Procedia Structural Integrity最新文献

Wire Arc Additive Manufacturing is an additive manufacturing process with a high rate of material deposition capable of producing near-net shape parts. This process involves the reduction of production costs (material and lead times) and considers innovative designs. However, the deposition technique induces heterogeneities in the material, in particular the presence of porosity and a degraded surface finish. The process-induced surface asperities have a first-order influence on the fatigue life of as-built parts as they act as stress raisers. Various finishing treatments can be considered to reduce the criticality of the surface finish influence over crack initiation and propagation: conventional ones such as hammer, laser or shot peening and some specially developed for Additive Manufacturing (AM) processes such as in-situ cooling or hot rolling. The multitude of AM parameters and the different finishing surface post-treatments entail many configurations that will modify fatigue properties. For this reason, rapid fatigue evaluation methods are an asset for process evaluation.

Thermo-elastic Stress Analysis (TSA) is a non-contact technique for measuring the distribution of stress at the surface of a component subject to cyclic loading using an infrared camera. The analysis of the thermo-elastic coupling amplitude maps allows the detection of initiation and monitoring of crack propagation. A four-point bending fatigue test protocol is conducted on CuAl9 WAAM specimens take in different direction for the deposition direction. Then failure mode and life duration are compared for the 2 directions.

Crane runway girders are subjected to cyclic loading and must therefore be designed against fatigue failure. For existing structures, however, there are no standards for handling pre-damaged components at the end of their calculated lifetime. The Institute for Material and Building Research of the University of Applied Sciences Munich examines different approaches on how to deal with existing welded steel structures. The research project addresses the following questions:

How can pre-damaged components without visible cracks be strengthened?

How can components with visible cracks be repaired in a fatigue-proof manner?

How can pre-damaged components be reinforced through a low-notch application of steel cover plates?

In order to answer these questions several numerical and experimental investigations are carried out. Different innovative fastening techniques like lockbolts, adhesives and self-tapping screws for attaching reinforcement cover plates are tested on small and large specimens. Within this paper the research project will be presented and previous results on components without visible cracks will be summarized.

One of the simplest and most efficient ways to design lightweight structural components is the combination of welding and aluminum alloys. However, welded joints are extremely sensitive to fatigue failure and making accurate lifetime predictions is still challenging when Variable Amplitude (VA) loading conditions are involved. Among all design criteria available in the literature, the present investigation focuses on the Peak Stress Method (PSM), an engineering finite element (FE)-based approach to rapidly assess the fatigue strength of welded joints. In more detail, the PSM suggests modelling both weld toe and weld root as sharp V-notches having null tip radius and correlates their fatigue strength using the intensity of the local linear elastic asymptotic stress distributions described by the Notch Stress Intensity Factors (NSIFs). The theoretical formulation of the PSM for the fatigue strength assessment of welded joints subjected to VA loadings has been recently proposed by combining its Constant Amplitude (CA) formulation with the Palmgren-Miner's cumulative linear damage rule. Such VA formulation has been successfully validated against a large bulk of experimental fatigue results generated by testing welded joints made of structural steels under uniaxial as well as multiaxial loadings. In the present investigation, the VA formulation of the PSM has been further validated against experimental data relevant to welded joints made of aluminium alloy under VA loadings.

In order to predict fatigue lifetime of cableway installations components, EN 1993-1-9 standard (Eurocode 3) is commonly used. However, EN 1993-1-9 is based on a major hypothesis : stress state has to be uniaxial. But for some components, this uniaxial hypothesis is not verified (for example : fixed grip of chairlift submitted to horizontal tightening stress and vertical load stress, or chairlift structure stressed by gravity and lateral shake when passing a tower). It then appears important to use a fatigue criterion taking into account multiaxial stress state. Our research work proposes to apply for fatigue study the Dang Van criterion, which takes into account multiaxial stress state. Results are then compared to Eurocode SN curves, representing a huge experimental database on unixaxial fatigue, that we propose to use in multiaxial fatigue thanks to an appropriate recalibration. The scientific originality of this work lies in the justification of that Dang Van criterion's recalibration. Therefore, a theorical study ensures that our use of Dang Van's multiaxial criterion is still consistant with the Eurocode SN curve even if these SN curves are experimentally performed under uniaxial stress state. Finally, that hitherto unpublished use of the Dang Van criterion makes it possible to consider the entire database of detail categories defined in NF EN 1993-1-9, by generalising their use for components under multiaxial fatigue.

