1129215501isrm Sm Blast Vibration Monitoring - 1992

Int. J. Rock Mech. Min. Sci. & Geomech. Abstr. Vol. 29, No. 2, pp. 143-156, 1992 0148-9062/92 $5.00 + 0.00 Printed in Great Britain. All rights reserved Cop~r:ght © 1992 Pergamon Press Ltd

I N T E R N A T I O N A L S O C I E T Y F O R R O C K M E C H A N I C S

C O M M I S S I O N O N T E S T I N G M E T H O D S

S U G G E S T E D M E T H O D FOR BLAST V I B R A T I O N M O N I T O R I N G

C O N T E N T S

Introduction ........................................................................................................................... 144

Scope ..................................................................................................................................... 145

Character of Blast Excitation ................................................................................................. 145

Measurement Techniques and Instruments .............................................................................. 149

Evaluation of Measurements ................................................................................................... 152

References .............................................................................................................................. 156

Appendix: Permanent Degradation and Displacement of Adjacent Rock ............................. 156

Coordinator C. H. Dowding (U.S.A.)

143

144 ISRM: BLAST VIBRATION MONITORING SUGGF:-TED METHOD

INTRODUCTION

The President of the Commission on Testing Methods appointe?, the Coordinator to organize a Working Group to draft a Suggested Method for blast vibration monitoring on I December 1988. Since that appointment, the working group has reviewed three successivel? narrowed guidelines dated Spring 1989, Summer 1990 and Spring 1991. This guideline fail~ under Category II: Engineering Design Tests. within the In S i tu Group, I tem 8 of Table 1. The purpose of this method is to specify procedures, and to achieve some degree of standardization without inhibiting the development or improvement of techniques.

Any person interested in these recommendations and wishing to suggest additions or modifications should address his remarks to the Secretary General, International Society for Rock Mechanics, Lab6ratorio Nacional de Engenharia Civil, Avenida do Brasil, Lisboa 5, Portugal.

Table 1. Test categories--priority order for standardization ~

Category I: Classification and Characterization

Rock material (laboratory tests): (I) Density, water content, porosity, absorption (2) Strength and deformability in uniaxial compression: point load

strength (3) Anisotropy indices (4) Hardness, abrasiveness, attrition, driUability (5) Permeability (6) Swelling and slake-durability (7) Sound velocity (8) Micro-petrographic descriptions

Rock mass (field obserrations ): (9) Joint systems: orientation, spacing, openness, roughness, ge-

ometry, filling and alteration (10) Core recovery, rock quality designation and fracture spacing (I 1) Seismic tests for mapping and as a rock quality index (12) Geophysical logging of boreholes

Category II: Engineering Design Tests

Laboratoo': (1) Determination of strength envelope and elastic properties

(triaxial and uniaxial compression: tensile tests) Direct shear tests Time-dependent and plastic properties

(2) (3)

In situ: (4) (5) (6)

Deformability tests Direct shear tests Field permeability, ground-water pressure and flow monitor- ing: water sampling

(7) Rock stress determination (8) Monitoring of rock movements, support pressures, anchor

loads, rock noise and vibrations (9) Uniaxial, biaxial and triaxial compressive strength

(I0) Rock anchor testing

aThis Table will be superseded as the Commission updated the priorities--J. A. Hudson, Commission President and Journal Editor.

ACKNOWLEDGEMENTS

C. H. Dowding (U.S.A.) coordinated the working group and prepared the 1st, 2nd and 3rd drafts. Extensive written comments were received from B. New (U.K.), F. Ouchterlony (Sweden), D. Siskind (U.S.A.), K. Sassa (Japan). Written comments were received from J. Esteves (Portugal), E. Fernandez (Spain), O. Mueller (Hungary), A. Ghose (India). Copies of the drafts have been sent to D. Beitzer (Fed. Rep. Germany), J. Brinkman (South Africa), P. Calder (Canada), T. Li and T. Xu (P.R. China), V. Rosai and M. Lern (Mexico), and A. Schwenzfeier (France).

ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD 145

Suggested Method for Blast Vibration Monitoring

Scope

1. This guideline is separated into three main sections. The first section, Character of Blast Excitation, defines the terminology necessary to describe blasting vibrations and the associated air over pressure. Importance of dominant frequencies of excitation and structural re- sponse is introduced here and is emphasized throughout the document. The second section, Measurement Tech- niques and Instrumentation, describes generic attributes of instruments necessary to measure time histories of the blast-induced disturbances. Special emphasis is placed on computerized systems. Guidance is given for the choice and deployment of instruments, both at the beginning and continuation of a project. The third section, Evaluation of Measurements, presents defi- nitions of structural response. Explanation is given for the need of studies with immediate pre- and post-blast inspection to separate weather- and blast-induced re- sponse. Guidance is also given for monitoring response of rock masses and buried structures.

(a) While the subject of this guideline is the measure- ment of blast-induced, transient or vibratory displace- ment, effects of blast-induced permanent displacements are included in the Appendix for completeness as they are associated with significant transient effects at rela- tively small distances. Whenever vibration response is a legitimate concern, these permanent displacements can be more important than the vibrations.

(b) Transient effects result from the vibratory nature of the ground and airborne disturbances that propagate outward from a blast. In this discussion, it is assumed that no permanent displacements are produced in or on the rock or soil mass surrounding the blast. Thus the only effects are those associated with the vibratory response of facilitities. Transient means that the peak displacement is only temporary (i.e. lasts less than one-tenth of a second) and the structure or rock mass returns to its original position.

(c) This document implicitly separates measurement of vibration to control cosmetic cracking from that to reduce human response by presenting only studies of blast-induced cosmetic cracking. Differing cultures have differing thresholds of the toleration of vibration. Some have so little that urban blasting is prohibited altogether. Others have a great deal more than the regularly allowed 5 cm/sec maximum particle velocity at high excitation frequencies. Since it is unlikely that the physics of cracking changes at national borders, these national variations are certainly influ- enced by several factors in addition to the crack suscep- tibility of structures. Additional factors such as human response as well as administrative and political expedi- ency must be recognized as separable from cracking in the measurement and evaluation of ground motions.

Character of Blast Excitation

2. As shown in Fig. 1, both the ground and airborne disturbances (upper-four time histories) produce struc- ture response (lower-four time histories). Because of the importance of excitation frequency in determining this structural response, the full waveform or time history of the motions should be recorded. When a critical location in a structure is known, blast response is best described by measurement of the strain at that location. Alterna- tively, excitation particle velocity (that shown in Fig. l) can be measured outside the structure of concern (U.S.) or on the structure's foundation (Europe); however, many recent cracking studies have correlated visual observations of cracking with excitation particle velocity measured in the ground.

