Project Participants: General Dynamics Electric Boat
Project Start: November 2004
The costs required to effectively deal with weld distortion and/or correct the resulting negative effects are significant during submarine construction. Weld distortion impacts the fabrication costs of all welded parts, however, the larger the structure, the greater the impact on construction schedule. What are absent from the overall manufacturing process are computational tools that can accurately predict weld distortion without being biased by intuition. Software tools that can predict weld distortion have been identified, namely Virtual Fabrication TechnologyTM and SYSWELD®. This project evaluated and validated these tools’ abilities to predict weld distortion likely to occur during submarine pressure hull construction.
Electric Boat assembled a team of experts to tackle this difficult problem. EB’s team of Battelle Memorial Institute, ESI Group and Edison Welding Institute provided state-of-the-art software tools and material characterization and consultation.
The project consisted of three phases. Phase I verified that the existing software can correctly predict weld distortion for welded submarine structures. Although the selected analysis codes have been around for a number of years and used in other applications, they have not been used under the size, materials and welding process conditions of submarine hull construction.
Phase II of the project demonstrated that the verified method is valid for predictions of VIRGINIA-Class prototype hull circularity. During this phase, weld distortion analysis was performed on a model of an actual VIRGINIA-Class hull cylinder. This phase tested the ability of the software to efficiently analyze large models. Using the methods developed in Phase I, the results of Phase II analyses were compared to distortion measurements gathered during construction.
Thus far, the results have supported the value in these predictions, and that the associated cost avoidance to the VIRGINIA submarine program would exceed $4M during the first four years following full scale implementation. By integrating automation and other streamlining measures, the tool could be further developed to provide real-time feedback during the fabrication process.