Numerical assessment of the influence of preheating and boundary conditions on welding residual stresses and deformations

Zbog jednostavnosti izrade, niskih troškova, pouzdanosti i brzine izvedbe zavarivanje je jedan od najčešće primjenjivanih postupaka u mnogim industrijskim područjima. Zbog velikog lokaliziranog unosa topline dolazi do taljenja elektrode i bliskog područja osnovnog materijala te do pojave zaostalih naprezanja i deformacija. Vlačna zaostala naprezanja u kombinaciji sa radnim opterećenjem doprinose smanjenju radnog vijeka u slučaju prisutnosti pukotina, a trajne plastične deformacije predstavljaju problem pri montaži. Uklanjanje zaostalih naprezanja i deformacija mehaničkim ili toplinskim postupcima predstavlja dodatni trošak i produljuje vrijeme izrade proizvoda. Iz tih razloga u današnje vrijeme primjenom numeričkih metoda nastoji se doći do rješenja već u fazi projektiranja konstrukcije koje će minimalizirati zaostala naprezanja i deformacije. U ovom radu provedena je numerička simulacija zavarivanja dviju čeličnih ploča u obliku T-spoja pri čemu je za problem prijenosa topline spregnutog s nelinearnom mehaničkom analizom korištena sekvencijalno spregnuta metoda. Rad se sastoji od 9 poglavlja. U poglavlju 1 dan je kratak uvod u temu i definirana je problematika procesa zavarivanja. U poglavlju 2 objašnjena je teorija zaostalih naprezanja, deformacije i ponašanje elastoplastičnog materijala nakon čega su u trećem poglavlju dane teorijske osnove termodinamike zavarivanja. Nadalje u poglavlju 4 prikazana je toplinsko-mehanička analiza nakon koje je opisana primjena metode konačnih elemenata na simulaciji procesa zavarivanja. U istom poglavlju također su opisani konačni elementi korišteni u toplinskoj i mehaničkoj analizi. U poglavlju 5 dana je verifikacija numeričkog modela usporedbom numeričkih rezultata referentnog modela (M1) na primjeru ploča zavarenih u obliku T-spoja, s dostupnim rješenjima iz literature. U poglavlju 6 obrađen je utjecaj toplinskih rubnih uvjeta na progibe horizontalne ploče T-spoja i na zaostala naprezanja u pravcu zavarivanja. U radu je pretpostavljeno hlađenje zavarene konstrukcije zrakom, prirodnom konvekcijom, na temperaturu okoline. Referentni model M1 s definiranim konstantnim koeficijentom konvektivnog prijelaza topline uspoređen je s modelom M2 gdje je koeficijent konvektivnog prijelaza topline definiran kao temperaturno ovisan. U okviru ovog poglavlja zaključeno je kako način modeliranja koeficijenta konvektivnog prijelaza topline nema veliki utjecaj na raspodjelu temperatura, progiba i zaostalih naprezanja. U poglavlju 7 promatran je utjecaj mehaničkih rubnih uvjeta na progibe i zaostala naprezanja za model M3 u kojem je ukrućenje stalno prisutno, te modele M4 i M5 gdje se mehanički rubni uvjeti uklanjaju nakon što se model ohladi na temperaturu okoline (model M4) ili odmah nakon zavarivanja (M5). Pokazano je da su progibi slobodno zavarenog (bez ukruta) modela M1 puno veći nego je to slučaj kod modela koji su mehanički ukrućeni u toku zavarivanja. Također je uočeno veće naprezanje kod stalno uklještenog modela u odnosu na ostala tri razmatrana modela. U osmom poglavlju istražen je utjecaj temperature predgrijanja konstrukcije na progibe horizontalne ploče T-spoja i na zaostala naprezanja u pravcu zavarivanja. Referentni model M1 uspoređen je s modelima M6 i M7. Temperature predgrijanja iznosile su 100 °C (M6) i 150 °C (M7) te je pokazano kako predgrijanje ima pozitivan utjecaj na smanjenje navedenih rezultata.Due to the simplicity, low costs and reliability the welding process is one of the frequently used engineering methods of joining structural components in many industry fields. The large localized heat generation during welding leads to residual stresses and deformations. Tensional residual stress along in combination with additional stress due to exploitation reduces the durability of the construction, while plastic deformations cause inconveniences during the structure assembly. Removal of residual stress and deformation, using conventional post-weld thermal or mechanical treatments, represents an additional cost and prolongs the production time. Therefore, striving to reach a solution in construction phase, which will reduce the residual stress and deformation, different numerical methods are being used. In this paper a numerical simulation of a welding process of a T-joint fillet weld was carried out. To solve the problem of the heat transfer combined with nonlinear mechanical analysis the sequentially coupled method was used. The paper consists of nine chapters. Chapter 1 is a introduction with the defined welding process problem. In chapter 2 the theory of the residual stress, deformation and the behaviour of elastoplastic material was explained which follows to chapter 3, where are given the basics of thermodynamics of the welding process. Furthermore, the chapter 4 shows the heat-mechanical analysis after which the application of the finite element method on the simulation of the welding process is described. In the same chapter the finite elements used in the heat and mechanical analysis are described. Also, the same chapter gives the explanation of the verification proces by comparing the numerical results of the reference model (M1) on the example of T-joint panels with available soultions from the literature. The chapter 6 deals with the impact of thermal boundary conditions on deformations (deflections) and the longitudinal residual stress of the horizontal T-shaped panel. In this paper the cooling of welding is through natural convection on the environment temperature. The reference model M1 with defined convective heat transition coefficient is compared to the M2 model where the convective heat transition coefficient is defined as depending on temperature. Within this chapter it is concluded that the way of modeling of the convective heat transition coefficient does not have a big influence on the distribution of the temperature, deflection and residual stress. Chapter 7 observes the influence of mechanical boundary conditions on deflection and longitudinal residual stress for model M3 in which the encastre is present during the whole simulation and the models M4 and M5 where the mechanical boundary condition, encastre, is excluded after the model is cooled on the environment temperature (model M4) or right after welding (model M5). It is concluded that the deflections of free welded model M1 are much bigger than in case of the models that are encastred only during the welding process. Also, a bigger residual stress is noticed in permanently encastred model (M3) in comparison with the other three models. In chapter 8 the construction was preheated on a previously defined temperature. The impact of the temperature of preheating is expressed on deflections of the T-joint horizontal plate and on the longitudinal residual stress. The reference model M1 is compared with models M6 and M7. The preheating temperatures were 100°C (M6) and 150°C (M7) and it is showed how the preheating has an positive influence on reduction of the mentioned results