Luleå University of Technology Publications
Not a member yet
39356 research outputs found
Sort by
to meet each other, to know and grow, and to have a good time: insights from swedish pupils with intellectual disabilities who participated in philosophical dialogues
No full textThis research project aims to answer the following questions: (1) What experiences do pupils with intellectual disabilities (ID) report after participating in education based on philosophical dialogue? and (2) How do these experiences relate to the needs of pupils with ID? To address these questions, an interview study was conducted with 12 pupils, aged 13 to 15 years, with primarily mild ID, attending the Swedish Compulsory School for Pupils with Intellectual Disabilities. The pupils participated in a small-scale intervention program consisting of 12 sessions of philosophical dialogues over 6 weeks, guided by two experienced facilitators. Toward the end of the program, the children took part in semi-structured interviews to share their experiences of the philosophical dialogues. The responses to these interviews were analyzed in relation to identified needs of pupils with ID, specifically the need for cognitively stimulating activities, communication and social skills, and decision-making skills. Thematic analysis of the data revealed the following themes: To meet each other, to know and grow, and to have a good time. The results indicate that philosophical dialogue holds promise in addressing the identified needs of pupils with ID. Notably, the role of humor in the children’s experiences during the philosophical dialogues was of particular interest.Validerad;2025;Nivå 1;2025-11-06 (u2);Full text license: CC BY</p
Friktion och nötning mellan differentialkomponenter i Scanias elektrifierade drivlina
No full textDetta examensarbete är utfört åt Scania CV AB. Scania som många andra företag håller på att ställa om till elektriska fordon, härnäst är tunga lastbilar på väg. Detta medför nya drivsystem med elmotor vilket påverkar differentialen som detta arbete tittar på. Scania har tidigare haft problem med bland annat differentialen med rotation av sfärisk brickorna i differentialkomponenterna. Detta nöter bort på huset och i värsta fall kan det leda till totalhaveri. Detta område är svårsmort med höga laster, höga temperaturer och låga varvtal vilket är tuffa förhållande för tribologiska kontakter. Vanlig bakaxelolja är korrosiv mot koppar och gör den ej kompatibel med elmotorer. Scania introducerar en ny olja som är kompatibel med elmotorer vilket då blir mindre kompatibel med differentialer. Denna olja är tunnare och ej innehåller additiv som hjälper i de tuffa förhållanden differentialen utsätt för. Detta examensarbete ska utreda friktionen i kontaktytan mellan differentialdrev och bricka. En jämförelse har gjorts mellan EV oljan och bakaxeloljan. Detta har utförts på LTU i tribolab. En kortare studie har utförts av ytfinheten före och efter testade brickor och differentialdrev. Resultatet i friktionskoefficienten visar för EV oljan på 0,01–0,1 och för bakaxeloljan 0,01–0,02. En 4–5 gånger högre friktion vid användning av EV olja
Structural purification of technical lignins via fractional dissolution using non-azeotropic solvent mixtures
No full textTwo technical lignins, a softwood kraft lignin (SKL) and a wheat straw organosolv lignin (WSOSL) were fractionated using a Soxhlet extractor that was connected to a piston pump for solvent movement such that Soxhlet extraction using non-azeotropic solvent mixtures was feasible. Fractionation of the lignins using such solvent mixtures that could be tuned in terms of hydrogen-bond acceptor and donor characteristics and polarities yielded novel fractions not accessible in standard Soxhlet-based fractionations. Two SKL fractions could be obtained applying aqueous acetone that displayed homogeneous structural characteristics while differing significantly in molecular weights. WSOSL could be gradually purified, allowing for the generation of a rather pure lignin carbohydrate complex (LCC) fraction and a purified high molecular weight lignin fraction.Funder: European Union – NextGeneration EU;Fulltext licence: CC BY-NC</p
Structured temporal representation in time series classification with ROCKETs and hyperdimensional computing
No full textTime series classification poses significant challenges due to the inherent temporal order of the data points and the existence of sequential dependencies between them. The ROCKET family, featuring methods like MiniROCKET, MultiROCKET, and HYDRA, is currently a leading approach in this domain, leveraging convolution kernels to aggregate temporal features into encodings for linear classifiers. However, these models encode temporal features over short temporal windows and then aggregate them as an unordered set of encodings over the longer temporal window of the entire data sequence. This prevents these models from capturing any longer sequence structure. To address this design drawback, we propose integrating hyperdimensional computing into ROCKET methods to explicitly incorporate temporal order of the short-term features within the entire time series. This approach enhances the discriminative power of encodings generated by MiniROCKET, MultiROCKET, and HYDRA where longer-term structure exists in the data, leading to increased classification performance with minimal computational overhead. More specifically, we introduce a method to represent time series as high-dimensional vectors through multiplicative binding of ROCKET encodings with encodings representing temporal order, applying this approach across various ROCKET methods. Additionally, we explore