Non-conservative Predictions Research Articles

A continuum approach to multiaxial high-cycle fatigue modeling for ductile metallic materials

In this work we propose a generalization of the endurance function employed in the continuum approach to high-cycle fatigue modeling by Ottosen et al. (2008). The von Mises-type backstress-modified effective stress term of the original formulation is replaced by the non-quadratic Hershey-Hosford function, which introduces flexibility with respect to the predicted fatigue limit in pure shear. While the original model assumes a fixed ratio between the fatigue limits in fully reversed torsion and fully reversed bending of 0.5774 for all materials, the modification enables the ratio to take on values on the interval from 0.5 to 0.5852 by calibration to experimental data that is readily available in the literature. Like the original formulation simplifies to the invariant-based Mises-Sines fatigue limit criterion for proportional stress histories, the current proposal can be written as a Hershey-Hosford-Sines criterion. A systematic validation to experimental multiaxial infinite fatigue life data from eight different mild and semi-hard metals with fully reversed torsion and bending fatigue limit ratios lower than 0.6 demonstrates that the proposed modification increases the accuracy and reduces the number of non-conservative predictions.

A comparative study on characterizing the fracture toughness variations in thick-wall hydrogenation reactor welded joint by spherical indentation tests

This study provides an investigation on the existing models in predicting the fracture toughness distributions in thick-wall hydrogenation reactor welded joint from SITs. Three criterions, namely the critical stress, the critical strain, and the critical damage variable, were used in determining the critical indentation depth in the IEF model, together with the ERR model. Experimental results proved that none of the existing four criterions (or models) works for all zones in the welded joint. The critical stress criterion well reproduces the decrease of KIC in the weld (compared with that of the base metal) and the KIC distributions along the thickness, but may provide highly non-conservative predictions (more than 100% over-estimation). Highly stable KIC predictions were obtained from the critical strain criterion and the ERR model, but the critical strain criterion failed in characterizing weakest parts of the welded joint (i.e., the HAZ and weld) as assuming all zones with the same plasticity, and the ERR model failed in characterizing one of the weakest parts (i.e., weld) due to the existence of initial damage. Compared with the critical strain criterion and the ERR model, highly instable prediction results were observed in the CDV criterion, with the maximum error exceeds 50%. Source of errors in the existing fracture toughness prediction models were extensively explored through the physical basis of each criterion (or model) and from the solidarity of the phenomenologically summarized criterions.

Probabilistic investigations on the elastic stiffness coefficients for suction caisson considering spatially varying soils

The importance of accurate estimation of elastic stiffness coefficients of suction caisson has been already highlighted in previous works. Nonetheless, studies related to the spatial non-uniformity of the clayey seabed are somewhat limited, and current theoretical solutions are usually established using deterministic soil properties, overlooking the inherent uncertainty of the soil. This study focuses on the elastic performance of suction caisson in spatially random soils using three-dimensional (3D) small strain random finite element analyses within a Monte-Carlo framework. The non-uniformity of soil is mapped by modeling the shear modulus as a random field. The influences of soil profiles, embedment ratios and random field parameters on the elastic stiffness coefficients are comprehensively quantified. Results demonstrate that the spatially varying shear modulus notably alters the failure mechanism, which in turn significantly affects the elastic stiffness coefficients, ignoring soil variation will provide non-conservative predictions. To deal with the uncertainty of soil shear modulus, a series of expressions are developed by adding reliability-related parameters to the existing theoretical solutions, which can be adopted to calculate quantiles of stiffness coefficients corresponding to several given probabilities. The findings of this research may potentially facilitate the elastic performance assessment of suction caisson.

