Ultrasonic Signal Evaluation Used to Detect and Localize Temperature Changes in a Concrete Specimen Caused by a Heating Cartridge

Transcription

1 September 5-7, 5, Berlin, Germany More Info at Open Access Database Ultrasonic Signal Evaluation Used to Detect and Localize Temperature Changes in a Concrete Specimen Caused by a eating Cartridge Sven GROTE, Chinaemerem KANU BAM Federal Institute of Materials Research and Testing Division 8. Non-destructive damage analysis and environmental measurement methods Berlin, Germany, Phone: , sven.grothe@bam.de Colorado School of Mines, Golden, USA, ckanu@mymail.mines.edu Abstract Ultrasonic measurement evaluation methods have been proven to be effective for detection of subtle changes, caused by temperature, load or moisture. To detect and localize temperature changes, a concrete block of 5.8 m, including a heating cartridge and multiple temperature sensors, has been set up to change the temperature and monitor the temperature distribution in a certain area inside the specimen. An ultrasonic monitoring system with ultrasonic sensors ( transmitters, receivers, 5 kz central frequency) has been implemented on the specimen. Data from sensor combinations was collected over the whole period of the experiment in an interval of minutes. Quantitative methods (Coda Wave Interferometry and Time of Flight method) were used to evaluate the changes in ultrasonic travel-time caused by the heating period, when the cartridge was active, and the cooling period after turning off the cartridge. Furthermore the travel-time changes from all sensor combinations were used to locate the heating cartridge. Keywords: Ultrasound, Monitoring, Concrete, Time of Flight Method, Coda Wave Interferometry, Temperature. Introduction Detection of subtle changes in concrete structures is an important subject to evaluate damages over time (e.g. freeze-thaw attacks, alcali silica reaction). By means of ultrasonic signals it is possible to detect not only the direct signal path, but also to observe the area around, due to signal reflections in the concrete structure. This makes it possible to monitor structures in larger areas. Changes caused by temperature changes have been analysed already in laboratory environments (e.g. in []). Therefore the whole specimen was heated up and cooled down in a climatic chamber. Furthermore in [] a specimen was compressed locally at a certain point to investigate the changes caused by the load cell and the area of influence from the load cell. To combine these results an experiment with a heating cartridge inside a specimen has been set up to cause temperature changes in a specimen locally and observe them with an ultrasonic monitoring system. The goal is to observe the changes and the area of influence caused by the heating cartridge and approximately pinpoint it with the help of the gathered ultrasonic data.. Experiment The used specimen is a 5.8 m concrete block that has been cast at BAM Test Site Technical Safety for various experiments with ultrasonic sensors outside of laboratory environments. A total of ultrasonic sensors have been attached around the specimen ( sensors on each side of the specimen: north and west side contain transmitting sensors, south and east side contain receiving sensors). One multiplexer connects all transmitting sensors to a rectangular transmitter (5 kz), the second multiplexer connects all the receiving sensors to an analog to digital converter for storage of the collected data on the PC. Additionally the receiving signals were preamplified by factor 6 to ensure a good signal quality on the high distance between transmitting and receiving sensors. The measurement setup is shown schematically in Figure. This setup makes it possible to track transmitter-receiver combinations within only a few minutes, regarding to the number of repetitions. Data was collected in minutes intervals during the whole experiment. The heating cartridge was set at 5 C for exactly hours. A temperature sensor in cm distance from the heating cartridge detected a temperature increase from C to7 C during the heating period. After turning off the heating cartridge, data was still collected for a few more days in the same interval to evaluate data from the cooling period.

2 September 5-7, 5, Berlin, Germany MUX Transmitter Wavelet TTL Trigger PC Specimen Preamplifier (x6) MUX Receiver Storage A D Figure : Measurement setup (just half of the sensors is shown for clarity) S S S S S R.5 S R9 R8 y (m) S.5 R7 R6 R5.5 R R R.5 S R R R R R R5 R6 R7 R8 R9 R 5 6 Figure : Arrangements of sensors and the heating cartridge. Theory To evaluate the velocity changes of ultrasonic signals two methods were used. Both methods used a reference time series and compared the velocity changes against the reference.. Time of Flight Method The Time of Flight method describes the time that the pressure wave needs to travel from the transmitting sensor to the receiving sensor through the medium. Therefore the velocity of a reference time series v Ref is calculated directly by knowing the distance s between the two sensors and the elapsed time t Ref between sending of the trigger pulse and the first impact in the receiving signal through: v Ref = s t Ref ()

