PSI - Issue 19

Hugo Heyraud et al. / Procedia Structural Integrity 19 (2019) 566–574 H.Heyraud et al. / Structural Integrity Procedia 00 (2019) 000–000

574

9

Good correlation between the proposed model and the reference model is observed for the 3 welded structures. The maximum di ff erence is 2,7% for structure 1, loaded in the y direction. These results are consistent with the Fayard model and the Lohr model. However, the results from the model currently used by Manitou show a maximum di ff erence greater than 10% for 4 of the configurations studied. The di ff erences between the 4 models can be explained by the level of complexity used to model the welded assemblies. The local sti ff ness of the weld is not considered in the model currently used by Manitou, and it is idealized by specific meshing rules in the Fayard and the Lohr models. In the proposed approach the local geometry of the weld is fully considered at the expense of greater calculation time. However it should be noted that this is mainly pre-processing necessary to establish a library of weld geometries. The proposed model seems to be a good alternative to the current ones, however, other welded assemblies should be tested in order to confirm these initial results. This work focuses on the fatigue design of welded assemblies. For this, a numerical strategy to consider the impact of the welds on the overall sti ff ness of these assemblies has been proposed. The proposed method to model welded structures is based on a combined surfacic-volumetric approach. Weld areas are modelled using solid elements in order to take into account the real sti ff ness of the weld while steel sheets are modelled using shell elements. In order to limit computational costs, the local models of the welds are condensed into an equivalent sti ff ness matrix and then, the equivalent sti ff ness matrix is inserted into the global shell elements model. After calculation, from displacements at the connecting nodes of the sti ff ness equivalent matrix, the local stress state at weld toes and weld root can be determined. To validate the sti ff ness behaviour of this method, three elementary welded assemblies have been modelled using solid elements as the reference case and two other methods from the literature. The results show good correlation between the proposed approach and the reference models for three di ff erent structures. These comparisons highlight the improvement that can be easily achieved in terms of the model currently used by Manitou. Future work will focus on the fatigue evaluation of welded assemblies, textcolorblueonce the weld sti ff ness and consequently the local stresses are accurately known. Fatigue tests on 4 elementary welded structures are being carried out to estimate the influence of the stress gradient at the hot spot as well as size e ff ects. Then, from an analysis of the experimental data and by knowing the local stress state at the hot spots, an equivalent stress will be defined in order to obtain a single ’master’ Wo¨hler curve taking into account the stress gradient, the size e ff ect and the stress multiaxiality. The validation process of the method will include a fatigue test on a full scale chassis structure. 5. Conclusion [1] Bennebach, Dimensionnement des assemblages soude´s en fatigue. e´tat de l'art sur les me´thodes et techniques de mode´lisations e´le´ments finis. confrontations de certaines approches a` des re´sultats d'essais, Cetim, N / Re´f : 2018 / 4S / AC / MBEN / CTHIl (2018). [2] Osawa, Sawamura, Shota, Suzuki, Study on the relationship between shell and solid stresses in the vicinities of web sti ff ened cruciform joints, IIW, XIII-2288-09 (2009). [3] Ho ff , Mueller, Wallerstein, Weld modeling with msc.nastran (2000). [4] Fricke, Iiw guideline for the assessment of weld root fatigue, IIW-Doc. XIII-2380r1-11 / XV-1383r1-11 57 (2013) 753–791. [5] Fayard, Dimensionnement a` la fatigue polycyclique de structures soude´es, Ph.D. thesis, Palaiseau, Ecole polytechnique (1996). [6] Turlier, Klein, De Noni, Lawrjaniec, Fea shell element model for enhanced structural stress analysis of seam welds, Procedia Engineering (2011). [7] Dang Van, Bignonnet, Fayard, Janosch, Assessment of welded structures by a local multiaxial fatigue approach 24 (2001) 369 – 376. [8] Ferme´r, Andre´asson, Frodin, Fatigue life prediction of mag-welded thin-sheet structures, SAE International (1998). [9] Klein, Turlier, Ober, Fea structural stress of modalohr system for semi-trailer rail transportation: Weld root fatigue focus, Procedia Engineering 66 (2013) 79 – 87. [10] Fezans, Analyse line´aire et non line´aire ge´ome´trique des coques par e´le´ments finis isoparame´triques tridimensionnels de´ge´ne´re´s, Theses, E´ cole Nationale Supe´rieure de l'Ae´ronautique et de l'Espace de Toulouse (1981). References