BS 1881 : Part 203 : 1986UDC 666.972.017 : 691.32 : 620.1

British Standard

Testing concretePart 203. Recommendations for measurement ofvelocity of ultrasonic pulses in concrete

Essais du b&onPartie 203. Recommandations pour la mesure de la vitesse de dbplacementdes impulsions ultrasonores dans le b&on

Prtifverfahren von BetonTeil 203. Empfehlungen zur Messung der Geschwindigkeit von Ultraschallimpulsenin Beton

NO COPYING WITHOUT BSI PERMISSION EXCEPT AS PERMITTED BY COPYRIGHT LAW

BS 1881:Part203: 1986

Foreword d

This Part of BS 1881 has been prepared under the directionof the Cement, Gypsum, Aggregates and Quarry ProductsStandards Committee. It supersedes BS 4408 : Part 5 : 1974,which is withdrawn. All aspects of testing concrete arebeing included as Parts of BS 1881, from sampling freshconcrete to assessing concrete in structures. Part 201 givesgeneral guidance on the choice of nondestructive testmethods and should be consulted for advice on methodswhich can be used to complement the measurement ofultrasonic pulse velocity.

Ultrasonic testing of concrete has been in use in a largenumber of countries for many years, in laboratories, inprecast works and on site.

The technique generally used is the measurement ofultrasonic wave group velocity from the transit time of apulse between separate transmitting and receivingtransducers through a known path distance in the concrete.This provides a measurement of the mean ratio of elasticstiffness to density along the path and has been found to bea useful index of concrete quality.

The velocity is a function of the composition, degree ofcompaction, maturity and free water content which areinherent in concrete products and structures. Under certainconditions and for properly defined ranges of material,useful correlations may be established between the velocityand properties such as the modulus of elasticity andstrength.

This ultrasonic technique is completely distinct from thepulse-echo technique commonly used for the detection oflocalized defects in metals. The high frequency wavesnecessary for metals cannot travel through concrete as theyare rapidly scattered and dispersed by the aggregateparticles and pores. However, individual defects can bedetected if they are sufficiently extensive to influencepulse transit times by increasing the effective path lengthor cutting out the signal altogether.

All ultrasonic pulses passing through concrete are dattenuated by an amount depending on their frequencyand the properties of the concrete. This attenuation is noteasy to measure and techniques based on this aspect ofpulse propagation have not been widely used, but helpfuladditional information may be obtained under certaincircumstances. Specialist literature should be consulted.

This standard is concerned with the measurement of thevelocity of transmission of longitudinal ultrasonic pulsesthrough concrete and it is hoped that better understandingof the measuring technique will result from its use. Itincludes some recommendations on the interpretation oftest results and their correlation with elastic modulus andstrength tests.

Compliance with a British Standard does not of itselfconfer immunity from legal obligations.

L 0BNID Amendment No. 1

AMD 6659

published and effective from 28 February 1991toBS1881 :Part203: 1986

Testing concrete

Part 203. Recommendations for measurement of

velocity of ultrasonic pulses in concrete

Revised text

----------------___-------------------------------

AMD 6669 Clause 5.5 Setting the zero for the timing equipmentFebruary 1991

In paragraph 2 delete the second sentence and substitute the following:

‘A bar having a pulse transit time of about 25 ps is suitable(see also 5.6) .’

L

AMD 6669February 1991

Clause 5.6 Checking the accuracy of transit time measurementsIn paragraph 2, line 3 delete ‘accurately’ and substitute the following:

‘to an accuracy of + 0.2 vs. This performance check shouldbe carried out at intervals no longer than 5 years or if theinstrument is damaged or malfunctions in such a way thatit has to be returned to the manufacturer for repair. In allcases the performance check should be carried out andcertified either by the manufacturer or by an accreditedlaboratory, in both cases using reference bars of which thetransit times are traceable to the National Standard ofTime.’

Delete note 1.

Delete paragraph 3 and note 2 and substitute the following:

‘The two reference bars should have pulse transit times ofabout 25 ps and 100 ps, respectively, and these timesshould be marked on the bars to the nearest 0.1 vs. Theshorter of the reference bars should first be used to set thezero for the equipment, as explained in 5.5. The longer ofthe reference bars should then be used to check the accuracyof pulse transit time measurement by the equipment.Measurements are made on these reference bars by placing atransducer at each end and taking the transit time readingas described in 5.4. The measurement obtained should notdiffer from the known value for the longer reference bar bymore than f 0.5 %.’

--------------------------------------------------

AhlD 6669February 1991

Clause 9.3 Estimating the depth of a surface crackDelete equation 6 and substitute the following:

p; 3Tz2+2T3 ’ -25 ,T2 T3 ) 1

AMD 6669 Clause 12 Determination of the modulus of elasticity and dynamic Poisson’s

cFebruary 1991 ratio

Delete equation 10 and substitute the following:

‘(1 +u) (1 - 20) = 4n2 L2 1O-6 ’

(1 -u) V2--________-_________---------------- -------- _-----

L AMD 9999February 1991

Table 3. Values of dynamic Poisson’s ratio

Delete the heading to column 1 and substitute the following:

‘n L lOma ’

V

------_--__--_______------------------------------

c

9102 CAB/4

3

Amendment No. 2published and effective from 30 August 1991

toBS1661: Part203: 1966

Testing concrete

Part 203. Recommendations for measurement of

velocity of ultrasonic pulses in concrete

AM0 6766

Correction

Note. This amendmant corrects an error introduced by Amendment No. 1.

------------em- --------me------ ------------------.

AMD 9799August 1991

Clause 9.3 Estimating the depth of a surface crack (as amended by Amendment No. 1)

Delete equation 6 and substitute the following.

----------------------w--m-e- ---------------------

9109 CAB14

BS 1881:Part203: 1986

Contents

y:*

Foreword Inside front coverCommittees responsible Back cover

Recommendations

1 scope2 Definitions3 Applications4 Principle5 Measuring apparatus6 Determination of pulse velocity

Page

7 Factors influencing pulse velocitymeasurements

8 Determination of concrete uniformity9 Detection of defects

10 Determination of changes in concreteproperties

11 Correlation of pulse velocity and strength12 Determination of the modulus of elasticity

and dynamic Poisson’s ratio13 Report

71010

1313

1415

Appendix

A Bibliography 16

Tables

1 Effect of temperature on pulse transmission 72 Effect of specimen dimensions on pulse

transmission 73 Values of dynamic Poisson’s ratio 144 Empirical relationship between static and

dynamic moduli of elasticity and pulse velocity 15

Figures

1 Methods of propagating and receivingultrasonic pulses 4

2 Pulse velocity determination by indirect(surface) transmission 6

3 Influence of steel reinforcement on pulsevelocity: bar parallel to pulse path 9

4 Influence of steel reinforcement on pulsevelocity: correction factors for bars parallelto pulse path 9

5 Influence of steel reinforcement on pulsevelocity: bars normal to pulse path 10

6 Arrangement for determination of crack depth 12

c

1

BS1881:Part203:1986

Recommendations

1 Scope

This Part of BS 1881 gives recommendations on thenondestructive testing of concrete in the form of plain,reinforced and prestressed test specimens, precastcomponents and structures by the measurement ofultrasonic pulse velocity [ 1,2] .NOTE 1. The numbers in square brackets refer to the bibliographicreferences in appendix A.NOTE 2. The titles of the publications referred to in this standardare listed on the inside beck cover.

