Chalmers Open Digital Repository
Not a member yet
26247 research outputs found
Sort by
Numerical simulations of highly perturbed lean hydrogen-air flames. A detailed investigation on strained lean hydrogen combustion characteristics
No full textPremixed or partially premixed fuel configuration with turbulent combustion is widely used in all combustion technologies and applications available in the industry. Though there are many interesting alternatives exist to combustion technologies (e.g., battery technologies), combustion technologies still become relevant as of today because of the existing infrastructure built around the combustion applications and the ease of application. Many popular companies (e.g., Volvo Car, Siemens, etc) and research enthusiasts in Sweden and worldwide are developing interesting strategies to mitigate the global warming, which is one of the downsides of combustion technologies. Among various renewable fuels, hydrogen becomes one of the important carbon free fuels of interest. Also compared to the other fuels, hydrogen fuel has very good combustion characteristics (high laminar flame speed, wide flammability limits, low ignition energy, etc.) and can be produced using various technologies and renewable methods. It is important for researchers to develop efficient combustion models to make the application of hydrogen fuel more practical and adaptable in various applications. This can be done only by using the data obtained by experimental results and Computational Fluid Dynamics (CFD) tools effectively. However, there is not enough research available which can accurately predict the burning rate and combustion characteristics of turbulent hydrogen – air mixture. Hydrogen has a significant high burning rate which is usually controlled by local mixture composition changes due to higher molecular diffusivities of hydrogen air mixture. The most popular approach is based on the hypothesis that the entire turbulent flame regime is controlled by local flamelets which are highly perturbed. This thesis work mainly covers the study of such highly perturbed local flamelets. Eventually, the data can be further used to design or develop hydrogen-air turbulent combustion applications.
CHEMKIN-PRO is the main tool which is used in this project to simulate combustion characteristics of hydrogen – air mixture. CHEMKIN PRO gives flexibility to the user to setup specific boundary conditions based on user needs to obtain a desired output. This project can be divided into two parts, the first part – simulation of laminar hydrogen air flames and the second part – simulation of strained laminar flames. In laminar flame simulations, the main goal is to study the effect of different input configurations like inlet temperature, pressure, diffusion and transport model effects, gird resolution and curvature and different combustion mechanisms on laminar flame speed for different equivalence ratios () ranging from 0.5 to 2.9. It is observed that the flame speed increases rapidly up to =1.5 and then reduces sharply. In the second part, strained laminar flame simulations, the goal is to study the strain rates at which the maximum consumption velocity and the extinction has attained for lean hydrogen air flames ranging from =0.36 0.80, using Oppdif module in CHEMKIN PRO. For strained flames, consumption velocity becomes more relevant as it represents the global burning rate of the mixture. Later, the effects of different mechanisms, inlet temperature, different equivalence ratios, pressure and flame thickness are studied and presented at the end of this thesis. A significant amount of time is also spent on extracting and visualizing the relevant data in MATLAB from the output obtained by CHEMKIN PRO
Enhancing Supply Chain Forecasting with Machine Learning
No full textAccurately forecasting unconstrained demand is essential for effective supply chain planning and inventory management. This thesis examines the performance of both traditional statistical methods and modern machine learning models in predicting unconstrained demand, with a particular focus on comparing global and local modeling strategies across multiple products. Using real-world sales data, models such as ARIMA, XGBoost, and LightGBM were evaluated based on their ability to produce accurate forecasts.
The comparative analysis of forecasting models across multiple products reveals that model performance varies significantly depending on product characteristics and the chosen modeling approach. Overall, modern machine learning approaches, particularly LightGBM and XGBoost, consistently outperformed the traditional ARIMA benchmark in terms of predictive accuracy. However, their effectiveness varied by modeling approach and product type.
