Luxembourg National Research Fund

Abstract

Gasoline Direct Injection engine technology will continue to play a major role in passenger car powertrain for many years to come. The fuel injection systems have a direct impact on the engine efficiency, noise generation, emissions, but also the system durability. Following the industry 4.0 megatrend to digitalize engineering processes the proposed project aims to develop a complex real-world physics modelling tool for the simulation of the complex multiphase flow in a high-pressure fuel pumps. This tool will help to reduce the needs of prototype hardware manufacturing and tests in product engineering and support knowledge-driven product development. The complex flow phenomena to be considered for such pumping systems are cavitation accompanied by degassing, pressure waves that may interact with the aforementioned cavitation phenomena and thermal dynamic effects like e.g. fuel heating by compression and John Thomson cooling at fuel expansion.

The development and assessment of related injection system today still mainly relies on thermo hydraulic testing of prototypes. Problems being solved by trial and error approaches increasing tremendously the testing expenses as well as the development time. Even thought significant progress in the related physical modelling has been achieved in recent years most of the related models are not yet considered in fluid dynamic codes or other engineering tools because of the lack of validation to real engineering problems.

The proposed project is to develop a practical CFD simulation methodology by integrating the recent physics modelling progress available in literature into a commercial software ANSYS Fluent for the analysis of the above-mentioned complex multiphase flow phenomenon in a high-pressure GDi fuel pump in a full cycle operation. A homogeneous mixture model consisting of fuel liquid, vapor, and non-condensable gas, will be applied for the description of the multiphase/multicomponent flow. The compressibility of all phases will be considered. The turbulence modelling will adopt the URANS approach based on the k-omega SST model in order to be feasible for daily industrial simulation. A previous Fluid Model implementation for thermal hydraulic analysis of injector flow will be revised for the pump flow simulation.

A cavitation model implementation using a Equation of State for Barotropic fluids and a cavitation erosion index developed from a recent PhD program at BorgWarner will be extended to include the thermal dynamic effects. The cavitation model validation will be based on the fundamental test case data available in the literature and based on the critical cavitation point measurement with fuel temperature variations. For the cavitation erosion modelling, the erosive power index will be evaluated for the pump flow together with other index parameters.

The prediction results will be compared with micro-visualization of the pump erosion damages from durability test. The CFD model will be coupled to a 1D system model to account for the pressure wave impact. A feasible moving mesh model setting will provide as a starting point for this project. Overset mesh technique, 6 degree of Freedom (6DOF) and adaptive mesh refinement will be introduced in order to deal with and the valve opening and closing operations. The degassing modelling will be developed within the project.

The validated calculation methodology supports the developments of new high-pressure fuel injection pumps by allowing to pre-calculated the performance of different prototypes thus minimizing the testing effort and cost. In addition, the simulation tool can be applied to investigate the local flow details in the pump to foster the deeper understanding of the complex physics of the flow allowing to further develop the related technology and enhance the performance.

Abstract

Thermoelectric devices have been a subject to intensive investigation in recent years, thanks to rapid progress in novel materials development and discoveries of contributing topological effects. The Smart Heat FluxSensing Platform project (SELENA) aims to develop a smart heat flux measurement solution using anisotropic mono-crystalline as well as multilayer composite materials.In collaboration with private sector partners, the project will investigate in depth the underlying physical phenomena and practical platform application, assuring its relevance to both basic research and industry. By combining numerical and experimental methods SELENAaims to deliver a complete physical model, novel sensing platform design and a scalable manufacturing process description, to forge a path for further adoption and development of thermoelectric devices.

Abstract

Electric circuit boards used in the expanding 5G communication market, rely on the use of high-quality copper foils to ensure reliable high-rate data transfers. It is a high value market, where a Luxembourgish copper foil manufacturer is one of the leaders. For 5G applications, it is necessary to manufacture foils with ultra-flat surfaces to ensure low signal loss due to skin effect occurring at high frequencies. Up to now, CFL carried out corrosion and adhesion promoting treatments using hazardous substances and performed careful wastes treatment. To remain competitive, with 6G technology in sight, and to satisfy the constantly evolving environmental considerations and regulations about substances handling, the use of promising environmentally friendly plasma technologies is considered for copper surface treatment with high control over the treatment parameters.

