Bundesanstalt für Materialforschung und -prüfung

International Symposium (NDT-CE 2003)

Non-Destructive Testing in Civil Engineering 2003
Start > Contributions >Posters > Ultrasonic: Print

Use of low Frequency Ultrasound Echo Technique to Determine Cavities in wooden Construction Composites

Andreas Hasenstab, Martin Krause
Federal Institute for Materials Research (BAM), Berlin, Germany
Carsten Rieck, Bernd Hillemeier
Technical University of Berlin, Germany

1 Introduction

The life cycle of wood can be very long if it is protected against damaging environmental influences. A special problem is heart rot, which is difficult to recognize and can cause the sudden collapse of constructions. In Fig. 1 and Fig. 2 the examples of a damaged roof are shown.

Fig. 1, lateral view of a beam damaged by fungal Fig. 2 bottom view of the beam of Fig. 1 with decay

There are several minor destructive and non-destructive techniques to detect the seventy of the damage. Common minor destructive techniques are drilling resistance, trial bearing and the measuring of the pullout resistance (BAM 2000, Eckstein 1994, Görlacher 1990, Wenzel 1999). With these methods very exact information about the damage at the location of the measurement can be obtained. The disadvantage is that an already damaged construction will be even further damaged by the measurements. A large construction also requires numerous measurements that take time. Occasionally it is also difficult to assess the damage to the entire construction on the basis of data obtained from a few holes with a diameter of 3 mm.

Non-destructive testing methods to detect hidden cavities in wood are high-level radiation, usually X-rays, and ultrasound (BAM 2000, Combe 1997, Habermehl 1992, Hasenstab 2002). Being non-destructive, these methods permit unlimited numbers of tests. Some of the methods work fast, allowing testing of large structures in a short time. Most of the non-destructive methods are based on the principle of transmission, requiring that both sides of the specimen are accessible. In contrast to this, ultrasound echo technique only needs access to one side. For a long time, low frequency ultrasonic echo technique (80 to 200 kHz) has already been used for non-destructive testing of concrete, e.g. to measure thickness or to localize compaction faults or tendon ducts (Krause 1999, Schickert 1999, Wollbold 1999, Krause 2001). Fundamental experiments with ultrasonic echo technique applied to wooden test specimens have been carried out at the Federal Institute for Materials Research and Testing (BAM) at division IV.4 in cooperation with the Department of Construction Materials of the Technical University of Berlin. The tests aimed to find back walls or hidden cavities.

2 Principle of Ultrasound Echo Technique

Ultrasonic technique uses a transmitter that sends the sonic waves and a receiver that measures the sound. Some transducers require a coupling media like Vaseline, ultrasonic gel or glycerine. Theoretically, sharp echo will result from high frequencies. However, attenuation rises with frequency, permitting only low frequencies. The transmission technique evaluates the time of flight (TOF) of the transmitted pulse in the range of 100 kHz through the specimen. It is however difficult to locate the position of damage within the specimen, as the TOF resulting from damage close to the surface can identical to the TOF resulting from deeper damage (Bekhta 2002). In wood it can also be difficult to distinguish between one large knot or a cluster of small ones. The echo technique allows localizing directly a reflector, like a back wall or an in homogeneity. After many tests it can be assumed that a clear echo from the back wall shows that the specimen is free of defects. An echo with a short travel time can be the result of a defect. If the velocity and dimensions of the specimen are known, the defect can be located. To obtain a good echo, broad beam transducers with high attenuation in the transducer and short signals are necessary. Ultrasonic echo results are often represented as A-scans or B-scans. An A-scan (Fig. 3) shows the intensity of a testing pulse along the time of flight. A B-scan (Fig. 4) is a composite of a number of A-scans along the line of measurement. It is a two-dimensional cross section through the specimen and shows the time of flight (TOF) along the line of measurement. If the TOF is multiplied with the velocity the result is ways of flight of the sound wave. So a B-scan can show directly the position of in homogeneities.

Fig. 3. A-scan (intensity along the time) of a back wall measurement Fig. 4. B-scan of the back wall of an un destroyed specimen Fig. 5 radial and tangen-tial cut of a four year old pine (Scholz 1999)

3 Characteristics of wood

Fig. 5 shows that wood is a an isotropic material. One can distinguish between three different directions - in fibre direction (axial), perpendicular to fibre (radial) and tangential to fibre. In addition to the dependence of the acoustic velocity to fibre direction, attenuation is high due to the high amount of air in wood, requiring powerful low frequency transmitters.