Development of a strain-based fracture assessment approach for X80 steel pipe welded girth by modified Mohr-Coulomb model

In the study, a strain-based fracture assessment scheme with the Modified Mohr-Coulomb (MMC) model is established to numerically characterize the tensile strain capacity (TSC) of X80 steel pipe welded girth. The TSC at crack initiation and local instability are predicted based on a lode-dependent MMC model accounting for crack onset and propagation phases. The combined experimental-numerical calibration of the MMC criterion for crack onset is undertaken for X80 pipeline steel, where seven geometrical specimens are designated to represent various stress states from low to high triaxiality, and the parameter determination of the MMC criterion for crack growth is carried on by fitting the single-edge notched tension tests. Finally, the comparison between the present work and PRCI and ExxonMobil in terms of TSC is performed. A relatively good agreement can be achieved, whereas PRCI gives non-conservative predictions of TSC at crack initiation when crack depth a/t is longer than 0.4 and ExxonMobil gives conservative predictions of TSC at local instability when internal pressure is small.

Shear performance and design rules for welded stainless steel I-shaped members with web openings

The use of stainless steel I-shaped members as primary load-bearing members has become increasingly popular, particularly in acidic environments, owing to their corrosion resistance and other advantages. Openings on the web are often introduced for ease of installation. The presence of web openings can reduce the shear resistance and produce high stresses around the openings, limiting their application. However, no previous study has investigated the shear resistance and design rules of such welded stainless steel I-shaped members with web openings. This work presents a detailed numerical analysis to examine the shear performance of perforated stainless steel members. Finite element (FE) models considering material nonlinearity and initial geometric imperfections were developed and verified using corresponding test data presented by the authors and many researchers. A parametric study including 434 FE models was undertaken to analyse the influence of web openings on the shear resistance of such members. The results suggested that for members with a hole diameter–web height ratio of 0.7, web openings can reduce the shear resistance by 61.4% on average, when compared to unperforated members. The generated numerical results were used to evaluate the current design recommendations given in the Eurocode (EN 1993-1-4+A1) (2015), American standard (ANSI/AISC 370-21) (2021), and those design rules proposed by Chen et al. (2018), although they were designed only for unperforated members. The comparisons demonstrated that the current design guidelines provided non-conservative predictions for calculating the shear resistance of perforated members. New design formulas were proposed using the strength reduction factor. A reliability analysis was undertaken, which indicated that the new design formulas were reliable for predicting the shear resistance of such members.

Earthquake Response Modeling of Corroded Reinforced Concrete Hollow-Section Piers via Simplified Fiber-Based FE Analysis

The effect of corrosion-induced damage on the seismic response of reinforced concrete (RC) circular bridge piers has been extensively investigated in the last decade, both experimentally and numerically. Contrarily, only limited research is presently available on hollow-section members, largely employed worldwide and intrinsically more vulnerable to corrosion attacks. In this paper, fiber-based finite element (FB-FEM) models, typically the preferred choice by practitioners given their reduced computational expense, are validated against previous shake-table and quasi-static cyclic tests on hollow-section RC columns, and then used to investigate the influence of corrosion-induced damage. To this end, modeling strategies of varying complexity are used, including artificial reduction of steel rebar diameter, yield strength and ductility, as well as that of concrete compressive strength to simulate cover loss, and ensuing dissimilarities quantified. Pushover and incremental dynamic analyses are conducted to explore impacts on collapse behavior, extending experimental results while accounting for multiple corrosion rates. Produced outcomes indicate a minimal influence of cover loss; substantial reductions of base shear (up to 37%) and ultimate displacement capacity (up to 50%) were observed, instead, when introducing relevant levels of deterioration due to corrosion (i.e., 30% rebar mass loss). Its predicted impact is generally lower when considering more simplified assumptions, which may thus yield non-conservative predictions.

