System redundancy of bridges is studied by comparing load-carrying capacities between a bridge system and individual girders. Nonlinear 3D FE models of existing composite simple I-girder bridges and a non-composite continuous I-girder bridge, which were designed according to a previous design code, were developed to investigate the effect of bridge types on the system redundancy. A simple method is proposed to quantify system redundancy as a system factor for load rating. The proposed system factor facilitates accounting for the system redundancy into evaluating load-carrying capacity for existing bridges by using conventional linear elastic structural analyses.

This paper discusses compressive strength of box section members with artificial uniform section loss portions. FE elasto-plastic analyses were conducted using solid element models for 112 cases with parameters of depth, area of section loss and slenderness ratio in order to clarify the effect of the section loss on the elasto-plastic behaviors/ultimate strength. As results, effects of section loss on the ultimate strength were numerically assessed and practical equations for estimating ultimate strength of box section members with uniform section loss were proposed.

In this study, the stress transmission between main and patch plates in the patch plate repair method by the combined joint of adhesive and high-strength bolt friction joints was focused on. The uniaxial tensile tests and finite element analysis were carried out for the defective steel plates repaired by the patch plates. As a result of the tensile tests and FE analysis, it was found that the adhesive joints can transmit the stress on the main plate to patch plates mainly even in the combined joint of adhesive and high-strength bolt friction joints until debonding. Therefore, adhesive and high-strength bolted combined joint might be considered as the adhesive joint until debonding.

From around 2010, the replacement of deteriorated RC slab in composite girders is began to be implemented largely. One of the main problems of replacement work is the lack of ultimate strength of compressive flanges when the deteriorated RC slabs are removed. Therefore, it is required to strengthening of the compressive flange before removing the slab. In this study, in order to improve the load-carrying capacity of the compressive flange, that is the outstanding plate which simply supported at three edges, the angle lumber strengthening of the free edge in the outstanding plate by using Thread Rolling Screws, TRS. The proposed method was verified by the compression tests of the cruciform-sectional column specimen strengthening by the angle lumber with TRS.

Bridges are composed by many structural members which interact with each other to resist against various load combinations. In the present study, full-scale FE analyses have been conducted for a steel I-girder bridge system with corrosion damages which have been removed each member. From the analytical results, it was found that when corrosion damage occurred at the end of the outer girder and the girder was loaded, the influence on the maximum load carrying capacity is large for the deck slab and the cross frames. When the corrosion damage ratio is low, the load sharing function of the cross frames increased near the damaged girder. However, when the corrosion damage ratio is high, the deck slab became effective, and the load sharing path changed depending on the degree of corrosion damage.

In this paper, we investigated the ultimate behavior of the end sway bracing in a steel truss bridge. It is revealed that the buckling strength and deformation of the bracing member is very similar to those of the ompressed fixed support column by cyclic loading experiments using about half-size specimens of the panel structure of an actual truss bridge. It was also confirmed that the load-displacement hysteresis curve of the panel specimen was a spindle shaped loop. Furthermore, in order to perform a rational seismic retrofit design for the truss bridge, we proposed an analysis model using fiber elements and confirmed that the behavior of the experiment could be reproduced.

In this paper, we studied on the method of seismic reinforcing the gusset plate joint in the existing upper steel arch bridge. Because the bracing member and joints of arch bridges constructed in the past are designed only with seismic force by a level 1 earthquake motion, by a level 2 earthquake motion the bracing member is not only damaged but also the gusset plate joint is damaged. Therefore, we conducted experiments on the panel including bracing members and the joints using 1/2 scale specimens of actual bridges, we made clear the ultimate behavior of the bracing member and the joint designed by level 1 earthquake motion. Furthermore, we proposed a method for reinforcing the joint, and confirmed the effectiveness of the method by experiments using 1/2 scale specimens and FEM analysis.

