In this work, the finite element method is employed to gain an understanding of the behaviour of a cracked bridge roller bearing in service. The cracked roller is considered as an edge-cracked disk (two-dimensional plane strain system) subjected to a radial compressive line load. The crack parameters K_I and K_II are calculated for the relevant load configuration and angle of disk rotation. The calculated data are also used to check the accuracy of approximate SIF solutions reported earlier [1] and [2]. For plain Mode I loading very good agreement is found between the obtained results and data presented in Schindler and Morf (1994). 

The paper is aimed at finding the likely failure mechanism of a bridge roller bearing made of high strength martensitic stainless steel. Spectroscopy and finite element stress analysis of the roller indicated that an initial radial surface crack, found at an end face of the roller and close to the contact region, was induced by stress corrosion cracking (SCC). The initial crack subsequently changed shape and increased in size under growth through fatigue and finally formed a quarter-circle radial crack centred on the end face corner of the roller. Numerically computed stress intensity factors for the final crack showed that crack loading was predominantly in Mode II. For a crack size as observed on the fracture surface, the maximum service load, as specified by the manufacturer, enhanced by a certain roller bearing misalignment effect, was sufficient for failure through fracture.

With the innovation of elastomeric bearings in the mid-1950s steel bearings lost their interest and significance both in research and development and subsequently even in application. Steel bearings were gradually abandoned in bridges, followed by the technical literature and design standards. However, a great number of steel bearings remain today in service world-wide and will pose their particular challenges in the future. To the author’s knowledge, just in Sweden, high strength stainless steel bearings still exist in no less than some 650 bridges. In recent years, a large number of such bearings have failed with an alarmingly high frequency in Sweden during a period of six to twenty years after installation making them a serious maintenance cost issue.

After a brief summary of the history of high strength stainless steel bearings, the paper reviews service experience of such bearings in Sweden and elsewhere. Accompanying finite element analyses were performed in order to gain insight into the likely failure mechanism. Finally, this comprehensive review leads to a conclusion that identifies the causes of the failures occurred and makes some recommendations.

Although previous investigations of the stainless steel bearings have not been able to clearly identify the cause(s) of the failures occurred, it is found that the failures primarily occurred due to initiation of cracks through stress corrosion cracking followed by fatigue crack growth requiring a certain stress range and a sufficiently large number of cycles until final failure ensued through sudden and instable fracture after fatigue growth to a critical crack size.

This thesis is concerned with failure analysis of high strength steel bridge roller bearings.Paper Adescribes how the commonly used Hertz formulas for contact stresses underestimate the actual stresses in practice due to temperature differences, misalignments and other construction-related conditions. In this paper, finite element analyses of bridge roller bearings were carried out to investigate the accuracy of the traditional roller bearing design rules in view of issues such as girder deformability, misalignment imperfections and material nonlinearity. The results first indicated that roller bearings develop contact stress concentrations at the outer edges of the rollers. Second, it was shown that the contact stresses are very sensitive to misalignment imperfections between the bridge girder and the abutment. Third, it was shown that the roller bearings develop inelastic deformation at relatively low loads in relation to the design load.In Paper B, the finite element method was employed to gain an understanding of the behaviour of a cracked bridge roller bearing in service. The cracked roller was considered as a two-dimensional edge-cracked disk subjected to a diametrical compressive line load. The crack parameters, stress intensity factor Mode I, K_I and Mode II ,K_II were calculated for the relevant load configuration and angle of disk rotation. The calculated data for KIwere also used to check the accuracy of approximate stress intensity factor solutions reported earlier for Mode I. For plain Mode I loading very good agreement was found between the obtained results and data presented in Schindler and Morf (1994).

Paper Cis aimed at finding the likely failure mechanism of a bridge roller bearing made of high strength martensitic stainless steel. Spectroscopy and finite element stress analysis of the roller indicated that an initial radial surface crack, found at an end face of the roller and close to the contact region, was induced by stress corrosion cracking (SCC). The initial crack subsequently changed shape and increased in size under growth through fatigue and finally formed a quarter-circle radial crack centred on the end face corner of the roller. Numerically computed stress intensity factors for the final crack showed that crack loading was predominantly in Mode II. For a crack size as observed on the fracture surface, the maximum service load, as specified by the manufacturer, enhanced by a certain roller bearing misalignment effect, was sufficient for failure through fracture.InPaper D, after a brief summary of the history of high strength stainless steel bearings, the paper reviews service experience of failed bearings in Sweden and elsewhere. Accompanying finite element analyses were performed in order to gain better insight into the likely failure mechanism. Finally, thiscomprehensive review leads to a conclusion that identifies the causes of the failures occurred and makes some recommendations.

