Infrastruktur
Filtern
Erscheinungsjahr
Dokumenttyp
- Vortrag (1007)
- Zeitschriftenartikel (553)
- Beitrag zu einem Tagungsband (390)
- Posterpräsentation (176)
- Buchkapitel (50)
- Sonstiges (24)
- Forschungsbericht (23)
- Beitrag zu einem Sammelband (19)
- Dissertation (17)
- Forschungsdatensatz (11)
Sprache
- Englisch (1501)
- Deutsch (779)
- Mehrsprachig (9)
- Tschechisch (2)
- Französisch (1)
- Polnisch (1)
Schlagworte
- Korrosion (126)
- Concrete (118)
- Corrosion (102)
- Ultrasound (56)
- Monitoring (55)
- Potentialfeldmessung (49)
- Structural health monitoring (48)
- Beton (43)
- Alkali-activated materials (40)
- LIBS (39)
Organisationseinheit der BAM
- 7 Bauwerkssicherheit (1213)
- 8 Zerstörungsfreie Prüfung (606)
- 2 Prozess- und Anlagensicherheit (328)
- 8.2 Zerstörungsfreie Prüfmethoden für das Bauwesen (275)
- 7.2 Ingenieurbau (260)
- 3 Gefahrgutumschließungen; Energiespeicher (250)
- 7.6 Korrosion und Korrosionsschutz (245)
- 7.5 Technische Eigenschaften von Polymerwerkstoffen (233)
- 2.1 Sicherheit von Energieträgern (231)
- 7.4 Baustofftechnologie (176)
Paper des Monats
- ja (26)
The carbonation resistance of alkali-activated materials (AAMs) is a crucial parameter for their applicability in concrete construction, yet the parameters influencing it are insufficiently understood to date. In the present study, the carbonation resistance of alkali-activated concretes with varying fractions of ground granulated blast furnace slag (GGBFS) and calcined clay (i.e., high, intermediate, and low Ca contents) were assessed under natural and accelerated conditions. Corresponding hardened AAM pastes were studied using X-ray diffraction, thermogravimetry, Raman microscopy, and mercury intrusion porosimetry. The carbonation resistance of the concretes at natural CO2 concentration depended principally on their water/(CaO + MgOeq + Na2Oeq + K2Oeq) ratio. The remaining variability for similar ratios was caused by differences between the pore structures of the AAMs. For concrete with favorable water/(CaO + MgOeq + Na2Oeq + K2Oeq) ratio and pore structure, the carbonation resistance was comparable to that of Portland cement concrete. The relationship between carbonation coefficients obtained under accelerated and natural conditions differed for concretes with high and low fractions of calcined clay, indicating that accelerated carbonation testing is less suitable to study the carbonation of low-Ca AAMs.
Although numerous investigations study the improvement of flame retardancy of epoxy resins using additives, maintaining the flame retardant (FRs) modes of action present in the resins upon transfer to composites is challenging. In this study, ammonium polyphosphate (APP) and inorganic silicate (InSi) are loaded at 10%, 30%, and 50% by weight, in a diglycidyl ether of bisphenol A (DGEBA) resin cured with dicyandiamide and transferred to bidirectional (BD) glass fiber (GF) composites. Although a 50% loading of the FRs impacts the curing kinetics of the resin system, the effect on the glass transition temperature of the resin system remains negligible compared to reactive FRs in the state of the art integrated into the resin's chemical structure. Increasing the FR content improved the heat release characteristics in both the resins and composites. However, the charring mode of action is completely suppressed in the formulation with 10% APP + InSi. A 30% concentration of the FRs restored the charring action in the composite and the GFs provide increased protective layer action upon transfer to the composites. This study highlights the importance of accounting for the changing dynamics related to processing and flame retardancy upon transferring FRs from resins to composites.
Die zerstörungsfreie Prüfung von Eisenbahnschienen auf betriebsbedingte Schädigungen wird mit Schienenprüfzügen mittels Ultraschall- und Wirbelstromprüfverfahren durchgeführt [1]. Die Auswertung der Prüfdaten erfolgt überwiegend manuell, die eingesetzte Software unterstützt die Auswertenden lediglich durch eine Vorauswahl relevanter Anzeigen. Die Überprüfung der Ergebnisse erfolgt anschließend vor Ort mittels handgeführter Prüfgeräte. Instandhaltungsmaßnahmen werden auf Basis der Befundung vor Ort abgeleitet.
Ziel des durch das Bundesministerium für Digitales und Verkehr (BMDV) im Rahmen von mFUND unter dem Förderkennzeichen 19FS2014 geförderten Vorhabens AIFRI (Artificial Intelligence For Rail Inspection) ist es, durch den Einsatz von KI-Methoden den Automatisierungsgrad des Prüfprozesses von der Auswertung der Daten bis hin zur Planung von Instandhaltungsmaßnahmen zu erhöhen. Die Genauigkeit der Fehlerdetektion soll gesteigert werden, um eine automatisierte Einstufung der aufgefundenen Anzeigen in Risikoklassen zu ermöglichen. Hierfür werden Daten sowohl von Wirbelstromprüfungen als auch Ultraschallprüfungen verwendet.
