| Other contributing persons, institutions or organisations | dfg - Funder | |
| Other contributing persons, institutions or organisations | Müller-Pabel, Michael - ILK (ORCID: 0000-0003-3624-3242) - Researcher | |
| Other contributing persons, institutions or organisations | Grüber, Bernd - ILK (ORCID: 0000-0002-8504-2095) - Researcher | |
| Other contributing persons, institutions or organisations | Gude, Maik - ILK (ORCID: 0000-0003-1370-064X) - ProjectLeader | |
| Person(s) who is (are) responsible for the content of the research data | Preiß, Gina - ILK | |
| Person(s) who is (are) responsible for the content of the research data | Löffler, Markus - DCN | |
| Person(s) who is (are) responsible for the content of the research data | Koch, Ilja - ILK (ORCID: 0000-0002-8122-4158) | |
| Person(s) who is (are) responsible for the content of the research data | Geske, Vinzenz - ILK | |
| Person(s) who is (are) responsible for the content of the research data | Danczak, Marek - ILK | |
| Person(s) who is (are) responsible for the content of the research data | Gude, Maik - ILK (ORCID: 0000-0003-1370-064X) | |
| Description of further data processing | For analyzing the micro- and meso-structure of EPP bead foams Trainable WEKA Segmentation provided by the open source package FIJI is used. Three classes are defined: "cell walls", "cell volume" and "background". The classes are selected with the onboard selection tools of FIJI in one image of the individual image stacks, the classifier is subsequently trained and the resulting model is applied to the rest of the image stack gaining probability maps of the different classes. In a final step the probability maps are converted to segmented image stacks by the pixel wise allocation of the class of highest probability.
The high resolution scans are segmented using the standard filter settings and for the coarse scans the most suitable results come from application of the training features Gaussian blur, Hessian, Membrane projections, Sobel Filter, Difference of gaussians and Neighbors. Further missegmentations of the class "background" are suppressed by filling the cylindrical specimen volume with the grey value of class "cell volume" consistently. | |
| Type of data acquisition | Experiment: The image acquisition is performed on two very different X-ray computed tomography systems. Scans of the whole specimens, further on referred to as "coarse" (CS) in the file names, were performed using a v|tome|x L 450 system from GE located at the Institute of Lightweight Engineering and Polymer Technology at TU Dresden with a microfocus tube of 300 kV maximum acceleration voltage and a wolfram target. After scanning, the CS data has been imported to VGStudio, cropped to the specimen volume and exported as image stack of the circular cross section with a plane resolution of 26 µm under preservation of 7 µm resolution in axial direction.
Additional scans of the identical specimens with a high resolution X-ray computed tomography system Zeiss Versa 520 were carried out at the Dresden Center for Nanoanalysis (DCN). The results are further on referred to as "high resolution" scans (HRS) in the file names. Due to technical and financial limitations, selected cylindrical regions of interest (ROI) with a diameter and height of approximately 1 mm of each sample were scanned. | |
| Research objects | Substance: In this work, the morphology of three EPP bead foams with closed cells and different densities is studied using x-ray computed tomography. The specimens of 10 mm diameter and 10 mm length were cut from steam-chest-molded plates of 40 mm thickness. The plates are based on beads of different bulk densities (Neopolen (BASF) P92HD 105, 92HD 130, 92HD 180) and welded together in an Energy Foamer 5.0 of Kurtz Ersa using the crack-fill-method. | |
| Abstract | In this work, the morphology of three EPP bead foams with closed cells and different densities is studied using x-ray computed tomography. The specimens of 10 mm diameter and 10 mm length were cut from steam-chest-molded plates of 40 mm thickness. The plates are based on beads of different bulk densities (Neopolen (BASF) P92HD 105, 92HD 130, 92HD 180) and welded together in an Energy Foamer 5.0 of Kurtz Ersa using the crack-fill-method.
The image acquisition is performed on two very different X-ray computed tomography systems. Scans of the whole specimens, further on referred to as "coarse", were performed using a v|tome|x L 450 system from GE located at the Institute of Lightweight Engineering and Polymer Technology at TU Dresden. Additional scans of certain regions of interest of the identical specimens with a high resolution X-ray computed tomography system Zeiss Versa 520 were carried out at the Dresden Center for Nanoanalysis (DCN). The results are further on referred to as "high resolution" scans. For analysing and statistical describing the micro- and meso-structure the cell walls are further on segmented from the different CT-volumes using WEKA-segmentation provided by the open-source package distribution FIJI.
The provided data set enables the direct comparison of the results of different tomography setups at selected regions of interest and the deeper understanding of the multi-scale morphology of bead foams. | |
| Applied methods and techniques | The main image aquisition parameters for the coarse scans (CS, v|tome|x L 450) are:
Acceleration voltage: 50 kV
Beam current: 100 μA
Focal spot: ≈7μm
Number of projections: 720
Exposure time: 4000 ms
The main image aquisition parameters for the high resolution scans (HRS, Zeiss Versa 520) are:
Acceleration voltage: 30 kV
Beam current: ≈ 67 μA
Number of projections: 3201
Exposure time: ≈ 24.00 s | |
| Additional keywords | particle foam | |
| Additional keywords | x-ray tomography | |
| Additional keywords | Partikelschaum | |
| Additional keywords | structure analysis | |
| Language | eng | |
| Year or period of data production | 2021-2022 | |
| Publication year | 2022 | |
| Publisher | Technische Universität Dresden | |
| Publisher | Institute of Lightweight Engineering and Polymer Technology - ILK | |
| References on related materials | IsPartOf: 123456789/5755 (Handle) | |
| Content of the research data | Image, Dataset: - *.pca- and *.pcr-files (readable by any editor) containing parameters for image aquisition and reconstruction for coarse scans (CS)
- x-ray computed tomography data of the coarse scans (CS) in form of zip-files including *.vol-Files (raw pixels encoded as integers; volume generation using ROI_Size from *.pcr-file)
- x-ray computed tomography data in form of txm-Files by Zeiss (ole-format, also readable by FIJI/ImageJ using XRM-Reader Plugin)
- ROI-definitions (*.zip) from FIJI/ImageJ ROI-Manager plugin defining the reagions of interests used for segmentation and high resolution scans
- volume image data of segmented cell walls in form of image stacks *.tif readable in FIJI/ImageJ | |
| Holder of usage rights | Technische Universität Dresden | |
| Usage rights of the data | CC-BY-NC-SA-4.0 | |
| Software | Resource Processing: FIJI 1.53v | |
| Discipline(s) | Engineering | de |
| Discipline(s) | Materials Science | de |
| Title of the dataset | Computed tomography data of different resolution of expanded polypropylene foams of different density | |