GROUND MOTION

3. Ground motion can be described by three mutually perpendicular components labelled L (longitudinal), T

(transverse) and V (vertical) in Fig. 1. The L and T directions are oriented in the horizontal plane with L directed along the line between the blast and recording transducer. When a study focuses upon structural re- sponse, axes can be labelled HI, H2 and V, with H 1 and H2 oriented parallel to the structure's principal axes.

(a) Variation of peak motions in each component (L, V and T in Fig. 1) has led to difficulty in determination of the most important. Horizontal motions seem to control the horizontal response of walls and superstruc- tures, and vertical motions seem to control the vertical response of floors. In an absolute sense, the peak ground motion and thus ground strain is the maximum vector sum of the three components, which usually occurs at the largest peak of the three components, the dashed line in Fig. 1. This true maximum vector sum is not the pseudo-maximum vector sum calculated with the max- ima for each component (dots in Fig. 1) no matter their time of occurrence. The pseudo-maximum vector sum

146 ISRM: BLAST VIBRATION M O N I T O R I N G S U G G E S T E D M E T H O D

Long.

Trans.

VerL

Air

dl(L)

dl(T)

d,(L)

maximum maximum ground motion superstructure resp.

I •

, , , . V v v , ~ v v u V v v ~ - A ''u" 9 ,¢' B . . . . ~ , ^ a , A A / ~ ,,,,, r~ ^ . - ' - v , t r ' v v V v , _ v " ' v

p^,,.a..~v p . . . ^ _ A ~ ^ ^ vV . . . . r . . . . ~ " V V "" ~ ,.r

, . . . . . . . . . . . ^ M^A, ' - - ' , . . . . . yv"

I

._^A^A^^ ^AA,^~5.,, ~ t ~ - ~ ^ ^ . h A . - ' v v ' T v "" v ' v " v ~ w " " v ' , v v v"

.~ , . . . . 1 , . , . . - . . ~ . , ^ ~ ^ ^ = ItlWln/!VVV"v'V v v V ~ - - v ---

v,

~ , - - y v - v v vV [ I I V V ~ v "-" I I

1/fa

3.20 mm/s

4.14

2.69

109 dB

excRatlon Ops)

velocity

8.18 MID

WALL 7.26

9.17 SUPER

STRUCTURE 9.24

(15-25 Hz)

(6-7 Hz)

relponl4 011a)

velc¢~

Time

Fig. 1. Comparison of blast excitation by ground and air-borne disturbances and residential structure response of walls and superstructure. Measurements were made some 2000 ft (600 m) from a typical surface coal mining blast (after Dowding [10]).

may be as much as 40% greater than the true maximum vector sum, which is normally 5-10% greater than the maximum, single-component peak.

(b) In general, experimental observations of threshold or cosmetic cracking, which form the basis of blasting controls in North America, have been correlated with the maximum single component regardless of direction. Therefore, use of the pseudo-maximum vector sum for control provides a large, unaccounted for, factor of safety.

(c) Two principal wave types are produced by blast- ing, body (P/S) surface (R) and are illustrated by the ground motion in Fig. 1 measured some 600 m from a typical surface coal mining blast. Body waves travel through earth materials, whereas surface waves travel close to surfaces and interfaces of earth materials. The most important surface wave is the Rayleigh wave, denoted R on the vertical trace in Fig. 1. Body waves can be further subdivided into compressive (compression/ tension) or sound-like waves, and distortional or shear waves, denoted as P/S on the vertical trace in Fig. 1. Explosions produce predominantly body waves at small distances which propagate outward in a spherical manner until they intersect a boundary such as another rock layer, soil or the ground surface. At this intersec- tion, shear and surface waves are produced. Rayleigh surface waves become important at larger transmission distances as illustrated in the vertical trace by the

relatively larger " R " amplitude compared to the "P/S" amplitude.

TRANSIENT NATURE OF BLAST MOTIONS

4. Great care should be taken not to confuse the effects of steady-state, single-frequency, harmonic motions with those of transient, irregular blast motions. As can be seen in Fig. I, the maxima of blast-induced motions last only one or two cycles at a relatively constant amplitude and frequency. Thus they are not continuous (last many cycles) or steady-state (have constant frequency and amplitude).

S I N U S O I D A L A P P R O X I M A T I O N

5. Typical blast vibrations, no matter the wave type, can be approximated as sinusoidally varying in either time or distance as shown by the time variations in Fig. 2a and b. This approximation is similar to the motion of a cork caused by a passing water wave. Displacement of the cork from its at-rest position is similar to the displacement u of a particle in the ground from its at-rest position. Similarly, the cork's velocity, as it bobs up and down ti is analogous to that of a particle in the ground, hence the term particle velocity. The wave shape that excites the cork can be described by its wavelength 2, the distance between wave crests; the wave

Time ( t ) ~j Distonce ( x )

Fig. 2. Sinusoidal approximations: (a) sinusoidal displacement at a fixed point (x = constant); and (b) sinusoidal displacement at one instant (t = constant) [1].

ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD 147

speed or propagation velocity c at which the wave travels o , past the cork; and the frequency f, or the number of times the cork bobs up and down in 1 sec. Frequency f o 3 is equal to l I T or the reciprocal of the period or time it takes the cork to complete one cycle of motion, o2 Frequency is measured in cycles per second or Hertz, Hz. o~

(a) The general form for the sinusoidal approxi- mation is best understood by beginning with the oo equation for sinusoidai displacement u, when there is a ~ 03 single dominant frequency:

o 0 .2 u = U s i n ( 2 n f t ) (1 ) o

.=

where U is maximum displacement, f is frequency and t ~ o 1 tZ is t i m e a n d

Urea x ---- U

f~m~ = U 2 n f = 2nfUm,

t2m~ = U4n2f' = 27rfZim~

(max displacement),

(max particle velocity),

(max acceleration).

0.0

0 3

(2) o.z

Usually, acceleration is measured in units of gravita- tional acceleration, where g = 9814 mm/sec-'. Therefore, an acceleration of 2000 mm/sec' is:

2000 9814 = 0.2g,

or two-tenths that of gravity. (b) Kinematic relations between particle displace-

ment, velocity and acceleration for complex waveforms are exactly related through integration or differentiation of any of the waveforms. For instance, an acceleration time history can be integrated once for a velocity time history, which in turn can be integrated for a displace- ment time history. Even though a particle velocity record can be differentiated to find acceleration, it is not recommended, as the procedure is sensitive to small changes in the slope of the velocity time history. Further discussion of the inaccuracies of differentiation and integration can be found in Dowding [1] and in texts devoted to interpretation of time histories (e.g. [2]).

ESTIMATION OF DOMINANT FREQUENCY

6. Adoption of frequency-based vibration criteria has made the estimation and calculation of the dominant or principal frequency an important concern. Dominant frequency can be estimated through: (1) visual inspection of the time history or calculated with (2) response spectra or (3) Fourier frequency spectra.

