Department of Construction

This study quantifies how beam-to-column joint damage alters the axial resistance and failure modes of cold-formed steel (CFS) uprights used in industrial storage rack systems. Current design-oriented assessments typically neglect the reduction of joint rotational stiffness and moment resistance caused by prior cyclic joint deterioration, which can compromise subsequent compression performance. To address this gap, six commercial perforated upright sections are tested under axial compression in intact and joint-damaged conditions. A unified framework combines experiments, nonlinear finite element analysis (FEA; ANSYS), and machine learning (ML) prediction using geometric descriptors (thickness, area, second moments of area, and radii of gyration) together with experimentally derived joint properties (rotational stiffness and ultimate joint moment). Among the evaluated regressors, the gradient boosting model provides the best predictive performance (cross-validated R2 and lowest root mean square error [RMSE]/mean absolute error [MAE]), showing close agreement with experimental outcomes and numerical trends. Results indicate axial capacity reductions of up to similar to 30% after joint degradation and a tendency toward coupled distortional and torsional-flexural (TF) instability. The findings highlight the engineering need to account for joint stiffness degradation in residual capacity checks and support rapid, data-informed decisions for inspection and retrofit of rack uprights. The study provides a practical predictive framework for postdamage capacity evaluation in industrial rack systems.

The development of a reliable analytical model for industrial steel pallet rack connections has long been a goal for researchers, as practitioners still rely primarily on costly and time-consuming experimental testing to assess structural behavior. The present study addresses this gap by developing an analytical framework that enables the prediction of moment-rotation behavior without extensive testing. The framework is calibrated for a specific upright profile within defined parameter ranges and is intended as a configuration-specific predictive tool to reduce reliance on testing within the validated geometric space. Eighteen specimens with varying connection configurations commonly used in practice were tested under monotonic loading in accordance with EN 15512. Major failure modes and the influence of configuration on behavior, particularly in terms of strength and stiffness, were identified. A four-parameter Richard Abbott analytical model was successfully calibrated using experimental data, achieving exceptional predictive accuracy, with R-2 > 0.94 across all configurations. Parametric equations expressing model parameters as functions of beam depth and tab count were developed, enabling direct prediction of behavior. A local parametric sensitivity analysis identified initial stiffness and reference moment as the most influential parameters, with average sensitivities of 16.3% and 18.7%. The validated model significantly outperforms previous approaches, offering a reliable design tool that removes the need for configuration-specific testing.

Macro-models based on equivalent diagonal struts are the only practical option for the structural-scale analysis of masonry-infilled reinforced concrete (RC) frames, but their parameters, above all the effective strut width, are empirical, were calibrated almost exclusively on solid units and cannot be measured directly; experimental calibration is scarce because each point requires a full-scale frame test. This article proposes simplified micro-modelling (SMM) of perforated block masonry as the technique that supplies them. Perforated clay blocks are replaced by equivalent solid units expanded by the mortar thickness, the joints are reduced to zero-thickness cohesive interfaces with tensile and shear yield criteria, and the continuum is governed by a concrete damage plasticity (CDP) law. Replacing the perforated unit removes the constraint its thin internal shells impose on the explicit time increment, while the expanded-unit properties come from small prism and wall-panel tests or from Eurocode 6 expressions and are used without user subroutines. The technique reproduces the capacity, stiffness and observed failure mechanisms of two quasi-static tests on perforated clay block infilled RC frames to within 10%-12%; a third was reproduced with the same technique in earlier work. The validated model is then the reference from which an equivalent-strut macro-model is calibrated: three parallel compression-only struts on a fibre frame in OpenSees. On Specimen 13 the envelope is identified from the measured backbone and the model returns the peak base shear to within 0.6% with a whole-curve error of 2.1%; in every validation case the envelope is generated from the masonry properties and one quantity only is fitted, the dimensionless strut width. Calibrated on a single one-bay frame, that parameter transfers unchanged to two- and three-bay frames with a mean absolute error of 2.5% in peak base shear and a maximum of 3.1%, verified on complete curves rather than peaks, and is conservative by 8%-16% when carried to six-storey frames. The calibrated width increases with the quality of the masonry, from 0.193 to 0.294 of the panel diagonal, whereas a width defined through a stiffness parameter alone falls from 0.152 to 0.129 over the same range; the three levels examined vary strength, modulus and strains together and cannot separate them. Column shear links following expression A.12 of EN 1998-3 show the configuration with the smallest drift to be the one whose columns fail, so for frames of these proportions the governing limit state is local; a sensitivity study over seven distributions of the infill thrust among the struts, and over +/- 20% on every calibrated quantity, leaves that conclusion standing and identifies the strut width as the one quantity it depends on. The reduced-scale study then delimits what a similitude comparison can demonstrate: a strut macro-model contains no length scale and satisfies the similarity relationships identically, whereas the micro-model departs from them by up to 4.5%, and that departure is the measurable size effect.

