UNIVERSIDADE ESTADUAL DE CAMPINAS
Faculdade de Engenharia Agrícola
MARIANA NAGLE DOS REIS
ASSOCIAÇÃO DE MÉTODOS NÃO DESTRUTIVOS PARA
INSPEÇÃO DE ÁRVORES.
ASSOCIATION OF NONDESTRUCTIVE METHODS FOR
TREE INSPECTION
CAMPINAS
2017
MARIANA NAGLE DOS REIS
ASSOCIAÇÃO DE MÉTODOS NÃO DESTRUTIVOS PARA
INSPEÇÃO DE ÁRVORES.
Dissertação
apresentada
à
Faculdade
de
Engenharia Agrícola da Universidade Estadual de
Campinas como parte dos requisitos exigidos para a
obtenção do título de Mestra em Engenharia
Agrícola, na área de Concentração Construções
Rurais e Ambiência.
Orientadora: Profa. Dra. Raquel Gonçalves
Co-orientador: Dr. Alex Julio Trinca
CAMPINAS
2017
AGRADECIMENTOS
Agradeço a professora Raquel pela dedicação, empenho e carinho, aos amigos do LabEnd por
toda ajuda nos experimentos e à minha família pelo apoio incondicional. Agradeço também
ao Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPQ pela bolsa de
estudos e à FAPESP (Proc. 2015/05692-3) pelo financiamento da pesquisa. Agradeço a
Diretoria de Meio Ambiente da UNICAMP pela doação das toras utilizadas nos ensaios e à
PD Instrumentos pelo auxílio e equipamento utilizado nos ensaios de resistência a perfuração.
RESUMO
A arborização é importante para propiciar equilíbrio ao ambiente, liberar oxigênio e
absorver gás carbônico, melhorar a qualidade do ar, ofertar sombra, absorver ruídos, fornecer
proteção térmica, quebrar a monotonia da paisagem, abrigar e alimentar a fauna e propiciar
bem-estar às pessoas. No entanto, as árvores que se localizam nas proximidades de
habitações, equipamentos urbanos ou estruturas agrícolas, podem representar riscos humanos
e financeiros quando seu estado fitossanitário está comprometido. Detectar o estado
fitossanitário de uma árvore nem sempre é possível utilizando-se somente de análises visuais
ou sinais externos de enfermidades ou ataques de fungos e organismos xilófagos, tornando as
técnicas de inspeção fundamentais. Essa pesquisa teve como objetivo avaliar a tomografia
ultrassônica e a resistência a perfuração, de forma isolada e associada, na detecção dos níveis,
das dimensões e da localização de deteriorações. Os ensaios foram realizados em toretes de 6
espécies de árvores, com diferentes tipos e níveis de deterioração. No caso da tomografia a
avaliação foi feita com base nas variações de velocidade fornecidas pela imagem tomográfica
e para a resistência a perfuração com base no gráfico de amplitude de resistência. Os
resultados permitiram concluir que a resistência a perfuração foi eficiente na detecção e na
obtenção da dimensão aproximada de ocos, uma vez que a amplitude nestes casos é zero. As
amplitudes em regiões deterioradas são inferiores (média em torno de 4%) às de zonas de
madeira sã, permitindo inferir a localização destas zonas. No entanto, a identificação do nível
da deterioração não é evidente, já que a amplitude varia entre cerne alterado e alburno e,
também, entre espécies. A resistência a perfuração é um ensaio pontual e, assim, sua
eficiência depende da localização adequada para ser executado. No caso da tomografia, as
zonas ocadas apresentam redução de velocidades superiores a 70%, enquanto as zonas
deterioradas por fungos começam a ser destacadas com reduções de 30% na velocidade. No
caso de fendas ou galerias o detalhamento da imagem depende da relação entre o
comprimento de onda e a dimensão destes defeitos. De maneira geral as imagens de
tomografia ultrassônica não permitem a obtenção do formato e da localização exata da área
deteriorada, mas permite aportar informações gerais da condição da madeira inspecionada. A
associação dos métodos é eficaz, pois a resistência a perfuração permite detalhar a condição
da madeira nas zonas destacadas pela tomografia como suspeitas.
ABSTRACT
The urban forestation is important to balance the environment, release oxygen and
absorb carbon dioxide, improve air quality, provide shade, absorb noise, provide thermal
protection, break the monotony of the landscape, as shelter and food to the fauna and provide
wellness to people. However, trees that are located near housing, urban equipment or
agricultural structures may pose human and financial risks when their sanity is compromised.
Detecting the health status of a tree is not always possible using only visual analyzes or
external signs of diseases or attacks of fungi and xylophage organisms, making very
important the improvement of inspection techniques. The aim of this research was to evaluate
ultrasound and drilling resistance, isolated and associated, in the detection of levels,
dimensions and location of decays. The tests were carried out on 6 tree species, with different
types and levels of deterioration. In the case of ultrasound tomography, the evaluation was
made based on the velocity variations provided by the tomographic image and for the drilling
resistance based on the graph of resistance amplitude. The results allowed concluding that the
drilling resistance was efficient in detecting and obtaining the approximate size of hollows,
since the amplitude in these cases is zero. The amplitudes in deteriorated regions are inferior
(average around 4%) to those of healthy wood zones, allowing to infer the location of these
zones. However, the identification of the level of decay is not evident, since the amplitude
varies between modified core and sapwood and also between species. The drilling resistance
is a punctual test and its efficiency depends on the proper location to be executed. In the case
of tomography, the hollow zones present reduction of velocities greater than 70%, while the
zones deteriorated by fungi begin to be highlighted with 30% velocities reductions. In the case
of cracks or galleries the image detailing depends on the relationship between the wavelength
and the size of these defects. In general, the ultrasonic tomography images do not show the
format and the exact location of the decayed area, but it provides general information of the
condition of the inspected wood. The association of the methods is effective, because the
drilling resistance allows detailing the condition of the wood in the areas highlighted as
suspect by the tomography.
Sumário
INTRODUÇÃO GERAL ... 9
ARTIGO 1: DRILLING RESISTANCE AMPLITUDE IN DIFFERENT TYPES OF
DETERIORATED LOGS OF TREES... 12
ARTIGO 2: ULTRASONIC TOMOGRAPHY IN LOGS OF TREES WITH DIFFERENT
TYPES OF DETERIORATION ... 26
ARTIGO 3: ASSOCIATION OF NONDESTRUCTIVE TOOLS FOR TREE INSPECTION ... 42
DISCUSSÃO GERAL ... 59
Resistência a perfuração ... 59
Tomografia ultrassônica ... 61
Associação da tomografia ultrassônica e resistência a perfuração ... 63
CONCLUSÃO GERAL ... 65
ANEXO I ... 66
Gráficos de Resistência à Perfuração em todas as rotas de medição dos toretes ensaiados ... 66
ANEXO II ... 85
Tabelas de dados utilizados para geração das imagens de tomografia ultrassônica ... 85
ANEXO III ... 92
INTRODUÇÃO GERAL
A importância da existência de arborização é indiscutível, sendo, portanto, um tema de
grande interesse e atualidade. Políticas de manutenção de áreas verdes têm grande relevância
social, estética, cultural e educacional, além de desempenharem significativo papel do ponto
de vista climático e ambiental.
