Taguchi optimization and ultrasonic measurement of residual stresses in the friction stir welding

The main goal of this study is optimization of residual stresses produced by friction stir welding (FSW) of 5086 aluminum plates. Taguchi method is employed as statistical design of experiment (DOE) to optimize welding parameters including feed rate, rotational speed, pin diameter and shoulder diameter. The optimization process depends on effect of the welding parameters on longitudinal residual stress, which is measured by employing ultrasonic technique. The ultrasonic measurement method is based on acoustoelasticity law, which describes the relation between acoustic waves and internal stresses of the material. In this study, the ultrasonic stress measurement is fulfilled by using longitudinal critically refracted (L_CR) waves which are longitudinal ultrasonic waves propagated parallel to the surface within an effective depth. The ultrasonic stress measurement results are also verified by employing the hole-drilling standard technique. By using statistical analysis of variance (ANOVA), it has been concluded that the most significant effect on the longitudinal residual stress peak is related to the feed rate while the pin and shoulder diameter have no dominant effect. The rotational speed variation leads to changing the welding heat input which affects on the residual stress considerably.     

Nondestructive Evaluation of Welding Residual Stresses in Austenitic Stainless Steel Plates

This article investigates the nondestructive capability of ultrasonic waves in residual stress evaluation of austenitic stainless steel plates (AISI 304L). Longitudinal critically refracted (L _CR ) waves are employed to measure the residual stresses. Measuring the acoustoelastic constant through the tensile test is eliminated on the main investigated sample to keep it intact. Another welded plate with the same welding specification, geometry, thickness, and the same material is used to extract tensile test samples. To find the acoustoelastic constant of the heat affected zone (HAZ), a metallographic investigation is done to produce microstructure similar to that of the HAZ in a tensile test sample. A finite element model of welding process, which is validated by hole-drilling method, is used to verify the ultrasonic results. The results show good agreement between finite element and ultrasonic stress measurements which is accomplished nondestructively.     

Nondestructive Evaluation of Welding Residual Stresses in Dissimilar Welded Pipes

This paper investigates the nondestructive capability of ultrasonic waves in residual stress evaluation of dissimilar welded pipes. Longitudinal critically refracted (L _CR ) waves are employed to measure the residual stresses in a pipe-pipe joint of stainless steel 304 and carbon steel A106. Measuring the acoustoelastic constant is usually accomplished through the tensile test which needs cutting the tested material to extract tensile test specimens. However, cutting the tested pipe to complement the nondestructive ultrasonic measurement was not considered here. Instead, a dissimilar welded plate with the same welding specification, joint geometry, thickness and the same dissimilar materials is used to extract tensile test samples. The measured acoustoelastic constant of the plate along with the measured time of flight of the L _CR wave on the pipe, are utilized for ultrasonic stress measurement. A finite element model of welding process validated by hole-drilling method is used to verify the ultrasonic results. The results show good agreement between finite element and ultrasonic measurements in the pipe measured without any destructive process.     

Comparison between contact and immersion ultrasonic method to evaluate welding residual stresses of dissimilar joints

This paper presents a comparison of contact and immersion waves used to measure residual stresses. The residual stresses are produced due to a dissimilar welding of stainless steel (304) to carbon steel (A106). Longitudinal critically refracted (L_CR) wave propagated by 2 MHz contact and immersion ultrasonic transducers is employed to measure the residual stresses. A Finite Element (FE) model of welding process, which is validated by hole-drilling method, is used to verify the ultrasonic results while an acceptable agreement is achieved. The best agreement is observed in the parent material while the maximum difference is measured in the heat affected zone (HAZ). The results show no considerable difference between using contact and immersion transducers in ultrasonic stress measurement of dissimilar joints.     

