^{1}

^{2}

^{3}

In this study, the effects of portable pneumatic needle-peening (PPP) on the bending fatigue limit of a low-carbon steel SM490A welded joint containing a semi-circular slit on the weld toe were investigated. PPP was applied to the specimens with a semi-circular slit with depths of a = 0.4, 0.8, 1.2, and 1.6 mm. Then, three-point bending fatigue tests were carried out under R = 0.05. The fatigue limits of low-carbon steel welded specimens containing a semi-circular slit were increased for peened specimens compared with non-peened specimens. Peened specimens having a semicircular slit with a depth of a = 1.2 mm had high fatigue limits, almost equal to those of the non-slit peened specimens. It was concluded that a semi-circular slit with a depth of less than a = 1.2 mm can be rendered harmless by peening. Then, the fatigue improvement by peening was predicted. The fatigue limits before and after peening could be estimated accurately by using a modified Goodman diagram considering the effects of residual stress, stress concentration, and Vickers hardness. Moreover, the maximum depth of a semi-circular slit that can be rendered harmless by PPP was estimated based on fracture mechanics assuming that the semi-circular slit was equivalent to a semi-circular crack. The prediction results were almost consistent with the experimental results.

Welded joints are often used in large steel structures such as bridges, ships, and pressure vessels. However, the welded joints reduce the fatigue limit of structures as a result of stress concentration caused by a discontinuous portion of the weld bead and tensile residual stress in the weld toe resulting from the effect of the heat accompanying the weld process. Fatigue cracks often initiate at the welded joint, ultimately leading to structural fracture. Therefore, non-destructive inspections are typically conducted in a regular manner, and detected cracks are treated appropriately. However, non-destructive inspection has a detection limit, i.e., small cracks below the limit cannot be detected. These undetected cracks are left until the next regular inspection, so these cracks not only result in a high cost for treatment in the future but also decrease the reliability of the welded structures. If these fatigue cracks could be rendered harmless, the structural integrity could be significantly improved and maintenance cost reduced.

In previous studies, research on rendering a surface crack harmless by shot peening was conducted. Shot peening is a surface treatment process used to produce a beneficial compressive residual stress layer and modify the fatigue strength of metals by impacting a surface with high-velocity spherical media [

The material used is low-carbon steel SM490A. The mechanical properties are shown in

Two kinds of peening were combined for the fatigue test, namely, portable pneumatic needle-peening (PPP) and ultrasonic needle-peening (UNP). PPP was performed at the weld toe, and UNP was conducted on the heat- affected zones of the specimens. The range of UNP treatment was 20 mm from the weld toe, as shown in

The residual stress was measured by the X-ray diffraction method. The X-ray conditions are a Cr-Kα beam X-ray spectrum and a 1.0-mm X-ray beam injection diameter. The residual stress distributions in the thickness

0.2% Proof stress [MPa] | Ultimate tensile strength [MPa] | HV |
---|---|---|

317 | 511 | 204 |

①PPP | ②UNP | ||
---|---|---|---|

Air pressure [MPa] | 0.5 | Frequency [kHz] | 20 |

Radius of needle tip [mm] | 1.5 | Amplitude [μm] | 40 |

Coverage [%] | >100 | Coverage [%] | >100 |

direction were obtained by alternately measuring the residual stress on the surface and then chemically etching to remove the surface layer.

The specimens were divided into four groups: 1) as welded, 2) having a semi-circular slit at the weld toe, 3) peening treated at the weld toe and 4) peening treated at the weld toe having a slit.

As shown in ^{6} cycles of stress.

The results of the fatigue tests are shown in _{a}) and the depth of the semi-circular slit (a) [^{6} cycles. The dotted lines represent the fatigue limit (σ_{w}). The asterisk symbols (*) indicate that the specimen fractured outside the semi-circular slit. It can be seen from

If the fatigue test results of a peened specimen with a semi-circular slit meet either of the following two conditions, the slit is considered to have been rendered harmless.

Condition (a): The fatigue limit increased up to more than 95% of that of the peened specimen without a semi-circular slit.

Condition (b): The specimen fractured outside the slit.

From

The longitudinal residual stress distributions for the welded specimen with peening and without peening are shown in

Measurement of Vickers hardness of the surface layer of the weld toe was performed. In the case of non-peened welded specimens, the value of the Vickers hardness was 280 HV. This was 37% larger than that before welding (204 HV). The reason for this is a refinement of the grains caused by welding. After peening, the value of the Vickers hardness was increased to 304 HV. This was 9% larger than that before peening. This was caused by work hardening resulting from peening.

Finite element method (FEM) analysis was conducted in order to calculate the stress concentration factor K_{t} of the weld toe before and after peening. The formula for K_{t} is:

where σ_{max} is the stress of the weld toe; σ_{nom} is the nominal stress, i.e., the stress of non-welded specimen at the same position as the weld toe.

The FEM analysis model is shown in _{x}, around the weld zone. _{t} can be calculated by means of comparing to the nominal stress (Δ): K_{t} before peening is 2.3, and K_{t} after peening is 2.0.