The present paper addresses how the commonly used Hertz formulas for contact stresses underestimate the actual stresses seen in practice due to temperature differentials, misalignments and other contruction-related defects. First, two failure cases of Swedish bridge roller bearings are analyzed and discussed; then, a detailed finite element (FE) model is used to investigate the accuracy of the traditional roller bearing design rules in view of issues such as abutment and girder deformability, misalignment imperfections and material nonlinearity. The bearing capacity of the studied rollers as provided by the manufacturer is used as reference. A rigorous FE model that accurately models girder, roller assembly and abutment provides the necessary information for the assessment of the related contact stresses, which were traditionally calculated by means of the Hertz analytical formulas. Numerical results first establish that roller bearings develop contact stress concentrations at the outer edges of the cylindrical drums. Second, it is established that the contact stresses are very sensitive to misalignment imperfections between the bridge girder and the abutment. Last, it is shown that the roller bearings develop inelastic deformation at relatively low loads in relation to the design load. These reasons, combined with the unlikelihood for roller bearings to shake-down, constitute the basis of the observed roller bearing failures.

In the present thesis, the application of fracture mechanics in design and assessment of steel structures has been studied.The first case study concerns a ring-flange connection used in wind turbine towers. The flange is rolled from straight steel profile into a complete ring. Subsequently, at the ends of the ring are welded by way of electron beam welding. This weld providing the integrity of the ring-flange was designed against fracture. To avoid potential failure by fracture and/or fatigue, the fracture mechanics approach for engineering assessments (as described in the Swedish handbook for safety assessment of structures containing defects) was implemented to predict the maximum allowable crack size. Material properties along with the design stress and crack size are the three variables that must be taken in account in order to design a structure against fracture. To this end, toughness and mechanical properties of the weld metal were experimentally determined. The results confirmed that the material behaves in a ductile manner and the electron beam welding has a positive effect on material properties. The fatigue performance of the flange was examined using the fatigue load histogram experienced by the flange in its 20years lifetime. Detailed finite element analyses of the flange connection were conducted to determine stress distribution at the most critical location along a hypothetical crack plane. The stresses obtained from the finite element analysis were used to assess the fracture stability of the flange. The fracture and fatigue assessment results confirmed that the flange in question is susceptible to fatigue failure which turns out to be the governing criterion in order to define the maximum allowable crack size. Furthermore, a parametric study was carried out to investigate the effect of parameters such as crack eccentricity, material properties and crack geometry on the integrity of the flange connection.The second case concerned the maintenance problem of roller bearings of relatively old bridges. These tend to rupture and no satisfactory explanation exists in the literature. Two broken specimens were supplied by the Swedish Transportation Authority (Trafikverket). These were first subjected to fractographic examination to determine the crack initiation location. Afterward, Charpy, compact tension, tensile and chemical composition specimens were machined out of the broken roller pieces and the respective tests were conducted. The results showed a very brittle material but no deviations from the prescribed material properties that were set forth in the 1967 certificate of application (Zulassung) of the rollers in question. To explain the observed failures a mechanical over-stressing condition was sought. Two such over-stressing sources were investigated. First the Hertz contact solution traditionally used in the design of roller bearings (both DIN 4141 and EN 1337-4) is valid for an infinite length cylinder. At the edges of the roller contact zone, stress singularities of the sort appearing in the case of a rectangular rigid body indenting an elastic half-space are likely to appear: three-dimensional finite element analyses were conducted and a 30% edge stress increase was established. As a second source of over-stressing the imperfections of welded I-beams according to EN 1090-2 were introduced into the FE model including, as before, the girder end, the roller assembly and the abutment. A wedge-type imperfection between the lower flange of the bridge girder and the lower support plate of the roller can create detrimental stress concentrations even at values no higher than 50% of the EN 1090 limits. Finally, the effect of the contact stresses in the roller was examined with a linear elastic fracture mechanics approach. It was found that surface cracks at the roller edge may become unstable in Mode-II when traversing the limits of the contact zone as this is the area of maximum shear stress: daily thermal cycles cause such cracks to go through a full plastic cycle as they are forced to swing past both edges of the contact area of the roller. This causes plastic strain accumulation at the crack tip of even small stable cracks: this mechanism causes them to grow until they become critical. A fracture assessment diagram for one of the broken rollers was computed.

A method for post-processing of finite element analyses to assess critical macroscopic voids is presented. One single, global analysis of the structure or component is carried out and then post-processed. By using a template model and the same constitutive model as for the global model, a damage scaling function is defined once and then applied to the result from the global model. The result is a database with critical relative sizes for different void rotations at the integrations points in the global model