2 Definitions

For the purposes of this Part of BS 1661 the definitionsgiven in BS 3663 : Part 4 and BS 6100 : Part 6 apply,together with the following.

2.1 transit time. Time taken for an ultrasonic pulse totravel from the transmitting transducer to the receivingtransducer, passing through the interposed concrete.

2.2 onset. Leading edge of the pulse detected by themeasuring apparatus.

3 Applications

Measurement of the velocity of ultrasonic pulses oflongitudinal vibrations passing through concrete may beused for the following applications, described in detail inclauses 6 to 12:

(a) determination of the uniformity of concrete in orbetween members [3,4,51, clause 6

(b) detection of the presence and approximate extent ofcracks, voids and other defects [6,7,6], clause 9

(c) measurement of changes occurring with time in theproperties of the concrete [9, 10, 111, clause 10

(d) correlation of pulse velocity and strength as ameasure of concrete quality (12, 13, 141 , clause 11

(e) determination of the modulus of elasticity anddynamic Poisson’s ratio of the concrete [ 15, 161,clause 12

The velocity of an ultrasonic pulse is influenced by thoseproperties of the concrete which determine its elasticstiffness and mechanical strength. The variations obtainedin a set of pulse velocity measurements made alongdifferent paths in a structure reflect a correspondingvariation in the state of the concrete (see clause 6).

When a region of low compaction, voids or damagedmaterial is present in the concrete under test, a corres-ponding reduction in the calculated pulse velocity occursand this enables the approximate extent of theimperfections to be determined (see clause 9). As concretematures or deteriorates, the changes which occur with timein its structure are reflected in either an increase or adecrease, respectively, in the pulse velocity. This enables

the changes to be monitored by making tests at appropriateintervals of time (see clause 10).

Pulse velocity measurements made on concrete structuresmay be used for quality control purposes. In comparisonwith mechanical tests on control samples such as cubes orcylinders, pulse velocity measurements have the advantagethat they relate directly to the concrete in the structurerather than to samples which may not be always trulyrepresentative of the concrete in situ.

Ideally, pulse velocity should be related to the results oftests on structural components and, if a correlation can beestablished with the strength or other required propertiesof these components, it is desirable to make use of it. Suchcorrelations can often be readily established directly for

precast units and can also be found for in situ work

(see clause 11).

Empirical relationships may be established between thepulse velocity and both the dynamic and static elastic

moduli and the strength of concrete (see clause 12).The latter relationship is influenced by a number of factorsincluding the type of cement, cement content, admixtures,type and size of aggregate, curing conditions and age ofconcrete. Caution should be exercised when attempting toexpress the results of pulse velocity tests in terms ofstrengths or elastic properties, especially at strengthsexceeding 60 MPa.

4 Principle

A pulse of longitudinal vibrations is produced by anelectro-acoustical transducer which is held in contact withone surface of the concrete under test. After traversing aknown path length L in the concrete, the pulse ofvibrations is converted into an electrical signal by a secondtransducer. Electronic timing circuits enable the transittime T of the pulse to be measured.

The pulse velocity v (in km/s or m/s) is given by

Ly=-T

d

where

L is the path length;

T is the time taken by the pulse to traverse that length.

A pulse of vibrations of ultrasonic rather than sonicfrequency is used for two reasons:

(a) to give the pulse a sharp leading edge;

(b) to generate maximum energy in the direction ofpropagation of the pulse.

When the pulse is coupled into the concrete from atransducer, it undergoes multiple reflections at theboundaries of the different material phases within theconcrete. A complex system of stress waves is developedwhich includes both longitudinal and shear wavespropagating throughout the concrete.

d

2

BSlB81 :Part203:1986

L

ci*

c

c

5 Measuring apparatus

5.1 General

The apparatus consists essentially of an electrical pulsegenerator, a pair of transducers, an amplifier and anelectronic timing device for measuring the time intervalbetween the onset of a pulse generated at the transmittingtransducer and the onset of its arrival at the receivingtransducer. Two forms of electronic timing apparatus anddisplay are available, one of which uses a cathode ray tubeon which the received pulse is displayed in relation to asuitable time scale, the other uses an interval timer with adirect reading digital display.

5.2 Performance characteristics

The apparatus should have the following characteristics.

(a) It should be capable of measuring transit time overpath lengths ranging from 100 mm to 3 m (see 5.7),to an accuracy of + 1 % as determined by the proceduredescribed in 5.6.

(b) The electronic excitation pulse applied should havea rise time not greater than one quarter of the naturalperiod of the transmitting transducer. This is to ensurea sharp pulse onset.

(c) The interval between pulses should be long enoughto ensure that the onset of the received signal in smallconcrete test specimens is free from interference byreverberations produced within the preceding workingcycle.

(d) The apparatus should maintain its performance overthe ranges of ambient temperature, humidity and powersupply voltage stated by the suppliers.

5.3 Transducers

5.3.1 Type. Any suitable type of transducer operatingwithin the frequency range stated in 5.3.2 may be used.Piezoelectric and magneto-strictive types of transducersare normally used, the latter being more suitable for thelower part of the frequency range.

5.3.2 Natural frequency of transducetx The naturalfrequency of the transducers should normally lie withinthe range of 20 kHz to 150 kHz, although frequencies aslow as 10 kHz may be used for very long path lengths and,at the other extreme, frequencies up to 1 MHz for mortarsand grouts.

High frequency pulses have a well defined onset but,as they pass through concrete, become attenuated morerapidly than pulses of lower frequency. It is, therefore,preferable to use high frequency transducers for short pathlengths and low frequency transducers for long pathlengths. Transducers with a frequency of 50 kHz to 60 kHzare suitable for most common applications.

5.4 Determination of the arrival time of the pulse onset

5.4.1 Genera/. The object of the determination is tomeasure the time taken for the onset of the pulse to passthrough the concrete. As explained in 5.3.2, the onsetbecomes less well defined as the pulse proceeds further

into the concrete. It is therefore necessary that theapparatus be able to detect the arrival of the earliest partof the pulse. Although it is technically possible to separatethe signal from background noise if the signal/noise ratiois less than one, the required accuracy of 1 % of themeasured transit time will only be achieved if this ratio isgreater than one.

5.4.2 Cathode ray oscilloscope. In the case of timingdevices in which a cathode ray oscilloscope display is used,the received pulse should be amplified to the maximumpossible level, limited only by the appearance of ‘grass’ onthe time-base trace of the display. The onset of the pulseshould be taken as the point of tangent of the signal curvewith the initial horizontal time-base line. Alternatively,some other well defined feature of the curve, such as thefirst crossover, may be used.

5.4.3 Digital instruments. For digital instruments, thereceived pulse should be amplified and shaped to the leveland rise time, respectively, required for triggering thedigital timer.

The timer should trigger from a point on the leading edgeof the pulse within a time corresponding to the range ofaccuracy specified in 5.2. However, the absolute accuracyof the instrument will always be limited by the signal/noiseratio.

When using a digital indicating device, there may beoccasions when the second wave of the pulse, rather thanthe first, triggers the instrument. The general pattern ofresults should be inspected to enable errors ofinterpretation from this cause to be eliminated.

