Structural Health Monitoring Techniques:The impact-echo method is a technique for flaw detection in concrete. It is based on monitoringthe surface motion resulting from a short-duration mechanical impact. One of the key features of the method is the transformation of the recorded time domain waveform of the surface motion into the frequency domain. The impact gives rise to modes of vibration and the frequency of these modes is related to the geometry of the test object and the presence of flaws mechanical impact.

Matlab MethodThe data from sensors are collected in the form of electrical signals in most cases. The signals are correlated with physical sources. However, most sensor signals are contaminated with noises, such as those from unwanted physical sources and electrical noises. The noise components should be cleaned, and accurate data should be obtained during cleansing curation. The system identification process estimates system models that co-relate input and output signals. From the identified models, features that have strong correlation with damage and deterioration are extracted to be used in diagnosis and prognosis. In the last process, pattern recognition tools such as support vector machines (SVM) and neural networks are frequently used.2. The data obtained at a building are automatically transmitted to the server through the internet when the monitored responses are requested to be transferred. The conditions to trigger the transfer and the other configurations of the system can be set in advance in the form of a database for each sensor. Thus, the users of this system do not need to configure the sensors for each building. Only when it is necessary, the set up conditions can be changed through the server using an easy-to-use user interface. As the database for the server, a relational database, Postgre SQL, was adopted. The user interfaces were coded using Hyper-text Pre-processor (php) to make the system flexible. The data needed for analyses are selected using a php interface and submitted for analyses by MATALB web servers.

Vibrating Wire Strain Gauge:Strain Gauges consist of a tensioned steel wire anchored at both ends into flanges.The wire is enclosed in a stainless steel tube. The internal parts of all MGSGeosense strain gauges are essentially identical, only the body geometry and theinclusion of additional springs change within the units with longer gauge lengths. Theconfiguration of the sensing elements may also vary slightly from model to model.Electromagnetic coils are located within the body close to the axis of the wire. Whena brief voltage excitation, or swept frequency excitation is applied to the coils, amagnetic field is induced causing the wire to oscillate at its resonant frequency. Thewire continues to oscillate for a short period through the field of the permanentmagnet, thus generating an alternating current (sinusoidal) output. The frequency ofthis current output is detected and processed by a vibrating wire readout unit, or by adata logger equipped with a vibrating wire interface, where it can be converted intoEngineering units of Strain.Forces within the structural element onto, or in, which the gauge is fixed, cause thelength of the gauge to change. This causes a change in the tension of the wire withinthe gauge. It is the tension in the wire that produces the value that can be convertedto strain.A change in length of the wire changes the tension of the wire which results in achange in resonant frequency of oscillation of the wire. The change in the square of frequency of oscillation is directly proportional to the change in strain in the structuralelement.Embedment Strain Gauges in Wall and SlabsEmbedment gauges in wall panels and floor slabs are usually installed in pairs. Onegauge would be installed close to one face and the other gauge close to the other face.This enables the detection of bending of the elements together with any axial strainchanges.

Figure a shows a typical arrangement ofembedment strain gauges for measuring the loadand bending in a concrete beam. Two gaugesinstalled in the compression zone and two in thetension zone. ( The red dots indicating thegauges positions ).

Fig. a

Figure b shows a typical arrangement forembedment strain gauges within a column.

Fig. b

Figure c shows a horizontal section through aconcrete wall in which strain gauges arepositioned to measure bending of the wall.Concrete elements that include strain gauges canbe either pre-cast or cast in-situ. Fig. c

