Universität Duisburg-Essen Home

How to find us
News / Conferences
External Links


Studies / Lectures
Final theses /
Project works: Topics

Heinz-Luck-Fire Lab

AUBE '17 / SUPDET 2017
ICOF 2016


Electrical and Electronics

Alumni Engineering Sciences
Engineering Sciences
Universität Duisburg-Essen

Mobile radio:
Smart Antennas
Adaptive MIMO
     communication systems

Filter bank multicarrier
    transmission systems

Adaptive multicarrier

Wideband GHz and THz
     radar sensors

Fire detection:
Automatic Fire Detection
Test devices for
     dust and water fog tests

     Aerosol Charactersiation


Fire Detection

Stand: 08.08.2016
Universität Duisburg-Essen, Campus Duisburg
Polarised Aerosol Characterisation

According to various statistics, over 40 % of the alarms of automatic fire alarm systems installed in buildings are caused by nuisance aerosols, especially due to dust and water droplets like steam or fog. Due to the higher sensitivity of aircraft fire detectors compared to detectors in buildings, aviation is faced with a false alarm rate of even 100:1.

Against this background the Fire Detection Research Group at the Chair of Communication Systems investigates since many years into new fire detection technologies with reduced false alarm ratio. One outstanding research result is the polarimetric smoke detection technology "POLARISE".

POLARISE explores the possibility of distinguishing nuisance aerosols by their particular optical light scattering characteristics, targeting a high specificity without reducing the sensitivity to smoke. Rather than only evaluating the particle size, POLARISE focusses on special morphological and chemical characteristics of the non-fire aerosols.

To investigate into the scattering characteristics of smoke, dust and water droplets like fog, simulation tools based on the Mie-Theory and the Rayleigh-Debye-Ganz approximation were developed and comprehensive simulations were carried out to explore the optimal scattering angles and polarisation directions. With these results different measurement systems have been developed and integrated into extensive measurement campaigns in the fire detection laboratory and in field experiments (e.g. stables).

First results are very promising. The discrimination between smoke and dust is based on the depolarisation effect caused by the compact and irregular structure of dust particles, which cause a partial depolarisation of the scattered light.

This effect can be observed by monitoring the light scattering of different aerosols illuminated by a polarised light source, as depicted in figure 1. In figure 1 also results with smoke from an open wood fire, a smouldering fire and dust is shown. From the left part of the camera pictures it can be seen that dust partially depolarises the incident light.

Figure 1      Depolarisation measuring set-up and first results.

Figure 2 shows the depolarisation ratio sdepol, i.e. the ratio between the depolarised and the polarised scattered light intensity of several different aerosols. TF1 – TF5 are different kinds of smoke arising from standard test-fires according to the EN54. These test-fires include open fires (TF1, TF4 and TF5) as well as smouldering fires (TF2 and TF3). Paraffin denotes the typical paraffin oil aerosol used in the standard smoke detector tests. UltraFine, Medium, Cellulose and Dolomit 10 are test dusts with different particle size distributions and particle shapes. These types of dust are used for the tests of smoke detectors in non-fire scenarios or the test of automobile and vacuum cleaners air filters. Figure 2 shows that the dusts show a distinctly higher depolarisation ratio than all other aerosols. The depolarisation ratio can be therefore implemented as an indicator for dust.

Figure 2      Depolarisation ratio of different fire and non-fire aerosols.

The discrimination of water droplets (fog or steam) and smoke is performed by the analysis of the polarisation ratio of the light scattered in a defined angle, similar to the rainbow effect.

Observations of the natural phenomenon of a rainbow show that they only occur at very pure water-droplets at an angle of approximately 42 degrees from the direction opposite to the light source. Figure 3 shows two images taken of a rainbow looking through a polarising filter. In the left image the transmitting axis of the polarising filter was rotated around 55°, while in the right image this value is about 145° (i.e. 55° + 90°). The comparison of both images show that the rainbow is highly polarised, as it can only be seen in the left image. This effect is particular for pure rain droplets.
A similar effect called "fog bow" can also be observed at the scattering of smaller particles like those in fog and water vapour. Smoke of smouldering fires also consists to a high degree of water, but the mixture with soot changes the chemical composition of the particles, impacting on their optical characteristics, so that no fog bow occurs.

Figure 3      Images of a rainbow through a polarising filter.

Figure 4 shows results of measurements of the polarisation ratio, i.e. the ratio of the scattered light with parallel polarisation to the scattered light with orthogonal polarisation at the angle of a fog bow. As described above, TF1 – TF5 are different kinds of smoke arising from standard test-fires according to the EN54. These test-fires include open fires (TF1, TF4 and TF5) as well as smouldering fires (TF2 and TF3). Paraffin denotes the typical paraffin oil aerosol used in the standard smoke detector tests.

Figure 4 shows that the polarisation ratio of fog and water vapour is much higher than the polarisation ratio of different kinds of smoke. Thus the polarisation ratio can very well be applied for the discrimination between fog, water vapour and smoke which is one of the major innovations in fire detection technologies in the last years.

Figure 4      Polarisation ratio of different fire and non-fire aerosols.

Current works under the project POLARISE focus on the implementation of the investigated methods in different applications and set-ups.