Optimization of Process Parameters for the Laser Polishing of Hardened Tool Steel.

In mold making, the mold surface roughness directly affects the surface roughness of the produced part. To achieve surface roughness below 0.8 µm, the cost of surface finish is high and time-consuming. One alternative to the different grinding and polishing steps is laser polishing (LP). This study investigates and models the LP of tool steel (X38CrMoV5-1-DIN 1.2343), typical for the mold industry, having an initial rough surface obtained by electrical discharge machining. The microstructures of the re-melted layer and heat-affected zone due to the LP process were also studied. Four parameters: the laser spot size, velocity, maximum melt pool temperature and overlapping were investigated via a design of experiments (DoE) approach, specifically a factorial design. The responses were line roughness (Ra), surface roughness (Sa), and waviness (Wa). The surface topography was measured before and after the LP process by white light profilometer or confocal microscopy. DoE results showed that the selected factors interact in a complex manner, including the interactions, and depend on the responses. The DoE analysis of the results revealed that the roughness is mainly affected by the velocity, temperature and overlap. Based on a first DoE model, an optimization of the parameters was performed and allowed to find optimum parameters for the LP of the rough samples. The optimum conditions to minimize the roughness are a spot size of 0.9 mm, a velocity of 50 mm/s, a temperature of 2080 °C and an overlap of 90%. By using these parameters, the roughness could be reduced by a factor of almost 8 from 3.8 µm to approximately 0.5 µm. Observations of the microstructure reveal that the re-melted layer consists of columnar grains of residual austenite. This can be explained by the carbon intake of the electro-machined surface that helps stabilize the austenitic phase.

Machine learning monitoring for laser osteotomy.

This work proposes a new online monitoring method for an assistance during laser osteotomy. The method allows differentiating the type of ablated tissue and the applied dose of laser energy. The setup analyzes the laser-induced acoustic emission, detected by an airborne microphone sensor. The analysis of the acoustic signals is carried out using a machine learning algorithm that is pre-trained in a supervised manner. The efficiency of the method is experimentally evaluated with several types of tissues, which are: skin, fat, muscle, and bone. Several cutting-edge machine learning frameworks are tested for the comparison with the resulting classification accuracy in the range of 84-99%. It is shown that the datasets for the training of the machine learning algorithms are easy to collect in real-life conditions. In the future, this method could assist the doctors during laser osteotomy, minimizing the damage of the nearby healthy tissues and provide cleaner pathologic tissue removal.

Author Correction: Supervised deep learning for real-time quality monitoring of laser welding with X-ray radiographic guidance.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Supervised deep learning for real-time quality monitoring of laser welding with X-ray radiographic guidance.

Laser welding is a key technology for many industrial applications. However, its online quality monitoring is an open issue due to the highly complex nature of the process. This work aims at enriching existing approaches in this field. We propose a method for real-time detection of process instabilities that can lead to defects. Hard X-ray radiography is used for the ground truth observations of the sub-surface events that are critical for the quality. A deep artificial neural network is applied to reveal the unique signatures of those events in wavelet spectrograms from the laser back-reflection and acoustic emission signals. The autonomous classification of the revealed signatures is tested on real-life data, while the real-time performance is reached by means of parallel computing. The confidence of the quality classification ranges between 71% and 99%, with a temporal resolution down to 2 ms and a computation time per classification task as low as 2 ms. This approach is a new paradigm in the digitization of industrial processes and can be exploited to provide feedbacks in a closed-loop quality control system.

Characterization of Ablated Bone and Muscle for Long-Pulsed Laser Ablation in Dry and Wet Conditions.

Smart laser technologies are desired that can accurately cut and characterize tissues, such as bone and muscle, with minimal thermal damage and fast healing. Using a long-pulsed laser with a 0.5-10  ms pulse width at a wavelength of 1.07  µm, we investigated the optimum laser parameters for producing craters with minimal thermal damage under both wet and dry conditions. In different tissues (bone and muscle), we analyzed craters of various morphologies, depths, and volumes. We used a two-way Analysis of Variance (ANOVA) test to investigate whether there are significant differences in the ablation efficiency in wet versus dry conditions at each level of the pulse energy. We found that bone and muscle tissue ablated under wet conditions produced fewer cracks and less thermal damage around the craters than under dry conditions. In contrast to muscle, the ablation efficiency of bone under wet conditions was not higher than under dry conditions. Tissue differentiation was carried out based on measured acoustic waves. A Principal Component Analysis of the measured acoustic waves and Mahalanobis distances were used to differentiate bone and muscle under wet conditions. Bone and muscle ablated in wet conditions demonstrated a classification error of less than 6.66 % and 3.33 %, when measured by a microphone and a fiber Bragg grating, respectively.