HomeMy WebLinkAboutRudenko Article re: Air Blast
.i
AIR BLAST
AN OFTEN OVERLOOKED CAUSE OF STRUCTURAL RESPONSE
by
Douglas Rudenko
Vibra- Tech Engineers, Inc.
Hazleton, Pennsylvania, USA
ABSTRACT
When blasting complaints come, as an industry we often immediately look to the ground vibration as the
source of the trouble. Often times we overlook or place less emphasis on the collected air overpressure
data. Air blast like ground vibration is an undesirable side effect of blasting and can produce
Considerable structural response.
As a conSulting seismologist, we often hear homeowners' claim that the quany blasted twice in rapid
succession and that the second blast was worst than the first. In this paper we will apply a simple,
reliable way of estimating structural response from recorded air overpressures. The technique developed
by others determines the air overpressure load using a superposition technique. Single-degree-of-
freedom dynamic response can then be calculated using the modified air overpressure time history.
In this presentation we will examine the primary factors that control structural response. Through the
use of a large array of digital seismographs, the effects of two production blasts on a community
surrounding a quany are studied. The predicted structural response resulting from the air blast for each
blast is calculated at 175 seismograph locations. The predicted response is then correlated to previous
research on human perceptibility to vibrations. The results of this analysis and correlation give a better
understanding of what the residents actually feel in their homes.
THE UNRECOGNIZED PROBLEM
As an industry, we have a predisposition to focus on the resulting ground vibration for our blasting
operations often times completely ignoring or giving only momentary pause to the recorded air
overpressure. Consulting seismologists are often called upon to visit neighbors of quarries against
whom complaints have been lodged. We try to gain an understanding of why the neighbors complain by
listening to what they say. The following are actual excerpts from a local complaint line for a quarry.
"Calling to report the blasting at 11:30 am. again today. It was so bad that my son came downstairs to
see if I hadfallen, and there were quite a few rumbles after it. "
"Ya know, I really don't mind the blasting that much, but why do you always shoot two shots one right
after the other? And you always load the second one heavier!"
"I'm calling with a blast complaint. The blast went off at approximately 11:25 a.m. today. There were
two separate blasts; the second one shook the living day lights out of my home. "
"Our home felt two shocks today. The first, a rumble under the home. The second felt as if our home
was struck by lightening. Our entire home felt the shock. It shook on all levels. We were all startled,
even the dog. A photo hanging in the front bedroom of the second floor fell to the floor but was not
broken. We demond a solution/ 1i!Jl!D!!"
AIR BLAST CHARACTERISTICS
The airblast produced by a typical qua..-ry blast is an impulsive sound generated by the explosive blast
and the resulting rock fragmentation and movement. There are four generally recognized sources of
airblast. These sources are
. The displacement of rock from the free face or mounding at the blasthole collar
. The movement of the soil from ground.vibration
. The venting of gas through rock fractures
. Gas escaping because of stemming ejection
These four contributing sources to the total air overpressure pulse are called the air pressure pulse, rock
pressure pulse, gas release pulse, and stemming release pulse respectively.
Each blasthole in a multi-hole production blast is a source of an air pressure pulse (APP). In a properly
designed blast the air pressure pulse will be the dominant contributor to the total air overpressure pulse.
The time .between holes in the front row can usually be detected by measurements taken in close
proximity to the shot. At large distances however, the individual air pressure pulses form a single very
low frequency overpressure due to dispersion and refraction.
The rock pressure pulse (RPP) is formed by the vertical displacement of the ground over an area. In
essence, the ground acts as a large piston pumping up and down pushing the air. This is typically
identified on the time history prior to the air pressure pulse and occurs at the same time as the ground
vibration that produces it. The rock pressure pulse is usually of high frequency but low amplitude. The
amplitude is roughly about 1/600th of the vertical peak particle velocity.
