Determination of optimal wood-dowel welding parameters for two North American hardwood species

Abstract

Rotational wood-dowel welding has been shown to rapidly produce wood joints of considerable strength without any adhesive. The technique offers an opportunity to increase productivity and reduce costs in the furniture industry. The objective of the study was to define optimal wood-dowel welding parameters for two North American hardwood species frequently used for indoor appearance products: sugar maple (Acer saccharum) and yellow birch (Betula alleghaniensis). Optimized parameters for individually studied species were determined using a rotational wood-dowel welding machine designed for the technique. A comparative analysis of wood-dowel welding parameters was performed. The investigated parameters for both species were grain orientation, rotational speed, and insertion speed. Temperature profile measurements at the interface during rotational wood-dowel welding were also carried out. Optimal welding mechanical properties were determined from the dowel withdrawal strength using a standard tensile strength test. Results revealed a significant interaction between species, rotational speed, and insertion speed. Sugar maple produced wood joints with higher withdrawal strength than yellow birch. The best results for sugar maple and yellow birch were obtained with a rotational speed of 1000 rpm. A 25 mm s⁻¹ insertion speed produced significantly stronger welded joints in sugar maple than at 12.5 mm s⁻¹. For yellow birch, a 16.7 mm s⁻¹ insertion speed provided the best results. Both species and rotational speed had a significant effect on peak temperature at the interface during welding. Peak welding temperatures with optimal parameters were 244 and 282 °C for sugar maple and yellow birch, respectively.

Keywords:

• sugar maple
• yellow birch
• rotational wood welding
• dowel
• wood joints
• welding parameters
• temperature measurement
• furniture application

Acknowledgments

The authors acknowledge the financial support of Le Fonds Québécois de la Recherche sur la Nature et les Technologies (FQRNT), the Natural Sciences and Engineering Research Council of Canada (NSERC), and FPInnovations. Thanks are extended to the Centre de Recherche Industrielle du Québec (CRIQ) for technical support. Special thanks are owed to Bibiane Desmeules and Gina Rodriguez for their assistance with the laboratory work.

Notes

aAsterisk indicates significance at α = 0.01.

aStandard error.

aAsterisk indicates significance at α = 0.01.

^aStandard error