5.5 Setting the zero for the timing equipment

It is necessary to establish the correct zero reading for theapparatus, since the indicated measurement is influencedby a time delay due both to transmission of the pulsethrough the transducer material and transmission of theelectrical signal along the transducer cables. The apparatusshould incorporate a suitable time delay adjustment so thatthe indicated readings may be made independent of thiseffect.

The time delay adjustment should be made while thetransducers are coupled to the opposite ends of a referencebar for which the transit time is accurately known.A suitable bar is described in 5.6. It is important always touse the same technique to place the transducers on thereference bar. A minimum amount of couplant should beused and the transducer pressed firmly against the end ofthe bar. Any other technique, such as sliding the trans-ducers onto the bar, may give a significantly different zeroreading and should be avoided.

Adjustments to the time delay to provide a correct zerosetting for the apparatus should be made each time theequipment is used, whenever transducers are interchanged,whenever different transducers are used and wheneverdifferent lengths of cables are used. It may also be necessaryto carry out more frequent checks on the zero setting,depending on the stability of the electronic circuits orcables.

3

BS 1881 : Part 203 : 1986

5.6 Checking the accuracy of transit time measurements IThe accuracy of the measurement of pulse transit time willdepend on the accuracy of the electronic device used formeasuring time intervals and also on its sensitivity indetecting the onset of the pulse.

Overall performance should be checked by makingmeasurements on two reference bars in which the pulsetransit times are known accurately.NOTE 1. If necessary, the accuracy of the transit times indicatedon the reference bars may be checked by the National PhysicalLaboratory.

The two reference bars shouid have pulse transit times ofabout 25 us and 100 us, respectively. The shorter of thereference bars should be used to set the zero for theequipment, as explained in 5.5. The longer of the referencebars should be used to check the accuracy of pulse transittime measurement by the equipment. Measurements aremade on these reference bars by placing a transducer ateach end and taking the transit time reading, as describedin 5.4. The use of the longer reference bar is not normallynecessary as it may be assumed that the instrumentmanufacturer will have checked his apparatus in thisrespect. The measurement obtained should not differ fromthe known value for the reference specimen by morethan + 0.5 %.

NOTE 2. Pulse transit times are stated by the reference barsupplier to an accuracy of f 0.2 ps.

5.7 Accuracy of path length measurement

The accuracy of measurement should be better than + 1 %.Where direct measurement of path length is not physicallypossible, the nominal dimension and its tolerance asspecified by the designer should be used and the factreported. Where path lengths of less than 300 mm aretested in this way, unacceptable errors will occur.

6 Determination of pulse velocity

6.1 Transducer arrangement

The receiving transducer detects the arrival of thatcomponent of the pulse which arrives earliest. This isgenerally the leading edge of the longitudinal vibration.

Although the direction in which the maximum energy ispropagated is at right angles to the face of the transmittingtransducer, it is possible to detect pulses which havetravelled through the concrete in some other direction.It is possible, therefore, to make measurements of pulsevelocity by placing the two transducers on either:

(a) opposite faces (direct transmission);

(b) adjacent faces (semi-direct transmission); or

(c) the same face (indirect or surface transmission).

These three arrangements are shown in figures 1 (a), 1 (b)and 1 (c). Figure 1 (a) shows the transducers directlyopposite to each other on opposite faces of the concrete.It is, however, sometimes necessary to place the transducerson opposite faces but not directly opposite each other.Such an arrangement is to be regarded as semi-directtransmission.

(b) Semi-direct transmission

(c) indirect or surface transmission

KWTransmitter IT)

Receiver (RI

Figure 1. Methods of propagating and receivingultrasonic pulses

BS1881:Part203: 1986

L

L

L

6.2 Determination of pulse velocity by direct transmission

Where possible, the direct transmission arrangement shouldbe used since the transfer of energy between transducers isat its maximum and the accuracy of velocity determinationis therefore governed principally by the accuracy of thepath length measurement. The couplant used should bespread as thinly as possible to avoid any end-effectsresulting from the different velocities in couplant andconcrete.

6.3 Determination of pulse velocity by semi-directtransmission

The semi-direct transmission arrangement has a sensitivityintermediate between those of the other two arrangementsand, although there may be some reduction in the accuracyof measurement of the path length, it is generally found tobe sufficiently accurate to take this as the distancemeasured from centre to centre of the transducer faces.This arrangement is otherwise similar to direct transmission.

6.4 Determination of pulse velocity by indirect or surfacetransmission

Indirect transmission should be used when only one face ofthe concrete is accessible, when the depth of a surface crackis to be determined or when the quality of the surfaceconcrete relative to the overall quality is of interest (seeclause 9).

It is the least sensitive of the arrangements and, for agiven path length, produces at the receiving transducer asignal which has an amplitude of only about 2 % or 3 %of that produced by direct transmission. Furthermore,this arrangement gives pulse velocity measurements whichare usually influenced by the concrete near the surface.This region is often of different composition from that ofconcrete within the body of a unit and the test results maybe unrepresentative of that concrete. The indirect velocityis invariably lower than the direct velocity on the sameconcrete element. This difference may vary from 5 % to20 % depending largely on the quality of the concrete undertest. Where practicable, site measurements should be madeto determine this difference.

With indirect transmission there is some uncertaintyregarding the exact length of the transmission path becauseof the significant size of the areas of contact between thetransducers and the concrete. It is therefore preferable tomake a series of measurements with the transducers atdifferent distances apart to eliminate this uncertainty.To do this, the transmitting transducer should be placedin contact with the concrete surface at a fixed point x andthe receiving transducer should be placed at fixedincrements x, along a chosen line on the surface. Thetransmission times recorded should be plotted as points ona graph showing their relation to the distance separatingthe transducers. An example of such a plot is shown asline (b) in figure 2.

The slope of the best straight line drawn through the pointsshould be measured and recorded as the mean pulse velocityalong the chosen line on the concrete surface. Where the

points measured and recorded in this way indicate adiscontinuity, it is likely that a surface crack or surfacelayer of inferior quality is present (see 9.4) and a velocitymeasured in such an instance is unreliable.

6.5 Coupling the transducer onto the concrete

To ensure that the ultrasonic pulses generated at thetransmitting transducer pass into the concrete and are thendetected by the receiving transducer, it is essential thatthere be adequate acoustical coupling between the concreteand the face of each transducer. For many concretesurfaces, the finish is sufficiently smooth to ensure goodacoustical contact by the use of a coupling medium and bypressing the transducer against the concrete surface.Typical couplants are petroleum jelly, grease, soft soap andkaolin/glycerol paste. It is important that only a very thinlayer of coupling medium separates the surface of theconcrete from its contacting transducer. For this reason,repeated readings of the transit time should be made untila minimum value is obtained so as to allow the layer ofcouplant to become thinly spread.

Where possible, the transducers should be in contact withconcrete surfaces which have been cast against formworkor a mould. Surfaces formed by other means, e.g.trowelling, may have properties differing from those of themain body of material. If it is necessary to work on such asurface, measurements should be made over a longer pathlength than would normally be used. A minimum pathlength of 150 mm is recommended for direct transmissioninvolving one unmoulded surface and a minimum of400 mm for indirect transmission along one unmouldedsurface.

