Publications by authors named "Edvard Govekar"
Article Synopsis
- This study investigates how acoustic emission measurements can help characterize biocomposites in aviation, using advanced fiber optic sensors for more accurate data collection.
- Researchers applied a convolutional autoencoder to extract both classic and deep features from AE signals, utilizing various machine learning methods for classification.
- Findings reveal that combining classic and deep features significantly boosts classification accuracy, with neural networks achieving the highest accuracy at 80.9%, demonstrating the effectiveness of complex models over simpler ones.
In the paper the results of the acoustic emission (AE) based detection and characterization of stress-corrosion cracking (SCC) in stainless steel are presented. As supportive methods for AE interpretation, electrochemical noise, specimen elongation measurements, and digital imaging of the specimen surface were used. Based on the defined qualitative and quantitative time and power spectra characteristics of the AE bursts, a manual and an automatic procedure for the detection of crack related AE bursts were introduced.
View Article and Find Full Text PDFWe propose an experimental setup allowing for the characterization of laser droplet generation in terms of the underlying dynamics, primarily showing that the latter is deterministically chaotic by means of nonlinear time series analysis methods. In particular, we use a laser pulse to melt the end of a properly fed vertically placed metal wire. Due to the interplay of surface tension, gravity force, and light-metal interaction, undulating pendant droplets are formed at the molten end, which eventually completely detach from the wire as a consequence of their increasing mass.
View Article and Find Full Text PDF