National Technical University of Athens (NTUA)
Ολοκληρωμένος χημικός, οπτικός και ακουστικός έλεγχος εσωτερικών χώρων για πρόληψη τεχνολογικού ατυχήματος
Abstract
dc:descriptionIn the present doctoral thesis an integrated method proposed, where, PSS-MS, acoustic, and video sensing technologies were used for monitoring lab scale indoor fires under controlled field simulated conditions. In this work, mass spectrometry was combined with optical and acoustical analysis. In particular, a commercial MS (Mass Spectrometer) combined with the in-house developed PSS (Pulsed Sampling System) were used with an optical and an acoustic sensor in order to monitor lab scale burning. The optical sensor was a CCD (Charged Coupled Device) camera operating in visual spectrum. The acoustic sensor was a shotgun condenser microphone with frequency range 20 – 20000 Hz. In addition, the proposed method includes all the techniques used for efficient data acquisition, which collected from sensory equipment. The developed, for this work, signal processing had the aim to reach to conclusions regarding the burning evolution process. Besides the processing of raw data, a quite effective approach proceeds with the modelling of the signal waveform through curve fitting and the utilization of model parameters as characteristic features. In the modelling process, the effort was focused to explore model parameters for the various (chemical or physical) signals in order to select unique characteristics during the burning of each different material. For implementing the experimental section of this thesis, the appropriate experimental set-up and protocol were developed. The final fire monitoring experiments were performed in two lab series and a scale-up experiment in open field. In the first series, two types of fire experiments were carried out; direct and indirect fire monitoring experiments. The main objective of indirect fire monitoring experiments was to simulate the case where physical obstacles in the field prevent the microphone and the camera from direct fire monitoring. In these experiments, audio and video signals either reaffirmed the chemical detection inferences or they provided complementary information when limitations in chemical detection occurred. Similarly, chemical detection in molecular base but strongly related to macroscopic phenomena, supplemented audio and video monitoring when macroscopic scale observations were absent. Audio and video signals proved to provide useful information about the fire event even when physical obstacles prevented the direct audio and video monitoring of the fire. Reflected audio and video signals on metallic surfaces or audio signals from the interaction of the fire with the surrounding materials were recorded and combined with PSS-MS measurements in those cases. In the next series of fire monitoring experiments the burning of different common materials such as cotton textile, inkjet white paper, and oak wood was investigated in lab environment under controlled field-simulated conditions. The burning of combustible materials took place on an inclined metallic plate, with angular orientation 30o with the horizontal level. Integration of mass spectrometry with audio and video signals enabled combined monitoring of the event in both molecular and macroscopic scale. The synergistic effects of the three sensing technologies supplemented by signal processing and modelling appeared to be significant for studying the fire patterns of different materials. Chemical, video and acoustic signal processing took place as an exploratory work in order to identify modalities during burning process of different materials, through the modelling and curve fitting techniques. For the chemical and optical responses the temporal models with curve fitting were used, whereas for the acoustic signal the spectral analysis was used for quantifying harmonics. The extrapolations of this work were ensued with initial models, characteristic for each material studied. xviii In the scale-up experiment the proposed method, developed for monitoring lab scale indoor fire, was applied to monitor a large scale fire of dried forest material in the open field. More specifically a large fire was on-site monitored by chemical, optical and acoustical sensors. The chemical analysis was performed by gas exhaust analyser MRU 1600V Delta, as well as video and acoustical signal were provided a thermographic camera and a shotgun condenser microphone, respectively. Despite difficulties, originated by field conditions, the proposed integrated method was successfully employed. In conclusion, the proposed method presented the capability to effectively monitor the burning process, under lab and field conditions. The analytical capability of the combined action of the three sensing technologies has effected by signal processing and modeling to identify unique characteristics of each material’s combustion. The promising potentials of the present work, with further elaboration, might find field monitoring applications regarding the early warning and identification of hazardous and unsafe conditions.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2012
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Stavrakakis, Panagiotis
- Σταυρακάκης, Παναγιώτης
Subjects
dc:subject × 20- Χημική ανάλυση πεδίου
- Οπτική ανάλυση
- Ακουστική ανάλυση
- Παλμικό δειγματοπηπτικό σύστημα - Φασματόμετρο μάζας ( PSS - MS)
- Ολοκλήρωση αισθητήρων
- Μοτίβο καύσης
- Πτητικές οργανικές ενώσεις
- Μοντελοποίηση κυματομορφής σήματος
- Field chemical analysis
- Optical analysis
- Acoustical analysis
- Pulsed sampling system - Mass spectrometer ( PSS - MS)
- Sensors integration
- Burning pattern
- Volatile organic compounds
- Signal waveform modelling
- Φυσικές Επιστήμες
- Χημεία
- Natural Sciences
- Chemical Sciences
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/33318