Chalmers Open Digital Repository
Not a member yet
26247 research outputs found
Sort by
A scaled axisymmetric finite element model for heat flow in ventilated brake rotors
No full textThe aim of this thesis was to develop a finite element model of the heat flow in a ventilated disc brake that is both accurate and computationally efficient. The model is intended to be used during the early stages of development for brake components. It must capture the out-of-plane effects on the non-axisymmetric ventilation layer while remaining two dimensional to allow for fast simulations. To accomplish this an enriched two-dimensional model was developed where each edge and face are scaled based on their size in the out of plane direction. The scaling factor was acquired by doing a geometry mapping of each surface and volume on a three-dimensional model of the brake disc, using it to integrate the weak-form over the out of plane dimension. The scaled axisymmetric model was validated against the result of a high-fidelity three-dimensional model of the brake disc supplied by Volvo Car Corporation.
The difference between the simulation result of the scaled two-dimensional axisymmetric model and the three-dimensional model was only about 1% when considering heating of an insulated wheel. When adding boundary condition pertinent to the convection over all surfaces, the difference in simulation results becomes about 5%. In both simulation cases, the scaled two-dimensional axisymmetric model accurately captures the average temperature distribution of the three-dimensional model.
This accuracy is good considering the difference in computational cost between the models. The scaled two-dimensional model takes minutes to compute while the three-dimensional model takes several hours, which is a good trade off during the early development stages of a brake rotor
Life Cycle Assessment of Organic Waste Treatment by Composting and Anaerobic Digestion for the City of Cuenca – Ecuador
No full textThis thesis explores the environmental performance of anaerobic digestion (AD) and composting as waste management solutions for the organic fraction of municipal solid waste (OFMSW) in Cuenca, Ecuador. Using Life Cycle Assessment (LCA) methodologies, it evaluates key environmental impacts such as global warming potential and photochemical oxidant formation, aiming to identify the most suitable method for reducing the amount of organic waste sent to landfills.
The results indicate that anaerobic digestion has strong potential to reduce greenhouse gas emissions, particularly when biogas is effectively captured and used. On the other hand, composting, while less effective in terms of emission reductions, is more cost-efficient and can play a valuable role in improving soil health. Despite these benefits, both technologies face considerable challenges in the Latin American and Caribbean (LAC) region, where the focus is often still on meeting basic waste management needs, such as establishing sanitary landfills. The high costs and limited development of AD in the region also raise concerns about its viability.
The study highlights the importance of understanding the composition of OFMSW, as this has a direct impact on the efficiency of AD systems. Future research should explore the potential of co-digesting OFMSW with other types of organic waste to boost biogas production. Additionally, factoring in socio-economic aspects—such as the costs of implementation, potential economic gains from energy generation, and effects on local communities—will provide a fuller picture of the sustainability and practicality of these technologies.
Ultimately, the findings suggest that while both AD and composting offer clear environmental benefits, their adoption must be guided by local priorities and infrastructure readiness. Addressing these regional factors is crucial for optimizing resource recovery and minimizing environmental impacts
Developing Effective Long-Haulage Infrastructure in Sweden through the Lens of TEN-T: Insights into stakeholder preferences for truck stop safety, security, service, and alternative fuels
No full textThis thesis investigates how rest and refueling infrastructure for long-haul heavy goods vehicles (HGVs) in Sweden can be developed to better meet regulatory requirements and stakeholder needs, particularly in the context of the Trans-European Transport Network (TEN-T). The study applies a mixed-method approach combining a literature review, expert and haulage company interviews, and a driver survey. Key themes include safety and security, service availability, and the implementation of alternative fuels such as electricity and hydrogen.
Using the Kano model and Multi-Criteria Decision Analysis (MCDA), the research identifies which resting area features add the most value to truck drivers. The findings reveal that basic hygiene and health services, and secure parking features are considered essential. These preferred features differ slightly from the required features that the EU has defined in their safe and secure certification requirements. Moreover, demographic factors such as gender, nationality, and exposure to crime, affect how features are perceived and prioritized. These preferences were compared with qualitative data from expert and haulage company interviews, and literature. Slight variations were found in what the qualitative and quantitative data suggests as important features.