In this paper, the authors focus on the repair design of thickness-reduced steel members by patch plates with high-strength bolts. In the previous papers, it has been reported that the axial-force sharing in thickness-reduce steel plate does not correspond to that calculated by the composite theory of the thickness-reduced plate and the patch plates. However, so far, the dead load in steel members and slip load of patch plates are not considered in the previous studies. In this research, the effect of dead load and slippage of patch plates are considered to the patch plate repair method for the thickness-reduced plate. Finally, the authors discussed how to design the patch plates to enable or disable the thickness-reduce plates.

This study focused on the stresses in a reduced-thickness plate repaired by a onesided patch plate with a high-strength bolted joint under axial loading. First, the stresses in the thickness-reduced plate repaired by the one-sided patch plate were evaluated by finite element analysis. As a result, it was found that the stress distribution in the thickness-reduced plate was different from that in the composite theory of defected section with patch plate. The stresses in the thickness-reduced plate repaired by the one-sided patch plate under axial loading were estimated by the frame structure analysis, and the results showed a similar tendency to that calculated by FE results. Finally, the tensile tests were conducted to verify each analysis result.

Aluminum alloy is a light-weight material with excellent corrosion resistance but low rigidity. When the aluminum alloy is used to a girder bridge, it takes high costs owing to the increment of its stiffness. Therefore in order to reduce a material cost, the cost minimization problem was performed on beam string structure (BSS) made of the aluminum alloy material. We focused on the layout of the BSS and diameter of the cable. The conducted simulation made clear the effectivity of the BSS to the aluminum alloy material for a reduction of material cost and increment of the beam span.

In this study, the authors conducted tensile coupon tests of an aged built-up member and residual stress measurement of channel-shaped steels removed from a truss bridge, which has passed 115 years since the completion. In addition, the authors numerically analyzed the compression strength of builtup columns with disappeared lacing bars. As a result, specimens exhibited low tensile strength compared to current structural carbon steel, JIS SS400. The channel-shaped steels showed irregular residual stress distribution, whereas the magnitude of residual stress was equivalent to the existing measurement results. Moreover, the vanishment of a lacing bar does not significantly affect the compression strength of the built-up column.

Due to the rapid spread of road bridges about sixty years ago, noticeable deterioration has occurred in the existing bridges in japan. The purpose of this paper is to evaluate the remaining load carrying capacity of steel plate girders with a fatigue crack. In this study the plate girder was modeled by FEA and various proportions of bending and shear load were applied analytically. The results of the analyses found that in the case of shear-dominant loading, the cases with cracks are not safe. Moreover the result indicated that when there is no significant difference in the ratio of bending to shear loading, the shear load capacity decreases and the bending load capacity varies a little as the crack length increases.

Chloride attack, which is a typical aged deterioration of RC structures, significantly reduces the bearing capacity of RC structures due to rebar corrosion. The repairs for the deteriorated structures are generally carried out by qualitative judgment based on inspection. In the inspection data, the corrosion crack width is quantitative data related to the rebar corrosion. If the deterioration of the structures can be quantitatively estimated from the corrosion crack width, the existing RC structures can be reasonably maintained. The authors collected a lot of rebar corrosion experimental data and studied the relationship between the corrosion crack width and the amount of rebar corrosion, and its uncertainty. Then using the reliability method, we estimated the limit state probability for the model case of RC structures.

The aim of this paper is to quantify the effect of track irregularity on the dynamic response of railway bridges during train passage. Literature survey results found that the standard deviation of the vertical track irregularity of recent general track condition in Japan is about 3 mm for conventional lines and about 1.5 mm for Shinkansen. Numerical simulations considering vehicle / bridge dynamic interactions and the track irregularity as a stochastic model using Monte Carlo method were carried out. An impact coefficient of track irregularity was proposed as a function of the speed parameter and the irregularity amplitude so that the non-exceeding probability is approximately 95% for the calculated results.