The results indicate that local models, especially those based on LightGBM, tend to perform better for products with lower demand, as they are more capable of capturing product-specific patterns and fluctuations. In contrast, global models show stronger performance for high-demand products, where shared patterns across products can be more effectively leveraged. Although global models may not always match the accuracy of well-tuned local models, they offer significant practical advantages, including simplified model management, reduced training time, and better scalability. These findings highlight the trade-offs between accuracy and efficiency, and provide valuable insights into when global forecasting models may serve as viable alternatives to individualized local models
A parametric approach to environmental impact assessment of structural systems in early-stage design: Development of a user-oriented tool for intuitive feedback and informed decision-making
No full textByggbranschen spelar en avgörande roll i att minska utsläppen av växthusgaser i en
tid då klimatfrågan är mer aktuell än någonsin. Den bärande stommen kan stå för
60% av en byggnads utsläpp under byggskedet. Därför är det viktigt att kunna bedöma
miljöpåverkan från olika stomlösningar redan i tidiga skeden av designprocessen, när
möjligheten att påverka är som störst, men osäkerheten också hög. Trots detta inkluderas
konstruktionsperspektivet ofta sent i dagens arbetsflöden, vilket försvårar integrerade
och klimatmedvetna beslut. Det finns därför ett tydligt behov av verktyg som
möjliggör tidig utvärdering av bärande system i samspel med arkitektonisk utformning.
I detta examensarbete har ett parametriskt verktyg för snabb utforskning och jämförelse
av olika stomalternativ i tidiga skeden utvecklats. Verktyget kombinerar parametrisk
modellering, konstruktionsprinciper och livscykelanalys för att möjliggöra en effektiv
och transparent bedömning av klimatpåverkan. Utvecklingen har skett utifrån användarnas
perspektiv med definierade användarprofiler, iterativt prototyparbete och scriptbaserad
implementering. Verktyget är främst riktat till konstruktörer men underlättar
också samarbete med arkitekter och andra aktörer i designprocessen. Användaren kan
generera olika stomlösningar och få sammanställda resultat kring klimatpåverkan och
dimensioner för bärande element. Fokus har varit på användarvänlighet, tydlig visuell
återkoppling och hög prestanda. Verktyget är modulärt uppbyggt för att hantera flexibla
geometrier och dimensioneringen baseras på tabellvärden lagrade i en utbyggbar
databas. Verktyget är baserat på ett antaget system av fritt upplagda bjälklag, balkar och
pelare. En lastnedräkning utförs av verktyget och har verifierats med handberäkningar.
Två fallstudier har genomförts för att vägleda arbetet och utvärdera verktygets praktiska
användbarhet. Den första fokuserade på den parametriska funktionaliteten och samlade
in feedback från handledare inom akademin och branschen. Den andra utgick från ett
verkligt byggprojekt i tidigt skede och genomfördes som en workshop där arkitekter och
ingenjörer samarbetade. Verktygets potential att stödja tvärdisciplinär dialog och snabb
iteration av designalternativ kunde verifieras. Arkitektens användning av resultaten i
kommunikationen med beställaren understryker dess praktiska relevans.
Framtida utveckling kan bygga vidare på det implementerade verktyget och att djupare
studera hur det kan stödja tvärdisciplinärt samarbete och informerade beslut. Potentiell
utveckling av verktyget kan inkludera en utökad databas och optimeringsfunktioner
CFD-Informed Neural Networks for Centrifugal Fan Design: Combining Simulation and Deep Learning to Predict the Performance of Parametrically Varied Fan Blades
No full textThis project presents the development of a surrogate model for centrifugal fan performance,
focusing on the Same Sky CBM-97S series DC fan. The objective is to develop a validated
CFD model of the centrifugal fan and to use the model to train neural networks that can
predict the performance of varying blade geometries. To achieve this a steady state CFD
model was developed in STAR-CCM+ using k − ω turbulence model and a moving reference
frame to simulate the rotating impeller. After validation against experimental data, a
simplified fan geometry was parametrized, and a dataset consisting of 200 simulations with
varying blade count and blade angle of attack was generated.
Two different supervised Neural Networks were then trained: the first predicts the pressure
and velocity fields across a 2D plane section, while the other estimates performance
parameters such as outlet mass flow and pressure based on the geometric inputs. Both
models demonstrated high accuracy. The flow field model achieved R2 values above 0.94.
The fan performance model showed mass flow and pressure predictions with errors below 6%.
The project shows that the neural networks trained on the CFD simulation data are able to
accurately predict a two dimensional flow field and performance variables. Future work may
involve extending to 3D field predictions using a physics-informed neural network, incorporating
additional parameters such as blade length, and optimizing network architecture for
enhanced performance
Modeling and Evaluation of the Olshammar engine. A simulation based approach of a five-stroke turbocharged engine using Siemens Amesim.