The report presents three development scenarios ranging from a core model focused on driver satisfaction to more advanced modules including alternative fuel infrastructure and a higher degree of safety and functionality. By comparing stakeholder preferences with EU certification criteria and funding opportunities, the study outlines practical and strategic recommendations for infrastructure investments. It concludes that aligning infrastructure development with user needs and funding eligibility is crucial to achieving both operational relevance and long-term sustainability in the Swedish transport sector
Recurrent Neural Networks for Lagrangian Tracking of Bacteria
No full textResearch in microbiology is crucial for development of antibiotics, vaccines and other
medicines that cure diseases and prevent spread of viral infections. One method
of studying microorganisms is Lagrangian tracking, where the movement of single
microorganisms, such as bacteria, is tracked over long periods of time, which is
important when studying for example chemotaxis. Lagrangian tracking has previously
been implemented using deep learning, showing promising results. However,
the model, a convolutional neural network (CNN), struggled to handle overlapping
bacteria, which resulted in failure of entire experiments when the model switched
which bacterium was currently being tracked. This thesis aimed to create a model
for Lagrangian tracking that could accurately track over long periods of time as well
as handle overlapping bacteria. The method included simulation of fluorescence microscopic
data as well as design, training and evaluation of recurrent neural networks
(RNNs) using the simulated data. The results showed that the RNNs gave lower
error distributions and were able to handle overlapping bacteria better compared to
the CNNs implemented for benchmarking. An analysis of the importance of features
of the bacteria for tracking indicated that the tracking was harder when surrounding
bacteria were close to the focal plane or had higher intensity compared to the bacterium
that was currently tracked. Although testing the RNNs in an experimental
setup remains, the results suggest that replacing a CNN with an RNN can improve
the accuracy of the Lagrangian tracking and to greater extent avoid losing the bacterium
during an overlap. In turn, improving the accuracy of Lagrangian tracking
contributes to the possibility of tracking single microorganisms over long periods of
time and gain more knowledge about for example chemotaxis
CFD-Based Sensitivity Study of Flow and Design Parameters in Multiphase Flow Meters. Analyzing the Impact of Variable Conditions on Homogeneity and Measurement Accuracy
No full textIn subsea oil- and gas applications, multiphase flow meters (MPFM) are used to measure volumetric flow rates of oil, gas and water produced from a well, without first separating the phases. When well productions decline the flow may become unstable, requiring additional controls. In these cases MPFMs are useful instruments to detect well instability, blockages and disturbances so that the well can be controlled and stabilized in real time. Variations in flow regime and homogeneity in the Venturi-based MPFM may affect the accuracy of measured volume flow. Additionally, understanding slip velocity between phases is crucial for formulating an accurate slip model to compute phase volume flow rates.
In this project, the sensitivity of the MPFM to flow- and geometrical parameters is studied by modeling and simulating the MPFM using CFD. This gives insight into how MPFM design and operating conditions influence measurement certainty on a macroscopic scale, while also allowing for the investigation of smaller scale phenomena. Time-averaged mean values and periodic behaviors of the flow parameters are evaluated to give insight into the flow behavior and the MPFM sensitivity to flow and design parameters. The homogeneity and mixing of the flow before entering the MPFMs is evaluated, to understand how operating conditions and geometry changes affect the flow characteristics considering MPFM accuracy. Additionally, a
suitable CFD modeling technique is found to aid in the design and development of future MPFMs. The modeling technique is evaluated in terms of accuracy, quality and computational expense.
In this study, a suitable modeling method was identified using Eulerian Multiphase models. These models have high accuracy and is effectively capturing the flow behavior while being computationally efficient. The slip-ratios evaluated showed that for increased pressure and viscosity the liquid film, dispersed, and overall slip decreased while for design changes the slip had more mixed results. The sensitivity study revealed that the MPFM’s sensitivity to geometrical parameters, such as blind-T depth and vertical entrance length, was minimal. Operating conditions, especially pressure and liquid viscosity, play a major role in shaping the flow regime and phase mixing. These factors can significantly affect MPFM accuracy if they are not properly accounted for in the interpretation models
Large Scale Efficient Data Readout for Vehicle Fleets
No full textAs vehicles become more technologically advanced, the data generated by a single vehicle reach significant amounts. Diverse data has high potential in use cases such as machine learning by providing insights into different conditions. Currently, there is no clear solution for collecting this data as vehicle systems are restricted in terms of compute, memory, storage, and bandwidth. This thesis investigates the problem of large scale vehicle data readout and presents a solution to it, providing a significant increase by leveraging lossless streaming based compression at low cost. Furthermore, it addresses the architecture necessary in order to sufficiently process the data globally and how best to integrate this efficiently with a massive number of vehicle systems. Lastly, a generalized model is formulated at the micro scale, which establishes the requirements in terms of compute and memory on a single vehicle system based on the findings presented. At the macro scale, the infrastructure required to support the solution is discussed
Machine Learning for Wind Power Prediction. A Comparative Analysis of Traditional Machine Learning Models and Graph Neural Network for Wind Power Prediction and Forecasting in Wind Farms
No full textThis thesis presents a comparative study of machine learning (ML) models and the deep learning (DL) model graph neural network (GNN) for wind power prediction and short- to medium-term forecasting in wind farms. Using high-resolution SCADA data from a 16-turbine onshore wind farm in Sweden, along with re-analysis and forecast weather datasets, various models including Random Forest (RF), eXtreme Gradient Boosting (XGBoost), k-nearest neighbours (kNN), Multi-Layer Perceptron (MLP), and GNN were trained and evaluated. Two baseline approaches, the farm’s
theoretical power curve and FLORIS wake model, were used as references. Results show that ML models outperform baseline models in predicting wind power output, with GNN achieving the best overall performance, although all ML models perform similarly. The ability of the models to generalize from wind power prediction to forecasting is however limited. The findings indicate that re-analysis data with low spatial resolution fails to adequately capture local weather conditions necessary for accurate power prediction. The study also investigates the effects of input feature selection, temporal resolution, and multi-task learning on model performance.
