To the depths of weld integrity . . . and beyond
  from: Offshore Engineer
  by: Dr Simon Rees
  Wednesday, March 19, 2008

Weld simulation techniques developed by Frazer-Nash Consultancy and taken from nuclear primary reactor components are now being applied in subsea and deepwater situations. The Frazer-Nash Consultancy’s Dr Simon Rees explains why.

With deeper and higher pressure operations, underwater components must not only be able to handle an increasingly extreme environment, but also need to have a longer life. Repair and replacement of these components is challenging enough in the shallows let alone in deepwater.

Ultimately improved designs and the accuracy of testing those components are becoming increasingly essential, and not unlike similar demands being placed on the nuclear industry. Unsurprisingly therefore, those improved designs, and techniques for high-integrity assessment, are being borrowed from the world of nuclear engineering, where the demonstration of ‘incredibility of failure’ has been a standard approach for many years.

A key to this whole issue is the demonstration by analysis that welded structures are able to survive the worst conceivable operating scenarios. Complex welds of thick structures invariably introduce residual stresses that have a material effect on the strength and life of the component. Historically, allowances for these stresses have been made by using crude empirical approximations, which have often led to inadequate designs.

Weld simulation techniques developed by Frazer-Nash Consultancy and taken from nuclear primary reactor components are now being applied to deepwater situations. Not only do such techniques accurately predict the residual stresses in the component – and hence the structural integrity – but they are also capable of determining the optimum welding strategy in order to examine repair weld options.

Cracking up

The tensile residual stresses induced in heavy-duty subsea components by welding inevitably reduce their service life.

Tensile residual stresses, combined with in-service loads and high temperature gradients, lead to reduced crack-initiation life, faster growth of existing or in-service defects, and considerably increase the susceptibility of a structure to catastrophic failure by fracture.

However, compressive residual stresses, if generated, can actually increase service life. However, predicting the generation of these stresses and characterising their effects is notoriously difficult. Traditionally, designers have relied on bounding residual stress profiles, such as those published in the R6 and API579 standards. However, these standards are extremely conservative and in some key areas do not even agree. The option of measuring the residual stresses is often not available for subsea equipment, and so Frazer-Nash Consultancy has been studying the use of finite element techniques originally developed for highintegrity applications in the nuclear industry.

There are many similarities between the nuclear and subsea industries. In both cases large steel structures are subject to extremes of pressure and temperature gradient, specialist materials are used, and failure is to be avoided at all costs. The successful application of these techniques to subsea situations has been built on research undertaken for the nuclear world, and from which the offshore industry can now benefit.

Hot stuff

To understand how weld simulation came about, and why it is driving forward improvements for the oil and gas industry, a quick science lesson may help.

Residual stresses are originally induced by the intense heat generated by the welding process and the constraints surrounding the weld. In order to adequately model the stresses it is first necessary to accurately model the temperature history of the weld.

As the weld is built up from individual passes – and in complex welds there can be many dozens – the moving torch deposits material and heats the surrounding structure as it travels along the line of the weld. There are also transient effects associated with each end of the pass, and the torch lingers briefly before commencing or ending that pass.

The heat transfer mechanism from the torch to the surrounding material is not easily modelled. At the high temperatures encountered conduction and radiation effects dominate, and the properties of the materials themselves become highly nonlinear. As part of the research work that led to the development of this technique – and which lasted over ten years – a heat source modelling tool (HSMT) was perfected. The HSMT used the best available experimental data to refine the material and heat transfer models, using measured values for the fusion boundary location – where the surrounding material melts into the deposited weld material – to calibrate results. After repeated validation studies, the tool reached the point where it could repeatedly and correctly predict the location of the fusion boundary.

The HSMT itself was a research tool used to refine our understanding of the physics of welding. Once it represented a reliable model, it was used to generate appropriate parameters for the thermal modelling element of the transient finite element analysis (FEA) used to predict the residual stresses.

Under pressure

The FEA models of the welds are built up by representing each pass as a separate line of elements within the model. These are activated in turn as the moving torch is modelled, with the appropriate thermal terms applied to replicate the temperature history predicted by the HSMT.

The elements themselves are not conventional. Not only do they have to cope with high levels of plastic deformation, the weld passes are restrained by the surrounding material, and so a hydrostatic pressure results in increased stresses.

The other unconventional aspect of the mechanical analysis is the treatment of the materials. At these elevated temperatures, all mechanical properties become temperature dependent, and these variations must be considered all the way up to melting point.

At the end of the process, a complete 3D map of the residual stress field in the component is produced. This can be used either to modify the welding and post-weld treatment to ensure a better, more compressive, stress map, or to assess the life of the component when subjected to aggressive environmental or chemical conditions.

Back to business

This technique has already found a number of applications in the subsea world. For example, it has been used to help assess the integrity and life of reeled pipe systems. Residual stress predictions were generated for the pipe girth welds in an X-series steel, which were in-turn combined with the stresses induced by the reeling processes to establish whether there was a relaxation in the residual stress levels. In another case, different weld layouts were compared to optimise the welding strategy on an underwater structure.

In all cases, the better understanding given by the finite element analysis has led to a product with a longer in-service life than those designed using conventional methods. Because of the high cost of manufacturing and installing subsea components, and the difficulty in undertaking any subsequent repair and life extension, the numerical method invariably leads to cost savings.

There is still more research to be done. There are many different welding processes, only some of which have been modelled, and a proliferation of modern, high performance steels whose material properties are still not fully explored. However, these techniques offer designers and operators the chance to extend equipment lives considerably, with all the cost savings that result. OE

Click here to register to receive your own copy of Offshore Engineer each month.