that results in 2 times increase of impact toughness together with simultaneous increaseof yield strength. It is shown that modifying of weld metal by titanium and aluminum nanooxides reducessolidification interval, that can indicate their effect as 2nd type inoculants. 15 Ref., 3 Tables, 6 Figures.

Keywo r d s : weld, low-alloy steel, nanopowder,nanooxides, structure, non-metallic inclusions, mechani-cal properties, solidification interval

Analysis of investigations and publications of re-cent years indicates a role of non-metallic inclu-sions (NMI) as factor for regulation of structureand properties of cast metal. It is shown [1, 2]that inclusions engineering can be used for steelmicrostructure optimizing. Inclusions (oxides,sulfides, carbides) of <1 μm size, promoting nu-cleation of acicular ferrite (AF), are referred toa separate group due to their small size, and theyare called dispersoids having no negative effecton reduction of mechanical properties, but de-termining the conditions of metal microstructureformation.

Works [3—6] studied effect of NMI size onheterogeneous nucleation of AF structure as wellas content and peculiarities of distribution ofNMI in presence of aluminum oxides, titaniumoxides and nitrides.

Systematic researches in this direction weremade by staff members of the E.O. Paton ElectricWelding Institute. Works [7—13] experimentallystudied and theoretically grounded effect of car-bides, nitrides and oxides on level of AF forma-tion and increase of mechanical properties of weldmetal of low-alloy steels.

Peculiarity of works indicated above is studyof role of NMI, including of nanosize range, beingformed in weld metal at corresponding changesof concentration of introduced elements and theirreactions with oxides, nitrides, carbides forma-tion, as well as when not more than 1 μm size

elements are introduced via charge of flux-coredwire and in form of nanosize powder inoculants.

Development of works in this direction re-quires accumulation and analysis of experimentaldata applicable to range of low-alloy steels aswell as schemes of introduction of nanosize par-ticles in the weld pool.

Aim of this work is investigation of effect ofNMI on structure and properties of weld metalsat direct introduction of nanooxide powders inthe weld pool.

Investigations for detection of general depend-encies were carried out in welding of low-alloysteels 09G2S and 10G2FB with Sv-10KhGN2SMFTYu wire as well as steel A-514(18GSKhNF) with Sv-09G2S wire. In all thecases welding was performed in Ar + 28 % CO2gas mixture at 12.3 kJ/cm heat input.

Nanocomponents were introduced in the weldpool in form of addition alloy after pressing andbaking of homogeneous mixture of iron powderof 40 μm fraction and nanosize powders of alu-minum or titanium oxides (27—41 nm) with setvolume relationship.

Prepared addition alloy was used as consum-able electrode of set length and diameter, embe-ded in a groove along the butt length beforewelding [14]. In this case effect of processes,related with passing of nanopowders via arc, waseliminated.

Examination of structure and NMI were car-ried out by methods of optical and electron mi-croscopy using Neophot-30 microscope, electronscanning microscope JSM 35CF with attachmentfor local X-ray analysis INCA Energy 350 as well© V.D. KUZNETSOV and D.V. STEPANOV, 2015

10 11/2015

as computer programs developed at the E.O. Pa-ton Electric Welding Institute for analysis ofmicrostructure constituents and distribution ofNMI by size and composition. Thermal-physicalcharacteristics of cast metal were investigatedusing NETZSCH thermal analyzers DSK 404F1.

Investigations showed that amount ofnanopowder introduced in the weld pool resultsin change of amount of NMI in the weld metal.

Dependence of general volume fraction ofNMI on content of aluminum nanooxides in theweld are given in Figure 1.

If NMI fraction in the reference weld makes0.45 % then introduction in the weld pool ofnanopowder of aluminum oxide in 0.5 % amountpromotes for rapid increase of fraction of NMIto 0.65 %. Introduction of aluminum oxidenanopowder in 0.5—2.5 vol.% range does not havesignificant effect on volume fraction of NMI. Fur-ther growth to more than 2.5 % of nanopowderresults in rise of NMI volume fraction from 0.70to 0.86 %.

However, results of evaluation of structuralchanges for indicated range of introducednanopowders showed that exceed of volume frac-tion above 0.5 % is not accompanied by signifi-cant structural changes.

Figure 2 for example shows structure of weldmetal in welding of 09G2S steel for indicatedrange of changes of volume fraction of aluminumoxide nanopowder.

Formation of structure, the main constituentsof which are precipitation of polygonal ferrite,acicular and lamellar ones with ordered and dis-ordered secondary phases, takes place in the weldmetal under initial conditions. The peculiarityof this structure is presence of coarse constituentsof indicated morphological forms of ferrite (Fi-gure 2, a). Microhardness of constituents changesfrom HV 145 to HV 187.

Microstructure of weld metal with additionof Al2O3 nanopowder in 0.5 % amount has refineddispersed structure mainly consisting of upperbainite, partially lower and acicular ferrite (Fi-gure 2, b). Microhardness of constituents changesfrom HV 264 to HV 304.

Weld metal with addition of Al2O3 nanopow-der in amount of 2.5 % volume fraction has mi-crostructure of intra-granular ferrite and ferritewith ordered and disordered secondary phases

Figure 1. Dependence of NMI volume fraction on contentof aluminum nanooxides in weld metal in welding of steel09G2S using Sv-0KhGN2SMFTYu wire

Figure 2. Microstructure of reference weld (a) and weld with addition of nanopowder Al2O3 in amount of 0.5 (b), 2.5(c) and 4.5 (d) %

11/2015 11

(Figure 2, c). Microhardness of constituents vir-tually does not change and stays in HV 180—189limits.

Addition of Al2O3 nanopowder in amount of4.5 % results in formation in the structure ofintra-granular ferrite with precipitates of AF andferrite with ordered secondary phases (Figure 2,d). Microhardness of constituents changes fromHV 188 to HV 236.

It can be concluded that the main role in for-mation of structure with increased strength andtoughness properties plays nanooxides being intro-duced only in small amount (0.5 %), and increaseof their concentration, apparently, promotes forcoagulation and coalescence with NMI of materialduring solidification of the weld pool and has nosignificant effect on structure formation.

In fact, computer processing of results of dis-tribution of NMI by size allowed selecting threesize ranges from total data array, i.e. inclusionsof to 0.3 μm, from 0.3 to 0.8 μm and more than0.8 μm.

The peculiarity of NMI in the weld metal istheir smaller size range at nanooxides introduc-tion into the weld pool (Figure 3).

Thus, if from 5 to 10 inclusions of the referenceweld fall for 0.10—0.25 μm and from 5 to 40inclusions for 0.5—1 μm size range, then in thecase of nanooxide introduction the first rangecovers from 8 to 22 inclusions and the secondfrom 5 to 27 inclusions.

Found dependencies are also verified by theresults of processing of size of only spherical in-clusions by index of diameter of equivalent cir-cumference. For example, Figure 4 gives the hy-

grograms on volume content and distribution ofsuch inclusions in the weld metal for initial stateand with inclusion of nanooxides.

The main part of spherical inclusions from 4to 6 % falls for the range of up to 0.3 μm size aswell as for the ranges of 0.3—0.8 μm and morein the initial state without addition of nanopow-der oxide. At that, up to 9 % precipitation ofparticles of 0.31—0.37 μm size are observed (Fi-gure 4, a).

Addition of TiO2 nanopowder in 0.5 vol.%amount promotes for rise from 6 to 14 % of partof spherical inclusions of 0.3 μm as well as 0.30—0.55 μm size. At that, no inclusion of more than0.8 μm size can be virtually observed (Fi-gure 4, b).

Results of local spectral analysis of chemicalcomposition of the inclusions showed that sig-nificant increase of carbon, oxygen, aluminum,sulfur, titanium, manganese concentration is ob-served in each of them independent on size. Thisis particularly obvious for carbon, oxygen andsulfur (several orders).

Introduction of nanooxides, in keeping thegeneral pattern on content of indicated elements,somewhat changes it to the side of larger contentof oxygen, aluminum and titanium, that indicatespresence of oxides of these elements in the inclu-sions.

Analysis of triple diagrams of oxides of SiO2—Al2O3—MnO and TiO2—Al2O3—MnO systems forthe most typical range of inclusion size 0.3—0.8 μm also verifies increased content of Al2O3and TiO2 oxides (Figure 5).

Figure 3. Histogram of NMI distribution by size: a – in reference weld; b – in weld with addition of aluminum oxidein 0.5 % amount

12 11/2015

Figure 4. Histogram of distribution of inclusions in weld metal by diameter of equivalent circumference d: a – in initialcondition; b – weld with 0.5 vol.% TiO2 nanooxide in welding of 10G2FB steel

Figure 5. Triple diagrams of oxides of SiO2—Al2O3—MnO and TiO2—Al2O3—MnO system: a, b – reference weld; c, d –weld with 0.5 % TiO2

11/2015 13

Thus, if significant part of inclusions of oneas well as another systems in the reference weldcontains from 50 to 90 % MnO, up to 40 % ofAl2O3 and to 20 % TiO2 (Figure 5, a, b), thenintroduction in the weld pool of 0.5 % TiO2 pro-vides for high MnO concentration only in sepa-rate inclusions, and presence of Al2O3 as well asTiO2 in the inclusions exceeds to 55 and 40 %,respectively (Figure 5, c, d).

Thus, complex analysis of inclusions indicatessignificant difference in their sizes, distributiondensity and content in metallic matrix at nanoox-ide introduction, which has effect on weld metalstructure.

In fact, results of metallographic analysisdetermined that the most wide-spread morpho-logical forms of ferrite in the weld metal struc-ture are block ferrite (BF), lamellar ferrite(LF), intra-granular AF, Widmanstatten fer-rite (WF), upper (UB) and lower bainite (LB).Percent content of each of forms in the inves-tigated welds in welding of 10G2FB steel isgiven in Table 1.

Initial structure of welds is characterized byincreased content of brittle constituents (BF,WF, UB) and formation of AF to 10% with highcoefficient of shape (L/B of 4—7) and acicularsof up to 20 μm length.

Microstructure of weld metal has sufficientlyhigh content of intra-granular polygonal ferritewith precipitation along the boundaries of grainsof xenomorphic ferrite (Figure 6, a—c). Weldswith such structure are characterized by low levelof toughness and ductility of metal. Values ofmicrohardness of structural constituents varyfrom HV 231 to HV 253.

Weld metal structure with TiO2 in 0.5 %amount is characterized by reduced content ofbrittle constituents (BF, WF, UB) and increasedAF content to 40 % with more favorable coeffi-cient of shape (L/B of 3—5) and aciculars’ lengthto 5 μm in comparison with reference structure(Figure 6, d—f).

Welds with such structure are characterizedby combination of sufficiently high level of in-dices of toughness, ductility and strength. The

Table 1. Typical morphological forms of ferrite in weld metal structures

Studied weldConstituents of microstructure of weld metal, %

BF LF AF UB LB WF

Reference weld without nanooxides Up to 10 10—20 Up to 10 20—40 20—40 Up to 35

Weld with 0.5 % TiO2 nanooxide Up to 10 Up to 10 20—40 10—20 10—15 Up to 15

Figure 6. Typical structure of weld metal without nanooxides (a—c) and with 0.5 vol.% TiO2 (d—f)

14 11/2015

measurements show that microhardness of struc-tural constituents varied from HV 230 to HV 250.

Thus, introduction of nanooxides, in particu-lar titanium, in the weld pool results in positivestructural changes from the point of view of for-mation of tough morphological forms of ferrite,that promotes for increase of weld metal mechani-cal properties (Table 2).

As follows from data of Table 2, increase ofyield strength for 157 MPa and strength for105 MPa is observed at introduction of titaniumnanooxides in the weld pool. At that, impacttoughness also increases 2 times. Thus, the com-mon point, independent on grade of investigatedlow-alloy steel and type of introduced nanoox-ides, is change of size, distribution density, com-position of the inclusions as well as their positiveeffect on structure and mechanical properties ofweld metal.

Role of introduced nanooxides in structure-forming is also shown by data of thermogramsof differential scanning calorimetry. The inves-tigations found the differences in melting tem-peratures as well as in solidification of metalmodified by nanooxides (Table 3).

The general dependence is some reduction ofweld metal melting temperature at nanooxidesintroduction. It is the well-known fact thatnanostructural materials differ by significantlylower melting temperature [15], therefore, somereduction of liquidus temperature can be relatedwith presence of nanooxides in the weld metal.Reduction of melting temperature is developedin different ways for titanium and aluminum ox-ides. Thus, if melting temperature of the refer-ence metal is 1543.4 °C, then its reduction to1535.8 °C is observed at introduction of 0.5 %TiO2 and to a greater extent, i.e. 1522.5 °C, at1 % Al2O3.

At the same time, increase of solidificationtemperature of weld metal, modified by nanoox-ides, except for 1 % Al2O3, is observed. At that,the general dependence is in reduction of intervalof solidification of modified metal, independenton investigated range of changes of volume frac-tion of introduced nanooxides.

Thus, it makes 43.9 °C for the reference metal,then observed reduction for the investigatedrange of changes of each of nanooxides is virtually

the same and it is ΔT = 23.3 °C for aluminumoxide and ΔT = 11.8 °C for titanium oxide.

It can be concluded based on this that effectof nanooxides can be observed already in solidi-fication stage, that changes its conditions to theside of quicker propagation, may be as a conse-quence of appearance of additional centers of nu-cleus formation on the inclusions, i.e. their effectas 2nd type inoculants.

This stage of investigations does not allowdetermining one-valued role of nanooxides inmechanism of regulation of weld metal structureformation. They also can play a role of 3rd typeinoculants. In the melt they can draw heat forown heating and reduce melt temperature undercertain conditions of melting, that results ingrowth of solidification rate. Besides, as it wasdetermined in works of the E.O. Paton ElectricWelding Institute, their effect can also be ob-served at the second stage of the secondary so-lidification effecting austenite transformation re-sistance.

Accumulation and analysis of experimentaldata in this direction allows outlining the mostsignificant sides of NMI effect, including fornanosize range, on weld structure formation.

Conclusion

It is shown that increase from 0.5 to 4.5 % ofvolume fraction of nanooxides introduced in theweld pool results in increase from 0.46 to 0.87of general NMI portion in the weld metal, at thatsignificant structural changes take place at in-troduction of nanooxide, volume fraction ofwhich does not exceede 0.5 %. Typical in thiscase is increase of fraction of inclusions of small

Table 2. Results of mechanical tests of weld metal in welding of A-514 (18GSKhNF) steel using 09G2S wire

Studied metalYield strength σy,

MPaTensile strength

σt, MPaRelative

elongation, %Reduction in area,

%Impact toughness

KVC, kJ/m2

Without nanopowders 357 542 21 61 4.6

With TiO2 nanooxide in 0.5 % amount 514 647 18 54 9.3

Table 3. Thermal-physical characteristics of weld metal

Metal content TL, °C TS, °C ΔT, °C

Reference 1543.4 1499.5 43.9

0.5 % TiO2 1535.8 1524 11.8

1 % TiO2 1540.6 1527.7 12.9

0.5 % Al2O3 1541.9 1518.6 23.3

1 % Al2O3 1522.5 1499 23.5

Note. TS – solidus temperature; TL – liquidus temperature;ΔT – solidification interval.

11/2015 15

size range 0.07—0.50 μm. Inclusions in the weldmetal modified by nanooxides have increasedcontent of oxygen, aluminum and titanium, thatindicates their prime composition from oxides ofthese elements.

Introduction of titanium nanooxide promotesfor reduction of brittle constituents of ferrite andincrease of content of acicular ferrite. This resultsin increase of mechanical properties of the weldmetal, in particular impact toughness. Introduc-tion of nanooxides reduces weld metal solidifi-cation range, that can indicate their effect as 2ndtype inoculants.

1. Takamura, J., Mizoguchi, S. (1990) Roles of oxidesin steels performance – Metallurgy of oxides insteels. In: Proc. of 6th Int. Iron and Steel Conf.(Tokyo, Japan, 1990), Vol. 1, 591—597.

2. Grong, O., Kolbeinsen, L., van der Eijk, C. et al.(2006) Microstructure control of steels through dis-persoid metallurgy using novel grain refining alloys.ISIJ Int., 46, 824—831.

3. Lee, T.K., Kim, H.J., Kang, B.Y. et al. (2000) Ef-fect of inclusion size on the nucleation of acicular fer-rite in welds. Ibid., 40, 1260—1268.

4. Yamamoto, K., Hasegawa, T., Takamura, J. (1996)Effect of boron on intra-granular ferrite formation inTi-oxide bearing steel. Ibid., 36, 80—86.

5. Vanovsek, W., Bernhard, C., Fiedler, M. et al.(2013) Influence of aluminium content on the charac-terization of microstructure and inclusions in high-strength steel welds. Welding in the World, Issue 1,57, 73—83.

6. Seo, J.S., Kim, H.L., Lee, C. (2013) Effect of Ti ad-dition on weld microstructure and inclusion charac-teristics of bainitic GMA welds. ISIJ Int., 53(5),880—886.

7. Golovko, V.V., Grigorenko, G.M., Kostin, V.A.(2011) Influence of nanoinclusions on formation ofweld metal structure of ferritic-bainitic steels (Re-view). Zbirnyk Nauk. Prats NUK, 4, 42—49.

8. Pokhodnya, I.K., Golovko, V.V., Stepanyuk, S.M. etal. (2012) Study of effect of titanium nanocarbideson formation of microstructure and properties ofweld. FKhMM, 6, 68—75.

9. Golovko, V.V., Stepanyuk, S.M., Ermolenko, D.Yu.(2012) Study of influence of nanoformation in metalon microstructure formation of weld and its mechani-cal properties. In: Transact. on Building, materialsscience, machine-building, Issue 64, 155—159.

10. Golovko, V.V., Pokhodnya, I.K. (2013) Effect ofnon-metallic inclusions on formation of structure ofthe weld metal in high-strength low-alloy steels. ThePaton Welding J., 6, 2—10.

11. Golovko, V.V., Stepanyuk, S.N., Ermolenko, D.Yu.(2014) Technology of welding of high-strength low-alloy steels with introduction of titanium-containinginoculants. In: Nanosized systems and materials inUkraine, 395—399. Kiev: Akademperiodika.

12. Golovko, V.V., Stepanyuk, S.M., Ermolenko, D.Yu.(2015) Effect of titanium-containing inoculants onstructure and properties of weld metal of high-strength low-alloy steels. The Paton Welding J., 2,14—18.

13. Grigorenko, G.M., Kostin, V.A., Golovko, V.V. et al.(2015) Effect of nanopowder inoculants on structureand properties of cast metal of high-strength low-alloysteels. Sovr. Elektrometallurgiya, 2, 32—41.

14. Kuznetsov, V.D., Loboda, P.I., Fomichov, S.K. et al.Method of electric arc welding with introduction ofnanocomponents into weld pool. Pat. 98985 UA. Int.Cl. B 23K 9/16. Fill. 15.12.14. Publ. 12.05.15.

15. Ragulya, A.V., Skorokhod, V.V. (2007) Consolida-ted nanostructural materials. Kyiv: Naukova Dumka.

Received 21.10.2015

16 11/2015

EFFECT OF COOLING MODE AFTER DIFFUSIONWELDING AND BRAZING ON RESIDUAL STRESSES

IN GRAPHITE—COPPER EDGE JOINTS

V.V. KVASNITSKY1, G.V. ERMOLAEV2 and M.V. MATVIENKO3

1NTUU «Kiev Polytechnic Institute»37 Pobeda Ave., 03056, Kiev, Ukraine. E-mail: kvas69@ukr.net

2National Shipbuilding University9 Geroiv Staliningrada Ave., 54025, Nikolaev, Ukraine. E-mail: welding@nuos.edu.ua

3Kherson Branch of National Shipbuilding University44 Ushakov Ave., 73022, Kherson, Ukraine. E-mail: matvienkomv@i.ua

Studied is stress-strain state at different modes of cooling of cylindrical assemblies from graphite and copperin plastic and elastic stages considering short-term plastic deformations as well as creep. Effect of coolingmode on plastic deformations in butt zone and axial stresses on assemblies’ surface, determining possibilityof fracture of brittle graphite was jound. 7 Ref., 3 Tables, 12 Figures.

Keywo r d s : diffusion welding, brazing, graphite—copper joint, stress-strain state, cooling modes, residualstresses, mathematical modeling

Graphite products are used in combination withmany metals in current equipment. Graphite hav-ing high electric and heat conduction is widelyused in development of current-carrying or cur-rent-collecting devices of different assembliesand machines, graphitized electrodes, sealing ar-rangements etc., where joining with metals isnecessary. Brazing and diffusion welding [1, 2]are used for that. The general problem for bothmethods of joining are residual stresses causedby different thermal coefficients of linear expan-sion (TCLE) of materials to be joined. Residualstresses in ductile materials being joined usuallyresult in degradation of shape and size, and canpromote crack formation and fracture in graphite,which is a brittle material.

Metals with close to graphite TCLE are se-lected for reduction of stresses in metal-graphiteassemblies. Many assemblies use titanium, whichis not equal to graphite on electric and heat con-duction. The best choice for indicated devices iscopper, but it has significantly larger TCLE thangraphite, that often results in crack formation ingraphite after assembly cooling.

Our earlier researches [3, 4] were dedicatedto investigation of dependencies of formation ofstress and deformation fields in the butt zone indiffusion welding (DW) and brazing of partsfrom dissimilar materials. This information isnecessary for updating assembly structure andproduction of quality joint. Some general de-

pendencies of formation of residual stresses werereceived for metal—ceramics joints in bush-to-bush joint assemblies, in which Young’s modulesof joined materials are equal [5]. But they arean order different in graphite and copper, thatcan have significant effect on stress-strain state(SSS) of the assembly in their joining. Therefore,investigation of SSS formation in copper-non-metallic assemblies is a relevant task.

Researches of graphite (ceramics)-to-metalsjoints showed that fracture is initiated by cracksappearing in the butt in non-metallic part [1].Metal surface after complete fracture includesthin graphite layer as shown in Figure 1.

Aim of the investigation is study of peculiari-ties of SSS formation in cooling after DW andbrazing of metal-graphite assemblies and effectof graphite and copper joining technology on itby computer modeling.

Investigations were carried out by computermodeling method using license software complexANSYS (vers. 10). Axially symmetric problems

Figure 1. Fracture surface of metal (1) to graphite (2) joint© V.V. KVASNITSKY, G.V. ERMOLAEV and M.V. MATVIENKO, 2015

11/2015 17

with finite elements of PLANE 182 type weresolved.

Edge joints, which are usually used in cur-rent-carrying devices, were taken for investiga-tions. Modeling was carried out for bush-to-bush(B—B) assembly. This model in comparison withbar or cylinder is more informative since has two-side surfaces and being easily transformed in cyl-inder with inner radius equal zero, and rules ofSSS formation in cylinder—cylinder type assem-blies are also preserved for bar—bar type assem-blies as shown in works [6, 7].

View of B—B assembly and section of its finiteelement model are shown in Figure 2. This modelcan be used for joints made by pressure brazing,eliminating brazing metal interlayer due to itssmall thickness.

Table 1 shows thermal-physical properties ofcopper and graphite, taken in calculations.

Assembly cooling in 900 to 500 °C range, inwhich creep resistance has a significant effect onstress relief, was considered for investigation ofgeneral dependencies of SSS formation.

Modeling was carried out for 5 variants oftemperature reduction in indicated interval:

• elastic solution at temperature reduction to500 °C (variant 1);

• elastic-plastic solution at quick reduction oftemperature to 500 °C (variant 2), where creepdeformations are negligibly small;

• elastic-plastic solution considering creep atgradual reduction of temperature to 500 °C incourse of 60 s and further holding during 6000 sat 500 °C (variant 3);

• elastic-plastic solution considering creep atgradual reduction of temperature to 500 °C dur-ing 600 s and further holding at 500 °C up to6000 s (variant 4);

• elastic-plastic solution considering creep atgradual reduction of temperature to 500 °C dur-ing 6000 s (variant 5).

Figure 3 provides for thermal cycles for con-sidered variants.

Copper creep rate in variants 3—5 was deter-mined by equation

ε. = C1σ

C2e

C3

T ,

Figure 2. View of B—B assembly (a) and its finite elementmodel (b): 1 – copper bush; 2 – graphite bush

Figure 3. Thermal cycles of cooling for studied five variants(1—5)

Table 1. Thermal-physical properties of materials joined

Material Temperature, °CYoung’s modulus

E⋅103, MPaPoisson’s coefficient TCLE⋅106, 1/deg

Yield strength,MPA

Strengtheningmodulus⋅103, MPa

Graphite 20 9.3 0.18 4.8 — —

200 9.4 5.0 — —

400 9.8 5.1 — —

700 10.3 5.5 — —

900 10.8 5.7 — —

Copper 20 125 0.34 16.7 69 1.5

200 110 17.2 60 1.3

400 100 17.8 45 0.9

700 60 19.4 17.3 0.8

800 40 20.5 11.5 0.7

900 38 19.8 8.5 0.6

18 11/2015

where C1 = 1.67⋅10—30, C2 = 5, C3 = 25872 arethe equation coefficients, received by us experi-mentally in investigation of copper creep.

Fields and diagrams of distribution of stresses,deformations and movements at different stagesof cooling, including the moments of end of tem-perature reduction (all variants) and end of hold-ing at 500 °C (variants 3 and 4), were analyzed.Further only fields and diagrams of axial stressesare given. They, as shown in work [5], are themain reason of brittle fracture of materials withlow TCLE.

Analysis of stress fields for different coolingvariants showed that all stress constituents, in-cluding equivalent ones, significantly change atcooling rate variation, in particular in copper

(Figure 4). Further holding at 500 °C (variants3 and 4) promotes more changes in them (Fi-gure 5). At that radial, circumferential, tangen-tial and equivalent stresses are concentrated closeto the butt, and axial stresses are in internal andexternal surfaces of the bush.

Plastic deformation are distributed in accord-ance to stresses, i.e. from copper side in the vi-cinity to the butt, to the most extent at distanceof around 1/4 of bush thickness from its externalsurface (Figures 6 and 7).

Analysis of fields and values of plastic defor-mations (Table 2) showed that instantaneous de-formations in copper reduce from 1.4 % in variant1 to zero in variant 5 with increase of coolingtime. Creep deformations at temperature reduc-

Figure 4. Axial stresses after cooling to 500 °C in variants 1 (a), 2 (b), 3 (c), 4 (d) and 5 (e)

Figure 6. Equivalent deformations of instantaneous plasticity after cooling to 500 °C in variants 2 (a), 3 (b), 4 (c) and5 (d)

Figure 5. Axial stresses after cooling and holding in variants 1 (a), 2 (b), 3 (c), 4 (d) and 5 (e)

11/2015 19

tion stage, on the contrary, significantly increasewith rise of cooling time to 500 °C (from 0.4 %in variant 3 to 1.05 % in variant 5). Total plasticdeformations at cooling stage reduce from 1.4 %in variant 2 to 1.05 % in variant 5.

Creep deformations show insignificant rise(0.19 and 0.22 %) in process of holding at con-stant temperature 500 °C (variants 3 and 4). Asa result, total creep deformations for the wholeperiod of cooling and holding at 500 °C increasewith rise of cooling time and corresponding de-crease of time of holding from 0 (variant 2) to1.05 % (variant 5), whereas sum ones (instanta-neous and plastic) reduce from 1.4 % (variant 2)to 1.05 % (variant 5).

Thus, increase of cooling rate is more efficientthan increase of time of holding after temperaturedecrease from the point of view of developmentof short-term plastic deformations in copper.And, vice versa, slow cooling is more efficientfrom the point of creep deformations.

Nature of fields of short-term and creep plasticdeformations are also distinguished (Figures 6 and7). The short-term deformations are mostly con-centrated close to the butt in its part adjacent tothe external surface (Figure 6). The creep defor-mations cover all butt area at distance comparablewith bush thickness (Figure 7). At that, they sig-nificantly reduce (Figures 6, b and 7, a) in processof long-term holding at 500 °C (variant 3).

Effect of plastic deformations in copper onaxial stresses appearing on graphite side surfaceis caused by two mechanisms. On the one hand,in accordance with the general principles of me-chanics, plastic deformation of copper providesfor reduction of level of stresses in it and, respec-tively, preserve assembly equilibrium conditionsin graphite. On the other hand, they effect bendshape of generatrix, which also influences thelevel and nature of stress distribution in copper[5]. Nature of this effect is ambiguous. Plasticdeformations in copper can increase as well asreduce surface curvature, respectively increasingor reducing stress level.

Elastic and plastic deformation of bush mate-rial result in change of its surface shape. Analysisof diagrams of radial movements of points of ex-ternal surface showed (Figure 8) that the natureof generatrix bending is almost kept the same,as at equal rigidity of materials being joined [5].However, their skew-symmetries relatively tobutt plane at elastic loading (variant 1) are sig-nificantly violated. The generatrix from the side

Figure 7. Equivalent creep deformations after cooling and holding in variants 3 (a), 4 (b) and 5 (c)

Figure 8. Diagrams of radial movements of points of bushexternal surface after cooling to 500 °C (a) and after coolingand holding (b) for variants 1—5 (1—5)

Table 2. Maximum plastic (equivalent) deformations in copper,%

Variant

End of coolingEnd ofholding

SumInstanta-neous

deforma-tions

Creepdefor-

mationsSum

Creepdeforma-

tions

2 1.40 0 1.40 0 1.40

3 0.90 0.40 1.30 0.59 1.49

4 0.23 0.75 0.98 0.97 1.21

5 0 1.05 1.05 1.05 1.05

20 11/2015

of more rigid material (copper) shows less bend-ing, approaching to free contraction state (seedashed line in Figure 8). The situation is oppositeand the bend is significantly increased from theside of less rigid material (graphite).

Bend in more rigid, but ductile material (cop-per) increases and that in less rigid elastic ma-terial (graphite) reduces, shape of bend of thegeneratrix approaches to symmetric one in appear-ance of plastic instantaneous and creep deforma-tions (variants 2—5). In other words, the short-termplastic deformations, developing at quick cooling,and creep deformations (variants 3—5) compensatehigher rigidity of copper in elastic state in com-parison with graphite and equalize deformationsof external surface in the butt region.

At that, the diagrams of radial movements ofpoints of external surface after cooling and hold-ing are virtually matched in variants 3—5 regard-less some differences in value of maximum plasticdeformations in copper (see Table 2).

Nature of radial movement of the points ofinternal side surface is more complex (Figure 9).A convex part of the graphite surface (materialwith lower TCLE) in immediate proximity tothe butt (1—2 mm) gradually passes into a con-cave one with the removal from it. The situationis different in copper (material with largerTCLE), in which concave surface close to thebutt gradually passes into a convex one with theremoval from it.

Such a complex bend shape of the generatrixon internal surface of the bush is explained in work[5] by oncoming effect of two factors. On the onehand, this is a difference of mutual shift of upperand lower parts of the bush due to different changeof their diameters, and, on the other hand, it isvarious change of section width at temperaturecontraction of materials of upper (metal) and lower(graphite) parts of the assembly.

Mutual shift of upper and lower parts of thebushes results in formation of convexity on theexternal and concavity on the internal surface ofmaterial with lower TCLE (graphite). Differentchange of width results in formation of convexityon the both sides of this material. As a conse-quence of simultaneous effect of both factors anature of bend of the external surface, i.e. outsideconvexity, is preserved and the internal one beingsignificantly changed.

Presence (effect) of two mechanisms of move-ment is also verified by absence of symmetry ofthe fields of plastic deformations and diagramsof tangential stresses relatively to middle of thebush thickness. Quantitative characteristic of ef-fect of shift of the bush parts can be a level of

tangential stresses in the middle of bush thickness(around 20 MPa).

Nature of bend of the internal surface is pre-served at change of cooling rate, but value ofmovement is reduced at decrease of cooling rate(variants 3—5). Difference in movements is re-duced at further holding (see Figure 9).

Axial stresses are distributed in accordancewith surface deformation (Figures 10 and 11).They are compressive on the external surface inupper copper bush (with larger TCLE) and ten-sile ones in graphite lower part (with lowerTCLE) (see Figures 4, 5 and 10).

After cooling tensile stresses for all variantsreach maximum value (around 14 MPa) close tothe butt (at 2—3 mm distance), and show gradual2 times reduction at approaching to it (Figure 10,a). Further holding at 500 °C in variants 3 and4, when creep deformations are added to short-term deformations, results in the fact that maxi-mum stresses in graphite near the butt, on thecontrary, rapidly increase to 30—35 MPa (Fi—gure 10, b) and maximum point is right up tothe butt.

The distribution is much more complex (seeFigure 11) on the internal surface. Axial stressesin the lower graphite bush are of the tensile nature,they reduce to the zero step-by-step (at 2—3 mmdistance from the butt) and transform in the com-pressive ones at removal from the joint, and thenthey again reduce to the zero at the edge.

Figure 9. Diagrams of radial movements of points of bushinternal surface after cooling to 500 °C (a) and after com-plete cooling and holding (b) for variants 1—5 (1—5)

11/2015 21

Table 3 and Figure 12 for convenience of com-parison of the variants provide for the values anddiagrams of maximum tensile stresses in graphiteafter cooling to 500 °C and holding at this tem-perature during 6000 s.

Analysis of the Table and Figures show thatthe internal surface is exposed to the most danger

from the point of view of crack formation ingraphite after cooling. At that, for all variantsexcept for variant 5, the maximum tensilestresses reach bending strength of graphite(45 MPa for MPG-8 graphite and 34.3 MPafor MPG-6 and MPG-7 graphite). Variants 3and 4 are dangerous for the external surface.Using them holding after cooling will result inlarger tensile stresses.

Thus, the optimum, from the point of view ofreduction of danger of crack formation in graph-ite after cooling to 500 °C, shall be variant 5(gradual reduction of temperature from 900 to500 °C during 6000 s) at which tensile stressesdo not exceed 25 MPa.

Variants with quick cooling and further hold-ing are not reasonable, since large stresses aredeveloped on the external surface of graphite atthem.

Further cooling of the assembly to 20 °C takesplace under conditions of growth of yield point

Figure 10. Diagrams of axial stresses on external surface ofgraphite after cooling (a) and after cooling and holding at500 °C (b) for models 1—5 (1—5)

Figure 12. Maximum tensile stresses on external (a) andinternal (b) surfaces of graphite in variants 1—5 after coolingto 500 °C and holding at this temperature (variants 3 and 4)

Figure 11. Diagrams of axial stresses on internal surface ofgraphite after cooling (a) and after cooling and holding at500 °C (b) for models 1—5 (1—5)

Table 3. Maximum tensile stresses in graphite, MPa

Variant On external surface On internal surface

End of coolingEnd ofholding

End of coolingEnd ofholding

1 14.2 — 35.7 —

2 13.7 — 34.2 —

3 14.0 44.7 33.3 31.0

4 14.5 39.0 29.9 28.5

5 14.3 14.3 25.0 25.0

22 11/2015

and creep resistance. Creep rate reduces and re-sidual stresses exceed graphite strength limit(more than 65 MPa) even in cooling from 500to 20 °C during 3.8 h.

Thus, regardless the effective influence of de-formations of instantaneous plasticity and creepin 900—500 °C temperature interval, applicationof DW or brazing of copper (silver) braze alloyscan result in assembly fracture.

Sufficiently high temperatures of DW are tobe noted when considering technological variantsof graphite-to-copper joining. Work [2] investi-gates three variants of graphite-titanium DW,namely with nickel interlayer of 10—30 μm thick-ness deposited on graphite by galvanic method,with nickel foil of 10 μm thickness, and directlygraphite to titanium. Temperature of 850 °C wastaken in welding with nickel interlayer, and thatwithout interlayer made 1100 °C. Such high tem-peratures of DW do not allow using these meth-ods in given assemblies.

Brazing allows regulating a stress level bychanging temperature interval of cooling via se-lection of braze alloys with necessary meltingtemperature, providing serviceability of the as-semblies under specific conditions. The calcula-tions showed that application of low-temperaturebraze alloys with melting temperature up to250 °C guarantees production of defect-freejoints. For example, POS-50 braze alloy withmelting temperature 209 °C provides for tensilestrength equal 36 MPa, i.e. at graphite level.Taking into account lead toxicity, it is reasonableto use substitutes of tin-lead braze alloys, forexample, low-temperature ones developed at theE.O. Paton Electric Welding Institute, in whichlead is replaced by small amounts of silver. Ifgraphite wetting by braze alloy requires highertemperature, than brazing is reasonable to be car-ried out by two-step technology with pretinningof graphite surface. Thus, graphite-to-copperbrazing with further slow cooling is more prac-tical. Time of cooling is to be calculated for spe-cific assembly using copper creep parameters.

Conclusions

1. All constituents of stresses and deformationsare significantly changed with cooling rate vari-

ation. Further holding at 500 °C changes themmore.

2. Reduction of cooling rate from the point ofview of development of plastic deformations ismore efficient than increase of holding time aftertemperature reduction to 500 °C. This promotesfor reduction of the level of stresses in the as-sembly at cooling.

3. Internal surface has the highest risk of crackformation after cooling, where in all variants,except for variant 5, the maximum tensile stressesreach graphite bending strength.

4. Variant 5 (gradual reduction of temperaturefrom 900 to 500 °C during 6000 s), at whichtensile stresses do not exceed 25 MPa, is consid-ered the most optimum from the point of viewof reduction of danger of crack formation ingraphite after cooling to 500 °C.

5. Variants with quick cooling and furtherholding are not reasonable since large stresses onthe external surface of graphite are developed atthem.

6. Brazing with low-temperature lead-freebraze alloys is the most practical for productionof copper-graphite assemblies from low-strengthgraphite.

1. Ermolaev, G.V., Kvasnytsky, V.V., Kvasnytsky,V.F. et al. (2015) Brazing of materials: Manual. Ed.by V.F. Khorunov at al. Mykolaiv: NUK.

2. Kazakov, N.F. (1976) Diffusion welding of materi-als. Moscow: Mashinostroenie.

3. Kvasnitsky, V.V., Ermolaev, G.V., Matvienko, M.V.(2008) Influence of plastic deformation on stress-strain state in diffusion welding of dissimilar metalsapplicable to cylinder—cylinder and bush—bush assem-blies. Zbirnyk Nauk. Prats NUK, 1, 100—107.

4. Makhnenko, V.I., Kvasnitsky, V.V., Ermolaev, G.V.(2008) Stress-strain state of diffusion bonds betweenmetals with different physical-mechanical properties.The Paton Welding J., 8, 2—6.

5. Kvasnitsky, V.V., Ermolaev, G.V., Matvienko, M.V.(2009) Influence of strength and creep resistance onresidual stress-strain state of metal-ceramic joints.Zbirnyk Nauk. Prats NUK, 3, 83—92.

6. Kvasnitsky, V.V., Ermolaev, G.V., Matvienko, M.V.(2008) Influence of geometry of dissimilar materialsparts on stress-strain state in diffusion welding.Ibid., 5, 42—46.

7. Kvasnitsky, V.V., Matvienko, M.V., Ermolaev, G.V.(2009) Influence of the ratio of dimensions of cylin-drical parts from dissimilar materials on their stress-strain state in diffusion welding. The Paton WeldingJ., 8, 17—20.

Received 21.07.2015

11/2015 23

THERMAL-PHYSICAL PECULIARITIESOF GAS-SHIELDED PULSE-ARC WELDING

USING NON-CONSUMABLE ELECTRODE (Review)

A.A. SLIVINSKY, L.A. ZHDANOV and V.V. KOROTENKONTUU «Kiev Polytechnic Institute»

37 Pobeda Ave., 03056, Kiev, Ukraine. E-mail: o.slyvinsky@gmail.com

The work studies the main problems of development and possible perspectives for further improvement ofinert-gas non-consumable pulse-arc welding. Main technological possibilities and thermal-physical peculi-arities of the process are given depending on variation of welding mode parameters. Considered are thedirections for development of methods of mathematical modeling of welding process and possible ways oftheir improvement. Necessity in further investigations of the process for determination of parametric de-pendencies and development for engineering and scientific purposes of optimum modes of pulse-arc weldingusing tungsten cathode was grounded. 41 Ref., 7 Figures.

Keywo r d s : pulse TIG welding, thermal cycles,mode rigidity, gas-dynamic characteristics, transitionprocesses, discretization, precision, thermocouple, finiteelement method.

TIG welding starts active development in periodof aircraft and space engineering, namely in the1960s of the last century. Today TIG welding isone of the most wide-spread methods for joiningof metals and alloys. It is widely used in industrybranches with high requirements to welds andstructures, particularly in aircraft and rocket pro-duction [1—6].

Equipment for alternating current weldingwith pile-up of pulses for aluminum joining wasdeveloped till the end of 1960s. It was basedon mechanical contactors and performed non-consumable pulse-arc welding. At that, low-amperage stationary arc was kept between thepulses, that allowed for significant process sta-bilization and increase of welded joint quality[7—9].

Development of TIG welding together withapplication of different consumables and auxil-iary materials (fluxes, pastes, filler wires and gasmixtures) is based on methods and means of changeof process energy characteristics, that provides forefficient influence on heat input in welding [10].Thus, in proper time application was found forsuch methods of welding as step-arc welding [11]and modulated current welding [12].

Further development of element base, powerelectronics as well as microprocessor and digitalequipment resulted in expansion of amount ofcontrolled values and regulated parameters ofthe mode. It is 7—10 for pulse-arc welding in

modern current power sources. At that, pulsecurrent (Ip), break current (Ib), duration of pulse(tp) and breaks (tb), time of pulse growth (tup)and time of pulse drop (tdown) and shape of pulsehave direct effect on peculiarities of existence ofarc discharge. Arc voltage is set by length ofinter-arc gap, i.e. set by operator.

Obviously that correct selection of weldingmode is a complex and long process. Therefore,synergetic power sources, in which a programsets the optimum relationship of parameters ofmode in the power source, find more and moreapplication at current time. In order to receivecorresponding software program, providing nec-essary mode of power source, an operator shallset material, its thickness, electrode diameter,welding position and welding current (stationaryor pulse mode). At that, application of advancedcomputer technologies using PC and communi-cation technologies, namely Ethernet and WiFi,are provided for easy selection of welding modeparameters. Development of PC—welding powersource systems has already allowed tracing allparameters of welding mode as well as carryingtheir analysis and adjustment in automatic modedirectly via software set on PC (ewm Xnet tech-nology) [13] from EWM and Fronius DigitalRevolution [14]).

Number of investigations was carried out andlarge amount of welding equipment with differ-ent capabilities was developed from the appear-ance of non-consumable pulse welding up to pre-sent moment. However, large number of parame-ters of welding mode regulation complicates itsselection and designation for specific task. It is

© A.A. SLIVINSKY, L.A. ZHDANOV and V.V. KOROTENKO, 2015

24 11/2015

particularly relevant for sheet metal, where in-crease of efficiency is important, but formationof burning through and significant deformationof structure shall not be allowed. Modern ad-vanced technologies of regulation of process ofpulsed-current argon-gas tungsten-electrode arcwelding (PTIGW) do not always allow achievingoptimum result, first of all, due to large set ofprocess regulated parameters and their relation-ships. Therefore, investigation of their effect ongeometry of welded joint, which depends on weldpool shape and being determined by physicalcharacteristics of plasma of arc discharge, is stillrelevant task.

Balance between efficiency and quality ofwelded joint is to be kept in welding of sheetmetal. Heat supply in pulse mode input allowssignificant expansion of possibilities of regula-tion of heat input into base metal and its pene-tration. In general, base metal penetration andweld formation can be represented using the fol-lowing scheme.

Input of heat Q takes place in course of currentpulse of tp duration. It provokes melting of basemetal with formation of weld pool in form ofpoint Q = qatp, where qa is the arc effectivepower, and tp is the duration of pulse of principlecurrent. The weld pool is cooled till complete orpartial solidification in course of tb. The nextheat pulse will melt the same spot weld pool atS = v(tp + tb) distance from previous point, wherev is the welding speed.

In such a way, the weld is formed by meansof periodical melting of separate points with setstep S. Level of points overlapping is determinedby their size and step. Amount of heat per unitof weld length makes Q/S = qatp/v(tp + tb),where qatp/tp + tb is the average power of peri-odic arc.

Study [15] shows that process of pulse arcwelding to a first approximation can be consid-ered as welding process with direct current arcof qr/qa(1 + tb/tp) power. Calculated currentat that makes Ir = Ip/(1 + tb/tр). Relationshiptb/tp characterizes current frequency and stipu-lates penetration ability of periodic arc. Thegreater it is at the same calculated current Ir(average power), the more pulse current (pulsepower) is. Dimensionless tb/tp value is a tech-nological parameter characterizing ability topenetration of periodic (pulse) arc at set pulseenergy and cycle duration. It is called rigidity ofmode of pulse arc welding (G).

Thus, pulse arc penetration ability can be ef-fected by change of mode rigidity and absenceof variation in current value. Comparative inves-

tigations on stationary and pulse arc welding of3 mm thick 12Kh18N9T steel, given in [15],showed that increase of rigidity promotes for riseof value of total heat efficiency (Figure 1) underconditions of equal average rate of heat inputand welding speed.

The attention is also paid to the fact [15] thatincrease of G under condition of similar penetra-tion provides for reduction of zones of plasticdeformations (heating above 600 °C) (Figure 2).