Furthermore, it identifies challenges related to input data quality, particularly in the estimation of global wind conditions from SCADA-based measurements. These results underscore the potential of ML methods for wind power applications and highlight the critical importance of accurately representing global weather data, as well as accounting for discrepancies between training data and forecast data
How the distance between the fan outlet and the surrounding walls affects the fan efficiency. An Investigation of the Influence of Outlet-Wall Distance on Fan Efficiency using Computational Fluid Dynamics
No full textThis thesis investigates the influence of the outlet-wall distance on the aerodynamic performance of centrifugal fans used in Air Handling Units (AHUs). The study was conducted using steady-state Computational Fluid Dynamics (CFD) simulations with a realizable k–ε turbulence model and Moving Reference Frame (MRF) methodology. The aim was to identify how variations in the height, width, and depth of the outlet enclosure affect fan efficiency and static pressure rise.
A mesh independence study was first carried out to ensure result validity, followed by a systematic variation of outlet geometry. Results showed that increasing the outlet enclosure dimensions led to improved flow uniformity, reduced turbulence, and enhanced fan efficiency. However, the improvements diminished beyond a certain geometric threshold, indicating an optimal enclosure size for performance. Additional modifications such as cone extensions, cylindrical blockages, and rounded corners were also tested and showed further performance gains.
The results offer design guidance for optimizing outlet-wall distances in spaceconstrained AHU systems. While the findings are based on numerical simulations, they lay the groundwork for future experimental validation and further CFD studies using transient solvers
Optimization of Fan Blade Design Using CFD and Reinforcement Learning
No full textThe design of fan blades has undergone significant advancements over the past century. However, it is important
to consider whether conventional design approaches may unintentionally constrain the range of possible blade
geometries. This paper investigates the potential of integrating Machine Learning and CFD to further improve
and accelerate the fan blade design process.
To achieve this, the study focuses on two main tasks: developing an accurate numerical model of a
centrifugal fan in STAR-CCM+ and creating a Reinforcement Learning (RL) framework which implements the
STAR-CCM+ model to optimize the fan blade geometry. CFD simulations were performed using the k − ω
SST solver, and a mesh convergence study was performed. The RL framework for blade optimization was based
on the Deep Q-Network (DQN) algorithm, implemented in Python using Pytorch.
The CFD Model validation was carried out by comparing the performance curve obtained from STAR-CCM+
simulations with the manufacturer’s fan curve. The results indicate that the developed model accurately
predicts the flow field generated by the fan.
The static pressure rise across the fan serves as the primary performance metric for evaluating design
improvements. The Reinforcement Learning (RL) approach successfully produced new and improved blade
designs in each iteration. However, none of the generated designs outperformed the original fan blade. Despite
this, the approach shows strong potential for improving blade design given more time and computational
resources.
For future studies, additional methods can be incorporated to better evaluate blade design. For instance,
investigating other Reinforcement Learning methods or alter the current environment to reduce its design
constraints
Abaqus Based Modelling of Ball Bearings
No full textThis thesis investigates whether deep groove ball bearings can be accurately modeled in Abaqus by replacing rolling elements with non-linear springs. The main focus is on the transfer of radial, axial, and combined loads. Existing analytical stiffness models, such as those by Harris and Hamrock, were found inadequate for general use, prompting the development of a new stiffness model. This was achieved through a curve-fitting approach that correlates bearing geometry (e.g., ball diameter, pitch diameter, number of balls, radial play) with displacement data from reference simulations derived using the commercial gearbox software SABR. The final model showed an error margin of less than 5% for most load cases and misalignments. The outcome suggests that this method provides a flexible and accurate tool for simulating bearing stiffness, with potential for broader implementation and further refinement