4 Experiments

Measuring procedure The experiments were carried out with longitudinal waves and transversal waves. The longitudinal waves were optimised at 100 kHz, the transversal waves at 55 kHz. For the longitudinal waves transducers (D=45 mm) were used in impulse echo technique with separate transducers (Fig. 6), coupled by Vaseline or ultrasonic gel. The transversal transducers were also used with separate transducers but did not require any coupling media. Fig. 7 shows the aperture of the measurement. The function generator sends the signal to a preamplifier, which sends the amplified signal to the transducer. The received signal was also amplified and then saved at a PC. An Oscilloscope was used to convert display both signals as an A-scan.

Fig. 6 adjustment of the probes on a specimen; sound wave between the probes perpendicular to fibre Fig. 7 aperture with oscilloscope, functions generator, amplifier, PC

Specimens The dimension of the specimens and the way of damage were designed according to previous publications (Erler 1993, Kothe 1998) and discussions with specialists of wood protection (Panzer 2001). After testing a large pine specimen was cut into three smaller specimens. One of these is specimen P2 (Fig. 8). To imitate heart rot, cores were inserted in direction of fibre.

Fig. 8: Bottom view of pine specimen P2 (36cm x 20cm x 9,5cm) with a single core (D=30mm) and a group of small cores (D=10mm) to imitate heart rot

5 Results

Tests with longitudinal waves result that the velocity rises from tangential to radial up to axial to fibre (Fig. 9). This well-known result (Niemz 1999) doesn't confirm with transversal waves. There the measured velocities vary in one area. In some measurements the velocities of all directions were very similar to each other.

Fig. 9 Velocity of longitudinal waves in dependence of direction of fibre Fig. 10 Pine specimen P2 with measuring lines 1 to 5 and cores imitate damages

The following ultrasound echo measurements were carried out on specimen P2 along the measuring lines as shown in Fig. 10. Fig. 4 shows a B-scan of a specimen with a very clear echo of the back wall showing that the travel time is homogeneous. There are no other echoes. A visual check after cutting of the specimen in smaller ones confirmed that there were no in homogeneities inside. The shorter travel times at the end of the specimen could have been caused by the higher velocity of waves in dry wood (Kabir 1998) and special effects caused by the close edge of the specimen.

The B-scan shown in Fig. 11 is the result of the measurements along line 1 (Fig. 10) with the transducers coupled perpendicular to fibre (Fig. 6). There are clear echo with 153µs and 71µs. To make the evaluation easier, a recalculated travel time was added in the picture.

Another B-scan along line is shown in Fig. 12. Here, the transducers are coupled parallel to fibre. Measured and the recalculated TOF were again very similar. The difference between Fig. 11 and Fig. 12 is a result of the different coupling of the transducers. The known effect of early echo by coupling parallel to fibre is caused by surface waves. This effect causes problems in the A-scan only - in the B-scan it can be recognized easily. Furthermore, the angle of reflection of the beam, the dimension of the transducers and the specimen and the dispersion influence the results.

Fig. 11 specimen P2, measuring along line 1, ducts coupled like in Fig. 6, way of sound perpendicular to fibre, theoretical TOFs added in red, measuring point every 2cm. Fig. 12 same measurement as in Fig. 11, with different way of the sound waves (parallel to fibre), Fig. 13 echo of back wall and ducts, measurements every 1cm

Similar to the other pictures, the red bar represents the recalculated TOF to make a comparison easier. Fig. 13 shows the result of a measurement along line 5 perpendicular to fibre. Small cores (D=30mm) in direction of fibre are visible.

First tests with transversal waves were performed on a pine specimen without damage. The result of this test is shown in Fig. 14 as a B- scan. The specimen did not require a coupling medium.

Fig. 14 B-scan of a measurement with transversal waves (pine specimen d=9,9cm) c Trans= 1430m/s

6 Conclusion

The results of the measurements with ultrasonic echo are very promising. It is possible to detect the back walls of the specimens as well as inserted cavities inside the structures. These results show the possibility of testing at wooden specimen. More detailed investigation are planed on damaged wooden construction due to heart rot. The relative results can be converted into absolute values by measuring the dimension of the specimen at the end of the beam or by drill resistance.

So it is possible to check the thickness of a beam even though the back wall is not accessible. It is therefore possible to determine the dimensions and inner structure of large construction. Detailed tests are only necessary in areas with no or an unexpectedly early echo from the back wall.

The ultrasonic echo technique can be utilized in the preservation of large structures, e.g. monuments and historic buildings (frameworks, churches,...). The early detection and containment of defects makes it possible to precisely locate hidden damage that would normally only be discovered during restoration work. The possibilities of discovering hidden damage before commencement of work permits a more precise estimation of costs, and therefore a more economic allocation of funds.