When the concrete surface is very rough and uneven, thearea of the surface where the transducer is to be appliedshould be smoothed and levelled. Alternatively, a smoothingmedium such as quick-setting epoxy resin or plaster maybe used, but good adhesion between the concrete surfaceand smoothing medium has to be ensured so that the pulseis propagated correctly into the concrete under test. It isimportant to ensure that the layer of smoothing mediumis as thin as possible. If it is necessary to make a significantbuild-up, then the pulse velocity of the smoothing mediumhas to be taken into account.

In order to avoid the problems of obtaining good acousticalcontact between the transducers and a surface which is notsmooth enough to allow the use of a thin couplant, specialtransducers have been developed which impart or pick upthe pulse through integral probes having 6 mm diametertips. A receiving transducer with a hemispherical tip hasbeen found to be very successful but the transmitter cannotimpart enough energy into the concrete surface to ensuretransmission over any but short path lengths. Other trans-ducer configurations have been developed to deal withspecial circumstances. It should be noted that a zeroadjustment will almost certainly be required when specialtransducers are used. Unsatisfactory seating of thetransducers is often indicated by an excessive degree of‘hunting’; with a satisfactory degree of coupling, thedigital reading will settle rapidly.

5

BS 1881 : Part 203 : 1986

4lb)

Distance X (mm)

(a) Results for concrete with the top 50 mm of inferior quality (see BA).(b) Results for homogeneous concrete (see 6A).

figure 2. Pulsa velocity determination by indiract (surface) transmission

6

BS 1881:Part203:1986

L

I=iI=

i-

CA*

L

c

7 Factors influencing pulse velocitymeasurements

7.1 General

In order to provide a measurement of pulse velocity whichis reproducible and which depends essentially on theproperties of the concrete under test, it is necessary toconsider the various factors which can influence pulsevelocity and its correlation with various physical propertiesof the concrete.

7.2 Moisture content

The moisture content has two effects on the pulse velocity,one chemical, the other physical. These effects areimportant in the production of correlations for theestimation of concrete strength. Between a properly curedstandard cube and a structural element made from the sameconcrete, there may be a significant pulse velocitydifference, Much of the difference is accounted for by theeffect of different curing conditions on the hydration ofthe cement while some of the difference is due to thepresence of free water in the voids. It is important thatthese effects are carefully considered when estimatingstrength.

7.3 Temperature of the concrete

Variations of the concrete temperature between 10 ‘C and30 “C have been found to cause no significant changewithout the occurrence of corresponding changes instrength or elastic properties. Corrections to pulse velocitymeasurements should be made only for temperaturesoutside this range, as given in table 1.

r Table 1. Effect of temperature on pulse transmission

Correction to the measured pulse velocity

Temperature Air-dried Wetratumtedconcrete concrete

“C % %

60 +5 +440 +2 +1.720 0 0

0 -0.5 -1- 4 -1.5 - 7.5

indicate a tendency for velocity to reduce slightly withincreasing path length. This is because the higher frequencycomponents of the pulse are attenuated more than thelower frequency components and the shape of the onset ofthe pulse becomes more rounded with increased distancetravelled. Thus, the apparent reduction of pulse velocityarises from the difficulty of defining exactly the onset ofthe pulse and this depends on the particular method usedfor its definition. This apparent reduction in velocity isusually small and well within the tolerance of timemeasurement accuracy given in 5.2.

7.5 Shape and size of specimen

The velocity of short pulses of vibrations is independent ofthe size and shape of specimen in which they travel, unlessits least lateral dimension is less than a certain minimumvalue. Below this value, the pulse velocity may be reducedappreciably. The extent of this reduction depends mainlyon the ratio of the wavelength of the pulse vibrations tothe least lateral dimension of the specimen but it isinsignificant if the ratio is less than unity. Table 2 gives therelationship between the pulse velocity in the concrete,the transducer frequency and the minimum permissiblelateral dimension of the specimen.

If the minimum lateral dimension is less than thewavelength or if the indirect transmission arrangement isused, the mode of propagation changes and, therefore,the measured velocity will be different. This is particularlyimportant in cases where concrete elements of significantlydifferent sizes are being compared.

Table 2. Effect of specimen dimensions on pulsetransmission

Transducerfrequency

Pulse velocity in concrete lin km/s)

VC = 3.5 VC = 4.0 VC = 4.5

Minimum permissible lateralqwcimen dimension

kHz mm mm mm

24 146 167 18854 65 74 8382 43 49 55

150 23 27 30

7.4 Path length

The path length over which the pulse velocity is measuredshould be long enough not to be significantly influencedby the heterogeneous nature of the concrete. It isrecommended that, except for the conditions stated in 7.5,the minimum path length should be 100 mm for concretein which the nominal maximum size of aggregate is 20 mmor less and 150 mm for concrete in which the nominalmaximum size of aggregate is between 20 mm and 40 mm.The pulse velocity is not generally influenced by changes inpath length, although the electronic timing apparatus may

7.6 Effect of reinforcing bars

7.6.1 General. The pulse velocity measured in reinforcedconcrete in the vicinity of reinforcing bars is usually higherthan in plain concrete of the same composition. This isbecause the pulse velocity in steel may be up to twice thevelocity in plain concrete and, under certain conditions,the first pulse to arrive at the receiving transducer travelspartly in concrete and partly in steel.

The apparent increase in pulse velocity depends upon theproximity of the measurements to the reinforcing bar,the diameter and number of bars and their orientation withrespect to the propagation path 1171, The frequency of

7

BS 1881 : Part203 : 1986

the pulse and surface conditions of the bar may both alsoaffect the degree to which the steel influences the velocitymeasurements. Corrections to measured values to allow forreinforcement will reduce the accuracy of estimated pulsevelocity in the concrete so that, wherever possible,measurements should be made in such a way that steel doesnot lie in or close to the direct path between transducers.Electromagnetic cover measuring devices (see BS 1881 :Part 204’) should be used to assist in locating steel.

7.6.2 Axis of reinforcing bar parallel to direction ofpropagation. The position of paths along which pulses arepropagated should be chosen, whenever possible, so as toavoid the vicinity of reinforcing bars parallel to these paths.If this cannot be achieved, the measured values of pulsevelocity should be corrected to take into account thepresence .of steel. The correction will depend on thedistance between the line of the path and the edge of thenearest bar, the bar diameter, and the pulse velocity inthe surrounding concrete.

The pulse velocity in the concrete v, (in km/s) is given by

2 av,VC =

J (4a2 + (TV, - L)‘)(2)

provided that v, > v,

where

v, is the pulse velocity in the steel bar (in km/s);

a is the offset, measured as the distance from thesurface of the bar to the line joining the nearest pointof the two transducers (in mm) (see figure 3);

T is the transit time (in ps);

L is the length of the direct path between transducers(in mm).

The influence of the steel disappears when

J

(v, - VC)f>X -

(v, + VC)

and equation 2 is no longer applicable. The zone withinwhich the steel may influence measurements thus dependsupon the relative values of pulse velocity within the steeland concrete but an upper limit of a/L of about 025 maybe expected for large diameter bars in low quality concrete.For high quality concrete the limiting value of a/L isunlikely to be greater than 0.15 but may be considerablyless for bar diameters of 12 mm or below. Bars of 6 mmdiameter or less will be virtually impossible to detect inpractical situations and may be ignored.

