Mark H. Hammond

Early warning fire detection system using a Probabilistic Neural Network

The Navy program, Damage Control-Automation for Reduced Manning is focused on enhancing automation of ship functions and damage control systems. A key element to this objective is the improvement of current fire detection systems. As in many applications, it is desired to increase detection sensitivity and,more importantly increase the reliability of the detection system through improved nuisance alarm immunity. Improved reliability is needed, such that fire detection systems can automatically control fire suppression systems. The use of multi-criteria based detection technology continues to offer the most promising means to achieve both improved sensitivity to real fires,and reduced susceptibility to nuisance alarm sources. A multi-criteria early warning fire detection system, has been developed to provide reliable warning of actual fire conditions, in less time, with fewer nuisance alarms,than can be achieved with commercially available smoke detection systems. In this study a four-sensor array and a Probabilistic Neural Network have been used to produce an early warning fire detection system. A prototype early warning fire detector was built and tested in a shipboard environment. The current alarm algorithm resulted in better overall performance than the commercial smoke detectors, by providing both improved nuisance source immunity with generally equivalent or faster response times.

Multivariate statistical process control for continuous monitoring of networked early warning fire detection (EWFD) systems

Abstract Sensor array networks provide much information concerning an environment if the data is adequately used. However, monitoring large networks of sensor arrays can be data intensive. Multivariate statistical process control (MSPC) methods allow monitoring of an entire system at a supervisory level. These methods are demonstrated for fire detection using the early warning fire detection (EWFD) system. The EWFD system comprised of four sensors that included photoelectric and ionization smoke detectors, as well as carbon monoxide and carbon dioxide sensors. Fourteen sensor arrays were distributed in 10 compartments and passageways throughout the ex-USS SHADWELL, the advanced damage control fire research platform of the Naval Research Laboratory. Data was collected for a series of smoldering and flaming fire sources as well as nuisance sources in different compartments over two decks. The network of sensor responses, location and temporal data are used to identify events, determine source location and monitor fire rate of growth. Hotelling’s statistic and the Q-statistic are employed initially for event detection. Subsequently, contribution plots are used to determine source location, rate of growth and to discriminate between actual fires and their byproducts in adjacent compartments. The response times of the MSPC method are compared to the commercial smoke detectors co-located with the EWFD systems. MSPC is shown to be an efficient method for continuous monitoring of EWFD systems.

Numerical simulation of an optical chromatographic separator.

Optical chromatography achieves microscale optical manipulation through the balance of optical and hydrodynamic forces on micron sized particles entrained in microfluidic flow traveling counter to the propagation of a mildly focused laser beam. The optical pressure force on a particle is specific to each particles size, shape and refractive index. So far, these properties have been exploited in our lab to concentrate, purify and separate injected samples. But as this method advances into more complex optofluidic systems, a need to better predict behavior is necessary. Here, we present the development and experimental verification of a robust technique to simulate particle trajectories in our optical chromatographic device. We also show how this new tool can be used to gather better qualitative and quantitative understanding in a two component particle separation.

Continuous flow, explosives vapor generator and sensor chamber.

A novel liquid injection vapor generator (LIVG) is demonstrated that is amenable to low vapor pressure explosives, 2,4,6-trinitrotoluene and hexahydro-1,3,5-trinitro-1,3,5-triazine. The LIVG operates in a continuous manner, providing a constant and stable vapor output over a period of days and whose concentration can be extended over as much as three orders of magnitude. In addition, a large test atmosphere chamber attached to the LIVG is described, which enables the generation of a stable test atmosphere with controllable humidity and temperature. The size of the chamber allows for the complete insertion of testing instruments or arrays of materials into a uniform test atmosphere, and various electrical feedthroughs, insertion ports, and sealed doors permit simple and effective access to the sample chamber and its vapor.

Trace explosives sensor testbed (TESTbed)

A novel vapor delivery testbed, referred to as the Trace Explosives Sensor Testbed, or TESTbed, is demonstrated that is amenable to both high- and low-volatility explosives vapors including nitromethane, nitroglycerine, ethylene glycol dinitrate, triacetone triperoxide, 2,4,6-trinitrotoluene, pentaerythritol tetranitrate, and hexahydro-1,3,5-trinitro-1,3,5-triazine. The TESTbed incorporates a six-port dual-line manifold system allowing for rapid actuation between a dedicated clean air source and a trace explosives vapor source. Explosives and explosives-related vapors can be sourced through a number of means including gas cylinders, permeation tube ovens, dynamic headspace chambers, and a Pneumatically Modulated Liquid Delivery System coupled to a perfluoroalkoxy total-consumption microflow nebulizer. Key features of the TESTbed include continuous and pulseless control of trace vapor concentrations with wide dynamic range of concentration generation, six sampling ports with reproducible vapor profile outputs, limited low-volatility explosives adsorption to the manifold surface, temperature and humidity control of the vapor stream, and a graphical user interface for system operation and testing protocol implementation.

Analysis of ammonium nitrate headspace by on-fiber solid phase microextraction derivatization with gas chromatography mass spectrometry

A novel analytical method has been developed for the quantitation of trace levels of ammonia in the headspace of ammonium nitrate (AN) using derivatized solid phase microextraction (SPME) fibers with gas chromatography mass spectrometry (GC-MS). Ammonia is difficult to detect via direct injection into a GC-MS because of its low molecular weight and extreme polarity. To circumvent this issue, ammonia was derivatized directly onto a SPME fiber by the reaction of butyl chloroformate coated fibers with the ammonia to form butyl carbamate. A derivatized externally sampled internal standard (dESIS) method based upon the reactivity of diethylamine with unreacted butyl chloroformate on the SPME fiber to form butyl diethylcarbamate was established for the reproducible quantification of ammonia concentration. Both of these compounds are easily detectable and separable via GC-MS. The optimized method was then used to quantitate the vapor concentration of ammonia in the headspace of two commonly used improvised explosive device (IED) materials, ammonium nitrate fuel oil (ANFO) and ammonium nitrate aluminum powder (Ammonal), as well as identify the presence of additional fuel components within the headspace.

Multisensor data fusion with disparate data sources

In many instances, sensing tasks are best addressed with multiple sensing modalities. However, fusion of the outputs of disparate sensor systems presents a significant challenge to forming a cohesive sensing system. A discussion of strategies for fusion of disparate sensor data is presented and illustrated with examples of real time and retrospective data fusion for multisensor systems. The first example discussed is a real-time system for situational awareness and the detection of damage control events in ship compartments. The second example is a retrospective data fusion framework for a multisensor system for the detection of buried unexploded ordnance at former bomb and target ranges.