Work [16] according to the results of meas-urement of thermal cycles showed that rise ofmode rigidity at the same currents significantlyreduces HAZ (temperature in test points beingat 2 mm distance from welding axis decreasedper 200 °C at rise of G from 0.3 to 3). The resultspublished in [17] (Figure 3) verify this. Thus,it can be stated that mode rigidity has decisiveeffect on form of welding thermal cycles and timeof weld pool existence and can serve as generalintegral parameter of PTIGW process.

Studying PTIGW processes, in addition totime characteristic of pulses, requires outliningsuch parameters of modes as pulse current (Ip),base current value (Ib) and welding speed (v),which determine a rate of energy input. If Ip isthe parameter, first of all, effecting the base met-al penetration, then Ib, according to [18], is usedfor stabilizing arc discharge. Thus, Ib zero valuepromotes for rise of possibility of so-called erraticarc, that results in undetermined location of areasof arc excitation on electrode and reduces arcprocess stability.

Another peculiarity of PTIGW is deformationof shape of arc discharge with possible laggingof anode spot. Author of study [19] explains this

Figure 1. Mode rigidity versus total heat efficiency [15]

Figure 2. Calculation isotherms T = 600 °C at differentmodes of welding using pulsed and stationary arc [15]

11/2015 25

effect by the fact that evaporation of base metal,taking place during current pulse, supplies to arcdischarge the atoms of lower ionizing potentialthan that of shielding gas atoms, but simultane-ously promotes for decrease of level of heat con-traction of arc column, that, in turn, results inhigh values of voltage drop during the break.This generates contradiction between availablefavorable conditions of arc existence on weld met-al melted by current pulse and necessity of itsmovement in less favorable conditions of coldmetal.

Based on principle of Steenbeck thermody-namic minimum [20], lags of anode spot shallappear on molten metal from time to time inprocesses of welding, since the molten metal pro-vides for minimum arc voltage. The anode spotmoves synchronously with molten metal pool,lagging from electrode movement before voltageof stretched arc do not become equal to anodeprocess excitation voltage. Due to this arc move-ment can be of stepwise nature. Such a lag ofanode spot results in significant for oscillationof arc voltage per 10—15 V [21], that, in turn,provokes oscillations of heat power and as a con-sequence unstable base metal penetration.

Thus, it can be noted that spatial instabilityof moving arc significantly limits area of allow-able parameters of welding mode. Therefore, sup-porting of arc column ionizing requires keepingsome arc current before the next pulse within the

time gaps between pulses. Thus, stability ofPTIGW process can be reached namely with thehelp correct selection of base current value, keep-ing arc column ionization, and frequency ofpulses.

A great deal of current investigations [22—25]is dedicated to selection of optimum parameters.Quantitative verification of investigations is rep-resented in overwhelming majority by mathe-matical dependencies being received usingTaguchi method. Effect of parameters of modeof pulse-arc welding on base metal penetrationfor the range of current pulsing frequency up to10 (PTIGW) and 100—250 Hz (HF-PTIGW) wasevaluated in [26] by given procedure. It can beseen from diagram (Figure 4) that tp/tb rela-tionship (value reverse to mode rigidity) and cur-rent pulsing frequency f-p have the main effecton process of penetration and formation ofwelded joint in PTIGW as well as HF-PTIGW.At the same time, such parameters of mode asbase and pulse currents have the least effect onpenetration. However, these are the parametersinfluencing welding process stability and, respec-tively, quality of welded joint formation.

There is a great deal of works dedicated toPTIGW process simulation. At that, physical-mathematical modeling of characteristics of arcpulse discharge is usually a basis for experimentaldata, received with the help of arc soundingmethod [27, 28].

Results of mathematical modeling of arc dis-charge between tungsten cathode and copperwater-cooled anode, given in [29], showed thatexistence of arc discharge in welding using pulse-periodic mode is accompanied by significantchanges of electromagnetic, thermal and gas dy-namic characteristics of arc plasma.

Presented results of simulation show that tran-sition processes play an important role in heat inputdistribution over the base metal. Their experimen-tal investigation can be carried out based on analy-sis of electron oscillograms of energy parametersof arc discharge, i.e. current and voltage.

Selection of characteristics of primary probesand registration devices, namely analog-digitaltransducers (ADT), is very important when per-forming such oscillographic testing. Procedure onselection of the primary probes for generation ofsuitable digital signal image is described in detailsin work [30], and principles of selection of ADTfor oscillographic testing are given in [31].

Current approaches to simulation of physicalprocesses, taking place in arc discharge plasmaduring PTIGW, described in works [32—36], pro-vides for application of finite element method

Figure 4. Diagrams of distribution of effect of main pa-rameters of PTIGW (a) and HF-PTIGW (b) on base metalpenetration [26]

Figure 3. Thermal cycles measured in welding with pulsed(1—3) and stationary (4—6) arc for points at 4, 6 and 8 mmdistance from weld axis [17]

26 11/2015

(FEM). Thus, results, received in [36], are basedon application of infra-red chamber and numeri-cal FEM-simulation using ComsolScript com-plex. At that, calculation of temperature distri-bution in cathode and anode area and in plasmaarc is carried out using a system of main classicMaxwell’s equations. Simulation of magnetic-hydro-dynamic processes showed that differentvarious take place in weld pool at different mo-ments of time and they significantly depend onelectrodynamic characteristics of plasma columnas well as its temperature (Figure 5).

It was determined that the maximum tempera-ture of tungsten cathode made 3500 K in pulsemoment and 3200 K at break moment. The maxi-mum temperature of plasma area was 14250 K inbreak moment and 16540 K at pulse current ef-fect. It is the well-known fact that these valuessignificantly depend on material of anode, valueof welding current and distance between the elec-trodes. Results of calculation also show natureof change of penetration depth and width de-pending on amount of pulses at the same pointof metal. This allows talking about heat satura-tion depending on amount of pulses (Figure 6).

The main thesis of results presented in [36] isa statement of authors that no change of shapetakes place in cross-section of the weld pool in

pulse arc welding even at significant differencein base and pulse values of current, that allowsfor regulating these parameters in a wide rangeduring sheet metal welding. At the same time,this work assumed the following simplificationsduring simulation, namely thermodynamic equi-librium takes place in arc plasma, and effect ofmetal vapors is not considered. The latter as-sumption significantly reduces efficiency of ap-plication of this model on practice. In generalcase, it should be noted that the model is designedfor spot welding, i.e. when power source is sta-tionary. This also introduces significant errors inprediction of base metal penetration.

It is the well-known fact that structural trans-formations in metal and stress-strain state ofwelded joint and structure in whole have decisiveeffect on welded joint mechanical properties.These factors are determined by nature of weldingheat processes. And if simulation of stationarywelding process is sufficiently well described inscientific sources, then of information on calcu-lations of heat processes in non-consumablepulse-arc welding is not enough. It is the well-known fact that thermal-physical situation inpulse-arc welding has significant differences fromthat in stationary arc welding, i.e. process of heatdistribution becomes non-stationary, cyclic

Figure 5. Results of simulation of magnetic-hydro-dynamic processes in weld pool for periods of effect of base current(left) and pulse current (right) at time moment of 1 (a), 5 (b), 10 (c) and 15 (d) s from the PTIGW beginning at Ip == 160 A, Ib = 80 A and f-p = 1 Hz [36]

11/2015 27

change of processes of heating and cooling ineach point of weld and near-weld zone takesplace.

Schemes of spot, linear and volumetric powersources are used for calculation of heat processesin TIG welding. The basics for calculation ofheat processes in welding were founded, as eve-rybody knows, by Rozental and Rykalin. Seriesof simplifications are mainly used in applicationof Rykalin calculation schemes for non-consum-able pulse-arc welding. Works [37, 38] can bean example. Their results are rather qualitativethan quantitative, and, besides, such an approachdoes not allow predicting the weld pool shapeand it makes it inconvenient for determinationof pulse welding optimum parameters.

Application of volumetric heat sources [39] isnecessary for more accurate prediction of pene-tration depth as well as thermal condition of basemetal. Procedure of FEM-simulation of heatprocesses is the most widely used at current time.It considers volumetric heat sources of ellipsoid

form and in first approximation recall a weldpool shape [40]. Work [41] provides for the re-sults of PTIGW simulation for heat source rep-resented in form of parabola of rotation withnormal (Gaussian) power distribution. At that,size of the source was experimentally determined,using images received with the help of speed in-fra-red CCD-camera. The simulation was carriedout in two different statements, namely whenthe heat source is set in form of parabola of ro-tation at the moment of base current influence,and in form of Gaussian model (Figure 7, dy-namic model 1) at the moment of principle cur-rent influence, as well as in the case when it isrepresented in form of Gaussian model for themoment of break and moment of pulse of heatsource (Figure 7, dynamic model 2).

It can be seen from given thermal cycles thatboth dynamic models provide for satisfactoryconformity with experimental thermal cycle onmaximum temperatures, but that can not be saidabout nature of oscillation of temperatures re-ceived experimentally and mathematically. It isobvious that welding with periodic change ofpower, typical for PTIGW process, also promotesfor periodic change of metal temperature, thatcan’t be observed on experimental curve receivedby authors of work [41]. Probably, such mis-match is related with peculiarities of measure-ment equipment. Except for similarity of tem-perature calculation and experiment, an impor-tant factor is a prediction of penetration shape,that is difficult to realize using proposed proce-dure.

Conclusion

It is determined based on results of reference dataon peculiarities of gas-shielded nonconsumable-electrode pulse-arc welding that the main prob-lem with practical application of this method is

Figure 7. Welding thermal cycles, received experimentallyand at FEM-simulation, for points located at 40 mm distancefrom beginning of welding, 6 mm from weld axis and at4 mm depth for Iр = 160 A, Ib = 80 A, f-p = 1 Hz, tp == 0.5 s and tb = 0.5 s [41]

Figure 6. Comparison of geometry characteristics of weld pool for DC welding and PTIGW [36]

28 11/2015

absence of possibility for prediction of effect ofwelding mode parameters on physical and tech-nological properties of arc discharge and, as aresult, weld geometry characteristics.

It is shown that the main technological char-acteristic of this method of welding is mode ri-gidity.

It is found that such parameters of modes ascurrent and voltage in the pulse as well as basevalues of current and voltage, particularly,short-term changes of energy parameters of arcdischarge, have to be outlined in addition topulse time characteristics during PTIGWanalysis.

Requirements to ADT, which is used for reg-istration of energy parameters of PTIGW processand transition processes, were formulated basedon analysis of reference data and Nyquist theo-rem. Besides, application of synchronized videoregistration of pulse arc discharge is relevant forfurther decoding of oscillograms.

Results of simulation of physical processes inplasma and magnetic-hydro-dynamic processes inthe weld pool show that various processes takeplace in the pool at different moments of timeand they significantly depend on electrodynamiccharacteristics of plasma column as well as onits temperature.

Finite-element modeling of heat processes oftungsten-electrode pulse-arc welding requires ap-plication of calculation scheme of volumetric heatsource of ellipsoid form and presence of calcula-tion-experimental time dependencies of the mainmacroparameters for specific mathematical de-scription of time variable of arc power.

Thus, available experimental data and resultsof simulation of PTIGW process reflect existenceof arc discharge plasma at moments of currentpulse generation and its effect on behavior ofweld pool metal. Processes, accompanying ap-pearance of the main arc discharge in pulse andits extinction, i.e. transition short-term proc-esses, are not considered due to absence of accu-rate experimental data. This verifies relevance ofdevelopment of optimum procedure for furtherinvestigation of these processes under PTIGWconditions, which allows development of moreaccurate procedure for prediction of effect ofwelding mode on geometry and technologicalcharacteristics of welded joint.

1. Brodsky, A.Ya. (1956) Tungsten-electrode argon-arcwelding. Moscow: Mashgiz.

2. Petrov, A.V. (1961) Technology of inert-gas arcwelding: Reference book on welding, Vol. 2, 327—375. Ed. by E.V. Sokolov. Moscow: Mashgiz.

3. (1985) Actual problems of welding of non-ferrousmetals. In: Proc. of 11th All-Union Conf. Kiev:Naukova Dumka.

4. (1985) Metals science of aluminium alloys. Ed. byS.T. Kishkin. Moscow: Nauka.

5. Russo, V.L. (1962) Inert-gas welding of aluminiumalloys. Leningrad: Sudpromgiz.

6. Shakhanov, S.B. (2007) Theory and technology ofwelding production in rocket industry: Manual. St.-Petersburg: Balt.GTU.

7. Gurevich, S.M., Zamkov, V.N., Kushnirenko, N.A.(1965) Increase of penetration efficiency of titaniumalloys in argon-arc welding. Avtomatich. Svarka, 9,1—7.

8. Alov, A.A., Shmakov, V.M. (1962) Argon-arc weld-ing with auxiliary argon flow. Svarochn. Proizvod-stvo, 3, 13—16.

9. Terry, S.A., Tyler, W.T. (1958) Inert-gas tungsten-arc welding. Welding and Metal Fabr., 2, 58—61.

10. Paton, B.E. (1963) Further development of systemsof automatic control and regulation of welding proc-esses. Avtomatich. Svarka, 5, 1—6.

11. Razmyshlyaev, A.D., Patrikeev, A.I., Maevsky, V.R.(1988) Calculated definition of pool contour in step-by-step-arc welding of thick metal. Svarochn. Proiz-vodstvo, 5, 35—37.

12. Vagner, F.A. (1980) Equipment and methods ofpulsed arc welding. Moscow: Energiya.

13. Network-based welding processes analyzing, control-ling and managing. pdf. https//www.ewm-group.com/ru/service/downloads/ brochures-handouts-and-manuals/2698-ewm-xnet-brochure/download.html

14. (2002) Weld+vision FRONIUS MAGAZINE, 11.pdf. http://www.fronius.com/cps/rde/xbcr/SID-76553F1D-04EC16FD/fronius_brasil/4000062136_weld_ vision_Nr 9_en.pdf

15. Petrov, A.V., Slavin, G.A. (1966) Study of techno-logical features of pulsed arc. Svarochn. Proizvod-stvo, 2, 1—4.

16. Slyvinsky, O.A., Bojko, V.P., Prepiyalo, A.O.(2013) Mathematical modeling of heat processes inargon-arc welding of thin stainless steel of ferriteclass. In: Proc. of 6th All-Ukr. Interbranch Sci.-Techn. Conf. of Students, Post-Graduate Studentsand Researches on Welding and Related Processesand Technologies (Kiev, 29—31 May 2013), 14.

17. Slyvinsky, O.A., Korotenko, V.V. (2014) Mathemati-cal modeling of heat processes of pulsed tungstenargon-arc welding of thin stainless steel. In: Proc. of7th All-Ukr. Interbranch Sci.-Techn. Conf. of Stu-dents, Post-Graduate Students and Researches onWelding and Related Processes and Technologies(Kiev, 14—16 May 2014), 19.

18. Kovalev, I.M. (1973) Some methods for stabilizing ofunstable arc with non-consumable electrode.Svarochn. Proizvodstvo, 6, 3—5.

19. Kovalev, I.M. (1972) Space stability of moving arcwith non-consumable cathode. Ibid., 8, 1—3.

20. Finkelnburg, V., Mekker, G. (1961) Electric arc andthermal plasma. Moscow: IL.

21. Kozakov, Yu.M., Stolbov, V.I., Koryagin, K.B.(1986) Lagging of anode spot of moving welding arc.Svarochn. Proizvodstvo, 10, 19—21.

22. Kim, D., Rhee, S. (2001) Optimization of arc weld-ing process parameters using a genetic algorithm.Welding J., July, 184—189.

23. Giridharan, P.K., Murugan, N. (2009) Optimizationof pulsed GTA welding process parameters for weld-ing of AISI 304L stainless steel sheets. Int. J. Ad-vanced Manufact. Technology, Issue 5, 40, 478—489.

24. Babu, S., Senthil Kumar, T. (2008) Optimizingpulsed current gas tungsten arc welding parametersof AA6061 aluminium alloy using Hooke and Jeevesalgorithm. Transact. of Nonferrous Metals Soc. ofChina, 18, 1028—1036.

25. Chakravarthy, M.P., Ramanaiah, N., Sundara SivaRao, B.S.K. (2013) Process parameters optimization

11/2015 29

for pulsed TIG welding of 70/30 Cu—Ni alloy weldsusing Taguchi technique. Int. J. Mechanical, Aero-space, Industrial, Mechatronic and Manufact. Eng.,7(4), 342—348.

26. Arivarasu, M., Devendranath Ramkumar, K., Ari-vazhagan, N. (2014) Comparative studies of high andlow frequency pulsing on the aspect ratio of weldbead in gas tungsten arc welded AISI 304L plates.Proc. Eng., 97, 871—880.

27. Dyatlov, V.I. (1961) Volt-ampere characteristic of con-stricted electric arc. Avtomatich. Svarka, 1, 23—26.

28. Rabkin, D.M., Ivanova, O.N. (1968) Investigation ofarc in tungsten arc welding. Ibid., 5, 16—20.

29. Krivtsun, I.V., Krikent, I.V., Demchenko, V.F.(2013) Modelling of dynamic characteristics of apulsed arc with refractory cathode. The Paton Weld-ing J., 7, 13—23.

30. Zhdanov, L.A., Slyvinsky, A.M., Kotyk, V.T. et al.(2003) Possibility of analog-digital converter applica-tion for study of alternating current welding arc.Mashynoznavstvo, 2, 38—41.

31. Zhdanov, L.A., Slyvinsky, A.M., Kopersak, V.M. etal. (2004) Study of alternating current welding arcby personal computer. Nauk. Visti NTUU KPI, 3,49—55.

32. Wu, C.S., Gao, J.Q. (2001) Analysis of the heat fluxdistribution at the anode of a TIG welding arc. Com-put. Mater. Sci., 24, 323—327.

33. Lu, F., Tang, X., Yao, Yu.S. (2006) Numerical simu-lation on interaction between TIG welding arc andweld pool. Ibid., 35, 458—465.

34. Tanaka, M., Lowke, J.J. (2007) Predictions of weldpool profiles using plasma physics. J. Phys. D: Appl.Phys., 40, 1—23.

35. Traidia, A., Roger, F., Guyot, E. (2010) Optimal pa-rameters for pulsed gas tungsten arc welding in par-tially and fully penetrated weld pools. Int. J. Therm.Sci., 49, 1197—1208.

36. Traidia, A., Roger, F. (2011) Numerical and experi-mental study of arc and weld pool behavior forpulsed current GTA welding. Int. J. Heat and MassTransfer, 54, 2163—2179.

37. Petrov, A.V. (1967) Application of sources methodfor calculation of heat processes in pulsed-arc weld-ing. Fizika i Khimiya Obrab. Materialov, 5, 15—25.

38. Vagner, F.A. (1975) Calculation of temperatures inproduct during pulsed welding with exponentialpulse shape. Avtomatich. Svarka, 7, 13—15.

39. Kiselev, S.N., Kiselev, A.S., Kurkin, A.S. et al.(1998) Modern aspects of computer modeling of heat,deformation processes and structure formation inwelding and related technologies. Svarochn. Proiz-vodstvo, 10, 16—24.

40. Goldak, J.A., Chakravarti, A., Bibby, M. (1984) Anew finite element model for welding heat sources.Metallurg. Transact. B, Vol. 15, 229—305.

41. Zhang Tong, Zheng Zhentai, Zhao Ru (2013) A dy-namic welding heat source model in pulsed currentgas tungsten arc welding. J. Materials Proc. Tech-nology, 213, 2329—2338.

Received 28.07.2015

30 11/2015

MAIN TENDENCIES IN DEVELOPMENTOF PLASMA-ARC WELDING OF ALUMINIUM ALLOYS*

A.A. GRINYUK2, 3, V.N. KORZHIK1, 2, V.E. SHEVCHENKO1, 2, A.A. BABICH2,S.I. PELESHENKO4, V.G. CHAJKA2, A.F. TISHCHENKO2 and G.V. KOVBASENKO2

1Chinese-Ukrainian E.O. Paton Welding Institute(Guangdong General Research Institute of Industrial Technology)

(Guangzhou Research Institute of Non-Ferrous Metals, PRC)2E.O. Paton Electric Welding Institute, NASU

11 Bozhenko Str., 03680, Kiev, Ukraine. E-mail: office@paton.kiev.ua3NTUU «Kiev Polytechnic Institute»

6/2 Dashavskaya Str., 03056, Kiev, Ukraine. E-mail: andrey_grinyuk@ukr.net4South China University of Technology

510641, Guangzhou, PRC. E-mail: sviatoslav@qq.com

Publications, describing the characteristic technologies of aluminium alloy welding by an arc constrictedby high-velocity inert gas flow were analyzed. It is shown that plasma-arc welding (PAW) is furtherdevelopment of the process of nonconsumable-electrode inert-gas welding. It is established that duringdevelopment of PAW of aluminium alloys, there was a transition from alternating sinusoidal current toreverse polarity direct current, and furtheron to variable polarity asymmetrical current with rectangularcurrent waveform. A more promising direction of improvement of PAW equipment is transition fromspecialized power sources to modular design of PAW system, based on power sources applied for noncon-sumable electrode welding and plasma modules. Further path of improvement of the processes of aluminiumalloy PAW is combined or hybrid application of several heat sources, including the constricted arc andconsumable-electrode arc. In the authors’ opinion, hybrid consumable-electrode PAW with hollow anodeand axial wire feed in the most promising variant. 41 Ref., 17 Figures.

Keywo r d s : plasma-arc welding, aluminium alloys,alternating sinusoidal current, variable polarity asym-metrical current, hybrid plasma-arc welding, consumableelectrode

Aluminium and aluminium alloys take up thesecond place after steel as to production and con-sumption. Owing to a set of physical and me-chanical, corrosion and technological properties,aluminium-based light alloys are successfully ap-plied not only in development of flying vehicles,but also in other sectors of industry and construc-tion. Volumes of aluminium alloy application inmilitary equipment items, in shipbuilding, in fab-rication of automotive and railway vehicles, inelectrical engineering, in manufacture of cryo-genic and chemical apparatuses, in agriculturaland food industry are considerable. Alongside amultitude of diverse welding processes, arc weld-ing processes are more widely applied to producealuminium alloy permanent joints, which can bedivided into nonconsumable and consumableelectrode welding. Nonconsumable electrode

welding is applied in fabrication of structuresfrom semi-finished products of aluminium alloys2—12 mm thick. In its turn, nonconsumable elec-trode welding is performed by freely expandingand constricted arc.

The first mention of plasma jet energy appli-cation for metal treatment dates back to 1950s[1]. Nowadays the plasma jet is used for welding,cutting, coating deposition, surface hardening,etc. Transferred and non-transferred constrictedarc is applied for metal treatment. At transferredarc application the electric discharge runs be-tween the electrode and the item, and in the caseof non-transferred arc the discharge runs betweenthe electrode and the plasma nozzle. In this case,the treated item is exposed only to a plasma jetblown out through small diameter nozzle by inertgas flow.

