Acoustic emission (AE) testing is used for the continuous evaluation of structural integrity and the monitoring of damage evolution in structural components and materials. During operation, the environmental and loading conditions of metal structures can result in corrosion and surface wear damage. The early detection of surface degradation flaws is crucial, as they can serve as local stress concentration points, leading to crack initiation and failure. In this work, the effectiveness of AE in monitoring corrosion and surface wear flaw formation was experimentally evaluated. AE sensors were installed on steel test plates during the artificial induction of corrosion and surface wear in order to detect and record the generated AE signals. Corrosion-related AE signals typically exhibit low amplitude, count, and energy values. The direct detection of active corrosion can be challenging in noisy environments, but it can be carried out under certain conditions using dedicated AE sensor groups. Surface-wear-related AE signals exhibit high amplitude, energy, and count values, with long duration values that are associated with wear and grinding conditions. It was found that AE sensors can be utilised to detect corrosion and surface degradation events. The effectiveness of the AE method in detecting surface degradation in noisy environments can be improved by implementing a filtering methodology. This will limit the recording of noise-related signals that can mask out actual surface degradation AE events.

Acoustic emission (AE) testing is a widely used technique for the continuous evaluation of damage initiation and propagation in structural components. AE testing can be applied to a wide range of materials, such as metals and fibre-reinforced composites (FRCs). During loading, multiple failure mechanisms can become active, especially in composite materials, resulting in the generation of AE signals with distinctive waveforms and statistical characteristics. By evaluating AE signals based on their characteristics it is possible to group them into clusters and improve the effectiveness of AE in monitoring structural degradation. The clustering process can be effectively carried out using automated clustering algorithms. The applicability of various clustering algorithms for AE data clustering has been considered in several studies. In this paper, the effectiveness of the iterative self-organising data analysis technique (ISODATA) clustering algorithm is evaluated. The AE data considered have been acquired during tensile and flexural tests on glass fibre-reinforced composite samples. The results from automated clustering are compared with manual filtering of the recorded AE signals.

Disclosure: R.D. Paparodis: None. I. Androulakis: None. D.P. Askitis: None. I. Perogamvros: None. N. Angelopoulos: None. A. Rizoulis: None. S. Livadas: None. A. Boniakos: None. Background: The increasing availability of aggressive lipid lowering therapies (LLTs) worldwide is expected to produce a significant decline in cholesterol concentrations worldwide and a more successful attainment of LDL lowering goals (LLGs), especially in those who need them most: patients with established atherosclerotic cardiovascular disease (ASCVD). Aims: To assess the long-term changes in the rate of attainment of LGs according to the European Society of Cardiology Guidelines nationwide, with the use of different LLTs during 12 consecutive years. Methods: We reviewed the charts of all patients initially seen between 2010-2022 who returned for care follow up at our 8 busy Endocrinology clinics, located in 6 cities in Greece, between Sep-1-2022 and Nov-30-2022. We enrolled those with recorded diagnosis of ASCVD [coronary artery disease (CAD), peripheral arterial disease (PAD) or carotid stenosis >50% (CS)] prior to our initial visit. For the present work, we reviewed data on tobacco use, intake of LLTs, and LDL-cholesterol concentrations as recorded during our initial clinic visit. We excluded all patients referred to our clinics for lipids management. Data are presented as means ± standard deviation. Means are compared with Kruskal Wallis test and p<0.05 is deemed significant. Results: We enrolled n=890 consecutive patients, age 63.9 ± 11.4 years, n=339 were women (38.1%), n=529 (59.4%) had diabetes, n=672 (75.5%) had hypertensions and all had ≥1 forms of established ASCVD: n=630 with CAD, n=279 with CS and n=151 with PAD. Out of these, n=144 (16.2%) were on no LLTs, n=334 (37.5%) continued to smoke, n=381 (42.8%) were on guidelines-based appropriate LLT strategy, n=248 (27.9%) reached LLGs and n=318 (35.7%) fulfilled the year-specific guidelines requirements. Of note is, that this rate dropped to 2.4-20.4% in the years 2019-2022. The likelihood of achieving LDL goals was not statistically significantly different among patients with different forms of ASCVD (p=0.42). Overall, the mean LDL was 88.2 ± 32.1mg/dl, and was not different between patients with different forms of ASCVD (p=0.75). Mean LDL concentrations have only started to drop between 2020-2022 as compared to 2019 (p<0.001), but are not different compared to those seen in the years 2010-2018. The mean LDL dropped to 82.3 ± 30.5mg/dl in 2022 as compared to 102.2 ± 38.8mg/dl in 2019. Conclusions: Our study reveals a large care gap in patients with established ASCVD, which did not seem to improve significantly over the past 12 years, despite the excellent availability of aggressive LLT strategies in Greece. Intervention studies are urgently needed to enhance adherence to updated international guidelines and reduce the extremely high ASCVD risk in this fragile population. Presentation: Friday, June 16, 2023

Acoustic Emission (AE) monitoring offers high sensitivity in detecting faults and defects in rotating machinery. Hit-based AE acquisition systems produce only short, intermittent waveform signals rather than continuous recordings. This work addresses the challenge of fault detection from such fragmented sensor data. It is demonstrated that individual AE hits are insufficient for reliable fault detection, as waveforms from different fault conditions overlap substantially in feature spaces. To overcome this, a distributional feature aggregation approach is proposed. AE hit features are extracted from each waveform, grouped into non-overlapping sequential bags, and summarised through order statistics and distribution moments. Four one-class classifiers namely: (i) Isolation Forest, (ii) PCA one class classifier, (iii) K-nearest neighbour one class classifier, and (iv) Local Outlier Factor were evaluated on a Drivetrain Dimulator (DTS) test rig that simulates faulty gear and bearing conditions. AE signals coming from the healthy condition were used to train the classifiers. Signal deviations due to the presence of gear and bearing faults were subsequently identified. Fault detection results show that bag-level distributional features substantially outperform per-hit features: four of five faults achieve 100% single-bag detection under five-fold cross-validation, while per-hit detection rates remain below 50% for four of five faults. The PCA one class classifier achieves 100% detection for all faults individually but is excluded from the final ensemble due to a 99% false alarm rate caused by the limited healthy training sample in a high-dimensional feature space. A CUSUM sequential detector applied to a three-method ensemble (Isolation Forest, KNN, and LOF) is evaluated on a chronological 80/20 split of the healthy data, triggering alarms for all five fault conditions with a false alarm rate of 0% at the recommended setting. Four faults are detected with sustained alarm patterns (54–100% alarm rates), while bearing outer race, the most challenging fault, is detected intermittently (10.5% alarm rate), demonstrating that sequential evidence accumulation can identify faults that are invisible to single-bag thresholding.