The major difficulty in applying equation 2 lies in decidingon the value of v, since this is influenced both by the bardiameter and the pulse velocity in the surrounding concrete.A measure of this may generally be obtained by propagatinga pulse along the axis of the embedded bar, making dueallowance for any concrete cover at either end.

Equation 2 may conveniently be modified to give thefollowing

VC = k v,

where

v, is the measured apparent pulse velocity (L/T)(in km/s);

k is the correction factor given by

(3)

r+2 + J(l-yf,( >

in which

Typical values of 7 are plotted in figure 3 for a range ofcommonly occurring values of v, and bar diameter for apulse frequency in the order of 54 kHz. The value of-yobtained from this figure, for an assumed v,, may then beused in conjunction with figure 4 to provide an estimateof k for use in equation 3. An iterative procedure may benecessary to achieve a reliable estimate of v,.

These equations will only be valid for offsetsa which arein excess of approximately twice the end cover to the bar.For smaller offsets the pulse is likely to pass through thefull length of the bar. For bars lying directly in line withthe transducers, the correction factor is given by

d

k=,-L,‘ (1-Y)

where

(4)

L s is the length of‘the bar (in mm).

An estimate of v, is likely to be accurate within f 3 %provided that there is good bond between the steel andconcrete and that there is no cracking of the concrete inthe test zone.

Corrected pulse velocity measurements should be treatedwith caution since they relate only to the concrete in theimmediate vicinity of the transducers and the reinforcingbar, and do not represent the concrete along the linejoining the centres of the transducers.

d

7.6.3 Axis of reinforcing bar perpendicular to direction ofpropagation. The maximum influence of the presence ofthe reinforcing bars can theoretically be calculated assumingthat the pulse traverses the full diameter d of each barduring its passage. This is illustrated in figure 5. The effectof the bars on the pulse is complex and the apparent pulsevelocity in the steel will be reduced below that to beexpected along the axis of bars of similar size.

For practical purposes where 54 kHz transducers are beingused, bars of diameter less than 20 mm may be ignoredsince their influence will be negligible. An estimate of the

*Revision of BS 44Cl8 : Part 1 in preparation.

8

BS 1881:Part203:1986

(a) Cross section of concrete with lateral reinforcement

1.0

0.9

$ 0.8YII

)z 0.7

0.6

0.5 ,io zb io 4b To

Bar diameter (mm)0

(bj Relationship between bar diameter end velocitv ratio

Figure 3. Influence of steel reinforcement on pulse velocity: bar perallel to pulse path

Fkn 4. Infhenca of steel reinforcement on pulse velocity: correstion factors_.. ._ -_. __- ___for ban parallel to pulse path (e > 26) (see figure 3(a))

9

BS 1881:Part203: 1986

L 0.6L3.0 d

(a) Cross section of concrete with longitudinal reinforcement0 . 5 1

10 20 30 40 50Bar diameter (mm)

(b) Relationship of bar diameter to velocity ratio

Figure 5. Influence of steel reinforcement on pulse velocity: bars normal to pulse path

average influence can be obtained for well bonded bars ofbetween 20 mm and 50 mm diameter by considering themas equivalent longitudinal bars of total path length L’,(see figure 5). The method described in 7.6.2 for bars lyingdirectly in line with the transducers (see equation 4) maybe used for this purpose in conjunction with the value of-ydetermined from figure 5 which takes account of thereduced velocity in the steel. The influence of transversesteel will be reduced by bond deficiencies, and is difficultto assess with any degree of accuracy if the bars do not liedirectly in line with the path between transducers,

8 Determination of concrete uniformity

Heterogeneities in the concrete within or between memberscause variations in pulse velocity which, in turn, are relatedto variations in quality. Measurements of pulse velocityprovide means of studying the homogeneity and, for thispurpose, a system of measuring points which coversuniformly the appropriate volume of concrete in thestructure has to be chosen.

Variations in pulse velocity are influenced by magnitude ofthe path length because this determines the effective size ofconcrete sample which is under examination during eachmeasurement. The importance of variations should bejudged in relation to the effect which they can be expectedto have on the required performance of the structuralmember being tested. This generally means that thetolerance allowed for quality distribution within membersshould be related either to the stress distribution withinthem under critical working load conditions or to exposureconditions.

The number of individual test points depends upon the sizeof the structure, the accuracy required and the variabilityof the concrete. In a large unit of fairly uniform concrete,testing on a 1 m grid is usually adequate but, on small unitsor variable concrete, a finer grid may be necessary. It shouldbe noted that, in cases where the path length is the samethroughout the survey, the measured time may be used toassess the concrete uniformity without the need to convert

9 Detection of defects

NOTE. See reference [18].

9.1 General

The use of the ultrasonic pulse velocity technique to detectand define the extent of internal defects should berestricted to wellqualified personnel with previous

it to a velocity. This technique is particularly suitable forsurveys where all the measurements are made by indirecttransmission.

It is possible to express homogeneity in the form of astatistical parameter such as the standard deviation orcoefficient of variation of the pulse velocity measurementsmade over the grid. However, such parameters can only beproperly used to compare variations in concrete units ofbroadly similar dimensions.

d

10

BS 1881 : Part203 : 1986

L

y:*

L

L

experience in the interpretation of survey results. It isimportant that ultrasonic pulse velocity measurements aremade and interpreted by persons with experience andtraining in the technique. Attention is drawn to thepotential risk of drawing conclusions from single results.

When an ultrasonic pulse travelling through concrete meetsa concrete-air interface, there is negligible transmission ofenergy across this interface. Thus, any air-filled crack orvoid lying immediately between two transducers willobstruct the direct ultrasonic beam when the projectedlength of the void is greater than the width of thetransducers and the wavelength of sound used. When thishappens, the first pulse to arrive at the receiving transducerwill have been diffracted around the periphery of thedefect and the transit time will be longer than in similarconcrete with no defect.

It is possible to make use of this effect for locating flaws,voids or other defects greater than about 100 mm indiameter or depth. Relatively small defects have little or noeffect on transmission times, but equally are probably ofminor engineering importance. Plotting contours of equalvelocity often gives significant information regarding thequality of a concrete unit.

In cracked members, where the broken faces of themembers are held tightly together in close contact bycompression forces, the pulse energy may pass unimpededacross the crack. As an example, this may occur in crackedvertical bearing piles. If the crack is filled with liquid whichtransmits the ultrasonic energy, e.g. in marine structures,the crack is undetectable using digital reading equipment.Measurements of attenuation may give valuable informationin these cases.

Recommendations on the interpretation of the results oftests made to detect defects are given in 9.2 to 9.4 but itis strongly emphasized that the assumptions made havebeen simplified.

9.2 Detecting large voids or cavities

A grid should be drawn on the concrete member with itspoints of intersection spaced to correspond to the size ofvoid that would significantly affect its performance. Asurvey of measurements at the grid points enables a largecavity to be investigated by measuring the transit times ofpulses passing between the transducers when they are placedso that the cavity lies in the direct path between them.The size of such cavities may be estimated by assumingthat the pulses pass along the shortest path between thetransducers and around the cavity. Such estimates are validonly when the concrete around the cavity is uniformlydense and the pulse velocity can be measured in thatconcrete.

