Technical Insight: Residual Stress Measurement

Residual stresses are self-balanced stresses within a component or structure that are caused by incompatible internal strains. Residual stresses can be generated or modified at any point in a component’s lifecycle, but welding is one of the most significant causes of residual stresses.

Welding typically causes large tensile stresses on the weld, balanced by lower compressive residual stresses elsewhere. The compressive stresses are generally beneficial, although they can decrease the buckling load. Tensile residual stresses, by comparison, can reduce the performance or cause creep failure in manufactured parts. They can lead to an increase in the rate of damage caused by creep, environmental degradation and fatigue. Tensile stresses may also reduce the load capacity by contributing to failure by brittle fracture, or cause other forms of damage such as shape change or crazing.

The implications of residual stresses has led to TWI undertaking a great deal of work related to their measurement over the decades, including original research projects to better understand the impact of residual stresses and how to mitigate against them, and investigations into specific cases of failure that included residual stress as a concern.

Core Research Programme (CRP)

Through the core research programme (CRP), TWI conducts projects to address challenges faced by our Industrial Members, advancing technologies and processes, and finding new solutions for industry problems. These projects are jointly funded by our Industrial Members, with the outcomes made available to all our Members.

- Residual Stress and Fatigue Strength of Slag Defect Welds

Dating back to 1977, this CRP project aimed to determine the influence of residual stress on the fatigue behaviour of manual metal arc butt welded joints that contained buried slag-line defects. Samples were created and tested under axial pulsating tension loading to establish S-N relationships and determine the influence of residual stress on the fatigue behaviour of the defective joints. The research work indicated that residual stress had no effect upon the fatigue strength of slag-bearing butt welds. The test results were found to be consistent with comparable data obtained from other sources, and were quantified using a proposed method of determining acceptable defect sizes for fatigue service.

- Residual Stresses and Fatigue of Fillet Welded Steel Joints

This 1980 project aimed to investigate the effect of tensile residual stresses on the fatigue behaviour of fillet welded joints in steel under different load ratios and the relevance of the tensile strength of the steel. To achieve this, fatigue tests were conducted on fillet welded joints in four steels under different applied load ratios. While some specimens were stress-relieved, the remainder were spot-heated to ensure that high tensile residual stresses, as would be present in welds in real structures, were present in the specimens.

- Residual Stresses in 25mm Thick Weld Metal COD Specimens

This 1981 CRP project made through-thickness and surface measurements of residual stresses transverse to the welding direction in 25mm thick crack opening displacement (COD) specimens in the as-welded and locally compressed states.

- Heat Treatment and Mechanical Properties/Residual Stress

By 1984 TWI’s experts had turned their attention to the effects of post-weld heat treatment (PWHT) on the mechanical properties and residual stress levels of steels. We undertook a study of the effects of temperature, time and repeated PWHT cycles on the strength, toughness and residual stress levels of a continuously cast C-Mn-Nb-Al steel. The study determined that temperature had a more significant effect on properties than time across all parameters. Degradation of properties with increasing time and temperature was shown to be more rapid in the HAZ at the plate mid-thickness than elsewhere. However, in general, excellent HAZ toughness results were obtained with only occasional low values.

- Submerged Arc Weld Residual Stress During Postweld Heating

Continuing the investigation into post-weld heat treatment (PWHT) on residual stresses, this 1985 project studied submerged arc welds in a 50mm thick C-Mn-Nb-Al parent steel. Our experts measured the surface and internal residual stresses in the welds after the application of PWHT at different temperatures and hold times as well as for repeated PWHT cycles. Isothermal and anisothermal stress relaxation test procedures were used to create supporting measurements of the relaxation properties of all weld metal tensile specimens. These stress relaxation tests showed that the weld metal was more resistant to stress relief and that its relaxation properties were more variable than for typical C-Mn parent steels. The maximum longitudinal residual stresses were up to 50N/mm² greater than the upper bound of the weld metal stress relaxation data for corresponding PWHT procedures, whilst the maximum transverse residual stresses were fairly low and were insensitive to the heat treatment conditions. The results suggested that residual stresses parallel to the welding direction after PWHT may be significantly greater than those predicted to occur using the accepted methods of the time.

