Can sonic tomography predict loss in load-bearing capacity for trees with internal defects? A comparison of sonic tomograms with destructive measurements

Autor: Daniel C. Burcham, Brian Kane, Nicholas J. Brazee, Robert E. Marra
Rok vydání: 2019
Předmět:
Zdroj: Trees. 33:681-695
ISSN: 1432-2285
0931-1890
Popis: Sonic tomography can be used to examine reductions in the load-bearing capacity of tree parts with internal defects, but the limitations of sonic tomography and mathematical methods must be considered. The measurement and assessment of internal defects is an important aspect of tree risk assessment. Although there are several methods for estimating the reduced load-bearing capacity of trees with internal defects, the advancement of these methods has not kept pace with improvements to methods used to measure the internal condition of trees, such as sonic tomography. In this study, the percent reduction to the section modulus, ZLOSS (%), caused by internal defects was estimated using 51 sonic tomograms collected from three tree species, and the accuracy of measurements was assessed using the destructively measured internal condition of the corresponding cross sections. In tomograms, there was a repeated underestimation of the percent total damaged area, AD (%), and a repeated overestimation of the offset distance between the centroid of the trunk and the centroid of the largest damaged part, LO (m). As a result, ZLOSS determined using tomograms was mostly less, in absolute terms, than that determined from destructive measurements. However, the accuracy of these estimates improved when using colors associated with intermediate sonic velocities to select damaged parts in tomograms, in addition to the colors explicitly associated with the slowest sonic velocities. Among seven mathematical methods used to estimate ZLOSS, those accounting for LO were more accurate than others neglecting it. In particular, a numerical method incorporating greater geometric detail, called zloss, gave estimates that were consistently better than six other analytical methods.
Databáze: OpenAIRE
Externí odkaz: