For convenience of calculation, the values of effective shrinkages in Table 1 and 2 should be introduced into the model with positive sign, and the effective temperatures with negative sign.
Two thermo-elastic calculations have been performed with a conventional FEM code (PERMAS version 4.0). For information, the equivalent needed CPU-time is less than 15 minutes on recent SGI computer (ORIGIN 200) and the necessary disk-place, about 2 Go for each calculation [10].
Results of both models are resumed in Table 3.
It is seen on Table 3 that two conditions required by the procedure (effective buckling limit and limit tolerance) are entirely satisfied for both two models.
The corresponding calculated deformed shapes and longitudinal compressive stress distributions on top surface layer of two PE models are presented in Fig.8 and 9.
Since these obtained results and taking into account the criterion of minimum structural weight, the Constructor-Engineer has finally decided to choose the first solution: Model (1, 1).
5. DISCUSSION
5.1 Linear Heat Input Energy
Through the above example, it is seen that in a welding process, heat input energy is an important factor influencing the deformation of a welded assembly. So, that could be to say "Decreasing the linear Energy decreases the Distortion".
5.2 Optimum Thin Plate Panel
In a large sense, the so-called "optimum" thin plate panel ought to fit up three main conditions ensuring:
→ the minimum of structural weight, (i.e of welding operations),
→ the low deflection level induced by welding effects,
→ the maximum "no risk" in buckling mode under external loadings in rough sea state conditions of navigation and particularly under regular sagging wave loadings prescribed as rules by all the ship classification societies.
In the above example, taken account of the calculated maximum compressive stress value induced by welding effects, the selected Model (1, 1) remains itself more than 50 % of its effective resistant capacity against buckling load due to such any sagging wave bending moment of the hull girder.
And therefore, that could be to say also "Decreasing the Distortion decreases the Residual stresses".
5.3 Optimization at Design Stage
In addition to the basic concept of optimization by strength of the structure subjected to regular conventional loadings, a good knowledge in advance of the approximate levels of welding residual stresses could be helpful to the Designer-Engineer in determining main parameters of the size of ship structure members with regard to the minimum structural weight and the high buckling strength of thin plate panels.
5.4 Optimization at Fabrication Stage
By using the procedure, the Constructor-Engineer could choose the appropriate parameters of the welding process in the required conditions to get higher quality of fabrication and higher productivity.
5.5 Accuracy in Fabrication Chain
High-Technology today needs more accuracy in automation production chain by use of robot welding processes. The proposed procedure may enable shipbuilder to keep lower the tolerance within the Quality Standard limits of welded structures realized by the mechanized block assembly.
6. CONCLUSION
In this paper, an example was presented suggesting to optimize a type of thin stiffened plate panel in shipbuilding.
As the principles are available for all thermal weldings, the multipurpose usefulness of the method may be applied to arc-welding as well as to laser beam welding for further applications.
So, in standard and high-tech shipyards, the method should be considered as an useful numerical approach tool that would be helpful for shipbuilder to evaluate, to verify and to control the distortion levels of thin welded assemblies before taking himself his judicious decisions in the choice of several design and fabrication parameters for any new shipbuilding project.
Furthermore, the global performance of the structure mainly depends on the local behaviours of the elementary plates of welded structure. Thus, the harmful effects of these local imperfections have a great influence on the strength of ship structure such as Ultimate strength, static Fracture, Fatigue behaviour.... and so on the Life-Cycle Service of the Ship.
Henceforward, for these problems, much more further investigations in analytical studies and experiments could be thoroughly developed taking into account the calculated values of the welding residual stresses and deflections obtained by using the IRCN numerical method for evaluating the welding effects on such any welded ship structure assembly.
ACKNOWLEDGMENT
Homage to Shipbuilders for their courage in attempting to overcome the welding effects.
Thanks to the French Ministry of Industry and Research and to French Shipyards (C.A. Chantiers de l'Atlantique and ex-ACH Ateliers et Chantiers du Havre).
REFERENCES
[1] Navy Joining Center NJC operated by Edison Welding Institute EWI, Welding Journal (August 1995), p99.
[2] Chau, T.T. and Masson, J.C., Association Technique Maritime et Aeronautique (ATMA), Vol.92 (1992), pp 207-233, 19-21 rue du Colonel Pierre Avia, 75015 Paris, France - Fax: (33) 1 40 93 57 72.
[3] Chau, T.T., Paradis, A. and Masson, J.C., 3rd International Conference on Marine Technology, ODRA'99, (11-13 October 1999), Szczecin, Poland, pp485-530, WIT Press. E-mail: witpress@witpress.com.
[4] Leblond, J.B. et al, J. Mech.Phys.Solids, Vol.34, n°4 (1986), pp395-409.
[5] Welding Metallurgy, Chapter 4, American Welding Handbook, Vol.1 (1987), pp90-124, AWS.
[6] Adams, C.M. Jr, Welding Journal Research Supplement (1958), pp210-215.
[7] Jhaveri, P. et al., Welding Journal, 41 (1), (1962), pp12-16.
[8] Heat Flow in Welding, Chapter 3, American Welding Handbook, Vol.1 (1987), pp65-87, AWS.
[9] Adam, A. and Simonneau, J., Soudage et Techniques Connexes, (Jan-Feb. 1991), pp32-37.
[10] Chau, T.T., Final Report (complement), IRCN/SEA5249 (Nov.1995) - E-mail: ircn@ircn.asso.fr.