- Repair Weld Residual Stress Measurements in Pressure Vessels

This 1986 CRP project measured surface and internal residual stresses at repair welds in C-Mn-Si-Al, C-Mn-Si and Mn-Ni-Mo steel test panels with thicknesses ranging from 50-105mm. The repair welds were deposited by manual metal arc and pulsed metal inert gas processes using a two-layer technique to control the HAZ microstructure where the repair cavities were located. The test panels, which were restrained by strongbacks during repair welding and residual stress measurement, were not subject to post-weld heat treatment.

- Measuring Residual Stress and Effects of Prior Deformation

In 1987, TWI’s experts investigated the use of the neutron diffraction technique to non-destructively measure residual stresses. Earlier residual stress studies using the neutron diffraction technique concentrated on establishing the validity of the method by comparison with established mechanical techniques. This later study extended the earlier research to consider both the residual field and the prior plastic deformation history of an aluminium alloy and A533B steel. The neutron diffraction method was also used to obtain the three-dimensional residual stress distribution in an aluminium alloy weld before a destructive 'layering' method was applied to the weld to provide a means of direct comparison.

- Residual Stresses in MIG Welded Aluminium Alloy Panels

Also in 1987, we produced a report related to a programme of surface and internal residual stress measurements in MIG welded, 10 and 50mm thick aluminium alloy panels. For this work we used the following parent/weld metal combinations: 5456-H116 with 5556 welding wire, 6061-T651 with 5356 and 4043 wire, 5083-0 with 5183 and 2219-T87 with 2319 wire. The residual stress measurement methods used were centre-hole rosette gauges, sectioning and neutron diffraction. The measured residual stresses was used in the assessment of wide plate fracture test specimens extracted from nominally identical test panels.

- Relief/Prediction of Residual Stresses in 2 1/4 Cr-1Mo Welds

This project, completed in 1989, sought to determine the effects of hold temperature, heating and cooling rates, hold time and external restraint on the relief of residual stress in 2¼Cr-1Mo steel weldments. Our project team also investigated the relationship between stress relief in weldments and stress relaxation in test specimens subject to corresponding heat treatment conditions.

- Residual Stresses in Prior Loaded Cracked Specimens

This 1991 CRP project involved the measurement and prediction of residual stresses in a preloaded, cracked bend specimen made from A533B steel. The preloading or overstressing of flawed components can induce residual stresses near the tips of flaws. These stresses can be beneficial to the structural integrity of the components if they are compressive. Our team used the non-contact, neutron diffraction method to obtain the experimental results while numerical predictions were also obtained via finite element analysis. Although the results from the various techniques were compared there was no initial assumption as to which technique was deemed more ‘correct.’

- Fusion Welding Models to Predict Residual Stress/Distortion

Also in 1991, this project aimed to address fragmentation in welding process modelling at the time, whereby arc physics and weld pool dynamics were not being modelled in conjunction with simulations to predict residual stresses and distortion. This work mainly considered computer numerical techniques for the modelling and prediction of residual stresses and distortion, detailing different welding processes with an assessment of accuracy.

- Residual Stresses in Double Tension Crack Arrest Specimens

This 1992 report described the results of surface and bulk residual stress measurements in a remelted crack starter section and the adjacent region of test material in an embrittled electron beam melt run introduced at the edge of a plate.

- Mean/Residual Stress and Welded Joint Fatigue Strength

At the time of this 1992 project a substantial number of fatigue tests had been performed on as-welded joints under variable amplitude loading. However, few comparable tests had been carried out on stress relieved joints. This work not only obtained directly comparable results for as-welded and stress relieved joints for stress ratios R = 0 and R = 1 but also made comparative tests on as-welded joints both at R = 0 and under stalactitic loading. In order to enable a relatively high peak stress to be used, these specimens were fabricated from high tensile steel (RQT 700).