The most undesirable part of an airblast is the gas release pulse (GRP) and stemming release pulse
(SRP). The gas release pulse and the stemming release pulse resuh from a blowout. These pulses are
high amplitude, high frequency spikes that are typically superimposed on the air pressure pulse.
Fortunately, they are controllable since they are often related to improper burden or spacing, inadequate
stemming, detonation velocity, or timing issues. Figure 1 below is an example of an air overpressure
time history containing three of these characteristics.
2
,-
Figure 1. Example of Air Overpressure Time History
RPP\ ,I ~. .~PP
-~ "~\f~SIlP --.r.~.~V~ ·
.h..,..~
FACTORS CONTROLLING STRUCTURAL RESPONSE
The air vibrations produced by blasting cause the nonnal air pressure to fluctuate. Changes in normal
air pressure due to the airblast are referred to as overpressure, as in pressure over atmospheric pressure.
The typical shape of an airblast wave is an initial positive overpressure phase followed by a negative
phase in which the pressure falls below atmospheric pressure. When air pressure changes rapidly
different pressures can result on both the inside and outside of a structure. A change in pressure that
occurs between locations is often called a pressure' gradient. A pressure gradient from the inside of a
structure to the outside produces forces, which are exerted over the structure's exterior surfaces. These
forces, if large enough, can cause the structure to respond causing building movement and noise within
the structure.
Structural response is the real cause of why people complain about blasting. All blast vibration
complaints are due to how much your neighbor's house shakes, not how much the ground shakes. There
are three factors of air vibrations that determine how much your neighbors' houses vibrate. They are the
amplitude of the air overpressure (dB), air vibration duration, and air vibration frequency.
In order to understand how these three factors control how a house vibrates, let us think of a more
familiar model - a swing. A swing is a, single-degree-of-freedom vibration model that behaves in a
manner similar to a house. Think of the air overpressure amplitude as how hard you push the swing. If
you push it harder, it goes higher. Structural response therefore, is directly and linearly proportional to
air overpressure amplitude. If you reduce your psi by half: stIUctw'al response will be cut in half.
Air vibration duration is an equally important parameter in considering structural response. Using the
swing analogy, you can easily make a swing go higher without pushing it harder, by simply pushing it
again. The more times you push the swing the higher it will go. Longer air vibrations continue to shake
the house causing a greater amplitude of structural response.
Frequency is the most important of the three factors of air vibration. Air vibrations at the fundamental
frequency of a house are like pushing a swing whenever it comes back to you. If you push the swing at
any other time, you disrupt its rhythm. If a house is exposed to air vibrations near its fundamental
frequency, the house will amplify the vibration.
3
DETERMINING THE FUNDAMENTAL FREQUENCY OF OUR NEIGHBORS' HOUSES
In 1976, Ken Medearis published a report to the National Crushed Stone Association on the
development of rational damage criteria. In this report, Medearis determined the fundamental frequency
and damping ratio of 63 residential structures subjected to micro-vibration testing. He found that the
height of the structure primarily determined its fundamental frequency. The fundamental frequencies of
the structures in this study were found to range from 4 to 18 hertz (Hz), with an average fundamental
frequency of9.6 Hz. A summary of the results from the Medearis study is shown in Figure 2.
Fipre 2. Fundamental frequencies of structures.
~
...
6 0.25
l 0.2
t> 0.15
a
-= 0.1
1
.t 0.05
~
6 0.25
1 0.2
=
t> 0.15
is
-= 0.1
.a
o
It 0.05
Fundamental Frequencies of Residential Structures
One Story Structure
Mean =It.l1 :az
~
...
60.25
] 0.2
t> 0.15
~
w 0.1
~ 0.05
o
5 10 15 20
Frequency, Hz
5 10 15 20
Frequency, Hz
o
.AB~
~ 0.3
6 0.25
I 0.2
t> 0.15
i
III 0.1
1
l5; 0.05
o
o
MQn = '" l&
MelIJl= '.6l&
5 10 15
Frequency, Hz
20
o
5 10 15
Frequency, Hz
20
In the 1970s, the United States Bureau of Mines (USBM) also studied the fundamental frequency of
residential structures. The USBM determined the dynamic properties of 23 residential structures
subjected to actual blasts rather than impulse loading as in the Medearis study. The values for frequency
and damping computed from actual blasting records tended to be slightly lower than the Medearis study.