Failure modelling of glass plates in biaxial loading: using flaw-size based weakest-link systems

Experimental strength tests are performed on two series of nominally equal plate specimens of annealed soda-lime glass subjected to either ring-on-ring or ball-on-ring bending. The Weibull effective area which represents a fictitious surface area exposed to uniform tension is calculated using closed-form solutions. Finite-size weakest-link systems are implemented numerically in a computationally intensive procedure for random sampling of plates extracted from a virtual jumbo pane whose surface area contains a set of stochastic Griffith flaws. A non-linear finite element analysis is conducted to compute the bending stresses. The glass surface condition is represented in different flaw-size concepts that depend on a truncated exponentially decaying flaw-size distribution. Stress corrosion effects are modelled by implementation of subcritical crack growth. The effective ball contacting radius is determined in a numerical computation. The results show that surface size effects in glass are not only a matter of strength-scaling, as also the shape of the distribution changes. While the lowest strength value, as per the major in-plane principal stress at the recorded fracture origin, in the respective data sets is very similar, the strongest specimen observed in ball-on-ring testing is over 70% stronger than the correspondingly strongest specimen observed in ring-on-ring bending. The Shift function is used to make visual comparisons of the difference in quantiles in the observed data sets. Use of an ordinary Weibull distribution leads to non-conservative strength predictions on smaller effective areas, and to too low strength predictions than are viable for glass design on larger areas. The numerical implementation of finite-size weakest-link systems can produce better predictions for the strength-scaling compared to a Weibull distribution, in particular when the flaw-size concept is modified to include a doubly stochastic flaw-size distribution or a random noise added to each subdivided region of the discretized surface area. The simulated ball-on-ring fracture origins exhibit greater spread from the centre point than otherwise observed in laboratory tests. It is indicated that the chosen representation of surface condition may not be accurate enough for the modelling of all fracture origins in the ball-on-ring setup even though acceptable results are obtained with the ring-on-ring model. There is a need for more insight into the surface condition of glass which can be conducive to the development of flaw-size based weakest-link modelling.

Evolution of metal-loss severity criteria: Gaps and a path forward

This paper considered integrity management (IM) based on the convergence of inspection technologies with those for defect assessment, as occurred circa 2000. A reformulation of the assessment criterion PCORRC, termed R-PCORRC, was presented, which now embeds a function of the steel’s strain-hardening response. Relative to full-scale tests for Grades from B up through and beyond X100 (L690) this criterion was shown accurate within 1%, with a low coefficient of variation of just 0.077. In contrast, B31G and Modified B31G showed significant bias, and scatter. The highly selective sometimes significant effect of width was shown to depend on the structural stiffness within the metal loss as compared to the stiffness of the pipe surrounding it. Defect assessment based on ILI of real corrosion was considered next, to identify technology gaps in the practical applications of this technology. It was apparent that improvements in MFL signal interpretation will help to resolve the issues with boxing criteria. While such concerns currently can be managed using the concept of plausible paths, as feature size and interaction are better characterized this concern will diminish. In such cases the predictive outcome of the plausible paths concept will in the limit tend toward that of B31G.The need for and benefits of accurate precise predictive schemes for metal-loss severity were illustrated. It was shown that a conservative prediction of failure pressure gives rise to a non-conservative prediction of defect size, leading to a non-conservative re-inspection interval for a pipeline. Finally, it was also shown that using an accurate scheme can result in significantly reduced maintenance, with a related benefit in a reduced scope of field digs to demonstrate the viability of inspection-based IM.

Effect of steel tube thickness on the behaviour of CFST columns: Experimental tests and design assessment

In this paper, the results of experimental tests conducted on concrete-filled steel tubular (CFST) columns are presented. There is currently a deficit of data available that can be used to evaluate current guidance documents and provide assessment to improve their accuracy when considering the behaviour of CFST columns filled with high strength concrete. Thus, this paper aims to increase the volume of experimental data available with a series of groups of tests on stub CFST columns subject to axial compression. Among the specimens of the same group only one dimension changes, the steel tube thickness. The columns are filled with both normal and high strength concrete for comparison purposes. The use of high strength concrete in circular tubes increases the concrete contribution, but this effect does not fulfil in rectangular specimens where also the confinement is less effective than in circular CFST even when the former have thin steel tubes and high strength concrete. The specifications of four commonly used codes are discussed. Comparison of their predictions with the experimental data collected shows that AISC is conservative but EC4, DBJ and AS produce similar non-conservative predictions.

