The effect of a shutter on the wind induced loads on a window and wind driven rain intrusion into the building through experiments at the Wall of Wind Experimental Facility
M. Moravej a, B. Hajra b, A. G. Chowdhury a*, I. Zisis a, P. Irwin
bInternational Hurricane Research Center, Florida International University, USA
In the event of a storm that normally includes high wind speeds and rain, damage to a window can occur if the wind loads exceed the design loads of the window. This damage can result in increased internal wind induced pressure that can increase the uplift forces on the building roof, besides wind driven rain (WDR) intrusion into the building causing damage to the building interior. For many years, shutters fixed to a window have been thought to reduce wind induced loads on the window and protect them from wind borne debris. The present paper focuses on an experimental investigation of these issues through large scale testing at the Wall of Wind (WOW) Experimental Facility at Florida International University (FIU), USA. Two types of shutters, namely: shutters with bracket and shutters without bracket were considered for wind directions varying from -90 to 90 degrees. For each of these cases, the mean and peak pressure coefficients were measured on the exterior and interior surfaces of the window and surrounding wall, as well as the volume of water that accumulated inside the building. Results show that the mean and peak pressure coefficients on the window do not change markedly for cases with and without the shutter. However, the shutter greatly helps in preventing WDR intrusion into the building for all the cases tested. Future research in this area devoted to the effect of window size and multiple wall openings is needed.
Keywords: Wind loads; Wind driven rain; Wall of Wind, Shutter; Window; Peak pressure coefficient.
A window is an essential component of a building, since it enhances occupant comfort through proper ventilation, by allowing the entry of air and light into the building. Proper ventilation in a building is necessary to provide a comfortable temperature and maintain indoor air quality for the occupants (Mochida et al., 2005). However in the event of a hurricane involving high wind speeds and rain, a damage to the window can lead to the damage of the building in two ways: increase in internal pressure leading to high uplift forces on the roof, and WDR intrusion into the building interior. Shutters made of plywood or metal are very often used, especially in hurricane prone areas to prevent damage on windows in the event of high wind speeds (Beason et al., 1984; Kordi et al., 2010). Based on a study in Houston, Texas, Beason et al., 1984 found that glass damage to windows was mostly due to windborne debris, and these damages could be avoided by improving the existing building design guidelines. Indeed, the role of the shutters is important since they are known to protect the window from windborne debris, besides preventing WDR intrusion (Minor, 2005).
In fact, many loss modelling techniques assume that a shutter can reduce the wind loads on a
8th International Colloquium on Bluff Body Aerodynamics and Applications Northeastern University, Boston, Massachusetts, USA June 7 - 11, 2016
window in the event of high wind speeds (Fernandez et al., 2010). In the absence of available literature on the effect of shutters on the wind loads on a window and WDR intrusion, it is necessary to investigate these aspects experimentally.
This paper presents results from an experimental study carried out at the Wall of Wind (WOW) experimental facility at Florida International University (FIU), Miami, USA. In this context, wind directions were varied from -90 to 90 degrees for two different cases: shutters without bracket and shutters with bracket. The ‘shutters without bracket’ were directly fixed outside the window. The ‘shutters with bracket’ were installed by sliding the shutters through the bracket fixed on top of the window. Measurements of mean and peak pressure coefficients at various internal and external locations around the window were made, as well as the volume of water that accumulated inside the building. These measurements for both the cases were compared to the case of “window without shutters”. Section 2 describes the WOW experimental set up and various cases. The results and discussion and conclusions form part of sections 3 and 4 respectively.
2. WALL OF WIND EXPERIMENTAL SETUP AND VARIOUS CASES
The Wall of Wind (WOW) experimental facility at FIU consists of 12 fans placed in a two–row by six-column pattern, capable of producing a wind field of 6 m wide and 4.25 m high (Aly et al., 2011). The WOW can generate up to a Category 5 Saffir–Simpson Scale hurricane wind speed that reasonably replicates mean wind speed and partial turbulence characteristics of real hurricane winds (Mooneghi et al., 2014). The velocity and turbulence characteristics corresponding to an open terrain were produced by spires and roughness elements.
Three different cases were considered:
Case 1: Unprotected window (no shutter installed);
Case 2: Window protected by ‘shutter with brackets’ (Figure 4 (b));
Case 3: Window protected by ‘shutter without brackets’ (Figure 4 (c)).
For each test, the external and internal wind induced pressures at various tap locations were measured using a ZOC 33 Scanivalve system (Scanivalve Corporation, 2013) at a sampling rate of 512 Hz for one minute duration. A tubing transfer function was used to correct the distortion effects (Irwin et al., 1979). The tests were found to be repeatable within ± 5%.The peak pressure coefficients were estimated using the partial turbulence simulation (PTS) method.
This paper presents result from an experimental study to determine the effectiveness of shutters in reducing wind loads on the window while preventing WDR intrusion into the building. The study was carried out at the WOW experimental facility at FIU, USA. Based on observations, it was noticed that the shutter (with and without bracket) does efficiently prevent WDR intrusion into the building. However, considering the mean and peak wind loads on the window, experimental results indicated that any wind pressure reduction due to the installation of the shutter, irrespective of the wind direction and the shutter installation method is minimal. Although a similar trend was observed among the net loads on the window, but the net loads across the shutter itself do reduce significantly as a result of pressure equalization. This study shows that the assumption often made in loss modelling techniques of reduced wind loads on the window due to a shutter is questionable. Future research in this area directed at the effect of shutter type, window size and multiple wall openings is needed.