Ninth International Symposium on Nondestructive Testing of Wood Application of ultrasound nondestructive evaluation to grading pallet parts

High-quality pallets are easier to repair and last longer. To obtain the high-quality parts necessary to build durable pallets requires grading and sorting by an automated inspection system. Ultrasonic sensing was selected for this application because 1) it is relatively inexpensive; 2) it can penetrate and characterize internal features of wood specimens; and 3) its effectiveness has been demonstrated for certain features of wood. This study has sought to 1) determine the ultrasonic parameters that are most sensitive to pallet part defects; and 2) specify which aspects of ultrasonic response signals best characterize defect size and location. Time-of-flight measurements appear reliable for delineating knots from clear wood. Also, preliminary indications are that peak amplitude may be sensitive to areas with splits/ checks/shake. Results from this work and subsequent studies will be applied to the development of an industrial prototype that automatically detects defects and grades parts according to established visual grading rules. The authors are, respectively Res. Forest Prod. Technologist, USDA Forest Serv., Brooks Forest Prod. Center, VPI&SU, Blacksburg, Vs.; Professor, Dept. of Engineering Sci. and Mechanics, Materials Sci. and Engineering, VPI&SU; Laboratory Technician, Dept. of Engineering Sci. and Mechanics, VPI&SU; and Laboratory Technician, Dept. of Wood Sci. and Forest Prod., Brooks Forest Prod. Center, VPI&SU. This work was partially supported by the Southeastern Forest Expt. Sta., USDA Forest Serv. Introduction Wood pallets are the largest single use of sawn hardwood logs in the United States. Unfortunately, millions of wood pallets are discarded annually due to damage or because their low cost makes them readily disposable. High-quality wood pallets, on the other hand, have a much longer life-cycle and promote reuse. However, building durable pallets requires high-quality pallet parts. Because manual grading and sorting of pallet parts is not feasible, we have a long-term goal to develop an automated pallet-part inspection system. We envision that it will scan moving parts in a production environment, locate and identify defect areas, and grade the parts according to established visual grading rules, Most wood pallets are constructed from two types of pallet parts (Fig. 1): 1) stringers, the structural center members that support the pallet load; and 2) deckboards, the topand bottomfacing materials that provide dimensional stability and product placement. There are many variants of this basic design, but most pallets contain solid wood components that are produced from lumber or from the center cant material of logs. Cant material has a high percentage of defect area and is generally not highly valuable for other solid wood products. Ultrasonic testing has been used by other investigators to examine wooden objects. Previous work has established that knots can be detected in hardwood and softwood lumber (5,6), decay is detectable in structural timbers (’7,9), beams can be graded for strength in situ (8), dry-coupled transducers can be used to detect knots and decay Schmoldt, Duke, Morrone, and Jennings — 183 in timber bridge members (2,3), knots, decay, and cross grain are detectable in small laboratory samples (4), and wooden art objects can be examined for small cracks and other degradation (1). In most of these studies, softwood species, such as southern yellow pine (Pinus spp.), Suisse spruce, Douglas-fir (Pseudotsuga menziesii (Mirb.) France), and white fir (Abies concolor (Gord. and Glend.) Lindl. ex Hildebr.), were examined. With some exceptions (2,3 ,5,6), reports on the application of ultrasound to hardwood species is limited. Physical and mechanical properties of hardwoods differ from softwoods, and previous ultrasonic examination methods for softwoods may not directly translate to hardwood applications. Because many different hardwood species are used in pallets, inspection methods must be adaptive to interspecific dissimilarities. Additionally, pallet parts are rough cut, and therefore wood surfaces can vary greatly in smoothness and in reflectance of elastic waves.