Fusion welding with heating by constrictedplasma arc is a further development of inert-gasnonconsumable-electrode welding process [2].Another inner water-cooled copper nozzle is

© A.A. GRINYUK, V.N. KORZHIK, V.E. SHEVCHENKO, A.A. BABICH, S.I. PELESHENKO, V.G. CHAJKA, A.F. TISHCHENKO and G.V. KOVBASENKO, 2015

*Work was performed with financial support within Program of Foreign Experts in PRC #WQ20124400119, R&D Project of InnovationGroup of Guangdong Province #201101C0104901263 and International Project of Ministry of Science and Technology of PRC# 2013DFR70160.

11/2015 31

added to the torch design. Gas, blown outthrough inner nozzle hole of small diameter (from1 to 5 mm) forms a layer of thermal and electricinsulation around the arc, preventing its free ex-pansion. Welding zone is protected by inert gasflow from larger diameter outer nozzle (Fi-gure 1), and arc column temperature and its volt-age rise [2].

On the whole, analysis of the experience ofplasma arc application for welding showed thatit is a more versatile source of metal heating inwelding, compared with electric arc in noncon-sumable and consumable electrode welding.Compared to these methods, plasma providesdeeper penetration of metal at simultaneous re-duction of its melting volume, and it also has thefollowing advantages:

• higher plasma arc temperature (up to18,000—25,000 K);

• smaller arc diameter;• arc cylindrical shape (unlike the regular

conical shape);• plasma arc pressure on the metal 6—10 times

higher than that of a regular one;• ability tо maintain the arc at low currents

(0.2—30 A).Plasma arc application in welding also ensures

absence of metal spattering, smaller heating ofbase metal and, consequently, its smaller defor-mation, no need for edge preparation (grooving)in many cases due to high penetrability of theplasma arc, higher cost effectiveness and effi-ciency (by 2—3 times and more), reduction of thecost of machining at butt preparation for weld-ing, and of weld treatment after welding by 3—5times and more, lower consumption of weldingconsumables by 3 to 5 times or more, weld quality

improvement and possibility of 100 % automationof the processes of welding the longitudinal andcircumferential welds.

According to GOST 2601—84 «Metal welding.Terms and main definitions», fusion welding withheating by an arc constricted by gas flow, is calledplasma welding. In our opinion, such a definitionis more suitable to the process of welding by non-transferred constricted arc, which is usually notused for welding. Therefore, it is more correct toapply the term «plasma-arc welding», which speci-fies participation of both the arc energy, and plasmajet energy, in part heating.

Plasma-arc welding with alternating sinusoi-dal current. A feature of aluminium alloy arcwelding is the need to break up the surface re-fractory oxide film. For its effective breaking up,plasma-arc welding (PAW) should be performedat reverse polarity direct current or at variablepolarity current [3].

Similar to the case of nonconsumable elec-trode welding with freely expanding arc, the firstexperiments on PAW of aluminium alloys wereconducted with application of alternatingsinusoidal current of 50 Hz frequency. At theend of 1960s, PWI performed intensive studiesof this process [4, 5]. To ensure constricted arcrunning, a pilot arc module was added to a stand-ard machine for nonconsumable electrode alu-minium welding. The pilot arc ran between theelectrode and plasma-forming nozzle, ionizing thegap between the electrode and the item. At appli-cation of sinusoidal AC, oxide film breaking upproceeds in reverse polarity half-wave at electrodepositive. Up to 70 % of the arc heat evolves in theelectrode. In 1970—1971 the method of PAW withsinusoidal AC was applied in ship-building. Anadvanced automatic machine «Aluminij-1» fornonconsumable electrode welding of aluminium al-loys and UDG-701 power source were proposedfor its realization [6]. This welding process did notbecome widely accepted later on. There is no in-formation in publications about batch-producedmachines for PAW at sinusoidal AC.

Plasma-arc welding at reverse polarity directcurrent. PAW at reverse polarity was appliedfor the first time for joining 6.35 mm aluminiumalloy sheets by the staff of Thermal DynamicsCorporation [7]. The first studies in the SovietUnion on application of DCRP date back to thebeginning of 1970s [8, 9].

The best cathode breaking up of the film (cath-ode cleaning) in nonconsumable electrode weld-ing is achieved at reverse polarity. Plasmatronelectrode assembly is the anode, and the itemproper is the cathode. PAW at DCRP is easierto implement, compared to sinusoidal AC weld-ing. For instance, standard welding rectifiers can

Figure 1. Schematic of the process and shape of weld pene-tration in nonconsumable electrode arc (a) and plasma-arc(b) welding: 1 – tungsten electrode, 2 – shielding gasnozzle; 3 – welding arc; 4 – plasma-forming nozzle

32 11/2015

be used as the constricted arc power source. None-theless, this PAW process has a significant draw-back – a considerable amount of heat evolvesin the electrode assembly at reverse polaritywelding, leading to electrode premature failure.

In order to extend the service life of electrodeassembly, its intensive cooling is performed, tung-sten rods of large diameter (8—10 mm) or specialcopper electrodes with inserts of refractory mate-rials (tungsten, hafnium) are used. To extend plas-matron life, its components are made massive, thatincreases its overall dimensions.

Stability of running of a constricted directcurrent arc at reverse polarity depends on thequality of part surface preparation for welding.Presence of unremoved oxide films on the surfaceof items being welded can cause wandering ofthe heating spot, because of different emissivityof the oxide film and cleaned base metal [8, 9].At small currents this wandering will becomemore pronounced.

The process was widely applied for manufac-turing items from aluminium alloys for cryogenicapplications [10]. UPS-301 and UPS-503 ma-chines were manufactured in the USSR for PAWat DCRP, and welding could be also performedwith surfacing machine UPNS-304.

Studies of this variant of PAW of aluminiumare intensively pursued in Russia. Company«Shipbuilding and Ship Repair Center» devel-oped PPN-200 semi-automatic machine for PAWof aluminium at DCRP. Nonconsumable elec-trode welding machine Master-2500 of Kemppiis used as the power source [11]. Company«Plazmek» Group developed a versatile PAW sta-tion, including an inverter power source, modulefor pilot arc generation and plasma gas feeding,independent water-cooling unit and proprietaryplasmatrons. List of companies, manufacturingequipment for direct current PAW is extensive: itcan include companies from North and South Amer-ica, Europe and Asia. The following equipmentmodels can be noted: DigiPLUS A7P andDigiPLUS A7PO of IMC Engenharia de Soldagem(Brazil), Ultima150 of Thermal Arc (USA), PI400Plasma of Svejsemaskinefabrikken MigatronicA/S (Denmark), Plasmaweld 402 of L-TECSchweisstechnik (Germany), PSI 350plus of Kjell-berg Finsterwalde Schweisstechnik und Ver-schleissschutzsysteme (Germany), Tetrix 552 RCPlasma CW of EWM Hightec Welding Automat-ion (Germany), PW400 of Arcraft Plasma Equip-ment PVT. Ltd (India), Plasma 500XP of PowwelCo. Ltd (Korea), and others.

To increase the effectiveness of PAW atDCRP, Russian scientists proposed pulsed feedof plasma gases [12—14].

One of the negative points of DCRP applica-tion for PAW of aluminium alloys is a strongproneness of this welding process to hydrogengas porosity in the weld metal [15].

Plasmatron welding process proposed by Ino-con Technologie GmbH (Austria) is one of thevariants of PAW of aluminium alloys at directcurrent [16]. An ingenious shape of the weldingtorch is a feature of the welding process (Fi-gure 2). As reported by Inocon TechnologieGmbH staff, such a scheme of the welding processallows temperature in the active heating spot tobe increased (Figure 3), and welding to be per-formed without an additional module of pilot arcgeneration.

Plasma-arc welding of aluminium alloys withalternation of electrode and plasma-formingnozzle polarities. To reduce the adverse impactof heat evolved in the electrode at current passageduring reverse polarity half-wave, and preservesufficient cathode cleaning of the weld, PWI staffproposed an original method for applying voltageto the plasmatron [17]. During straight polarityhalf-wave the arc ran between the tungsten elec-trode and the item, and during the reverse po-larity half-wave the arc was excited between theitem and copper plasma-forming nozzle. Owingto such a schematic of energy application to theplasmatron, good penetration was achieved,tungsten electrode resistance was improved, andthe arc running during reverse polarity half-wave, ensured effective breaking up of the oxidefilms. However, this equipment for PAW of alu-minium alloys was not put into batch production.At present, German scientists revisited the ideaof separate feeding of energy to the electrode andthe plasma-forming nozzle. Their machine for

Figure 2. Appearance of Plasmatron welding torch (InoconTechnologie GmbH) [16]

11/2015 33

PAW with alternation of the electrode and theplasma-forming nozzle polarity also is still at thelaboratory stage.

Plasma-arc welding by variable polarityasymmetrical current with rectangular wave-form. Process of PAW by variable polarity asym-metrical current with rectangular current wave-form, called Variable Polarity Plasma Arc(VPPA) welding, was registered by HobartBrothers Inc. in 1978 [18]. A special module,forming an additional pulse at transition fromstraight to reverse polarity, was incorporated intothe design of welding machines generatingsinusoidal welding current, to ensure continuousarcing. Rectangular waveform of welding current

allowed ensuring reliable arc excitation at polar-ity reversal without application of a special mod-ule. To lower the thermal impact on the torchelectrode assembly, PAW by variable polaritycurrent with rectangular waveform was per-formed with prevalent flowing of straight polar-ity current. Minimum admissible duration of cur-rent flowing at reverse polarity is selected toensure effective cathode breaking up of the oxidefilms. The ratio of duration of current flowingat straight and reverse polarity was equal to ap-proximately 3:1. Current amplitude at reversepolarity was approximately by 30—40 A higherthan at straight polarity (Figure 4) [19]. Reduc-tion of the amount of heat evolving in electrode

Figure 4. Amplitude and duration of current flowing at straight and reverse polarity for VPPA welding aluminium alloysof different composition [18]

Figure 3. Temperature distribution in traditional torch for PAW (a) and Plasmatron welding torch of Inocon TechnologieGmbH (b) [16]: a – heat flow surface density of 103—105; b – 106 W/cm2

34 11/2015

assembly allowed increasing plasmatron perform-ance and reducing its geometrical dimensions.

In 1972, Sciaky proposed to Boeing to usevariable polarity asymmetrical current with rec-tangular waveform for PAW of aluminium alloys[20]. Sciaky power source, however, did not pro-vide arcing stability at variable polarity asym-metrical current. In 1974, Hobard Brothers Inc.started development of a power source and setof equipment for PAW at variable polarity asym-metrical current of rectangular waveform forBoeing. In 1979, the first such set of equipmentfor PAW was supplied to Marshall Space FlightCenter. The set was based on VP-300-S powersource, actual values of welding current weremonitored with Cyber-Tig 11 Series 800 Pro-grammer. This set of equipment and its sub-sequent analogs were widely used for fabricationof fuel tanks for liquid hydrogen and oxygen ofSpace Shuttle system [21—23]. In 1990s, HobartBrothers Inc. stopped manufacturing equipmentfor VPPA welding process. At the same time, thedemand of US space industry in the field of PAWof aluminium alloys by variable polarity asym-metrical current is satisfied by products of Li-burdi, Canada (LTP400-VP power source) (Fi-gure 5) and AMET, USA (VPC450 powersource). High cost of equipment and complexityof operation did not promote wide acceptance оfthis welding process by industry.

VPPA welding at asymmetrical current wasthe most effectively applied for gravity weldingof joints without using backing elements withforming grooves. Such a process was called «key-hole» (Figure 6). In this process the plasma jetpenetrates right through the weld pool moltenmetal, forming a through-thickness channel,which exists during the entire period of welding.Through-thickness penetration of the plasma jetcompletely eliminates oxide film inclusions inthe weld, and also reduces item deformation, dueto homogenizing of heating of the parts beingwelded by height. 100 % argon was used as plasmagas, and 100 % helium was the shielding gas.

More active application of PAW of aluminiumalloys by variable polarity asymmetrical currentbegan from 2000s. In Europe, a number of suchcompanies as Merkle (Germany), EWM (Ger-many), Castolin (Switzerland), SBI (Austria),Migatronic Automation (Denmark) startedbatch production of equipment for PAW by vari-able polarity asymmetrical current with rectan-gular waveform. Some of these systems – Tetrix350 AC/DC Plasma (EWM) and PMI-380AC/DC (SBI), can be used both for automaticand for manual PAW of aluminium alloys.

At the start of 1990s, PWI performed inves-tigations of PAW of aluminium alloys by variablepolarity asymmetrical current with rectangular

waveform using thyristor power sources with in-ductive energy storages. Investigation resultsshowed that just temporary asymmetry is suffi-cient for sound weld formation, i.e. prevalenceof the time of straight polarity current flowingand equality of straight and reverse polarity cur-rent amplitudes [24, 25].

In 2001, Fronius (Austria), based on PWItechnical assignment developed a set of equip-ment for PAW of aluminium alloys by variablepolarity asymmetrical current with rectangularwaverform. The complex includes MW450 powersource, FPM plasma module, filler wire feed

Figure 5. Appearance of LTP400-VP power source (Liburdi)for welding by VPPA process [18]

Figure 6. Block diagram of PAW in «keyhole» mode (ISFAachen scheme) [22]: 1 – plasma torch; 2 – weld; 3 –weld surface; 4 – weld root

11/2015 35

mechanism KD 4000, plasmatron displacementsystem, welding process control system FPA2003-Plasma, and independent water coolingunit (Figure 7). This equipment complex wasused in PWI to develop a technology of PAWof aluminium—lithium alloys by variable polarityasymmetrical current [26, 27].

Unlike equipment for PAW of aluminium al-loys of other companies, PAW system fromFronius is made by power source—plasma modulescheme. Such a design allows selection of a powersource of required power to solve the definedtasks. If not more than 150 A welding current isapplied during operation, it is possible to usepower sources of up to 220 A nominal power inthe PAW system. All Fronius batch-producedpower sources for nonconsumable-electrode ar-gon-arc welding by variable polarity asymmetri-cal current (power sources of MW series) can beused as power sources for PAW.

Application of FPA 2003-Plasma system forwelding process control allows significant expan-sion of the capabilities of PAW of aluminium

alloys by variable polarity asymmetrical current.So, the possibility of pulsed feed of plasma gasallows effectively preventing porosity in alu-minium alloy welding, and welding currentmodulation and filler wire feeding at variablerate provides sound formation of welds in orbitalwelding of position butt joints. Investigationsconducted at PWI confirmed that more preciseregulation of parameters of PAW by variable po-larity asymmetrical current, using FPA 2003-Plasma welding process control system, allows ap-plication of this welding process to produce per-manent joints of thin (1—2 mm) aluminium semi-finished products (Figure 8), as well as to performwelding in the mode of through-thickness penetra-tion of the plasma jet (Figure 9) [27].

Availability of machine and manual plasma-trons in the equipment range, as well as switchingof plasma module operation allows equally effi-cient application of Fronius equipment for bothautomatic and manual plasma welding.

At present FPM plasma module has alreadybeen removed from production, and the new Plas-maModule-10 with digital control, unfortu-nately, does not provide PAW by variable po-larity asymmetrical current.

Over the recent years several approaches haveemerged to development of equipment for PAWof aluminium alloys by variable polarity asymmet-rical current, namely development of specializedPAW systems and plasma modules (consoles) forconnection to batch-produced power sources fortungsten electrode welding. In our opinion, appli-cation of additional plasma modules for creationof PAW system based on power sources for non-consumable electrode welding will allow reductionof costs for setting up a welding station for con-stricted arc welding, improvement of versatility of

Figure 7. Fronius equipment for PAW of aluminium alloysby variable polarity asymmetrical current

Figure 8. Macrosection of welded joint of D16 alloy 2 mmthick produced by PAW in downhand position

Figure 9. Macrosection of welded joint of 6 mm alloy 1420(Al—Li—Mg) produced by downhand PAW on a backingwith forming groove (a) and in keyhole mode of plasma jetthrough-thickness penetration (b)

36 11/2015

equipment complex application, variation of thevalue of required power of welding sources, morecomplete realization of the technological capa-bilities of modern power sources for nonconsu-mable electrode welding.

Laser-plasma welding of aluminium alloys.Laser beam, being electrically neutral, can com-bine readily with consumable and nonconsu-mable electrodes at running in a common pool.Laser beam combines equally well in hybridwelding processes with nonconsumable-electrodeconstricted arc, running both at DCRP and vari-able polarity asymmetrical current. Hybrid la-ser-plasma welding of aluminium alloys allowsnarrowing the weld, increasing the penetrationdepth and welding speed, compared to regularPAW [28, 29].

Plasma-arc welding with powder applicationas filler. Aluminium alloy wire was traditionallyused as filler material in PAW of aluminium al-loys, both by direct and variable polarity current.Scientists of Chemnitz and Ilmenau Universities(Germany), suggested application of aluminiumpowder as filler material in PAW of aluminiumalloys [30]. Aluminium powder application inPAW has several advantages, namely, interfer-ence from filler wire guides at movement ofrobotic «arms» is eliminated, it is possible toselect filler material composition in a broadrange, and quickly change the welding directionwithout turning the plasmatron with wire feedsystem fastened on it. The main disadvantage ofapplication of aluminium powder as filler mate-rial is the large area, covered by oxide film, com-pared to filler wire. Extensive powder surfacecoated by oxide film, limited application of thiswelding process for joining less than 2 mm thickparts. The oxide film does not effectively breakup in the arc at small values of welding current.At this moment the possibility of increasing arcpower through application of helium-containingmixtures is being studied [31].

Two-sided plasma-arc welding and noncon-sumable electrode welding. US scientists pro-posed joint application of plasmatron and non-consumable electrode torch for aluminium alloywelding (Figure 10). In such a scheme the non-consumable electrode torch is connected to theitem instead of the cable mass [32]. Constrictednonconsumable electrode arc and freely expand-ing nonconsumable electrode arc, powered by onesource, are simultaneously applied to the itemfrom different sides. Such simultaneous arcingincreases the voltage of both the arcs and pene-tration depth, and also improves the stability ofthrough-thickness channel in welding in themode of through-thickness penetration of theplasma arc [33].

Combined application of plasma-arc weldingand consumable electrode welding. Modern de-velopment of fabrication of aluminium struc-tures, in particular, for extended constructionsof ground and sea transportation, requires in-creasing the welding speed. One of the methodsto solve such a task can be combined applicationof several welding processes, for instance, PAWand consumable electrode welding. So, in 2002—2005, PWI developed a technology of combinedapplication of PAW by variable polarity asym-metrical current without filler material and con-sumable-electrode argon-arc welding (Fi-gure 11). The distance between the two heatsources was 65 mm, and common weld pool wasabsent. Nonconsumable electrode constrictedplasma arc performed item heating and partialremoval of hydrogen from the metal beingmelted. Weld convexity was formed due to elec-trode wire melting. Combined application ofPAW by variable polarity asymmetrical currentand consumable-electrode argon-arc welding at

Figure 10. Two-sided PAW and nonconsumable electrodewelding [32]

Figure 11. Position of plasmatron and consumable electrodetorch at combined application of PAW and consumable elec-trode arc welding

11/2015 37

achievement of the same penetration depth asthat in regular consumable-electrode argon-arcwelding, allows reducing the width of the weld

and base metal softening zone in welding, as wellas the degree of this softening (Figure 12).

Hybrid consumable-electrode plasma-arcwelding. One of the promising directions of de-velopment of the processes of aluminium alloyPAW is application of simultaneous arcing fromseveral heat sources into a common pool, such asnonconsumable electrode constricted arc and con-sumable electrode arc, to form the weld pool(Figure 13). It was called Plasma-MIG abroad.This process was mentioned in 1970s, and thefirst patent for this welding process is owned byPhilips Corporation [34]. Consumable electrodearc runs inside a nonconsumable electrode con-stricted arc. At present there exist two main sche-matics of this process realization: with annularanode (Figure 14) and with lateral position ofthe anode (Figure 15). Such a hybrid applicationof both the arcs allows reducing electrode metalspatter, increasing penetration depth and weld-ing speed, and applying the process for weldingthick items with groove preparation.

Investigations of hybrid consumable-electrodePAW are pursued in many universities in theworld, in particular, at Chemnitz Technical Uni-versity (Germany). A device was developed there,providing switching on and simultaneous operationof power sources for nonconsumable-electrodePAW at DCRP and power sources for consumableelectrode welding at DCRP [35]. Investigationsin this direction are also conducted at SLVMuenchen (Germany). Similar research is per-formed at Perm State University (Russia) [36, 37].In Ukraine consumable-electrode PAW was stud-ied at Priazovsky Technical University [38]. Fea-tures of the process of consumable-electrodePAW were also studied in China, Japan and Bra-zil [39, 40].

TBI (Germany) set up manufacture by indi-vidual orders of PLM 500 and PLM 600 torcheswith annular anode for consumable-electrodePAW, which can stand up to 250 and 300 Acurrent load of DCRP for the torch plasma partand consumable electrode assembly, respectively(see Figure 14).

Plasma Laser Technologies Ltd. (Israel) pro-posed a process of hybrid consumable-electrodePAW, called Super-MIG, which involves simul-taneous running of nonconsumable electrode con-stricted arc and consumable electrode arc into acommon pool. Consumable electrode arc runs tothe side of nonconsumable electrode arc (see Fi-gure 15) [41].

Despite the high interest to hybrid consum-able-electrode PAW with axial feed of electrodewire, batch-production of equipment for realiza-

Figure 13. Schematic of the process (a) and appearance ofconsumable electrode welding arc: 1 – constricted plasmaarc power source; 2 – MIG-arc power source; 3 – non-consumable electrode; 4 – stabilizing nozzle; 5 – shieldingnozzle; 6 – consumable electrode; 7 – current conduit;8 – constricted plasma arc; 9 – MIG arc; 10 – item

Figure 12. Hardness of deposits on 12 mm AMg6N alloyproduced at combined application of PAW and consumableelectrode welding (1) and at regular consumable electrodewelding with filler wire of SvAMg6 grade (2)

38 11/2015

tion of such a process has not been organized.Such equipment has a limited range of functionsthat significantly narrows the technological ca-pabilities and application of this process.

In 2014—2015, PWI within a cooperation pro-ject with Chinese-Ukrainian E.O. Paton Weld-ing Institute developed a versatile complex ofequipment PLAZER PW-HYBRID TC forplasma-arc, combined and hybrid welding.Equipment manufacturer is Science-ProductionCenter «Plazer» (Figure 16). This equipmentcomplex allows realizing a broad range of plasmaand arc processes: performing PAW with fillerwire at DCSP and DCRP and variable polarityasymmetrical current, hybrid consumable-elec-trode PAW with annular anode of the plasmatronand with axial feed of electrode wire, combinedwelding by constricted arc and consumable elec-trode in «soft plasma» mode (soft plasma arcwelding), and automatic consumable and non-consumable electrode welding.