ICT & Space: 3 projects 

Abstract

Car maintenance has traditionally been driven by OEM-centered maintenance programs, designed in a way that vehicles of a given type with a given engine should follow scheduled preventive maintenance events in order to guarantee a high level of availability and longevity. However, it is known that for a given vehicle, being used in different conditions and intensities may make those pre-scheduled events come either too early or, in the worst case, too late, leading to breakdowns with their associated loss of productivity. Throughout this project, we will conduct research on the usage of connected car and augmented telematics data to predict vehicle health status and potential breakdown events. Through an industrial collaboration, we intend to design and develop a number of predictive models for different vehicle systems and subsystems, integrating them into the industrial partner’s platform for real usage.

Abstract

The use of autonomous and connected shuttles is a great solution to complement existing transport services, proven to be flexible and easily deployable, with over 60% of the world’s test sites being in Europe. However, the services that are relying on these vehicles are usually constrained to fixed, restrictive routes. COMBO focuses on the optimal integration of these new connected and autonomous shuttles to complement the existing, more conventional transport services. A multidisciplinary approach combining optimisation, networks and AI for digital twins will be adopted, with the aim of proposing a decision support system to better plan the future operations of such shuttles. The expected result is to drastically facilitate the creation of flexible services for the transport of passengers but also of goods, to ensure first and last mile delivery. This project is a collaboration between LIST and CFL. Real experiments using shuttles deployed in the Belval campus are planned.

Abstract

BGL BNP Paribas gets many operational benefits from the automated services they built using Machine Learning (ML) technologies. However, a challenging issue is that the effectiveness of their ML systems declines over time due to distribution drifts. Data scientists then have to manually re-engineer the ML system, which requires considerable human and computational effort. The objective of TIMELESS is to develop automated solutions to accommodate the distribution drifts.

Our solutions lean on three research innovations: (1) data selection methods to make ML systems quickly adapt to distribution drifts through sample-efficient re-training, (2) data augmentation methods to make systems resistant to noise preemptively, and (3) an ensemble architecture to enable the combination of alternative models.

These three innovations will together contribute to a drastic improvement of the generalization ability of the ML systems that BGL developed and, in turn, increase the systems’ lifespan.

Abstract

Immunotherapy, in particular immune checkpoint inhibitors (ICI), represent one of the major recent advances in oncology. Since the microbiome has been identified as a potential underlying cause for resistance to ICI, researchers are currently developing innovative strategies to harness the microbiome to improve the response rate to ICI. Diet is a known microbiome regulator and may therefore play a significant role in therapy efficiency by modulating the immune system. However, dietary guidelines are poorly implemented in today’s cancer treatment plans. Here we will use a dietary fiber supplement, ImunoBran®, as a tool compound to modulate the microbiome. We will perform a clinical study in cancer patients undergoing ICI, as well as use humanized mouse cancer models to understand the effect of a fiber supplement on ICI therapy efficiency. Altogether, our study will help identify the benefits of fiber supplements for cancer patients undergoing ICI treatment.

Abstract

Diet is known to have a major impact on our overall health. In recent years, however, it has become clear that the bacteria inhabiting our gut, collectively called the “gut microbiome”, largely mediate the impact of certain dietary components. The microbiome directly breaks down specific food components to produce molecules that influence the integrity of the intestinal mucus barrier and the function of our immune system. Given that various diseases are linked to defined changes in the microbiome composition, it is reasonable to use diet as a lever to influence the microbiome and its effects on our health. Nonetheless, the use of diet to alter the microbiome inpatients with microbiome-related diseases is largely unrealized because we still do not know which bacterial targets within the microbiome are the most relevant for influencing disease progression. In this project, we aim to identify and isolate these bacteria and develop dietary strategies to modulate them for our benefit.