- Ageing Properties/Residual Stresses in Thermoplastic Welds

This 1996 CRP project saw our experts turn their attention to polymeric materials. These materials are susceptible to degradation during processing or in service, which can lead to a deterioration in properties and premature failure. The long-term performance of polymers can be further impacted by welding where the viscous polymer is subjected to elevated temperatures and considerable mechanical forces that can generate thermal and residual stresses and initiate degradation reactions, thus affecting the long-term performance of the welded components. At the time of this project there was no established technique available for the measuring of residual stresses in thermoplastic welds so, to close this gap, TWI developed a technique to measure residual stresses at welds in polymers. We were then able to generate data on residual stresses at welds in polyetheretherketone (PEEK), polyethersulphone (PES), polyetherimide (PEI), acrylonitrile-butadiene-styrene (ABS) and glass- reinforced (GR) polyester. Our materials experts also worked to determine the effect of ageing on residual stresses, mechanical properties and toughness in PEEK.

- Yield Strength Mismatch and Welding Residual Stresses

PD 6493: 1991 provided recommendations for dealing with the welding residual stresses and tensile property variations across the joint during structural integrity assessments. By 1998 the efficacy of these recommendations had been assessed by a number of publications. Our experts undertook a literature review in relation to the effect of welding residual stresses and weld mismatch on flaw assessments before commenting on how this published work affected PD 6493: 1991.

- Branching Out - Local Post-Weld Heat Treatment

This 2003 CRP project addressed a demand from industry for a reduction in the size of theheat-treated zone (HAZ), either to protect existing connections and heat sensitive attachments or to reduce energy consumption. This project used finite element analysis to investigate the effect of patch heat treatment on residual stresses. Our experts determined the magnitude of residual stresses resulting from local patch PWHT and compared them with those stresses arising from circumferential band PWHT. TWI also determined suitable procedures for local patch PWHT at branch connections and compared results between elastic analysis and elastic-plastic and relaxation analysis.

- Residual Stresses in Steel Girth and Butt Welds

By the time of this 2010 CRP project, it was known that welding residual stresses could reach the yield strength of the surrounding material. This often led engineering critical assessments (ECAs) to assume that the residual stresses are of yield magnitude. However, while residual stresses parallel to the welding direction can be very high, the values transverse to the weld are often much lower. In addition, it had been demonstrated that welding residual stresses vary with plate thickness, which created a need for methods to be developed for determining residual stresses through the plate thickness as well as on the plate surfaces. Such a determination would reduce the conservatism in ECAs arising from the assumption of a residual strength of yield strength and thereby create a more realistic evaluation of crack driving forces. Our experts developed finite element analysis (FEA) procedures for modelling welding residual stresses that were then validated against measurements. We also examined the effects on welding residual stress of geometric variations, restraints, and external static and cyclic loadings, and evaluated fracture mechanics parameters (K and J) characterising a crack in a welding residual stress field.

- Residual Stress in Clad Pipe Girth Welds

Fitness-for-purpose procedures such as BS 7910 (BSI, 2007), R6 (EDF, 2011) and API 579-1/ASME FFS-1 (API, 2007) presented residual stress profiles for use in engineering critical assessments (ECAs). However, review work on cladding residual stresses showed that none of the residual stress profiles given in the assessment procedures was developed for clad pipes. TWI created a CRP project in 2012 to measure the welding residual stresses in clad pipe girth welds using various measurement techniques before developing and validating finite element analysis (FEA) models for simulation of welding residual stresses in clad pipe girth welds. Our teams also assessed the effects of heat input, material hardening models, yield strength, and weld start/stop on welding residual stresses. This allowed TWI to provide a recommendation for the consideration of residual stresses in clad pipe girth welds.

