of 18
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Experimental Validation of Smoke DetectorSpacing Requirements and the Impactof these Requirements on DetectorPerformance Christopher L. Mealy* and Daniel T. Gottuk, Hughes Associates, Inc., 3610Commerce Drive, Suite 817, Baltimore, MD 21227, USA Received:  9 September 2008/ Accepted:  12 September 2009 Abstract.  Changes were made to the smoke detector siting requirements for beamedceilings in the 2007 edition of NFPA 72, the National Fire Alarm Code. These chan-ges were based on a modeling studying. A series of full-scale fire tests were conductedto validate the findings of the modeling study for level corridor ceilings with andwithout beams. This paper provides an overview of the experimental testing con-ducted, presents a brief summary of the findings of the experimental validation, andaddresses the impact of deep-beam profiles on level ceilings with respect to smokedetector response. Furthermore, the impact of deep-beamed ceilings on detectorresponse is evaluated with respect to the citing requirements of the 2002 and 2007editions of NFPA 72. Keywords:  Smoke detection, Spacing requirements, Model validation, Smoke detector performance 1. Introduction A modeling study [1] performed in 2006 under a research grant from the Fire Pro-tection Research Foundation used the computational fluid dynamics (CFD) modelFire Dynamics Simulator (FDS) [2, 3] to evaluate smoke detector spacing require- ments with respect to level ceilings with deep-beam profiles. This study was under-taken to evaluate the appropriateness of the NFPA 72 (2002 edition of theNational Fire Alarm Code) prescriptive provisions for non-uniform (i.e., beamed,waffled) ceilings. A similar research effort had been undertaken over a decadebefore by the National Institute of Standards and Technology (NIST) [4] in whichCFD modeling was used to understand the impact that sloped, beamed ceilingshave on detector and sprinkler response. Based upon the findings of the2006 modeling study, several changes to the code language were made in the 2007edition of NFPA 72. The most significant code change based upon the findings of the modeling study was the adoption of smooth ceiling spacing for beamed ceilingconfigurations that previously required detectors in every beam pocket. * Correspondence should be addressed to: Christopher L. Mealy, E-mail: cmealy@haifire.comFire Technology, 46, 679–696, 2010   2009 Springer Science+Business Media, LLC. Manufactured in The United StatesDOI: 10.1007/s10694-009-0108-6  1  Neither of these previous research efforts included full-scale testing to validatethe results of the modeling simulations. Consequently, a subset of the model simu-lations developed in the 2006 research was reproduced via full-scale experimentaltesting to validate the findings of the modeling study. The primary focus of theexperimental test program was to document the response times of smoke detectorsinstalled in various beamed corridor configurations relative to the response of detectors installed on an open, smooth ceiling. The experimental response datawas then compared to the model predictions and the NFPA code changes devel-oped from the relative responses in the 2006 modeling study [1]. The experimentalresults were also used to provide a direct comparison of fire conditions (tempera-ture, smoke, and velocity) with the output of the modeling study.The previous modeling study [1] evaluated multiple corridor configurations con-sisting of corridor widths of 1.5 m (5 ft) and 3.7 m (12 ft); ceiling heights of 2.7 m(9 ft), 3.7 m (12 ft), and 5.5 m (18 ft); beam spacing of 0.9 m (3 ft) and 2.7 m(9 ft); and beam depths of 0.3 m (1 ft), 0.6 m (2 ft), and flat smooth ceilings. Thebeams were modeled as 0.15 m (6 in.) thick, solid rectangular obstructions. Amulti-block grid was utilized in which 38 mm (1.5 in.) cells were used around thefire plume and along the ceiling to capture the flow between and below beams. Acourser grid of 152 mm (6 in.) cells was used for the rest of the domain. FireDynamics Simulator, version 4, was used [2, 3] for all simulations. The modeling simulations used an instantaneous 100 kW fire centered in a beam bay (i.e.,between two beams) at one end of the corridor. The base of the fire was0.31  9  0.31 m (1  9  1 ft) and was located on the floor. The exposure fire was pre-scribed with a soot yield of 2.2%. Instrumentation clusters were specified at vari-ous locations down the centerline of the corridor at the bottom of beams andcentered in the beam bays on the ceiling. A complete and detailed description of model inputs and analysis including the modeling domain, prescription of theexposure fire, model solution procedure, mesh analysis, and boundary conditionassumptions is provided in the modeling study report [1].This paper provides an overview of the experimental testing conducted, presentsthe validation results of the 2006 modeling study, and evaluates the impact of the2006 modeling-based code changes (i.e., between the 2002 and 2007 editions of NFPA 72 [5, 6]) on the response of smoke detectors on beamed ceilings. 