In recent years, works to improve existing structures and strengthen their seismic resistance have increased. Pile construction in narrow spaces is constrained by the site and the construction process. Therefore, a construction method of soil-cement composite pile using a mechanical agitator was developed. This paper briefly summarizes the construction method, then discusses the results of static load tests, and finally summarize the results of simulation by beam-soil spring model. The paper suggests the evaluation method of vertical bearing characteristics for the soil-cement composite pile.

In recent years, since it has been confirmed that the piers of old bridges collapse and the entire bridge is washed away, it is necessary to quantitatively evaluate the fluid force acting on the piers and bridge girders during floods. In two-phase flow, the hydraulic force greatly depends on the water surface shape of the surface of the structural member, the river volume inhibition rate, and the degree of inundation. In this study, the finite volume method is analyzed using OpenFOAM, and while comparing single-phase flow and two-phase flow, inundation with respect to the water surface shape, river volume inhibition rate, and girder height of the pier surface is performed. It was found that the drag coefficient of the structur al member is constant in the region.

In this paper, analytical study was conducted for the purpose of understanding the seismic behavior of railway viaducts located near the strike-slip fault. A non-linear static analysis was first carried out using a 3-dimensional frame model of a railway viaduct. The nonlinear dynamic behavior was then investigated considering effects by both inertial force and fault dislocation induced by a strike-slip fault. As a result, the response curvature of RC members drastically increased compared to the response in case of considering only each effect. It follows that it is necessary to calculate the response value in consideration of actions of both inertial force and fault dislocation if a railway viaduct is constructed near the fault.

The current seismic design specification for highway bridges is based on the low-level reliability theory, and there might be insufficient consideration of structural uncertainties. In this study, the uncertainty of seismic response was investigated by considering material properties and curing conditions using a pier model. The simulation results point out that material strength has a significant influence on the seismic response, but the uncertainties of response don’t follow the assumed normal distribution. In the analysis focusing the curing conditions, the results indicate that it is worthwhile to consider not only the material strength but also the effects of cracking and drying shrinkage during curing as probabilistic factors in seismic design for comprehensive reliability design.

Consideration of a hardening phenomenon of rubber bearings against seismic motions stronger than a design seismic motion is important for seismic isolated bridges. When the hardening phenomenon occurs, bearing stiffness suddenly increases and a large inertia force acts on the pier. This study investigated an active vibration control method using fuzzy control for the seismic isolated bridge to prevent the hardening phenomenon. The results showed that the controlled bridge prevented the rubber bearing from hardening in the event of an earthquake motion stronger than the design seismic motion.

An evaluation method for the residual performance of a supplemental facility on structures was proposed to quantify its recoverability after an earthquake. In this study, two indices were proposed. The one was Rr, expressing the amount of deterioration of the seismic performance after an earthquake. The other one was Rd that is the intensity ratio of second to first earthquake, assuming the structure endures under first motion but reaches its limit state under the second motion. The proposed indices were evaluated for an electric power pole on top of railway structure. It consequently follows that proposed methods would successfully help to determine the necessity of repair of power poles immediately after an earthquake.

Seismic performance of masonry piers retrofitted with prestressed PC bars was evaluated in experimental investigation and in numerical simulation with Applied Element Method (AEM). For a masonry pier with a weak joint in the middle height, sliding of masonry bed joints could be controlled and rocking behavior of the masonry pier was obtained using appropriate PC bar retrofitting method. The structural behaviors of masonry piers with and without retrofitting were well simulated by AEM. Furthermore, appropriate edge distance for PC bars to prevent damage of cover zone was proposed based on simulation results.