Numerical investigation on ultimate shear strength of long steel plate girder web panels at high temperatures

In this article, elastic shear buckling and ultimate shear strength of long steel plate girder web panels subjected to pure shear loading are investigated at both ambient and elevated temperatures by the finite element method (FEM). Ninety-six plate girders with compact, non-compact and slender long web panels are numerically analyzed and compared. From linear eigenvalue analysis, it is shown that AISC 360-16 predictions are more conservative than FEM results such that the difference is about 40% for all ranges of web panel slenderness. According to nonlinear analysis results at ambient temperature, there must be reduction factors applied to AISC 360-16 non-conservative predictions for considering the strength degradation caused by large slenderness values and effects due to initial geometrical imperfection. Furthermore, it is observed that the adopted equation of AISC 360-16 for fire situations yields values that are more non-conservative such that the difference reaches almost 18%. In this regard, the FEM results are used to develop a new equation for predicting the ultimate shear strength of long web panels by taking into account strength degradation caused by large slenderness values, high temperature and initial geometrical imperfections.

Development of cohesive zone models for the prediction of damage and failure of glass/steel adhesive joints

The use of mild steel/tempered glass adhesive joints has increased rapidly over recent years. Cohesive zone modelling (CZM) is used extensively for the numerical analysis and failure prediction of adhesive joints. The bonding to the glass surface is generally weaker than the bonding to metal substrates, and therefore the development of cohesive laws by testing on different substrates generally leads to overoptimistic and non-conservative predictions. However, the interface characterisation using standardised methods for glass/steel joints is complicated due to the relatively low strength of the glass substrate leading to premature failure. This paper presents modifications proposed for the Double Cantilever Beam (DCB) and End Notched Flexure (ENF) tests bonded with dissimilar glass/steel adherends used to extract traction-separation laws in fracture modes I and II. For this relatively small coupon size, an in-house glass heat strengthening process was developed. The cohesive laws were validated by comparing the numerical predictions for two different adhesives with experimental test data for double lap shear joints subjected to four different load cases.

Multiaxial fatigue experiments and life prediction for silicone sealant bonding butt-joints

Uniaxial and multiaxial fatigue experiments were conducted on the adhesively bonded hollow cylindrical butt-joint specimen under load-controlled mode, and several prediction criteria were used for multiaxial fatigue life prediction. The experimental results showed that the fatigue life of uniaxial tensile cycle was shorter than that of uniaxial torsional cycle, and the fatigue life of the non-proportional circular loading path was shorter than that of the proportional loading path. In addition, the multiaxial fatigue life prediction results showed that Smith-Watson-Topper criterion based on the maximum principal strain as the critical plane had a conservative life prediction for the non-proportional circular loading path. Fatemi-Socie criterion mainly considered the influence of shear strain which yielded the non-conservative prediction for some data points of proportional loading path. Chen-Xu-Huang criterion combined energy density and critical plane achieved better fatigue life predictions than those from the previous two criteria. Furthermore, considering general failure modes of the adhesively bonded butt-joints, the modified Chen-Xu-Huang model was proposed, and the predicted results are good agreement with the experiments.

A coupled crack initiation and propagation numerical procedure for combined fretting wear and fretting fatigue lifetime assessment

This work presents a numerical procedure for combined fretting wear and fretting fatigue problems. It combines the Archard wear model and the Smith-Watson-Topper fatigue indicator parameter with the Miner damage rule for crack nucleation estimation. The eXtended Finite Element Method (XFEM) is then used in combination with the Archard wear model for crack propagation prediction. This procedure was validated in 12 experimental tests taken from the literature, and a good correlation was found. The procedure allows a detailed study of the interaction between the crack and fretting contact in combination with wear. The results show that the relative slip decreases and changes its distribution unevenly when the crack grows, and highlight the key importance of accounting for wear in order to obtain non-conservative predictions.