9.3 Estimating the depth of a surface crack

An estimate of the depth of a crack visible at the surfacemay be obtained by measuring the transit times across thecrack for two different arrangements of the transducersplaced on the surface. One suitable arrangement is shownin figure 6(a) in which the transmitting and receivingtransducers are placed at a distance x on opposite sides of

the crack and equidistant from it. Two values of x arechosen and transit times corresponding to these aremeasured; convenient values of x are 150 mm and 300 mm.If these values are used, the depth of the air-filled crack C(in mm) is given by:

c= 150J(4t,Z - rz21

(fz2 - r121 (5)

where

t1 is the transit time when x is 150 mm [in us);

f2 is the transit time when x is 300 mm (in us).

Equation 5 is derived by assuming that the plane of thecrack is perpendicular to the concrete surface and thatthe concrete in the vicinity of the crack is of reasonablyuniform quality.

A check may be made to assess whether or not the crack islying in a plane perpendicular to the surface by placingboth transducers near to the crack (as shown in figure 6(b))and moving each in turn away from the crack. If a decrease inin transit time occurs when one transducer is moved itindicates that the crack slopes towards that transducer.

In an alternative arrangement, the transmitting transduceris placed at a distance of 2.5~ from the centre of the crackand three readings of the transit time taken with thereceiving transducer at distances of y, 2y and 3y from thetransmitter in the direction of the crack. The transit timesare plotted against distance as in figure 6(c) in which y is150 mm. If the projection of the straight line through thepoints (y, T1 ) and (2y, T2) passes through 0 there are nohidden cracks and the depth of the visible crack C (in mm)is given by: -

where L -1

T2 is the transit time for distance 2y (in us);

Ts is the transit time for distance 3y (in us).

Figure 6(c) shows therway in which the transit timegradually returns to the value expected in untrackedconcrete as the receiver is moved further from the crack.

(6)

9.4 Estimating the thickness of a layer of inferior qualityconcrete

Concrete may be suspected of having a surface layer ofpoor quality. This may occur during manufacture or ariseas a result of damage by fire, frost, sulphate attack, etc.The thickness of such a layer of concrete can be estimatedfrom ultrasonic measurements of transit times along thesurface. The procedure described in clause 10 should beused and the results plotted on a graph as in figure 2. At theshorter distance of separation of the transducers, the pulsetravels through the surface layer and the slope of theexperimental line gives the pulse velocity in this surfacelayer. Beyond a certain distance of separation the firstpulse to arrive has passed along the surface of the under-lying higher quality concrete and the slope of theseexperimental points gives the velocity in that concrete.

11

BS1881:Part203:1986

-Ti R

XI-,. .,::~. ., 1.: *... ::.. -..: ; . . . . . .‘,x:‘...‘( .:~~o~..,o.~:.:~.~.‘:.~.:~~.~:~

5

,; .,.. ,. :, ;, *...:.:.. .o-..::. .o. 0.;

.o..‘I.. ..‘.o’.:..: . . .,

-o’.,‘..?: :? _....* . . . .._ ‘0..‘..~,~~_..~ : .(.. *.. .P.,:;:.&,...”~‘~‘:~.,q~~~;,.b’.:‘o . .._..: .I’ C,.’ ::. . . . ‘...* .:.. .‘.o..-.6‘.: p.:,‘o:::~, f .o. .:’ *.. ; . ...o... ..:. ,..*.,., o::::- &.‘.I. . . . ‘.. 0:;. -

(a) Crack perpendicular to surface

(b) Oblique crack

300 -

2M-

I I I I I I Iloo 200 300 400 500 600 700 800 900 1000 1100 1200

Stationary transducer Crack Distance (mm)position position

(c) Effect of crack on surface measurement

F’wre 6. Arrengament for detemination of crack depth

12

BS1881:Part203:1986

L

ci*

c

L

The distance x0 at which the change of slope occurs(see figure 2) together with the measured pulse velocitiesin the two different layers of concrete, enables an estimateof the thickness t (in mm) of the surface layer to be madeas follows:

(v,- VCI)+$! -J (VI + Vd)

(7)

where

vd is the pulse velocity in the damaged concrete(in km/s);

v, is the pulse velocity in the underlying sound concrete(in km/s);

x, is the distance from the transmitter at which theslope changes (in mm).

The method is applicable to extensive surface areas in whichthe inferior concrete forms a distinct layer of fairly uniformthickness and when v, is appreciably smaller than v,.

Localized areas of damaged or honeycombed concrete aremore difficult to test but it is possible to derive anapproximate thickness of such localized poor qualitymaterial if both direct transmission and surface propagationmeasurements are made.

10 Determination of changes in concreteproperties

Changes occurring with time in the properties of concretecaused by the hydration process, by the influence ofenvironmental effects or by overloading, may be determinedby repeated measurements of pulse velocity at differenttimes with the same transducers in the same position.Measurements of changes in pulse velocity are usuallyindicative of changes in strength and have the advantagethat they can be made over progressive periods of time onthe same test piece throughout the investigation.

Pulse velocity measurements are particularly useful forfollowing the hardening process, especially during thefirst 36 h. Here, rapid changes in pulse velocity areassociated with physiochemical changes in the cementpaste structure and it is desirable to make measurementsat intervals of 1 h or 2 h if these changes are to be followedclosely. As the concrete hardens these intervals may belengthened to 1 day or more after the initial period of36 h has elapsed.

11 Correlation of pulse velocity andstrength

11.1 General

The quality of concrete is usually specified in terms ofstrength and it is, therefore, sometimes helpful to useultrasonic pulse velocity measurements to give an estimateof strength.

The relationship between ultrasonic pulse velocity andstrength is affected by a number of factors including age,curing conditions, moisture condition, mix proportions,type of aggregate and type of cement. If an estimate ofstrength is required it is therefore necessary to establish a

correlation between strength and velocity for the particulartype of concrete under investigation. This correlation hasto be established experimentally by testing a sufficientnumber of specimens to cover the range of strengthsexpected and to provide statistical reliability. The confi-dence that can be ascribed to the results will depend onthe number of samples tested. It is possible to establisha correlation between ultrasonic pulse velocity and strengtheither as measured in accordance with any of the testmethods described in BS 1881 : Parts 116 and 119 or bycarrying out tests on a complete structure or unit.

The reliability of the correlation will depend on the extentto which the correlation specimens represent the structureto be investigated. The most appropriate correlation will beobtained from tests in which pulse velocity and strength aremeasured on a complete structure or unit. It is sometimesmore convenient to prepare a correlation using tests onmoulded specimens. It should be noted that experience hasshown that correlations based on moulded specimensgenerally give lower estimates of strength than would beobtained by cutting and testing cores.

11.2 Correlation using moulded specimens

The strengths of a particular mix of cement and aggregatemay be varied by altering either:

(a) the water-cement ratio, or

(b) the age at test.

The method used for varying the strength of the specimensinfluences the correlation. It is therefore essential that onlyone method of strength variation be used for a particularcorrelation and that it be appropriate to the applicationrequired. The correlation of pulse velocity with strength isless reliable as the strength of concrete increases. A corre-lation obtained by varying the age of the concrete isappropriate when monitoring strength development butfor quality control purposes a correlation obtained byvarying the watercement ratio is preferable.