- Post Weld Heat Treatment Effect on Welding Residual Stresses

This 2016 CRP project sought to assess how post weld heat treatment (PWHT) effects welding residual stress by investigating a number of standards including BS 7910, R6 and PD 5500 with regards to both local and furnace PWHT. This research uncovered variations in the codes and a lack of clarity in relation to some recommendations within the codes. Because the level of residual stress depends on a wide range of varying factors such as the temperature attained, the width of the heated band and the width of the insulated region, no general recommendations were and it was advised that expert advice is sought, otherwise conservative assumptions should be made.

- Effects of Mechanical Loading on Residual Stress and Fracture: A Re-examination of the BS 7910 Rules

Also in 2016, this CRP project addressed advice on the assumptions about the magnitude and distribution of welding residual stress in as-welded joints according to BS 7910 at the time, which had not been amended since 2005 and whose rules dated back to PD6493:1991, one of the forerunner documents to BS 7910. At the time of the project, R6 had warned of the ‘limited validation’ of BS 7910 approaches for stress relaxation, suggesting they should be used ‘with caution.’ TWI’s experts re-examined previous research that was undertaken around the time of the 2005 amendments and found nothing to contradict the recommended approach.

- Residual Stress/Mechanical Properties of Ti6Al-4V Parts

Selective laser melting (SLM) was the subject of 2017 CRP research to investigate the development of high internal (residual) stresses caused by the high temperature gradients that result from the build cycle’s heating and cooling process. The project team researched residual stress and consequential component distortion and performance restrictions in titanium alloy Ti6Al-4V gr23, which was extensively used within the aerospace and biomedical industry sectors and gaining increasing use for other general engineering applications. The aim of this project was to address industry concerns and thereby increase adoption of the SLM process.

- Improved Finite Element Predictions of Residual Stresses

This project investigated the use of finite element simulation to predict stress evolution during welding. This work was done to address the challenges associated with measuring residual stresses to help ensure welded structures are able to resist certain failure mechanisms. Our experts produced a report incorporating modelling approaches using finite element modelling to predict microstructure and residual stress in a ferritic benchmark weld. These models were compared with published measurements for the specimen.

- Biaxiality and Residual Stress Relief/Fracture in Joints

Also taking place in 2017, this CRP project assessed TWI’s database of full-scale fracture tests for wide plate test results carried out on well characterised welded joints with and without the application of a warm prestress (WPS). The objectives of this work included to demonstrate the effect of WPS on residual stress under different biaxiality conditions, to calculate the effect of biaxiality on failure in terms of fracture assessment procedures, analyse the effect of biaxiality on the relief of welding residual stresses during WPS, and compare these results with experimental measurements, as well as demonstrating the effect of using different reference stress/limit load solutions on failure assessment.

- Warm Prestress and Residual Stress/Fracture in Welded Joints

Warm prestress (WPS) was also the subject of this 2017 project, which delved into the TWI fracture database for examples of wide plate tests on both plain steel plates and welded joints, with and without application of a warm prestress (WPS). Since these tests were conducted, the R and BS 7910 fracture assessment procedures had been developed, which provided an opportunity to validate both procedures for the treatment of welding residual stress, the enhancement of fracture toughness due to WPS, the effect of weld strength overmatch, and the treatment of plastic collapse.

- Residual Stress in Girth Weld Pipe After Plastic Deformation

Also in 2017, TWI investigated residual stresses in girth weld pipe after plastic deformation with, for example, pipe reeling operations and other high applied strains. This research provided information to improve the accuracy of fitness for service assessments, potentially improving safety and reducing overdesign.

- Validation of Process Models for Additive Manufacturing

Working alongside the SIMULIA brand of Dassault Systèmes, TWI aimed to validate finite element modelling (FEM) techniques to accurately simulate additive manufacturing processes to provide improved part design and process set-up. We produced double cantilever parts (Figure 1) from Ti-6Al-4V Grade 23 metal powder with a powder particle size range between 15 and 45μm, using a 90° alternating scan strategy comprising a series of parallel hatch lines and four boundary contours, rotated by 90° every layer. FEM formulations showed temperature-dependent material properties for the heat transfer and stress analysis simulations, including the residual stresses after the wire cutting (Figure 2). The FEM predictions were compared with measured deflections from the test pieces (Figure 3), showing a strong agreement between the predictions and measurements, and leading to confidence in the use of this new modelling approach.

