Ben T. Yen

Fatigue resistance of welded details enhanced by ultrasonic impact treatment (UIT)

Abstract Enhancement of the fatigue resistance of welded transverse stiffeners and cover plate details by ultrasonic impact treatment (UIT) was evaluated in 18 full-scale W27×129 rolled beam specimens. Fatigue tests were conducted under constant amplitude loading at various stress range levels and at two minimum stress levels simulating the effect of sustained load. The test specimens were investigated for fatigue crack initiation and propagation. Distributions of residual stresses adjacent to the weld toe were determined before and after the treatment. Test results indicated that UIT enhanced the fatigue performance of all treated details by improving the weld toe profile, changing microstructure and introducing beneficial compressive residual stresses at the treated weld toe. The treatment effectively elevated the fatigue crack growth threshold and the fatigue limit without changing the slope of the S–N curve.

Fatigue of Thin-Walled Plate Girders

For thin-walled plate girders, the factors to prevent fatigue cracks due to out-of-plane deformation of the web under in-plane loading are presented. The cracks are caused by plate-bending stress which is generated by out-of-plane deformation of the web. It is shown that for girders in bending, the deformation-induced cracks do not occur below the allowable fatigue stress range for the cracks at the end of vertical stiffeners which are classified as stress category C in the AASHTO Specifications. It is also shown that for girders in shear, the shear buckling strength for a rectangular plate with simply-supported conditions along the four edges gives a conservative guideline for the prevention of the deformation-induced cracks.

Fatigue strength of welded AL-6XN superaustenitic stainless steel

Full scale tests of welded AL-6XN superaustenitic stainless steel I-beams were carried out to assess the fatigue behavior of three weld details. A total of 66 full-scale I-beams were tested under constant amplitude loading. The test program was developed to assess the effects of minimum stress, stress range, and residual stress on the fatigue behavior of the longitudinal fillet weld detail, transverse groove weld detail, and a simulated bulkhead attachment detail. Weld defects that initiated fatigue crack growth were characterized. Internal discontinuities included porosity, entrapped oxides, and lack of fusion sites located in the longitudinal fillet welds and the groove welds. Surface cracks initiated at microdiscontinuities at the toe of the attachment fillet welds and at the toe of the groove welds.

Minimizing Fatigue and Fracture in Steel Bridges

Steel bridges are subjected to live loads which produce variable stress ranges in bridge components. At welded bridge details, the existence of initial defects and residual stresses eliminate the initiation stage of fatigue crack growth, and stress range is found to be the controlling factor for crack propagation. Laboratory tests have resulted in stress range-fatigue life relationships for various bridge details. These data correlate well with fracture mechanics theory and with field data. Limits on live load stresses have been adopted for steel bridges. Coupled with material fracture toughness requirements, the stress range limits minimize the probability of fatigue and fracture in steel bridges.

Theoretical Development of Lower Bound S-N Fatigue Curves

AbstractFatigue strength of welded structures is typically estimated through several approaches, including experimental S-N curves, the hot spot stress approach, and fracture mechanics. Recent work on the local stress field singularity of the weld toe has provided analytical expressions for the stress intensity factor of a crack emanating from the weld toe. Invoking the Paris law for fatigue crack growth together with the stress intensity factor expressions allows fatigue life to be evaluated in closed form. The fracture mechanics solution takes the same power-law form as the S-N curves and produces accurate lower bound estimates to the experimental fatigue data of stiffeners and attachments. The results show a strong relationship between the fracture mechanics, S-N curve, and hot spot approaches to fatigue that is based on two or more parameters from the local weld toe stress field. The two parameters may be tabulated as functions of weldment geometry and loading conditions for improved fatigue design an...

The effect of welding discontinuities on the variability of fatigue life

Large-scale I-section beams were welded from A710 (HSLA-80) steel plates. There were plain welded beams, similar beams with two attachment details fillet welded to the tension flange, and similar beams with transverse groove welds. The welding procedures used produced a wide range of welding discontinuities, particularly at groove welds which intersected the longitudinal fillet welds without a cope hole. The 162 beams were fatigue tested in four-point bending in a test matrix which included various stress ranges and minimum stress levels. The discontinuities at the origin of each fatigue crack were identified. These initiating discontinuities included microscopic weld toe discontinuities, porosity, inclusions, lack of penetration and hydrogen cracks. Fracture mechanics fatigue crack growth models were used to calculate the fatigue lives based on the initial discontinuity size. Various methods of idealizing the discontinuities as initial cracks were examined.

Fatigue Resistance Enhancement and Residual Stress Modification of Welded Steel Structures by Various Post-Weld Treatments

1 Three post-weld treatemnt methods (UIT, Air Hammer Peening, Shot Peening) are discussed in this study. Fatigue tests carried out on AL-6XN stainless steel specimens treated by these techniques showed that fatigue resistance has been significantly enhanced compared with the as-welded conditions. The enhancement is comparably effective for these three treatment methods in circumstance of this study. 2 Residual stress measurements were carried out using X-ray and neutron diffraction techniques to characterize the magnitude and subsurface distribution of the compressive stresses induced by UIT and shot peening treatments. Both treatments induced very high compressive stresses near the treatment location and the material was plastically deformed. 3 The compressive strength region by UIT alone (base metal) was 1.5–1.7mm in depth and ∼±15mm in width. For the welded specimen, this compressive layer could be reduced to 1mm deep due to existence of tensile weld residual stress. 4 The magnitude of compressive stress induced by shot peening was comparable to UIT, but the compressive layer was about half the depth (0.8mm) of that for UIT.