Stress Intensity Factor Equations for the Evolution of Surface and Corner Cracks to Through Cracks

Part-through surface or corner 2D cracks are commonly found in structural components, even because practically all fatigue cracks tend to start this way. It is a reasonable hypothesis to model them assuming the shape of their fronts is approximately an elliptic arc, as supported by many fractographic observations. But their transition to a 1D through-crack, an important issue in many practical applications, is normally not properly addressed in fatigue life predictions. Although experimental results reveal that the frontier of surface cracks essentially retain their elliptical shape as they gradually grow into a through-crack, it is usual to assume they are immediately transformed into a through-crack when their depth reaches the cracked component thickness. This crude approximation may create a large jump in stress intensity values, leading to excessively conservative fatigue crack growth predictions, or else the crude shape jump hypothesis may induce false overload events that can much affect fatigue crack growth retardation models, leading to inadmissible non-conservative life predictions. To minimize such problems, an improved model to describe the transition of 2D surface cracks to 1D through-cracks is proposed and verified by crack propagation tests in two different materials, 4340 steel and polycarbonate (PC). Moreover, fatigue life predictions based on this improved model are compared with experimental results.

Rotational demand and capacity of conventional single-plate shear connections subjected to gravity loading

This study aims to investigate the adequacy of the ductility requirement presented in the 15th edition of AISC Steel Construction Manual for conventional single-plate (CSP) connections when the threads of the used bolts are excluded from the shear plane (i.e., using X-bolts). For this purpose, by changing different parameters, 72 specimens were designed in accordance with the AISC Manual. All specimens were analyzed based on finite element (FE) method using ABAQUS software. FE results indicated that the most effective parameters which affect the connection rotational demand are loading type and beam length-to-beam depth ratio while connection rotational capacity is mostly affected by the number of bolts, beam depth and bearing strength of beam web/shear tab. To make a more accurate judgment about connection ductility, first, the existing equation for estimating rotational demand, usually leading to a non-conservative prediction, was modified. Then, using the FE results, three predictive models for connection rotational capacity was proposed. Comparing these models showed that the one which was developed on the basis of the artificial neural network is of the highest level of accuracy. Finally, using Monte Carlo method, a reliability analysis was conducted which showed that the ductility requirement of AISC Manual can be too conservative in most cases when the number of bolts exceeds five.

Taking into account the non-proportional loading effect on high cycle fatigue life predictions obtained by invariant-based approaches

Industrial structures are often subjected to multiaxial fatigue loadings. If the multiple stress signals are not synced the loading is said to be non-proportional. Most of the multiaxial fatigue criteria give highly inaccurate lifetime predictions when used in the case of such loadings. The scalar equivalent stress defined by the criteria does not take into account the non-proportional nature of the multiaxial loading and leads to non-conservative predictions. Moreover a multiaxial fatigue criterion can only be applied on a stress cycle which has no clear definition when multiple unsynced signals are to be considered. This study addresses these issues by proposing a correction of an invariant based multiaxial fatigue criterion through the definition of a non-proportional degree indicator. A definition of multiaxial cycle is also given based on the Wang-Brown method. Finally a complete chain of invariant based lifetime prediction for non-proportional multiaxial fatigue is validated.

Multiaxial fatigue analysis of complex welded joints in notch stress approach

The notch stress method is often recommended for fatigue assessment of complex welded joints as it can consider the influence of weld toe geometric parameters on local stress. However, the application of the notch stress method is still in the exploratory stage in multiaxial fatigue compared with the traditional nominal stress method and structural stress method. Considering the geometric effect at the weld toe, the notch stress method describes the local stress with an accurate geometry model and the complex stress states at the damage point under multiaxial loading are determined in the local coordinate system. Using the notch stress method, six classical multiaxial fatigue criteria are re-evaluated for a series of typical welded joints in the literatures under both proportional and non-proportional loads. In order to evaluate the different criteria quantitatively, a statistical analysis is performed and the probability of non-conservative prediction is determined for each multiaxial fatigue criterion. The results show that the S-N curve for notch stress under multiaxial loading has a lower inclination and higher fatigue class compared with the uniaxial curve recommended by IIW.