- Residual Stress for Integrity of Steel Structures

This 2019 report provided a comprehensive overview of practices to assess and treat residual stress for the integrity of welded steel structures. Findings determined that there was no single method capable of providing a full range of measurement capabilities, however there were several methods available to reduce welding residual stress, including post-weld heat treatments (PWHT). Finite element analysis was shown to be capable of predicting residual stress, although there was limited experimental data available on the relaxation of residual stress due to thermal or mechanical loading. Tensile residual stress is detrimental to the fatigue life of a structure so that modifying tensile residual stresses to compressive residual stresses will improve fatigue life of a welded joint.

- Treatment of Residual Stresses Considering Elastic Follow-Up

Structural integrity assessment codes such as API 579, BS 7910 and R6 all classified residual stresses contributing to plastic collapse of a structure as primary stresses and those that do not contribute to plastic collapse as secondary stresses. In these codes, residual stresses are usually treated as secondary stresses, but it is noted that long-range residual stresses, such as the uniform tensile stress introduced into a pipe due to the boundary conditions, are associated with significant “elastic follow-up” and therefore may be classified as primary stresses. Elastic follow-up is the follow-up elasticity which tends to increase the permanent plasticity at a localised region of an engineering component, but there was no detailed guidance defining what level of elastic follow-up was deemed to be significant at the time of this project, making it unclear whether welding residual stresses or long-range residual stresses should be considered to be primary, secondary or somewhere between the two. Our experts applied theoretical models and finite element analysis (FEA) to quantify elastic follow-up in three-bar structure models containing fit-up residual stresses and a plate model containing welding residual stresses. This allowed the influence of elastic follow-up on the redistribution of welding residual stresses to be examined and the implications of residual stress treatment in the context of fracture assessment were explored (Figure 4).

- Study of Use of Ultrasonic Waves for the Detection of Residual Stresses

Although a number of methods had been devised and validated for the measurement of residual stress, such as contour mapping, hole drilling and X-ray diffraction, they are often costly, time-consuming and require a lab to perform. This created a need for a portable method to characterise residual stresses non-destructively. This 2024 project investigated the application of ultrasonic critically refracted longitudinal waves (LCR) for the measurement of residual stresses within a component, with the primary objective of developing process parameters and manufacturing a defect-free, transparent, diffusion bonded micro-channelled thermoplastic demonstrator.

- Integrating Probabilistic Approach into British Defect Assessment Procedure: Application of Probabilistic Treatment of As-Welded and PWHT Residual Stresses in BS 7910 Context

This 2024 CRP project addressed the lack of guidance on the treatment of welding residual stress in single-pass narrow gap joints, such as electron beam welds, at the time. As well as providing more structured guidance for the assessment of residual stress in EB welds, this project explored the possibility of treating welding residual stress probabilistically by extending the approach in NSIRC Project 32688 (Figures 5-7). TWI’s experts also delivered a thorough statistical analysis of residual stress data determined from joints subject to post weld heat treatment (PWHT) in order to strengthen the BS 7910 rules.

Joint Industry Projects

Our joint industry projects (JIPs) allow interested parties to come together as project sponsors, pooling resources for greater research outcomes, while also being given exclusive access to the results and the opportunity to guide the direction of the projects themselves. These projects typically focus on challenges faced by specific industry sectors, providing industry-relevant information, guidance and solutions:

- Procedures for Improved Fatigue Life Assessment by Prediction of Fatigue Crack Growth in a Residual Stress Field

This 2009 JIP combined experimental and modelling activities to define an effective stress range to be applied when accounting for the evolution of residual stress with proof loading, fatigue load and crack growth. Our experts created guidance notes based on the use of finite element analysis to predict fatigue damage in a residual stress field as well as for the influence of material properties and loading history of fatigue assessment results. The aim was to allow for the life extension of engineering structures, increased safety and improved designs to save energy and materials.

