Forschungsdatensätze der BAM
Filtern
Dokumenttyp
- Forschungsdatensatz (18)
Sprache
- Englisch (18)
Referierte Publikation
- nein (18)
Schlagworte
- Ultrasound (4)
- Artificial neural networks (3)
- Concrete (3)
- Nondestructive testing (3)
- Validation (3)
- Automation (2)
- Elastic wave (2)
- Non-destructive testing (2)
- Online NMR spectroscopy (2)
- Process industry (2)
Organisationseinheit der BAM
- 8 Zerstörungsfreie Prüfung (18) (entfernen)
X-Ray computed tomography (XCT) scan of 11 individual metallic powder particles, made of (Mn,Fe)2(P,Si) alloy. The data set consists of 4 single XCT scans which have been stitched together [3] after reconstruction. The powder material is an (Mn,Fe)2(P,Si) alloy with an average density of 6.4 g/cm³. The particle size range is about 100 - 150 µm with equivalent pore diameters up to 75 µm. The powder and the metallic alloy are described in detail in [1, 2].
WEBSLAMD
(2022)
The objective of SLAMD is to accelerate materials research in the wet lab through AI. Currently, the focus is on sustainable concrete and binder formulations, but it can be extended to other material classes in the future.
1. Summary
Leverage the Digital Lab and AI optimization to discover exciting new materials Represent resources and processes and their socio-economic impact.
Calculate complex compositions and enrich them with detailed material knowledge. Integrate laboratory data and apply it to novel formulations. Tailor materials to the purpose to achieve the best solution.
Workflow
Digital Lab
Specify resources: From base materials to manufacturing processes – "Base" enables a detailed and consistent description of existing resources
Combine resources: The combination of base materials and processes offers an almost infinite optimization potential. "Blend" makes it easier to design complex configurations.
Digital Formulations: With "Formulations" you can effortlessly convert your resources into the entire spectrum of possible concrete formulations. This automatically generates a detailed set of data for AI optimization.
AI-Optimization
Materials Discovery: Integrate data from the "Digital Lab" or upload your own material data. Enrich the data with lab results and adopt the knowledge to new recipes via artificial intelligence. Leverage socio-economic metrics to identify recipes tailored to your requirements.
Data set of low-field NMR spectra of continuous synthesis of nitro-4’-methyldiphenylamine (MNDPA). 1H spectra (43 MHz) were recorded as single scans.
Two different approaches for the generation of artificial neural networks training data for the prediction of reactant concentrations were used: (i) Training data based on combinations of measured pure component spectra and (ii) Training data based on a spectral model.
Synthetic low-field NMR spectra
First 4 columns in MAT-files represent component areas of each reactant within the synthetic mixture spectrum.
Xi (“pure component spectra dataset”)
Xii (“spectral model dataset”)
Experimental low-field NMR spectra from MNDPA-Synthesis
This data set represents low-field NMR-spectra recorded during continuous synthesis of nitro-4’-methyldiphenylamine (MNDPA). Reference values from high-field NMR results are included.
SLAMD-FIB-Case-Study
(2022)
With 8% of man-made CO2 emissions, cement production is an important driver of the climate crisis. By using alkali-activated binders part of the energy-intensive clinker production process can be dispensed with. However, because numerous chemicals are involved in the manufacturing process here, the complexity of the materials increases by orders of magnitude. Finding a properly balanced cement formulation is like looking for a needle in a haystack. We have shown for the first time that artificial intelligence (AI)-based optimization of cement formulations can significantly accelerate research. The „Sequential Learning App for Materials Discovery“ (SLAMD) aims to accelerate practice transfer. With SLAMD, materials scientists have low-threshold access to AI through interactive and intuitive user interfaces. The value added by AI can be determined directly. For example, the CO2 emissions saved per ton of cement can be determined for each development cycle: the more efficient the AI optimization, the greater the savings. Our material database already includes more than 120,000 data points of alternative cements and is constantly being expanded with new parameters. We are currently driving the enrichment of the data with a life cycle analysis of the building materials. Based on a case study we show how intuitive access to AI can drive the adoption of techniques that make a real contribution to the development of resource-efficient and sustainable building materials of the future and make it easy to identify when classical experiments are more efficient.
This dataset contains raw data acquired in ultrasound measurements on a reference specimen made of concrete at Bundesanstalt für Materialforschung und -prüfung (BAM), Berlin (Germany). The internal specimen identifier is “Pk266”. The measurements were conducted using the pulse-echo method. The upper surface of the specimen was defined as measuring area. The aim of the measurements is to determine both the geometrical dimensions (thickness) and the position of tendons to the measuring area. In addition to this, a second dataset of a second specimen with identifier is existing named “Pk050” has been acquired. Pk050 has the same geometrical dimensions and concrete recipe as Pk266 recipe but does not contain tendons [Reference: https://doi.org/10.7910/DVN/9EID5D].
This dataset contains raw data acquired in ultrasound measurements on a reference specimen made of concrete at Bundesanstalt für Materialforschung und -prüfung (BAM), Berlin (Germany). The internal specimen identifier is “Pk050”. The measurements were conducted using the pulse-echo method. The upper surface of the specimen was defined as measuring area. The aim of the measurements is to determine the geometrical dimensions (thickness) of the specimen “Pk050”. In addition to this, a dataset of a second specimen with identifier “Pk266” has been acquired. Pk266 has the same geometrical dimensions and concrete recipe as Pk050, but contains tendons [Reference: https://doi.org/10.7910/DVN/NUU0WZ].
AI-reflectivity is a code based on artificial neural networks trained with simulated reflectivity data that quickly predicts film parameters from experimental X-ray reflectivity curves. This project has a common root with (ML-reflectivity)[https://github.com/schreiber-lab/ML-reflectivity] and evolved in parallel. Both are linked to the following publication:
Fast Fitting of Reflectivity Data of Growing Thin Films Using Neural Networks A. Greco, V. Starostin, C. Karapanagiotis, A. Hinderhofer, A. Gerlach, L. Pithan, S. Liehr, F. Schreiber, S. Kowarik (2019). J. Appl. Cryst.
For an online live demonstration using a pre-trained network have a look at github.
This data set contains three different data types obtained from concrete specimens. For each specimen, the rebound numbers, ultrasonic data (ultrasonic velocity, time of flight), and destructive concrete strength are given. Two kind of specimen geometries were tested: cubes and drilled cores. The files are labeled according to the specimen geometry as "cube" or "core" and the type of measurement data as "compressive_strength", "rn_R" and "rn_Q" for rebound numbers as well as "us" for ultrasonic data. The ultrasonic data were generated by six independent laboratories, the rebound numbers by five independent laboratories and the destructive tests by one laboratory. The designation of each specimen establishes the relationship between the different data types.