Data Determination And Calculation:The tension of a sensor wire can be measured by detecting the frequency (note) atwhich it naturally vibrates. The following is a description of the units commonly usedby the instrumentation industry.Frequency Units ( Hz ). If the wire is excited electronically the frequency at which itvibrates can be measured. The units used to express frequency are Hertz (Hz) orKiloHertz (kHz).The disadvantage of these units is that there is no linear conversion from change inHertz to change in wire tension.Linear Digits ( B ). In order to overcome the problem of a linear conversiondescribed above, the frequency value can be squared, thereby rendering it linear, butquite large. To reduce its size, it is often divided by 1000 (or multiplied by 103). Theexpression Hz2/1000 (or Hz2 x 103) is the most commonly adopted as a linear digitaloutput.Period Units ( P ). Some readout devices utilise the counter function available inmany common integrated circuits.Period Units represent the time taken for the wire to vibrate over one full oscillation,expressed in seconds. Due to the very small size of the number generated mostequipment manufacturers display the unit multiplied by either 1000 ( 103 ) or10000000 ( 107).The relationship between Period units and Frequency units is expressed asP = 1FrequencyPeriod units are convenient to measure but do not have a linear relationship to thechange in wire tension.Calibration Factor. Each instrument is supplied with a Calibration Factor ( orsometimes called Gauge Factor ), to enable conversion from the raw data ( in theunits described above ) into engineering units such as Micro-strain.The value of the calibration factor will vary depending upon the engineering units intowhich the raw data is to be converted.Some instruments have "Generic" calibration factors and others are calibrated to generate an individual calibration factor.Readings from VW Strain Gauges are typically in a form that is a function offrequency ( as above ) rather than in units of strain.To convert the readings to units of strain, calibration factor(s) must be applied to therecorded values. For Vibrating Wire Strain Gauges the calibration factors are uniqueto each model. A certificate of conformance, detailing the applicable calibrationfactor will be supplied with all Geosense Vibrating Wire Strain Gauges. In addition, aBatch factor may be included on the certificate. This factor accommodates any slightvariations in the model specific calibration factor and must also be applied in thecalculation.If a readout display is in Period units ( e.g. 0.03612 or 3612 depending upon thereadout used ) a calculation must first be performed to convert the reading fromPeriod units to Linear Digits ( Hz2/1000 ) units.Two examples of this can be seen below. The first (1) where readout includes adecimal point and displays the Period in Seconds2 and the second (2) where thereadout displays the Period in Seconds-7(1) Linear Hz2/1000 = ( 1 / 0.03612 x 102 ) 2 / 1000= 7664.8(2) Linear Hz2/1000 = ( 1 / 3612 x 10 7 ) 2 / 1000= 7664.8If the readout displays Frequency values, ( e.g. 2768.5 Hz ) only a simple calculationis required to convert the reading to Linear Digits.Linear Digits ( Hz2/1000 ) = ( 2768.5 ) 2 / 1000= 7664.6Certain data loggers store their Vibrating Wire data in Linear Digits but divided by afurther 1000. Obviously these data would have to be multiplied by another 1000 tomaintain the standard data format for the conversion to engineering units.There are many ways to achieve the conversion from recorded data to usefulengineering values. The following is included as a guide only and as a basis foralternative approaches.Linear CalculationThis is the most reliable and straight forward calculation to convert raw data toengineering units. It can be easily carried out using a simple calculator. It assumesthat the reading is in ( or has been converted to ) Linear Digits ( Hz2/1000 ). For mostapplications this equation is perfectly adequate and is carried out as followsConverting from a frequency reading in Hz to microstrainUse the following equation to convert a reading in Hz to microstrain (): = ( F2 / 1000 ) x Gauge-Factor x Batch- FactorWhere: = microstrainF is a reading in Hz.Gauge-FactorBatch-Factor ( if applicable )Calculating a change () in microstrainChange in strain is calculated by subtracting the initial strain from the current strain:Change () = current - initialor = ( F2 /1000 current - F2 /1000 initial) x Gauge-Factor x Batch- FactorWhere: = microstrain = change in microstrainF is a reading in Hz.Gauge-FactorBatch-Factor ( if applicable )To check whether the gauge is in Tension or Compression:Positive indicates tensile strain. Negative indicates compressive strain.