(a) The accuracy or difficulty of visually estimating the dominant frequency depends upon the complexity of the time history. The type of time history record with the most easily estimated dominant frequency is one with a single dominant pulse like that shown in the inset in Fig. 3. The dominant frequency of a single pulse is the inverse of twice the time interval of the two zero crossings on either side of the peak.

(b) The most difficult type of record to interpret is that which contains nearly equal peaks at two dominant frequencies such as that in Fig. 1. The two dominant

Surface co(~l mine bla~n(~

L f = principal frequency

: 1.~ t 1

Quarry blasting

r - I I I

Construction blastin9

o .o 20 40 6 0 a o l o o 12o

P r i n c i p a l f r e q u e n c y ( H z )

Fig. 3. Dominant frequency histograms at nearest structures catego- rized by industry. Dominant frequency is defined in the inset (after

Siskind et aL [15]).

frequencies are the initial 15-20 Hz portion (peak A) and the later 5-10 Hz portion (peak B). As can be seen in the figure, the initial portion produces the highest wall response while the second produces the greatest super- structure response. For the best frequency correlation of both types of response, both frequencies should be calculated.

(c) The best computational approach to determining the dominant frequency involves the response spectrum. The response spectrum is preferred over the Fourier frequency spectrum because it can be related to struc- tural displacement and thus strains [1]. A compromise approach is to calculate the dominant frequency associ- ated with each major peak by the zero crossing approach described above.

(d) Since many time histories do not contain as broad a range of dominant frequenices as that in Fig. 1, most approaches require only the calculation of the frequency associated with the maximum particle velocity for blasts that produce low particle velocities. The more complex frequency analyses need to be employed only when peak particle velocities approach control limits.

(e) As shown in Fig. 3, the relatively large explosions produced by surface coal mining, when monitored at typically distant structures, tend to produce vibrations with lower dominant frequencies than those of construc- tion blasts. Construction blasts involve smaller ex- plosions, but the typically small distances between a structure and a blast as well as rock-to-rock transmission paths tend to produce the highest dominant frequencies [3]. Such high-frequency motions associated with con-

148 ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD

struction blasts have less potential for cracking adjacent structures than do lower frequency mining blasts [1].

PROPAGATION EFFECTS

7. Ground motions decrease in amplitude with in- creasing distance. Effects of constructive and destructive interference and geology are included within the scatter of data about the mean trend of the decay in amplitude with distance. While this scatter is large, the associated decay with distance is observed in all blast-vibration studies. Typical examples of this decay are shown in Fig. 4 where maximum particle velocity is plotted as a function of square-root scaled distance from the blast.

(a) Square-root scaling, or plotting peak particle vel- ocity as a function of the distance R, divided by the square root of the charge weight R / W ~/2, is more tra- ditional than the cube-root scaling, which incorporates energy considerations [4]. Both square or cube-root scaling can be employed to compare field data and to predict the attenuation or decay of peak particle vel- ocity; however, square-root scaling is more popular. Site specific scaling is sometimes employed where scaled distance takes the form of R / W " , where n is determined empirically by curve fitting [3].

(b) Several square-root attenuation relations em- ployed in the U.S are shown in Fig. 4. They are banded to reflect scatter, which is typical of blasting operations. Curve P should be used for presplitting, cratering and beginning new bench levels. It is also the basis for the U.S. Office of Surface Mining (OSM) regulations for conservative shot design when monitoring instruments are not employed.

1.000

0.500

c

> O . t O O

0.050

0,010

O.OOS t0

10

R/W "~( m/kg w2)

t00

100

sn

E E

100 1000

Square roo't scaled distance R/W~/Z(ft/ Ib ~'~)

Fig. 4. Attenuation relations showing scatter from geological and blast design effects as well as high expected velocities from confined shots,

such as presplitting (after Siskind et al. [15]).

(c) Dominant frequencies also tend to decline with increasing distance and with increasing importance of surface waves. At larger distances typical for mining, higher frequency body wa~es begin to have relatively lower peak amplitudes than the lower frequency surface waves, as shown in Fig. 1. Since lower frequencies can elicit greater structural response [5] as shown in Fig. !, OSM scaled-distance limits decline with increasing ab- solute distance.

BLAST-INDUCED AIR OVER-PRESSURES

8. Although technically airborne disturbances are not directly related to ground motion, these air over-press- ures generated by blasting intensify human response and thus need to be documented. Previous researchers have found that response noise within a structure (from blasting and sonic booms, respectively) is the source of many complaints. The audible portion of the over-press- ure produces direct noise, while the less audible portion by itself or in combination with ground motion can produce structural motions that in turn produce noise. Over-pressure may crack windows; however, it would have to be unusually high for such cracking.

(a) Just as with ground motions, blast-induced air over-pressure waves can be described with time histories as shown in Fig. 1. The higher frequency portion of the pressure wave is audible sound. While the lower fre- quency portion is less audible, it excites structures, which in turn causes a secondary and audible rattle within the structure and is the source of many complaints. The air-blast excitation of the walls can be seen by comparing air-blast excitation and wall response in the rightmost portion of the time histories in Fig. 1 where there is no ground motion. Unlike ground motions, air over-press- ure can be described completely with only one trans- ducer, since at any one point air pressure is equal in all three orthogonal directions.

(b) Propagation of blast-induced air over-pressures has been studied by numerous investigators and is generally reported with cube-root rather than square- root scaled distances. Peak pressures are reported in terms of decibels, which are defined as:

d B = 20 logt0 , (3)

where P is the measured peak sound pressure and P0 is a reference pressure of 2.9 x 10 -9 lb/in.2 [20 x 10 -~ (P,)].

(c) Figure 5 summarizes the effect of two important instrumentation and shot variables. First, the effect of the weighting scales is dramatically evident. "C" weight- ing greatly reduces the recorded peak pressure at any scaled distance. This does not mean the peak is reduced by changing instruments, but rather that the "C" weight- ing system does not respond to the low-frequency press- ure pulses. These low-frequency pressure peaks excite structures and occupants whether or not they are sensed by the measurement instruments. The other (5 and 0.1 Hz) labels denote the lower-frequency bounds of the recording capabilities of these so-called "linear" systems.

ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD 149

(d) Second, the effect of gas venting caused by in- adequate stemming in shot holes can be observed in Fig. 5 from the higher average pressures produced by the parting shots at any scaled distance. Parting shots are detonated in thin rock layers between coal strata in surface mines. Consequently, there is less hole height available for stemming, and these shots many times eject the stemming and thereby produce abnormally high air over-pressures. The unconfined relation should be used for demolition of structures after modification for effects of weather and ground reflection.

(e) An air temperature inversion causes the sound

pressure wave to be refracted back to the ground and at times to be amplified in isolated locations about 16-acres in size. Such an inversion occurs when the no ,rm, al decrease in temperature with altitude is reversed because of the presence of a warmer upper layer. Schomer et al. [6] have shown that for propagation distances of 3--60 kin, inversions produce zones of intensification of up to three times the average, attenuated or low air over-pressures at those distances, with an average in- crease of 1.8 times (5.1 dB). At distances less than 3 km, where high air over-pressures are likely to occur, his measurements show no inversion effects.