For realization of such a range of technologicalcapabilities, this equipment is made to have modu-lar design and includes the following main units:

• welding process control system, providingcontrol of power sources, filler wire feed mecha-nisms, independent cooling modules, versatilemulti-position manipulator for performance ofvarious welding processes;

• inverter power source for welding plasmatron;• inverter power source for consumable elec-

trode;• module for coordinating the operation of

plasma welding and consumable electrode weld-ing power sources in «Plasma + MIG» mode (hy-brid and combined consumable-electrode PAWmodes);

• plasmatron module for machine hybrid PAW«Plasma-MIG» with axial wire feed with cablehosepack;

• monoblock for combined «Plasma + MIG»welding with cable hosepack;

Figure 14. Schematic of the process of hybrid consumable-electrode PAW with annular hollow anode and axial feed ofelectrode wire (a) and PLM 500 welding head of TBI for this method of welding (b)

Figure 15. Schematic (a) and welding torch for Super-MIG process (b) [41]

11/2015 39

• filler wire feed mechanism for hybrid PAW«Plasma-MIG»;

• electrode wire feed mechanism for hybridconsumable-electrode PAW «Plasma + MIG»;

• independent cooling module for plasmatronand consumable electrode torch;

• multiposition welding manipulator for hy-brid plasma welding in the downhand position,on vertical plane (including performance of ver-tical and horizontal welds on vertical plane), aswell as for welding cylindrical parts.

In addition, the functions of this equipmentprovide the possibility of operation in a complexwith welding robot instead of welding manipu-lator. For this purpose an interface is envisagedin the machine control system and in power

sources for connection to a robot with communi-cation protocols applied for the main types ofadvanced welding robots.

This equipment complex PLAZER PW-HY-BRID TC (see Figure 16) allows making weldsin the downhand position, vertical and horizontalwelds on vertical and inclined planes. Availabil-ity of rotating welding positioner in the systemdesign allows performance of circumferentialwelds by the above techniques. The complex isfitted with plasmatrons for nonconsumable andconsumable electrode welding, developed byPWI (Figure 17).

In conclusion the following can be noted.PAW of aluminium alloys was developing as acontinuation of nonconsumable-electrode argon-arc welding with sinusoidal AC. The first appa-ratuses were based on power sources for noncon-sumable electrode welding with alternating cur-rent. However, instability of running of sinusoi-dal AC arc, when passing through zero line, andconsiderable heating of plasmatron electrode as-sembly limited the possibility of wide applicationof PAW at sinusoidal AC of commercial fre-quency, both in FSU territory and abroad.

The next stage of development of PAW ofaluminium alloys is application of DCRP forpowering the constricted arc. This process hasseveral advantages: equipment design simplicity,good cleaning of the surface of aluminium alloysand weld pool from oxide film inclusions. Largewidth of welds, higher susceptibility to hydrogenporosity formation, considerable overheating ofplasmatron electrode assembly, leading to in-crease of plasmatron overall dimensions, shouldbe regarded as disadvantages of this process, aswell as greater wandering of DCRP constrictedarc at small current amplitudes. At this momentthis welding process is used occasionally, bothfor production and for research purposes.

Development of the next generation of powersources with rectangular current waveform forPAW allowed regulation in a broad range ofstraight and reverse polarity current amplitude,duration of straight and reverse polarity currentflowing, as well as frequency of variable polaritycurrent. Welding current form close to a rectan-gular one, simplified the processes of nonconsu-mable electrode arc excitation during reverse po-larity half-wave at its transition through zero.Appearance of power sources with the possibilityof creation of amplitude and time asymmetry ofwelding current provided a new impetus for de-velopment of PAW of aluminium alloys. In par-ticular, VPPA welding process was the main onein manufacture in 1980—1990s of cryogenic fuel

Figure 16. Appearance of versatile equipment complexPLAZER PW-HYBRID TC for plasma-arc, combined andhybrid welding in different positions

Figure 17. Plasmatrons developed by PWI for nonconsu-mable-electrode (a) and hybrid consumable-electrode (b)PAW

40 11/2015

tanks from high-strength aluminium—lithium al-loys for Space Shuttle system.

Simplification of the design of equipment forPAW by variable polarity asymmetrical current,as well as its cost reduction, is a favorable factorfor introduction of this method not only into aero-space industry, but also into other mechanical en-gineering industries. Transition from specializedsystems for PAW to modular design of the complexof equipment for PAW based on batch-producedsources for nonconsumable electrode welding alsohas a positive effect. Modular design in develop-ment of the equipment complex allows more preciseadjustment for production engineering tasks.

Further promising direction of developmentof PAW of aluminium alloys is combined andhybrid application of two and more heat sourcesat welded joint formation. This will allow in-creasing the welding speed, reducing electrodemetal spatter, and lowering the level of weldedstructure distortion. One of the promising fieldsis hybrid consumable-electrode PAW with plas-matron hollow annular anode and axial feed ofelectrode wire.

1. Gage, R.M. Arc torch and process. Pat. 2806124 US.Publ. 10.09.1957.

2. Lathi, K., Jenstroem, P. (1999) Plasma welding alu-minium. Svetsaren, 3, 26—28.

3. Dzelnitzki, D. (2000) Aluplasmaschweissen: Gleich-oder Wechselstrom? Technica (Suisse), 49(10), 44—53.

4. Dudko, D.A., Lakiza, S.P., Vinogradsky, F.M. et al.(1966) Alternating current constricted arc welding.Avtomatich. Svarka, 7, 47—49.

5. Dudko, D.A., Kornienko, A.N. (1967) Thermal effec-tiveness of alternating current plasma arc weldingprocess. Ibid., 11, 27—30.

6. Rubinchik, Yu.L. (1974) Mechanized welding ofhull structures from aluminium alloys. Leningrad:Sudostroenie.

7. Cooper, G., Palermo, J., Browning, J.A. (1965) Re-cent developments in plasma welding. Welding J.,44(4), 268—276.

8. Bykhovsky, D.G., Belyaev, V.M. (1971) Specifics ofweld formation in plasma (constricted) arc welding ofreverse polarity. Svarochn. Proizvodstvo, 9, 25—26.

9. Bykhovsky, D.G., Belyaev, V.M. (1971) Power char-acteristics of plasma arc at reverse polarity. Avto-matich. Svarka, 5, 27—30.

10. Nekrasov, S.A., Salkin, G.P., Bychkov, A.S. et al.(1976) Application of plasma arc welding in produc-tion of cryogenic equipment from aluminium alloys.Svarochn. Proizvodstvo, 4, 16—17.

11. Gorbach, V.D., Bochkarev, V.P., Nazaruk, V.K.(2009) Technology of plasma welding of aluminiumalloys. Mir Svarki, 3, 22—25.

12. Kiselev, G.S. (2010) Development of plasma weldingtechnology of AMg5 aluminium alloy with pulsedfeed of plasma-forming gases: Syn. of Thesis forCand. of Techn. Sci. Degree. Tula.

13. Tatarinov, E.A., Kiselev, G.S. (2009) To calculationof volt-ampere characteristic of plasma welding withpulsed feed of argon or helium. Svarka i Diag-nostika, 5, 11—14.

14. Bychkovsky, S.L., Novikov, O.M., Radko, E.P. etal. Method of plasma arc welding. Pat. 2351445Russia. Appl. 2007121870/02, 14.06.2007.

15. Ruge, J., Lutze, P., Norenberg, K. (1989) Eignung vonAluminiumdruckguss zum Plasma- und Elektronen-strahlschweissen – Entgasungmechanismen undNahtgute. Schweissen und Schneiden, 43(7), 327—332.

16. Schweinkhart, G. Plasma welding torch. Pat.6215089B1 US. Pat. 10.04.2001.

17. Paton, B.E., Dudko, D.A., Gvozdetsky, V.S. et al.Method of plasma welding. USSR Author’s cert.221477. Int. Cl. B 23K9/16. Appl. 1164345/25—27,17.06.67.

18. Cert. 73735584 of VPPA trade mark registration.http://www.tmfile.com/mark/?q=737355845. Lat-est appeal 16.07.2015.

19. Micheli, J., Pilcher, C. (2000) Advanced variable po-larity plasma arc welding. The Fabricator, November.

20. Nunes, A.C., Bayless, O.E., Jones, C.S. (1983) Thevariable polarity plasma arc welding process: Its ap-plication to the Space Shuttle External Tank: 1st re-port, June. Marshall Space Flight Center.

21. Clower, F.R. (1980) Welding of the external tank ofthe Space Shuttle. Welding J., 54(8), 17—26.

22. Tomsis, M., Barhorst, S. (1984) Keyhole plasma arcwelding of aluminium with variable polarity power.Ibid., 63(3), 25—32.

23. (1998) Lockheed Martin Michoud Space Systems.Weld. Des. and Fabr., 71(11), 32.

24. Brik, E.Yu. (1989) Alternating current plasma arcwelding of aluminium alloys. In: Abstr. of 1st Int.Conf. of Junior Sci. in Welding and Related Tech-nologies (Kiev, 16—20 May 1989), 91—92.

25. Zaparovany, A.G., Ignatchenko, G.N., Yarinich,L.M. et al. (1989) Power supply of pilot arc for vari-able polarity current plasma welding of aluminium.Avtomatich. Svarka, 9, 73—74.

26. Labur, T.M., Grinyuk, A.A., Poklyatsky, A.G.(2006) Mechanical properties of plasma welded jointson aluminium-lithium alloys. The Paton Welding J.,6, 32—34.

27. Labur, T.M., Grinyuk, A.A., Taranova, T.G. et al.(2007) Features of micromechanism of fracture injoints of aluminium-lithium alloys produced byplasma welding. Ibid., 9, 11—16.

28. Krivtsun, I.V., Shelyagin, V.D., Khaskin, V.Yu. etal. (2007) Hybrid laser-plasma welding of aluminiumalloys. Ibid., 5, 36—40.

29. Ternovoj, E.G., Shulym, V.F., Khaskin, V.Yu. et al.(2007) Properties and structure of hybrid laser-plasmawelded joints in aluminium alloys. Ibid., 11, 10—15.

30. Wesling, V., Schraem, A., Bock, A. et al. (2004)Plasmapulververbindungsschweissen von Aluminum-blechen. Praktiker, 10, 288—294.

31. (2005) Untersuchung der metallurgischen Grund-lagen zum Plasma-Pulver-Verbindungsschweissenduenner Aluminiumbleche. Schlussbericht fuer denZeitraum: 1.6.2003 bis 31.5.2005. AiF Vorhaben13.770 B/4. Technische Universitaet Ilmenau.

32. Zhang, Y.M., Zhang, S.B. (1998) Double side arcwelding increases weld joint penetration. WeldingJ., 6, 57—61.

33. Moulton, J.A., Weckman, D.C. (2010) Double sidearc welding of 5182-O aluminum sheet for tailoredwelded blank applications. Ibid., 1, 11—23.

34. Method of plasma-MIG welding. Pat. 3809824 U.S.Pat. 24.06.75. U.S. Philips Corporation.

35. Matthes, K.-J., Kusch, M. (2000) Plasma-MIG-Schweissen. Praktiker, 5, 182—188.

36. Shchitsyn, Yu.D., Tytkin, Yu.M. (1986) Consumableelectrode plasma welding of aluminium alloys.Svarochn. Proizvodstvo, 5, 1—2.

37. Shchitsyn, Yu.D., Shchitsyn, V.Yu., Herold, H. etal. (2003) Plasma welding of aluminium alloys.Ibid., 5, 36—42.

38. Makarenko, N.A., Nevidomsky, V.A. (2003) Thermalcycles in plasma-MIG surfacing. The Paton WeldingJ., 1, 43—45.

39. Yan, B., Hong-Ming, G., Ling, Q. (2010) Droplettransition for plasma-MIG welding on aluminium al-loys. Transact. of Nonferrous Met. Soc. China, 20,2234—2239.

40. Tiago Vieira da Cuhna, Jair Carlos Dutra (2007)Processo plasma-MIG contribuicao do arco plasma nacapacidade de fusao do arame. Soldagem Insp. SaoPaulo, 12(2), 89—96.

41. (2007) Hybrid welding: An alternative to SAW.Welding J., 10, 42—45.

Received 21.07.2015

11/2015 41

APPLICATION OF N—O—C—H GAS SYSTEMSFOR SYNTHESIS OF STRENGTHENING COMPONENTS

IN PLASMA COATINGS

V.N. PASHCHENKONTUU «Kiev Polytechnic Institute»

37 Pobeda Ave., 03056, Kiev, Ukraine. E-mail: vn.paschenko@ukr.net

Possibility of strengthening component synthesis during plasma spraying with application of plasma-chemicalreactor is considered. Thermodynamic analysis of systems based on bottled gases N2, CO2, C3H8—C4H10,CH4, as well as air (N2 and O2 mixture) and a number of powder materials was performed. Iron-basedmaterials and powders with sufficiently high titanium content were selected as model ones. Fundamentalpossibility of producing carbides, nitrides and oxides in the condensed state is demonstrated, ranges ofthermodynamic parameters, in which they exist, dependence of synthesized compound content in the systemon process temperature, pressure and quantity of initial solid product were established. It is found thatorganization of spraying process with concurrent synthesis of strengthening components is impossible withoutavailability of objective data on energy capabilities of plasma equipment, consisting of plasmatron—plasma-chemical reactor complex. A series of experiments were performed on plasma coating spray-deposition andtesting them for abrasive wear. 8 Ref., 6 Figures.

Keywo r d s : complex gas systems, plasmatrons,plasma-chemical reactor, plasma spraying, strengtheningcomponent synthesis

Rising requirements to service properties of partsand structures necessitate application of morecomplex compositions of materials, from whichthe functional surface layers are formed. Tradi-tionally, initial material chemical composition ismade more complicated by introduction of alloy-ing elements or composite material application[1, 2]. Combination of plasma spraying with pur-pose-oriented chemical transformations can be analternative [3].

Application of complex plasma-forming mix-tures provides a potential possibility for runningof reactions of interaction of plasma active gascomponents with the initial material that mayresult in synthesis of strengthening componentsin the produced coating [4, 5]. For instance, in[6] ultrafine particles of titanium nitride werefound in splats produced with plasma-chemicalreactor application under the conditions of su-personic spraying in nitrogen atmosphere. A num-ber of powders were selected, after analyzing thepossible variants of formation of strengtheningcomponents, which can be synthesized in plasma-forming media of N—C—H—O, N—C—H or N—Osystem, and taking into account the condition ofavailability of initial materials, on the base ofwhich synthesis will be performed. These mate-rials are batch produced and are available in the

market of Ukraine [7]. They can be conditionallysubdivided into four main groups:

• iron-based materials (PZh R3, self-fluxingiron-based), in which iron carbides and, par-tially, silicon and chromium carbides (in the caseof presence of these elements in the initial ma-terial) can form; alternatively, the initial mate-rial can be strengthened by synthesized iron ox-ides;

• nickel-based materials (PG-10N-03,PKh40N60, PG-SR3, PG-SR4, PG-19N-01),which have sufficiently high content of carbide-forming components – chromium, silicon, boronin their composition;

• materials, containing considerable amountof titanium (PT65Yu35, PN55T45), and in whichnitride and carbide synthesis is possible;

• aluminium-based materials (PAD, ASD-T,PT65Yu35, PN70Yu30), in which strengtheningcomponents of aluminium oxide can form undercertain conditions.

Thermodynamic calculations were conductedwith application of TERRA software package.Systems, based on the above spraying materialsand N2, CO2, C3H8—C4H10, CH bottled gases, aswell as air (N2 and O2 mixtures), were studied.Fundamental possibility of producing carbides,nitrides and oxides in the condensed state wasestablished, as well as the range of thermody-namic parameters, in which they exist, and de-pendence of these compounds content in the sys-tem on process temperature, pressure and quan-tity of initial solid product.© V.N. PASHCHENKO, 2015

42 11/2015

Special attention was given to two materials,which had been selected as model ones: iron andnickel—titanium. We will assign Fe3C iron car-bide as the target product, which will be synthe-sized at spraying of iron-based powder. Fe3C syn-thesis is possible in the presence of a sufficientamount of carbon in the system and preventionof its binding by plasma medium oxygen. Amongthe possible variants of reaction medium compo-sition, which were studied, the compositionformed by plasma-forming nitrogen and propaneturned out to be the most effective. Propane wasfed into pre-formed gas-powder flow within theplasma-chemical reactor.

Figures 1 and 2 show the dependencies of theyield of iron carbide condensed phase on processtemperature and pressure in the reaction volume.Presented dependencies reveal that the tempera-ture range of target product (Fe3C condensedphase) yield depends on the content of hydrocar-bon gas in the reaction medium. Within this tem-perature range, the level of target product yieldpractically does not change.

Temperature range is maximum wide in thecase of hydrocarbon gas content of 19—25 wt.%and becomes narrower in the case of going beyondthis interval with appearance of explicit maxi-mum of product yield at 1800—1900 K. Exceeding2500 K temperature level markedly reduces theproduct yield. It is found that pressure in thezone of plasma-chemical reaction running prac-tically does not influence the target product yieldwithin the process working temperature range,and has a significant influence on temperaturerange boundaries. Figure 3 shows the depend-encies of target product yield on the quantity ofloaded initial material under the condition ofunchanged quantity of the gaseous phase, and oftarget product yield on the quantity of hydro-carbon gas, fed into the reaction zone.

Increase of hydrocarbon component contentfrom 8.6 to 12.4 wt.% increases the carbide yieldpractically 2 times, but further increase of hy-

drocarbon gas quantity in the system practicallydoes not lead to increase of synthesis process ef-ficiency.

Increase of the quantity of loaded initial ma-terial, proceeding from calculation results, re-duces the quantity of synthesized product (pro-vided the values of other mode parameters remainunchanged).

Presence of a small (up to 45 wt.%) quantityof titanium in PN55T45 alloy composition createsthe prerequisites for possible synthesis of tita-

Figure 1. Dependence of target product (40.39 % Fe,35.38 % N2, 24.23 % C3H8) yield on process temperature:1 – pressure of 0.1; 2 – 0.3; 3 – 0.3; 4 – 0.4 MPa

Figure 3. Dependence of synthesized carbide quantity onpropane content in reaction medium (a) and quantity ofloaded dispersion material (b)

Figure 2. Dependence of target product (40.39 % Fe,35.38 % N2, 24.23 % C3H8) yield on pressure in reactionspace: 1 – T = 1000—2000; 2 – 2500; 3 – 3000 K

11/2015 43

nium nitride (TiN). High activity of titaniummakes higher requirements to reaction mediumcomposition. Preliminary estimates show the ra-tionality of application of plasma-forming nitro-gen, in order to create the conditions for titaniumnitride synthesis, with subsequent addition ofreaction nitrogen into the pre-formed gas-powderflow within the plasma-chemical reactor.

According to calculations, temperature rangeof target product (TiN condensed phase) yielddepends on the ratio of interacting solid and gase-ous phases, and in the studied range of this ratiochange it becomes narrower at lowering of gase-ous phase content. Increase of nitrogen contentfrom 75.77 to 86.22 wt.% widens the temperaturerange by 100 K. In this case, absolute quantityof synthesized nitride will rise.

Pressure increase in the reaction space widensthe temperature range of existence of titaniumnitride condensed phase. Thus, it is rational toconduct the process of heating of dispersednickel—titanium material within the plasma-chemical reactor at nitrogen content of 75—85 wt.%, in the temperature range of 2000—2400 K and about 0.1 MPa pressure in the reactionmedium. Under these conditions, required fixingof synthesized product is performed by feedinghardening nitrogen into the reactor end part.

Process of coating deposition with simultane-ous synthesis of strengthening components is im-plemented in a special spraying device, whichconsists of plasma sprayer and plasma-chemicalreactor, tightly and rigidly combined in one unit.

Initial material together with carrier gas isfed into plasmatron arc channel. A jet of high-temperature gas flows out through outlet elec-trode orifice into reaction space (within theplasma-chemical reactor), in which the dispersedmaterial is heated and accelerated.

During heating the material of initial particlesurface layer evaporates with temperature risewith formation of a vapour cloud around theliquid or ductile core. Additional gas is fed intoreaction space behind plasmatron nozzle edge. Asa result of interaction with it, synthesis of ap-propriate chemical compound begins in the va-pour phase surrounding the particles. Part of syn-thesized strengthening component can condenseon the surface of carrier-particles, and the otherpart is in the gaseous phase around the core.

At reaction zone outlet hardening gas is fedinto the unit, which, while lowering system tem-perature, creates the conditions for condensationof formed refractory compound on the surface ofcarrier-particles. Here, the particle proper re-mains in the liquid or ductile state, that promotes

the process of mixing of the particle liquid surfacelayer with the condensed compound and coagu-lation of synthesized ultrafine particles withthose of initial material.

Strengthening component composition de-pends on chemical composition of initial mate-rial, plasma, carrier, reaction and hardeninggases.

Process of synthesis of coating strengtheningcomponents during heating and acceleration ofinitial material envisages application of plasmagenerators, capable of creating flows of low-tem-perature plasma from complex, reactive gaseousplasma-forming mixtures in a broad range of vari-ation of energy parameters, and, in particular,of specific enthalpy. Plasma flow componentsshould take an active part in synthesis, or, atleast, not hinder its running.

Successful practical implementation of theprocess is possible in the case of availability ofsufficiently complete data on energy charac-teristics of the used plasmatron and their inter-relation with plasma generation mode parame-ters. Comprehensive studies of the above char-acteristics of plasmatrons, generating plasma ofN—O—C—H system [8], allowed application ofexactly this gas system for producing strength-ened plasma coatings. In the specific case,straight polarity two-electrode plasmatron withstepped anode of up to 35 kW total power wasused, which is capable of stable operation in ni-trogen, air and mixtures of air with hydrocarbongases (methane, propane, butane).

Proceeding from the results of these studies,criterial dependence of specific enthalpy on modeand geometrical parameters of plasma sprayer op-eration was established. This dependence is thebasis for preliminary calculation of mode parame-ters of the process, at which strengthening com-ponents synthesis is possible with maximum yieldof synthesized product during coating deposition.

In the simplified variant (for the applied de-sign) in the case of air application, the criterialdependence has the following form:

ε = 3.984I0.648Qair—0.33d1

0.092d2—0.035 p0.357 ,

(1)

where I is the arc current, A; Q is the plasmagas flow rate, m3/h; d1 and d2 are the diametersof the narrow and wide part of outlet electrodearc channel, mm; p is the pressure at plasmatroninlet, N/m2.

For nitrogen:

ε = 3.684I0.66QN—0.318 d1

0.092d2—0.058p0.357.

