Documente Academic
Documente Profesional
Documente Cultură
Background HKIE Cavern Engineering Group formed to promote the general use of cavern facilities in Hong Kong Common perception in the industry (contractor, consultant and some GEOs) vibration limit for slope (being considered as sensitive receiver) is being set too conservatively low No case record of slope failed/moved/distressed due to blasting Checking of slope stability usually follows the approach in GEO Report No. 15
2
Background Maximum vibration limit allowed for slope/utilities Rule of Thumb/Empirical Maximum vibration limit being further reduced Implication to cost and program to tunneling
Terms of Reference:
1. 2. 3.
Consolidate current knowledge, experience and practice in cavern and tunnel engineering in Hong Kong. Identify key issues on and actions required for appropriate use of caverns and tunnels in Hong Kong. Promote awareness among professionals, developers and the relevant authorities of the potential and benefits of caverns and tunnels in Hong Kong. Make recommendations to the HKIE GDC on follow-up actions to be taken for promoting cavern and tunnel engineering good practice in Hong Kong.
4
4.
Approaches to Study:
Based on field instrumentation, calibrate numerical model which is based on established theory and assess the effect of blast-induced vibration on slopes Evaluate the vibration limit based on the above findings Depending on whether the vibration limit being set is conservative or un-conservative, propose alternative vibration limit based on the calibrated model Continue field monitoring and refine the model as blasting continue Allow more charge weight per delay improve blasting cycle and construction program
6
50 mm/s
25 mm/s
Humans are very sensitive to vibrations. Low frequency vibrations are more readily felt than high frequency vibration.
Blasting Range
Hertz
Restriction on Peak Particle Velocity and Vibration Amplitude : Utilities and Other Installations
Maximum Allowable Vibrational Amplitude (mm) 0.2 0.1 0.1 0.2 0.1 0.2 0.2 0.1
9
Utility
Installation
Gas: Hong Kong & China Gas Company Electricity: China Light and Power Company, Hong Kong Electric Company Water: Water Supplies Dept., Hong Kong Government
All Power Stations Sub-Stations (major) Sub-Stations (minor) Underground Cable Joints Underground Cables & Pylon Foundations Non Water Retaining Structures & Water Mains Water Retaining Structures
Restriction on Peak Particle Velocity and Vibration Amplitude : Utilities and Other Installations
Installation Utility Maximum Allowable Peak Particle Velocity (mm/s) Maximum Allowable Vibrational Amplitude (mm) 0.2 0.2 0.2
Drains & Sewers: Drainage Services Dept., Hong Kong Government Telephone: Hong Kong Telephone Company Estates/Structures/ Schools/Private Properties Computers Fresh Concrete: Less than 2 days old: 2 and 8 days old: More than 8 days old:
25 25 25
All
0.2
All
5 25 50
50 mm/s
25 mm/s
10 mm/s
Blasting Range
11
Li U. and Ng S. (1992), Prediction of blast vibration and current practice of measurement in Hong Kong, Asia Pacific quarrying the rim : proceedings of the Conference Asia Pacific - Quarrying the Rim, held in Hong Kong between 7 and 10 April 1992. Hong Kong: Institute of Quarrying (Hong Kong branch)
D PPV = K 0.5 W
Site Specific Constants PPV: Peak Particle Velocity D: Distance W: Charge weight per delay
12
50%
Mines Dept.
84%
Mines Dept.
1.22
95%
Mines Dept.
MTRC
D PPV = 994 0.5 W
D PPV = 1079 0.5 W
1.22
84%
1.07
84%
13
100.00
y = 644x - 1 . 2 2
Arbitrary Limit without data, and without knowing its actual effect
y = 1032x - 1 . 2 2
y = 642.33x - 1 . 2 3 4 2
10.00
y = 516.3x 84% Line
-1.2342
95%
1.00
90% (1.6449SE)
y = 368.95x - 1 . 2 3 4 2 R 2 = 0.8057 50% Line
95% (1.9599SE)
14
All blasts are controlled to less than 10 mm/s at sensitive structures. Therefore real responses are never known.