Ultrasonic inspection of a welded stainless steel pipe to evaluate residual stresses through thickness

This study investigates ultrasonic method in axial and hoop stress measurement through thickness of an austenitic stainless steel pipe. Longitudinal critically refracted (L_CR) waves are employed to measure the welding residual stresses while outer and inner surfaces of the pipe are inspected by using different frequency range of ultrasonic transducers. The acoustoelastic constant is measured on a plate with the same material and thickness of the investigated pipe to keep the pipe intact. Welding process of the pipe is simulated by a 3D finite element (FE) model which is validated by hole-drilling method performed on 25 points. The residual stresses calculated by FE simulation are then compared with those obtained from the ultrasonic measurement while a good agreement is observed. It is demonstrated that the residual stresses through thickness of the stainless steel pipe can be evaluated by combining FE and L_CR method (known as FEL_CR method).     

Grain size correction of welding residual stress measurement in a carbon steel plate using the critical refraction of longitudinal waves

In this, a method to measure welding residual stress in butt-welded joints of carbon steel plates using longitudinal critically refracted wave (L_cr wave) is proposed. Cross-correlation was employed to calculate the difference in time of flight between L_cr wave, and the optimal step length for the measurements is discussed. To determine L_cr wave acoustoelastic coefficient of the heat affected zone (HAZ), the relationship between the L_cr wave acoustoelastic coefficient and the grain size is established. The results show that one cycle is the optimal step length for the difference in the time-of-flight calculation, and with increasing grain size increase, L_cr wave acoustoelastic coefficient decreases in the form of a power function. In addition, grain size can be determined by using amplitude of the L_cr wave, so that the measured value of welding residual stress in HAZ can be corrected. The welding residual stress in melted zone (MZ) is corrected by calibrating acoustoelastic coefficient of the MZ. The acoustoelastic coefficient of the MZ is larger than that of parent material (PM). At last, welding residual stress in the butt-weld joint is measured and corrected with the L_cr wave technique. The results are verified by the hole drilling method.     

Using finite element and ultrasonic method to evaluate welding longitudinal residual stress through the thickness in austenitic stainless steel plates

This paper uses a 3D thermo-mechanical finite element analysis to evaluate welding residual stresses in austenitic stainless steel plates of AISI 304L. The finite element model has been verified by the hole drilling method. The validated finite element (FE) model is then compared with the ultrasonic stress measurement based on acoustoelasticity. This technique uses longitudinal critically refracted (L_CR) waves that travel parallel to the surface within an effective depth. The residual stresses through the thickness of plates are evaluated by four different series (1 MHz, 2 MHz, 4 MHz and 5 MHz) of transducers. By combining FE and L_CR method (known as FEL_CR method) a 3D distribution of residual stress for the entire of the welded plate is presented. To find the acoustoelastic constant of the heat affected zone (HAZ), a metallographic investigation is done to reproduce HAZ microstructure in a tensile test sample. It has been shown that the residual stresses through the thickness of stainless steel plates can be evaluated by FEL_CR method.     

Residual Stress Measurement in Steel Plates and Welds Using Critically Refracted Longitudinal (LCR) Waves

Abstract

The application of acoustoelasticity using critically refracted longitudinal (L_CR) waves is described for measuring residual stress in welded steel plates. Residual stresses are self-equilibrating and may exist in a material that has been deformed in a nonhomogeneous manner. When unknown residual stress is present in a structure, the true stress may become significantly greater than the working stress. In a corrosive environment, highly stressed areas that have not been properly stress relieved are prone to stress corrosion cracking. Areas near welds are particularly susceptible to stress corrosion cracking.

Two welded plates were investigated for the present work: one hot-rolled and the other cold-rolled. Residual stresses are usually greatest after welding. Further, longitudinal stress (i.e., stress parallel to the weld bead) is typically greater than the component transverse to the weld bead. Since the acoustoelastic behavior of the L_CR wave is largest when propagated parallel to a uniaxial stress, the L_CR wave traveling parallel to the weld bead was used to investigate the stress changes after stress relieving of the welded plates. Both 1 MHz and 2.25 MHz probe frequencies were used in this study. The stress changes in the welds and in the cold-rolled plate were clearly indicated by the L_CR data.