(2)

44 11/2015

For a mixture of hydrocarbon gases with air(without combustion within the plasma-chemicalreactor):

ε = 0.273I0.6QΣ—0.214d1

—0.297 ×

× d20.303p0.633(l/l1)

—0.034n0.099 .(3)

For a mixture of hydrocarbon gases with air(with running of the process of mixture compo-nent combustion within the plasma-chemical re-actor):

ε = [0.273 I0.6QΣ—0.214 d1

—0.297 ×

× d20.303p0.633(l/l1)

—0.034n0.099] +

+ [0.0895 aQ a.gI—0.132QΣ

—0.69d1—0.119 ×

× d2—0.059p0.241 (l/l1)

0.079n0.064].

(4)

Plasma-chemical reactor is a system of indi-vidual sections sequentially and tightly con-nected to each other. Two kinds of sections wereused: sections for feeding reaction gases and feed-through sections, the purpose of which is creationof a certain reaction volume. Each feed-throughsection has individual water cooling. Sections forgas feeding are cooled by those gases, which arefed through them. Geometrical dimensions andconfiguration of all the sections (except for thefirst one) are unified, that allows changing thereactor overall length by increasing the numberof feed-through sections. Accordingly, locationof sections for gas feeding within the reactor canbe also changed, depending on process require-ments.

Longitudinal geometrical size of reactor innerspace (time of material particles staying in thereaction zone) is determined by number of sec-tions, used in the structure.

Reaction zone diameter was selected constant,proceeding from preliminary information aboutthe angle of opening of the heterogeneous flowduring spraying with the plasmatron of the useddesign (proceeding from the condition of elimi-nation of dispersed material deposition on thereactor inner wall during the process). Overallschematic of the technological process of sprayingpractically does not differ from the traditionalprocess of plasma spraying, although its individ-ual stages have certain specific behavioural fea-tures.

Typical preparation of initial material usuallyenvisages powder drying and its sieving to removepossible contamination and separation of frac-tions of a certain range, suitable for forming thecoating. The proposed process envisages applica-tion of essentially narrower range of possible di-mensions of individual particles. Range narrow-

ing improves the stability and predictability ofthe process of strengthening component synthe-sis, as it runs in the vapour cloud around theparticles. Reduction of initial powder particlesize can lead to complete evaporation of the par-ticle, and its increase – to its insufficient heatingand slowing down of the process of surface layerevaporation.

Requirements to temperature mode of the basebecome more stringent. Thermal insulation of theplasma jet from the environment by reactor wallspromotes preservation of high level of heat flowsto the base at spraying distances. This leads tobase overheating, acceleration of the processes ofoxide film growing on item surface and, conse-quently, to lowering of the strength of coatingadhesion with the base. In the case of applicationof gas mixture hydrocarbon components, thefragments of which burn down at reactor outletduring sucking of oxygen from the air, additionalmeasures are required for more intensive coolingof the item as natural cooling becomes insuffi-cient. Here, higher requirements are made to sta-bility of parameters of plasma sprayer and initialmaterial feed system. Random change of carriergas flow rate and dispersed material quantitydisturbs the spatial arrangement of gas-powderflow within the plasma-chemical reactor, leadingto powder particle deposition on reactor walls,and destabilizes the process of plasma-chemicalsynthesis.

The synthesis process is also sensitive tochange of working medium energy level and itcan be disturbed as a result of instability ofsprayer input parameters and electrode fracturebecause of erosion. The required nominal modeparameters of heterogeneous flow generation (arccurrent, plasma gas flow rate and its chemicalcomposition, gas pressure, arc channel geometri-cal dimensions, reaction and hardening gas flowrates) are calculated beforehand.

The main thermodynamic parameters, deter-mining the probability of running of plasma-chemical reaction of strengthening componentsynthesis (in the condensed form), are pressureand weight-average temperature in the reactionspace. Temperature, in its turn, is the derivativeof the amount of energy, applied to a unit ofvolume (mass) of reaction medium (specific en-thalpy).

Schematic of preliminary determination ofmode parameters for conducting the process canbe as follows:

• proceeding from analysis of chemical com-position of spraying material and assignedstrengthening component, element composition

11/2015 45

of gas system, in which the process is to be im-plemented, is determined;

• element composition of gas system is pro-vided with application of the respective plasmagas, which contains the chemical elements fromthe required list, and reaction gas, which withits chemical composition complements the systemas to chemical element list and their content;

• thermodynamic calculations of created sys-tems (allowing for solid phase presence) are con-

ducted in a broad range of thermodynamic pa-rameters variation – process pressure and tem-perature;

• complete range of the above parameter val-ues, in which formation of target product is pos-sible, and values, at which product yield will bemaximum, are established;

• required value of considered gas system spe-cific enthalpy is determined (by dependenciessimilar to those given in Figure 4), which pro-vides the required process temperature level (incase of ensuring appropriate pressure in reactionspace);

• equations of (1)—(4) type are used to deter-mine the required values of mode and geometricalcharacteristics of the process.

The process of coating spray-deposition beginsfrom starting the plasma sprayer with completeplasma gas mixture or with its main component,at arc current, usually, below the nominal value.After the plasma sprayer has reached the workingmode in terms of current, composition and flowrate of plasma gas, feeding of dispersed materialinto the reaction space begins.

Appearance of heated particle flow at the out-let of plasmatron—reactor system is the signal tostart feeding reaction and hardening gases intothe reactor. Here, the dimensions and shape ofgas flow at reactor outlet change visually (Fi-gure 5).

Functional properties of coatings from ironpowder, which at this moment is the cheapestinitial material for plasma spraying, can be es-sentially improved, by adding to coating compo-sition iron carbides or its oxides synthesized dur-ing spraying. Selection of reaction mediumchemical composition is performed, dependingon chemical composition of target product. Ther-modynamic calculations confirm the theoreticalpossibility of carbide synthesis during sprayingin a mixture of carbon-containing gases.

Figure 5. Process of coating spray-deposition

Figure 4. Dependence of plasma jet weight-averaged tem-perature on specific energy input into plasma gas: 1 – Ar;2 – Ar + 15 % H2; 3 – Ar + 25 % H2; 4 – Ar + 50 % H2;5 – N2; 6 – air; 7 – air + CH2, α = 1; 8 – air + CH4,α = 0.4; 9 – air + CH4, α = 0.8; 10 – air + CH4, α = 0.6;11 – CO2; 12 – products of carbon dioxide conversion ofnatural gas; 13 – NH3; 14 – products of steam conversionof natural gas; 15 – H2O

Figure 6. Microstructure (×1000) of coating, spray-depos-ited with PZh R3 material in a complex gas system usingplasma-chemical reactor

46 11/2015

Similar calculations show the theoretical pos-sibility of oxide synthesis in the case of the changeof spraying conditions.

In order to produce coating with synthesizediron carbide in iron matrix, the spraying processwas conducted in nitrogen plasma with feedingof carbide-forming propane-butane into the reac-tion space.

Plasma gas flow rate was equal to 3.5 m3/h,reaction gas flow rate was 0.7 and 1.5 m3/h, arccurrent was 150 A. At these parameters voltagedid not exceed 105—110 V. Coating materialspraying efficiency was 5 kg/h, main size of dis-persed material fraction was 63—100 μm.

Samples for spraying were mounted practi-cally at the edge of plasma reactor last section.Spraying distance was equal to 230—250 mm, thatis essentially greater than the optimum value inspraying by the traditional schematic. Increaseof the distance during plasma-chemical sprayingpractically does not influence the value of par-ticle velocity at the moment of colliding withthe base surface and greatly increases particletemperature, as a result of limiting the air suctionfrom the environment.

Figure 6 shows the microstructure of the pro-duced coating. Coating has a characteristic lami-nated structure, which forms from molten parti-cles. Presence of a certain quantity of particlesof a globular shape, which have passed into thecoating from unmolten state, is observed at thesame time. All kinds of boundaries and pores arevisible.

In addition to compounds, which were syn-thesized in the gas phase during gas medium in-teraction with material vapour phase and whichcoagulate with carrier-particles, intensive gas ab-sorption by molten particle material occurs dur-ing spraying. Under the conditions of high cool-ing rates (about 105 K/s) gases dissolved in par-ticles are fixed in the solid solution with forma-tion of oversaturated solid solutions. Relaxationof metastable state leads to precipitation of ul-trafine, nanosized excess phases in the coolingparticles.

Quantitative microanalysis of sample mi-crosection reveals considerable increase of carboncontent: from 0.05 % in the initial material upto 5.0—5.6 % in the coating. Samples with iden-tical coating were studied in parallel under theconditions of abrasive wear in LKI-3 instrument.The load was 150 N; fused corundum with mainparticle size of 300—400 μm was used as abrasive.

It is found that resistance of a sample withcoating, containing a strengthening component,under identical test conditions is 7—7.5 timeshigher than that of a sample, sprayed withoutrunning of plasma-chemical reaction (mass losswas equal to 0.02583 g/m compared to0.00343 g/m in a sample with strengtheningcomponent).

Conclusions

1. Use of plasma-forming media of N—O—C—Hsystem creates prerequisites for synthesis ofstrengthening components (carbides, nitridesand oxides) during plasma spraying of protectivecoatings. Thermodynamic analysis of the respec-tive systems allows evaluation of quantitativecharacteristics of target product yield and energyconditions of the process.

2. Practical realization of the technology ofspray-deposition of coatings with concurrent syn-thesis of strengthening components is possible incase of availability of objective data on energycharacteristics of applied plasma equipment andregularities of these characteristics connection toprocess mode parameters.

3. Results of studying coatings from PZh R3powder reveal an abrupt (by 2 orders of magni-tude) increase of carbon content in the coating,compared to initial material, and respective wearresistance increase (by more than 7 times) underconditions of abrasive wear.

1. Smirnov, I.V., Selivestrov, I.A., Chyorny, A.V. et al.(2011) Improvement of wear resistance of plasmacoatings based on composite powder with SiO2nanoparticles. Visnyk NTU «KhPI», 2, 70—74.

2. Borisov, Yu.S., Borisova, A.L., Kolomytsev, M.V. etal. (2015) Supersonic plasma gas air spraying of cer-met coatings of the (Ti, Cr)C—NiCr system. The Pa-ton Welding J., 2, 19—25.

3. Ronald, W.S., Emerging, U.S. (1993) Thermalplasma materials processing research and develop-ment. In: World Progress in Plasma ApplicationsConf. Proc. (Palo Alto, Febr. 9—11, 1993), 1—16.

4. Pashchenko, V.M., Kuznetsov, V.D. (2002) Plasmaheat sources in surface engineering processes.Problemy Tekhniky, 3, 26—32.

5. Borisov, Yu.S., Petrov, S.V. (1993) Application ofsupersonic jets in thermal spraying technology. Avto-matich. Svarka, 1, 24—34.

6. Petrov, S.V. (1997) Plasma synthesis in thermalspraying technology. Poroshk. Metallurgiya, 9/10,34—38.

7. Borisov, Yu.S., Kharlamov, Yu.A., Sidorenko, S.L.et al. (1987) Thermal coatings from powder materi-als: Refer. Book. Kiev: Naukova Dumka.

8. Pashchenko, V.N. (2009) Effect of the compositionof plasma air-gas mixture on parameters of the plas-matron jet. The Paton Welding J., 4, 27—31.

Received 28.09.2015

11/2015 47

EVALUATION OF THERMAL STRESSED STATEIN WELDED JOINT OF ALLOY Inconel 690

N.O. CHERVYAKOVE.O. Paton Electric Welding Institute, NASU

11 Bozhenko Str., 03680, Kiev, Ukraine. E-mail: office@paton.kiev.ua

The results of evaluation of thermal stressed state in welded joint of nickel alloy Inconel 690 are presented.The experimental-calculation analysis of thermal processes and mathematical modeling of stress-strain statein single-pass welding was performed. The verification of the data by comparing the experimental thermalcycles with the calculated ones showed a good similarity of results of both as to the sizes of a weld as wellas to the thermal cycles in the zones at the different distance from the fusion line. The kinetics of stress-strainstate in different zones of welded joint was analyzed considering the probability of hot crack formation.It is shown that under the conditions of modeling linear one-pass welding of plates of 3 mm thickness withenergy input of 304 J/mm the arising stresses and deformations do not exceed the critical values and, atthe same time do not create the conditions for hot crack formation. The applied mathematical model canbe used in evaluation of kinetics of stress-strain state for different variants of welding (change in energyinput, different filler material, etc.). 7 Ref., 1 Table, 5 Figures.

Keywo r d s : nonconsumable-electrode arc welding,nickel alloys, thermal processes, thermal cycles, mathe-matical modeling, stresses, deformations

The nickel-based alloys play an important roleas corrosion-resistant materials in the nuclear andchemical industries, especially under conditionsof operation at the high temperatures, consider-able stresses and aggressive environments. Thewidely used nickel-based alloys at the presenttime are alloys of Ni—Cr—Fe alloying system. Oneof such alloys is Inconel 690 with high chromiumcontent (30 %) [1]. The welded joints of In-conel 690 with appropriate filler materialsshowed a high sensitivity to hot crack formation,particularly under the conditions of welding ofthick-walled structural elements. The alloysmanifest the highest tendency to the formationof ductility-dip cracks [2, 3].

As was shown previously [4], the ductility-dipcracks in welding of alloys of alloying systemNi—Cr—Fe occur at the temperature range of1050—650 °C, the value of critical deformationof crack formation εcr is about 1.2 %.

The thermal processes in fusion welding havea decisive influence on the nature and kineticsof changes in the stress-strain state, structure andproperties of weld and HAZ. In this regard, theanalysis of temperature fields in welding of alloysof Ni—Cr—Fe alloying system with the aim offurther investigations of kinetics of stress-strainstate and evaluation of tendency to hot crackformation is an urgent problem.

The aim of the work was a comparative ex-perimental and calculation analysis of tempera-

ture fields and modeling of stress-strain state inone-pass argon-arc welding of alloy Inconel 690.

The computer prediction and calculation al-low determining the temperature fields for dif-ferent values of thermal power of the arc. Usingthe calculation procedure the efficiency coeffi-cient of the arc is established, which varies de-pending on the method and conditions of weld-ing, etc. To specify the input data of mathemati-cal model the experimental plotting of thermalcycles and their comparison with the calculatedvalues is required.

The recording of welding thermal cycles wascarried out on the Inconel 690 specimens of 170 ×× 40 × 3 mm size. TIG welding in argon withoutfillers was carried out at the following conditions:Iw = 72 A, Uw = 10.5 V, vw = 7.2 m/h. The gasconsumption in welding amounted to 5 l/min.

For measurements of thermal cycles of weld-ing in the HAZ the thermocouples of type K(chromel-alumel) were used. The diameter ofthermoelectrodes was 0.2 mm, the diameter ofhot junction was smaller than 0.5 mm. The junc-tion of thermocouple was produced by fusion ofthermoelectrodes using manual argon arc weld-ing. The calibration of thermocouples was carriedout according to the melting temperatures of com-mercially pure aluminum and copper.

The record of thermal cycles was carried outin real time using ADC Expert 9018R and spe-cialized software PowerGraph 3.0 allwoing theautomatic convertion of the values of thermalEMF to the temperature.

To the surface of plates the junctions of ther-mocouples were fixed by capacitor-type welding,that ensured a permanent stable contact of the

© N.O. CHERVYAKOV, 2015

48 11/2015

junction with the metal in a wide temperaturerange. The appearance of the specimen withwelded-in thermocouples after welding is shownin Figure 1.

After welding the weld and thermocouples inthe HAZ were photographed in binocular micro-scope with 25-fold magnification and the meas-urement of distances from the fusion line to thethermocouple junctions was carried out. In ad-dition, on the macrosections the weld width andpenetration depth were measured.

The calculated evaluation of thermal stressedstate was performed using the specialized soft-ware «Welding of nickel alloys», created on thebasis of algorithms developed at the E.O. PatonElectric Welding Institute, and describing thedistribution of temperatures and stress-strainstate near the weld pool of complex-alloyed al-loys based on nickel. For the considered plate,in the process of welding heating the temperaturefield T(x, y, z, t) in time was determined. Then,according to the average values of temperatureT(x, y, t) the problem of kinetics of the stressesand deformations was solved [5, 6]. It was basedon the method of successive tracking in time with

pitch Δt and the finite element method in space,i.e. the considered area was represented by a to-tality of elementary volumes of hx × hy × hz sizes.The size of a single element in the model amountsto 0.5 × 0.5 × 0.5 mm.

For evaluation of the regularities of tempera-ture distribution in the weld and HAZ jointsusing calculation method the temperature de-pendencies of change in heat capacity and ther-mal conductivity of the investigated materialwere used. These values for alloy Inconel 690were chosen according to the data of literaturesources, and also considering the own investiga-tions [1, 7].

The arc efficiency factor η, used in the calcu-lations, for TIG process amounts to 0.50—0.75.The temperature fields at the values of this factorof 0.5, 0.6 and 0.7 were calculated. Figure 2shows that there is a significant influence of arcefficiency factor both on the heat distribution,as well on the shapes of isotherms. In modelingη = 0.70—0.75 was selected, as far as the appro-priate temperature field correlates well with theexperimental results.

The first stage of modeling was the determi-nation of temperature fields, taking into account

Figure 1. General view of specimen with weld and thermo-couples after welding

Figure 2. Influence of values of arc efficiency factor η ontemperature distribution in the high-temperature region

Temperature dependence of thermophysical properties of alloy Inconel 690

Temperature, °C Young’s modulus, MPa Yield strength, MPaCoefficient of thermal

expansion, 1/degThermal conductivity,

W/(cm⋅deg)Heat capacity,

J/(g⋅deg)

20 205,000 348 0.00001406 0.115 0.453

100 204,000 281 0.00001406 0.130 0.474

200 193,000 248 0.00001431 0.155 0.496

300 187,000 240 0.00001453 0.173 0.523

400 185,000 231 0.00001480 0.190 0.557

500 177,000 225 0.00001519 0.212 0.574

600 167,000 215 0.00001570 0.227 0.608

700 158,000 200 0.00001618 0.250 0.633

800 152,000 190 0.00001660 0.269 0.660

900 145,000 175 0.00001701 0.285 0.690

1000 135,000 150 0.00001741 0.303 0.715

1100 125,000 125 0.00001779 0.318 0.739

1200 100,000 100 0.00001800 0.332 0.769

11/2015 49

the temperature-dependent properties of material(specific heat, thermal conductivity) and tempera-ture losses as a result of convection and radiation.

The temperatures of melting and solidificationwere taken as 1377 and 1343 °C, respectively [1].

Figure 3 shows the results of calculation of tem-perature distribution along the surface of sourcemovement at the steady state in the area of weldpool and HAZ. Due to symmetry of the solvedproblem a half of the plate is shown in this Figure.

The verification of the calculated tempera-ture fields with the thermal cycles and actualsizes of the weld, measured using thermocou-ples, were carried out. The comparison of thedata showed a good similarity of results bothas to the sizes of weld (weld width and pene-tration depth in the experiment were 4.2 and1.3 mm, respectively, in modeling they were4.1 and 1.4 mm) as well as to the thermal cycles

at the points at different distances from the fusionline. Thus, the validity of using the initial ther-mophysical characteristics for calculation of tem-perature distribution in welded joint in weldingof alloys of Ni—Cr—Fe alloying system was con-firmed.

The modeling of stress-strain state was per-formed taking into account the temperature de-pendence of physical and mechanical propertiesof the alloy, including modulus of elasticity,yield strength and coefficient of thermal expan-sion. All these characteristics were determinedexperimentally within the temperature range ofup to 1100 °C. The yield strength and modulusof elasticity were determined in installationMTS-810, the evaluation of linear expansion co-efficient was performed in the contact-free laserdilatometer. The data on thermal and mechanicalproperties are presented in the Table.

Figure 3. Distribution of temperature fields in the plane XY and across the plate thickness

Figure 4. Kinetics of change in temperature T, longitudinal σxx and transverse σyy stresses and longitudinal plasticdeformations εxx along the weld axis during welding

50 11/2015

Taking into account the calculated tempera-ture fields the modeling of stress-strain state dur-ing welding was performed. The distributions oftemporary and residual stresses and deformationsfor different zones of welded joint were obtained.

The investigations of weldability of alloys ofNi—Cr—Fe alloying system showed [4] that thecrack formation in the weld and HAZ is possibleat distance of 0.5—3.0 mm from the fusion line. Inthis regard, the kinetics of thermal stressed statewas considered in the weld and HAZ at distanceof 0.5 mm from the fusion line (Figures 4 and 5).

At the beginning of heating cycle, the metalis subjected to compressive longitudinal σxx andtransverse σyy stresses. The longitudinal plasticdeformations εxx are also compressive, their valuereaches 0.9 %. At the subsequent cooling, thetensile stresses of both longitudinal and trans-verse nature occur. The longitudinal stressesreach 215 MPa and are significantly larger thanthe transverse ones, which are equal to 40 MPa.The longitudinal stresses reach maximum valuesduring cooling to 650 °C. With the further cool-ing, the longitudinal stresses remain at a constantlevel, and the transverse ones are gradually de-creasing. The longitudinal plastic deformationsduring cooling have positive increments to 0.5 %.

Basing on these results it can be concludedthat for the given conditions of modeling of sin-gle-pass welding, the maximum increments ofplastic deformation coincide with the tempera-ture range of brittleness of the alloy and arepropagated simultaneously with increase in thelongitudinal tensile stresses, but they are not suf-ficient for the ductility-dip crack formation.

Conclusions

1. Using experimental and calculation procedurethe evaluation of temperature fields in single-passwelding of nickel alloy Inconel 690 was carriedout. The comparison of the data showed a good

similarity of results of both as to the sizes of theweld and also as to the thermal cycles at thepoints at different distance from the fusion line,that confirms the validity of using the initialthermophysical characteristics of metal for cal-culation of temperature distribution in thewelded joint.

2. Taking into account the experimentally de-termined physical and mechanical properties ofthe alloy, the mathematical modeling of stress-strain state was performed and the kinetics ofchanges in stresses and deformations was consid-ered for different zones of welded joint. It isshown that under the conditions of modeling oflinear single-pass welding of plates of 3 mm thick-ness at energy input of 304 J/mm the occuringstresses and deformations do not exceed the criti-cal values, and the conditions for hot crack for-mation are not created.

1. (2009) Inconel® alloy 690. URL: http://www.special-metals.com/documents/Inconel%20alloy%20690. pdf

2. Kiser, S.D., Zhang, R., Baker, B.A. (2009) A newwelding material for improved resistance to ductility dipcracking. In: Proc. of 8th Int. Conf. on Trends in Weld-ing Research (Pine-Mountain, GA, 2009), 639—644.

3. Nishimoto, K. (2006) Microcracking in multipassweld metal of alloy 690. Pt 1: Microcracking suscep-tibility in reheat weld metal. Sci. and Techn. ofWelding and Joining, 11(4), 455—461.