This research aims to understand the tensile force acting on the fixing bolts caused by the bending-shear deformation of an actual super-high damping rubber bearing. Loading tests were conducted on a developed loading system that allows bending-shear deformation of an actual rubber bearing for simulating tensile behavior of the fixing bolts. The tests achieved the bending-shear deformation on an actual rubber bearing. The tensile force acting on the fixing bolts increased nonlinearly from the bolts on the loading side caused by the tensile force generated at the rubber end. The maximum value was reached at the end bolt farthest from the loading side. In addition, numerical analyses were performed to grasp the effect of tensile force generated at the rubber end on the tensile behavior of the fixing bolts.

Two transmission towers, namely No.78 and No.79 located at the Kiuchi power line in Japan, collapsed due to strong winds caused by the typhoon Faxai in 2019. To study the collapse mechanism, a time-dependent wind response analysis was conducted on the basis of wind load conditions estimated in accordance with the quasi-steady theory. By comparing the obtained results of analysis with the actual damage, it was found that a main member of No.78 buckled first, whereby No.78 tilted on to the side of No.77 owing to a large difference between the wind loads at the front and rear spans. Furthermore, the obtained results indicate that the decline in the resistance of the bottom panel of No.78 was caused by deformations of the bracing members and auxiliary members present on the same panel.

In this study, wind tunnel tests were conducted on vehicles on a girder simulating the Tokyo Bay Aqua Bridge for various cases of lane, vehicle type, relative angle, and airflow. The pressure distribution and aerodynamic force coefficient acting on the vehicles were calculated. As a result, the differences in the pressure distribution and aerodynamic coefficients according to the lane, vehicle type, relative angle, and airflow were significant. Therefore, it indicates the possibility of applying reasonable traffic regulation standards depending on vehicle type and lane instead of the current regulation standards where no differences are considered.

Rapid structural evaluation of bridges after earthquakes is imperative to recover quickly. While structural displacement is an essential indicator, direct measurement is difficult. Indirect estimation of displacement by numerical integration of measurement acceleration is inaccurate due to integration errors. In this research, estimation of displacement by only using acceleration measurements is proposed. Extended Kalman filter (EKF) assuming a simple non-linear hysteretic behavior and a genetic algorithm to identify parameters of the non-linearity are combined to achieve the displacement estimation. The displacement estimation capability of the proposed method is confirmed with simple models and a nonlinear RC pier model simulating an E-defense shaking table specimen.

In recent years, studies have been reported which aim to detect bridge damage from changes in vibration characteristics. In those studies, variable characteristics are estimated from bridge dynamic responses by utilizing structural identification methods such as ERA and stochastic subspace identification, where estimation accuracy generally decreases due to the influence of noises in the response data. This study aims to improve the estimation accuracy of natural frequencies by processing response data for inputting. Verification using analytical and actual response data showed improvement of identification in accuracy for lower-order modes and robustness for higher-order modes.

This study investigates feasibility of VBISI method, which can simultaneouslly estimate all mechanical parameters (mass, damping and stiffness) of vehicle and bridge and road roughness, only from the vehicular vibration and position data. The proposed VBISI method in this study consists of Karman Filter and Adaptive Nelder-Mead method. To show the efficiency of this idea, the shapes of objective functions are also illustrated. The obtained results show that the estimated values for the mechanical parameters are accurate, even in the cases considering noises. However, the robustness to various noises needs to be investigated in the future.

Fourier transform is an essential procedure to obtain vibration characteristics of structure. Fast Fourier Transform (FFT) Algorithm is often used to calculate Fourier spectrum. However, if FFT algorithm is used for a real-time continuous monitoring of Fourier spectrum, calculation for every new sample occurs a waste of computational resources. Sequential Fourier Transform is basically based on discrete Fourier transform by storing of samples in ring buffer. Difference of Fourier spectrum is calculated on every single sample updating. Therefore, the number of multiplications on the single sample updating are minimized compared with FFT. This paper describes outline of Sequential Fourier Transform technique, cumulative Fourier spectrum, real time filtering and examples of implementation.