A falta de planejamento da arborização urbana no passado e as atuais mudanças
climáticas podem provocar declínio nas árvores, com alterações de dimensão, de formato e de
posição em relação ao eixo, tornando estas árvores mais propícias a problemas patológicos e
posteriormente à queda. Adicionalmente, as árvores são seres vivos e, como tal, têm um ciclo
de vida que envolve o início do crescimento, a fase juvenil, a fase adulta e a morte, mesmo
sem nenhum agente externo causador.
Assim, a manutenção de áreas verdes requer ações políticas, administrativas e técnicas
e, do ponto de vista da coexistência das árvores com as estruturas que envolvam riscos
humanos ou financeiros, devem ser estabelecidos critérios que permitam a aplicação de novas
metodologias que minimizem os problemas.
A inspeção visual dirigida à poda, já praticada, muitas vezes não contempla a
identificação de anomalias que sejam indicativas de fragilidades ou instabilidades internas. A
alternativa vislumbrada nesse projeto é o uso de tecnologias que, associadas à inspeção visual,
permitam monitorar a condição interna da árvore, de forma a ser possível, em uma fase
posterior, ligar esse conhecimento ao risco de queda desse indivíduo arbóreo.
Na madeira, diferentes tipos de deteriorações provocam, como consequência,
alterações em sua estrutura. Por outro lado, a propagação de ondas mecânicas em meios
materiais é afetada, por diferenças de propriedades elásticas, por diferenças de impedância
entre os meios e por alterações de percurso. Por essa razão, existe a viabilidade de se utilizar
variações na velocidade de propagação de ondas como parâmetros de inferência de
modificações estruturais na madeira (alterações na rigidez ou na estrutura anatômica e zona
ocadas).
A tomografia utilizando métodos sônicos tem sido adotada com frequência para
avaliação da condição interna de árvores em diversos países, principalmente como ferramenta
auxiliar na avaliação do risco de queda. Existem no mercado vários equipamentos comerciais
de tomografia acústica, utilizando, principalmente, propagação de ondas de tensão
(denominados tomógrafos sônicos), mas ainda tem havido muita pesquisa na área, fruto da
necessidade de melhor entendimento do significado da imagem gerada, bem como de seu
alcance em termos de precisão.
Da mesma forma, a resistência a perfuração tem sido reconhecida como técnica que
apresenta bons resultados em inspeções de árvores e de estruturas, embora de forma muito
localizada. Considerando que o equipamento mede a resistência que a madeira oferece à
perfuração, é esperado, também, que seus resultados possam ser correlacionados com perdas
de resistência devidas a deteriorações. Assim, tem havido vários estudos que visam avaliar as
correlações entre a amplitude da resistência a perfuração, obtida pelo equipamento, e a
resistência da madeira ou sua eficácia na detecção de deteriorações em estruturas, toras ou
árvores.
Considerando os aspectos mencionados o objetivo geral desta pesquisa foi avaliar a
tomografia ultrassônica e a resistência a perfuração, de forma isolada e associada, na detecção
dos níveis, das dimensões e da localização de deteriorações. No caso da tomografia
ultrassônica é importante salientar que os ensaios foram realizados com equipamento e
software desenvolvidos no grupo de pesquisa da Faculdade de Engenharia Agrícola
(FEAGRI) da Universidade Estadual de Campinas (UNICAMP).
A dissertação foi redigida em formato alternativo previsto pela Intuição, no qual após
o capítulo de Introdução Geral da pesquisa são apresentados os artigos submetidos à
publicação, o capítulo de Discussão Geral, que aglutina os resultados dos artigos, e o capítulo
de Conclusões Gerais, que estabelece o vínculo dos artigos com a hipótese e objetivo
principal da pesquisa.
Os dois objetivos específicos compuseram os temas dos dois primeiros artigos, que
tratam, isoladamente, da avaliação da resistência à perfuração e da tomografia ultrassônica,
respectivamente. O terceiro artigo trata da associação destes métodos não destrutivos para a
avaliação interna de árvores. Artigos possuem lógica e formatação específicas, exigindo
apresentação resumida e focada de resultados, sendo assim, foram inseridos em anexos os
resultados completos obtidos durante a pesquisa.
No Anexo I são apresentados todos os gráficos de resistência à perfuração obtidos nos
ensaios com resistógrafo, os quais foram a base das análises envolvidas nos artigos 1 e 3.
No Anexo II são apresentadas as tabelas de dados utilizados para geração das imagens
de tomografia ultrassônica, utilizadas para a elaboração dos artigos 2 e 3.
No Anexo III apresentam-se as imagens de tomografia ultrassônica utilizadas como
base para a elaboração dos artigos 2 e 3.
ARTIGO 1:
D
RILLING RESISTANCE AMPLITUDE IN DIFFERENT TYPES OF
DETERIORATED LOGS OF TREES
D
rilling resistance amplitude in different types of deteriorated logs of trees
.
Authors and Afilliations:
Mariana Reis (1), Raquel Gonçalves (2), Gustavo Garcia (3), Leandro Manes (4)
(1) (3) Master Student, School of Agricultural Engineering, Nondestructive Testing Laboratory, University of Campinas, Campinas, SP, Brasil
(2) Professor, School of Agricultural Engineering, Nondestructive Testing Laboratory, University of Campinas, Campinas, SP, Brasil. E-mail: raquel@agr.unicamp.br
(4) Undergraduate Student, School of Agricultural Engineering, Nondestructive Testing Laboratory, University of Campinas, Campinas, SP, Brasil
Author Contribution Statement:
(1) This paper is part of her Master degree research. Worked in the tests, in discussions of results and in text writing.
(2) Supervisor of the Master student (1). Responsible for the research idea. Worked supervising the developing of the research, discussing the methodology and results and on the final correction of the text.
(3) Part of the research group, collaborated in the tests, prepared the images and participated on discussions of the results.
(4) This paper is part of his Undergraduate research. Collaborated with author (1) during the tests and also participated of the discussion of the results.
Key Message
This paper contributes, with theoretical bases discussion and statistical analysis, for interpretation of drilling technology results that, although widely used, has no the directly interpretation released by manufacturers.
Abstract
The drilling resistance test has been widely used in tree inspections and structures. As the needle pierces the wood, the equipment registers, in amplitude graphs (%), the resistance to drilling offered by the wood. So, it´s expected that in the presence of a hollow, the amplitude obtained in the graph is zero, allowing its detection. However, when the wood doesn´t have a hollow, the results of the amplitude doesn´t have an obvious interpretation and can be influenced by different factors. The objective of this research was to use discs, from different tree species and with different types of deterioration, to study the behavior of the amplitude of the drilling resistance. In this research, the amplitude ranged from 0% to 40%, with the 0% sections corresponding to hollows, allowing the confirmation that this tool is adequate for the detection of the location and approximate size of hollows. In the heartwood, the values of amplitudes were superior and statistically different from those obtained in the sapwood. There were statistical differences between drilling amplitude among the studied species. The variance analysis showed that the amplitude in deteriorated regions was always lower (average around 4%), except in the wood attacked by Coleoptera, whose deterioration was not captured by the equipment. In some discs, there were peaks of amplitude of the drilling resistance, probably associated to compartmentalization zones promoted in the tree tissues to protect themselves from the advance of the deteriorated zone. This behavior may affect the interpretation of the inspection.
