13979 research outputs found
Sort by
Functional DSL Compilers with Leightweight Proofs
No full textAn essential component of programming language design, is to walk the trade-off between re-using as much as possible from already existing languages while carefully questioning core design choices to overcome fundamental problems. To investigate compilation techniques for high-level programming languages, we present a series of compilers covering a wide range of language features. This thesis is structured into two parts: (i) compiling and enforcing protocols, and (ii) intermediate representations and optimizations.
We begin the first part, by comparing two emerging trends in programming languages for distributed protocols – choreographic and multitier programming. Choreographic and multitier programming are both global programming models, in the sense that a single program specifies the behavior of all participants in the distributed protocol. To investigate similarities and differences, we present an algorithm to translate a restricted version of choreographic programming to multitier programming, and vice versa. Then, we present our own global language called Prisma, for writing programs that run partially on a smart contract. Besides splitting the source program into a program for the smart contract and a program for its clients, Prisma also enforces the control flow of the contract even in the presence of clients who deviate from the protocol.
In the second part, we investigate intermediate representations and optimizations. In functional compilers, A-normal form is a common intermediate representation, whose design can be derived from the concept of laws of monads. However, monads are usually associated with sequentiality and bringing evaluation into a total order. We present a notation par-seq that compiles to monadic operations to express dependencies between terms, and applicative functor operations to express independent terms. Then, we present the indexed A-normal form (AiNF), an extension of A-normal form for array operations, based on the idea that arrays are dual to functions, and show how it can create optimization opportunities in combination with partial evaluation and common subexpression elimination. Finally, we present the density compiler repot. It takes a probabilistic program as input and then searches for a program to compute the corresponding probability density function to output. A key part of this search is the use of conditional A-normal form (canf) to perform program inversion
Pionierin ohne Sogwirkung. Vor 100 Jahren schloss an der TH Darmstadt erstmals eine Frau ihr Studium mit Ingenieur-Titel ab
No full textAm 18. Juli 1913 verlieh die Technische Hochschule Darmstadt zum ersten Mal den Titel Diplomingenieur an eine Frau. Von einer damaligen Zeitung wurde sie als „Fräulein Ingenieur“ und die erste Dame bezeichnet, die in Deutschland diesen Titel erhalten habe. Der Bezeichnung Ingenieurin war 14 Jahre nach der Einführung des Dipl.-Ing. wohl noch zu ungewöhnlich, um Jovanka Bontschits auch so zu nennen
Neubauten für die TH Darmstadt in der NS-Zeit
No full textBetrachtet man die historischen Bauten der Technischen Universität, so fällt der Blick an einigen Gebäuden auf die Überreste der nur teilweise entfernten NS-Staatssymbole. Sie sind bis heute sichtbare Spuren der Erweiterung der TH Darmstadt in der Zeit der NS-Diktatur und prägen im Herrngarten, am Ballonplatz und auf dem Griesheimer Sand auf markante Weise das Erscheinungsbild der TU
Two-photon polymerization for inertial fusion energy target fabrication
No full textThis study focuses on optimizing the fabrication of foam targets for laser-driven inertial fusion energy using two-photon polymerization (2PP). 2PP offers nanoscale precision, enabling the fabrication of foams with precise control over structure, strut thickness, and pore size distribution. Due to its repeatable and deterministic fabrication, 2PP is a promising technique for studying crucial target parameters on the path towards a fusion power plant. However, 2PP faces challenges such as stitching errors in large prints, data limitations, and extended printing times, which limit it to single-piece production. We investigate the impact of foam structures and printing parameters on the foam density and demonstrate the successful fabrication of a foam-shell target for the proton fast ignition approach. By optimizing the printing path, we achieved a reduction in file size by one order of magnitude and printing time by a factor of four for the fabrication of a hexagonal honeycomb structure. Additionally, this research identifies and addresses key limitations and provides a detailed outlook for future improvements
Physical modeling of conjugate heat transfer for multiregion and multiphase systems with the Volume-of-Fluid method
No full textThe Volume-of-Fluid (VOF) method is commonly used for numerical simulations of phase change phenomena, such as nucleate boiling or droplet evaporation. A key issue with the standard VOF method is the averaging of the liquid and vapor properties in interface cells, which causes non-physical conjugate heat transfer with a solid wall. Therefore, we aim at a physical model for conjugate heat transfer between a solid and a multiphase fluid. The first measure for higher quality simulations is the splitting of the single temperature field in the fluid region into separate liquid and vapor temperature fields. The second measure is the development of a new, more physical temperature boundary condition for conjugate heat transfer between a solid region and a multiphase fluid, based on experimental results, theoretical models and theoretical considerations. In interface cells, the vapor phase is excluded from the conjugate heat transfer because only heat transfer to the liquid phase occurs resp. dominates. Additionally, the conjugate heat transfer between solid and liquid in the interface cells is performed with virtual subcells, depending on the respective volume fraction of the liquid phase. This new approach (we name it distinctive approach) is successfully validated for energy conservation, and stability issues are discussed for the first time. Significant differences to simulations with averaged properties are observed due to the (now) physically correct modeling of conjugate heat transfer. In our boiling cases, the more accurate numerical simulations lead to considerably larger bubble growth rates. Higher quality simulations are also expected for nearly all applications, where there is a three-phase contact line, be it vapor bubbles in nucleate boiling or droplets impacting on a heated surface
Large Eddy Simulation of the Piston Boundary Layer Evolution During the Compression Stroke in a Motored Internal Combustion Engine