Measurement Techniques and Instruments

9. This section describes characteristics of instruments that measure the ground motions (acceleration, velocity, displacement) and air blast (air over-pressure). Since there are many excellent sources for information on instruments, the principal characteristics of available systems will be summarized rather than exhaustively reviewed. The most complete single reference for de- tailed instrumentation information that is updated periodically is the Shock and Vibration Handbook [7]. Specific information on blast vibration monitors is con- tained in recent publications by the U.S. Bureau of Mines and the OSM (i.e. [8]).

(a) An idealized, field-portable blast monitoring sys- tem operating on a 12 V battery is illustrated in Fig. 6. It consists of transducers (1) that convert physical motion or pressure to an electrical current, which is transmitted

R/WV3(rn/kc~ *s )

10 100 1,000 10"I 1 ~ I t51

\

141 \

10 "z 131 - s , . ~ O . l . , X \ \

','41 \

, x\ r.J, ' \\ .

-~ - - 101 I .~_ ,~ \ C- slow

10"" C- s l o w ~ I \ \ 9'1

HighwoII 81 PaRing

Unconfined

10 "~ I I t t t l l l l i I I I t l t t l i i i I l i t

10 100 1,000 10,000

Cube roo1~ scolecl ¢listonce, R/W'/~(ft,/Ib I/3) Fig. 5. Attenuation relations for air over-pressures produced by confined (highwall) and partially-confined (parting) surface coal min-

ing blasts as well as unconfined blasts [29].

through cables (2) to an amplifying system (3); and a magnetic tape, paper or computer digital recorder (4) that preserves the relative time variation of the original signal for eventual permanent, hard-copy reproduction by a pen recorder light-beam galvanometric recorder or dot matrix printer (5). There is an almost endless variety of configurations of these five basic components. How- ever, the best involve microprocessors (computers) for data acquisition, storage and reproduction.

(b) While particle velocity is the traditional measure- ment of choice, structural strains control cracking. They should be measured directly from relative displacements on structures or within rock masses when critical lo- cations are known (i.e. pipelines and unusual opening geometry) and can be obtained with a variety of strain and relative displacement gauges [9, 10]. Unfortunately, these critical locations may be either unknown or too many in number to economically measure. Therefore some means of estimation is necessary.

(c) Ground motion and air over-pressure time histories can be employed to calculate the relative displacement of structural components with a knowl- edge of the responding structure's dynamic response characteristics [l]. These relative displacements can in turn be employed to calculate strains. The accuracy of these estimates is limited by the degree to which the structure behaves as a single-degree-of-freedom system and the accuracy of the estimate of the dynamic response characteristics.

®

® ®

q Velocity (3 orthogonal) ond sound pressure transducers

2 Cobles 3 Amplifier 4 Recorder (tope, disk or memory) 5 Light beam oscilloSCope or dot maitrlx printer

Fig. 6. Idealized, field-portable, blast monitoring system that shows the schematic relation of the five principal components [1].

150 ISRM: BLAST VIBRATION MONITORING SLGGESTED METHOD

APPROPRIATE MEASUREMENT OF PARTICLE VELOCITY

10. While any of the three kinematic descriptors (displacement, velocity or acceleration) could be em- ployed to describe ground motion, particle velocity is the most preferable. It has the best correlation with scientific observation of blast-induced cracking, which forms the basis of vibration control. Furthermore, it can be integrated to calculate displacement. If acceleration is desired, it should be measured directly to avoid differen- tiation of the particle velocity time history. Integration after vectoral addition of components should be con- ducted only after possible phase shifts have been taken into account.

(a) The location for measurement varies throughout the world. In North America, the excitation or ground motion is measured on the ground adjacent to the structure of interest. In Europe, the excitation motion is measured on the structure's foundation. The difference stems from historical precedent and location of trans- ducers during scientific observation of cracking rather than difference in philosophy. In North America, many times it is impossible to place transducers on adjacent property owned by a party not involved in the project. Furthermore, if it is desired to describe the excitation motions, then those motions should be measured outside of and not on the structure. If it is desired to measure structural response motions,, then they should be measured on the most responsive structural members, which are not the basement or foundation walls because of the restraint provided by the ground.

(b) Time histories of the three components of motion should be measured because of the importance of exci- tation frequency. Recording only the magnitudes of peak motions will not yield information about the dominant frequency and time history details that control structural response and rock mass strains. Peak motions and dominant frequency can be employed to describe low-level, non-critical motions. Therefore machines em- ployed to monitor critical motions (Type I below) should be capable of recording time histories of selected critical motions. Machines that record only peak motions (Type II below) can be employed with those that record time histories to provide redundant measurement where fre- quency content does not vary widely and where particle velocity is low.

TRANSDUCER RESPONSE FREQUENCY

11. Frequency response is the frequency range over which the transducer's electrical output is constant with a constant mechanical motion. This constancy is nor- mally expressed in terms of decibels (dB). For instance, linear within 3 dB between 2 and 200 Hz means that the transducer produces a voltage output that is constant within 300 between 2 and 200 Hz. Generally, it is better to request a transducer's response spectrum (such as those shown in Fig. 7) to determine the frequencies where this difference occurs. Many manufacturers of

4O

30

20

= ~0

> ~ 6

0 3

4 # : o , r

2

I I I B I I I I I I I I I I I 0 2 0 .4 0.6 ~ 2 ~, 4 6 e 10 20 a o

o 3 0.8

Frequency (Hz)

Fig. 7. Example response spectra of a velocity transducer with differ- ing pereenta~s of damping. With 70% of critical damping this system

is × 3dB ( × 30%) down 1 Hz [I].

blast monitors are electronically amplifying transducer output at low-frequency excitation to allow use of smaller, high-frequency transducers. Instruments with such electronic amplification should be physically cali- brated as described below.

(a) Proper frequency response for blast vibration transducers is dependent upon two considerations: measurement of the "true" phenomena, and efficient measurement of important characteristics. Unfortu- nately the entire range of frequencies necessary to de- scribe true blast phenomena is too large for any one transducer. Blast-induced delayed gas pressure pulses occur at frequencies of less than 1 Hz, and close-in accelerations have been measured above 1000 Hz. There- fore it is necessary to compromise the goal of defining the true phenomena when only one transducer type is employed, and the optimum choice is dependent upon the important motion characteristics.