To ensure reliability of hybrid cylindrical roller bearings each ceramic roller is inspected, and scrapped if a surface imperfection above the critical size is detected. This inspection aims to reduce the potential root cause of Rolling Contact Fatigue (RCF) which can occur during operation, shortening the bearing life. The rejection criterion is based on experimental and theoretical knowledge, which for the last decade was developed in SKF for the material imperfections mainly located on the rolling element raceway. These imperfections are subjected to high contact pressure and therefore are considered as the primary root cause of RCF failure. Regarding rollers, however, imperfections can be present beyond the raceway, i.e. at the roller chamfer where the lower risk of RCF is expected, because the edge imperfections are typically out of the rolling contact zone. Nevertheless, the risk associated with these features should be assessed too, chiefly because the size of edge imperfections can be rather large. In our previous study, the imperfection termed as a Missing Material was studied, combining the semi-analytical tool for the contact mechanics and the Finite Elements (FE) method for the stress analysis. In the current work, another imperfection type is considered, and this is a surface crack located at the chamfer of ceramic roller. The RCF analysis is based on the semi-analytical evaluation of the rolling contact pressure (between a ceramic roller and a steel inner ring), and computational fracture mechanics for the estimation of fatigue crack propagation.

The frequency effect is a commonly encountered challenge in ultrasonic fatigue testing (UFT) of low-carbon, ferritic steels, wherein factors such as the increased strain rate and reduced test duration change the apparent fatigue resistance of the tested material. The usability of UFT for rapid f atigue testing of these materials is therefore limited as the results cannot be directly compared to conventional fatigue results. In this investigation, fatigue curves were evaluated at frequencies of 20Hz and 20kHz for two comparable grades of ferritic structural steels: Q355B and S355JR, using different conventional frequency specimen geometries. Methods to evaluate the frequency sensitivity of the steels based on the finite life regime were adapted from previously proposed models in literature to produce corrected curves and to allow comparison to similar steels in literature. It was found that previously reported results may be overestimating the frequency sensitivity due to the influence of size effects. It was also found that these models are of limited use for producing corrected SN curves based on UFT data.

Among the energy-based approaches to estimate the fatigue life of steel specimens, the experimental method based on the heat dissipation (or intrinsic dissipation) per cycle, Q, proved effective for correlating the effects of geometrical stress concentrations, uniaxial and multiaxial loadings, and mean stress. The mean stress effect requires a properly defined temperature-corrected parameter Q. The parameter Q is readily evaluable using temperature measurements and in this investigation it has been employed for fatigue strength assessment of plain specimens, extracted from a 42CrMo4 Q&T connecting rod big end of a marine engine. Completely reversed, strain-controlled, constant amplitude fatigue tests were carried out and the Q parameter evolution was monitored during each test by suddenly stopping the fatigue test several times and measuring the cooling gradient of material temperature. As result, besides the traditional strain-life (εa-2Nf) curve, the Q-Nf curve was also obtained, which is expected to be applicable for correlating notch and mean stress effects in future investigations.

Bogies of rail vehicles for passenger coaches and traction units commonly contain air spring systems as secondary spring stages. In the development and design of spring stages, it is necessary to ensure precise knowledge about the material properties and fatigue behaviour of the air spring bellows.

The aim of this work is to systematically investigate the damage mechanisms evaluated at air spring bellows on sample level and to analyse the fatigue strength of the base material under different load conditions. The specially developed small-scale sample is biaxially loaded and different layups are examined at varying load levels. In the tests with different parameters, an increase in the mean value of the longitudinal displacement by 20 % has proved to act as a suitable failure criterion. In addition to the purely optical damage analysis, micro computed tomography analysis was carried out.