In this study, we investigated the feasibility of using bridge-weigh-in-motion (BWIM) as a feature for bridge anomaly detection in machine learning. To balance additional influence line due to bridge damage, the proposed BWIM approach evaluates actual and virtual wheel loads. The displacements used in BWIM were estimated from video footage using a deep learning method. The weight assigned to the virtual wheel has been considered as a feature in machine learning based anomaly detection. Damaged model bridge experiments showed the proposed method’s ability to detect bridge anomalies and its sensitivity to a damaged position.

This study investigates feasibility of bridge damage detection using BWIM and deflection of the bridge. A virtual axle is introduced to the vehicle so that increase of identified wheel loads caused by changes in deflection influence line due to damage can be distributed to the virtual axle. Investigating the identified wheel load of the virtual axle which should be zero in a pristine condition of the bridge theoretically, we can examine the possibility of damage in the bridge. Both model bridge damage experiments and simulations showed changes in the wheel load of the virtual axle due to damage in the bridge. Observations imply that selecting a proper location of the virtual axle would improve the accuracy of delivering increased virtual wheel load due to bridge damage.

One of the causes of fatigue damage in steel bridges is high stresses due to traffic loads. The Bridge Weigh-in-Motion (BWIM) is a system to measure the traffic loads in the bridge. The response induced by a vehicle which does not enter the measurement section may be sometimes observed, even though the target bridge is not a continuous girder bridge. Based on synthesized observation data, the effect of considering the pre-entry response on the results of axle weight estimation is studied first. Then using the measured data, the effect of pre-entry is examined. The relative error of estimated axle weight with the consideration of the pre-entry response was about 20% smaller than that without the consideration.

Relaxation of high strength bolted joints causes large bolt pretension loss at an initial stage of relaxation. However, the detailed characteristics over a long time period had not been clarified. In this study, relaxation tests were carried out in high strength bolted joints. Then, we investigated characteristics of bolt pretension loss of relaxation over a long time period by literature search. From these results, we discussed the testing period for relaxation tests in high strength bolted joints.

Railway steel structures are often erected in a short time at night. Therefore, it is required to shorten the erection time by reducing the number of bolts of the friction type of high strength bolted connections and improve the workability by reducing the weight of the splice plate. As a method of reducing the number of bolts and the weight of the splice plate, there is a method of adopting a large diameter bolt and narrowing the bolt pitch. In this study, we perform tensile tests and FEM analysis of friction type of high strength bolted connections to clarify the effects of bolt diameter, splice plate thickness, and bolt pitch on the slip coefficient.

There are some examples of studies focused on the slip coefficients in high-strength bolted frictional joints with various holes. It found that the slip coefficients in the joints with oversized holes and slotted holes were less than the coefficients in the joints with normal holes. However, in Japan, there is no rule regarding the slotted holes, and the use is not accepted. In this study, FEM was carried out and the mechanism by which the slip coefficients were decreased was investigated in the joints with slotted holes and oversized holes varied the slip/yield resistance ratio and the washer thickness. Therefore, the mechanism of the slip coefficient reduction is clarified.

Huck High-Strength One-Side Bolt (hereinafter, referred to as “MUTF”) is one of the most-used blind bolts for repaired and strengthened of members in Japanese existing structures. However, there are few examples of studies focused on contact pressure and ultimate limit state in frictional joints with MUTF. In this study, slip and after-slip behavior in the joints were examined by contact pressure tests and tensile tests. As a result, the slip coefficients in the joints with MUTF and thin splice plate are about 0.10 smaller than the others because the contact pressure is higher. Also, the ultimate strengths in the joints which have a shear surface on the bulb sleeve are much more than the designed value.

When repairing the plate of a member with a closed cross-section, a singe-sided construction bolt is applied. The authors proposed the Mechanical Bearing Blind Rivet Bolts, which can be installed on single side, and transmits the load by the bearing force acting by the contact between the bolt and the bolt hole wall and the friction force acting on the joint surface. The load-displacement relationship of the Mechanical Bearing Blind Rivet Bolts up to the sliding load is the same as that of the friction joint using M22F8T, and it is possible to apply the same use limit to Mechanical Bearing Blind Rivet Bolt joints as to friction joints.