different high-dimensional vector representations of temporal order, yielding diverse similarity kernels that enhance classification accuracy. Through experiments on synthetic datasets, we highlight the limitations of ROCKET methods in handling temporal dependencies and show how the methods based on hyperdimensional computing overcome these limitations. Furthermore, our extensive experimental evaluation with real-world datasets included in the recent UCR archive, validates the advantages of our approach, consistently achieving classification improvements across all ROCKET methods that integrate hyperdimensional computing. Notably, our best model achieves a relative error rate reduction of over 50% compared to the best ROCKET model on several UCR datasets.Validerad;2025;Nivå 2;2025-11-24 (u4);Funder: National Research Fund of Ukraine (NRFU)(2023.04/0082);Fulltext license: CC BY</p
Numerical assessment of the influence of residual rock mass properties on destress blasting in deep mine
No full textDestress blasting is a widely adopted technique for mitigating rockburst hazards in deep underground excavations. However, its effectiveness is closely linked to the mechanical degradation characteristics of the rock mass after failure, which are often overlooked in design practices. This study presents a numerical assessment of how residual rock mass properties influence the performance of destress blasting, focusing specifically on residual cohesion, residual friction angle, and critical plastic strain. Three models were developed using the 3DEC code to represent (1) a conventional excavation without boreholes, (2) excavation with relief holes but no blasting, and (3) excavation with relief holes subjected to dynamic loading. Each model was tested under three sets of residual parameters to simulate varying degrees of post-failure degradation. Simulation results show that boreholes without dynamic loading do not induce failure or stress relief, regardless of residual strength conditions. In contrast, when dynamic loading is applied, the extent and continuity of plastic zones, as well as the magnitude of stress redistribution, are significantly influenced by the residual parameters. Lower values of critical plastic strain result in greater post-yield stress reduction in the rock mass and lead to more pronounced stress relief near the excavation face. Additionally, the destressing effects are spatially non-uniform, with greater reductions observed near the center of the drift face where blast influence is strongest. These findings highlight the critical role of post-peak rock behavior in determining the effectiveness of destress blasting. Considering the residual mechanical properties of the rock mass after blast-induced damage is essential for optimizing blast design and enhancing excavation safety in deep mining operations.Full text license: CC BY 4.0;Funder: State Key Laboratory (2011DA105287-FW202409, IDMEKFJJB01); Rut and Sten Brand foundation</p
Advanced machine learning for pillar stress prediction and design optimisation in hardrock platinum mining: enhancing safety and sustainability on the Great Dyke of Zimbabwe
No full textThis study advances pillar stress prediction and design optimisation in hardrock platinum mining on the Great Dyke of Zimbabwe using advanced machine learning (ML) techniques, addressing significant gaps in traditional methods. Utilising Gradient Boosting Machine (GBM), XGBoost, NGBoost, Random Forest, and AdaBoost, the research evaluated a dataset of 503 observed practical insitu pillars, incorporating key features such as Depth Below Surface (DBS), Actual Panel Width, and Actual Extraction Ratio (AER). GBM and XGBoost emerged as top performers, achieving R2 scores of 99.58% and 99.44%, respectively, with GBM exhibiting an MSE of 0.3094 and RMSE of 0.5563. NGBoost added value with predictive uncertainty, enhancing risk management frameworks. The study also highlights feature importance, emphasising DBS, AER, and Actual Pillar Area as critical predictors, ensuring robust and site-specific design solutions. Practical outcomes include a 15% reduction in material overdesign and a 20% improvement in identifying high-risk pillars, contributing to safer and more efficient operations. Integration with real-time monitoring systems enabled dynamic adjustments, reducing pillar failure risks by 30% under evolving conditions. This research, the first of its kind on the Great Dyke, demonstrates the transformative potential of ML in mining engineering, providing a framework for safer, economically viable, and sustainable operations. This study paves the way for leveraging ML to tackle complex geological and geotechnical challenges in global mining projects by addressing predictive accuracy and uncertainty.Validerad;2025;Nivå 2;2025-11-27 (u5);Full text license: CC BY;Funder: University of Johannesburg, South Africa</p
Integrating Sustainability Criteria into a Decision Model for Reverse Logistics in Industrialized Housebuilding: A Field Study Approach
