Detecting Internal Decay In Trees/Logs Based On Acoustic Tomography And Ct Combining With Mechanical Property

Internal decay is a major structural defect of many tree species. The economic loss caused by internal decay is most significant, and tree decay is also a major concern in relation to human safety. Early detection of internal decay in trees could provide a significant benefit to the industry in terms of making accurate quality assessments and volume estimates and use of the resource. It can also help foresters in prescribing silvicultural treatments for improved management decision-making and thus help maintain a healthy forest, and reduce the accident caused by falling trees. The acoustic tomography and CT techniques were used to detect internal decay of trees/logs in this study. Acoustic tomography is an emerging NDT technology for tree decay detection in both urban community and production forest. It allows the users to visualize the velocity distribution of the acoustic waves as the waves propagate through the cross section of a tree. Computed tomography scanning with ionizing radiation provides three-dimensional information about the internal inhomogeneous structure of the specimen under test in a non-destructive, non-invasive and rapid manner. CT imaging is honored as the best way in wood non-destructive testing.Firstly, the decay types, the reasons cause decay, the hazard of decay, and the necessity of inspecting decay in their early stages were introduced. The achievements of CT and acoustic tomography in wood nondestructive testing (NDT) were introduced, and good application prospect of these two techniques in NDT of wood were put forward in this study. Secondly, the basic principle of CT imaging and factors affecting the quality of CT image were studied, and then the wood specimens were scanned using the suitable scanning parameters. Thirdly, the end-hardness of each grid on each specimen, basic density, compression strength, bending strength, and moisture content of small samples cut from each specimen were tested. A3D color mapping was then built using the values of local physical and mechanical properties. Based on the data and3D mappings, the prediction rule of physical and mechanical properties in trees/logs was given. Fourthly, to evaluate the reliabily and accuracy of TOF acoustic tomography for internal decay detection in trees, Matlab was adopted to build a2D mapping of mechanical property on each specimen, and comparing it to the acoustic tomogram. And then the quantitatively relationship between apparent acoustic velocity and local mechanical property of each disc was calculated. Fifthly, the Hounsfield profile was used to detect type, location, and size of defects. And then, the correlation between CT values under different scan parameters and hardness values/compression strength values were calculated to decide the lowest parameter for wood CT testing. Lastly, The CT images of specimens were processed using fractional Brownian function, and then the internal decay in each specimen was also analyzed according to the Hurst exponent.The following results were gotten:(1) Through the development research of acoustic tomography and CT all over the world, it have been found out these two techniques have very good application prospect in NDT of wood and wood physical properties.(2) In order to achieve fast and accurate wood CT detection, improve wood CT scanning efficiency, and save time and cost, the suitable CT scanning parameters for wood specimens were used:the scanning thickness is2.5mm-5mm, scanning voltage is80KV-120KV, scanning current is10mA-500mA, window level is-400Hu, window width is1200Hu.(3) The end-hardness of each grid on each specimen, basic density, compression strength, bending strength, and moisture content (MC) of small samples cut from each specimen were tested. A3D color mapping was then built using the values of local physical and mechanical properties. The3D mapping provided real defect conditions inside of each specimen. The prediction rule of physical and mechanical property in wood was then given based on the data and image analysis, which provides scientific evidence for evaluation of quality and safety of wood.(4) There showed similar defects information both in acoustic tomogram and mapping of mechanical property on each specimen by comparing them, but the tomogram's display results was not as good as mechanical mapping. Statistical analysis showed poor relationships between apparent acoustic velocity and mechanical property of individual small samples. This indicates that apparent velocity-does not fully reflect the true velocity of acoustic wave transmitting in a cross section of tree. And TOF acoustic tomography has limited capability in detecting early stage of decay in trees.(5) Different scan parameters were used to scan the specimens. The Hounsfield profile was adopted to analyze the defects on each CT image. The results showed the wood internal decay, different levels of decay, and other defects could be inspected by using Hounsfield profile. And then, the correlation between local mechanical properties and CT values were analyzed. The results showed the scan parameter of10mA80KV1s appeared to be a good choice, which used a relatively low radiation dose, and produce reasonable resolution for wood CT scanning. After that, a few relatively high scan parameters were used to scan wood specimen, the result shows the correlation between CT values and local mechanical properties were tend to remain same.(6) The fractional Brownian function was used to process CT images of wood samples, and the internal decay was also analyzed. The following results were gotten:the Hurst exponent (H) can be as the reference to distinguish the defects from the background. WhenO<H<1, their set was the background (sound wood area); when H>1, their set was the defects region.