Residual stresses are internal self-equilibrating stresses that remain in the component even after the removal of external load. The aforementioned stress when superimposed by the operating stresses on the offshore piping, enhance the chances of fracture failure of the components. Thus, it is vital to accurately estimate the residual stresses in topside piping while performing their fitness for service (FFS) evaluation. In the present work, residual stress profiles of girth welded topside sections of P91 pipes piping are estimate using a machine learning approach. The training and testing data for machine learning is acquired from experimental measurements database by Veqter, UK. Twelve different machine learning algorithms, namely, multi-linear regression (MLR), Random Forest (RF), Gaussian process regression (GPR), support vector regression (SVR), Gradient boosting (GB) etc. have been trained and tested. In order to compare the accuracy of the algorithms, four metrics, namely, Root Mean Square Error (RMSE), Estimated Variance Score (EVS), Maximum Absolute Error (AAE), and Coefficient of Determination (R^2) are used. Gradient boosting algorithm gives the best prediction of the residual stress, which is then used to estimate the residual stress for the simulated input parameter space. In the future work authors shall utilize the residual stress predictions from Gradient boosting algorithm to train the Bayesian Network, which can then be used for estimating less conservative through-thickness residual stresses distribution over a wide range of pipe geometries (radius to thickness ratio) and welding parameters (based on heat input). Furthermore, besides topside piping, the proposed approach finds its potential applications in structural integrity assessment of offshore structures, and pressure equipment’s girth welds.

Girth welded pipes, such as those located offshore on platforms in the North Sea, are subjected to highly corrosive environment. The need to consider welding residual stresses in the assessment of the fitness for service and damages to these pipes when investigating local corrosion damages across a welded region is therefore important for the operators of the platforms and the manufacturers of the pipes. This paper presents a review of work carried out to ascertain the welding residual stresses present within a thin-walled girth welded pipe mock-up made from steel API 5LX Grade 52 before and after reduction of the wall thickness. The mock-up was manufactured to replicate typical pipes used to convey gas, oil and water through the platforms. The mock-up was of diameter 30” and of thickness 19mm. The incremental deep hole drilling (iDHD), contour, hole drilling, XRD, and ultrasonic technique were applied to characterise the residual stresses in the weld and heat affected zone of the specimen. The residual stresses were then measured during the manufacture of a groove located on the weld at the ID and were compared to an FE prediction. Ultrasonic measurements were then carried out on the outer surface of the pipe and show a significant increase in the residual stress and could be used to monitor the changes in the residual stress caused by internal corrosion.

This paper presents the residual stress measurements carried out on a t-section representative of a ring stiffened cylindrical structures. This paper presents the work carried out to ascertain the residual stresses present within a T-plate section representative of a ring stiffened cylindrical structures. The contour, the deep hole drilling (DHD) and the neutron diffraction (ND) methods were applied to determine the longitudinal component of residual stress in the weld toe of the fillet weld in the as-welded condition. The results of these measurements are presented and compared to highlight agreements and discrepancies in the measured residual stress distributions using these different techniques. Finally, non-destructive residual stress measurement using the ultrasonic (US) technique was carried out on the component. The ultrasonic measurement provides a relative measurement and usually requires a tensile test in order to determine the acoustoelastic constant and the time of flight in a stress-free state. The tensile test requires some material to be extracted from the component. The tensile test can be avoided if other residual stress measurement techniques are used for the calibration. After the calibration the US technique can be deployed on a full-scale ring stiffened cylindrical structures to detect abnormal variation in the residual stress field.

