Chalmers Open Digital Repository
Not a member yet
26247 research outputs found
Sort by
Suspension vertical and longitudinal modeling and parameter estimation for ride comfort assessment
No full textRide comfort is a key vehicle attribute, strongly influenced by suspension design
and vital to both perceived quality and long-distance comfort. Setting meaningful
requirements early in development is challenging, as engineers must work with aggregated vehicle parameters. Existing high fidelity simulation tools, while powerful,
typically require detailed component data, limiting their usefulness in early-stage
development. At the same time, increasing pressure from electrification and competition makes it more important than ever to define clear ride requirements early, to
downstream effectively.
To address this challenge, this thesis investigates whether a simplified modeling
approach can aid early-stage ride comfort target setting, focusing specifically on
impact harshness. A modular quarter-car model was developed that includes both
vertical and longitudinal dynamics, along with a simplified tire model to represent
force distribution in both directions. Tailored for the concept phase, the model was
structured around system-level inputs rather than detailed component data. To capture the relevant high-frequency dynamics, vertical parameters were identified using
4-poster shaker rig data through nonlinear optimization. Kinematics & Compliance
(K&C) rig data supported both the estimation process and partial parameterization
of the longitudinal model. Data from three different vehicles were used to ensure
the model’s robustness across varying suspension setups. The model was evaluated
by simulating a cleat test and comparing its outputs with corresponding test track
measurements.
The nonlinear optimization of vertical parameters allowed the model to capture
ground-to-body acceleration with an average fit exceeding 80 percent across all
vehicle axles. Impact harshness metrics predicted from cleat simulations showed
deviations of up to roughly 50 percent across different axles, with some larger discrepancies in specific cases. These differences were primarily linked to the simplified
tire model’s limited ability to capture high-frequency dynamics
Diskret matematisk modell av humant glioblastom och dess invasionsmönster i grå och vit hjärnsubstans
No full textDetta arbete undersöker huruvida implementering av en mekanism för interaktion mellan celler
och deras omgivning kan förbättra en modell för tillväxt av hjärntumörer av typen glioblastom
(GBM). Modellen som tagits fram är diskret och stokastisk, och är baserad på versionen av
lattice gas cellular automata som formulerades av Tektonidis m.fl. Tillägget består i en fallenhet för celler att vandra längs vit hjärnsubstans före grå sådan, vilken integrerats genom
segmentering av verkliga hjärnor. Modellen bygger på ett antal parametrar knutna till sannolikhet. Dessa optimerades genom tillämpandet av approximativ bayesiansk beräkning (eng.
approximate Bayesian computation; ABC).
Både den ursprungliga och den vidareutvecklade modellen har jämförts med MR-bilder och
segmenteringar av riktiga GBM från två olika patienter. Datan med vilken simuleringarna
jämförts kommer från datamängden LUMIERE. Jämförelsen gjordes med hjälp av Jaccardindex, och resultaten tyder på att den utvecklade modellen kan prestera bättre. Dessa resultat
bör dock betraktas med försiktighet, eftersom de till följd av begränsningar i datan och de
biologiska antaganden som gjorts är förenade med viss osäkerhet. Detta diskuteras i rapporten.
Arbetet bidrar till att lägga grunden för vidare utveckling av modeller som tar hänsyn till den
påverkan omgivningen har på hur GBM växer
Leveraging Large Language Models for advanced analysis of crash narratives in traffic safety research
No full textFree-text crash narratives recorded in real-world crash databases have been shown to play a significant role in improving traffic safety. But they remain challenging to analyze at scale due to unstructured writing, heterogeneous terminology, and uneven detail. The development of Large Language Models (LLMs) offers a promising way to automatically extract information from narratives by asking questions. However, crash narratives remain hard for LLMs to analyze because of a lack of traffic safety domain knowledge. Moreover, relying on closed-source LLMs through external APIs poses privacy risks for crash data and often underperforms due to limited traffic knowledge. Motivated by these concerns, we study whether smaller open-source LLMs can support reasoning-intensive extraction from crash narratives, targeting three challenging objectives: the travel direction of the vehicles involved in the crash, identifying the manner of collision, and classifying crash type in multivehicle scenarios that require accurate per-vehicle prediction. In the first phase of the experiments, we focused on extracting vehicle travel directions by comparing small LLMs with 8 billion parameters (Mistral, DeepSeek, and Qwen) under different prompting strategies against fine-tuned transformers (BERT, RoBERTa, and SciBERT) on a manually labeled subset of the Crash Investigation Sampling System (CISS) dataset. The goal was to assess whether models trained on a generic corpus could approach or surpass the performance of domain-adapted baselines. Results confirmed that fine-tuned transformers achieved the best accuracy; however, advanced prompting strategies, particularly Chain of Thought, enabled some LLMs to reach about 90% accuracy, showing that they can serve as competitive alternatives.