15
Assumptions made in GEO Report No. 15 For soil slope (Pseudo-static Approach):
SDOF Simple Harmonic motion in bedrock rock Rigid body motion in soil mass (horizontal direction only) Movement in-phase with acceleration Constant damping ratio in soil, no damping in rock Constant peak acceleration at all period Constant dynamic pore pressure generated Fundamental period of slope material equal to 3.3H/S. H=height of slope. S=velocity of traveling wave (300 m/s) from study of dams where vibration is amplified at the crest (not applicable to slope which is confined and continuous) No phase shift between velocity and acceleration Single input frequency of 30 Hz is used Mode shape not verified
16
22 1
ym y=4.6
H
H=12 m
17
y = 4.6 m H = 12 m y = 0.383 H
K a = 0.75 F = 30 H z
PPVc =
1.22
That means 11 mm/s would cause FOS to reduce to 1.0 (17% drop) further reduce in design to consider Alert, Alarm and Action levels
1 0.61
1 0.61
11 = 112 2 644
= 16 kg / delay
K a = 0.75
y = 0.383 H
Ka vs y/H for Slopes with Horizontal Bedrock Based on 1-D Shear Response Model GEO Report No. 15
18
Dynamic Analysis reduces all the implicit assumption made in GEO Report No. 15 by considering:
MDOF Actual non-periodic time history of motion can be input Soil is a compliant system (different points within the system move differently and out-of-phase from each other)
Incremental equation of motion can be used to model the nonlinear soil behavior if necessary change in shear modulus and damping ratio varies with cyclic strain
19
D PPV = K 0.5 W
acceleration(m/s/s)
1 1
5 0 5 0 0 - 5 0 . 5 1 1 . 5 2
- 1 - 1 - 2
0 5 0 t i m e ( s )
Receiver 2
Receiver 1
D1
D2
Tunnel
Blasting
15 10 acceleration(m/s/s) 5 0 0 -5 -1 0 -1 5 -2 0 t im e ( s ) 0 .5 1 1 .5 2
(not to scale)
20
Numerical analysis Blasting carried out towards and away from the slope Different nodes on the soil slope and in the rock are monitored
Soil
Blasting point
6
Rock
21
Need to combine Dynamic Stability & Deformation Analysis QUAKE/W computes the static plus dynamic ground stresses at specified intervals during an earthquake. SLOPE/W can use these stresses to analyze the stability variations during the earthquake and estimate the resulting permanent deformation. This approach is based on the assumption that there will be some permanent movement of the sliding mass during short instances in time when the factor of safety falls momentarily below 1.0. An average acceleration for the entire sliding mass that makes the factor of safety equal to 1.0 is known as the yield acceleration. When the average acceleration is greater than the yield acceleration, the sliding mass will move.
22
Full frequency spectrum can be studied Deformation of slope can be modeled Factor of slope of critical slip surface can be calculated as a function of time
23
Average Mass Acceleration For each trial slip surface, SLOPE/W determines the total mobilized shear arising from the dynamic inertial forces. This dynamically driven mobilized shear is divided by the total slide mass to obtain an average acceleration. The procedure is to integrate along the average acceleration versus time record to find the velocity during the times that the acceleration is greater than or equal to the yield acceleration.
Diagram illustrating the Newmark method. (a) Acceleration versus time; (b) velocity versus time for the darkened portions of the acceleration pulses; (c) the corresponding downslope displacement versus time in response to the velocity pulses. (After Wilson and Keefer 1985.)
24
Numerical Simulation using Field Data from the Eagles Nest Tunnel
25
26
27
Slope Berm 2
South Portal
Tunnel Face
Slope Crest
Slope Berm 3
Slope Berm 2
HKU sensor
HKU sensor
Slope Berm 1
HKU sensor
29
Accelerometer
30
Slope Berm 3
Slope Berm 2
South Portal
1 Slope Berm
VE101 On Slope
Contractor sensor
31
North Portal
Measured PPVs are too low for analysis because of the WSD Bend B tunnel
32
Slope Crest
HKU sensor
Slope Berm 3
Slope Berm 2
Slope Berm 1
33
Contractor sensor
34
HKU sensor
35
36
Scrutinizing Data in relation to: Geology Tunneling Direction/Progress Boundary Condition Others so that only quality data (R2>75%) is used for comparisons with simulation that has similar boundary conditions
37
1050
1100
1150
1200
1250
1300
1350
100
10
29-Nov-04 08-Jan-05 17-Feb-05 Date 29-Mar-05 08-May-05
20
30
40
50
60
70
80
90
17-Jun-05
27-Jul-05
North Bound CH 1155 Closest Distance 10.5 m 16 March 2005 Charge Wt per Delay 0.2747 kg
1000 1010 1020 1030 1040 1050 1060 1070 1080 1090 1100 1110 1120 1130 1140 1150 1160 1170 1180 1190 1200 1210 1220 1230 1240 1250 1260 1270 1280 1290 1300 1310 1320 1330 1340 1350 1360 1370 1380 1390 1400 1410 1420 1430 1440 1450 1460 1470 1480 1490 1500
05-Sep-05
North Bound
South Bound
1000
y = 800.44x-1.4194 R2 = 0.6506
Scaled Distance Vs PPV North Bound 13 Nov. 2004 to 15 March 2005 Tunnel Approaches Receiver only