Optimising the structural integrity of an oil and gas pipeline is hugely important for its productivity and hence profitability. The structural integrity of a pipeline is influenced by factors such as: stress (i.e. applied and residual), material properties, environment, and the size and orientation of contained flaws. For example, whilst in operation, the integrity of a pipeline can be extended by reducing its applied stresses e.g. the flow and pressure of the oil and gas running within. Prior to operation however the integrity of the pipeline can easily be extended by reducing the residual stresses generated during installation or even “negatively pre-loading” the pipeline using residual stresses to help cancel out some of the applied stresses. Therefore understanding the distribution of residual stresses within a pipeline can be of great benefit to Oil and Gas engineers. In this paper, complementary residual stress measurement techniques are used to obtain near surface and through-thickness residual stress distributions in a fully circumferential butt welded pipe. The deep hole drilling (DHD) method was used to obtain the axial and hoop residual stresses along radial lines through the pipe wall. Near surface measurements on the outer surface of the pipe were obtained using the incremental centre-hole drilling (ICHD) method. The measurements were made only at limited points in and adjacent to the circumferential weld. In order to estimate the complete residual stress field present in the pipe, a mapping procedure utilising a finite element (FE) method was implemented. This entailed introducing the measured residual stresses into a FE model of the pipe as an initial condition and allowing redistribution. Naturally, the stresses at the measurement locations should remain at their initial values. Consequently, the method was developed to allow redistribution while retaining the measured values. The paper provides these estimates of the full residual stress state present in the pipe based on this mapping procedure. The FE model was then used to simulate the influence of various sizes of flaw on the mapped residual stresses field. An assessment of the acceptability of areas of loss of the wall thickness in internally pressurised pressure vessels was then performed.

Plastic deformation from peening induces a compressive residual stress on the treated surface which provides resistance to metal fatigue. Hence, peening is often used to extend the fatigue lives of welded components by reducing the effective tensile residual stress. This paper describes the influence of ultrasonic peening on the residual stress. For this study a four point plastically bent beam specimen, similar to the reeling process, was used. The specimen was made from 50D steel a material often used in offshore structural components. The residual stresses in the specimen were measured before ultrasonic peening with 5 different measurement techniques. After the ultrasonic peening treatment the residual stress was measured using the Incremental centre hole drilling and the ring core techniques. Measurements were carried within the homogeneously bent section location as well as outside. In order to see the variation of the results influenced by the gauge volume, strain gauges of three different sizes were used to provide results within 0.5mm, 1mm and 2mm depth. The measurements show that the ultrasonic peening surface treatment carried out induced high compressive residual stresses up to 2mm deep inside the specimen. Finally a compendium of residual stress profiles using different peening processes and materials is presented and discussed.

This paper presents results from a programme of through thickness residual stress measurements and finite element analysis (FEA) modelling carried out on a temper bead mock-up. Emphasis is placed on results comparison rather than the measurement technique and procedure, which is well documented in the accompanying references. Temper bead welding processes have been developed to simulate the tempering effect of post-weld heat treatment and are used to repair reactor pressure vessel components to alleviate the need for further heat-treatment. The Temper Bead Mock-up comprised of a rectangular block with dimension 960mm × 189mm × 124mm was manufactured from a ferritic steel forged block with an austenitic stainless steel buttering and a nickel alloy temper bead cladding. The temper bead and buttering surfaces were machined after welding. Biaxial residual stresses were measured at a number of locations using the standard Deep-Hole Drilling (DHD) and Incremental DHD (iDHD) techniques on the Temper Bead Mock-up and compared with FEA modelling results. An excellent correlation existed between the iDHD and the modelled results, and highlighted the need for the iDHD technique in order to account for plastic relaxation during the measurement process. Maximum tensile residual stresses through the thickness were observed near the austenitic stainless steel surface at 298MPa. High compressive stresses were observed within the ferritic base plate beneath the bimetallic interface between austenitic and ferritic steels with peak stresses of −377MPa in the longitudinal direction.

Published experimental data from residual stress measurements are generally limited or difficult to find. Experimental data in the offshore industry are even more scarce and difficult to access to. The oil and gas industry can benefit from research done in other industries. For example, the nuclear industry has published a multitude of residual stress measurements that could be beneficial to the offshore industry, gaining more understanding and confidence in the structural integrity of critical components. For the past year VEQTER Ltd has been developing a web access database for storing and comparing residual stress measurement and modelling results. This paper presents a comparison of through thickness residual stress measurement results that are published. The first case will show numerous measurements using different techniques on girth welded joints. Then, measurements on the temper bead welding processes which have been developed to simulate the tempering effect of post-weld heat treatment will be presented and compared with cladding. Finally, comparison of residual stress measurements on T-section plate fillet weld will be made.