The appropriate test specimens should be made inaccordance with the methods described in BS 1881 :Parts 108,109 and 110. At least three specimens shouldbe cast from each batch. The pulse velocity should bemeasured across a specimen between moulded faces. In thecase of beams, it is preferable to measure the pulse velocityalong their length to obtain greater accuracy. For eachspecimen there should be at least three measurementsspaced between its top and bottom (see table 2 for edgedistances). The variation between the measured transittimes on single test specimetis should be within + 5 % ofthe mean value of these three measurements, otherwisethe specimen should be rejected as abnormal. The specimensshould then be tested for strength according to the methodsdescribed in BS 1881 : Parts 116 and 119.

The mean pulse velocity and mean strength obtained fromeach set of three nominally identical test specimens providethe data to construct a correlation curve. A correlationcurve produced in this way relates only to specimensproduced, cured and tested in a similar way; differentcorrelation curves will be obtained for the same mixes ifair curing is substituted for water curing.

13

BS 1881:Part203:1986

11.3 Correlation by tests on cores

When making a correlation from tests on cores it will notgenerally be possible to vary the strength of the concretedeliberately. Pulse velocity tests should, therefore, be usedto locate areas of different quality and cores taken fromthese areas will give a range of strengths. The pulse velocitythrough the concrete at proposed core locations should beused for preparing the correlation. Pulse velocities takenfrom cores after cutting and soaking will generally be higherthan those in the structure and should not be used for directcorrelation. The cores should be cut and tested for strengthin accordance with the method described in BS 1881 :Part 120 and a correlation curve plotted as describedin 11.2. The shape of the correlation line is sensibly thesame for any given concrete notwithstanding the curingconditions. It is therefore possible to use the curve derivedfrom reference specimens to extrapolate from the limitedrange that will normally be obtained from core samples.

11.4 Correlation with the strength of precast units

When precast components are required to conform tostrength requirements it may be possible to establishcorrelations between the pulse velocity measurements andthe particular types of strength tests. This should be doneby making pulse velocity measurements on the componentsin the appropriate regions where the concrete would beexpected to fail under the test loading conditions. Theprocedure for obtaining a graphical correlation in suchcases should be as described in 11.2.

11.5 Combination of pulse velocity with othermeasurements

Some improvement in the accuracy of strength estimatesmay be obtained by combining ultrasonic pulse velocitymeasurements with rebound hammer measurements asdescribed in BS 1881 : Part 202 (see [19], 1201 andBS 1881 : Part 201). Greater improvement can be madeby combining ultrasonic pulse velocity with densitymeasurements. In a structure, the density measurementsshould ideally be made along the same line of sight as thepulse velocity measurements. The density should bemeasured either by the methods described in BS 1881 :Part 114 or by a gamma ray attenuation technique,provided that the influence of reinforcing steel can beavoided. A separate correlation curve may then beconstructed for each value of density in the required range.

12 Determination of the modulus ofelasticity and dynamic Poisson’s ratio

The relationship between elastic constants and the velocityof an ultrasonic pulse travelling in an isotropic elastic

medium of infinite dimensions is given by the followingequation:

Ed

cpv2 (1 +u) (1 - 24

(1 - u)(8)

where

Ed is the dynamic elastic modulus (in MN/m2 1;

u is the dynamic Poisson’s ratio;

p is the density (in kg/m3 1;

v is the pulse velocity (in km/s).

As stated in 7.5, the pulse velocity is not significantlyaffected by the dimensions of the test specimen exceptwhen one or more of the lateral dimensions is smallrelative to the wavelength of the pulse.

If, therefore, the values of p and u are known, it is possibleto use equation 8 to determine the value of Ed in concretesamples for a wide range of shape or size. Similarly u couldbe determined if the values of p and Ed are known.

For laboratory specimens the ratio of Ed/p may beobtained from the results of a longitudinal resonance testas described in BS 1881 : Part 5. In this test the value ofthe fundamental frequency of longitudinal vibration of aprism is determined. The ratio of Ed/p is given by

Ed- = 4”2~2 lo-6 (9)P

where

n is the resonant frequency (in Hz);

L is the length of test specimen (in ml.

Combining equations (8) and (9) we obtain

(1 t u) (1 - 2~) 4n2L2

( l - u ) =-vz

The value of u may be determined from table 3.

nL- UV

Table 3. Values of dynamicPoisson’s ratio

0.257 0.450.342 0.400.395 0.350.431 0.300.456 0.250.474 0.200.487 0.150.494 0.100.499 0.05

(10)

14

BS 1881 : Part 203 : 1986

L

L

The values of modulus of elasticity (both dynamic andstatic), Poisson’s ratio and density all vary from point topoint in a concrete structure.

It is not often possible to carry out resonance tests onstructural members in order to determine the values ofthese properties. It is, however, possible to use empiricalrelationships to estimate the values of the static anddynamic modulus of elasticity from pulse velocitymeasurements made at any point on a structure. Theserelationships are given in table 4 and apply to concretemade with most commonly used types of natural aggregate.The estimate of modulus of elasticity obtained from thistable will have an accuracy better than * 10 %.

Table 4. Empirical relationship between static anddynamic moduli of elasticity and pulse velocity

Pulse velocity

km/s

3.63.84.04.24.44.64.85.0

13 Report

T Modulus of elasticity

Dynamic St&c

MN/m2 MN/m2

24 000 1300026 000 1500029 000 1800032 000 22 ooo36 000 27 00042 000 3400049000 4300058000 52 000

1

The report should affirm that the ultrasonic pulse velocitywas determined in accordance with the recommendations

given in this Part of BS 1881 and should include thefollowing information:

(a) Date, time and place of the investigation.

(b) Description of structure or specimens tested.

(c) Nominal composition of the concrete, including:

(1) cement type;

(2) cement content;

(3) water/cement ratio;

(4) aggregate type and size;

(5) any admixtures used.

(d) Curing conditions, temperature and age of concreteat time of test.

(e) Specification of the environment for which theconcrete was designed.

(f) Sketch showing location of transducer position andpaths of pulse propagation. This sketch should indicatethe details of reinforcing steel or ducts in the vicinity ofthe test areas.

(g) Surface conditions at test points (e.g. smooth,trowelled, rough, presence of surface cracking orspalling as a result of fire damage).

(h) Estimated internal moisture conditions of concreteat time of test and long term curing conditions, whereknown, e.g. wet surface, surface dry (having not longbeen demoulded), or airdry (demoulded in dryingconditions for some time).

(i) Type and make of apparatus, its reading accuracy,pulse frequency and any special characteristics.

(j) Path lengths, methods of measurement and estimatedaccuracy of measurements.

(k) Measured values of pulse velocity.

(I) Pulse velocity values corrected for presence of steelreinforcement where necessary.

15

BS 1881:Part203: 1986

Appendix

Appendix A. Bibliography

1. JON ES, 9. Non-destructive testing of concrete,London: Cambridge University Press, 1962.

2. JONES, R. & FACAOARU, I. Recommendation fortesting concrete by the ultrasonic pulse method. In:RILEM MaterialsandStructures, Paris, 1969, 2, No. 19.

3. DRYSDALE, R.C. Variation of concrete strength inexisting buildings. In: Mag. of Concrete Res., 1973,25, No. 85.

4. DAVIS, S.G. & MARTIN, S.J. The quality of concreteand its variation in structures. C & C A PublicationNo. 42.487, 1973.