Evaluation of the bolt and weld eccentricities of conventional single-plate shear connections

This study aims to investigate the accuracy of bolt eccentricity (eb) and the required weld size for connecting shear tab to the support presented in 15th edition of AISC Steel Construction Manual for conventional single-plate (CSP) connections. In addition, the available equation for prediction of the weld eccentricity (ew) was evaluated. For this purpose, by changing different parameters, 72 specimens were designed in accordance with the AISC Manual. The designs are in a way that the bolt fracture becomes the governing limit state. Investigated parameters include loading type, bolt pretensioning force, beam span (Lb), beam depth (db), the number of bolts (n) and the spacing between bolt and weld lines (a). All specimens were analyzed through finite element (FE) method using ABAQUS software. Results indicate that the manual prediction for eb is non-conservative and inaccurate. Although the available equation for ew was able to reasonably predict ew, this equation yields to a non-conservative prediction in most cases. Furthermore, results indicate that among all investigated parameters, only n, Lb and a have remarkable effects on eb and ew. In order to propose alternative equations for eb and ew in terms of effective parameters, artificial neural network was utilized which resulted in predictive models with acceptable level of accuracy. Regarding the weld size, a procedure was proposed in which the weld must fail after the bolt fracture. Results show that, using this procedure, leads to a safer weld design in comparison with the use of AISC Manual suggestion for weld design (w = 5/8tp).

A simplified numerical model to simulate RC beam–column joints collapse

Post-earthquake reconnaissance confirmed that the high vulnerability of non-seismically detailed RC frame structures could be related to shear failure in the core of the beam–column joints which might cause the collapse of the structure. The main focus was given on developing a simplified numerical model to simulate RC beam–column joints collapse based on theoretical formulations and experimental observations. For this, a joint model has been proposed so that nonlinearities in the joint core were considered by two diagonal axial springs. According to the principal stress approach, a more refined calibration of principle tensile stress versus joint rotation relation was developed to calculate the characteristics of these springs. In the model, the effects of the main factors influencing the mechanical behavior of RC joints i.e. column intermediate bars, joint aspect ratio, joint shear reinforcements, type of beam bar anchorage, etc. were considered. To verify the simplified numerical model, it was vastly applied to experimental specimens available in the literature. Results revealed that the model was capable of estimating inelastic response of RC joints with reasonable precision. Furthermore, assuming the joint core to behave as a rigid body, even for joints reinforced by shear reinforcements might bring about non-conservative predictions in terms of strength and ductility capacities. Based on a parametric study, it was also concluded that the effectiveness of the influential factors of RC beam–column joints is noticeably a function of the level of the axial load applied on the column. Using experimentally computed factors and simple procedure to calculate joint characteristics could make the model properly suitable for practical applications.

A validated CFD model of plain and serrated fin-tube bundles

This work presents a Computational Fluid Dynamics model of helically wound fin tube bundles and demonstrates its predictive capability for thermal-hydraulic performance. A consistent validation against experimental data is given for four different fin tube geometries, two with plain fins and two with serrated fins. Predicted heat transfer and pressure drop data are within, or very close to, the experimental uncertainty, with maximum root mean square errors of 13.8% and 14.4% respectively. The modeled fin temperature distribution is used to evaluate three fin efficiency models, revealing that correction equations can be in significant error for tall plain fins. Three sets of semi-empirical correlations for Nusselt and Euler numbers are also evaluated, showing non-conservative predictions for several of the tested geometries. Results from the study confirm the efficacy of reduced domain modeling, whereby geometric periodicity of the heat exchanger array is exploited to reduce computational cost.