- Fatigue Design Guidance for Risers in Sour Service via Full Scale Testing

Residual stress measurement formed part of this wider 2010 JIP focused on the fatigue design of steel catenary risers (SCRs) used within deepwater oil and gas developments. While resonance fatigue testing of full-scale girth welds had become standard industry practice to demonstrate acceptable levels of performance, these tests failed to take account of aggressive service environments, such as sour production fluids. Full scale fatigue tests were performed on welds in pipes containing a sour environment. The findings were investigated via metallurgical examination, residual stress measurements, fatigue crack growth rate measurements and fracture mechanics assessments. This allowed TWI to present validated guidance for the design of risers based on a full-scale sour service riser fatigue design curve.

Public-Funded Projects

Our expert teams are also called to participate in public funded projects in partnership with other organisations from industry and academia to solve specific challenges, typically for particular industry sectors:

- Residual Stress and Structural Integrity Studies using Thermography

Funded by Innovate UK, the RESIST project proposed a new method for evaluating residual stress based on material and system models combined with data from a full-field, non-contact, non-destructive measurement technique. This approach was particularly suited to large or expensive components, where material removal is undesirable and where contact is impossible.

- Optipress: Stepped Change in Precision and Strength Optimisation Capability for Steel Fabrications

The OPTIPRESS project was created to innovate fuel rail designs by connecting design tools for finite element analysis, distortion prediction, and fatigue design to maximise the potential for autonomous design optimisation. Modelling and software was also used as part of the innovation to predict and control the heat distortion created by welding, including residual stresses.

Dedicated Industrial Member Support and Other Projects

Much of the work undertaken at TWI is conducted confidentially on behalf of individual Industrial Member companies. However, there are some examples of these types of project work that we are able to share with you, as follows:

- Robert Jenkins Pressure Vessel

TWI investigated the failure of a 34m long pressure vessel consisting of two parts joined by a flanged and bolted connection (Figure 8). The failure occurred in the second section of the vessel, which was 2.7min diameter with a wall thickness of 28mm. The vessel was constructed according to ASME Section VIII, with the failed section being in an as-welded condition. The failure was found to have initiated in the vicinity of a fillet weld of a manhole compensating plate on the main shell. The brittle fracture extended in two directions through the strake containing the manhole into the adjacent strake on one side and the dished end on the other before arresting. The overall length of the fracture was 4.3m with a subsidiary crack of 1.5m long present on the other side of the manhole. The initiating defect was thought to have been caused by hydrogen cracking in the hear affected zone (HAZ) of the fillet weld. Our experts undertook a hardness survey of the area of the toe of the weld as well as tests to assess the chemical composition of the initiation region of the failed plate, which found it was susceptible to liquid cracking. The investigation postulated that the initiating defect may have formed as a liquation crack, which extended by hydrogen cracking under the conditions of high restraint and residual stresses present at the fillet weld toe. It was determined that the toughness of the parent plate was low and it was susceptible to liquation cracking and the formation of high hardness zones in the HAZ, leading to probable hydrogen cracking.

- John Thompson Pressure Vessel

In December 1965, a large pressure vessel being manufactured by John Thompson (Wolverhampton) Limited for the ICI Immingham plant fractured during a hydraulic test. The damage was substantial, with four large pieces being thrown from the vessel, one of which – weighing around 2 tonnes – was ejected through the workshop wall, landing around 46m away. The failure was found to have occurred at the flange end of the vessel, with the flange forging being cracked through in two locations. The first two shell strakes on the structure broke into several pieces and cracking extended into the third strake. Pre-existing cracks were located in the HAZ of the submerged arc weld join between the flange end forging and the vessel shell. Inadequate heat treatment meant that the forging and shell plate were unable to arrest a running crack of a size equal to the weld cross section. Residual stresses were also considered a contributory factor to crack initiation at the relatively low applied stress level as the heat treatment conditions had not been sufficient for full relief of the residual stresses.