This study focuses on frictional load transfer of high strength bolted frictional joints in post-slip behavior. In a past experiment, the ultimate resistance of the bolted joint tightened without standard preload is almost the same as that of the joint tightened with the preload, although the slip resistance depends on the bolt preload. In this study, finite element analysis had been conducted to clear the mechanism. The bolt preload deceased according to the progress of post-slip behavior in the case the bolt tightened with the preload. In contrast, the bolt preload increased in the case the bolt tightened without preload. As a result, the ultimate resistances of both cases are almost the same. It is supposed the change of the bolt preload is caused by the progress of the bending deformation of the bolt shank.

The ultimate limit state of high-strength frictional bolted joints should be determined considering the resistance and ductility to enhance the resiliency of connected members. We proposed the deformed bearing limit state defined by bolt hole deformation and clarified the effect of various configurations on the bearing resistance. In this study, pure bending experiments of the girder bolted connections, which varies the cross-section classification of girder and the number of bolts of flange and web splices, have been conducted to evaluate the deformed bearing limit state of girder bolted connections. The ultimate strength was exceeded the designed full plastic moment, and the flange bolt hole deformation at the designed full plastic moment was less than 10% of the bolt diameter.

In the design and construction of high-strength bolt friction joints, an axial force reduction due to relaxation is known to occur. This reduction has been supplemented by tightening the bolts with an introduced axial force that is 10% higher than the design axial force. In this study, we focused on the residual axial force of frictional joints of high-strength bolts, and the extracted high-strength bolts from four steel bridges in service considering different bridge types, bolt types, and joint surface treatments to determine whether the residual axial force was maintained as designed. The results of the investigation showed that the residual axial forces from the four bridges were lower than the design values except for some.

For the maintenance of high strength bolted connections, it is required to clarify the relationship between corrosion behaviors of spliced plates and slip resistance. Therefore, various tensile tests had been performed on slip specimens with corroded spliced plates. However, it have been not yet clear that the slip resistance in high strength bolted connections with corroded spliced plates on both sides. Therefore, in this study, tensile tests are performed in order to clarify the slip resistance of high strength bolted connections with corroded spliced plates on both sides. And from the tests results, the diagnostics for corroded spliced plates on both sides of high strength bolted connections are proposed.

Epoxy adhesive is often used for making the recess of corroded plate flat in the repair of the plate using high-strength bolts. The slip strength and mechanical behavior of the combined joints with high-strength bolts and adhesive is unclear. In this study, tensile tests of the combined joints, which vary a bolt pitch, have been conducted to clarify the slip strength and slip mechanism. FEA was carried out to evaluate quantitatively the slip strength of the combined joints. In result, the slip strength of the combined joints is larger than frictional joints even if narrow bolt pitch. As the bolt pitch and the thickness of connecting plates increase, the slip coefficient becomes larger.

This study focuses on connections of steel earth-retaining beams used in underground construction. Although these connections have been used for decades, the yield and maximum bearing capacity and collapse process are not cleared. Therefore, this study conducted four-points loading test to obvious these mechanical behaviors. The experiment results show that the yielding of end-plate joint is dominant in the yield capacity of the earth-retaining-beam connections. Further, the presence or absence of a splice plate on the compression side of the retaining member has no effect on the yield and maximum bearing capacity, indicating that the splice plate on the compression side can be omitted under pure bending conditions.

The cast iron deck is a lightweight deck slab with excellent fatigue resistance. The cast iron decks are joined by high-strength bolts at the end ribs. However, there is a problem that the deck is separated when negative bending acts on the joint. As a countermeasure, we proposed a structure that uses both friction joint by the splice plate to the upper surface of the deck and tension type joint of end ribs. In this study, this new joint structure was analyzed by FEA to clarify the effect on separation of deck and the interaction between friction and tension type joint. Furthermore, we investigated how to improve the performance of the joint.