No full textThis work examines the momentum boundary layer evolution on the piston top of the Darmstadt optically accessible Internal Combustion Engine (ICE). For this purpose, a 3D-CFD wall-resolved Large Eddy Simulation (LES) under motored conditions was deployed. The piston wall is resolved down to 25 µm, corresponding to y⁺ < 1 . For statistical purposes and to compare with experimental data, 33 consecutive engine cycles are simulated. A large-scale tumble motion characterizes the flow field. This flow impinges on the piston on the exhaust side, it moves along the flat piston wall and detaches on the intake side. The near-wall velocities of the simulations align well with the experiment. Analysis revealed regions of Favorable Pressure Gradient (FPG) on the exhaust side and Adverse Pressure Gradient (APG) on the intake side, separated by a sharp pressure inversion zone. The near-wall flow accelerates and then decelerates until detachment. Analysis of the non-dimensional u⁺ - y⁺ profiles reveals the absence of a logarithmic region in the boundary layer. This scaling procedure is sensitive to thermo-physical properties like density and viscosity that vary across the boundary layer, which complicates comparisons with canonical studies. The shape factor of the boundary layer suggests a fully turbulent state despite the low momentum thickness-based Reynolds number. The boundary layer height increases from the exhaust towards the intake side, especially in the presence of strong pressure gradients. Pressure gradients acting perpendicular to the boundary layer are observed. The comparison of ensemble-averaged and single-cycle instantaneous data shows high levels of cyclic fluctuations
On the gluing and interlocking of asphalt on concrete
Full text linkPaving asphalt on existing concrete is very common for mass transit roads, bridges, and highway constructions. For this, detailed knowledge of material behavior and bond interactions is vital for understanding the adhesion mechanism of these construction methods. A thorough comprehension of the bond mechanism is highly significant for assessing the service life of such traffic pavements. In this study, the influence of gluing and interlocking on the bond layer was examined through a close-range measurement method consisting of photogrammetry evaluated by a digital image correlation method. This testing method allowed a detailed detection of local failure points at the bond layer between asphalt and concrete. Tests were carried out on different asphalt-concrete combinations where the combination with a milled base showed the largest strain absorption at failure. Test specimens with bitumen emulsion exhibit higher strains at the boundary layer before failure than those without bitumen emulsion. In addition, the test specimens with a milled grounded substrate achieved higher shear values than those with a smooth substrate. Based on these experimental findings a finite element model was developed to show the effects of the local stress distribution and degree of gluing on the layer bond behavior, which opens possibilities for material optimization
Relevance of Local Dispersion on Mixing Enhancement in Engineering Injection and Extraction Systems in Porous Media: Insights from Laboratory Bench-Scale Experiments and Modeling
Full text linkThis work investigates the dynamics of flow, transport and mixing in subsurface porous media during an engineered injection–extraction (EIE) system. We perform laboratory bench-scale experiments mimicking an EIE system in an unconfined aquifer, and we explore the role of local dispersion on mixing enhancement. The experimental setup is equipped with four wells operated in a sequence, one at a time, creating transient flows and a fluctuating water table impacting the transport dynamics of an injected dye tracer plume. A high-resolution imaging technique is applied to monitor the spatial and temporal evolution of the plume concentration. The experiments are performed in porous media with fine and coarse grain sizes and considering two different sequences of injection and extraction. The plume spreading and mixing are quantified by computing the spatial moments and the plume area, respectively. The Okubo–Weiss parameter is calculated over the plume area to correlate mixing enhancement with changes in flow topology. The results indicate that the operation of EIE system significantly enhances mixing and spreading, particularly when the effective Okubo–Weiss parameter is higher. Furthermore, the mixing enhancement is larger in the experiments performed in the coarse porous media, indicating the importance of local dispersion as a factor for mixing enhancement in EIE systems
Adoption drivers and barriers of digital freight transport platforms—An intermodal case study
No full textIncreasing environmental pressure urges firms to decarbonize their supply chains by reducing emissions caused by freight transport. This puts intermodal freight transport (IFT) on the agenda. IFT combines the ecological advantages of rail transport with the flexibility of road transport. However, it increases supply chain complexity by creating additional interfaces between the actors involved. This hampers efficiency and calls for automation through digital platforms. By contextualizing the Technology-Organization-Environment (TOE) framework and applying a multiple-case study approach, we aim to investigate why users opt for or against adopting IFT platforms and how adoption can be fostered. Among 30 adoption factors identified, we find that sellers of IFT services fear increased market transparency and interface standardization through platforms, while demanders of IFT services favor these attributes. We contribute to the extant literature by providing a nuanced understanding of the underlying decision rationales from the perspectives of platform users and providers and derive nine levers suitable to increase platform adoption and, hence, supply chain automation
Towards High-performance and Trusted Cloud DBMSs
No full textCloud Database Management Systems (DBMSs), such as cloud-native analytical or serverless DBs, are experiencing rapid growth in adoption due to their flexibility and scalability. However, recent incidents with cloud providers show that the traditional model of a trusted provider/admin no longer applies to protect the customers’ data. One promising solution that can prevent a sole reliance on cloud and database service providers are trusted execution environments (TEEs). While past TEEs had many limitations and caused high performance overheads, recent work shows that the support of TEEs like Intel SGX for DBMS workloads improved significantly. Thus, it is time to actively integrate TEE technologies into cloud DBMSs to achieve better security that does not rely on the cloud provider. In this paper, we discuss directions for how recent TEEs can be used to build efficient and secure databases. We summarize the recent results on Intel SGX’s performance for DBMS workloads and lay out the remaining research challenges that must be addressed to achieve optimal performance and thus minimize the performance cost for additional security