Micromechanical fracture analysis of high strength steel weldments

Industrijska primena celika povišene cvrstoce u elementima zavarenih konstrukcija cini neophodnim poznavanje ponašanja spojeva ovih materijala pri žilavom lomu. Stoga, procena integriteta zavarenih struktura je potrebna da bi se obezbedio potrebni nivo sigurnsti i pouzdanosti, imajuci u vidu uticaj ogranicenog deformisanja i heterogenosti na ponašanje pri lomu kriticnih zona spoja: zone uticaja toplote (ZUT) i metala šava (MŠ), koji cesto imaju nižu žilavost i višu prelaznu temperaturu. Takode, veoma je važno da procena integriteta bude uradena realno i da ne bude previše konzervativna, kako bi se sprecilo povecanje mase strukture i obezbedilo ekonomicno korišcenje materijala. U ovoj disertaciji, mikromehanicki pristup je korišcen za analizu uticaja mehanicke heterogenosti i ogranicenog deformisanja na žilav lom zavarenih spojeva celika povišene cvrstoce. Ovaj pristup je korišcen kao rešenje za problem prenosivosti parametara klasicne mehanike loma. Takode, motiv za njegovu primenu je i to što standardni parametri mehanike loma: faktor intenziteta napona, otvaranje prsline i konturni Jintegral, ne mogu na odgovarajuci nacin opisati odgovor materijala sa prslinom na dejstvo spoljnog opterecenja u svim uslovima, kao što su izraženo plasticno tecenje (large scale yielding), razliciti uticaji heterogenosti, obika i geometrije zavarenih konstrukcija u eksploataciji. Parametri mehanike loma, odredeni laboratorijskim ispitivanjem epruveta, nisu direktno prenosivi na komponente i stoga se moraju uzeti u obzir dodatni faktori (kao što je uticaj ogranicenog deformisanja). Takode, cilj ove disertacije je odredivanje mehanickih osobina zona zavarenog spoja korišcenjem kombinovanog eksperimentalnonumerickog postupka, što je narocito važno kod zona male širine u okviru ZUT koje su izložene opterecenju u transverzalnom pravcu. Rad na disertaciji je podrazumevao primenu metode konacnih elemenata i eksperimentalna ispitivanja. Eksperimentalna ispitivanja su uradena na zavarenim glatkim epruvetama (uz korišcenje ARAMIS stereometrijskog mernog sistema) za odredivanje mehanickih osobina, kao i epruvetama za savijanje u tri tacke i epruvetama za zatezanje sa pocetnom prslinom u metalu šava i zoni uticaja toplote za analizu ponašanja pri lomu. J-R krive i vrednosti parametra mehanike loma koje odgovaraju pocetku rasta prsline odredene su eksperimentalno i numericki na epruvetama sa pocetnom prslinom u ZUT i MŠ. Numericka analiza elasto-plasticnih modela konacnih elemenata (2D i 3D) je uradena u programskom paketu Abaqus, a mikromehanicki kompletni Gursonov model (CGM) je primenjen preko korisnickog potprograma UMAT (autor Z.L. Zhang). Ograniceno deformisanje oko vrha prsline i promena troosnosti u ligamentu su numericki analizirani na epruvetama za savijanje u tri tacke i onim za zatezanje, da bi se analizirala prenosivost mikromehanickih parametara oštecenja sa jedne epruvete na drugu. Takode, mikromehanickim pristupom je odreden uticaj geometrije i rezultati su uporedeni sa eksperimentalnim podacima. Najveci deo numerickih rezultata dobijenih korišcenjem CGM modela pokazuje dobro slaganje sa eksperimentalnim rezultatima...The industrial application of high strength steel in structural welded components has increased the demand for understanding the ductile failure behavior of this type of welded materials. Therefore, integrity assessment of those welded structures is required in order to ensure a certain level of safety and reliability, having in mind the effects of constraints and heterogeneity on fracture behavior of crucial regions such as: heat affected zone (HAZ) and weld metal (WM) which usually have low toughness and higher transition temperature. It is also essential that the assessment is done in a realistic and not too conservative way in order not to increase the mass of the structure or impair the economic efficiency too much. In this thesis, micromechanical approach has been used to study the effect of mechanical heterogeneity and constraints on ductile fracture of high strength steel weldments. This approach has been used as a solution for the transferability problem of conventional fracture mechanics parameters. It has also been used on basis of that fracture mechanics parameters recommended by standard, such as: stress intensity factor, crack opening displacement and contour J-integral, cannot reliably describe the reaction of a pre-cracked material to the effects of external loading under all conditions such as: large scale yielding, various effects of heterogeneity, shape and geometry of real welded structures. The fracture mechanics parameters, determined from laboratory scale experiments are not also directly transferable to components and hence additional considerations (like constraint effects etc.) need to be taken care of. In addition, the aim of the thesis was to estimate precise mechanical properties using a combined experimental and numerical procedure for various welded joint regions, especially for narrow HAZ regions, when they are subjected to transversally applied load. The study was carried out using finite element method and experiments. Experimental analysis was carried out on: welded smooth tensile specimen with ARAMIS measuring system for estimation mechanical properties, welded single-edge notched bend and flat tensile specimens with pre-cracks in weld metal (WM) and heat-affected zone (HAZ) for studying the ductile fracture behavior. J-R curves and crack growth initiation iii values of fracture mechanics parameter were experimentally and numerically obtained for specimens with a pre-crack in HAZ and WM. Numerical analysis of elastic-plastic finite element models (2D and 3D) was performed in software package Abaqus, with micromechanical complete Gurson model (CGM) applied through user subroutine, UMAT (author: Z.L. Zhang). The crack tip constraint and variation of stress triaxiality in ligament were numerically analyzed on single-edge notched bend and tensile specimens to analyze the transferability of micromechanical damage parameters from one specimen to another. In addition, the geometry effects were also studied by the micromechanical approach and the results were compared with those of the experiments. Most of numerical results obtained with CGM model are in good agreement with the experimental results..