(b) Monitoring ground motion to control cosmetic cracking in low-rise structures is typically accomplished by measurement of ground, or particle, velocity over a frequency range of 2-200 Hz. This range ensures proper recording of amplitudes at excitation frequencies which: (1) encompass fundamental frequencies of structures; and (2) are associated with the peak velocity that produces the greatest response displacement (i.e. are dominant). Typical structure fundamental frequencies are 5-10Hz for two- and one-storey structures and 10--30 Hz for walls and floors. Some mechanical systems may have fundamental frequencies near 100 Hz, but they are usually attached to and excited by the lower fre- quency walls and floors. Typical dominant excitation frequencies range from 5 to 100 Hz as shown in Fig. 3. If it becomes necessary to monitor situations with un- usually low or high dominant frequencies, special trans- ducers should be employed that are linear in the range of interest [30].

TRANSDUCER ATTACHMENT

12. One of the most critical aspects of vibration monitoring is the mounting of the transducers in the field. The importance of mounting is a function of the

ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD 151

particle acceleration of the wave train being monitored. The type of mounting on a horizontal surface is the least critical when the vertical maximum particle accelerations are less than 0.2g. In this range, the possibilities of rocking the transducer or the transducer package are small, and the transducer may be placed upon a horizon- tal measurement surface without a device to supply a holding force. When the maximum particle accelerations fall between 0.2 and 1.0g, the transducer or transducer package should be buried completely when the measure- ment surface consists of soil [11]. Mounting of transduc- ers on spikes in soil is discouraged because the free response of the mounting system may effect the recorded motion. When the measurement surface consists of rock, asphalt or concrete the transducers should be fastened to the measurement surface with either double-sided tape, epoxy or quick-setting cement (hydrocal or other gyp- sum based cements set within 15-30 min). If the above methods are unsatisfactory or accelerations exceed 1.0g, only cement or bolts are sufficient to hold the transducer to a hard surface. All transducers mounted on vertical surfaces should be bolted in place.

(a) Air over-pressure transducers should be placed at least 1 m above ground, pointed downward (to prevent rain damage) and covered with a wind screen to reduce wind excitation-induced false events.

DIGITAL, TAPE AND HARD-COPY RECORDERS

13. Microprocessor (computer) or digital recording systems now dominate new sales of technical recording devices because of the ease of data acquisition and computer linkage. The signal is sampled at a certain rate, say, 1000 times/sec, and each sample is converted to a single magnitude. Digital recording has several advan- tages. It is very accurate, as variation in tape speed has no effect if cassette tapes are employed as the storage medium, and records can be directly accessed by a computer. Details of the digitization process are dis- cussed elsewhere [1].

(a) Of those blast-monitoring systems with tape recorders, most employ compact FM cassettes, but the best employ digital recording techniques. Many of the tape systems involve separate record and reproduction modules to reduce the complexity of recording. Care should be exercised to determine the exact details of the system before purchasing, as tape recorder performance varies at low temperature.

(b) A permanent record or "hard copy" of the vi- bration time history is usually made on photographic film, floppy disc, battery-powered memory chips or paper. Almost all present film-based recorders employ field-developable, u.v. light-sensitive paper in combi- nation with light-beam galvanometers to record high- frequency motions. The newest generation recorders employ dot matrix printers and/or floppy discs with microcomputers. Unfortunately, those that automati- cally print after a vibration event may not be recording another event while printing. If multiple shots are likely, this reset time should be determined. Furthermore,

RMMS 29/2--E

printer behaviour in cold weather is variable and should also be investigated.

(c) Most recorders can be bought as either single- or multichannel units. A four-channel unit is necessary in blast monitoring to record simultaneously the three components of the ground motion (L, V and T) and the air blast. The present trend in vibration equipment is to include a signal-conditioning amplifier in the recorder to allow flexible amplification of the signals.

(d) Frequency analysis of records requires a time history and thus some form of permanent record. Instru- ments recording only peak particle velocities will not allow a frequency analysis. Sending permanent records through the mail for interpretation, results in a delay of 5 days, and sometimes up to 1 month. Systems with light-sensitive paper or dot matrix printers allow im- mediate interpretation of frequency without additional costly equipment.

CALIBRATION

14. It is obvious that the entire vibration measure- ment system should be calibrated, as it is futile to record data if they cannot be exploited because of a lack of reference. Manufacturers supply calibration curves with their instruments that are similar to the response spectra for transducers shown in Fig. 7. Recalibration or check- ing requires special platforms where frequency and displacement are controlled, and in the field, a calibrat- ing circuit to pulse the magnetic core of the geophone [12].

NUMBER OF INSTRUMENTS

15. While the smallest number of instruments or triaxial transducer locations for recording blast exci- tation motions is one, two triaxial positions would provide a more thorough documentation of the spatial distribution of effects. If only one instrument is em- ployed, then it should be located at the nearest or most critical receiver. This single, Type I instrument should record time histories of the three axes of particle velocity as well as air over-pressure. Since it must monitor continuously, it must trigger (begin recording) automati- cally, and be capable of monitoring even while printing or communicating results. When blasting will occur at more than one general location (i.e. involve different nearest structures separated by hundreds of metres), then two and four are the smallest and optimum number of instruments, respectively. A third and fifth should be available but not deployed to insure continuous cover- age in case of instrument failure.

(a) The second and fourth instruments in the situ- ations described above may provide a lower level of information and will be termed Type II. They must at least continuously record the peak particle velocity in one axis and may or may not measure air over-pressure. The best axis is the vertical, since no horizontal direction decision is required and surface waves usually involve a significant vertical component regardless of the direction of the maximum horizontal component. These

152 ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD

instruments should be located at a greater distance than the nearest structure to monitor a large area.

(b) The third or spare instrument can be either Type I or II. Where air over-pressures will be problematic or frequencies critical, the spare should be Type I. This spare instrument can also be employed to monitor sites where complaints develop. Such public relations moni- toring of vibrations at locations associated with com- plaints is essential in North America where lawsuits arise even when all blast effects comply with regulatory guide- lines.

(c) The above approach describes the least number of instruments. Applicable regulations and mining or con- struction schedules may require a larger number. Measurement of structural response (in addition to excitation) may require more instruments; however, control limits are based upon excitation and not re- sponse motions.

INSTRUMENT DEPLOYMENT DURING TEST BLASTS

16. When blasting projects begin, when geological conditions change radically or when new initiation sys- tems are introduced, test blasts should be conducted to minimize the number of instruments necessary to moni- tor production blasts. Instrument locations should be chosen to produce project-specific attenuation relations for both air over-pressure and ground motion. Such relations vary from project to project because of changes in geology and blasting practices. Additionally, the test blasts allow the determination of the frequency content of motions at different scaled and absolute distances.

(a) The attenuation relation is not solely a site prop- erty. Although it is dependent upon geology, it is also

heavily dependent upon the blast geometry and timing. For instance, wi:h the s:lme weight of explosive deto- nated at any ins:..r,t of time, a blast with a larger burden will produce :~.:: .'.::nuation relation parallel to that in Fig. 4 but wi~.h a larger intercept on the velocity axis. Furthermore. differing initiation timing will produce changes in the time history, both length and frequency content.