2. Experimental Approach The corridor apparatus used in this work was designed such that a large subset of the configurations examined in the modeling study could be evaluated experimen-tally. Such configurations included corridor widths of 1.5 m (5 ft) and 3.7 m(12 ft); ceiling heights of 2.7 m (9 ft), 3.7 m (12 ft), and 5.5 m (18 ft); beam spac-ing of 0.9 m (3 ft) and 2.7 m (9 ft); and beam depths of 0.3 m (1 ft), 0.6 m (2 ft),and flat, smooth ceilings. Ceiling heights greater than 5.5 m (18 ft) were not inves-tigated due to the constraints of the laboratory in which the corridor apparatuswas constructed. The corridor apparatus was 14.6 m (48 ft) long, 3.7 m (12 ft)wide and consisted of a steel support structure and 6.35 mm (0.25 in.) gypsum 680 Fire Technology 2010  wall board (GWB). The corridor had 6.35 mm (0.25 in.) GWB walls extending1.2 m (4 ft) below the ceiling. The remainder of each wall extending to the floorwas constructed of 0.1 mm (4 mil) polyethylene plastic sheeting. The corridorbox-beams were constructed from steel stud framing and 6.35 mm (0.25 in.)GWB. The beams were 0.15 m (6 in.) thick and either 0.3 m (1 ft) or 0.6 m (2 ft)deep. The corridor ceiling was leveled to within  ± 5.1 cm (2 in.) both laterally andlongitudinally.Figure 1 shows a diagram of the corridor apparatus with beams in place, andFigure 2 shows a photograph of the setup. Figure 1 presents the locations of spot smoke detectors, optical density meters and velocity probes. Twenty-four Ga,Type K, bare-bead thermocouples were installed at every detector location andwere positioned 19 mm (0.75 in.) from the ceiling, beam, or wall surface to whichthe detector was mounted. Most devices were centered down the corridor on theceiling or the bottom of a beam, except for a number of devices mounted to thewall, as shown by the symbols near the walls in Figure 1.Optical density meters were constructed in general accordance with the specifi-cations of UL 268. [7] The optical density meters used were comprised of a 6VGeneral Electric sealed beam light source, and a Huygen Model 856 RRV, photo-voltaic cell. The path length for all ceiling mounted ODM’s was 1.52 m (5 ft). Inorder to accommodate the installation of beams within the corridor, it was neces-sary to decrease the path length of the wall-mounted ODM. This ODM had apath length of 0.6 m (2 ft). Figure 1. Diagram of corridor test apparatus with instrumentation. Experimental Validation of Smoke Detector Spacing Requirements 681  Velocity measurements were collected at two locations using Applied Technolo-gies Sonic Anemometer/Thermometer  Model SPA5/2Y   (courtesy of NIST BFRL).The velocity probes were located in the 4.6 m (15 ft) and 6.4 m (21 ft) beam bayson the right side of the corridor. The probes were able to measure velocities inboth the longitudinal and lateral directions. They were installed such that veloci-ties were measured approximately 19 mm (0.75 in.) from the ceiling of the corri-dor. The velocity probes were capable of measuring velocities ranging from 0 to10 m/s with a resolution of 0.01 m/s.Two different spot detection technologies (i.e., ionization and photoelectric)from two different manufacturers were installed within the corridor. Smoke detec-tor clusters were installed on the ceiling of the corridor, on beam bottoms, and atvarious elevations along the walls of the corridor. Detector clusters were installedat the locations illustrated in Figure 1. Wall-mounted detectors were installed atlocations of 0.07 m (3 in.) and 0.3 m (1 ft) to center below the corridor ceiling ateach of the locations presented in Figure 1.Photoelectric detectors from the same manufacturer (Mfg A) were installed atall locations along the length of the corridor as well as on the walls of the corri-dor. The installation of the same technology from the same manufacturer on bothsides of the fire source provided a means of verifying corridor symmetry. Photo-electric detectors from Manufacturer B were installed on the right hand side(RHS) of the corridor and ionization detectors from Manufacturer B were instal-led on the left hand side (LHS). The installation of the same detector technology(i.e., photoelectric) from two different manufacturers on the same side of the cor-ridor (i.e., RHS) provided a means of assessing whether the results of the studywere independent of manufacturer. Furthermore, the installation of differentdetector technologies (ion and photo) on the same side of the corridor (i.e., LHS)provided a means of evaluating the different detector technologies when subjectedto comparable exposures. Figure 2. Photograph of corridor apparatus. 682 Fire Technology 2010

Smoke Detectors

Jul 23, 2017


Jul 23, 2017
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks