Conference paper Rust, Göcke & Liebisch (2024) im Proceedings-Band des 23rd International Nondestructive Testing and Evaluation of Wood Symposium, Campinas. USFS Treesearch / FPL-GTR-305 · Volltext (NDT.net).
We introduced acoustic tomography to tree care in 1999. Since then, it has become the standard for advanced tree assessments. However, tomographs still rely solely on travel time. It is often unclear whether a low velocity is caused by a crack, rot or a cavity. This is because transit time reveals only part of the picture.
Our “impactor” prototype takes things three steps further:
-
Automatic tapping with a defined force, which is reproducible rather than manual. The tapping force measurably affects the transit time, frequency and signal quality. This problem is solved by constant excitation.
-
Full signal: not just travel time, but also amplitude, frequency, attenuation and energy are recorded. This makes multivariate analysis possible.
-
Sensor positioning via pattern recognition using a smartphone or tablet instead of tedious manual measurement. With complex cross-sections in particular, positional accuracy determines the velocity estimate, the tomogram shape and the defect size (see Rust (2017) (https://doi.org/10.13141/jflr.v1i1.449)).
Preliminary results: The dominant frequency differs significantly among tree species. In beech trees infected with the fire blight fungus (Kretschmaria deusta), the signals are weaker and more chaotic, with high frequencies (>2500 Hz) absent. Without time-of-flight information, an XGBoost model can detect fire blight infestation in the signal with 98% accuracy. Amplitude and energy are particularly helpful for this purpose.
There is also a practical advantage: Automatic excitation and position tracking enable measurements to be taken from a work platform or by tree climbers — situations in which manual tapping and callipers are impractical.
In short, after 25 years, the question is no longer just ‘How fast does the impulse arrive?’, but ‘What information is contained in the entire signal?’, and ‘How can we take repeatable measurements even under difficult conditions?’.
Find out more: Rust, S.; Göcke, L.; Liebisch, J. A New Tomographic Device for the Non-Destructive Testing of Trees—25 Years On. In: Gonçalves, R.; Wang, X.; Ross, R.J.(eds.) Proceedings: 23rd International Nondestructive Testing and Evaluation of Wood Symposium. FPL-GTR-305. Madison, WI: USDA Forest Service, Forest Products Laboratory, 2024, S. 151–158. https://research.fs.usda.gov/treesearch/69321