Two verification methods were used: hole drilling (HD) and neutron diffraction (ND). The stress relief was verified by the hole-drilling technique. While the HD technique showed about the same stress magnitude as found by the L_CR results, the orientation was reversed. The stress orientation was probably caused by the grinding process used to flatten the weld bead. Texture was also investigated using a neutron diffraction (ND) technique on the (001)[110] texture. The through-the-thickness technique yields an average of the orientation distribution of the (110) planes. At locations in the parent metal and in the weld, the distribution was found to be very similar, indicating uniform texture throughout the weld and parent metal zones.     

Residual stress measurement in steel plates and welds using critically refracted longitudinal (LCR) waves

The application of acoustoelasticity using critically refracted longitudinal (L_CR) waves is described for measuring residual stress in welded steel plates. Residual stresses are self-equilibrating and may exist in a material that has been deformed in a nonhomogeneous manner. When unknown residual stress is present in a structure, the true stress may become significantly greater than the working stress. In a corrosive environment, highly stressed areas that have not been properly stress relieved are prone to stress corrosion cracking. Areas near welds are particularly susceptible to stress corrosion cracking.Two welded plates were investigated for the present work: one hot-rolled and the other cold-rolled. Residual stresses are usually greatest after welding. Further, longitudinal stress (i.e., stress parallel to the weld bead) is typically greater than the component transverse to the weld bead. Since the acoustoelastic behavior of the L_CR wave is largest when propagated parallel to a uniaxial stress, the L_CR wave traveling parallel to the weld bead was used to investigate the stress changes after stress relieving of the welded plates. Both 1 MHz and 2.25 MHz probe frequencies were used in this study. The stress changes in the welds and in the cold-rolled plate were clearly indicated by the L_CR data.Two verification methods were used: hole drilling (HD) and neutron diffraction (ND). The stress relief was verified by the hole-drilling technique. While the HD technique showed about the same stress magnitude as found by the L_CR results, the orientation was reversed. The stress orientation was probably caused by the grinding process used to flatten the weld bead. Texture was also investigated using a neutron diffraction (ND) technique on the (001)[110] texture. The through-the-thickness technique yields an average of the orientation distribution of the (110) planes. At locations in the parent metal and in the weld, the distribution was found to be very similar, indicating uniform texture throughout the weld and parent metal zones.     

Characterization of the Critically Refracted Longitudinal (LCR) Waves and Their Use in Defect Detection

The critically refracted longitudinal (L_CR) waves are used in various domains of nondestructive evaluation, especially for residual stress measurements. However, few works have characterized the associated ultrasonic beam. In this article, the characterization of the L_CR ultrasonic beam, both numerically and experimentally, is first clarified in order to provide some answers to questions that arise about its behavior in elastic solids. The aim of the second part of this work is to investigate the use of the L_CR waves for the detection of surface defects of different sizes in aluminum. For that, the effect of defects at fixed depth (5 mm) with various diameters (2, 4, 6, and 8 mm) and fixed diameter (8 mm) with various depths (5, 7, and 10 mm) in an aluminum sample have been investigated at frequencies around 1 MHz. It has been experimentally found that the amplitude of the (L_CR) wave decreases and its frequency spectrum changes with a given defect. The rate of the decrease of the amplitude and the change in the spectrum is related to the increase of the defect depth. The study shows that the effect on the propagating of the L_CR waves is larger as the defect depth increases.     

Effect of post-weld heat treatment and electrolytic plasma processing on tungsten inert gas welded AISI 4140 alloy steel

Post-weld heat treatment (PWHT) is commonly adopted on welded joints and structures to relieve post-weld residual stresses; and restore the mechanical properties and structural integrity. An electrolytic plasma process (EPP) has been developed to improve corrosion behavior and wear resistance of structural materials; and can be employed in other applications and surface modifications aspects. In this study the effects of PWHT and EPP on the residual stresses, micro-hardness, microstructures, and uniaxial tensile properties are explored on tungsten inert gas (TIG) welded AISI-4140 alloys steel with SAE-4130 chromium–molybdenum alloy welding filler rod. For rational comparison all of the welded samples are checked with nondestructive Phased Array Ultrasonic Testing (PAUT) and to ensure defect-free samples before testing. Residual stresses are assessed with ultrasonic testing at different distances from weld center line. PWHT resulted in relief of tensile residual stress due to grain refinement. As a consequence higher ductility but lower strength existed in PWHT samples. In comparison, EPP-treated samples revealed lower residual stresses, but no significant variation on the grain refinement. Consequently, EPP-treated specimens exhibited higher tensile strength but lower ductility and toughness for the martensitic formation due to the rapid heating and quenching effects. EPP was also applied on PWHT samples, but which did not reveal any substantial effect on the tensile properties after PWHT at 650 °C. Finally the microstructures and fracture morphology are analyzed using scanning electron microscopy (SEM) and optical microscope to study the evolution of microstructures.     

