Long-term behaviour of welded structures at high temperature

Advances in technology often require lighter structures and higher operating loads and temperatures. For example, it is important to keep the weight of an aero engine to a minimum and the temperature and pressure in a steam power plant to a maximum in order to increase efficiency, resulting in power, cost and environmental benefits in terms of reduced emissions. Therefore, new advanced materials are necessary and these require assessment before use, which presents new challenges. More accurate, novel life predictions can ensure that the maximum life is obtained before component replacement or repair is required, and they can reduce maintenance costs and increase operating flexibility. Currently, this is essential for the UK’s ageing power plants, to allow maximum use and ensure a safe and reliable energy supply before the next generation of plants are built.

It is important that the lifespan of a structure is considered during the design phase to ensure that the appropriate materials and operating stresses are chosen. Therefore, for accurate life predictions, the material behaviour needs to be well-defined for the service stresses, temperatures and time.

Repeated load cycles can cause exhaustion or ‘fatigue’ failure of materials, and at high temperature, materials may continue to strain or ‘creep’ under constant load. Creep, fatigue and combined creep-fatigue failure mechanisms govern the ultimate life of many structures. Welding can critically reduce the resistance to these sorts of failure (even if the weld contains no defects), since the process introduces residual stresses and distortions, and causes detrimental changes in the material microstructure. Life predictions of welded structures are particularly difficult to deduce with sufficient accuracy, since the effects mentioned above cannot be easily quantified and incorporated in lifing models, nor are they evenly distributed throughout a weldment.

The authors have conducted extensive research in order to achieve accurate life predictions of both new and existing structures, particularly those containing welds, operating at high temperature and stresses for long periods of time (such as aero engines and power generation plants); this work is ongoing and past, present and future work is described here. The methods used aim to cover the entire life of a structure, including the manufacturing process, which can have a major influence on the lifespan. Finite element (numerical) welding simulations have been used to predict weld-induced residual stresses, distortions and microstructural changes. These models can help to design optimum welding procedures, bypassing the need for multiple expensive weld trials. The models can be extended to include heat treatment and machining operations with the results used as input data for life assessment. The most accurate life assessment models make use of the finite element method in conjunction with experimental testing, which is necessary to derive material behaviour parameters and to validate the lifing models. Small-scale models are developed first to compare with laboratory tests then full complex three-dimensional structures can be modelled to predict the ‘holistic’ lifespan of a real structure operating under realistic conditions.

The main result of the lead author’s work, completed thus far (during his PhD 2005-2009), has been to improve the understanding of the mechanical effects of welding, in order to mitigate distortions in test specimens. This in turn allowed more reliable high temperature material data to be obtained and applied to demonstrate more accurate life assessments for an aero engine welded shell structure.