7 Acknowledgements

Special thanks go to Mr. Mielentz and Mr. Mielmann for the support with the experiment and Mrs. Streicher and Mr. Schaurich for the help with the graphics. Mr. Untergutsch assisted with the translation.


  1. Bekhta P., Niemz P., Kucera L. J.: Untersuchung einiger Einflussfaktoren auf die Schallausbreitung in Holzwerkstoffen. Holz als Roh- und Werkstoff Jh. 60 (2002), S. 41-45 Bundesanstalt für Materialforschung und -prüfung: ZfPBau-Kompendium. url www.bam.de/zfp-kompendium.html (1999)
  2. Combe J: Bestimmung der Holzqualität mit Ultraschall. AFZ/Der Wald, Band 52 (1997) Heft 3, S. 135-137
  3. Eckstein D., Saß U.: Bohrwiderstandsmessung an Laubbäumen und ihre holzanatomische Interpretation. Holz als Roh- und Werkstoff Jh. 52 (1994), S. 279 - 286
  4. Erler K.: Alte Holzbauwerke beurteilen und Sanieren. Berlin: Verlag für Bauwesen Berlin, München, (1993)
  5. Görlacher R., Hättrich R.: Die Bohrwiderstandsmessung. Bauen mit Holz (1990), H. 6, S. 455-459
  6. Habermehl A., Ridder H.-W.: Methodek der Computer-Tomographie zur zerstörungsfreien Untersuchung des Holzkörpers von stehenden Bäumen. Holz als Roh- und Werkstoff 50 (1992), S. 465-474
  7. Hasenstab A.: Die Hohlstellenortung bei Holzbauteilen mit dem Ultraschallverfahren, Diplomarbeit an der Technischen Universität Berlin, unveröffentlicht, (2002)
  8. Kabir M.F., Daud W.M., Khalid K. Sidek H. A. A: Dialactric and ultrasonic properties of rubberwood. Effect of moisture content, gain direction and frequency. Holz als Roh- und Werkstoff 56 (1998) S. 223-227
  9. Kothe E.: Auswirkung von Holzschäden durch Pilze und Insekten auf die Standsicherheit von Holzbauwerken - eine Bestandsaufnahme. Bautechnik 75 (1998), H 8, S. 552-558
  10. Krautkrämer J. und H.: Werkstoffprüfung mit Ultraschall, Berlin, 5. Auflage S.198 , Springer Verlag (1986)
  11. Krause M., Mielentz, Milman, Wiggenhauser, Müller: Spannkanal-untersuchung mit bildgebenden Ultraschallecho-Verfahren. Deutsche Gesellschaft für zerstörungsfreie Prüfung, Berichtsband 66-CD: Plakat 2 (1999)
  12. Krause M., Mielentz, Milman, Wiggenhauser, Müller, Schmitz: Ultrasonic image of concrete members using an array system. NDT & E International 34 (6) (2001) pp. 403-408
  13. Niemz P., Kucera L.J.: Untersuchung zum Einfluss des Faserwinkels auf die Ausbreitungsgeschwindigkeit von Schallwellen in Holz. Holz als Roh- und Werkstoff 57 (1999) S. 225
  14. Panzer T., Sachkundiger für bekämpfenden Holzschutz. Persönliche Mitteilung, (2001)
  15. Schickert, M.: Ein empirisches Modell der Dispersion von Ultraschall in Beton in: DGZfP-Jahrestagung 10.-12.05.1999 in Celle, DGZfP-Berichtsband BB 68, Berlin (1999)
  16. Scholz Wilhelm: Baustoffkenntnisse. Düsseldorf. Werner Verlag (1999)
  17. Wenzel F., Kleinmanns J.: Sonderforschungsbereich 315, Historische Holztragwerke, Untersuchungen, Berechnungen und Instandsetzen. Universität Karlsruhe (1999)
  18. Wollbold, F., Neisecke, J., Hillger, W., Kapphahn, G., Krause, M., Kroggel, O., Müller, W., Roye, W. und M. Schickert: DGZfP-Merkblatt B4 in aktualisierter Form: Merkblatt für Ultraschallverfahren zur Zerstörungsfreien Prüfung mineralischer Baustoffe und Bauteile in: DGZfP (Hrsg.); Fachtagung Bauwerksdiagnose - Praktische Anwendungen Zerstörungsfreier Prüfungen, 21.-22. Januar 1999 in München, DGZfP-Berichtsband 66-CD (1999) Poster 26
STARTPublisher: DGfZPPrograming: NDT.net