Sub-surface stress measurement of cross welds in a dissimilar welded pressure vessel

Manufacturing process of pressure vessels used in high-temperature applications normally needs employing dissimilar metal welds (DMWs) between austenitic and ferritic steel. However, high amount of thermal stresses at the welds are induced due to differences in coefficient of thermal expansion between the types of steel. This study investigates the evaluation of welding residual stresses in a pressure vessel manufactured by DMW of 316 L stainless steel shell to A106 carbon steel caps as well as similar welding of stainless steel shells. By using longitudinal critically refracted (LCR) ultrasonic waves, the residual stresses are experimentally measured. The ultrasonic method is based on acoustoelasticity law by which the ultrasonic wave velocity could be connected to the material stress. By changing the frequency in which the ultrasonic transducers are working, the LCR waves are able to penetrate in various depths of the material in order to measure the sub-surface residual stresses. Hence, four main aspects are considered in this study: (I) stress evaluation of the DMW; (II) sub-surface stress measurement; (III) stress evaluation of longitudinal weld and (IV) stress measurement in cross weld. It is demonstrated that the residual stresses of the DMW pressure vessel could be comprehensively evaluated by using the LCR method.     

Investigation of Stress and Temperature Effect on the Longitudinal Ultrasonic Waves in Polymers

This work aims at establishing the effect of stress and temperature on the velocity of ultrasonic longitudinal waves in typical engineering polymers, and evaluating the potential of ultrasonic stress measurement in the evaluation of residual stresses in polymer parts. In order to estimate the effect of material morphology, two amorphous and two semicrystalline polymers have been considered. A series of tests are implemented, to determine the acoustoelastic constants and temperature constant of materials, by using the designed transducer fixtures for in situ measurement of longitudinal wave velocity. As expected, the velocity changes linearly with stress and temperature, and the temperature effect is as important as the acoustoelastic effect. It shows that this kind of nondestructive method is a valuable quantitative tool to estimate the residual stress in polymer products, but the material temperature influence must be considered during the estimation.     

Ultraschallbasierte in situ Vorspannkraftermittlung an Schrauben ohne Referenzmessung im nicht verspannten Zustand durch Kombination von Longitudinal- und Transversalwellen

Ultrasonic based bolt preload evaluation is commonly performed using the mono-wave method. This method works by measuring the time of flight of longitudinal waves. Here, a reference measurement in the unloaded condition is necessary for each bolt. In this publication the longitudinal wave is complemented by another type of ultrasonic wave - the transverse wave. This method does not require a reference measurement in the unloaded condition for each bolt. Moreover, an analytic method for determining the bolt-specific K-value is introduced, which is needed for the ultrasonic bolt preload determination. The analytically calculated K-values are compared with experimental K-values from tensile tests. The influence of material, bolt property class and surface protection system were determined with cylindrical specimens. In component tests, the bolt preload was evaluated using the bi-wave method, a possible influence of the bolt assembly method was investigated and the results were interpreted regarding their accuracy.     