Introduction
The drilling resistance is a semi-destructive test in which a drilling needle is inserted in the wood under inspection. The needle must be very thin to capture tangential density differences and minimize material damage and, on the other hand, can´t be so thin that could suffer from resonance or buckling problems (Tannert et al., 2013). In general, the needle has a 3 mm diameter tip where the cutting edges are located, and 1.5 mm diameter shaft (Nutto & Biechele 2015, Tannert et al. 2013). With the rotation, the needle cuts the wood and perforates the material as it advances (Nutto & Biechele 2015, Botelho 2006). Some equipment allows the variation of the advance speed of the needle (cm / min) as it´s rotation speed (turns / min), and it´s already known that this adjustment must be done according to the hardness or the density of the wood (Nutto & Biechele 2015). Besides the velocities, sharpening of the needle tip is also very important to reduce friction, which is one of the responsible for interpretation problems of this technique´s results, especially in dense woods (Nutto & Biechele 2015, Tannert et al. 2013). While drilling, the needed energy is measures depending on the drilling depth of the needle, and registers this parameter in a percent-amplitude graph. The more resistance the wood offers to drilling, the greater the energy required. Some manufacturers offer new versions featuring two amplitude curves, one representing the resistance of the drill to the rotation and another that represents the resistance to drilling, this last one less affected by friction.
It is expected that in the presence of a hollow, the amplitude obtained in the drilling resistance graph is zero or close to zero, allowing its detection. However, when the wood doesn´t have a hollow, the meaning of the amplitude of drilling resistance has no obvious interpretation.
Considering that the equipment measures the resistance offered by the wood to drilling, it´s also expected that its results can be correlated with resistance losses due to deterioration. So, there has been several studies aiming the evaluation of the correlations between the amplitude of the drilling resistance obtained by the equipment and the wood resistance (Botelho Jr., 2006) or its effectiveness in detecting deteriorations in wood structures (Brashaw et al. 2011; Tannert et al. 2013, Rinn 2012), logs (Wang et al. 2005 apud Nutto & Biechele 2015) or trees (Johnstone et al. 2010; Kubus 2009, Johnstone et al. 2007, Wang et al. 2008).
In order to contribute with information that would allow the interpretation of the drilling resistance test, the objective of this work was to use real images of wood discs of different species and with different types of deterioration to study the behavior of the drilling resistance amplitude.
Methodology
Sampling was composed of logs collected from six tree species: Centrolobium sp., Tabebuia ochracea, Liquidambar styraciflua, Platanus sp., Poencianella pluviosa and Copaifera sp., widely spread in the urban arborization of São Paulo State, Brazil. From these logs were removed 6 discs approximately 200 mm high. Eight equidistant points were marked on each discs perimeter for measurements with the drilling resistance
equipment (IML F400, Germany) - Figure 1. The measurements were performed in the perpendicular direction to the grain in the 8 marked points (measurement routes), obtaining 48 graphs of amplitude of drilling resistance. To help with the test, the discs were fixed to a concrete table with sergeants (Fig 1).
Fig 1 Drilling Resistance Test in discs
The species used in the research had different densities and deterioration conditions (Table 1). The model of equipment used allows the choice of different levels of feed speed (from 15cm / min to 200cm / min) and needle rotation (1500 revolutions per minute at 5000 rotations / minute). This adjustment is usually made according to the density of the wood. In light woods the feeding speed must be higher to allow sufficient amplitude so that the curve variations can be seen. In denser woods the feed speed must be reduced and the rotation increased so drilling is possible. Considering these aspects of the equipment, feeding and rotation velocities were adopted according to the wood density, and it was also necessary to consider the deterioration condition of the disc, avoiding that it would break with the needle entry (Table 1).
The wood condition influenced the choice of feed velocity, although the species Liquidambar and Platanus has practically the same density, the feed velocity used in the Liquidambar was lower than in the Platanus (Table 1). The feed speed used in the Copaifera sample could also be 100 cm / min, but it was necessary to reduce it and increase the rotation so that the drill inlet did not break the disc (Table 1).
Table 1 Values of basic density (bas), feed velocity (VA) and rotation speed (VR) adopted for each species
Species bas kg.m-3 VA cm/min VR rot/min Liquidambar styraciflua 490* 100 2500 Platanus sp. 500** 150 2500 Copaifera sp. 575*** 50 3500 Centrolobium sp. 660**** 100 2500 Tabebuia ochracea 840*** 100 2500 Poencianella pluviosa 890**** 50 3500 *Lima et al. 2015; **Freitas 2012; ***IPT http://www.ipt.br/informacoes_madeiras3.php?madeira=38; ****Remade http://www.remade.com.br/madeiras-exoticas/116/madeiras-brasileiras-e-exoticas/arariba
After the drilling resistance tests were completed, the discs were cut to the exact position of the drill travel on the different routes. These sections were detailed through photographic record in order to compare the amplitude plots of drilling resistance with the different zones of the wood through which the drilling occurred. The equipment model used in this research provides two amplitude curves, one representative of the power supply (power needed for drilling) and the other for the drilling resistance. However, older models provide only the feed curve. Nutto and Biechele (2015) observed that in high density wood the feed amplitude tends to increase with the drilling depth due to friction. This increase in amplitude may cause wrong interpretation of the results, since it can camouflage the amplitude drop that would be caused by deterioration.
The discs were macroscopically analyzed to identify the heartwood, sapwood and deteriorated areas. Descriptions of the species were obtained in the literature (Lima et al., 2015, Freitas 2012 and IPT and Remade sites) as well as experienced person support in visual assessment. For the species in which it was possible to distinguish the heartwood and sapwood regions, the average amplitude of drilling resistance obtained separately in the bark, heartwood, sapwood and deterioration zones was determined.
For all discs, the drilling resistance graphs obtained on each route were compared with the photographic images, by image overlay. In the case of discs with heart and sapwood zones, the average amplitudes of drilling resistance were statistically analyzed using multivariate variance analysis. In this research, only the amplitude plots of drilling resistance were used because the use of the two graphs (power amplitude and resistance to drilling) made the image overlay very loaded with information.
Results and Discussion
In the Centrolobium, Platanus, Poencianella and Copaífera discs, it was possible to define the heartwood and sapwood zones from coloring distinction (Fig 2), whereas at the Tabebuia and Liquidambar species this distinction was not possible (Fig 3). The characteristics of the heartwood and sapwood differentiation as a function of color differentiation were similar to those proposed in the literature (Chudnoff 1980; Lima et al., 2015, Freitas 2012 and IPT and Remade sites). Not all of the discs contained hollowed areas and all differed in types and levels of deterioration (Table 2).
Fig 2 Images of the Cetrolobium sp. (a), Platanus sp. (b), Poencianella pluviosa (c) and Copaifera sp. (d) discs with color distinction between heartwood and sapwood
Fig 3 Images of the discs of the species Tabebuia ochracea (a) and Liquidambar styraciflua (b), without distinction of color between heartwood and sapwood
Table 2 Short description of disc deterioration
Species Description
Centrolobium sp. Hollow caused by termites Tabebuia ochracea Coleopterans attack
Liquidambar styraciflua Near the pith there is a small area with an early stage of fungal decay and lateral cracking from the pith to the bark
Platanus sp. Most of the wood shows signs of fungi attack and there are some hollowed areas caused by termites
Poencianella pluviosa Fungi attack at the center of the disc
Copaifera sp. Hollow caused by termites and fungi deterioration around them
Considering the methodology, it was possible to obtain 8 graphics of drilling resistance amplitude from each disc, corresponding to the 8 measurement routes, which were overlaid with the photographic images (Example in Fig 4). The same result observed in the example of Fig 4 was obtained for the six-species studied (48 overlapped images).
