An algorithm to estimate the cooling rate of welding seamson the shell plating of a ship, below the waterline, while it is onvoyage has been derived. The demand for this technique hasarisen from the wish of ship operators to make it possible forthe safe repair of ship structures without taking them out ofoperation. [1] The strength of the shell plating after welding isdetermined by its metallurgic structure, which is dependent onthe cooling rate, its chemical composition and the original grainsize of the base material. [2] The cooling rate for this type ofwelding seam depends on the velocity of the water flow, thedistance from the bow, the thickness of the plate, and the heatfrom the heat input of the welding. The algorithm makes itpossible to calculate the cooling rate for a base material affectedby a forced flow of fluid by means of Rosenthal’s equation andthus enabling suitable welding parameters to be determined.As the welding parameters can be chosen to fit the specificrepair to be made, it is now possible to determine the suitabilityof a welding procedure in advance. The algorithm is applicablewhen determining welding parameters at Hot-Tappingoperations as well, where the base material is affected by aforced flow of fluid. A number of experiments have beenperformed and the results support the theoretical model. Theresearch project continues with the aim of finding an algorithmto include the enhanced cooling rate due to the layer of boilingfluid on the back of the base material. A method to improve themeasurements of the most important parameter in the algorithmhas been developed and makes it possible to build up aquantitative database of typical values for various configurations.

The strength of a weld joint is determined by its geometry and its metallurgic structure, which is dependent on the cooling rate, its chemical composition and the original grain size of the base material. During in-service welding of structures affected by a forced flow of fluid on its reversed side the cooling rate depends on the fluid’s boundary layer, the material’s thickness and the heat input of the welding process. Currently, the calculation of the cooling rate during in-service welding is made by means of numerical methods such as the Finite Element Method, FEM. Through the introduction of an apparent thermal conductivity, kPL, it possible to determine the cooling rate for specific welding parameters by means of Rosenthal’s equation. This can be done with a standard pocket calculator.An experimental rig for measurement of the heat transfer during the in-service welding of structures affected by a forced flow of fluid on its reversed side has been designed and built. The physical principles of welding on plates affected by a forced flow of fluid on their reverse side are the same as for welding on the circumference of a pipe containing a forced flow of fluid. In the rig, the required boundary layer is built up in a pipe system by means of a pump. As the flow and the temperature of the fluid can be controlled to simulate the specific heat transfer, it is now possible to verify the values of the apparent thermal conductivity, kPL, that were calculated

values of the apparent thermal conductivity, k_PL, for various configurations.For the purpose of evaluation and qualification of in-service Welding Procedures Specifications, WPS, the sponsors of the research project use the experimental rig.

An algorithm for heat transfer prediction of in-service welding operations in a forcedflow of fluid is presented. The algorithm presented is derived from Rosenthal’s 3D heatflow equation and boundary layer approximations. This was possible by the introductionof an apparent thermal conductivity kPL, which is a function of the boundary layer’s heattransfer coefficient f and the base material’s thickness . This implies that a weldcooling time tT1 /T2 in a forced flow of fluid can now be calculated by an ordinaryengineering calculator and thus enabling suitable welding parameters to be determined.The magnitude of kPLf , was established by regression analysis of results from aparametric finite element analysis series of a total number of 112 numerical simulations.Furthermore, the result of the regression analysis was validated and verified by a weldingexperiment series accomplished on an in-house designed and constructed in-servicewelding rig. The principle design of the welding rig as well as its instrumentation, a PCbased Data Acquisition system, is described. In addition, a method to measure the weldmetals cooling time tT1 /T2 by means of thermocouple elements is described. Finally,the algorithm presented in this study proved feasible for industrial in-service weldingoperations of fine-grained Carbon and Carbon–Manganese steels with a maximum CarbonEquivalent (IIW) (CE) of 0.32.