3 September 5-7, 5, Berlin, Germany By knowing the velocity of the reference signal it is possible to compare the relative velocity change v of a recorded time series towards the reference by using the time lag t between the first impact of the reference and the first impact of the recorded time series. v = t v Ref t Ref () The disadvantage of this technique is that the limit of detection is restricted to the resolution of the receiving signals and the distance between the sensors. A sampling frequenzy of Mz and a sensor distance of 5 m is limited to detect velocity changes not smaller than approximately.8 % (in concrete).. Coda Wave Interferometry For the detection of very small velocity changes Coda Wave Interferometry (CWI) is a suitable technique. Small velocity changes can cause a phase shift in the later stage of a time series (Coda). The CWI expands the cross-correlation, which is used two compare two time series x(t),y(t), by a small factor ν that compresses or stretches the reference time series to maximise the cross-correlation. This negates the phase lag and the relative velocity change can be calculated out of this. This technique was described in []. CC(ν) := ν = t t x(t) y(t ( ν)) dt x(t) dt y(t) dt = v v = max () The disadvantage of this technique is, that larger velocity changes lead to larger phase shifts. This can cause multiple local maximums in the cross-correlation, so the time factor cannot be allocated clearly anymore. But on the other hand the CWI can detect very small velocity changes up to approximately %.. Imaging Algorithm The localized velocity change induced by the heating cartridge got imaged by solving for the velocity change in the following inverse problem [5]: K T d = [K T K +βi] m () where K is a discretized version of the sensitivity kernel of the travel-time change, due to the localized change in the concrete block, from the time-lapse coda waves, β is a regularization parameter and m defines the inverted change in the scattering medium which is the change in localized velocity in the concrete block. Equation is based on [6], showing a relationship between a localized fractional velocity change δv/v (x ) at x and the estimated travel-time changes τ(t) induced by the velocity change: τ(t) = K(s,x,r,t) δv v (x ) dv(x ) (5) V where K(s,x,r,t) is the sensitivity of the travel-time change to the localized velocity change in the scattering medium.

4 September 5-7, 5, Berlin, Germany. Results. Influence of the heating cartridge Figure shows the influence of the heating cartridge towards the ultrasonic velocity. Both methods show similar results. The heating period is slowing down the ultrasonic velocity by approximately %. After turning off the heating cartridge the ultrasonic velocity starts to increase immediately until the original value is reached after one week. This follows approximately the expected exponential loss of temperature after turning off the heating cartridge until the temperature inside the specimen is in thermal equilibrium with the surrounding environment. The results are overlaid by temperature day and night cycles mentioned in [], because the specimen is located in an outdoor field and exposed to weather conditions. Delta v/v velocity changes caused by local temperature changes from a heating cardridge velocity change pressure wave velocity change CWI date Temperature in C Outdoor temperature during experiment Temperature date Figure : velocity changes caused by local temperature changes from a heating cartridge evaluated with Time of Flight Method (red) and Coda Wave Interferometry (blue) (left image) and outdoor temperature during the experiment recorded 5 m away from the specimen (right image). Localization To localize the heating cartridge with the help of the ultrasonic velocity all available transmitter-receiver-combinations were used. Figure shows the estimated relative temperature distribution on the left side and the evaluated ultrasonic velocity from the measured data on the right side. The data shows an ultrasonic velocity drop in the area around the heating cartridge. The velocity change in the image is slightly shifted from the heating cartridge to the side face of the specimen. This could be affiliated to the sensor arrangements, that are also S S.5 S S S S R R9 R8.8 S S S S S S R R9 R8.5 y (m) S R7 R6 R5 R R R.6.. z (m) S R7 R6 R5 R R R.5 S R R R R R R5 R6 R7 R8 R9 R 5 S R R R R R R5 R6 R7 R8 R9 R 5 Figure : estimated relative temperature distribution inside the specimen at the end of the heating period (left image) and evaluated ultrasonic velocity change (right image)

5 September 5-7, 5, Berlin, Germany only located on the side face of the specimen, causing a larger velocity change on the nearest transmitting sensor. 5. Conclusion The evaluated data shows that the heating cartridge has a significant impact on the ultrasonic signals. A two hour heating period, which changed the temperature from C to 7 C in cm distance caused velocity changes up to %. Both evaluation methods provided similar results that can be compared to temperature experiments obtained in laboratory in [], showing ultrasonic velocity decreases on temperature increases and vice versa. The imaging algorithm was able to locate the heating cartridge and its area of influence almost correctly. References. E Niederleithinger and C Wunderlich, Influence of small temperature changes on the velocity of ultrasound in concrete, Proceedings of QNDE, American Institute of Physics, pp 9 97,. E Niederleithinger, C Sens-Schönfelder, S Grothe and Wiggenhauser, Coda Wave Interferometry used to localize compressional load effects in a concrete specimen, Proceedings of EWSM, Inria, pp 7, July. R Snieder, The theory of Coda Wave Interferometry, Pure applied geophysics, vol 6, Birkhäuser, pp 55 7, 6. S Grothe, Ultrasonic signal evaluation used to analyse changes in a concrete specimen caused by weather-related temperature changes, same volume, 5 5. C Kanu and R Snieder, Time-lapse imaging of a localized weak change with multiply scattered waves using numerical-based sensitivity kernel, Journal of Geophysical Research: Solid Earth, 5 6. C Pacheco and R Snieder, Time-lapse travel time change of multiply scattered acoustic waves, The Journal of the Acoustical Society of America, vol 8(), pp, 5