- Union Oil Amine Absorber Tower

An explosion st the Union Oil Co. refinery near Lemont, Illinois, USA took place on 23 July 1984, killing seventeen employees and causing over $100 million in property damage (Figure 9). The explosion was caused by the ignition of a large cloud of flammable gas (a mixture of propane and butane) that had leaked from a ruptured amine-absorber pressure vessel. An operator had noticed gas escaping from a 150mm horizontal crack near the bottom of the vessel and tried to close the main inlet valve. When the crack reached 600mm, the operator called for an evacuation of the area. As firefighters reached the vessel, the tower cracked further and a large quantity of gas was released. This gas ignited in an explosion that sent the upper part of the tower into the air, landing over a kilometre away. The explosion was felt over 20 kilometres away and the blaze which followed sent flames 150m into the sky. Hydrogen blisters had been found in the lower two courses of the shell along with laminations in the steel soon after it had been brought into service in 1970. The hydrogen blistering continued before the second course of the tower was replaced using manual metal arc welding with no preheat or post-weld heat treatment. In 1976 a Monel liner to reduce corrosion was fitted in the bottom head and first course of the tower but it did not cover the repair section.

The failure investigation determined that the tower had fractured at the circumferential weld between the replacement ring and the lower course, with four large cracks having been present in the HAZ prior to failure. These cracks had initiated via hydrogen cracking and then progressed by hydrogen-induced stepwise cracking (HISC).

It was concluded that the failure was a result of the welding procedure used when replacing a section of the vessel, which led to the formation of a hard microstructure in the weld’s HAZ. This hard section was susceptible to hydrogen assisted cracking resulting in growth of large cracks in the vessel. The uncracked material in the vicinity of the existing cracks had low toughness due to hydrogen embrittlement and failed at the applied CTOD in the vessel because of the operating pressure and residual stresses associated with the weld.

This failure showed the importance of controlling weld properties for operation in corrosive conditions to avoid the formation of high hardness microstructures, although welding residual stresses also made a significant contribution to the applied CTOD.

- Ashland Storage Tank

In January 1988, a 16000m ³ capacity fuel oil storage tank at Floreffe, Pennsylvania, USA suffered a brittle fracture failure leading to major environmental pollution. Having originally been built at Whiskey Island in Ohio during the 1930s-1940s, the tank was taken down by oxyacetylene cutting adjacent to the original welds in 1986, and then reassembled by welding in Floreffe, keeping the plates in the same order. As the tank was being filled to capacity for the first time since it was moved, the tank shell parted from the bottom plate at the connecting welds and, under the force of the escaping oil, moved sideways about 35m. The tank roof to shell joint remained sufficiently intact for the roof to move with the shell. The escaping oil flowed over the surrounding dykes damaging an adjacent tank and passed through storm sewers into the Monongahela River and then the Ohio River. The total spillage was estimated at 15.2 million litres, causing severe harm to the environment and affecting the drinking water supply. TWI experts examined the fracture faces in the tank and found the characteristic signs of brittle fracture (Figure 10). Metallographic studies, Charpy V notch tests, drop weight tests, engineering defect assessments and soil foundation analyses were all conducted, along with welding residual stress measurements. It was finally concluded that the failure was due to the material immediately surrounding the flaw being of particularly low toughness, with crack initiation occurring under the combined effect of hydrostatic and residual stresses. As the tank was operating below the NDT temperature of the shell plate, the crack emerging from the locally embrittled area could not be arrested.