Explicit understanding of the residual stress field of primary submarine pressure hull induced during fabrication will improve the fidelity of numerical analysis and experimentation. Hence, supporting operational envelope and design life extension initiatives. The fatigue lifetime of a submarine hull depends on the loads generated by hull contraction under the effect of hydrostatic pressure and the residual stresses existing in the absence of external loading. The use of numerical simulation allows a straightforward calculation of the stresses induced by the hydrostatic pressure. The effect of residual stress could be determined using the current failure assessment procedures, like BS7910 and R6. However it is more intricate to determine the residual stresses resulting from the sheet bending process combined with the sheet assembly using a multipass welding process. There are several measurement techniques available to measure residual stresses. They are often classified by their level of destructiveness and their penetration.In order to compare the different measurement techniques an elastic-plastic bent beam sample has been chosen as it is very comparable to the residual stress field induced during the sheet bending process used in the submarine structure. Four bent beams have been measured using five different techniques: Incremental centre hole drilling, ring core, neutron diffraction, slitting and deep hole drilling technique. The results from measurement techniques show an excellent agreement when compared with the FEA. In order to measure a full scale Rubis class submarine hull a limited number of techniques can be used, as the technique needs to be portable. The Deep Hole Drilling (DHD) technique was chosen because the neutron diffraction would require extracting a small test sample of about 400mm × 400mm, hence redistributing the residual stresses that were intended to be measured. Six measurements were carried out at different angular positions to detect variability in manufacture on a Rubis class submarine and a probabilistic calculation was done using all six DHD measurements. The Rubis class measurement results are also compared with two other submarine types, found in the literature. Understanding the three-dimensional behaviour of residual stress in this type of structure provides a valuable resource to the numerical modelling community. The results can also support fatigue and fracture experimental work and may help increasing the operating life of 28 year old French nuclear submarine.

AREVA has developed narrow gap weld techniques to perform junctions between low alloy steel heavy section components and austenitic stainless steel piping systems. In parallel, for a good understanding of welding and post weld heat treatment consequences, numerical welding simulation has already demonstrated its relevance to predict residual stress fields in welded components [1]. This paper presents Finite Element (FE) simulations of a 29″ multipass narrow gap Dissimilar Metal Weld (DMW) configuration, the welding simulation including non linear kinematic hardening models, phase transformations and visco-plastic calculations for reproducing the post weld heat treatment. The numerical results are compared to measurements obtained by the deep hole drilling technique [2]. This work gives another evidence of the relevance of the numerical welding simulation. Particularly, the comparison with a 14″ configuration [3] gives some elements to assess on the validity of both numerical and experimental techniques and on the weld thickness effect.

Prior knowledge of the magnitude and distribution of residual stresses in welded components is essential if a cost effective analyses of the integrity of the components is to be made. AREVA NP has recently developed, for EPR applications, narrow gap welding techniques, for joining ferritic low alloy steel heavy section components to austenitic, stainless steel piping systems, in nuclear reactors. An appraisal of available measurement methods was carried out and two residual stress measurement techniques were used to obtain through-thickness residual stress distributions in a fully circumferential narrow gap welded pipe, the neutron diffraction, which is not presented in this paper and the deep hole drilling (DHD) method. The DHD method was used to obtain the axial and hoop residual stresses along the weld centreline and on the heat affected zone in the ferritic and stainless steel sides up to depths of about 40mm from the outer surface of the pipe. The measured residual stress distribution in the weld centreline is compared with representative residual stress distribution provided in UK safety assessment procedures. It is found that the current safety assessment procedures BS 7910:2005 and R6 are conservative. The DHD measurements were made only at limited points in and adjacent to the circumferential weld. In order to estimate the complete residual stress distribution present in the pipe, a measurement mapping procedure using finite element (FE) analysis was implemented. Therefore this paper also provides the estimates of the full residual stress state present in the pipe based on the mapping procedure.