Im Rahmen des IT-orientierten Projektes werden relevante Schienenschädigungen und in der Schiene vorhandene Artefakte analysiert und in einen parametrierbaren digitalen Zwilling übertragen. Mit diesem digitalen Zwilling werden virtuelle Schädigungsbilder generiert, mit denen KI-Algorithmen auf die Defekterkennung und -klassifizierung trainiert werden. Insbesondere werden hierbei Synergien genutzt, die durch die Verknüpfung von Daten der Wirbelstrom- und Ultraschallprüfung bei einer gemeinsamen Bewertung entstehen. Mit Hilfe von Zuverlässigkeitsbetrachtungen werden die entwickelten und trainierten Algorithmen hinsichtlich der Detektion und Charakterisierung von Schienenschädigungen bewertet.
Im Verlauf des Projektes soll mit dem entwickelten IT-Werkzeug ein Demonstrator aufgebaut und im Feld mit realen Datensätzen getestet werden
Bias Identification Approaches for Model Updating of Simulation-based Digital Twins of Bridges
(2024)
Simulation-based digital twins of bridges have the potential not only to serve as monitoring devices of the current state of the structure but also to generate new knowledge through physical predictions that allow for better-informed decision-making. For an accurate representation of the bridge, the underlying models must be tuned to reproduce the real system. The updated model can be eventually used for the extension of the service life of the bridge based on an accurate description of the structure. Nevertheless, the necessary assumptions and simplifications in these models irremediably introduce discrepancies between measurements and model response. We will prove that quantifying the extent of the uncertainties generated by said discrepancies provides a better understanding of the real system, enhances the model updating process, and creates more robust and trustworthy digital twins. Among others, we identify that the inclusion of the explicit bias term through a Bayesian inference framework corrects the tuned parameters, allows the identification of non-prescribed noise sources and enables the introduction of additional information in the system without modifying the simulation model. The performance of selected model bias identification approaches will be compared in the context of digital twins of bridges. The different methods will be applied to a representative demonstrator case based on the Nibelungenbrücke of Worms. The findings from this work are englobed in the initiative SPP 100+, whose main aim is the extension of the service life of structures through monitorization and digitalization, especially through the implementation of digital twins.
The changes in the use and maintenance of track systems poses new challenges for the periodic mechanized in-service testing of rails using ultrasound and eddy current. The methods currently applied have been used since decades with only minor changes. To face the new challenges generated by modern drive systems, higher speeds, heavier loads adapted techniques have to be developed to detect new defect types and artefacts generated by new production methods. Especially the area where rolling contact fatigue takes place is under focus.
Going beyond the standard conventional ultrasound setups used since the 1950 enables a more detailed detection and classification of rail defects and size estimation. Eddy current methods are applied for surface crack detection and head check depth quantification at the gauge corner of railway tracks. An extension of the tested zone to the running surface uncloses rail defect signal types other than head checks to be detected and estimated in type and size.
For the automated evaluation of the recorded data algorithms based on artificial intelligence being trained based on simulation will be applied. Typically, the testing parameter vary depending on the track condition and the probe wear. To identify variables and parameters which have a significant influence on the overall performance of the test run modelling of the setup can be used.
Actual developments will be presented in this talk
Die mechanisierte zerstörungsfreie Prüfung von Eisenbahnschienen auf betriebsbedingte Schädigungen wird mit Schienenprüfzügen mittels Ultraschall- und Wirbelstromprüfverfahren durchgeführt. Die Auswertung der Prüfdaten erfolgt durch Auswerter, die Überprüfung der Ergebnisse erfolgt am identifizierten Schienensegment vor Ort mittels handgeführter Prüfgeräte. Instandhaltungsmaßnahmen werden anschließend auf Basis der Befundung der zerstörungsfreien Prüfung abgeleitet. Im Rahmen eines vom BMVI geförderten Verbundvorhabens im Programm „Digitale, datenbasierte Innovationen und Ideen für die Mobilität 4.0“ (mFUND) soll dieses Konzept modernisiert und weiterentwickelt werden.
Ziele des Vorhabens AIFRI (Artificial Intelligence For Rail Inspection) sind es, durch den Einsatz von KI-Methoden den Automatisierungsgrad des Prüfprozesses zu erweitern, die Genauigkeit der Fehlerdetektion zu erhöhen und eine automatisierte Einstufung der aufgefundenen Anzeigen in Risikoklassen zu ermöglichen. Weiterhin wird auf dieser Basis ein risikobasiertes Instandhaltungskonzept erarbeitet, welches zukünftig das bisherige präventive Vorgehen ablösen kann.