- Residual Stress: Elastic Shakedown in Fillet Welded Plate

This project researched the elastic shakedown of welding residual stresses in fillet welds, which are common in offshore applications. This research built upon similar studies conducted on the effect of shakedown on the fatigue behaviour of welded plate. Elastic shakedown is characterised by plastic deformation during the first few load cycles followed by an elastic response associated with a limit called the shakedown limit, it is considered as one of the reasons for stress relaxation or redistribution. To address a gap in the knowledge, our experts used experimental testing to study the effect of elastic shakedown in the redistribution of pre-existing residual stress fields in a fillet weld plate, manufactured in line with ship design and welding procedures using DH36, a ship building steel (Figure 11). This project determined that the shakedown limit of fillet welded geometry can be estimated based on a simplified method using plastic work done as a shakedown criterion and also that experimental measurement of residual stress redistribution after three load cycles is able to show that there is only minimal redistribution / relaxation in the transverse residual stress component even though the load applied was along this component. Finally, based on the experimental evidence, our experts decided that the conservative level of the residual stress relaxation rule in BS 7910 may require re-investigation.

- NDT: Practical Quantification of Residual Stress Measurement

This project was created to develop and investigate the capability, accuracy and limitations of non-destructive ultrasonic techniques for the quantification and monitoring of residual stress in metallic materials and then develop a procedure for using the ultrasonic method to assess the residual stress of materials and welds on-site (Figure 12). This work included the generation of tables with acoustoelastic constants and calibration blocks for the most common materials, to be used in combination with the NDT procedure. Finally, TWI worked to use the method to validate weld repair regions, in addition to providing further data for use in validating the stress relaxation clauses stipulated in structural integrity standards.

- Study to Determine Residual Stress on a Pipe Spool

TWI worked with ISIS on a collaborative study to quantify the relaxation of residual stresses on a pipe spool subjected to 3% strain. The study, which involved work at both the ISIS neutron diffraction facility at the Rutherford Appleton Laboratory near Harwell, Oxford and at TWI’s headquarters near Cambridge, produced results expected to improve existing fracture integrity assessment by challenging the treatment of residual stresses at the time (Figures 13-14). This had particular benefits for the oil and gas industry, where pipe spools are used for hydrocarbon flowlines and pipelines, as well as for the nuclear industry. The results were interpreted by TWI and ISIS in relation to current models of residual stress relief by mechanical loading. The findings, which had relevance for the future treatment of fracture mechanics assessment in the oil and gas and nuclear sectors, were reported to the R6 Panel in November 2012.

- Functional Data Analysis for Residual Stress Modelling

This project sought to address methods of modelling residual stress that had an undefined level of conservatism and could not be adjusted for welding procedure, geometry, or level of conservatism. To solve this issue TWI studied two methods for transforming discrete data points into functional data, creating a statistical framework for analysing this data. The first method was an interpolation method based on cubic smoothing splines while the second method was an extrapolation method based on the self-equilibrium of residual stress. By applying these two methods to low heat input residual stress data, residual stress data after post-weld heat treatment, and electron beam residual stress data, we aimed to provide a clearer picture of residual stress modelling for industry, creating a more accurate probabilistic model with less inherent conservatism. The project investigated probabilistic modelling techniques for three residual stress databases; arc welds for low heat inputs, arc welds post-weld heat treatment, and electron beam welds. The probabilistic models were calculated and applied to all three sets of data and any model assumptions were tested for suitability (Figures 15-16).

You can find out more about our support for residual stress measurement across industry and how we can help you with your challenges, here:

https://www.twi-global.com/what-we-do/services-and-support/asset-management/engineering-critical-assessment/residual-stress

Figure 1. SLM double cantilevers after wire cutting

Figure 2. Image of residual stress after cutting

Figure 3. Comparison between experimental measurements and predictions of deflection

Figure 4. Three-bar assembly with varying cross sections in the central bar

Figure 5: The mean curve for low heat input residual stress along with the 95% one-sided confidence interval and the BS 7910 level 2 upper bound fit

Figure 6: The mean curve for PWHT residual stress along with the 95% one-sided confidence interval plotted with the BS 7910 (BSI, 2019) 20% YS rule

Figure 7: The mean curve for EB plate residual stress along with the 95% one-sided confidence interval plotted with the proposed upper bound profile by Hurrell et al (2014)