For the second and third tasks, to bridge domain gaps, we apply Low-Rank Adaption (LoRA) fine-tuning to inject traffic-specific knowledge. Experiments on the CISS dataset show that our fine-tuned 3B models can outperform GPT-4o while requiring minimal training resources. Further analysis of LLM-annotated data shows that LLMs can both compensate for and correct limitations in manual annotations while preserving key distributional characteristics. The results indicate that advanced prompting techniques and fine-tuned open-source models prove effective in large-scale traffic safety studies
Optimisation of plasma enhanced chemical vapour deposition for silicon nitride photonics using optimal design of experiments
No full textThere have been many advances in silicon nitride based integrated photonics, enabling
a variety of interesting applications. The deposition of high quality and low
loss silicon nitride (SiN) during the fabrication of waveguides is essential for creating
useful devices. Low temperature alternatives to low pressure chemical vapour deposition
(LPCVD) such as plasma enhanced chemical vapour deposition (PECVD)
are required to enable back-end-of-line (BEOL) integration.
The difficulty with optimising PECVD deposition of silicon nitride is that it includes
many process parameters that affect the deposition and resulting properties of the
film. Thus, optimisation of the PECVD recipe requires a strategic approach to the
experimental design.
In this work, a technique called optimal design of experiments (DoE) is used to
obtain an overview of the PECVD factor’s influence on silicon nitride properties,
and to predict the optimal PECVD recipe. With the help of the statistical software
JMP the optimal DoE for PECVD deposition of SiN created, significant factors
and correlations identified, and optimal PECVD factor combinations predicted. Ellipsometry
measurements provide data regarding the responses of interest, namely
thickness uniformity, refractive index, and extinction coefficient of the SiN film.
It is found that the refractive index is correlated with the ammonia gas flow rate and
the extinction coefficient. Most PECVD factors appear to be relevant, in particular
the ammonia gas flow rate for the refractive index and the extinction coefficient,
and the frequency mode for the thickness uniformity. In addition, two-factor interaction
effects are present in the PECVD process and affect the relationship between
process parameters and responses. Furthermore, in terms of stoichiometry, silicon
nitride films with a refractive index close to two are found to be silicon rich. The
testing of predicted recipes shows that the three responses, i.e. refractive index,
extinction coefficient and thickness uniformity, cannot be optimised simultaneously
to the desired outcome. However, when considering a stoichiometric Si/N ratio instead
of the refractive index, predicted recipes result in improved and equally good
responses compared to default PECVD and LPCVD recipes, respectively.
This approach of optimal DoE shows promising potential and could be interesting
to explore further, especially regarding the optimisation of other relevant fabrication
steps affecting the propagation losses in waveguides
Kriterieviktning för optimerad lokalisering av distributionslagersplats/crossdock för LTL-transporter: AHP och TOPSIS mot COG
No full textSummary
In modern supply chain management, the strategic placement of distribution warehouses
plays a critical role in minimizing transport costs and environmental impact, especially in
industries such as automotive manufacturing where supplier networks are complex and Just-In-Time logistics are central.
This thesis explores how mathematical decision-making models
can be used to optimized warehouse location, focusing on Less-Than-Truckload (LTL)
transport scenarios involving multiple suppliers and one final destination.
The study applies two Multi Criteria Decision Analysis (MCDA) methods, Analytical
Hierarchy Process (AHP) and Technique for Order of Preference by Similarity to Ideal
Solution (TOPSIS) and compares them with the more traditional Center of Gravity (COG)
method. A case study was conducted in collaboration with the automotive company Aurobay,
using real-world supplier and transport data to evaluate potential warehouse locations.
Through a combination of literature review and semi-structured expert interviews, four
quantifiable and relevant criteria were identified and weighted using AHP: proximity to
suppliers, proximity to logistic hubs, access to renewable energy and labor costs. These were
then applied in a TOPSIS analysis for the case study, where six European cities were
evaluated as potential warehouse locations.
The findings in the study show that methods like AHP and TOPSIS provide a more extensive
basis for decision-making by incorporating both economical and sustainability-related
factors. In contrast, the COG method which only accounts for geographical location and
weight, fails to reflect the full complexity of strategic logistics planning. The comparison
reveals differences in recommended locations depending on the chosen method. The results
emphasize that modern warehouse location decisions benefit from combining quantitative
modeling with expert judgement. The presented approach is adaptable and can support
companies in similar industries aiming for cost-efficient and environmentally conscious
supply chain decisions.