Fig 4 Superposition of the amplitude graph of drilling resistance to the photographic image obtained from the
disc, in the position of the needle passage of the resistograph. Species: Centrolobium sp.
The detailed analysis of the 48 superimposed images allowed to verify the existence of a pattern of behavior of the amplitude of drilling resistance against the different zones of the wood (healthy and with deterioration) traveled by the needle during the drilling.
For all studied species, the amplitude ranged from 0% to 40%, with the 0% sections corresponding to the hollows, as expected. This result confirms that this tool is adequate to detect the location and the approximate size of hollows. The same result was highlighted by Kubus (2009) when analyzing the condition of a large monumental tree in Poland. The author presents records of 10 graphs of resistance to drilling, obtained in different positions of the tree, in which there are several zones with amplitudes close to zero (probably hollowed zones) and, in the other regions, average amplitudes of resistance to perforation of 16%. In the bark area the average amplitudes were 6%. Brashaw et al (2011) obtained drilling resistance varying from 0 to 25% in zones with different levels of wood degradation.
In this research, in the species in which it was possible to define the heartwood and sapwood regions, there was amplitude variation in these regions (Examples in Fig 4). In the two species in which it was not possible to visualize the distinction between heartwood and sapwood (Tabebuia ochracea and Liquidambar styraciflua), the amplitude of the drilling resistance graph also does not show variations (Ex: Fig 5). Similar graphs, without heartwood and sapwood visual distinction, were obtained by Kubus (2009).
Fig 5 Superposition of the amplitude graph of drilling resistance to the photographic image obtained from the disc, in the position of the needle passage of the resistograph. Species: Liquidambar styraciflua
In the heartwood, the values of amplitudes were higher than those obtained in the sapwood. For the four species in which it was possible to define the heartwood and sapwood regions (Fig 2) this result was statistically demonstrated (Fig 6). The analysis of variance showed that amplitudes in deteriorated regions (with or without hollows) were always lower (average around 4%), always higher in the heartwood regions (average around 22%) and sapwood region (average around 15%) and of bark (average around 14%) intermediate and statistically equivalent (Fig 6).
Fig 6 Representative graph of the average values and the variability of amplitude obtained in the different zones of the disc
Considering only the heartwood and sapwood regions without apparent deterioration, there was a differentiation between species. The amplitudes of drilling resistance were higher (about 25%) and statistically different for Platanus sp. and the same (about 11%) for the other species. This result is not expected, since the Platanus sp. has the lowest density (Table 1) and the drilling resistance, in general, has a positive correlation with density (Couto et al., 2013, Costelo & Quarles 1999). However, it is important to note that the amplitude of drilling resistance can be affected by the advance velocity, which for Platanus sp. was higher (Table 1) and that, in order to properly interpret as well as to compare amplitude results obtained in the resistance test, it´s necessary to know, in advance, the profiles obtained in the whole condition (Martinez 2016, Matheny et al., 1999). For the Poencianella pluviosa hearthwood, denser wood (Table 1), was statistically different from the others, with a drilling resistance of about 30% compared to 22% of the others.
For the Tabebuia ochracea, with beetle insect attack (Fig 3), the amplitude of drilling resistance did not show variations that allowed, in an inspection, to visualize the wood state (Example in Fig 7).
Zone (1 = bark; 2 = sapwood; 3 = heartwood; 4 = decay wood
A m pl it ud e of D ri ll ing r esi st an ce (% ) 1 2 3 4 0 4 8 12 16 20
Fig 7 Superposition of the amplitude graph of drilling resistance to the photographic image obtained from the disc, in the position of the needle passage of the resistograph. Species: Tabebuia ochracea. Deterioration caused by Coleoptera insects
In the Liquidambar styraciflua disc there was a significant crack (Fig 3) that was not identified in any drilling resistance amplitude graph, because no needle route passed through the crack site (Fig 5). This result stands out the fact that this type of inspection is punctual, and it is important to know where it should be applied. Botelho (2006) and Tannert et al (2014) had the same conclusion.
It was possible to observe that in some discs there was, in the contour of the fungi deteriorated zone, a darker colored region. In these cases, an increase peak in the amplitude of the drilling resistance was observed (Fig 8). Similar result was obtained by Rinn (1996) and was explained by the compartmentalization process promoted in the trees tissue to protect them from the advance of the deteriorated zone. In this process the tree develops a thicker cell wall tissue, in addition to obstruct cell voids, causing this tissue to act as a deterioration barrier (Fraedrich 1999, Shigo 1977, Shortle 1979). The darker color is due to the antimicrobial substances produced (Shigo1977, Shortle 1979). The Platanus sp. disc was the only one that presented these peaks in the sapwood zone, which may have influenced the average value of amplitude of drilling resistance, making this species, although with lower density, present the highest values of amplitude.
Fig 8 Details of amplitude peak observed in zones close to fungal decay a) Centrolobium sp., b) Platanus sp., c) Poencianella pluviosa, d) Copaifera sp
Rinn (1996) observed that differentiated patterns of behavior of the drilling resistance plot may help the inspection interpretation. The authors point out that in fungi deteriorated wood, the amplitudes are low and approximately homogeneous, while in termite deteriorated areas or with cracks present the graph is more variable, with higher amplitudes (surrounding wood in good condition) followed by amplitude values much smaller and localized. In the Platanus sp. and Poencianella pluviosa discs, in which there was a large fungus deteriorated zone (Fig 2 and Table 2), the drilling resistance graph shows a continuous segment of low amplitudes (Fig 9a and b), according to the behavior indicated by Rinn (1996).
Fig 9 Superposition of the amplitude graph of drilling resistance to the photographic image obtained from the disc, in the position of the needle passage of the resistograph. Species: a) Platanus sp.., b) Poencianella pluviosa
Conclusion
For all studied species, the amplitude ranged from 0% to 40%, with the 0% sections corresponding to the hollows, as expected, allowing confirming that this tool is adequate for the detection of the location and the approximate size of hollows. In the heartwood, the values of amplitudes were superior and statistically different from those obtained in the sapwood. There were also statistical differences between species. The variance analysis showed that the amplitude in deteriorated regions (with or without hollows) was always lower (average around 4%), except in the Coleoptera attacked wood, whose deterioration was not captured by the equipment. In some discs, amplitude increase peaks were observed on the drilling resistance graph, probably associated to compartmentalization zones promoted in the tree tissue to protect themselves from the advancement of the deteriorated zone. This behavior, also observed by other researchers, can affect the inspection interpretation, hiding, for example deterioration.
Acknowledgements
The authors would like to thank the National Council for Scientific and Technological Development (CNPQ) for the scholarships and Sao Paulo Research Foundation (FAPESP) - Proc. 2015/05692-3 - for the research funding. They also thank the Environment Department of UNICAMP for donating the logs used in the tests and to PD Instruments Company for lending the equipment used in tests of drilling resistance.
Conflict of Interest
The authors declare that they have no conflict of interest.
References
BOTELHO JR, J.A. (2006) Avaliação não destrutiva da capacidade resistente de estruturas de madeira de edifícios antigos. Thesis, Universidade do Porto, Portugal
BRASHAW, B. K.; VATALARO, R.; ROSS, R. J.; WANG, X.; SCHMIEDING, S.; OKSTAD, W. (2011) Historic Log Cabin structural condition assessment and rehabilitation – A case study. In: International nondestructive testing and evaluation of wood symposium, 17, Hungria. Anais... Hungria: University of West Hungary, v.2, p. 505-512
COSTELO L.R. QUARLES, S.L. (1999) Detection of wood decay in Blue Gum and Elm: an evaluation of the resistograph and the portable drill. Journal of Arboriculture, EUA, 25(6), p. 311 – 318
COUTO AL, TRUGILHO RF, NEVES TA, PROTÁSIO TP, Sá VA. (2013) Modeling of basic density of wood from Eucalyptus grandis and Eucalyptus urophylla using nondestructive methods. Cerne, v.19, n.1, p.27-34
FRAEDRICH, B. R. (1999) Compartmentalization of decay in trees, TR-18, Technical Report, Bartlett Tree Research Laboratories, Charlotte, NC
FREITAS, T. (2012) Caracterização tecnológica da madeira de Platanus sp. Monografia, Departamento de Ciências Florestais e da Madeira, Universidade do Espirito Santo, ES
INSTITUTO DE PESQUISAS TECNOLÓGICAS, IPT. Informações sobre Madeiras. Disponível em <http://www.ipt.br/informacoes_madeiras/31.htm> Acessed 04 November 2016
JOHNSTONE, D. M.; ADES, P. K.; MOORE, G. M.; SMITH, I. W. (2007) Predicting Wood Decay in Eucalypts Using an Expert System and the IML- Resistograph Drill. Arboriculture & Urban Forestry, 33(2), p. 76-82
JOHNSTONE, D.; MOORE, G.; TAUSZ, M.; NICOLAS, M. (2010) The Measurement of Wood Decay in Landscape Trees, Arboriculture & Urban Forestry, 36(3), p. 121-127
KUBUS, M. (2009) The Evaluation of Using Resistograph when Specifying the Health Condition of a Monumental Tree. Notuale Botanicae Horti Agrobotanici Cluj-Napoca, v. 37, n. 1, p. 157–164
MARTÍNEZ, R. et al. (2015) NDT to identify biological damage in wood. In: International nondestructive testing and evaluation of wood symposium, 19, Brasil, Anais…Brasil: Rio de Janeiro, p. 453-461.
MARTÍNEZ, R. (2016) Métodos no destructivos de estimación de la densidad de la madera. Thesis, 210p. Universidad de Santiago de Compostela, Espanha
MATHENY, N.P.; CLARK, J.R., ATTEWELL, D. (1999) Assessment of fracture moment and fracture angle in 25 tree species in the United States using fractometer. Journal of Arboriculture, v.25-1, p.18 – 23
NUTTO, L.; Biechele, T. (2015) Drilling resistance measurement and the effect of shaft friction – using feed force information for improving decay identification on hard tropical wood. In: International nondestructive testing and evaluation of wood symposium, 19 Anais… Brasil, Rio de Janeiro, p. 156-160
RINN, F. (2012) Basics of Typical Resistance-Drilling for Timber Inspection. Holztechnologie, 53, p. 24-29
SHIGO, A.L.; MARX, H. (1977) Compartimentalization of decay in trees. (CODIT). Inf. Bull. U.S. Dep. Agric., 405, 73p
SHORTLE, W.C. (1979) Mechanisms of Compartimentalization of Decay in living trees. Symposium on wood decay, v.69, n.10, p.1147-1151
TANNERT, T.; ANTHONY, R.W.; KASAL, B.; KLOIBER, M.; PIAZZA, M.; RIGGIO, M.; RINN, F.; WIDMANN, R.; YAMAGUCHI, N. (2013): In situ assessment of structural timber using semi-destructive techniques. Materials and Structures DOI 10.1617/s11527-013-0094-5. July 2013
WANG, X.; ALLISON, R.B. (2008) Decay Detection in Red Oak Trees Using a Combination of Visual Inspection, Acoustic Testing, and Resistance Microdrilling. Arboriculture & Urban Forestry, 34(1), p. 1-4
WILCOX, W.W. (1978) Review of literature on the effects of early stages of decay on wood strength, Wood and Fiber, p. 252-257
ARTIGO 2: ULTRASONIC TOMOGRAPHY IN LOGS OF TREES
WITH DIFFERENT TYPES OF DETERIORATION
Ultrasonic tomography in logs of trees with different types of deterioration
Authors and Affiliations:
Mariana Reis (1), Raquel Gonçalves (2), Stella Stopa Assis Palma (3), Danilo Profeta Ziller (4)
(1) (3) Master Student, School of Agricultural Engineering, Nondestructive Testing Laboratory, University of Campinas, Campinas, SP, Brasil
(2) Professor, School of Agricultural Engineering, Nondestructive Testing Laboratory, University of Campinas, Campinas, SP, Brasil. E-mail: raquel@agr.unicamp.br
(4) Undergraduate Student, School of Agricultural Engineering, Nondestructive Testing Laboratory, University of Campinas, Campinas, SP, Brasil
Author Contribution Statement:
(1) This paper is part of her Master degree research. Worked in the tests, in discussions of results and in text writing.
(2) Supervisor of the Master student (1). Responsible by the research idea. Worked supervising the developing of the research, discussing the methodology and results and on the final correction of the text.
(3) Part of the research group, collaborated on discussions of the results and in text writing.
(4) This paper is part of his Undergraduate research. Collaborated with author (1) during the tests and also participated of the discussion of the results.
Key Message
This paper contributes, with theoretical bases discussion and statistical analysis, for interpretation of ultrasonic tomography technique that, although widely used, needs improvement to be more useful and safe for tree risk assessment.
Abstract
The mechanical wave propagation is affected by elastic properties differences and impedance differences on propagation medium. Deviations in the wave routes also affect propagation, and therefore velocity. The deteriorated wood will have its properties of strength and stiffness affected, therefore, causing variations in velocity of wave propagation when compared with sound wood. Hollows in the wood will cause changes in the wave path, which seek the material medium, also affecting the velocity by the change of course. To facilitate visualization, speed variations can be associated with color and, through interpolation software, images are constructed. This procedure, which can have different levels of sophistication, is called acoustic tomography, which the greatest limitation is the interpretation of the image produced. Considering the mentioned aspects, the objective of this research was to evaluate, qualitative and quantitatively, results of ultrasound tomography in face of different types and levels of deterioration. The tests were performed on logs of 6 tree species. Hollowed areas presented reduction of velocities greater than 70%, while zones deteriorated by fungi begin to be highlighted on the tomographic imagens with reductions of 30% in speed. In the case of cracks or galleries the image detailing depends on the relationship between the wavelength and the size of these defects. In general, the images of ultrasound tomography are adequate to provide important information for the tree risk assessment.
Introduction
Different types of decay can cause changes in the wood structure, affecting the wave propagation (Bucur 2006) and these changes can be used to detect the wood internal defects (Wang 2013, Brancherian et al. 2012, Wessels et al. 2011).
In case of fungi attack, the wood becomes less resistant and less rigid. Brazolin et al. (2014) demonstrated statistical differences between modulus of rupture and elasticity in bending in sound and in wood deteriorated by fungi. The reductions of strength and stiffness obtained by Brazolin et al. (2014) varied from about 70% in the wood with incipient deterioration to about 90% in wood severe decayed. The same result was obtained by Trevisan et al. (2007), with statistically significant strength differences in sound and wood decayed by fungus reached 42% in compression and in 52% in bending. Such reductions in strength and stiffness produce reductions in wave propagation velocity (USDA 2014, Deflorio et al., 2007, Ross et al. 1998).
In the case of wood decayed by termites or other xylophage’s insects, the wood presents galleries in its interior, but the surrounding material is generally sound. Weiler et al. (2013) showed statistically significant variations of modulus of elasticity (MOE) and rupture (MOR) when termite deterioration was between 20% and 100%.
The association of fungal attack, followed by termites may be responsible for the existence of hollowed areas in trees. In these cases (galleries and hollows), the wave propagation will suffer deviations, and consequent velocity reduction. Deviations occur because mechanical waves seek the material medium to propagate (Weiller et al. 2013, Secco et al. 2012, Najafi et al. 2009, Lin et al. 2008, Bucur 2006, Wang et al. 2007, Wang et al. 2004). Thus, if deteriorations or cavities in the wood reduce the velocity of wave propagation, the variations can be used as identifiers of changes in the material.
In order to obtain a scan in the inspected element, a measurement mesh (Divos and Szalai, 2002), called diffraction (Example in Fig. 1a) is used. Through this mesh it is possible to obtain measurement routes, whose quantity depends on the number of sensors used. To facilitate visualization, velocities ranges are associated with colors and, through interpolation software (Du et al., Feng et al., 2014, Zeng et al., 2013), images are constructed (Example in Fig. 1b). This procedure, which may have varying degrees of sophistication, is called acoustic tomography. The quality of the image is affected by the interpolation process used, but the accuracy of the results, also suffers interference from the power of the equipment and the quality of the data obtained in the field.
There are several commercial acoustic tomography equipment on the market, mainly using stress wave propagation (called sonic tomography), but there is still a lot of research in the area (Balázs & Divós 2015, Yamashita et al 2015, Trinca et al. 2015 a e b, Arciniegas et al. 2014, Trinca et al. 2013 a,b e c, Van Dijk et al. 2013, Turpening 2011, Secco et al. 2012, Sanabria et al. 2011, Gonçalves et al. 2011, Secco et al. 2011a e b, Secco et al. 2010, Kim et al. 2009, Batista et al. 2009, Brancheriau et al. 2008, Lin et al. 2008, Secco et al. 2004, Bucur 2002, Comino et al. 2000), proof of the need for a better understanding of the meaning of the generated image, as well as its reach in terms of accuracy. Pereira et al. (2007) summarizes this understanding by concluding that acoustic tomography is a developing technique and, therefore, lacking in studies. The same conclusion was presented by the USDA (2014), which discusses acoustic tomography in one of its chapters, presenting examples of tomographic images produced using commercial tomographs, but indicating that the greatest limitation of the method is still the interpretation of the tomographic image. Du et al (2015) also point out that the quality of reconstructed imagens by acoustic tomography can be improved.
Considering the aspects mentioned, the objective of this research was to evaluate, qualitative and quantitatively, the results of ultrasound tomography in face of different types and levels of deteriorations.
Methods
The wood samples used in this study were taken from trunks of trees of the species: Centrolobium sp., Tabebuia ochracea, Liquidambar styraciflua, Platanus sp., Poencianella pluviosa and Copaifera sp. In order to carry out the ultrasonic tests, discs with a minimum of 200 mm height were cut from those trunks.
The ultrasound tests were performed using the diffraction mesh (Fig. 1a) in which 8 measurement points were adopted for all species. The measuring points were approximately equidistant and positioned at the middle height of the disc. As used in the standing tree tests, at each measurement point, 3 mm holes were drilled for the introduction of the transducer tip, to ensure coupling of the transducer to the wood and not to the bark.
The tests performed on each diffraction mesh measurement route (Fig. 1a) were done using conventional ultrasound equipment developed by the research group in partnership with a technology-based company (USLab, Agricef, Brazil) and 45 kHz-frequency dry tips longitudinal transducers (Fig. 2). To obtain the tomography image we used software (ImageWood 2.0) also developed in the research group.
Fig. 2 Centrolobium sp. disc being tested by ultrasound
After the ultrasound tests, the discs were cut at the height where the measurements were taken and the surface was polished to be analyzed and photographed in detail. The species presented different conditions of deterioration and the decayed zones were recognized only macroscopically through a detailed visual analysis of its surfaces. The generated pictures were used to obtain representative mask of the deteriorated zone, using the free software ImageJ. This mask was used to calculate the percentage of deteriorated area.
The tomographic images were generated in two ways. In the first one, 6 velocity bands were used, based on the percentage of the maximum velocity obtained on discs (red up to 20%, orange 20% to 30%, yellow 30% to 50%, green 50% to 70%, blue 70% to 80% and black from 80% to 100%). According to a USDA publication (2014), a 50% reduction in velocity means that the region presents severe deterioration, so the second form of image generation was done considering only two colors: brown for zones with speeds above 50% of the maximum velocity obtained on discs and yellow for zones with speeds below 50% of maximum speed, considering as areas with severe deterioration. These two-color images were used to determine the percentage of impaired area inferred by tomography, also using the free ImageJ software. The images generated in different colors by the ultrasound tomography were visually compared with the pictures obtained from the surface. In addition to this visual analysis, the percentage of deteriorated area, inferred by tomographic images with two colors, was compared with the percentage of deteriorated area obtained from the mask created by the picture of the discs surface.
Results and Discussion
The visual analysis of the treated surfaces allowed detailing the deteriorated zones in the discs of the different species (Table 1)
Table 1. Summary description of the visual analysis of the deteriorations in the discs
Species Description
Centrolobium sp. Large hollow caused by termite attack
Tabebuia ochraceae Numerous galleries caused by Coleoptera attack Liquidambar
styraciflua Lateral crack from the pith to the bark caused by drying process
Platanus sp. Most of the wood shows signs of attack by fungi and there are some hollowed areas caused by termites Poencianella pluviosa Fungus attack in the center of the disc
Copaifera sp Large hollow caused by termites. In the surroundings of these hollows there is deterioration by fungi The images of Centrolobium sp. and Copaifera sp.discs (Fig.3a, 3f) are those were the red and orange colors appear, which represent zones with velocities lower than 30% of the maximum velocity (or velocity loss greater than 70%). These discs are the ones that have great hollowed regions. Thus, although the image does not represent the exact shape or size of the decayed region, it was efficient to assess the falling risk of these two logs. The images identified advanced deterioration inside the logs, which would be enough to indicate the risk involved. Differences in image precision can be obtained using different algorithmic to interpolate velocities in acoustic tomography (Du et al. 2015) but, in spite of these differences, none tomographic imagens constructed using interpolation methods proposed by Du et al. 2015, Feng et al. 2014, Zeng et al. 2013, showed precisely the shape or the dimension of the internal hollows. On the other hand, in the same way as discussed here, the representation was enough to indicate the degree and the extension of the decay. The use of more complete algorithmic or other data manipulation, as compensate radial/tangential velocities, can improve the correctly detection of healthy areas (Du et al. 2015), minimizing the overestimation of decayed areas mentioned by Wang et al. (2009) or the worse quality of the image near the sensor (Gilbert and Smiley 2004).