No full textInternal combustion engines (ICEs) have been used since the 19th century and remain relevant as of 2025, although the field faces major challenges. The future of ICEs will heavily rely on new inventions and technological advancements. This thesis addresses one such invention: the Olshammar engine, a five-stroke engine concept featuring a lowpressure exhaust cylinder. The purpose of this study is to examine how the Olshammar engine performs in comparison to a conventional four-stroke engine. There are several software tools available for modeling and simulating ICEs, and in this study, Siemens Amesim was used. Initially, a two-cylinder petrol baseline engine was modeled and optimized, providing the reference for modeling the Olshammar engine. The results show that the Olshammar engine reduces brake-specific fuel consumption (BSFC) across the operating range of 2000–5000 rpm, with the largest improvement at 3000 rpm, where BSFC is reduced by approximately 4 %. This corresponds to a fuel conversion efficiency of ηf = 35.2% compared to ηf = 33.7% for the baseline engine. Furthermore, the simulations indicate that the improvement in BSFC can be attributed to the power contribution from the exhaust cylinder, resulting from the recovery of expansion work. Additional assessment of key parameters contributing to performance improvements suggests that tuning the exhaust cylinder offset relative to the combustion cylinders, as well as adjusting the bore-to-stroke ratio, can further enhance overall performance. Siemens Amesim proved to be a well-suited tool for ICE modeling, to the extent that it could replace GTPower in the ICE course at Chalmers in the future. Finally, the findings of this study demonstrate how the design and optimization of the Olshammar engine contribute to improved fuel efficiency compared to conventional four-stroke designs
Framework Insights and Automated Attestation for Software Supply Chain Security
No full textThe escalating threat of software supply chain attacks necessitates robust security measures; however, current guidance is fragmented across numerous, often overlapping, frameworks. This thesis addresses this challenge through a dual approach. First, it conducts a systematic comparative analysis of five prominent software supply chain security frameworks - ESF, S2C2F, SCVS, SLSA, and an academic SOK taxonomy - by decomposing their 284 guidelines into 1,321 atomic, actionable statements. These statements were then thematically labeled and semantically compared to define framework scope, identify consensus areas, reveal gaps, and highlight specialized strengths. The analysis found ESF to be the most comprehensive, while S2C2F, SCVS, and SLSA offer significant depth in specific niches, such as consumption, component verification, and build integrity, respectively. This underscores that no single framework is universally optimal.
Second, this research develops and evaluates a Proof-of-Concept (PoC) system to demonstrate the feasibility of automating compliance attestation. The PoC automatically verifies a targeted subset of decomposed guidelines for selected open-source projects, embedding cryptographically signed conformance attestations - including claims, evidence, and targets - directly within a Software Bills of Materials (SBOM). A companion visualization tool enables human inspection and signature verification of these enriched SBOMs. A feasibility study confirmed the viability of this endto- end process, showcasing a practical pathway for integrating verifiable compliance into the software development lifecycle. Ultimately, this work provides a clearer map of the current guidance landscape and demonstrates a practical path to embedding verifiable compliance, advancing the automation and trustworthiness of software supply chain security
Exploration of AI-Powered Tools and UX Writing Solutions for Improved Content Design Process in a Robotic Software
No full textWhile previous research discuss the advantages of combining UX writing and AI, more research is needed regarding integrating AI in the content design process. This project was carried out in collaboration with ABB Robotics in Gothenburg and
investigates the possibility of integrating AI-powered tools into the content design process for the robotic software RobotStudio. Further, the research has explored what potential solutions there could be for the project members to create consistent and high-quality content for various UI elements. Previous Heuristic Evaluation identified inconsistency in the text and conducted interviews revealed a process of content design that varied, including the difference in AI usage. The study evaluated AI tools for their ability to generate guideline-compliant text, showing that while some AI tools generated high-qualitative content, none of them fully aligned with the provided guidelines, indicating that human oversight remains essential for contextspecific nuances and maintaining consistency. To address the variability in content creation, the study proposes incorporating UX writing guidelines and standardized AI prompts for project members to use. The guidelines were informed by interviews, external and internal company analysis, and user testing. The findings highlight the necessity of balancing automation with human judgment to achieve cohesive and effective UX writing outcomes