This Bachelor’s Thesis is written in Swedish
Sahlgrenska at Home: Improving Communication and Collaboration for Effective Program Development
No full textThis study investigates how various areas at Sahlgrenska University Hospital collaborate and
communicate in relation to the Sahlgrenska at Home model. The aim is to provide an overview
of how these factors operate both within individual departments and across departments. Addi-
tionally, the study evaluates how Sahlgrenska can learn from the practices of other international
hospitals that offer similar forms of care. Through interviews with people from the different areas
of Sahlgrenska, the main problems related to collaboration and communication around the service
were identified. Furthermore, inspiration and guidelines could be obtained through interviews
with external people from the Northern Ireland Hospital and Medtronic. The findings and rec-
ommendations that address the problems focus mainly on strategy, structure, processes, rewards,
and people. In order to achieve improved cooperation between different departments, motivation,
trust, and commitment are required from the employees. In addition, a clear structure is needed
where roles and responsibilities are defined. Improved communication can possibly be achieved
through a centralized communication platform. But also through regular meetings, where ongoing
feedback between teams and employees will help continuously refine the service. The study also
provides guidelines for continued work with Sahlgrenska at Home
Konstruktion av skjutrigg till 155mm - L52
No full textThis project focuses on the design of a firing rig intended for the test firing of highly loaded artillery components such as barrels, breech blocks, muzzle brakes, and recoil dampers. Initially an information-gathering phase was conducted by collecting requirements and needs from engineers at the company with experience in previous design projects. Their input from them formed the foundation of the work. The company aims to achieve a more efficient testing process by using a stationary firing rig that enables test firing of the recoiling components in complete assemblies. With such a rig, it becomes easy to mount heavily loaded components onto the rig, conduct test firings, and remove them for inspection. This setup makes it possible to stock pre-tested components that can be used by the production department during the manufacturing of artillery systems. The goal of the project was to use multiple simulations as decision support to arrive at a final concept that meets all requirements of the rig. Based on the project background, the objective was to design a stationary firing rig with a focus on structural strength to withstand the forces generated by the firing impulse. The rig should provide a good working height and accessibility, creating a practical and safe working environment for easy assembly and disassembly during test firing. The development of additional equipment for the rig, such as cradle extensions, split yokes, and foldable yoke supports, is not covered in this report.
The result of the project consists of a solution for a firing rig constructed from rectangular beam profiles with dimensions of 250x250 mm and a wall thickness of 15 mm. The rig is equipped with square and rectangular plates to allow stable and secure attachment between the firing rig and the track plate, specifically adapted to the dimensions of the track plate. A plate is welded on top of the firing rig with an interface adapted to the trunnion fork and bearing. The rig is fixed to the track plate using clamping brackets and a mechanical stop. A structural strength analysis using Ansys confirmed that the final design meets the requirements to withstand the stresses encountered during test firing, ensuring long service life, safety, and good ergonomics during firing
Várdobáiki - Landscape Model
No full textVárdobáiki Sami Center, located in Evenskjer, Norway, serves as a cultural hub dedicated to preserving and promoting Sami heritage. The center is housed in a renovated former shopping center, featuring a distinctive facade with colored wooden slats in red, green, and yellow. Inside, the center offers a gallery, museum, language center, and course rooms, all designed with subtle Sami elements and materials, creating a cohesive and inviting atmosphere
Characterization of Natural Killer and Innate Lymphoid Cell Immunity in Ovarian Cancer and Endometriosis
No full textOvarian cancer (OC) is the most lethal gynecological cancer, due to discovery at late stages, vague symptoms and yet no effective screening methods. A common side effect of OC is the buildup of peritoneal fluid or ascites, which contain a large number of immune cells and tumor cells and spheroids. Moreover the carcinoma is heterogeneous and includes various types of tumors with distinctly varying molecular pathways. Endometriosis is a common benign inflammatory disease and is highly associated with certain subtypes of ovarian cancers, including endometrioid and clear cell OC. However, the mechanism behind the transition from endometriosis to OC is not fully understood, but previous research suggests genetic mutations associated with immune escape might play a role.
This project explores the immune landscape in high grade ovarian cancer, with emphasis on phenotypes of natural killer (NK) cells and innate lymphoid cells (ILCs) in ascites and matched peripheral blood mononuclear cells (PBMCs). Using spectral flow cytometry with clustering and dimensionality reduction, distinct immune cell populations were identified. Two populations were found that were significantly enriched in ascites, including one clear population of NK cells with clear tissue resident properties recognized by high expression of markers CD49a, CD103 and CD69 as well as of NK cell receptors, including NKG2A which aligns with previous results in the research group.
In parallel, six NK cell receptor-related SNPs were analyzed using public GWAS datasets for endometriosis and endometrioid and clear cell OC. The results showed slight significance for SNPs in genes associated with the NK cell receptors NKG2A, NKp30, and HLA-B-21. These results may have relevance in the susceptibility for ovarian cancer and impaired NK cell regulation in endometriosis and ovarian cancer highly associated with endometriosis.
These findings further deepened the understanding of variances in immunity in ovarian cancer and endometriosis and could in the future potentially identify immunotherapeutic targets in tissue resident populations in ascites fluid