8.0 Abteilungsleitung und andere
Filtern
Dokumenttyp
- Vortrag (20)
- Posterpräsentation (4)
- Beitrag zu einem Tagungsband (1)
Referierte Publikation
- nein (25) (entfernen)
Schlagworte
- Thermography (8)
- Additive Manufacturing (6)
- Additive Fertigung (5)
- Computed tomography (4)
- Laser Powder Bed Fusion (4)
- AGIL (3)
- Additive manufacturing (3)
- Machine Learning (3)
- Neutron diffraction (3)
- Residual stress (3)
Organisationseinheit der BAM
- 8.5 Röntgenbildgebung (25) (entfernen)
Eingeladener Vortrag
- nein (20)
Laser powder bed fusion is one of the most promising additive manufacturing techniques for printing complex-shaped metal components. However, the formation of subsurface porosity poses a significant risk to the service lifetime of the printed parts. In-situ monitoring offers the possibility to detect porosity already during manufacturing. Thereby, process feedback control or a manual process interruption to cut financial losses is enabled.
Short-wave infrared thermography can monitor the thermal history of manufactured parts which is closely connected to the probability of porosity formation. Artificial intelligence methods are increasingly used for porosity prediction from the obtained large amounts of complex monitoring data. In this study, we aim to identify the potential and the challenges of deep-learning-assisted porosity prediction based on thermographic in-situ monitoring.
Therefore, the porosity prediction task is studied in detail using an exemplary dataset from the manufacturing of two Haynes282 cuboid components. Our trained 1D convolutional neural network model shows high performance (R2 score of 0.90) for the prediction of local porosity in discrete sub-volumes with dimensions of (700 x 700 x 40) μm³.
It could be demonstrated that the regressor correctly predicts layer-wise porosity changes but presumably has limited capability to predict differences in local porosity. Furthermore, there is a need to study the significance of the used thermogram feature inputs to streamline the model and to adjust the monitoring hardware. Moreover, we identified multiple sources of data uncertainty resulting from the in-situ monitoring setup, the registration with the ground truth X-ray-computed tomography data and the used pre-processing workflow that might influence the model’s performance detrimentally.
The development of more powerful and more efficient lithium-ion batteries (LIBs) is a key area in battery research, aiming to support the ever-increasing demand for energy storage systems. To better understand the causes and mechanisms of degradation, and thus the diminishing cycling performance and lifetime often observed in LIBs, in operando techniques are essential, because battery chemistry can be monitored non-invasively, in real time. Moreover, there is increasing interest in developing new battery chemistries. Beyond LIBs, sodium ion batteries (NIBs) have gained increasing interest in recent years, as they are a promising candidate to complement LIBs, owing to their improved sustainability and lower cost, while still maintaining high energy density.[1] Initial phases of NIB commercialisation have occurred in the past year. However, for the widespread commercialisation of NIBs, there are still challenges that need to be overcome in developing optimized electrode materials and electrolytes. For the development of such materials and greater understanding of sodium storage mechanisms, solid electrolyte interface (SEI) formation and stability, and degradation processes, in operando methodologies are crucial.
Among the techniques available for in operando analysis, nuclear magnetic resonance spectroscopy (NMR) and imaging (MRI) are becoming increasingly used to characterize the chemical composition of battery materials, study the growth and distribution of dendrites, and investigate battery storage and degradation mechanisms. In situ and in operando 1H, 7Li and 23Na NMR and MRI have recently been used to study LIBs and NIBs, identifying chemical changes in Li and Na species respectively, in metallic, quasimetallic and electrolytic environment as well as directly and indirectly studying dendrite formation in both systems.[2-4] The ability of NMR and MRI to probe battery systems across multiple environments can further be complemented by the enhanced spatial resolution of micro-computed X-ray tomography (μ-CT) which can provide insight into battery material microstructure and defect distribution.
Here, we report in operando 1H and 7Li NMR and MRI experiments that investigate LIB performance, and the identification of changes in the Li signal during charge cycling, as well as the observation of signals in both 1H and 7Li NMR spectra that we attribute to diminishing battery performance, capacity loss and degradation. Additionally, recent operando methodology are adapted and implemented to study Sn based anodes in NIBs. 23Na spectroscopy is performed to monitor the formation and evolution of peaks assigned to stages of Na insertion into Sn, while 1H MRI is used to indirectly visualize the volume expansion of Sn anodes during charge cycling. Battery operation and degradation is further explored in these NIBs, using μ-CT, where the anode is directly visualized to a higher resolution and the loss of electrolyte in the cell, during cycling is observed
The global demand for concrete is growing, and with it, its carbon footprint. Current literature proposes biochar, a product of pyrolysis, as a possible car-bon sink to reduce the carbon footprint of concrete. This work investigates the microstructure of Portland cement pastes with 0%, 5%, and 25% of the cement replaced with wood biochar, since this should influence its macro-scopic mechanical properties. MIP, gas sorption, NMR, and µ-CT were used to analyze the pore space of the three materials. The combination of these methods, each with different resolution, enables a multi-scale investigation of biochar impact on the microstructure of cement pastes. NMR confirmed that biochar can absorb moisture and, thus, reduces the effective water-to-cement ratio. MIP and gas sorption results show 0% and 5% volume re-placement have similar gel pore structure. The results from µ-CT investiga-tions suggest that biochar may reduce the formation of larger pores. The in-clusion of non-reactive porous particles such as biochar increase the porosity of the material and should act as a weakness in terms of mechanical proper-ties. Overall, this study highlights the need to carefully tailor replacement rates to control the impact of biochar on the microstructure concrete mixtures and sees a strong need for further studies on mechanical properties.