In the Liquidambar styraciflua disc (Fig. 3c) there´s also a small region with the orange color located at the edge, representing reduction of 70% to 80% of the velocity. This region corresponds to the zone of greater opening of the crack in the disc. The rest of the image is mostly black (velocity losses below 20%) with a zone with velocity loss between 30% and 50% (green), which is on the same side of the crack. The tomographic image analysis would not adequately identify the size and position of the defect (crack). The ability to detect defects using wave propagation is associated with the relationship between wavelength () and defect size (Bucur 2006). In general, the detection sensitivity is of /2 order (Bucur 2006). In the case of the Liquidambar styraciflua specimen, the maximum speed (integral area) was 2000 m/s, so 44 mm. Thus, defect detection from 22 mm is expected to be possible. The largest aperture of the crack, located at the edge of the disc and detected by the tomographic image (Fig. 3c) is 25 mm, consistent with the theoretical aspects of this test.
The image of the Tabebuia ochracea (Fig. 3b) specimen has only colors green and yellow. There are many yellow cords within the green zone. Since the cords are representative of interferences caused by the interpolation system used by the software (Trinca et al., 2015, Trinca et al., 2013 a, bc, Van Dijk et al., 2013, Secco et al. 2012, Secco et al. 2011a e b, Secco et al. 2010, Batista et al. 2009), the interpretation of this image would consider that, in the internal part of the piece, the loss of velocity would be between 30% and 50% which,
according to USDA (2014) would indicate strength loss about 50%. At the edges of the piece there are large areas in yellow, indicating speed losses between 50% and 70% that, according to USDA (2014), represent severe deterioration. The visual analysis shows that the piece is completely taken by cavities and at its extremities there is a more severe Coleoptera attack. Therefore, in this case, the diagnosis of the inspection would lead to consider the piece with considerable loss of resistance, although it is not possible to visualize the cavities in the image. The non-visualization of the cavities in the image was already expected since, in this case, the velocity in the sound wood is also about 2000 m / s ( 44 mm) and the cavities size are way below 22 mm (/2).
The image of Platanus sp. disc (Fig. 3d) indicates severe deterioration in practically the entire inspected area, as there is a large area in yellow (velocity loss between 50% and 70%) interspersed with green (velocity loss between 30% and 50%). This result was compatible with the real state of surface of the disc, detailed in visual analysis. The Poencianella pluviosa disc (Fig. 3e) also indicated, in visual analysis, a central region with fungal decay. However, the visual inspection show that the degree of deterioration was inferior to the one observed in the Platanus sp. (Fig. 3d) in which the wood was much more softened. This pattern was consistent with the images generated in these two discs, because in the case of the Poencianella pluviosa (Fig 3e) the velocity loss was up to 30% (blue and black colors) while in Platanus sp. (Fig. 3d) between 30% and 50%. It is much more difficult the identification of decayed zones filled (not hollowed), as the case of decayed by fungi, because the transmission velocity differences between sound and deteriorated wood is weaken (Du et al. 2015). In a sample with hollow filled with clay, Du et al. (2015) showed that their proposed interpolation method indicated, not exactly but more accurately than Feng et al. (2014) and Zeng et al. (2013), the position of the defect. The interpolation method proposed by Feng et al. (2014) showed the entire disc with red color (deteriorated) and so, no identification of the clay filled hollow. The method proposed by Zeng et al. (2013) indicated the location of the defect approximately, but, showed sound areas as decayed.
Fig. 3 Pictures of the surfaces of the discs from the different species and images of its ultrasonic tomography Legend: percentages of maximum speed: red up to 20%, orange 20% to 30%, yellow from 30% to 50%, green from 50% to 70%, blue from 70% to 80% and black from 80% to 100% %
For the Centrolobium sp. and Copaifera sp. discs, the percentage of decayed area (Table 2) obtained by the representative mask of zones with deterioration (Fig. 4) was approximately 30% lower than the decayed percentage area infered using the two colors tomographic image (Table 2). Wang et al. (2009) concluded that an internal defect in the tree trunk tends to be overestimated in it size using acoustic tomografy. Areas with hollows show loses in velocities superior to 70% (Fig. 3), but the two colors tomographic images adopted as reference the one proposed by the USDA (2014), corresponding to velocity loss around 50% (zones in yellow in Fig. 4). So, the area of this zone (yellow) also covered regions not affected by the hollow, explaining the differences between percentage of decayed areas (Table 2). Trinca et al. (2015) also obtained loses of velocity superior to 70% in holow zones.
On the other hand, in discs with zones deteriorated by fungi (Platanus sp. and Poencianella pluviosa) the tomographic image underestimated the real area deteriorated (Table 2). This can be explained because the
reduction of velocity is associated by the strength and stiffness loss which, in turn, depends on the level of the decay (Trinca et al., 2015) and the level of the decay was not considerated for mask creation. The mask (Figure 3) was constructed based only on size and not based on the deterioration level. The results of Trinca et al. (2015) indicate that for insipient decay in wood attacked by fungi the vlocity loss is only around 10%, reaching about 80% in wood with severe decay. As the Poencianella pluviosa disc show, in visual analysis, an insipient to moderate fungal decay, the velocity loss will be not represented in two colors tomography because it is less than 50%.
In the case of the Tabebuia ochracea specimen it was not possible to obtain a mask due to the difficulty in obtaining the sum of the areas of the numerous of small galleries, which were also superficial.
Table 2. Percentage of deteriorated area using the photograph of the discs surfaces and the image of ultrasound tomography and difference between these two percentages
Deteriorated area (%) Species Mask obtaine in
Surface picture
Tomographic image* Difference (%) Centrolobium sp. 48,09 78,34 -30,25 Copaifera sp 43,46 73,70 -30,24 Platanus sp. 66,60 53,64 12,96 Liquidambar styraciflua 1,83 2,29 -0,46 Tabebuia ochracea** - 32,14 - Poencianella pluviosa 24,87 0 24,87
* Areas of the image with more than 50% of velocity loss; ** It was not possible to calculate because it was composed of numerous small and superficial cavities
Applying the expected ultrasonic velocity variations to zones with hollows and with deterioration caused by fungi (Trinca et al., 2015), new two colors tomographic images were constructed (Fig. 4). For the discs with hollows (Centrolobium sp. and Copaifera sp), yellow represents areas with 70% velocity loss (Fig. 4a, 4f) and for disc deteriorated by fungi, the zone in yellow was considered to have 30% velocity loss (Fig. 4c, 4d, 4e). Of course, this discussion has no practical motivation, as it was considered from the knowledge of the actual condition of the discs.The purpose of this discussion was to analyse the ranges of velocity losses associated with hollows and zones decayed by fungi attack.
