By Dong Han, PhD, Mohammad H. Mehrabi, MS, and Lori Koch, MS, PE
Click here to view the article in the June 2024 digital edition of STRUCTURE magazine.
The past decade has witnessed an increasing adoption of mass timber buildings in North America. These innovative structures primarily feature large, engineered wood products, such as glued-laminated timber (glulam) and cross-laminated timber (CLT). Recognized for their sustainability, these materials align well with the growing emphasis on sustainable building methods. In addition to their environmentally friendly attributes, mass timber structures offer pleasing aesthetics and reduced construction timelines, setting them apart from conventional alternative materials.
Despite their advantages, mass timber buildings, as a new entrant, face design challenges that traditional construction methods have largely overcome during their decades-long evolution. Many of these challenges arise, in part, from a lack of prescriptive connection options, as well as testing and standardization of system solutions, for mass timber. Engineering costs may rise when additional time is necessary to design, specify, and detail custom solutions for every new mass timber project.
The 2024 International Building Code (IBC) references ASCE/SEI 7-22 Minimum Design Loads and Associated Criteria for Buildings and Other Structures for seismic drift requirements, with 2% being a common drift limit for Risk Category I or II structures. The most reliable method for quantifying the in-service drift performance of a given connection requires rigorous testing. Some municipal jurisdictions mandate cyclic deformation tests conducted at a drift of up to 4%, surpassing levels specified in national codes and standards. Significant challenges persist in accurately predicting drift performance for entire mass timber systems.
Common assumptions used in the design process may not always fully reflect real-world conditions, which can influence drift performance. In mass timber construction, most connections are modeled as pinned connections. In reality, these connections exhibit varying degrees of fixity, allowing them to resist rotation. As a result, factors like beam depth, lever arm length, and the gap between primary and secondary members can significantly influence connection performance. For example, the relative connector placement will influence the length of the lever arm created by the eccentricity between the extreme tension or compression face and the center of gravity of the connector. An increase in the length of the lever arm will intensify the prying effect at the connection and could lead to undesired deflections or failure of the connection (Figure 1).
Moreover, seismic drift, in tandem with other considerations such as gravity loading and fire protection, plays a vital role in determining the ideal placement of the connector on the beam section that minimizes fracture development perpendicular to the grain under loading, especially in deep cross sections facing an elevated risk of cracking. Simultaneously addressing multiple concerns—a necessity in many cases—requires reinforcement, often achieved by incorporating reinforcing screws (Figure 2). While effective in many scenarios, this method has limitations.
Custom-Designed or Pre-Engineered Connectors?
In many current mass timber buildings, structural elements are linked through custom-made connectors—connectors designed based on the requirements specified in standards such as the National Design Specification for Wood Construction (NDS). The NDS design approach is generally considered conservative, often resulting in low material utilization efficiency. On the other hand, the present dearth of testing hinders a comprehensive understanding of real-world connector performance in various aspects, including capacity and seismic resilience. An increasingly common alternative involves the use of pre-engineered beam hangers (PEBHs), an off-the-shelf product whose performance has been validated through manufacturer testing. (Not all PEBHs are rated for seismic drift, and they should be carefully selected based on the manufacturer’s specifications.)
Through full-scale monotonic and quasi-static cyclic pushover testing, a study at Oregon State University examined the seismic performance of three beam-to-column connections: one employing a PEBH and two featuring custom solutions, a bearing plate and a notched column (referred to as PEBH, BP, and N connections, respectively) (Figure 3). Of these tested configurations, the PEBH can resist shear and axial forces, whereas the two custom connections only provide bearing support without lateral capacity that is essential for transferring seismic forces. The PEBH consisted primarily of two aluminum connection plates and two clamping jaws connected via threaded rods. In each connection, a glulam beam with a CLT panel on top was connected to a glulam column using the respective connector. The monotonic and cyclic tests involved the use of a hydraulic actuator to apply a maximum lateral displacement of 10-7/8 inches and 5-1/2 inches, corresponding to interstory drifts of 7.2% and 4%, respectively.
The residual drift, or the permanent deformation, of a mass timber structure following a seismic event is a critical seismic performance metric for its connections. Under monotonic loading, the PEBH, BP, and N connections exhibited residual displacements of 5/16 inch, 3/4 inch, and 4-1/16 inches, respectively. A smaller residual drift generally indicates that the connector is more effective in restricting the permanent deformation of the connection, contributing to lower rehabilitation costs and improved occupant safety and comfort.
A load–displacement analysis, as shown in Figure 4, reveals two other crucial performance metrics, capacity and ductility. Of the three tested connections, the PEBH connection exhibited the highest load-carrying capacity, approximately 2.4 and 2 times that of the BP and N connections, respectively. (The significant drop in force at a displacement of approximately 4.4 inches for the PEBH connection resulted from the unloading and reloading of the hydraulic ram—used to simulate a gravity load—to prevent rotation.) All three connections displayed ductile behavior under loading. Compared to the custom connections, the PEBH connection delivered a significantly more pronounced initial stiffness, indicating a reliable and predictable linear and elastic performance prior to yielding, a desirable characteristic for mass timber connections.
Cyclic testing facilitated an understanding of the capacity of the three connections to dissipate energy. A high capacity to dissipate seismic energy is generally preferable, as it contributes to the reduction of vibration energy. In comparison to the two custom connections, the PEBH connection gradually dissipated more energy per cycle as the testing progressed, ultimately dissipating the most energy, as illustrated in Figure 5.
Conclusion
As mass timber buildings continue to gain traction across North America, including its seismically active zones, the limited guidance on the seismic performance of connections designed based on NDS provisions raises challenges in determining if code-specified drift targets are being met. Codified prescriptive options with tested or established performance have not yet been widely developed in the industry, and custom-designing bespoke connections for every new project adds project costs. Mass timber connections can generally be established using either custom-designed connectors or PEBHs. Monotonic and cyclic testing has suggested that PEBHs can deliver seismic performance comparable to, or better than, that of custom solutions, particularly in terms of capacity, ductility, and energy dissipation. PEBHs are an attractive option that can be considered not only for their proven structural performance but also their value-engineering benefits. ■