Acoustoelastic determination of residual stress by measurement of resonance peaks and phase shifts

We made ultrasonic measurements of residual stress in a well-characterized specimen: an aluminum shrink-fit plate. We used various electromagnetic-acoustic transducers (EMATs) to generate and receive both shear and longitudinal waves. From the difference of velocities of orthogonally polarized shear waves (acoustic birefringence) we obtained the difference of principal stresses. From ratios of velocities of longitudinal and shear waves we obtained the sum of principal stresses. From the sum and difference we determined the principal stresses and found good agreement with theory. To obtain our ultrasonic data, we used two different techniques: measurement of resonant frequencies and phase shifts. We compared these methods both for accuracy and ease of implementation. For our experiments, phase shift measurements were superior. National Institute of Standards and Technology Materials Reliability Division     

Ultrasonic Velocity and Attenuation in Epoxy Resin/Granite (Marble) Powder Composite

In this research, gamma radiation has been used to graft styrene and acrylic acid monomers onto marble (M) and granite (G) powder, with irradiation doses 30, 10 kGy, respectively. The grafted marble and granite powders were used as filler of epoxy composites. They added to an epoxy resin at 20, 60, and 100 wt% and then cured at room temperature. Ultrasonic velocities and attenuation measurements in epoxy composites at 4 MHz in the 25–300°C temperature range in addition to Fourier Transform Infrared (FTIR) analysis had been investigated. The ultrasonic compression wave velocity and the shear wave velocity measurements were performed using the pulse-echo technique, and then wecalculated elastic modulus, Young’s modulus, longitudinal and shear moduli, and Poisson’s ratio. In addition to attenuation at different temperatures of epoxy/granite composites, ultrasonic results indicated that the ultrasonic wave’s velocities V_L and V_s increased with the increase of (M, G) content and a linear relation was observed. Also, the attenuation of epoxy was stable and appeared to increase at temperature 385°C, while the addition M or G to epoxy composites increased the attenuation.     

FATIGUE STRENGTH OF NON-LOAD-CARRYING FILLET WELDED JOINTS: EFFECTS OF WELD RESIDUAL STRESSES AND STRESS CONCENTRATION

The fatigue strength of non-load-carrying fillet welded joints of KE36(TMCP) steel was studied. Both residual stress measurements and fatigue tests were carried out, with the plate thickness, the plate width and the heat input being varied. Specimens given a Post Weld Heat Treatment (PWHT) were also prepared. The plate width had no effect on the fatigue strength, because it hardly affected the transverse residual stresses at the weld toe. However, the heat input influenced the transverse residual stress distribution, and a significant difference in fatigue strength due to the heat input was observed, especially when N≥ 10⁶ cycles. It was also found that PWHT removed almost all the residual stresses at the weld toe, improving the fatigue strength drastically. In this study, the values of stress concentration factor K₂ were estimated by Machida's method and it was concluded that the thickness effect resulted from a combination of both stress concentration and residual stresses with the contribution of the latter being particularly significant for N≥ 10⁶ cycles.     

Multiscale fatigue crack growth modelling for welded stiffened panels

The influence of welding residual stresses in stiffened panels on effective stress intensity factor (SIF) values and fatigue crack growth rate is studied in this paper. Interpretation of relevant effects on different length scales such as dislocation appearance and microstructural crack nucleation and propagation is taken into account using molecular dynamics simulations as well as a Tanaka–Mura approach for the analysis of the problem. Mode I SIFs, K_I, were calculated by the finite element method using shell elements and the crack tip displacement extrapolation technique. The total SIF value, K_tot, is derived by a part due to the applied load, K_appl, and by a part due to welding residual stresses, K_res. Fatigue crack propagation simulations based on power law models showed that high tensile residual stresses in the vicinity of a stiffener significantly increase the crack growth rate, which is in good agreement with experimental results.     

声弹性法测量铝合金预拉伸板中的应力

研究温度和应力对7050铝合金预拉伸板中的超声纵波、偏振横波和临界折射纵波传播速度的影响,并分析不同频率临界折射纵波在梯度应力场中的传播规律。结果表明:温度对声速的影响大于声弹效应的影响;单轴拉伸时,沿轴向传播的临界折射纵波、平行轴向偏振的横波的速度降低,垂直应力方向偏振横波的速度升高,垂直轴向传播的纵波速度变化不大;临界折射纵波的频率越高,其所反映的应力越接近表面;声弹性法测得的应力是声传播路径上各点应力在超声波造成的质点振动方向上分量的综合反映。