Fatigue limit estimation before and after peening was conducted by using the modified Goodman diagram, which is based on the method proposed by Nose [_{w0} for a stress ratio of R = −1 and the ultimate tensile strength σ_{B}. Line (b) shows the modified Goodman relationship considering the stress concentration at the weld toe. Specifically, σ_{w0} when the stress ratio is R = −1 is divided by the value of K_{t} before peening (K_{t} = 2.3). Line (c) shows the relationship between the mean stress and stress amplitude when the stress ratio R is equal to 0.05. Considering the surface residual stress after welding, line (d) is obtained by compensating line (c). The residual stress at the surface of the weld toe is considered in line (c). Line (e) shows the yield strength σ_{Y} after welding.

It is well known that mechanical properties such as σ_{w0}, σ_{B}, and σ_{Y} are proportional to hardness [_{w0}, σ_{B} and σ_{Y} are increased by welding and peening with the same increasing rate of Vickers hardness.

The value of the estimated fatigue limit can be obtained from the intersection between line (b) and line (d). Therefore, the value of the estimated fatigue limit before peening was 135 MPa, which is 16% safer compared to the experimental value shown in

In _{w0} is divided by the value of K_{t} after peening (K_{t} = 2.0). From

These fatigue limit estimation results are close to the experimental results if the measurement errors are taken into consideration. Thus, the fatigue limits before and after peening can be estimated accurately.

The semi-circular slit size that can be rendered harmless by peening was evaluated based on fracture mechanics assuming that the semi-circular slit was equivalent to a semi-circular crack. In this study, the semi-circular slit size which can be rendered harmless by peening was evaluated by means of focusing on the stress intensity factor K. _{T} contributes to fatigue crack propagation. If the sum of the minimum stress intensity factor K_{min} and the stress intensity factor by residual stress (K_{r} ) becomes compressive, ΔK_{T} can be calculated by the following equation.

The values of K_{max} and K_{r} were evaluated by FEM analysis using a quarter of a specimen model that has a semi-circular crack, as shown in _{max} was calculated by applying the operating stress distribution after peening in _{r} was calculated by applying the residual stress distribution after peening in

_{T} and the semi-circular slit depth a. ΔK_{T,A} and ΔK_{T,C} represent ΔK_{T} at the deepest point and at the surface of a semi-circular slit respectively. ΔK_{T} was calculated with Equation (2) because the sum of K_{min} and K_{r} calculated by FEM analysis became compressive. The threshold stress intensity factor range (ΔK_{th}) depends on the crack size. In order to determine the relationship between ΔK_{th} and the lengths of cracks, the equation proposed by El Haddad et al. [

ΔK(L)_{th} is the threshold stress intensity factor range for a large crack. We used ΔK(L)_{th} = 8.4 MPa∙m^{1/2}, which was obtained by the uniaxial loading test for a specimen containing a large through-crack [_{w0} is the stress range at the fatigue limit for non-peened specimens. We used Δσ_{w}_{0} = 420 MPa, which was obtained from the modified Goodman diagram in

Assuming that the semi-circular slit is equivalent to the semi-circular crack, we can judge whether the semi-circular slit can be rendered harmless. If ΔK_{T} is smaller than ΔK_{th} , the semi-circular slit can be rendered harmless. Thus, the intersection between ΔK_{T} and ΔK_{th} gives the maximum semi-circular slit size a_{max} that can be rendered harmless.

From Figure13, the value of a_{max} was 1.37 mm. As mentioned in Section 3.1, a semi-circular slit with a = 1.0 mm was rendered harmless by peening. Thus, this estimation result was consistent with the experimental results as shown in

In this study, the influences of portable pneumatic needle-peening on fatigue limits and the semi-circular slit size that can be rendered harmless by peening for a butt-welded joint of low-carbon steel SM490A were investigated. Three-point bending fatigue tests were carried out under R = 0.05 using a welded specimen containing a semi-circular slit. Then, the fatigue limit before and after peening was predicted by the modified Goodman diagram. Moreover, the fatigue improvement by peening and the maximum depth of the semi-circular slit that can be rendered harmless were estimated. The results are as follows.

1) The fatigue limits of low-carbon steel SM490A welded specimens containing a semi-circular slit were improved by peening. The fatigue limits were increased 43% - 118% for peened specimens compared with non- peened specimens.

2) Peened specimens having a semi-circular slit with a depth of less than a = 1.2 mm had high fatigue limits, almost equal to those of the non-slit peened specimens. We conclude that a semi-circular slit under a = 1.2 mm in depth can be rendered harmless by peening.

3) After peening, compressive residual stress was introduced on the weld toe. The surface and the maximum compressive residual stresses were about 490 MPa and 550 MPa respectively.

4) The value of the Vickers hardness after peening was 9% larger than that before peening.

5) The stress concentration factor K_{t} of the weld toe was calculated using finite element method analysis. It was decreased from 2.3 to 2.0 by peening.

6) It became clear that the fatigue limits before and after peening at the weld toe could be estimated accurately by using the modified Goodman diagram considering the effects of residual stress, stress concentration, and Vickers hardness.

7) The semi-circular slit size rendered harmless by peening was evaluated based on fracture mechanics assuming that the semi-circular slit was equivalent to a semi-circular crack. The value of a_{max} was 1.37 mm. The predicted result was almost consistent with the experimental results.

The authors wish to acknowledge Prof. K. Ando, emeritus professor of Yokohama National University, for his valuable comments. The authors also express our appreciation to TOYO SEIKO Co., Ltd. for peening treatment and measurement of residual stress.