Friction joints are used to join the cast iron deck and the main girder. In this study, the effect of the structural type of the support member between the deck and the girder on the slip behavior is investigated by push-out shear test. In addition, the presence or absence of taper in the joint and the effect of the bolt hole shape on the slip resistance will be examined. As a result, it was found that when the T-type support member was used, the slip resistance was lower than when the L-type support member. In addition, no significant difference was observed when slotted holes were used, and it was found that the slip resistance decreased when the taper was used.

Outer girder of girder end parts, consisting of main I -girders, stiffeners above the supports, crossbeams and so on, in steel girder bridge had serious damages by lateral load during a strong earthquake. However, design method of the girder ends, having reinforcing steel members, against the lateral load is not detailed in design specifications. Experimental observations of a web gap, a clearance between the underside of the crossbeams and the lower flange plates of the main girders, will help to discuss that point, but a few experimental studies were carried out. Then in this paper, loading tests using partial scaled models of the outer girder end are carried out to reveal load carrying capacity against the lateral load .

To investigate the influence of high-frequency induction heating for paint-coating removal on deformation and residual stress of steel structural members, a series of experiment and analysis was carried out. The heating of 160 to 260 ℃ generated the out-of-plane deformation of 0.13 mm to 0.52 mm in the steel plates with the thickness of 9 mm, 16 mm, and 25 mm. The maximum tensile residual stress was around 200 N/mm² in the heating region. The residual stress distribution in the heating region was not uniform and varied largely. The stress distribution might be strongly influenced by the change in the running speed of heating device.

In this paper, the computational fluid dynamics analyzes behavior of temperature and humidity inside of the cable covered with polyethylene. First, the computational fluid dynamics analyzes temperature and humidity in the cross-sectional direction in the cable for which the temperature and humidity were measured. Next, the computational fluid dynamics analyzes temperature and humidity in the extension direction assuming installation of the cable in the actual bridge. As a result, it was found that the unique behavior of temperature and humidity inside of the cable can be reproduced by considering the delay of phase change due to the complicated cross-sectional configuration inside the cable as a coefficient.

In this research, the authors focus on the strain under unloading conditions to detect the fatigue crack in welded structures. The fatigue tests of out-of-plane gusset joint under stress ratio R = -1 were carried out by measuring the strain every 10 thousand cycles. As the fatigue crack initiates and propagates, the mean strain which corresponds with the strain under the unloading conditions measured near the weld toe becomes smaller. This is because that the release of the tensile residual stress introduced by the welding. On the other hand, it was shown that the mean strain measured at the compressive residual stress region gradually increases by the crack initiation and propagation. Additionally, the appropriate location of strain gauges was discussed.

Basic three factors affecting fatigue durability of welded joints are shape of welded joints, stress ranges and their number of cycles. Under the allowable stress design, members designed to less dead load tend to be subjected to larger live loads. It implies that they are often subjected to larger stress ranges, and they are clarified as fatigue prone members. Steel expansion joints is one example of the fatigue prone details. Referring a few examples of fatigue cracks of finger joints, fatigue assessments are carried out to show how we can verify fatigue durability of these members.

In this study, the authors propose a new technique of adding small-diameter high-strength bolts near the crack to reduce the stress concentration in the stop-hole repaired by the patch plates. In order to confirm the effectiveness of the proposed method, tensile tests and finite element analyses were conducted on specimens with slits and stop-holes in the main plate and patch plates applied from both sides. As a result, it is clarified that the axial load sharing of the patch plate in the stophole increases due to the effect of shortening the distance between the small-diameter bolts, and the stress concentration in the stop-hole wall can be reduced in the elastic behavior range by the proposed method.