4. Yushchenko, K.A., Savchenko, V.S., Chervyakov,N.O. (2011) Comparative evaluation of sensitivity ofwelded joints on alloy Inconel 690 to hot cracking.The Paton Welding J., 11, 2—7.

5. Makhnenko, V.I. (1976) Computational methods forinvestigation of kinetics of welding stresses andstrains. Kiev: Naukova Dumka.

6. Yushchenko, K.A., Makhnenko, V.I., Savchenko, V.S.et al. (2007) Investigation of thermal-deformation stateof welded joints in stable-austenitic steels and nickel al-loys. Welding in the World, 51(9/10), 51—55.

7. Yushchenko, K.A., Savchenko, V.S., Zvyagintseva,A.V. et al. (2014) Physical-mechanical characteristics ofIn690 type weld in high-temperature range of ductilitydeep. In: Proc. of 55th Int. Conf. on Actual Problemsof Strength (Kharkov, Ukraine, 2014), 189.

Received 28.09.2015

Figure 5. Kinetics of change in temperature T, longitudinal σxx and transverse σyy stresses and longitudinal plasticdeformations εxx in HAZ at 0.5 mm distance from the fusion line during welding

11/2015 51

PHYSICAL AND MECHANICAL PROPERTIESOF TRANSITION ZONE OF BIMETAL

PRODUCED BY AUTONOMOUS VACUUM BRAZINGOF COPPER ON STEEL

M.G. ATROSHENKO, M.A. POLESHCHUK, A.V. SHEVTSOV, A.L. PUZRIN,D.D. MISHCHENKO, I.P. SEREBRYANIK and A.I. BORODIN

E.O. Paton Electric Welding Institute, NASU11 Bozhenko Str., 03680, Kiev, Ukraine. E-mail: office@paton.kiev.ua

The physical and mechanical properties of the joining zone of metal produced by autonomous vacuumbrazing of copper M1 on steel billets were determined. During tensile test the fracture of standard specimensoccurred always in the copper part. At the same time, the strength properties of brazed copper layer in theinitial state exceeded the reference values for wrought and annealed copper M1. The ultimate tear strengthof brazed copper layer determined on the special specimens is equal to the temporary rupture strength ofsteel 20. The investigations of physical characteristics of the joining zone showed that during technicalcalculations the additional electrical and thermal resistances of the contact zone can be neglected. 13 Ref.,1 Table, 7 Figures.

Keywo r d s : autonomous vacuum brazing steel-cop-per bimetal, mechanical properties of joining zone, elec-trical and thermal resistance of contact

During the production of separate units of ma-chines and devices, in some cases the bimetalcopper-steel billets are used manufactured be-forehand. The use of such billets allows produc-ing parts operating reliably under high electrical,thermal and mechanical loads. During their op-eration the physical and mechanical propertiesof metal in the zone of copper with steel joiningare essentially important. The transition steel—copper zone should have a minimum electricaland thermal resistance and a sufficiently highmechanical strength.

It is obvious that the specific values of thesevariables depend on the method of manufactureof copper-steel billet.

The bimetallic strips of up to 4 mm thicknessused for production of electric engineering powerequipment are manufactured mainly by joint roll-ing of steel and copper strips or multilayer pack-ets [1, 2]. The billets of other shapes with a largecross-section can be produced by diffusion weld-ing in vacuum [3], explosion welding [4], frictionstir welding [5], and also high-temperature braz-ing [6]. For manufacture of large-size billets suchas water-cooled hearth electrodes of arc steel-melting furnaces of direct current, sliding bear-ings of large diameters, etc., different methodsof fusion welding are applied [7], including elec-

tro-slag welding [8], pouring of molten copperon steel heated to 900 °C is used [9].

Recently, the method was developed to pro-duce large-size copper-steel billets by meltingcopper in autonomous vacuum. The method isbased on using autovacuum effect, which consistsin spontaneous cleaning the surfaces of steel bil-lets from oxide films at their heating to the tem-perature above 1000 °C under the conditions ofa closed volume. In this case, oxygen in the com-position of oxide film diffuses deep into the steelbillet. At the same time, the oxidation of itssurface by oxygen from the closed volume occurs.The diffusion process continues as long, unlessall the oxygen of the closed volume is consumedand oxide film disappears. The time of oxide filmremoval from the surface of steel billet can bereduced by preliminary pumping out of air fromthe closed volume [10].

For manufacture of bimetal product, a closedvolume is created on the billet steel part, thefuture interface of bimetal. The copper is placedinto it in the amount necessary to produce thecopper part of the billet. After sealing the volumewith copper the assembly is heated to 1130—1150 °C. At this temperature the cleaning of steelsurface from oxide film occurs resulting in for-mation of a strong metallic bonding with steelafter copper melting [11].

The aim of this work is the determination ofphysical and mechanical characteristics of the

© M.G. ATROSHENKO, M.A. POLESHCHUK, A.V. SHEVTSOV, A.L. PUZRIN, D.D. MISHCHENKO, I.P. SEREBRYANIK and A.I. BORODIN, 2015

52 11/2015

steel—copper transition zone, produced by melt-ing copper in autonomous vacuum, namely, ofelectrical and thermal resistance, mechanicalstrength, tensile and tear strength.

To carry out investigations the hollow cylin-ders with a massive bottom of steel 20 were manu-factured. The cavities of cylinders were tightlyfilled with cuttings of copper sheet M1 of 2.5 mmthickness. The opening of the cavity was closedby steel cover, to the center of which a thin-walled steel pipe was welded-in. The cover wassealed by argon arc welding (Figure 1).

The cylinders prepared for testing were placedin the thermal furnace with a usual atmosphere.The branch pipe-vacuum conductor was broughtout to the outside and connected to the fore-vacuum pump through the gate valve with a ma-novacuum gauge (Figure 2).

Before heating, the evacuation was createdinside the cylinders by the forevacuum pump andthe gate valve was closed. The thermal furnacewas heated to 1150 °C and after 30 min of iso-

thermal holding it was switched off and the doorwas opened. The cylinders were left inside thefurnace until their complete cooling. Duringheating the evacuation inside the cylinders ac-cording to the indications of the manovacuumgauge was almost unchanged.

After cooling each cylinder was cut along theforming part into the four equal parts, of whichthe specimens for mechanical tests were manu-factured, the measurements of electrical resis-tance and study of microstructure of the transi-tion zone were carried out.

Figure 3 shows a photo of macrosection of thelongitudinal template of cylindrical copper-steelbillet. The visual inspection of macrosection us-ing magnifying glass of 10-fold increase did notreveal defects, violating the integrity of metal inthe brazed copper and in the zone of its joiningwith steel.

Figure 1. Sketch of pilot billet for brazing: 1 – hollowcylinder; 2 – melting component (copper); 3 – cover;4 – branch pipe for vacuuming; 5 – sealing welds

Figure 2. Scheme of carrying out autonomous vacuum braz-ing: 1 – thermal furnace; 2 – pilot billet; 3 – vacuumconductor; 4 – manovacuum gauge; 5 – gate valve; 6 –forevacuum pump

Figure 3. Macrosection of steel-copper bimetal joint: 1 –copper; 2 – steel

11/2015 53

To carry out mechanical tensile tests the speci-mens with 6 mm diameter of rupture part and40 mm length were manufactured. The steel—cop-per contact was located in the middle of the rup-ture part. During tests the fracture of all thespecimens occurred along the copper part (Fi-gure 4). This means that the strength of the cop-per—steel contact is higher than the tensilestrength of copper.

The test results are given in the Table.The given data evidence that the values of

mechanical properties of the tested specimens arehigher than those of the deformed and annealedcopper according to the reference values [12].

The main characteristic of quality of bimetaljoint is the tear strength of the brazed layer. Toobtain the reliable data on tear strength the speci-mens of a special type should be tested, whichhave a ring-type groove on a stronger joint ele-ment (Figure 5).

Such a groove reduces the area of contact ofcopper with steel, at the same time preservingconstant diameter of the copper part of the speci-men [13]. The tests showed that the fractureoccurred along the line of joining. The averagevalue of ultimate tear strength was 480 MPa,that corresponds to the value of ultimate rupturestrength of annealed steel 20 [12].

Thus, the results of above-mentioned mechani-cal tests evidence that the method of autonomousvacuum brazing allows producing the copper-steel bimetal with high strength of layer joining.

The other important characteristics, whichprovide a reliable operation of bimetal at highelectrical and thermal loads, are the values ofelectric and thermal resistance in the steel—cop-per contact. The lower are these resistances, thebetter is heat removal through the interface andlower the electrical losses of current passingthrough it. In the transition zone the change ofcoefficients of electrical and thermal conductiv-ity occurs from values, inherent to steel, to thoseof copper. At the absence of a large number ofpores, cracks and other defects in it violatingthe density of metal, these changes occursmoothly. Therefore, in the transition zone thevalue of these coefficients may be taken as anaverage value between the coefficients of cop-per and steel parts. On this assumption thetotal electrical and thermal resistances of tran-sition zone are determined only by its width.The transition zone of steel-copper bimetaljoints, produced using different methods ofwelding and melting, is determined as the areaof propagating the solid solutions of copper intosteel and steel into copper on both sides of theinterface. These solid solutions have lower val-ues of electrical and thermal conductivity thanthe pure copper [7]. Thus, by measuring thethermal or electrical conductivity of the bi-metal specimen in the direction perpendicularto the joining line, the width of the transitionzone can be experimentally determined.

In metallic conductors a direct connection be-tween electrical and thermal resistance exists,determined by the Wiedemann—Franz law in itsmodern interpretation. Thus, knowing one ofthese values it is always possible to determinethe other one. Methodically it is much easier tomeasure the electrical resistance. Therefore, weconducted its evaluation using the procedure,proposed in work [7].

Figure 4. Specimens after tensile tests

Figure 5. Sketch of specimen for testing on static tear: 1 –copper; 2 – steel part

Mechanical properties of steel-copper specimens as compared toproperties of copper

Specimen numberYield

strength, MPaTensile

strength, MPaReduction in

area, %

1 110.3 282.5 70

2 113.5 281.0 73.1

3 127.4 282.8 73.1

Average 117.06 282.7 72.06

Copper, wroughtand annealed [12]

74 216 75

54 11/2015

From the cylinders the specimens of 4 × 4 ×× 40 mm were manufactured so that the steel—copper interface was in the middle.

The measurements were carried out using theammeter-voltmeter method in the differentialvariant by moving the potentiometric fork withthe separation of electrodes of 2.3 mm at 1 mmpitch from the steel part to the copper one per-pendicular to the fusion line. The measurementresults are shown in Figure 6.

The diagram in this Figure shows that thechange in electrical resistance at the steel—copperinterface is concentrated on the length of about4 mm. However, the size of this length can notbe identified with the transition zone width ofour joint. In fact, it is much smaller. This isevidenced by the nature of curve of changes ofelectrical resistance in the area of steel—copperinterface.

In fact, judging from the diagram, the initialchange in resistance occurs smoothly from steelto copper during more than 2 mm and then de-creases sharply down to the value inherent tocopper. Such a form of the curve indicates thatthe width of transition zone is smaller than thedistance between electrodes of the potentiometricfork. Therefore, the resolving capacity of themethod proposed in [7], does not allow evaluat-ing the width of thin transition zones, but onlydetermining the average value of electrical resis-tance in it.

To reveal the sizes of transition zone is possibleby metallographic method through investigationof microsections. Figure 7 shows microstructureof bimetal joint produced by autonomous vacuumbrazing of copper on steel surface.

As a result of interaction of melt of copperM1 with steel 20 at 100 magnification the fusionline of increased etching is observed. At 1000magnification the initial stage of penetration ofcopper into steel is visible. The structure of steel(ferrite + pearlite) near the fusion line is notessentially different from the rest array. Aroundthe whole volume of the brazed copper the finenon-metallic inclusions occur, the concentrationof which increases near the interface at the dis-tance of 40—50 μm.

The microhardness was measured at 50 g loadin the direction perpendicular to the fusion line.On the side of steel a sharp increase in micro-hardness was noted. So, in the main array of steelthe microhardness amounts to about 1880 MPa.At the distance of 10—15 μm from the fusion lineand ahead of it the microhardness reaches to3000 MPa. On the side of copper the abruptchanges in microhardness are not observed.

Across the whole cross-section of the brazed cop-per it amounts to 920—1080 MPa.

Thus, studying microstructure of bimetal jointon both sides of interface two anomalous regionswere revealed, where the change of coefficientsof electrical and thermal conductivity may occur.On the side of steel this is the area of increased

Figure 6. Diagram of changes in electrical resistance oftransition zone of steel-copper bimetal produced usingautonomous vacuum brazing

Figure 7. Microstructure of bimetal joint produced usingautonomous vacuum brazing of copper on steel: a – ×100;b – ×1000

11/2015 55

microhardness, and on the side of copper this isthe area of high concentration of non-metallicinclusions.

The total length of these regions does not ex-ceed 70 μm. Therefore, the size of this area maybe identified with the width of transition zonefrom steel to copper.

This width is at least 3 times smaller than thewidth of each component of steel-copper bimetal.Consequently, the total additional electrical andthermal resistances of transition zone, deter-mined as the product of mean values of resistancesin it by its width, will be also 3 times lower thanthe resistance of each part of the bimetallic con-tact. Therefore, during engineering electrical andthermal calculations of bimetal steel-copperjoint, produced using autonomous vacuum braz-ing, these additional resistances can be neglected.

Conclusions

1. The contact between steel and copper in bi-metal joints, produced by autonomous vacuumbrazing-on of copper, during electrical and ther-mal calculations can be considered perfect.

2. During tensile test the fracture of standardbimetallic specimens occurs along copper. At thesame time, the strength properties of brazed cop-per layer exceed the reference values for deformedand annealed copper M1.

3. The ultimate tear strength of the brazedcopper layer determined on the special specimenscorresponds to the ultimate rupture strength ofsteel 20.

4. The technology of autonomous vacuumbrazing can be used to produce bimetal consistingof carbon steels and copper of other grades.

1. Kuniharu, M. et al. One-sided cladding of steelsheet. Pat. 52-123824 Japan. Int. Cl. 12A 220 C 23 C1/00. Fil. 15.10.77. Publ. 8.05.79.

2. Bukov, A.A. (1979) Corrosion-resistant bimetal sheetproducts. Stal, 6, 446—450.

3. Nikolaev, G.A., Olshansky, I.A. (1975) Specialmethods of welding. Moscow: Mashinostroenie.

4. Kudinov, V.M., Koroteev, A.Ya. (1978) Explosionwelding in metallurgy. Ed. by E.S. Karakozov. Mos-cow: Metallurgiya.

5. Nikityuk, Yu.A., Grigorenko, G.M., Zelenin, V.I. etal. (2013) Technology of restoring repair of slabmold of continuous-casting machine with friction stirsurfacing. Sovr. Elektrometallurgiya, 3, 51—55.

6. Paton, B.E., Rossoshinsky, A.A. (1982) Some prob-lems of development of brazing technology. In: Mod-ern methods of brazing, 3—12.

7. Lakomsky, V.I., Bogachenko, A.G., Mishchenko,D.D. et al. (2013) Welded joint of copper with steelin hearth-level electrode of direct current arc furnace.Sovr. Elektrometallurgiya, 4, 7—9.

8. Paton, B.E., Medovar, L.B., Shevchenko, V.E. et al.(2003) Electroslag technologies in production of bi-metal billets. Advances in Electrometallurgy, 4, 7—10.

9. Golovanenko, S.A., Meandrov, L.V. (1966) Produc-tion of bimetals. Moscow: Metallurgiya.

10. Puzrin, L.G., Bojko, G.A., Atroshenko, M.G. (1975)Autovacuum brazing. Kiev: Znanie.

11. Serebryanik, I.P., Buravlev, Yu.M., Ivanitsyn, N.P.et al. Method of production of cooled panels. USSRauthor’s cert. 1235075. Fil. 28.06.84.

12. Keloglu, Yu.P., Zakharievich, K.M., Kartashev-skaya, M.A. (1977) Metals and alloys: Refer. Book.Kishinev.

13. Ovsyannikov, V.G., Shejko, V.I., Malyshev, V.A. etal. (1976) Method of determination of tearing resis-tance of bimetal clad layer. Zavod Laboratoriya, 3,339—340.

Received 30.07.2015

56 11/2015

DETERMINATION OF FORCE CAUSED BY HEATINGOF RING-TYPE PRODUCTS IN FLASH-BUTT WELDING

A.V. MOLTASOV, K.V. GUSHCHIN, I.N. KLOCHKOV, P.N. TKACH and A.I. TARASENKOE.O. Paton Electric Welding Institute, NASU

11 Bozhenko Str., 03680, Kiev, Ukraine. E-mail: office@paton.kiev.ua

In flash-butt welding of products of a closed shape the force, which welding machine should provide forhigh-quality formation of welded joint in a solid phase, is determined not only by upsetting force but alsoby force spent for bending the part itself, and by force caused by heating of shunting part. Therefore, whenchoosing the equipment for welding of any product of closed shape it is necessary to determine the totalforce, which welding machine should provide. The experimental measurements of temperature in differentcharacteristic points at the final stage of the process of FBW of the ring of steel 20 in machine K-724 werecarried out. It was established that the temperature in the shunting part is changed both along the circum-ferential as well as along the radial coordinate. The existing procedure for description of temperature fieldwas improved so that in determination of the Fourier coefficients not only the experimental data were used,but also the data obtained by computer modeling. Basing on the postulates of the proposed model thefunction of general form was obtained, which allows determining the temperature at any point of theshunting part using an unlimited number of terms of approximated series. The calculation of power parametersof FBW with pulsed flashing of the investigated ring was carried out. As a result, it was established thatthe force caused by heating of the shunting part, amounted to 9.1 % of the total force required for qualityformation of the joint. The value of temperature force is connected with the value of equivalent temperaturedisplacement. The estimated value of this displacement, corresponding to the proposed model, correlateswell with the results of numerical modeling of temperature displacements in the welded billet. 11 Ref.,1 Table, 3 Figures.

Keywo r d s : flash-butt welding, ring-type products,shunting currents, temperature field, Fourier series, tem-perature force

The advantages of flash-butt welding (FBW) de-termine its wide application in manufacture ofstructural elements such as frame rings, bands,turntables, stiffening rings, wheel rims and otherproducts of a closed shape [1].

At the E.O. Paton Electric Welding Institutea number of machines for FBW of ring-type prod-ucts was developed, which are able to develop theforce from several dozens to several thousandskilonewtons [2], and some of them have also limi-tations on the minimum inner diameter of the prod-uct being welded [3]. Therefore, the producer en-counters often a difficult task to select the optimumequipment from the existing number of machinesfor welding the certain ring product.

In FBW of parts of an open shape the powerparameters are determined by a specific upsettingpressure of material of the part and the area ofits cross-section [4]. In the process of FBW ofring-type products the bending of the part itselfand the shunting of electric current through itoccur [5].

In work [6] it was proved that heating of theshunting part contributes to arising of thermal

stresses, which affect the volume of the forcerequired for formation of quality welded jointduring FBW. However, the procedure of descrip-tion of the temperature field, demonstrated inthe present work, allows using a limited numberof terms of the series, with the help of which itis not always possible to obtain the desired ac-curacy. Furthermore, no expression was givenfor the force generated by thermal stresses in theshunting part.

Therefore, the aim of this work is the improve-ment of the procedure for description of tempera-ture field and force caused by heating of theshunting part.

The experimental measurements of tempera-tures were carried out during FBW of the ringof steel 20 with cross-section of 55 × 25 mm2 andinner diameter of 270 mm (Figure 1, a) in fivecontrollable points (Figure 1, b).

According to the results of measurements itwas established that the temperature in theshunting part is changed both around the circum-ference θ, as well as around the radial coordinater. As the determination of force caused by heatingof the shunting part is based on solving the rele-vant problems of theory of elasticity, it is neces-sary to preset the temperature as a continuous

© A.V. MOLTASOV, K.V. GUSHCHIN, I.N. KLOCHKOV, P.N. TKACH and A.I. TARASENKO, 2015

11/2015 57

function of the coordinates. In solving the prob-lems in polar coordinates the most convenientfor description of change in temperature is theeven part of the Fourier series around the cir-cumferential coordinate θ:

T(θ) = A0 + ∑ 1

n

An cos nθ.(1)

To determine the unknown Fourier coeffi-cients An it is necessary to create the system ofequations for temperature values in the differentpoints around the circumferential coordinate θof the shunting part. As far as the experimentaldata allow presetting only three temperature val-ues at the outer and inner diameters, namely,T(π/6), T(π/2) and T(π), the other ones canbe determined by computer modeling of the tem-perature field (Figure 2, a). The number of equa-tions of the system determines the accuracy ofdescription of the temperature field. Using thefirst six terms of the series (Table) allows de-scribing the temperature field with the accuracyof up to 3 °C. You can be sure that the values oftemperatures, taken from the diagram (Figure 2,b), coincide with the results of computer mod-eling.

As is seen, the value of the coefficients de-creases according to the absolute value, so theintroduction of the following coefficients willnot make a considerable amendment to the valuesof temperatures.

The change in the temperatures in the radialdirection is determined from the differentialequation of thermal conductivity [7]:

1χ ∂T

∂t =

∂2T

∂r2 +

1r ∂T

∂r +

1

r2 ∂2T

∂θ2, (2)

where χ is the coefficient of thermal conductivity.As far as the temperature field is considered

at the moment when the temperature reaches itsmaximum value, but not its change in time, thenthe left side of equation (2) turns to zero (solu-tion of such equation is given in work [6]).

As a result the function of general view wasobtained corresponding to the proposed model,which allows determining the temperature at anypoint of the shunting part:

T(r, θ) = K0 = H0 ln r + ∑ 1

n ⎛⎜⎝

⎜⎜Knr

n + Hn

rn

⎞⎟⎠

⎟⎟ cos nθ, (3)

where

K0 = A0(a) ln b — A0(b) ln a

ln b — ln a;

Kn = An(b)b

n — An(a)an

b2n — a2n;

H0 = A0(b) — A0(a)

ln b — ln a;

Hn = anbn

b2n — a2n [An(a)b

n — An(b)an].

(4)

(For procedure of determination of stressescaused by the presence of temperature field de-scribed by the Fourier series see works [8, 9].)