Im Rahmen des IT-orientierten Projektes werden relevante Schienenschädigungen und in der Schiene vorhandene Artefakte analysiert und in skalierbare Modelle übertragen. Mit diesen Modellen werden virtuelle Schädigungsbilder generiert, mit denen KI-Algorithmen auf die Defekterkennung und -klassifizierung trainiert werden. Mit Hilfe von Zuverlässigkeitsbetrachtungen werden die entwickelten und trainierten Algorithmen hinsichtlich der Detektion und Charakterisierung von Schienenschädigungen bewertet.
Im Verlauf des Projektes wird mit dem entwickelten IT-Werkzeug ein Demonstrator aufgebaut und im Feld mit realen Datensätzen getestet.
Das Projekt wird durch das Bundesministerium für Verkehr und digitale Infrastruktur (BMVI) im Rahmen von mFund unter dem Förderkennzeichen 19FS2014 gefördert.
The creation and use of Digital Twins of existing structures, such as bridges, implies precise digital replicas that accurately mirror their physical counterparts. Ensuring the trustworthiness of Digital Twins and facilitating informed decision-making necessitates a robust approach to Uncertainty Quantification (UQ). A suitable model-updating scheme is key in preserving the quality and robustness of simulation-based Digital Twins. Model bias, stemming from discrepancies between computational models and real-world systems, poses a significant challenge in achieving this goal. This study delves into the challenges posed by model bias within Bayesian updating of Digital Twins of bridges. Two alternative model bias identification methods —a modularized version of Kennedy and O’Hagan’s approach and another one based on Orthogonal Gaussian Processes — are evaluated in comparison with the classical Bayesian inference framework. A key innovation lies in the modification of the aforementioned approaches to incorporate additional information into the Digital Twin framework via the bias term. This enables the extension of the model non-intrusively, leveraging large pools of data inherent in Digital Twins. The study showcases the potential of this approach to correct predictions, quantify uncertainties, and enhance the system with previously untapped information. This underscores the importance of everaging available data within Digital Twins to identify deficiencies and guide potential future model improvements.
The RILEM TC 304-ADC has set up a large interlaboratory study on the mechanical properties of 3D printed concrete (ILS-mech). The study was prepared in 2022 by a preparation group leading to a Study Plan which the TC approved on 29 November 2022 (https://doi.org/10.14459/2023mp1705940). The ILS-mech was performed in 2023. The data was collected using a pre-prepared spreadsheet template. For data management, a database was derived and set-up in openBIS. The underlying Postgres database of openBIS was exported to the here-published SQLite database for sharing without maintaining a server. The structure of the database is described in (doi). The results are discussed in three associated papers focusing on the overall outcomes and evaluation of the procedures (doi), the compressive test results (doi), and the tensile test results (doi).
Optimization Framework for Powder Bed 3D Printed Concrete Structures under Various Constraints
(2024)
The fusion of 3D printing with concrete has transformed large scale construction, accelerating creative design possibilities. In recent years, multiple printing techniques have been developed and are still under investigation. The powder bed technique enables overhang constructions, thereby increasing design flexibility. To maximize the advantages of those printing techniques, a holistic approach is required taking topology optimization through finite element analysis into account. The challenge is to consider a range of constraints, from material to geometrical to manufacturing.
The current research focuses on the development of a holistic approach to large-scale 3D powder bed printed concrete structure optimization, navigating the in-tricacies of constraints, leveraging the advantages of powder-based printers, and integrating essential information from material tests. A workflow starting from geometry input and ending with printing instructions is derived, minimizing user interaction, and offering fast and reliable results for different structures. Starting from a geometric model of the global design with the required load cases, the geometry is optimized with respect to reduce the mass and thus CO2 emissions. Thereby, the topology optimization includes various constraints: stress con-straints (mitigating tension and limit compressive stress using Drucker-Prager based models), geometric constraints (preserving specific surfaces from optimi-zation), and manufacturing constraints (controlling member size and avoiding trapped powder). Different objective functions, like minimizing compliance or mass are considered in this study. The process receives input from upstream ma-terial tests, offering insights into strength, stiffness and possible anisotropy. This integration of material testing enhances the accuracy and reliability of the opti-mization, aligning the design with real-world behavior of 3D printed concrete. An important aspect of the workflow is the segmentation of the optimized global structure into substructures, aligning with the size limitations of the 3D powder-bed printer. As final output of the developed workflow, printer specific fabrica-tion instructions are generated for each substructure. The developed optimization framework is demonstrated in the design of a multiperson shading element.
Extrusion based 3D concrete printing (3DCP) is a growing technology because of its high potential for automating construction and the new possibilities of design. In conventional construction methods, a sample is taken to be representative for one material batch. However, in 3DCP continuous mixing is used which results in variations during the mixing process. Therefore, one sample is not representative for the entire structure. This leads to the necessity of continuous and real-time process monitoring.
This study focuses on the variations of pressure and temperature which are caused by changes in the material due to the ongoing mixing process. Changes in material, which is transported downstream, are influencing sensor signals in different positions with a time delay. In the following, the data is analysed to investigate if the changing material and the so caused change in pressure can be used to calculate volume flow.