(b) During tes: blasts, a minimum of four instruments should be deployed to measure peak particle velocity along a single azimuthal direction at widely differing scaled distances for the same blast. Therefore, for any one blast design, parameters and initiation sequences are constant, and the resulting attenuation relation shows only the effect of distance, direction and/or geology. Seismographs and/or transducers should be placed along a single line with constant geology to determine best the attenuation relation, or at all critical structures to deter- mine the effects of direction and variable geology. Ide- ally, the linear orientation should be along a path with constant thickness of soil and not cross any large geologic discontinuities such as faults. If geology changes radically, then two such attenuation lines are necessary, but not necessarily with each blast.

(c) A number of approaches to blast design for vi- bration control are now available that employ a single- delay, single-hole test blast and a number of instruments to record the attenuation and frequency change around the site [13]. These single-time histories are then syn- thesized to reproduce the additive time history effects of multiple delay, multiple hole blasts at the differing instrument sites. Such synthesis of time histories to guide blast design has met with variable success but does not replace monitoring of blast effects at critical structures during production blasting.

Evaluation of

17. Direct regulation or specification of effects, rather than specification of blast design, is the most effective control from a regulatory viewpoint because effects are so dependent upon details of the shot geometry and initiation sequence. Such dependency renders control impossible by simple regulatory specification of two or three design parameters. For instance, consider control by specification of the maximum charge weight deto- nated per instant at given distances from the nearest structure. Even with such detailed specification, intended vibration levels at the structure may be exceeded because of poor choice in the location of holes and/or their relative time of initiation.

DEFINITIONS OF STRUCTURAL RESPONSE

18. Excessive structural response has been separated into three categories arranged below in the order of declining severity and increasing distance of occurrence [14, 15]. MAJOR--permanent distortion---occurs only at very high particle velocities and results in serious weakening of the structure (e.g. large cracks or shifting

Measurements

of foundations or bearing walls, major settlement result- ing in distortion and non-vertical walls). MINOR-- displaced cracks--includes surfacial cracking which does not affect the strength of the structures (e.g. broken windows, loosened or fallen plaster), hairline cracks in masonry. THRESHOLD--cosmetic cracking--occurs at the lowest velocities and only opens old cracks or produces hairline cracks in plaster walls or may dislodge loose objects (e.g. loose bricks in chimneys). Description of these responses collectively as "damage" blurs the distinction between cosmetic cracking and structural distress.

STATISTICAL ANALYSIS OF DATA WITH PRE- AND POST-BLAST INSPECTION

19. Unmeasurables in observation or crack documen- tation can be taken into account indirectly by consider- ing the appearance of cosmetic cracks as a probabilistic event. In order to investigate the effects of certain data sets on the overall conclusions, the probability compu- tations of threshold or cosmetic cracking at given

(c) Admissibility of Dvorak's data has been ques- tioned by the researchers reexamining the old data in the late 1970s because of the absence of time histories; some of the other studies, such as that by Langefors et al. [18], are also plagued by the unavailability of time histories. To resolve this difficulty, only the new U.S. Bureau of Mines observations have been included in a recomputa- tion of probabilities in Fig. 9. The observations include low-frequency motions associated with surface mining. Again there is a particle velocity, 0.79 in./sec (20mm/sec), below which no blast-induced cosmetic cracking was observed. Furthermore, this lower bound case was observed in response to a surface coal mine blast.

particle velocity levels have been made several times [15, 16]. All of the observations studied by Siskind involve both immediate pre- and post-blast inspection of walls in residential structures in both Europe and North America, many of which were old, distorted and whose walls were covered with plaster. Such immediate inspec- tion is mandatory to separate structural distortion caused by natural weather changes from that caused by blast vibration.

(a) Data from various sets of systematic crack obser- vations were analyzed with the assumption that every cracking observation excludes the possibility of non- cracking at a higher particle velocity (Siskind et al. [15], p. 55). If the probability of cracking is calculated as the percentage of observations at lower levels of velocity, the result is the log-normal scaled plot of the probability of cracking particle velocity in Fig. 8. This approach seems conservative as low particle velocity observations do not count non-cracking at higher levels.

(b) According to Fig. 8, there appears to be a lower limit of particle velocity of 12 mm/sec below which no cosmetic or threshold cracking (extension of hairline cracks) has been observed from blasting anywhere in the world. This observation includes data with unusually low frequencies that were collected by Dvorak [17]. His data are those that tend to populate the lower region of Fig. 8. High-frequency data ( > 40 Hz) show that a 5% probability of displaced cracking does not occur until particle velocities reach 75 mm/sec [15].

FREQUENCY CONTROL OF STRUCTURAL RESPONSE

99

20. Structures respond most to ground motions when the excitation frequency matches the structure's funda- mental frequency. As shown in Fig. 1, walls and floors respond more to the higher frequency (15-20 Hz) waves in the early portion of that time history, while the superstructure or overall skeleton of the structure re- sponds more to the last or lower frequency (5-10 Hz) portion.

(a) Differences in structural response such as that shown in Fig. 1 can be calculated from the ground motions if the natural frequency and damping of struc- tural components arc known or estimated. Langan [19]

lO 1ooo

i l l l l I I

95

90

8O

7O

:= 60 J~ o .o 50 o

" 40 g o 30 E o r~ 2o

10

Par t i c l e ve loc i t y ( m m / s e c ]

l oo

:/f, i i i i i i i i~ i i o o l

• T h r e s h o l d d a m a g e ~ . ~ L ~ Minor domo,e ~;j:/.z~'-~-

• Major damage

ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD 153

1 I I ~ I I l t l I i i i i t i l l 1 1

0.2 0.5 1 10

P a r t i c l e ve loc i t y ( i n . / s e c )

Fig. 8. Probability analysis of worldwide blast cracking data [15]. Threshold damage is the occurrence of hair-sized, cosmetic cracks similar to those caused by natural, environmentally-induced expansion

and contraction.

99

95

90

8o

• ~ 60 J~ 0 • g so & G) O) ~ 3o o

2O

lO

lO

5

1

0.2

P a r t i c l e v e l o c i t y ( m m / s e c ) too

I

o

o j ° O

O

o

I I I I I I 0.5 I 2 5 10 20 50

P o r t i c l e v e l o c i t y [ i n . l s e c ]

Fig. 9. Probability analysis of blast-induced threshold cracks observed by U.S. Bureau of Mines [16].

15,1 ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD

has shown that measured structural response has a higher correlation coefficient with calculated single- degree-of-freedom (SDOF) response than with peak ground motion. Therefore structural motions can be estimated more accurately by assuming that they are proportional to response spectrum values at the particu- lar structure's natural frequency than by assuming that they are proportional to the peak ground motion [1]. This improved correlation is largely a result of the consideration of excitation frequency.

