Case Study | 1.6.2017
Coupled Analysis Method to Evaluate Curtain Wall Performance during Extreme Loading Events

Categories
Protective Design
Computational Modeling
Key Technologies
Probabilistic Glass Fracture Model
User-Defined Material for PVB
Window Hazard Analysis Modeling (WHAM)
Team
Leveraging the modeling and analysis capabilities of the finite element code LS-DYNA, PEC engineers have developed an approach for evaluating curtain wall performance during dynamic events, such as blast loading, hurricane debris impacts, earthquake movement, and wind loading, using the Window (hazard) Analysis Modeling (WhAM) design approach. WhAM is a simplified modeling approach intended to maximize computational efficiency (i.e., minimize run time) with sufficient accuracy to capture key aspects of system behavior.
In each WhAM design model, structural mullions are represented through the use of Timoshenko-type beam elements while glass lites, interlayers, and IGU spacers are modeled with reduced-integration shell elements. Constant stress solid elements are used to represent the IGU air gaps and structural silicone. Using this mixed element approach, a design-level fidelity is achievable, facilitated through the use of user-defined constitutive models (UMATs) for glass, PVB, and silicone, formulated by PEC engineers.
Finite Element Representation of Curtain Wall Specimen
Illustration of Finite Element Details along Vertical Cross-Section through Centerline of Curtain Wall Specimen IGUs
Glass material behavior in WhAM is characterized with an UMAT model with an embedded flaw-based probabilistic failure criterion consistent with ASTM E1300 Standard Practice for Determining Load Resistance of Glass in Buildings. In this approach, maximum principal stresses are used to define glass-break, at which point the failed glass lite is removed from the computational domain. However, the momentum of the shards is maintained by retaining the nodes of the failed elements and partitioning the mass of the failed glass lite to each node. In this way, momentum can be transferred from a failed glass lite to the PVB interlayer through a kinematic-constraint-based tied contact definition.
In addition to glass material behavior, Hyper-elastic UMATs that incorporate dynamic strength increases associated with both strain and strain-rate are also used in WhAM to characterize the material behavior of PVB and silicone. To appropriately represent the interaction between glass shards and the interlayer material, the PVB UMAT also incorporates a strain-based stiffness increase due to the presence of adhered shards.
Importantly, the WhAM modeling approach has been validated for blast loading conditions by comparing simulation results to experimental data obtained from several shock tube tests sponsored by Enclos Research and Development and conducted by ABS Consulting. Traditional curtain wall systems (not systems designed for blast) with monolithic spandrel IGUs and laminated vision IGUs were provided by Enclos for testing. In each of these tests, digital image correlation (DIC) was employed in conjunction with speckle patterns and high-speed cameras to map displacement fields that were used to support validation.
(Shock tube testing sponsored by Enclos Research and Development and conducted by ABS Consulting)