The study examined the impact of incorporating pumice and rubber obtained from waste tires as aggregate in specific ratios on geopolymer concrete (GPC). For this aim, various combinations and replacement percentages of pumice and waste tire rubber (WTR) were tested in GPC. Pumice was replaced with coarse aggregate in varying proportions of 0%, 5%, 10%, and 15% and WTR was added into mixture with same proportions. A total of 144 GPC specimens were cast and tested. Tests were performed to measure the compressive strength (CS), splitting tensile strength (STS), flexural strength (FS) and dynamic drop-weight tests. Evaluations of the setting time and workability were also conducted. When natural coarse aggregate was substituted with pumice in concrete, the strength of the resulting mixtures fell. The second research varied the percentage of WTR in the concrete mixture from 0%, 5%, 10%, and 15% while keeping the pumice percentage stable. The study showed that when the pumice ratio was set as constant, there was a steady loss of strength as the percentage of WTR increased. The experimental results demonstrated that partial replacement of coarse aggregate with pumice (up to 15%) led to a progressive reduction in both CS (up to 18.17%) and STS (up to 37.13%) of GPC. This decline in strength became more pronounced with the concomitant utilization of WTR. The findings demonstrated that as the WTR content elevated from 0 to 15%, there was a consistent decline in FS across all levels of pumice replacement, with reductions ranging approximately between 20 and 26%. Additionally, the study created an empirical formula related to STS and CS and compared with the formulas previously developed by various researchers. Scanning electron microscopy (SEM), Energy dispersive x-ray (EDX), and thermal gravimetric analysis (TGA) are also conducted. According to the data gathered from the impact tests, adding pumice to concrete has a negative impact on its impact behavior; on the other hand, adding 10% WTR has a favorable impact.

As concrete is one of the most commonly used construction materials, there is a massive production of cement, which causes cement manufacturing to be an energy-intensive industry. A significant amount of the cost of cement production, ranging from 20% to 25%, is attributed to thermal energy. In addition, the action of mining and burning fossil fuels results in the unfavorable emission of hazardous compounds into the environment. Therefore, the switch from conventional fossil fuels to alternative fuels (AFs) in the cement manufacturing business has attracted attention due to environmental and financial concerns. In this paper, four commonly used AFs are discussed, which are waste tires, municipal solid waste, meat and bone meal, and sewage sludge. It is found that each AF has a unique calorific value and properties, attributed to its source, treatment, and technology. Furthermore, the availability of AF is important as the amount varies depending on the location. In addition, their effects on gaseous emissions from the cement plant and the quality of clinker are found to be inconsistent. Thus, there will not be a single best type of AF option to be used in the cement industry. A good AF should be able to provide sufficient thermal energy while reducing the environmental impacts and costs. A careful analysis and multicriteria decision-making approach are always vital when employing AFs in order to prevent environmental problems, cost increases, as well as clinker quality degradation.