(b) Figure I0 compares time histories and response spectra from the longitudinal components of a small, urban construction blast and a large, surface coal mine blast. The mining blast involved detonation of 12,600 kg of ANFO (ammonium nitrate fuel oil) with a planned maximum charge per delay of 60 kg some 825m from the recording instrument. The much smaller construction blast involved detonation of 9 kg of gelatin with a maximum charge per delay of 2.3 kg at a distance of only 15m. Although the peak particle velocities are similar: 3.8 mm/sec for the construction

blast A; and 3.3 mm/sec for the surface mining blast B; the response spectra differ radically. This difference is greatest in the range of natural frequencies of residential structures and their components, 5-20Hz. In this range the surface mining motions produce response velocities that are 10 times greater than the construction blast.

(c) This lower response of structures with natural frequencies of 5-20 Hz to high-frequency excitation shown in Fig. 10 has led to the adoption of frequency- based standards in Germany and the U.S. [20, 21]. While both of these standards allow greater particle velocities for high-frequency excitation, there is considerable dis- agreement over the allowable particle velocities as shown in Fig. 11, which compares various control limits. Limits are based upon particle velocity measurement in the ground (OSM) and on the foundation (DIN). Regardless of the difference in limits, the allowance of higher particle velocities in high-frequency excitation is the same. More work is necessary to reconcile these differ- ences in limits.

O.lg 1.0 in. 1 . ~ , 0.1 in. l O.~, , , , 10.00 : j

IOamping ! 5%)

>

=,.

0.10 0.05

0.01 1 2 4 6 8 10 20 40 60 8 0 1 0 0

Frequency, Hz

0.01 in

lOO I

¢D lO •

c

Z

@

E

"~ o O.lO 0.20 o

¢D

Q.

maximum particle velocity A 3.8 mm/s

A

I I I I 0.~ 0.~ 0.~ 0.~

B 3.3 mm/s

I I I I I I 0 0.40 0.80

I I I I 1.20 1.60 2.00

T i t ~ , $O¢

Fig. 10. Comparison of time histories and response spectra from construction and surface mining blasts respectively lasting 0.15 and 2.0 sec. Even though the particle velocities are approximately equal, responses in the 5--20 Hz frequency range differ

greatly.

ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD 155

' ° ° 1 = . . . . . . . . ' ' ' ' ' ' " ' 1 o I - . - . - : o r e (6s~ . . . . . I

~, 20 F (~) ¢ ~ i n ot,ea \ / "71 ® , .w" o.n

i 0 2 . w J I

0.1 j , , i i i t i l l l ! i i I , i i i

crack are compared in Fig. 12. The continuous and highly cyclical curve is that of displacements produced by environmental change. The small circles are the

• ~- maximum, zero-to-peak, dynamic displacements .~ recorded by the same gauge. Even though the maximum

recorded particle velocity was as high as 24 ram/see, the > ® maximum weather-induced displacements were three

times that produced by blasting. At other gauges, a. weather changes produced displacements that were 10

times greater than those produced by blasting. @ 25 e~

_o <

E E I[

1 4 10 20 30 100 Blast Vibration Frequency, Hz

Fig. 11. Frequency based blast vibration control limits: ( ) Office of Surface Mining [21]; (-- * --) Deutsche Normen [20]. Comers 2 and 3 of OSM are unverified. Upper and lower dotted lines have been employed safely for close-in construction blasting near engineered structures (E) and in urban areas near older homes and historic

buildings (U).

COMPARISON OF BLAST AND ENVIRONMENTAL EFFECTS

21. Crack width changes from ground motions less than 25 mm/sec are less than those caused by the passage of weekly weather fronts [1(3]. This conclusion was reached after measuring the displacement response of a poorly-built, non-engineered, wood-framed house to surface coal mining vibrations for some 8 months. Displacements were measured at l0 different wall pos- itions that included cracked and uncracked wall cover- ing. Weather and blast-induced crack displacements across the most dynamically responsive wall covering

RESTRAINED STRUCTURES AND ROCK MASSES

22. Capacity for free response allows above-ground structures such as homes and rock pinnacles to amplify selectively incoming ground motions; however, buried or restrained structures such as pipelines and rock masses cannot respond freely. Regardless whether response is restrained or free, cracks are initiated by strains. Whereas strains in a freely-responding structure are proportional to the relative displacement between the ground and the superstructure, strains in a restrained structure such as pipelines will usually be those of the surrounding ground and can be approximated as those produced by plane wave propagation and are:

E = - - and 7 - - - (4) C¢ Cs '

where ~ and 7 are axial and shear strains, c= and c, are compressive and shear wave propagation velocities, and

are maximum compressive and shear wave particle velocities, respectively [1]. This calculation of strain is approximate, especially when ~ is measured at the ground surface, and requires the measurement of c= and c, at the site. More work is required to improve this

2.0

~" 1.5 E

t43

~ 1.0 I

0.5

0.0 E

~ 4 ) . 5 ! • ~ -1.0

-1.5 L

-2.0 86.6

0

weather

blast

0 0 0

0

0 CD

86.8 87.0 87.2 87.4

Year Fig. 12. Comparison of crack displacements in a wood-framed house produced by weather-induced changes in humidity and

temperature ( ) with those produced by surface coal mine introduced ground motions (0).

156 ISRM: BLAST VIBRATION MONITORING SUGGESTED METHOD

a p p r o a c h to e s t i m a t i n g s t ra in . F o r cases i n v o l v i n g one cri t ical l oca t ion a l o n g a p ipel ine , the p ipe s t ra ins shou ld be m e a s u r e d direct ly o n the meta l . F o r cases i n v o l v i n g t u n n e l a n d / o r cave rn l iners, cr i t ical s t ra ins can be esti-

m a t e d t h r o u g h ca l cu la t ion o f the relat ive flexibility o f the rock a n d l iner [23].

Accepted for publlcat,)n 30 Oc;ober 1991.

REFERENCES

1. Dowding C. H. Blast Vibration Monitoring and Control. Prentice- Hall. Englewood Cliffs, NJ (1985).

2. Hudson D. E. Reading and interpreting Strong Motion Accelograms. Earthquake Engineering Research Institute, Berkeley, CA (1979).

3. New B. M. Ground vibration caused by civil engineering works. Transport and Road Research Laboratory LR53, TRRL, Crowthorne, U.K. (1986).

4. Hendron A. J. Engineering of rock blasting on civil projects. Structural and Geotechnical Mechanics (W. J. Hall, Ed.). Prentice- Hall, Englewood Cliffs, NJ (1977).

5. Medearis K. The development of a rational damage criteria for low rise structures subjected to blasting vibrations. Report to National Crushed Stone Assoc., Washington, DC (1976).

6. Schomer P. D., Goff R. J. and Little M. The statistics of amplitude and spectrum of blasts propagated in the atmosphere. U.S. Army Construction Engineering Research Laboratory, Technical Report N-13 (1976).