No full textQuestion: How can sustainability criteria be integrated into a decision model to enhance the application of reverse logistics in construction? Purpose: The study explored how economic, social, and environmental sustainability criteria could be integrated into a decision model for reverse logistics design. Research Method: A field study tested a decision model based on multiple-criteria decision analysis at an industrialized housebuilding firm in Sweden. Through theory, observations, interviews, and expert validation, a practical decision model was developed and tested to enhance transport and weatherproofing of modules. Findings: The theoretical framework primarily developed for the engineering and automotive industries can be adapted for reverse logistics in industrialized housebuilding by using a decision model based on multiple-criteria decision analysis. Limitations: Further research is needed to validate and refine the model, and to extend its application to other construction products and processes. The findings may be more specific to the process examined in the field study rather than across the industry. Implications: The criteria of the decision model support decision-making in the weather and transportation protection process, though they can be further refined to provide a more precise basis for decisions. Value for practitioners: Understanding the importance of systematic decision-making in reverse logistics design can improve sustainability and ensure compliance.Validerad;2025;Nivå 1;2025-08-11 (u5);Full text license: CC BY-NC-ND 4.0;Funder: Bo Klok; Ikano Bostad, NCC; Veidekke;</p
A new statistical fracture model for particles in unbound road materials
No full textFracture of rock particles is important in many applications like mining, mineral comminution, unbound granular materials (UGMs) for railway and road structures. The latter application is the main interest presently, as fracture of rock particles in UGMs affects the UGMs performance and may compromise structural integrity of a pavement, potentially leading to premature road failures. Therefore, it is important to assess their resistance to aggregate fracture accurately. In this study, a new statistical fracture model for particle fracture, based on the results of single particle crushing tests, is introduced to investigate aggregate fracture. The proposed model is tested for UGMs composed of three different aggregate types: brick, granite and a volcanic material and its results are compared with other widely used fracture force models. The performance of the models is also investigated by simulating uniaxial monotonic compression tests on UGMs with different aggregate size distributions using the Discrete Element Method (DEM) and comparing the results with experiments. Fracture at two different load levels for three different particle size distributions are investigated for each material. One particle size distribution at one load level is used to identify the contact law parameters for each material, and single particle breakage test are used to identify the fracture force model parameters. The DEM models with a new fracture force model agrees well with the macro-mechanical behaviour observed in experiments and exhibits the highest degree of correlation to fracture results obtained from experiments.Validerad;2025;Nivå 2;2025-10-17 (u8);Full text license: CC BY 4.0;</p
Analytical Health Indices: Towards Reliability-Informed Deep Learning for PHM
No full textDeep learning has demonstrated significant potential for prognostics in complex systems (Fink et al., 2020). Recent advances in physics-informed machine learning have integrated physics-of-failure principles within data-driven models (AriasChao, Kulkarni, Goebel, & Fink, 2022). Beyond physical laws, fleet-level time-to-failure (TTF) distributions provide valuable prior knowledge for individual asset life predictions.In this paper we derive a probabilistic analytical health index(HI) model based on power-law degradation, enabling a probabilistic description that reconciles individual variability with fleet-wide trends. We show that, under Weibull, Gamma, and Pareto-distributed TTFs, the HI evolution follows an analytical form, allowing explicit characterization of time to reach intermediate degradation levels. Therefore, this work provides a theoretical foundation for integrating reliability principles with deep learning, advancing towards Reliability-Informed Deep Learning (RIDL). The approach is validated on synthetic turbofan engine data and real-world battery degradation datasets. This work establishes a rigorous basis for embedding reliability engineering principles into deep learning, improving predictive maintenance and remaining useful life (RUL) estimation.Validerad;2025;Nivå 1;2025-08-15 (u5);Full text license: CC BY 3.0 US;</p
Efficient CO2 capture using deep eutectic solvent-activated carbon slurry systems
No full textThe need for effective CO2 mitigation has led to research into capture technologies using novel solvents, such as deep eutectic solvents (DES), which are known for their high CO2 capture capacity. However, their high viscosity and cost are the main challenges. In this study, we aimed to enhance the CO2 capture capacity of monoethanolamine chloride: ethylene diamine ([MEACl][EDA]) with (1:5) molar ratio, while addressing these limitations by developing a slurry. Water was added as cosolvent, and activated carbon was considered as the substrate for DES immobilization. The resulting slurry system demonstrated enhanced CO2 uptake and improved kinetics, with a viscosity within the same range as that of the aqueous DES. Upon capture, only a slight increase in viscosity was observed, maintaining comparable performance to the aqueous system. The optimal slurry demonstrated a CO2 capture capacity of 26.12 wt% at 22 ℃, with viscosities of 8.14 and 22.82 mPa·s before and after CO2 capture, respectively. The CO2 capture rate of the slurry reached 1.67 mol CO2/(kg sorbent·min) within the first 2 min at 22 ℃, surpassing the rate of the aqueous DES (1.24 mol CO2/(kg sorbent·min)). Furthermore, the prepared slurry maintained ∼90 % of CO2 uptake after five regeneration cycles and depicted almost the same performance after one month.Validerad;2025;Nivå 2;2025-10-02 (u5);Full text license: CC BY 4.0;</p