The USBM also determined that while houses vibrate as a single-degree-of-freedom between 4 and 12
Hz, the natural frequency of the house's midwall tends to occur between 12 and 20 Hz. In order to
control the response of a structure that has more than one fundamental frequency, you must control the
two lowest fundamental frequencies. For residential structures this means minimizing air overpressure
energy between 4 and 20 Hz.
4
MODELED RESPONSE DUE TO AIR OVERPRESSURE
A structure such as a house or even a wall is a complex multi-degree of freedom, dynamic system. In
order to model a structure analytically, the model must be idealized. A single-degree-of-freedom (SDF)
model is the simplest model that incorporates the amplitude, frequency, and duration of the excitation
pulse and the frequency and damping of the model. The SDF model will be utilized to predict the
structure response from dynamic loads in this paper.
Fulthorpe (1980) discusses the interaction of an airblast wave with a structure. Due to the wavelength of
the air wave and the dimensions of the structure it can be shown that the air wave quickly engulfs the
spucture and that the overpressure acts on the front and rear of the structure at the same time. The net
pressure that acts on a surface facing the blast therefore, will be less than the full measured overpressure.
In order to model this effect, the loads on the front and rear of the structure must be combined into a
single resultant front load for input into the SDF model. The resultant pressure that acts on the front
face of a structure is calculated by superimposing on the original overpressure time history; an equal, but
negative overpressure time history delayed by the time required for the wave to pass across the structure.
The original overpressure time history represents the pressure on the front of the structure. The equal,
but negative overpressure time history represents the pressure on the rear of the structure. This
technique utilized by Fulthorpe is basically a simplification of that employed by Wiggins for modeling
sonic booms.
In order to model the response as a SDF system, it was necessary to reduce the masses of the many
components of a structure into a single mass. For superstructure response, Fulthorpe (1980) assumes
that the mass of the house superstructure is equal to the total mass of the structure above mid-height of
the first floor walls. The area of the superstructure that the air blast acts on is the area above mid-height
of the first floor walls normal to the direction of the measured response. In the case of mid-wall
response, the mass of the wall is assumed to be equal to the total mass of the wall between restraints.
Restraints are the floors, ceilings, and internal and external walls perpendicular to the wall under study.
For mid-wall response the area of the wall is assumed to be the area between restraints. The average
mass and area for the five structures studied by Fulthorpe (1980) were used for this paper. Figure 3
below shows the effect of superposition on an air overpressure time history along with the modeled wall
response. The modeled wall responses were found to be a factor of two higher than the measured wall
response in the Fulthorpe (1980) study.
5
Figure 3. Air Overpressure Superposition and Modeled Response
SuperlmpositiottofFree-Field Air Blast
'i'
.&
J
Resultattt SuperimpOsed Air Pressure1ime Histoty
i
!
~
Predicted wan VeJodty (iDlsee)
As stated earlier, the USBM determined that the house superstructure resonates between 4 and 12 Hz
while the resonant :frequency of the walls was between 12 and 20 Hz. Assuming a resonant frequency of
10Hz for the superstructure and 20 Hz for midwall, Figure 4 below shows the combined effect of both
ground vibration and air overpressure from . a blast. The recording seismometer was located
approximately 3,050 feet from a multi-hole production blast. Maximum pounds per delay were 3711bs.
From this figure we see that the dynamic load applied by the air blast is over two times that of the
ground vibration. From the resulting structural response at 2.5 seconds, we see why our neighbors think
the second blast is always loaded heavier. The predicted midwa11 response at 20 Hz is tremendous.
6
Figure 4. Superposition or Ground and Air Vibration and Modeled Response
.
I AJr mast Tone ffistory
I AJr BJaot'otdDa Function
.IGr_d VibrlOlion nmellis10ry
.. tfJl'.~
~-
~~~
~-
.
I Com;=dForcmg~__
!ModeIed ~:: 1~n.
I Modeled R..,_1O n.
~."
-+----.-
THE HUMAN SEISMOGRAPH
Structural response plays a critical role in your neighbor's perception of blast vibrations. The tolerance
and reactions of humans to vibrations is another important aspect of understanding why people complain
about blasting. Humans notice and react to vibration levels much lower than the levels established as
structural damage thresholds. The authors of U.S. Bureau of Mines Report of Investigation 8507 (RI-
8507) state quite plainly that the "method of analyzing the damage potential of blasting vibrations has
the disadvantage of possibly underestimating annoyance reactions."
One study on human response to transient vibrations was the Wiss and Parmelee study. In this study the
responses of 40 people to damped, sinusoidal pulses of 5-second duration were recorded. Figure 5
shows the threshold response (dashed line) of the most sensitive people along with the mean response
(solid line) of the average subject. The difference in the threshold response to the mean response shows
7
the subjective nature of human response to vibrations. The results of the Wiss and Parmelee study were
reanalyzed in RI-8507 for duration-of-vibration effects. The results of this analysis showed that human
tolerance to vibration decreases the longer the vibration continues.
Figure 5. Human response to transient vibrations of various frequency and
duration (From Wiss & Parmelee, 1974).
10.000
". ::';..:,:..,.;:'i<,::'::':":,,:,'..:.'))L;j:;:~:':~".,' :.:::::;..::::"1:"'
.DiStinctlyPerceptible.
- ... -':-,:- - - - -"'..
1.000
CJ
~
:e
~
.Q 0.100
-
~
~
~
~
0.010
Barely Perceptible
Barelv Perceotible
.-...--..---
Response Level
- Mean - - · Threshold
0.001
1
10
Frequency, Hz
100
A DIFFERENT APPROACH
Based upon our understanding of what causes structural response and how people react to vibrations, let
us examine a few production blasts from a Mid-Atlantic aggregate operation. During this study 170
digital seismographs were placed within a 1.S-mile radius of the quarry to record the ground vibration
and air overpressure time histories resuhing from four different production blasts. The maps shown in
Figure 6 are color contoured to coincide with the four threshold response levels from the Wiss and
Parmelee study. Blue representing "barely perceptible" from 0.01 to 0.10 inIsec. Green rt!})resenting
"distinctly perceptible" from 0.10 to 0.40 inIsec. Yellow representing "strongly perceptible" from 0.40
to 1.0 inIsec. and red representing "severe" above 1.0 inIsec. The figure shows a comparison of peak
response velocities resulting from the ground vibration on the left and the air blast on the right. The
response is calculated for a wall assuming a 20 Hz system.
8
Figure 6. Modeled response for a 20 Hz structure ground only vs. air only to Wiss & Parmelee, 1974
Ground - Production Blast 1
Air - Production Blast 1
Response Velocity
inIsec
--;:."""-"" I I "11I ",,1 'I
.. ,- ....... ~ ""n'
,_",:c..,.;.r,~ ;/;,.,~~;- \.~.~ I
0.01
8.10
0.40
1.0
2.0
From the maps in Figure 6 it is easy to dismiss complaining neighbors as being oversensitive if we only
focus on the ground vibration. Ignoring the air overpressure and resulting response only heightens the
neighbors' intolerance to the mining operation.
Figure 7 below shows a comparison of the peak response velocities resulting from the air overpressure
of two different production blasts. Production Blast 1 was located a few hundred feet west of
Production Blast 2. Both production blasts are located on the same bench. Bench heights at this quarry
are between 43 and 59 feet. The following table gives pertinent blast design information about each
blast.