7. Harris C. M. and Crede C. E. (Eds) Shock Vibration Handbook. McGraw-Hill, New York (1976).

8. Rosenthal M. F. and Morelock G. L. Blasting Guidance Manual. Office of Surface Mining Reclamation and Enforcement, U.S. Department of the Interior, Washington, IX2 (1983).

9. Stagg M. S., Siskind D. E., Stevens M. G. and Dowding C. H. Effects of repeated blasting on a wood-frame house. U.S. Bureau of Mines, Report of Investigations 8896 (1984).

10. Dowding C. H. Comparison of environmental and blast induced effects through computerized surveillance. The Art and Science of Geotechnical Engineering at the Dawn of the 21st Century, R. B. Peck Honorary Volume (3,V. J. Hall, Ed.), pp. 143-160. Prentice- Hall, Englewood Cliffs, NJ (1988).

I 1. Johnson C. F. Coupling small vibration gauges to soil. Earthquake Notes, Vol. 33, No. 3, pp. 40-47. Eastern Section, Seismological Society of America (1962).

12. Stagg M. S. and Engler A. J. Measurement of blast induced ground vibrations and seismograph calibrations. U.S. Bureau of Mines, Report of Investigations 8506 (1980).

13. Anderson D. A., Winzer S. R. and Ritter A. P. Synthetic delay versus frequency plots for predicting ground vibration from blast- ing. Proc. 3rd Int. Syrup. on Computer Aided Seismic Analysis and Discrimination, pp. 70-74. IEEE Computer Society Press (1983).

14. Northwood T. D., Crawford R. and Edwards A. T. Blasting vibrations and building damage. The Engineer 215, (1963).

15. Siskind D. E., Stagg M. S., Kopp J. W. and Dowding C. H. Structure response and damage produced by ground vibrations from surface blasting. U. S. Bureau of Mines, Report of Investi- gations 8507 (1980).

16. Siskind D. E. Open file report of responses to questions raised by RI 8507. Available for inspection, U.S. Bureau of Mines, Minneapolis, MN (1981).

17. Dvorak A. Seismic effects of blasting on brick houses. Prace Geofyrikenina Ustance. Ceskoslovenski Akademie, Ved., No. 159, Geogysikalni, Sbornik (1962).

18. Lang*fors U., Westerberg H. and Kihlstr6m B. Ground vibrations in blasting. Water Power Sept, (1958).

19. Langan R. T. Adequacy of single-degree-of-freedom system mod- eling of structural response to blasting vibrations. M.S. Thesis, Department of Civil Engineering, Northwestern University, Evanston, IL (1980).

20. DIN. Deutsche Normen: Erschiitterungen im Bauwesen--Ein- wirkungen auf bauliche Anlagen. DIN 4150 (1983).

21. OSM, Office of Surface Mining. U.S. Dept. of Interior. CFR, Vol. 48, No. 46 (1983).

22. Dowding C. H. and Gilbert C. Dynamic stability of rock slopes and high frequency traveling waves. J. Geotech. Engng ASCE 114, 1069-1088 (1988).

23. Hendron A. J. and Fernandez G. Dynamic and static design considerations for underground chambers. Seismic Design ofEra- bankments and Caverns (r. Howard, Ed.). American Society of Civil Engineers, Special Technical Publication (1983).

24. Siskind D. E. and Fumanti R. Blast-produced fractures in Litho- nia granite. U.S. Bureau of Mines, Report of Investigations 7901 (1974).

25. Holmberg R. and Persson P. A. The Swedish approach to contour blasting. Proc. Fourth Conf. on Explosives and Blasting Techniques, pp. 113-127. Society of Explosives Engineers, Montville, OH (1978).

26. Roth J. A model for the determination of flyrock range as a function of shot conditions. Report prepared for the U.S. Bureau of Mines by Management Services Association, Los Altos, CA, NTIS, PB81-222358 (1979).

27. Lundborg N. The probability of flyrock. Report DS 1981:5, Swedish Detonic Research Foundation, Stockholm (I981).

28. Ivanov P. L. Compaction of noncohesive soils by explosions. Translated from Russian by the National Science Foundation and available from the library of the U.S. Water and Power Resources Services, Denver, CO, TA I0193 (1967).

29. Siskind D. E., Stachura V. J., Stagg M. S. and Kopp J. W. Structures response and damage produced by airblast from surface mining. U.S. Bureau of Mines, Report of Investigations 8485 (1980).

30. New B. M. The effect of detonator variability on explosively induced ground vibration. Int. Conf. Earthquake, Blast and tmpact, UMIST Manchester. Institution of Civil Engineers, London (1991).

31. Stachura V. J., Siskind D. E. and Engler A. Airblast instrumenta- tion and measurement techniques for surface mining. U.S. Bureau of Mines, Report of Investigations 8508 (1981).

APPENDiX--Permanent Degradation and Displacement of Adjacent Rock

23. Permanent effects, with the exception of fly rock, are encoun- tered only near shot holes and can be divided into degradation and displacement. Degradation is normally described by cracking intensity. Such blast-induced cracking has been observed experimentally to vary with hole diameter and rock type [24, 25]. Small-bole-diameter con- struction blasting has induced cracking at distances of 1-2 m, and larger-hole-diameter mining blasts are capable of producing cracks at distances of 10-15m. Careful blast design can reduce dramatically these maximum distances.

(a) Displacement can be produced by either delayed gas pressures (those that accumulate during detonation) or to a lesser extent by vibration-induced shaking. Delayed gas pressures have dislocated blocks as large as 1000m ~ during construction blasting [1]. Such movement is unusual but is associated with isolated blocks, leakage of gas pressures along open joints, and poor shot design with large burdens. Vibratory or shaking-induced displacement is normally as- sociated with unstable blocks in rock slopes and can occur wherever static factors or safety are low and ground motions produce permanent displacements that are larger than the first-order asperity wavelength

on the sliding joint or plane [22]. Gas pressure related displacement can occur out to 10s of metres.

(b) Fly rock is a special case of permanent displacement of rock by explosive expulsion from the top of the blast hole and has been propelled as far as 100-1000 m [26]. Statistical studies have shown that the probability of these extreme events are quite low under normal circumstances, 1 in 10,000,000 at 600m [27]. Since the probability increases with decreasing distance, blasting mats are required for any construction blasting in an urban environment to prevent all fly rock.

(c) Another special case of permanent displacement is the vibratory densification of a nearby mass of loose, clean sand. The propensity for such densification is a function of the soil's density, mineralogy and grain size distribution. Soils that are densifiable are loose sands, with less than 5% silt-size particles. These clean sands were densified out to distances of 20 m [28] after detonation of single, 5 kg charges within the loose sand mass itself. Soils that are either slightly cemented or contain more than 5% fines are a great deal less subject to vibratory densification from typical ground motions.