Create an Account
Historical published weathered annealed failure data consisting of 13 samples were collected and glass failure prediction model surface flaw parameters were estimated for each sample using a robust statistical method. Historically, surface parameters based on 60-sec load duration were estimated to be consistent with previous design standards, ASCE7. As present-day design standards are based on 3-sec load duration, surface flaw parameters for each sample are estimated based on 3-sec load duration. Load resistances based on the estimated surface flaw parameters for each historical sample are compared to ASTM E1300 using a series of selected geometries including variations of rectangular dimensions, nominal glass thickness, and aspect ratio. Estimated surface flaw parameters do not appear to be completely independent of in-service conditions, aspect ratio, and surface area.
Glass continues to be a popular component in building facades as it provides better transparency and durability compared to other materials. Over the past 30 years, engineers and designers in the United States and elsewhere have used ASTM E1300-16 (ASTM, 2016) to determine the load resistance (LR) of window glass used in buildings. The LR of glass is based on a glass failure prediction model (GFPM), advanced by Beason (Beason, 1980). The GFPM addresses the presence of flaws on the surface of the glass that create stress concentrations, thereby reducing the load resistance (Griffith, 1920). The effect of the surface flaws due to the in-service conditions of the glass is represented by two surface flaw parameters in the GFPM. The surface flaw parameters used in ASTM E1300-16 (ASTM, 2016) were agreed upon by the committee based on available historical experimental data of weathered annealed glass and expert engineering judgment. However, the standard does not specifically reference which historical experimental data was used, resulting in a lack of transparency on how the surface flaw parameters were chosen.
The authors collected the available historical data of weathered annealed (AN) glass samples tested to failure and the associated published surface flaw parameters. The surface flaw parameters were historically chosen based on summary statistics such as mean, standard deviation (std. dev.), and coefficient of variation (cov). Consequently, attempts to recreate the published surface flaw parameters result in variations as the selection methods used were not documented. The maximum likelihood estimator, (MLE) method, tailored to the GFPM, provides a repeatable method to estimate the surface flaw parameters allowing for consistent comparisons. Prior to the 1995 edition of ASCE7 (ASCE 7-95, 1996), the wind load provisions were based on a fastest-mile wind speed, which corresponds to a 60-sec load duration. Because the majority of the historical test data preceded the change from fastest-mile to 3-sec wind speeds, the published surface flaw parameters were based on a 60-sec duration. As such, the surface flaw parameters estimated using the MLE method for the GFPM will be based on a 3-sec load duration. Since the geometry (rectangular dimensions, thickness) vary for each historical record set, several intermediate geometries will be used to compare the LRs associated with the new surface flaw parameter fits and those currently defined in ASTM E1300-16 (ASTM, 2016).
Glass, a brittle material, does not have a ultimate failure stress. Therefore, the use of an allowable stress for design would require the selection of a relatively low value of
Thirteen sample sets with varying nominal dimensions and loading rate of rectangular weathered annealed (AN) monolithic glass were identified in the published literature. The sample sets consist of nine sets
The EFL3’s for each sample set were calculated using a FEM developed by Yew (Yew et al., 2019) thereby removing any variations that may arise from using different mesh sizes
In each comparison (Figures 1, 2, 3), the ASTM E1300 ( ASTM, 2016) LR appears to overestimate the LR of weathered glass sample sets at a probability of 0.008. Only
Abiassi, J. J. (1981). The strength of weathered window glass using surface characteristics (Doctoral dissertation).
ASCE 7-95. (1996). Minimum design loads for buildings and other structures (ANSI/ASCE 7-95). American Society of Civil Engineers. https://doi.org/10.1061/9780784400920
ASTM. (1989). ASTM E1300-89: Standard practice for determining the minimum thickness of annealed glass required to resist a specific load
ASTM. (2016). ASTM E1300-16: Standard Practice for Determining Load Resistance of Glass in Buildings. https://doi.org/10.1520/E1300-16
ASTM. (2018). ASTM C1239-13. (2018). Standard practice for reporting uniaxial strength data and estimating Weibull distribution parameters for advanced ceramics. https://doi.org/10.1520/C1239-13R18
Beason, W. L. (1980). A failure prediction model for window glass (Doctoral dissertation). Texas Tech University.
Brown, W. G. (1974). A practicable formulation for the strength of glass and its special application to large plates. National Research Council Canada.
Griffith, A. (1920). The phenomena of flow and rupture in solids: Phil. Trans. Roy. Soc. Lond. Ser. A, 221, 163-98.
Hsu, S. Y. (1987). Comparison of window glass sample strength estimates: Full scale versus concentric rings (Doctoral dissertation).
Natividad, K., Morse, S. M., & Norville, H. S. (2016). Fracture origins and maximum principal stresses in rectangular glass lites. Journal of Architectural Engineering, 22(2), 04015014.
Norville, H. S., & Minor, J. E. (1985). Strength of weathered window glass. American Ceramic Society Bulletin, 64 (11), 1467–1470
Sligar, J. W. (1989). Evaluation of weathered window glass strength (Doctoral dissertation).
Wang, B. Y.-t. (1981). Nonlinear analysis of rectangular glass plates by finite difference method (Doctoral dissertation). Texas Tech University.
Yew, G. Z., Norville, H. S., & Morse, S. M. (2019). Nonlinear finite-element analysis of laminated glass using the four-node reissner-mindlin formulation and assumed transverse shear strain fields. Journal of Engineering Mechanics, 145 (7), 04019042.