Using the same procedure to calculate deteriorated areas (mask and tomographic image), we can see that the hollowed areas (Centrolobium sp. and Copaifera sp.species) provided by the ultrasonic tomography are now closer to the actually affected areas (Table 3) than the first analysis (Table 2). In the case of the Platanus sp, the area deteriorated by fungi inferred by tomography, which was previously underestimated, surpassed the highlighted area by the mask in about 15%. In the Poencianella pluviosa specimen, whose deterioration by fungi seems to be in the initial phase, the use of 30% of speed loss begins to highlight some deterioration, but still inferior to that highlighted by the mask (Table 3).
Table 3 Percentage of deteriorated area using the photograph of the discs surfaces and the image of ultrasonic tomography, and difference between these two percentages
Deteriorated area (%)
Species Surface photograph Tomographic image Difference (%) Centrolobium sp. 48,09 38,74* 9,35 Copaifera sp 43,46 45,50* -2,04 Platanus sp. 66,60 82,15** -15,55 Tabebuia ochracea*** - 32,14 - Poencianella pluviosa 24,87 8,10** 16,77
* Areas of the image representing speed losses exceeding 70%; ** image areas representing speed losses of more than 30%; *** It was not possible to calculate because it was composed of innumerable small and superficial cavities
Fig. 4 Masks representative of the deteriorated zones based on the surface pictures (A1, B1, C1, D1 and E1) and ultrasonic tomography images considering two situations. Situation 1: yellow: velocity losses greater than 50% and brown: velocity losses less than 50% (A2, B2, C2, D2, E2). Situation 2: discs with hollow - yellow: velocity losses greater than 70% and brown: velocity losses of less than 70% (A3, E3) and discs with presence of zones deteriorated by fungi - yellow: velocity losses greater than 30% and brown: velocity losses of less than 30% (C3, D3)
The tomography image generated with different colors (Fig. 3) is a little more confusing for a layman, but conceptually allows better differentiation of velocity variation levels, and consequently the location of the zones with greatest loss of wood stiffness, regardless of the level of the decay.
Using the USDA recommendation (2014) to construct images with two colors (above and below 50% loss of velocity) the hollowed region shown is amplified (Fig. 2 and Table 2) and zones with deterioration by fungi will only be identified when this deterioration is not in its initial stage. However, despite having deficiencies to localize and to give the exact size of deteriorated areas, the ultrasonic tomography show to be an important tool for falling tree risk assessment purposes, since it allows inferring the existence of zones with loss of stiffness.
Conclusions
In this paper we evaluate, qualitative and quantitatively, the results of ultrasound tomography facing different decay condition in wood.
Considering as hollowed areas those with 70% of velocity loss, the percentage of decayed area predict by the tomography is closer to the actual condition. If 50% of velocity loss are considering as hollowed areas the image interpretation tends to overestimated the hollowed areas.
In zones with moderate to severe decayed by fungi the velocity loss is greater than 50% whereas velocity loss around 20% and 30% are associated with insipient to low decay.
The identification of cracks depends on the relationship between their size and the wave length, which in turn depends on the transducers frequency. In areas with great number of small and superficial holes (as in case of Coleoptera attack), although ultrasound tomography is not capable to identify the holes, the velocity loss (30% to 50%) in the whole disc is an adequate result to the actual situation, since despite the attack being superficial, it was numerous, so it is expected wood stiffness loss.
Ultrasound tomography images are adequate to provide important information for falling tree risk assessment.
Acknowledgments
The authors would like to thank the National Council for Scientific and Technological Development (CNPQ) for the scholarships and Sao Paulo Research Foundation (FAPESP) - Proc. 2015/05692-3 - for the research funding. They also thank the Environment Department of UNICAMP for donating the logs used in the tests.
Conflict of interest:
The authors declare that they have no conflict of interest
References
ARCINIEGAS, A.; PRIETO, F; BRANCHERIAU, L.; LASAYGUES, P. (2014) Literature review of acoustic and ultrasonic tomography in standing trees. Trees. 28(6): 1559-1567
BALAZS, M.; DIVOS, F., (2015) Glue laminated timber structure evaluation by acoustic tomography. In: International Nondestructive testing and evaluation of wood symposium, Anais…, v.19. p.462 – 466
BATISTA, F.; GONÇALVES, R.; CERRI, D.G.P.; SECCO, C.B., (2010) Reprodução da condição interna de peças de madeira através de imagens representativas da propagação de ondas, 06/2010, Científico Internacional, XXVIII Congresso Nacional de Ensaios Não Destrutivos e Inspeção - 14a. Conferencia Internacional sobre Evaluacion de Integridad y Extension de Vida de Equipos Industriales, Vol. 1, pp.1-12, Santos, SP, BRASIL
BATISTA, F.A.F; GONÇALVES, R.; CERRI, D.G.P.; SECCO, C.B. (2009) Reprodução da condição interna de peças de madeira através de imagens representativas da propagação de ondas. Madeira: Arquitetura e
Engenharia, v.10 (25), p 23-32
BRANCHERIAU L., LASAYGUES P., DEBIEU E., LEFEBVRE JP. (2008) "Ultrasonic tomography of green wood using a non-parametric imaging algorithm with reflected waves".Annals of Forest Science, 65(7):712-718
BRAZOLIN, S.; TOMAZELLO FILHO, M.; YOJO, T.; NETO, M.A.O.; ALBUQUERQUE, A.R.; JUNIOR, C.R.S; (2014) Propriedades físico-mecânicas do lenho deteriorado por fungos apodrecedores de árvores de tipuana tipu, Cerne, Lavras, v. 20, n. 2, p. 183-190
BUCUR, V. (2002) High resolution Imaging of Wood Structure. In: International Symposium on Nondestructive Testing of Wood. Berkeley, California
BUCUR, V. (2005) Ultrasonic techniques for nondestructive testing of standing trees. Ultrasonics. 43:237-239
BUCUR, V. (2006) Acoustics of wood. 2nd ed. New York: Springer-Verlag, 393p
COMINO E, MARTINIS R, NICOLOTTI G (2000) Low current tomography for tree stability assessment, 278, In: Backhaus G F, H Balder, and E ldczak (Eds). International symposium on plant health in urban horticulture, Braunschweig, Germany, 22-25
DEFLORIO, G; FINK, S; SCHWARZE, F.W.M.R.; (2007) Detection of incipient decay in tree stems with sonic tomography after wounding and fungal inoculation. Wood Sci Technol. 42:117-132.
DIVOS, F.; SZALAI, L. (2002) Tree evaluation by acoustic tomography. In: Proceedings of the 13th International symposium on nondestructive testing of wood; 2002 August 19. 21; Berkeley, CA. Madison, WI: Forest Products Society: 251.256
DU, X. et al. (2015) Stress Wave Tomography of Wood Internal Defects using Ellipse-Based Spatial Interpolation and Velocity Compensation. BioResources, v. 10, n. 3, p. 3948-3962
FENG, H. et al. (2014) Tomographic image reconstruction using an interpolation method for tree decay detection. BioResources, v. 9, n. 2, p. 3248-3263
GILBERT E, SMILEY E (2004). Picus sonic tomography for the quantification of decay in white oak (Quercus alba) and hickory (Carya sp.). Journal of the Arboriculture 30(5), 277-281.