Production Blast 1 Production Blast 2
Number of holes 54 72
Number of rows 4 4
Burden 11 ft. 11 ft.
SpacinJl; 13 ft. 13 ft.
Hole Diameter 4.5 in. 4.5 in.
Stemming (Crushed Stone) 8 ft. 10 ft.
HoleslDelay 2 1
Maximum lbs.ldelay 738 310
Timing 17 ms between holes 19 ms between holes
76 ms between rows 104 between rows
9
Figure 7. Modeled response of blasts for a 10 Hz structure to Wiss & Parmelee, 1974.
Production Blast 1
Production Blast 2
Response Velocity
inIsec
"B.ord, -.: __~
PUCl.jHibl(''' ;:," ~''':''~ ,~' __
0.01
G.10
0.40
LO
2.0
The response maps in Figure 7 are calculated for a superstructure assuming a 10 Hz system. From this
figure we see that the blast design in Production Blast 1 had a less desirable effect for the surrounding
community than the blast design in Production Blast 2. The entire community located west of the quarry
experiences 11lC1cing response velocities of 0.1 in/sec to 0.40 in/sec for Production Blast 1. Neighbors
located north and east of Production Blast 1 experienced racking response velocities of 0.40 in/sec to 1.0
in/sec. Video footage of each shot reveals venting from the toe along with violent stemming ejection for
a particular hole in Production Blast 1. Production Blast 2 had less violent stemming ejection and no
venting from the face. In addition, maximum pounds per delay for Production Blast 1 is twice that of
Production Blast 2. Timing down the face could be an additional factor.
Figure 8 shows the response maps for calculated wall response assuming a 20 Hz system. Since the air
overpressure from Production Blast 1 is richer in higher frequency energy the mid';'wall response from
this blast is greatly enhanced. Production Blast 1 subjects most of the community to mid-wall response
velocities greater than 1.0 in/sec. Velocities of this nature would be considered "severe" by most
people. According to Siskind (2000), a general criterion for vibrations sufficient to produce movement
of loose objects is an acceleration of 0.20 g to 1.0 g. Assuming the 20 Hz system in Figure 8, 0.61
in/see to 3.0 in/sec would be required to knock loose objects off a wall shelf.
Complaints from neighbors about blasting almost always come from people inside their homes.
Generally, they describe structural motion and the associated rattling and rumbling noises. People
almost always attribute this motion and noise to ground vibration. In reality, the ground vibration
arrived unnoticed several seconds before the air blast. In addition, most of the energy from the air blast
10
is below the audible range of human perception. However, this energy is capable of causing structural
response, particularly mid-wall response. This response produces higher frequency secondary noise
which the occupants of a structure can perceive. Not hearing the sound, people generally assume the
rattling is due to ground vibrations. The following excerpt from a blasting complaint line illustrates this
point.
"Another blast today at approximately 11:46 a.m. I was in the garage and there must have been a
vibration in the air because the garage door vibrated back and forth. I have carpeting in the garage,
and you couldfeel the rumble underneath. This blast was about a 4 on a scale of 1 to 5. "
Figure 8. Modeled response of blasts from air overpressure at 20 Hz to Wiss & Parmelee, 1974.
Production Blast 1
Production Blast 2
Response Velocity
in/seo
"B.IIl'l>, ..
PcrnpttlJl(''' p
> .,
,
0.01
0.10
0.40
LO
2.0
A SOLUTION TO THE PROBLEM
Although airblast is often overlooked and under diagnosed, it is one of the more controllable aspects of a
blast. Many of the variables that create air overpressure problems are under the direct control of the
blaster. The following is a short list of variables that can effect the resulting air overpressure.
. Maximum charge weight per delay
. Amount and type of stemming material
. Amount of burden and condition of face
. Exposed detonating cord on the surface
. Orientation of face relative to structures
. Delay interval and direction of initiation
. Amount of displaced rock
11
. Wind direction
. Atmospheric conditions
. Topography
The following paragraphs are some blast design parameters and considerations that may be helpful in
reducing air overpressure problems.