In this study, we present an enhanced deep learning framework for the prediction of porosity based on thermographic in-situ monitoring data of laser powder bed fusion processes. The manufacturing of two cuboid specimens from Haynes 282 (Ni-based alloy) powder was monitored by a short-wave infrared camera. We use thermogram feature data and x-ray computed tomography data to train a convolutional neural network classifier. The classifier is used to perform a multi-class prediction of the spatially resolved porosity level in small sub-volumes of the specimen bulk.
The appearance of irregularities such as keyhole porosity is a major challenge for the production of metal parts by laser powder bed fusion (PBF-LB/M). The utilization of thermographic in-situ monitoring is a promising approach to extract the thermal history which is closely related to the formation of irregularities. In this study, we investigate the utilization of convolutional neural networks to predict keyhole porosity based on thermographic features. Here, the porosity information calculated from an x-ray micro computed tomography scan is used as reference. Feature engineering is performed to enable the model to learn the complex physical characteristics of the porosity formation. The model is examined with regard to the choice of hyperparameters, the significance of thermal features and characteristics of the data acquisition. Based on the results, future demands on irregularity prediction in PBF-LB/M are derived.
Avoiding the formation of defects such as keyhole pores is a major challenge for the production of metal parts by Laser Powder Bed Fusion (LPBF). The use of in-situ monitoring by thermographic cameras is a promising approach to detect defects, however the data is hard to analyze by conventional algorithms. Therefore, we investigate the use of Machine Learning (ML) in this study, as it is a suitable tool to model complex processes with many influencing factors. A ML model for defect prediction is created based on features extracted from process thermograms. The porosity information calculated from an x-ray Micro Computed Tomography (µCT) scan is used as reference. Physical characteristics of the keyhole pore formation are incorporated into the model to increase the prediction accuracy. Based on the prediction result, the quality of the input data is inferred and future demands on in-situ monitoring of LPBF processes are derived.
The influence of the inter-layer-time and the scanning velocity on the surface and bulk residual stress in laser powder bed fused 316L specimens was investigated. This study combines X-ray and neutron diffraction results with the thermal history of the specimens acquired through in-situ process monitoring. The process parameter variations were observed to directly influence the thermal history, which gave new insights in the assessment of the residual stress results.
Design of freedom, performance improvement, cost reduction and lead time reduction are key targets when manufacturing parts in a layer-by-layer fashion using the laser powder bed fusion process (LPBF). Many research groups are focussed on improving the LPBF process to achieve the manufacturing of sound parts from a structural integrity perspective. In particular, the formation and distribution of residual stress (RS) remains a critical aspect of LPBF. The determination of the RS in LPBF benefits from the use of neutron diffraction (ND), as it allows the non-destructive mapping of the triaxial RS with a good spatial resolution. Two case studies are presented based on experiments carried out on the angular-dispersive neutron diffractometers Strain Analyser for Large Scale Engineering Applications (SALSA) (Institut Laue Langevin, Grenoble) and STRESS-Spec (FRM II, Garching). The RS in LPBF parts having a rectangular and more complex geometry (lattice structure) is analysed. The former example discusses the mapping of the RS in a rectangular body manufactured from stainless steel 316L. The manufacturing of these parts was monitored using an in-situ thermography set-up to link the RS to the thermal history. The latter discusses the RS in a lattice structure manufactured from the nickel base superalloy IN625. This geometry is challenging to characterise, and the use of a X-ray computed tomography twin is presented as tool to support the alignment of the ND experiment. The results from these case studies show a clear link between the thermal history and the RS magnitudes, as well as giving insights on the RS formation.
An unusual microstructure, inherent residual stresses and void formation are the three key aspects to control when assessing metallic parts made by LPBF. This talk explains an experiment to unravel the interlinked influence of the two mechanisms for the formation of residual stresses in LPBF: the temperature gradient mechanism and constricted solidification shrinkage. The impact of each mechanism on the shape and magnitudes of the residual stress distribution is described. Combined results from neutron diffraction, X-ray diffraction, computed tomography and in-situ thermography are presented.
Also, influence of scan strategies as well as surface roughness of subjacent layers on void formation is shown. Results from computed tomography and in-situ thermography of a specimen dedicated to study the interaction of the melt pool with layers of powder underneath the currently illuminated surface are presented.
Laser powder bed fusion is used to create near net shape metal parts with a high degree of freedom in geometry design. When it comes to the production of safety critical components, a strict quality assurance is mandatory. An alternative to cost-intensive non-destructive testing of the produced parts is the utilization of in-situ process monitoring techniques. The formation of defects is linked to deviations of the local thermal history of the part from standard conditions. Therefore, one of the most promising monitoring techniques in additive manufacturing is thermography. In this study, features extracted from thermographic data are utilized to investigate the thermal history of cylindrical metal parts. The influence of process parameters, part geometry and scan strategy on the local heat distribution and on the resulting part porosity are presented. The suitability of the extracted features for in-situ process monitoring is discussed.