Advancing Timber Truss Design Through Targeted Joint Optimization

A recent study published in the journal Structures examines how specific joint configurations and the thickness-to-diameter ratio influence the performance of steel-plate-bolted glued laminated timber trusses. Researchers Chao-yue Li, Ai-jun Chen, Hang Wang, Guo-jing He, and Hao-lei Wang conducted integrated experimental testing and finite element analysis on Northeast larch glulam specimens to identify configurations that enhance load-bearing capacity and deformation behavior while supporting sustainable construction practices.

The work focuses on four truss configurations fabricated with consistent spans, heights, member sections, steel plate thicknesses, bolt diameters, grades, loading schemes, and support conditions. Variations centered on web-member arrangements and joint connectivity, allowing direct comparison of how these elements affect overall structural response under mid-span concentrated loading.

Background on Glued Laminated Timber and Truss Applications

Glued laminated timber, commonly abbreviated as glulam or GLT, consists of multiple layers of dimension lumber bonded with structural adhesives under controlled pressure. This manufacturing process yields members with high strength-to-weight ratios, dimensional stability, and the ability to span longer distances than solid sawn timber. Glulam finds frequent use in roof systems, bridges, and large public buildings where both aesthetic appeal and structural efficiency matter.

Timber trusses assembled from glulam members offer material efficiency by placing wood primarily in tension and compression zones while minimizing waste. Steel-plate-bolted connections provide a practical means of joining members on site or in prefabrication shops, transferring forces through dowel action and bearing between bolts, plates, and timber.

Global interest in mass timber construction continues to grow because of its lower embodied carbon compared with steel and concrete alternatives. Optimized joint details can further improve resource use by allowing trusses to carry higher loads over longer spans without excessive deformation or premature failure.

Experimental Program and Specimen Details

The study fabricated twelve trusses divided into four groups, with three replicates per group. All specimens used fast-growing domestic Northeast larch (Larix gmelinii) processed into glulam members via finger-jointing for length, followed by planing, adhesive application, and cold pressing. Trapezoidal symmetric diagonal web geometries were employed, reflecting common practice in roof and bridge applications.

Key controlled parameters included span length, truss height, member cross-sections, steel plate thickness, bolt diameter, bolt grade, loading protocol, and support conditions. The primary variables were web-member arrangement (M-type versus P-type geometries) and the associated steel-plate-bolted joint construction at chord and web intersections.

Mid-span concentrated loading tests proceeded until failure, with continuous recording of load, displacement, and strain data. Minimal variation within each group supported statistical reliability of the observed trends.

Comparative Performance of Truss Configurations

Results indicated that M-type trusses outperformed P-type trusses in both ultimate load capacity and deformation performance. Among the configurations tested, the M-D-type truss delivered the most favorable overall mechanical response. The disconnected mid-span chord joint in this design altered the load-transfer path, promoting force redistribution through bolt groups and adjacent web members. This redistribution delayed premature cracking in the upper chord, ultimately increasing both bearing capacity and ductility.

Primary failure modes observed across specimens included timber splitting near upper chord bolt holes and dowel-bearing failure around bolt holes in lower chord and diagonal web members. These localized mechanisms highlight the importance of joint detailing in governing global truss behavior.

Photo by Oksana Demenko on Unsplash

Finite Element Modeling and Validation

Researchers developed finite element models in ABAQUS software to replicate experimental conditions and extend analysis beyond physical testing limits. The models incorporated nonlinear material behavior for timber, steel, and contact interactions at bolted interfaces. Load-vertical displacement curves from the simulations showed strong agreement with experimental measurements, confirming the models' predictive capability for stress distribution and failure progression.

Validated models enabled parametric studies that would be impractical to conduct solely through laboratory testing, particularly for varying geometric ratios across multiple truss topologies.

Parametric Analysis of Thickness-to-Diameter Ratio

A central contribution of the work lies in the systematic examination of the glulam member thickness relative to bolt diameter. As this ratio increased, bolt deformation modes transitioned from rigid behavior through single-hinged yielding to double-hinged yielding. The transition influences how loads distribute between timber and fasteners, affecting both strength and ductility.

The analysis identified an optimal thickness-to-diameter ratio of 6.7. At this value, glulam members and bolts demonstrated improved cooperative performance, balancing embedment strength in the timber with flexural yielding in the bolts. Ratios below or above this optimum shifted failure toward less desirable modes, either excessive timber crushing or excessive bolt rotation without full utilization of timber capacity.

Implications for Design Practice and Prefabrication

Findings underscore that truss configuration and joint connectivity should be considered together rather than in isolation. Designers can achieve measurable gains in capacity and serviceability by adopting M-type geometries with strategic chord joint disconnection, paired with a thickness-to-diameter ratio near 6.7 under similar connection details.

These insights support greater adoption of prefabricated steel-plate-bolted glulam trusses in regions seeking to expand use of locally sourced, fast-growing timber species. The approach aligns with broader goals of reducing construction waste, lowering transportation emissions through lighter assemblies, and accelerating project schedules via modular assembly.

Industry professionals may reference the study when updating connection design guidelines or when evaluating alternative truss topologies for specific span and load requirements. The work also provides benchmark data for calibration of design software used in timber engineering offices.

Broader Context in Sustainable Construction

Timber structures contribute to decarbonization strategies in the built environment by storing carbon and displacing more emissions-intensive materials. Optimized joints that maximize structural efficiency further amplify these benefits by reducing the volume of material required for a given performance level.

Research of this nature bridges laboratory findings with practical engineering needs, offering quantitative guidance that can inform building codes, project specifications, and educational curricula in structural engineering programs worldwide.

Photo by Planet Volumes on Unsplash

Future Research Directions

Additional investigation could explore long-term performance under cyclic loading, moisture cycling, and fire exposure. Scaling studies to larger truss spans or different timber species would broaden applicability. Integration of the validated modeling framework with optimization algorithms may yield further refinements to joint layouts and member proportions.

Collaboration between academic researchers and industry partners can accelerate translation of these results into updated design tools and construction standards.

Accessing the Full Study

The complete research appears in Structures, Volume 90, August 2026, as article 112284. Readers can view the abstract and access options through the publisher at https://www.sciencedirect.com/science/article/abs/pii/S2352012426012336. The authors are Chao-yue Li, Ai-jun Chen, Hang Wang, Guo-jing He, and Hao-lei Wang.