5. DAVIS, S.G. Further investigations into the strength ofconcrete in structures. C & C A Publication No.42.514, 1976.

6. KNAB, L.I., BLESSING, G.V. & CLIFTON, J.R.Laboratory evaluation of ultrasonics for crackdetection in concrete. In: AC/ Journal, Jan./Feb. 1983.

7. CHUNG, H.W. An appraisal of ultrasonic pulsetechnique for detecting voids in concrete. In:Concrete, 1978, 12, No. 11.

8. VOLD, R.C. 81 HOPE, B.R. Ultrasonic testing of deepfoundations. In: Brit. J. of Non-destructive Testing,1978,20, No. 5.

9. ELVERY, R.H.& IBRAHIM, L.R.M.Ultrasonicassessment of concrete at early ages. In: Msg. ofConcrete Res., 1976,28, No. 97.

10. LOGOTHETIS, L. & ECONOMOU, C. The influenceof high temperatures on calibration of non-destructivetesting of concrete. In: RILEM MaterialsandStructures, 1981, 14, No. 79.

11. ANDERSON, D.A. & SEALS, R.K. Pulse velocity as apredictor of 28 and 90 day strength. In: ACI Journal,March/April 1981.

12. TOMSETT, H.N. The practical use of ultrasonic pulsevelocity measurements in the assessment of concretequality. In: Mag. of Concrete Res., 1980, 32, No. 110.

13. FACAOARU, I. Non-destructive testing of concrete inRomania. In: Symp. on Non-destructive testing ofConcrete and Timber, June 1969, London: I.C.E., 1970.

14. ELVERY, R.H. & VALE, D.W. Development in themeasurement of ultrasonic pulse velocity in concrete.In: Proc. Seventh lnt. Conf. Nondestructive Testing,Warsaw, June 1973.

15. NWOKOYE, D.N. Prediction and assessment ofconcrete properties from pulse velocity test. In:Mag. of Concrete Res., 1973, 25, No. 82.

16. MULLER-ROCKHOLZ, J. Determination of theelastic oroperties of lightweight aggregate by ultrasonicpulse velocity measurement. In: The internationalJournal of Lightweight Concrete, 1979, 1, No. 2.

17. BUNGEY, J.H. The influence of reinforcement onultrasonic pulse velocity testing. In: ACI In situ/Non-destructive testing of concrete, SP-82, 1984.

18. CHUNG, H.W. 81 LAW, KS. Diagnosing in situconcrete by ultrasonic pulse technique. In: ConcreteInternational Design and Construction, 1983,5, No. 10.

19. REYNOLDS, W.N., WILKINSON, S.S. 81 SPOONER,DC. Ultrasonic wave velocities in concrete. In:Mag. of Concrete Res., 1978,20, No. 104.

20. ROTHING, H. Measuring ultrasonic attenuation inconcrete. In: Building Research and Practice,Jan./Feb. 1975.

16

Publications referred to8s 1881 Testing concrete

Part 5 Methods of testing hardened concrete for other than strengthPart 108 Method for making test cubes from fresh concretePart 109 Method for making test beams from fresh concretePart 110 Method for making test cylinders from fresh concretePart 114 Methods for determination of density of hardened concretePart 116 Method for determination of compressive strength of concrete cubesPart 119 Method for determination of compressive strength using portions of beams broken in flexure (equivalent cube method)Part 120 Method for determination of compressive strength of concrete coresPart 201 Guide to the use of nondestructive methods of test for hardened concretePart 202 Recommendations for surface hardness testing by rebound hammer

*Part 204 Recommendations for the use of electromagnetic cover measuring devices

8.S 3683 Glossary of terms used in nondestructive testingPart 4 Ultrasonic flaw detection

8S 6089 Guide to assessment of concrete strength in existing structures8s 6100 Glossary of building and civil engineering terms

Part 6 Concrete and plaster

*In preparation as the revision of 8s 4408 : Part 1.

BS 1881 : Part 203 : 1986

This British Standard, having been prepared under the direction ofthe Cement, Gypsum, Aggregates end Quarry Products StandardsCommittee, was published under the authority of the Board of BSIand comes into effect on 29 August 1986.

0 British Standards Institution, 1986

lSBNO580 152162

British Standards InstitutionIncorporated by Royal Charter, BSI is the independent nationalbody for the preparation of British Standards. It is the UK memberof the International Organiration for Standardization and UKsponsor of the British National Committee of the InternationalElectrotechnical Commission.

CopyrightUsers of British Standards are reminded that copyright subsists inall BSI publications. No part of this publication may be reproducedin any form without the prior permission in writing of BSI. Thisdoes not preclude the free use, in the course of implementing thestandard, of necessary details such as symbols and size, type orgrade designations. Enquiries should be addressed to thePublications Manager, British Standards Institution, Linford Wood,

Milton Keynes MK14 6LE. The number for telephone enquiries is0908 320033 and for telex 825777.

Contract requirrmentsA British Standard does not purport to include all the necessaryprovisions of a contract. Users of British Standards are responsiblefor their correct application.

Ravision of British StandardsBritish Standards are revised, when necessary, by the issue eitherof amendments or of revised editions. It is important that users ofBritish Standards should ascartain that thay ara in possession ofthe latest amandmonts or aditions. Information on all BSIpublications is in the BSI Catalogue, supplemented each month byBSI News which is available to subscribing members of theInstitution and gives details of new publications, revisions,amendments and withdrawn standards. Any person who, whenmaking use of a British Standard, encounters an inaccuracy orambiguity, is requested to notify BSI without delay in order thatthe matter may be investigated and appropriate action taken.

The following BSI references relate to the work on this standard:Committee reference CAB/4 Draft for comment 84/13382 DC

Committees responsible for this British StandardThe preparation of this British Standard was entrusted by theCement, Gypsum, Aggregates and Quarry Products StandardsCommittee (CAB/-) to Technical Committee CAB/4 upon whichthe following bodies were represented:

Association of Lightweight Aggregate ManufacturersAssociation of Metropolitan AuthoritiesBritish Aggregate Construction Materials IndustriesBritish Civil Engineering Test Equipment Manufacturers’ AssociationBritish Precast Concrete FederationBritish Ready Mixed Concrete AssociationBuilding Employers’ ConfederationCement Admixtures AssociationCement and Concrete AssociationCement Makers’ FederationConcrete SocietyCounty Surveyor’s SocietyDepartment of the Environment (Building Research Establishment)Department of the Environment (Property Services Agency)Department of Transport (Highways)

Department of Transport (Transport and Road ResearchLaboratory 1

Electricity Supply Industry in England and WalesFederation of Civil Engineering ContractorsGreater London CouncilInstitute of Concrete TechnologyInstitution of Civil EngineersInstitution of Highways and TransportationInstitution of Structural EngineersInstitution of Water Engineers and ScientistsRoyal Institute of British ArchitectsRoyal Institution of Chartered SurveyorsSand and Gravel AssociationSociety of Chemical Industry

The following bodies ware also represented in the drafting of thestandard, through subcommittees and panels:

British Nuclear Fuels LimitedUnited Kingdom Atomic Energy Authority

Amendments issued since publication

Amd. No. Date of issue Text affected

British Standards Institution . 2 Park Street London WlA 2BS . Telephone 01-629 9000 . Telex 266933

8608 -7 CAB14