In this study, four layered samples with a fibre angle of ±15, ±25 and ±35 degrees in respect to the longitudinal direction are examined. A global evaluation of the service life tests reveals that under comparable load conditions, the fibre angle exhibits a clear influence on the fatigue strength. The increase of 10 degree in fibre angle roughly results in a 15 % reduction of the tolerable lateral displacement amplitude at a number of fifty thousand load-cycles which commonly acts as design lifetime. In a second step, a local analysis based on an analytical approach is presented. With the help of the fibre strain amplitude calculated, all fatigue test data points can be unified to a master S/N-curve leading to an elaborated design model of cord rubber composite materials used in air spring bellows of rail vehicles.

With the help of the presented methodology utilizing the developed representative small-scale sample testing procedure and evaluation approach, a time- and cost-efficient fatigue design is facilitated.

Finite Element Analysis (FEA) is widely used to perform fatigue calculations for geometric singularities at welded components. The analysis methodologies are described in design codes and recommended practices such as DNV, IIW and Eurocode. The focus in the present study is the application of hot spot stress methodology on a weld detail located at the cope hole in a pile sleeve connection of a jacket substructure. Finite element analysis is used to calculate the geometric stress where the influence factor (INF) technique has been implemented to calculate the hot spot stress at the weld location. The INF methodology is used as the preferred approach compared to the traditional nominal stress method due to its ability to capture the stress response in complex welded details. Generally, a mid-surface shell model excluding the weld is used to model the welded components in FE analysis and a stress extrapolation method is applied to calculate the hot spot stress at the fatigue critical location. Here a full solid model of the cope hole detail including the weld geometry has been used for fatigue calculation as benchmark to calibrate the weld modeling techniques using shell elements for the analyses. The results confirmed that the weld geometry and stiffness has a significant influence on the hot spot stress calculation at the considered cope hole. Thus, the weld geometry and stiffness must be included into the finite element model for an accurate fatigue damage calculation of such details. The calibrated results showed that the mid surface shell model can still be used if an appropriate weld stiffness is included in the finite element model.

The need for more environmentally sustainable ways of transportation and for a reduction in emissions and fuel consumption make lightweight structures essential. Together with the use of lightweight materials and design optimisation, the use of hybrid joints represents one way to reduce the weight of components and it is becoming increasingly popular in the transportation industry. The name ‘hybrid joint’ refers to a connection where adhesive bonding is used in conjunction with traditional joining techniques, such as spot welds and rivets with the aim of combining and exploiting the advantages of the individual joining techniques. To optimise the design of hybrid joints and minimise the risk of in-service fatigue failures, the transportation industry needs efficient, robust, and easy-to-use approaches for the modelling and fatigue life estimation of hybrid joints.

This work presents two practical methodologies for estimating the fatigue life of hybrid joints that can be easily adopted by companies in the transportation industry. The first methodology neglects the life given by the mechanical joints after failure of the adhesive (the joint is considered failed when the adhesive fails), while the second methodology considers the life of both the adhesive and the mechanical joints. In the first methodology, just one configuration would need to be analysed (‘hybrid’ joint or ‘purely bonded’ joint, if this simplification is considered reasonable). In the second methodology, instead, the analysis of two configurations would be required (the previous configuration followed by a configuration where only the mechanical fasteners are considered). The second methodology would produce more realistic fatigue life estimations compared to the first methodology, but it would be more onerous both in terms of modelling and computationally. For both methodologies, FE modelling guidelines to recover the required stresses are suggested. These guidelines require limited changes to the typical FE modelling strategies currently used, especially in the automotive industry. Furthermore, the proposed modelling guidelines provide FE models that are not computationally too onerous, reasonably mesh insensitive and that do not require congruent meshes. The relevant stresses recovered from the FE model are then used as an input into nCode DesignLife to estimate the fatigue life of the hybrid joints in the analysed structure. The fatigue life estimation is carried out using standard Stress-Life (SN) based nCode DesignLife analysis engines and bespoke SN curves obtained through testing of hybrid joint specimens, representative of the joints in the production parts.