Maximum Charge Weight
The maximum amount of explosive per delay affects the air overpressure in a manner similar to the
effect on ground vibration. Most blasting research indicates that air overpressures drop 6 to 7 dB per
doubling of distance. For estimating the air overpressure for a particular charge weight at a given
distance the following formula can be applied for average burial.
i D )-1.1
PAD = l.'\.VW
where P AO = peak air overpressure (psi)
D = Distance to structure (feet)
W=charge weight (Ibs)
Importance of Confinement
The selection of stemming material is a key factor in controlling airblast. Many mining operations
outside of the aggregate industry use the least expensive material available to them - drill cuttings. Drill
cuttings are a poor choice for stemming material. A better choice for stemming material is angular,
crushed stone. The size of the stone should be around one-tenth of the blasthole diameter. If cost is a
factor consider using 4 to 5 feet of crushed stone above the explosive column followed by drill cuttings.
A simple formula for determining minimum stemming lengths is
SL =CDx24+12
where SL = stemming length (feet)
CD = explosive charge diameter (inch)
Besides stemming, other important factors that effect proper explosive confinement are sufficient burden
and spacing, accurate drilling, and proper timing. Siskind (2000) states that although a well confined
blast with fully coupled explosive charges may increase ground vibration by a factor of two times,
airblast variations may be on the order of 10 to 100 times.
Timing Considerations
Besides the relative importance of timing to under confined or over confined blastholes, timing can also
play an important role in the generation of airblast by inadvertently reinforcing the air waves of
successive charges. Supersonic detonation down the face is the result of explosive charges that detonate
too fast for the given distance between charges. Andrews (in ISEE 1998) recommends keeping the
12
REFERENCES
Blast Dynamics, Inc. (1991). Efficient Blasting Techniques. Author.
Dowding, C.H., Murray, P.D. & Atmatzidis, D.K. (1981). Dynamic Properties of Residential
Structures Subjected to Blasting Vibrations. Journal of the Structural Division. Proceedings of the
American Society of Civil Engineers. l07(ST'n, 1233-1249.
Floyd, J. (1995). Blast Design, Lesson Six. In International Society of Explosives Engineers,
Practical Blasting Fundamentals Levell (pp. 6-1- 6-34). Cleveland, OH: Author
Fulthorpe, C.S. (1980). Computer Modeling of Structural Response to Combined Air Blasts and
Ground Vibrations. Master's Thesis, Department of Civil Engineering, Northwestern University,
Evanston, Dlinois.
International Society of Explosives Engineers. (1998). Blasters' Handbook (17th ed.). Cleveland,
OH: Author.
Meadearis, K. (1977). The Development of Rational Damage Criteria for Low-Rise Structures
Subjected to Blasting Vibrations. Proceedings of the 18th U.S. Symposium on Rock Mechanics, 1-16.
Rudenko, D. (2000, July). Diagnosing and Solving Blasting Problems. Aggregates Manager.
25-29.
.-
Siskind, D.E. (2000). Vibrations From Blasting. Cleveland, OH: International Society of
Explosives Engineers.
Siskind, D.E., Stagg, MS., Kopp, lW. & Dowding, C.R. (1980). Structure Response and
Damage Produced by Ground Vibration from Surface Mine Blasting (Report of Investigation 8507).
U.S. Bureau of Mines.
Siskind, D.E., Stachura, V.l, Stagg, MS. & Kopp, lW. (1980). Structure Response and
Damage Produced by Airblast from Surface Mining (Report of Investigation 8485), U.S. Bureau of
Mines.
Wiss, J.F. & Parmelee, M (1974). Human Perception of Transient Vibrations. Journal of the
Structural Division. Proceedings of the American Society of Civil Engineers, 100(ST4). 773-787.
14