So, stress components have the following form:

σr = B0 1

r2 + 2C0 + D0(1 + 2 ln r) +

+ ⎛⎜⎝

⎜⎜2B1r —

2C1

r3 +

D1

r

⎞⎟⎠

⎟⎟ cos θ;

σθ = —B0 1

r2 + 2C0 + D0(3 + 2 ln r) +

+ ⎛⎜⎝

⎜⎜6B1r +

2C1

r3 +

D1

r

⎞⎟⎠

⎟⎟ cos θ;

τrθ = ⎛⎜⎝

⎜⎜2B1r —

2C1

r3 +

D1

r

⎞⎟⎠

⎟⎟ sin θ,

(5)

Figure 1. Ring workpiece in clamping devices of welding machine K-724 (a), and layout of points 1—5 (1—5) of temperaturecontrol (b)

Calculated Fourier coefficients in points 1—5 acc. to Figure 1, b

Coefficient A0 A1 A2 A3 A4 A5

Tin.dia, °C 373 —221 —122.5 —51 —51 —3

Tout.dia, °C 339 —192 —106.5 —48 —51 —3

58 11/2015

where

B0 = — αEH0

2(b2 — a2) a2b2 ln

b2; B1 =

αEH1

4(a2 + b2);

C0 = αEH0

4 b2 ln b — a2 ln a

b2 — a2 +

αEH0

8;

C1 = — αEH1

4(a2 + b2) a2b2;

D0 = αEH0

4; D1 =

αEH1

2,

where α is the coefficient of linear expansion; Eis the modulus of elasticity of material.

The relation of the displacement componentswith the stress components was found in [10].

Therefore, stresses (5) in the closed ring areequivalent to circumferential displacement of thering moving edge with a gap for the value of

δt = —(2π — β)α ⎛⎜⎝H0

a + b2

+ H1⎞⎟⎠. (6)

As far as in our case, according to (4), H0 andH1 are negative, then the displacement δt is di-rected towards the increase of coordinate θ, i.e.is equal to increase of the gap.

In our case the temperature displacementamounted to 2.88 mm. To confirm this result anumerical modeling of displacement fields in thering workpiece was carried out during heating,being similar to that realized in the ring shuntingpart in FBW (Figure 3).

According to the results of numerical model-ing it was established that the temperature move-

ment of the left edge amounted to 3.04 mm, thatis 5.3 % higher than the value obtained by ana-lytical calculation according to formula (6), cor-responding to the temperature field (3).

As far as this displacement is directed oppositefrom displacement of the moving clamping deviceof the welding machine, a part of the force isspent to overcome it. The relation between thedisplacement and the corresponding force wasestablished in work [10]. Thus, to prevent thedisplacement (6) it is necessary to apply the force

Pt = —αEt 2π — β

4π ×

×

⎛

⎜

⎝

⎜⎜⎜⎜H0

(b2 — a2)2 — 4a2b2

2(a + b)(b2 — a2) + H1

a2 — b2 + (a2 + b2) ln b2

a2 + b2

⎞

⎟

⎠

⎟⎟⎟⎟.

(7)

For the investigated ring the calculation ofpower parameters of FBW with pulsed flashing

Figure 3. Field of displacements in the ring (mesh indicatesthe deformed state)

Figure 2. Temperature fields in the ring (a), and diagrams of temperature distribution along the outer (1) and inner (2)diameter (b)

11/2015 59

was performed [11]. The force, spent for bendingof the part, is determined by the formula obtainedin [10]:

Pbend = — EδΣt

4π(a2 + b2) ⎡⎢⎣a2 — b2 + (a2 + b2) ln

ba

⎤⎥⎦, (8)

where δΣ is the sum of the initial gap and allow-ances for flashing and upsetting.

As a result, it was established that the thermalforce amounts to more than 9 % of the total force,which the welding machine should provide toform the quality welded joint in solid phase.

The calculation of power parameters of FBWwith pulsed flashing of the investigated ring wascarried out at the following initial data: innerdiameter 270 mm; cross-section F = 55 ×× 25 mm2; specific upsetting pressure 40 MPa;modulus of elasticity 210 GPa; coefficient of linearexpansion 1.4⋅10—5 1/deg; initial gap 2 mm; al-lowance for flashing 9 mm; allowance for upset-ting 7 mm.

Obtained were the dummy displacement fromheating δt = 2.88 mm; thermal force Pt = 14.8kN; upsetting force Pups = 55 kN; bending forcePbend = 92.6 kN; total force PΣ = = 162.4 kN;Pt/PΣ = 9.1 %.

Сonclusions

1. The procedure was improved, which allowsdescribing the temperature field in the shuntingpart in FBW of ring-type workpiece in the formof even part of the Fourier series around the cir-cumferential coordinate on the basis of discretedata obtained by experimental measurements andcomputer modeling.

2. In accordance with the postulates of theproposed model the function of general form was

obtained, which allows determining the tempera-ture at any point of the ring shunting part usingunlimited number of series terms.

3. For the first time an analytical expressionwas obtained for determination of the force actingon the butt caused by heating of the shuntingpart. For the ring of steel 20 with inner diameterof 270 mm and cross-section of 55 × 25 mm2 thepower parameters of FBW using pulsed flashingwere calculated. As a result, it was establishedthat the force caused by heating of the shuntingpart amounts to 9.1 % of the total force, whichwelding machine should provide.

1. Pavlichenko, V.S. (1964) Resistance welding ofclosed-shape products. Moscow: Mashinostroenie.

2. Kuchuk-Yatsenko, S.I. (1992) Flash butt welding.Kiev: Naukova Dumka.

3. Kuchuk-Yatsenko, S.I, Chvertko, P.N., Semyonov,L.A. et al. (2013) Flash butt welding of products ofhigh-strength alloys based on aluminium. The PatonWelding J., 7, 2—6.

4. Gelman, A.S. (1952) Technology of electric resis-tance welding. Moscow: Mashgiz.

5. Kabanov, N.S. (1973) Welding on resistance ma-chines. Moscow: Vysshaya Shkola.

6. Moltasov, A.V., Samotryasov, S.M., Knysh, V.V. et al.(2014) Influence of non-uniformity of heating on upset-ting force value and forging time in flash-butt weldingof flat ring. The Paton Welding J., 10, 11—14.

7. Karslou, G., Eger, D. (1964) Thermal conductivityof solids. Moscow: Nauka.

8. Timoshenko, S.P., Goodier, J. (1975) Theory of elas-ticity. Moscow: Nauka.

9. Boli, B., Ueiner, J. (1964) Theory of temperaturestresses. Moscow: Mir.

10. Chvertko, P.N., Moltasov, A.V., Samotryasov, S.M.(2014) Calculation of upsetting force in flash buttwelding of closed-shape products. The Paton Weld-ing J., 1, 46—50.

11. Kuchuk-Yatsenko, S.I., Didkovsky, V.A., Bogorsky,M.V. et al. Method of flash butt welding. Pat.46820 Ukraine. Publ. 17.06.2002.

Received 11.02.2015

60 11/2015

INFLUENCE OF METAL SHRINKAGEIN LONGITUDINAL WELDS OF SLEEVES

ON CONTACT PRESSUREIN MAIN GAS PIPELINE REPAIR

O.I. OLEJNIKE.O. Paton Electric Welding Institute, NASU

11 Bozhenko Str., 03680, Kiev, Ukraine. E-mail: office@paton.kiev.ua

Effectiveness of main pipeline repair in service is largely determined by contact pressure, created at sleevemounting on pipe defective section. At correct selection of contact pressure value, long-term performanceof the repaired section is ensured through partial unloading of the pipe wall. Such a result is achieved dueto lowering of pressure in the main line before repair, and tight pressing of the sleeve around the pipe witha special device. Arc welding of sleeve reinforcing elements also is of considerable importance, resultingin formation of contact pressure, caused by shrinkage of multilayer longitudinal butt welds. In the workcomputational method was used to study the influence of sleeve wall thickness and pipeline diameter onthe magnitude of contact pressure caused by deposited metal shrinkage. It is shown that contact pressurecaused by metal shrinkage during welding performance has the strongest influence on unloading of adefective pipe wall in pipelines of 1020 mm and smaller outer diameter. Further increase of contact pressuremagnitude caused by metal shrinkage can be achieved by increasing sleeve wall thickness. Such a techniqueis not effective for pipes of 1220 mm and greater diameter. 5 Ref., 5 Figures.

Keywo r d s : main gas pipeline, sleeve, repair, arcwelding, deposited metal shrinkage, contact pressure

Keeping Ukrainian gas transportation system inthe appropriate technical condition by repairwelding without taking the damaged sections outof service is an important scientific-practicaltask. Technologies involving application of rein-forcing welded sleeves currently prevail in prac-tical repair of corrosion damage of pipes of maingas pipeline linear part [1]. Such structures con-sist of two symmetrical half-shells, which arepressed around the pipe outer surface duringmounting to create tight contact (interference).As a rule, such operations are performed underthe condition of inner pressure lowering in themain line.

It is known that contact pressure ΔPl arisingat sleeve mounting on the pipe has an importantrole in ensuring reliable operation of repairedsection [2]. This is associated not only with forcepressing of half-shells during assembly by variousdevices, but also additional redistribution ofloads between pipe and sleeve walls at inner pres-sure increase from repair value Prep up to workingpressure P (Figure 1). As the assembly processinvolves multilayer welding of half-shells to eachother by two parallel longitudinal butt welds,contact pressure ΔPsh arises additionally, as aresult of deposited metal shrinkage.

The objective of this work consists in deter-mination through calculation of ΔPsh value forevaluation of the degree of its influence on totallevel of contact pressure ΔPl.

Calculation was based on the idea of determi-nation of reactive stresses σr in a closed shell,which arise after making the longitudinal weldsand act in the circumferential direction (Fi-gure 2). Assuming that stresses σr are uniformlydistributed across sleeve wall section, and con-

Figure 1. Schematic of edge preparation in sleeve half-shellsfor butt welding: 1, 2 – pipe and sleeve wall; 3 – backingplate; D, Dsl – pipe and sleeve outer diameters, respec-tively; P, Prep – pressure inside the pipeline in workingmode and during repair, respectively© O.I. OLEJNIK, 2015

11/2015 61

dition of dense contact (Λ >> 0) is fulfilled,average contact pressure ΔPsh can be found fromthe following formula:

ΔPsh = σr 2δsl

D + 2δsl

, (1)

where D is the pipeline outer diameter; δsl is thenominal thickness of sleeve wall.

Known from theoretical fundamentals ofwelded structure design [4] is the calculated de-pendence for calculation of reactive stresses forthe cases of transverse location of welds relativeto the action of reactive stresses:

σr = —μ′E qh.i

F,

(2)

where F is the welded element section in thedirection normal relative to weld axis; qh.i is theaverage welding heat input; E is the materialmodulus of elasticity; μ′ is the coefficient char-acterizing weld transverse shrinkage and deter-mined by graphic dependence μ′ = f(m) (Fi-gure 3).

For the sleeve the section will be defined as

F = π(D + 2δsl)δsl. (3)

Parameter m, which reflects relative rigidity,is calculated from the following formula for sucha structure:

m = 0.125 ⋅106 4F

vwqh.i2

, (4)

where vw is the welding speed.In butt welding in several passes the condi-

tional heat input is calculated by the followingformula:

qcon = 340 ⎛⎜⎝Fgr +

2cos θ/2

rδsl

⎞⎟⎠, (5)

where Fgr is the area formed by butt weld groove;θ is the total groove angle; r = 43.5⋅10—3√⎯⎯⎯⎯qh.i .

Considering that sleeve half-shells werewelded to each other by two butt welds, formula(2) can be written as

σr = —μ′E 2qcon

F. (6)

Let us consider an example of calculation ofthe magnitude of contact pressure ΔPsh for thecase of sleeve repair of corrosion damage withoverall dimensions of 250 × 150 mm in the lon-gitudinal S and circumferential C directions andmaximum depth a = 4 mm. Pipe of X60 strengthclass of D = 1020 mm, δ = 10.5 mm was usedwith sleeve wall thickness δsl = 10.5 mm (δ == δsl). For calculation we will assume the fol-lowing welding mode, characteristic for 4 mmelectrode: I = 140 A, U = 24 V, η = 0.7, vw == 0.2 cm/s.

Then, qh.i = 0.24IUη(vw)—1 = 2822 [cal/cm];r = 2.31 cm; Fgr = 0.44 cm2.

We will determine welded element section inthe normal direction from formula (3)

F = π(D + 2δsl)δsl == π(102 + 2⋅1.05)⋅1.05 = 244 cm2.

We will calculate conditional heat input byformula (5), considering that groove angle θ == 40° and qcon = 340(0.53 + 2⋅2.31⋅1.05(cos 20°)—1 == 1905 cal/cm.

Sleeve relative stiffness is equal to

m = 0.125 ⋅106 4F

vwqh.i2

= 108.

Then, the coefficient, which determines weldtransverse shrinkage, will correspond to valueμ′ = —4/97⋅10—6 cm/cal.

Reactive stress magnitude in keeping with (6)is equal to

Figure 2. Stress-strain state of pipe and sleeve walls underthe impact of inner pressure: δ, δsl – pipe and sleeve wallthickness, respectively; ΔPl – contact pressure of sleeveon the pipe; ΔPsh – contact pressure of sleeve on the pipecaused by weld shrinkage

Figure 3. Coefficient μ′ versus relative rigidity m

62 11/2015

σr = —μE 2qcon

F = 110 kg/cm2.

Magnitude of average contact pressure be-tween the sleeve and pipe ΔPsh is determined as

ΔPsh = σr 2δsl

D + 2δsl = 2.22 kg/cm2 or ΔPsh = 0.22 MPa.

Figure 4 presents in the graphic form the re-sults, obtained with application of the above cal-culation method, showing the dependence of ΔPshon sleeve wall thickness δsl and pipeline diameterD. To evaluate the degree of ΔPsh influence onlevel of contact pressure ΔPl it is necessary tohave the diagrams of admissible linear dimensionsof pipe wall thinning, the sample of which isshown in Figure 5, and criterion of groundedselection of ΔPl value. The latter has the follow-ing form [5]:

ΔPl ≥ P — [P] — χ1(P — Prep), (7)

where [P] is the admissible pressure for pipe wallthinning, determined by diagrams of admissiblelinear dimensions;

χ1 = ⎛⎜⎝

⎜⎜1 +

(0.5Dsl)2δ

(0.5D)2δsl

⎞⎟⎠

⎟⎟

—1

.

In view of the requirements of normative docu-ments on repair in operating gas pipelines thatPrep = 0.7P (P = 5.4 MPa), for the consideredexample [P] = 0.75P, χ1 = 0.49. Then, in keepingwith (7), ΔPl = 0.55 MPa. Comparing the cal-culated value with ΔPsh = 0.22 MPa, one can saythat in this case contact pressure caused by lon-gitudinal weld shrinkage makes up a significantpart (about 40 %) of minimum required contactpressure. Their total value contributes to the mar-gin for ensuring the performance of gas pipelinesection repaired using the welded sleeve.

At the same time, it should noted that withincrease of pipe diameter D, value of admissiblepressure [P] will decrease for a defect with thesame overall dimensions, that will require in-creasing the minimum needed contact pressureΔPl, accordingly. This leads to the conclusionabout a decreasing role of ΔPsh.

Thus, it can be stated that shrinkage in weld-ing of sleeve longitudinal welds on gas pipelineswith outer diameter of 1020 mm and less willhave the strongest influence on unloading of pipedefective wall. In repair of the main gas pipelineswith 1220 mm and greater outer diameters, at-tention should be focused on providing the re-quired value of contact pressure ΔPl, while in-fluence of value ΔPsh should be regarded as in-significant. A certain increase of ΔPsh value canbe achieved by increasing sleeve wall thicknessδsl. Such a technique, however, has quite limitedpractical application, as a result of deteriorationof mounting conditions, because of half-shell ri-gidity and complexity of ensuring the requiredvalue of contact pressure ΔPl.

1. GBN V.3.1-00013741-12:2011: Main gas pipelines.Repair by arc welding in service conditions. Kiev:Ministry of Energy and Coal Industry of Ukraine. In-trod. 06.09.2011.

2. Kiefner, J.F. (1977) Repair of line pipe defects byfull-encirclement sleeves. Welding J., 6, 26—34.

3. Aginej, R.V., Puzhajlo, A.F., Aleksandrov, Yu.V. etal. (2012) Methods of safety assurance of long-termservice gas pipelines prone to stress-corrosion. Kor-roziya Territorii Neftegaz, 23(3), 50—61.

4. Okerblom, N.O. (1964) Structural-technological de-sign of welded structures. Moscow-Leningrad:Mashinostroenie.

5. Makhnenko, V.I., Olejnik, O.I., Shekera, V.M.(2013) Determination of contact pressure of reinforc-ing sleeve in repair of pipelines with surface defects.The Paton Welding J., 6, 11—14.

Received 09.10.2015

Figure 4. Influence of sleeve wall thickness δsl on magnitudeof contact pressure caused by longitudinal weld shrinkageΔPsh depending on pipeline diameter D

Figure 5. Diagram of admissible linear dimensions of pipewall thinning depending on pipe wall minimum thicknessδmin for 1020 × 10.5 mm gas pipeline from steel of strengthcategory X60 with maximum working pressure P = 5.4 MPa

11/2015 63

NEW WELDING WIRE MANUFACTURER IN UKRAINE

Manufacture of special-purpose metal structures from high-al-loyed corrosion-resistant, high-temperature, heat-resistant andhigh-strength steels is developing at a faster pace all over theworld. This is due to advance of technology in petrochemicalindustry, power engineering, construction industry, intensifi-cation of transport operation, and designers’ desire to reducethe weight and overall dimensions of metal structures. Con-sumable-electrode arc welding, gas-shielded welding and sub-merged-arc welding still remain the main technologies in fab-rication and repair of such metal structures.

The structure of Ukrainian manufacturers of welding wiresfor all arc welding processes, also those for stick electrodemanufacture, includes only manufacturers of wires from low-carbon and some low-alloyed steel grades. Production of spe-cial-purpose wires: high-alloyed corrosion-resistant, heat-re-

sistant, nickel-based, high-strength, etc., was absent in Ukraine until recently. All the above-mentioned typesof wires were imported from China, India, Russia, Italy and other countries.

In 2010, a group of enthusiasts from Boyarka (Kiev suburb) took a decision to set up modern manufactureof special-purpose welding wires in their facility. The group included graduates of Welding Chair of KievPolytechnic Institute, having experience of working at PWI and in welding consumable production, busi-nessmen and financial people. After detailed study of Ukrainian market of high-alloyed corrosion-resistant,high-temperature and other special wires (including welding wires), preparation of detailed business plan,selection of production area, preparation of engineering lines and communications, the team began imple-mentation of the project. During project preparation the most recent advances of the technology of wireproduct processing were analyzed, namely features of wire rod preparation for drawing, influence of varioustypes of drawing tools and process materials on the quality of the produced wire, kinds of finished wiresurface treatment to meet the expectations of consumers, used to working with wire, complying with therequirements of European and US standards. At this stage of project preparation, special attention was givento power efficiency of production process, as well as minimizing or complete elimination of adverse environ-mental impact of future production. Technological scheme and set of equipment for process lines were selectedonly on the basis of achieving maximum possible consumer characteristics of finished wire.

As a result of analysis of equipment supplier proposals, one main manufacturer – well-known companyLamnea Bruek, and several sub-suppliers of a number auxiliary devices from Germany, Italy, Austria, USAand India, were selected from a number of potential manufacturers, proceeding from detailed technologicalscheme of production. It should be noted that various auxiliary devices are key elements of the technologicalchain in manufacture of high-alloyed and special-purpose wires. Without these specialized additional devicesthe drawing line can only support production of general-purpose wires.

All the potential manufacturers and suppliers of process equipment were first inspected (visiting andauditing the production), in order to obtain information on the engineering level and service properties ofthe proposed equipment, its reliability, ease and safety of service. Finally, it was possible to combine in oneproduction line all the best currently known engineering solutions in manufacture of special-purpose wires.

A.N. Alimov,Technical Director of Company VITAPOLIS

64 11/2015

Start of production was organized by VITAPOLIS company, which began manufacturing marketable wirein July, 2015. The wires were assigned registered KHORDA trade mark.

When putting the line into operation, we performed commissioning work, using Sv-08GS and Sv-08G2Swire rods. Testing of produced general-purpose wires showed that these wires produced by our process scheme,in terms of their welding-technological properties are in no way inferior to such known wire grades asAristoRod 12.50 (ESAB) and EMK-6 Top (Boehler).

Our production line is capable of preparing for drawing and cleaning in a sound and highly reproduciblemanner the surface of wire rods, having difficult-to-remove or slight unforeseen contamination, possibledeviations of diameter, ovality, etc. Here, already at the inlet of the first drawing block, the rod surface hascharacteristic uniform metal luster, as after grinding by an abrasive tool. Ovality of wire leaving the firstdrawing block with rotating die, is not more than 0.01 mm, and its surface is super smooth. Further wiredrawing is performed with application of water-soluble stearates of sodium, calcium and drawing tools ofBremer and Paramount companies. Finish preparation of drawing tools is performed in our die shop in Bremersemi-automatic machine.

Remains of drawing lubricant are removed from the surface of finished wire by hot water at high pressure,which is followed by further polishing of the wire and covering its surface with a thin layer of antifrictioncurrent-conducting coating, containing corrosion inhibitor. Finished wire is wound on standard consumerreels of 200, 300 or 415 mm diameter, packed into plastic bags and cardboard boxes. At the request of thecustomer, the wire can be additionally packed into vacuum bags.

It should be noted that manufacture of a wide range of special-purpose wires, including those designedfor welding low-alloyed and high-strength steels, high-alloyed corrosion-resistant and heat-resistant steels,nickel-based alloys and armour steels, requires close monitoring and control of all the technological processstages.

For these purposes, testing and measuring laboratory is functioning at VITAPOLIS enterprise from thefirst days of its operation. The laboratory includes a range of equipment and set of devices for mechanicaltesting, chemical analysis, metallographic examination of raw materials and finished products. Welding-tech-nological testing section is fitted with stations for gas-shielded semi-automatic welding and mechanizedsubmerged-arc welding.

Before the start of production, we paid special attention to training production personnel, realizing thateven in the best equipment high-quality products with appropriate consumer properties can be produced onlyin the case of its operation by trained, qualified workers. Our engineers provided detailed descriptions of allthe elements of technological process, working in close contact with equipment manufacturers, technologistsfrom Germany, Sweden and Austria. After competitive selection process, all the workers had productiontraining directly in our Enterprise, in accordance with developed technological and operating manuals.Personal responsibility of each worker for the performed technological operation was introduced, with re-cording and documenting of the performed operations. Quality management system, corresponding to ISO9001, is in force at the Enterprise.

In October 2015, VITAPOLIS presented their products at the International Exhibition «Arms and Safety-2015» (Kiev), which aroused considerable interest of representatives of defence industry enterprises of Ukraineand a number of foreign countries.

By now, the first orders have been fulfilled for supplying welding wires to«Frunze Elektrod», «Gefest»and other companies. At present the Enterprise production facilities allow producing about 100 tons of weldingwire of 15 different grades per month. Company investment program envisages putting into production about30 more wire grades in 2016.

A.N. Alimov, VITAPOLIS

11/2015 65