1000
For North Bound Tunnel, correlation increases if data approaching receiver is used only
10
y = 2620.4x 2 R = 0.893
-1.6842
39
Sensor
Solid Face
Solid Face
Free Face
40
100
10
20-O ct-04 29-Nov-04 08-Jan-05 17-Feb-05 Date 29-Mar-05
20
South Bound Top Heading
08-May-05
30
40
50
60
70
80
90
17-Jun-05
27-Jul-05
1000 1010 1020 1030 1040 1050 1060 1070 1080 1090 1100 1110 1120 1130 1140 1150 1160 1170 1180 1190 1200 1210 1220 1230 1240 1250 1260 1270 1280 1290 1300 1310 1320 1330 1340 1350 1360 1370 1380 1390 1400 1410 1420 1430 1440 1450 1460 1470 1480 1490 1500
05-Sep-05 North Bound South Bound
41
10
y = 556.72x 2 R = 0.3306
-1.3161
Scaled Distance Vs PPV South Bound 16 Nov. 2004 to 9 May 2005 Tunnel Approaches Receiver only
10
y = 5527.9x 2 R = 0.3905
-1.8843
For South Bound Tunnel, correlation does not increase even if data approaching receiver is used only
42
Date
Sensor
Free Face
A free face is always present in between the South Bound Tunnel and the Sensor
43
Date
100 90
Distance from Tunnel Face to Receiver (m)
80 70 60 50 40 30 20 10 0 1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 1200 1220 1240 1260 1280 1300 1320 1340 1360 1380 1400 1420 1440 1460 1480 1500
South Bound CH 1149.5 Closest Distance 17.8 m 9 May 2005 Charge Wt per Delay 0.549 kg VE101 Soil Slope S f li d t l
44
y = 51.603x 2 R = 0.1993
-0.6506
Scaled Distance Vs PPV South Bound 16 Nov. 2004 to 9 May 2005 Tunnel Approaches Receiver only
1000 VE101 Soil Slope Surface
For South Bound Tunnel, correlation increases if data approaching receiver is used only
100 P P V (m m /s)
10
y = 604.75x
2
-1.2328
R = 0.7504
1
45
Sensor
46
Date
80
Top Bench
70
Lower Bench
60 50 40 30 20 10
1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 1200 1220 1240 1260 1280 1300 1320 1340 1360 1380 1400 1420 1440 1460 1480 1500
47
10
y = 44.39x-0.7474 2 R = 0.3261
Scaled Distance Vs PPV North Bound 13 Nov. 2004 to 16 March 2005 Tunnel Approaches Receiver only
1000
10
y = 218x-1.0823 R2 = 0.3347
For North Bound Tunnel, correlation does not increase even if data approaching receiver is used only
48
Free Face
Sensor
A free face is always present in between the North Bound Tunnel and the Sensor 49
For VE106 (inside Brick Lined Tunnel) and VE103 (in front of Village House), vibration level recorded was too low for any good correlation.
50
Scaled Distance Vs PPV South Bound 16 Nov. 2004 to 9 May 2005 Tunnel Approaches Receiver only HKU Data Added
1000
Close range
100 PPV (mm/s)
VE101
Slope
y = 604.75x 2 R = 0.7504
-1.2328
Similar
Far range
10
HKU Data
2
y = 1482x
-1.2093
R = 0.7662
1
Scaled Distance Vs PPV South Bound 16 Nov. 2004 to 9 May 2005 Tunnel Approaches Receiver only HKU Data Added
1000
It could be very deceiving when a straight line is forced to go through data at different ranges in a loglog plot
Contractor+HKU Data
PPV (mm/s)
100
10
y = 2469.6x 2 R = 0.7781
-1.5665
0.1 1
10
100
1000
51
Scaled Distance Vs PPV North Bound 13 Nov. 2004 to 15 March 2005 Tunnel Approaches Receiver only
Receiver in Rock
1000
Range 10 20 mm/s
10
y = 2620.4x
2
-1.6842
R = 0.893
1
Scaled Distance Vs PPV South Bound 16 Nov. 2004 to 9 May 2005 Tunnel Approaches Receiver only HKU Data Added
It would also be very deceiving when simply look at the coefficient of a straight line without looking at the range of the data at different material
Similar
Receiver in Soil
1000
Range 10 100 mm/s
PPV (mm/s)
100
Contractor+HKU Data
Similar
10 1
y = 2469.6x 2 R = 0.7781
-1.5665
0.1 1
10
100
1000
52
Numerical Modeling should match the actual range of data in natural scale and examine the trend closely
PPV vs Scaled Distance
120
100
80 PPV (mm/sec)
60
40
20
53
110
y = 2620.4x
2
-1.6842
R = 0.893 y = 2556.2x
2 -1.2282
R = 0.9548
Coefficients appear good match, but the trend line in numerical model appears higher. This is due to controlling too small vibration in sensitive structures and not recording close range vibration in rock.
54
y = 2469.6x
-1.5665
R = 0.7781
PPV (mm/sec)
80
QuakeW Generated in Soil
2
y = 2090.1x
-1.6434
R = 0.7662
Coefficients and trend line appear good match. This is due to large vibration obtained through HKU data (sensors placed at close range).
55
20 PPV (mm/sec)
y = 604.75x 2 R = 0.7504
-1.2328
15
y = 289.03x 2 R = 0.5344
-1.1499
10
Coefficients and trend line appear good match with Mines and Quarries Data. i.e., Mines and Quarries Data did not measure the full range of blasting vibration
1.22
50 Percentile
56
Stiffness Model Constant Young's Modulus (E) Poisson's Ratio 0.33 20000 kPa
Material Model
Stiffness Model Constant Young's Modulus (E) Poisson's Ratio 0.33 2000000 kPa
57
Ensure Initial Stresses are correctly generated based on unit weight of material
58
Change boundary condition for dynamic analysis Ensure boundary is far away from input vibration source (tunnel blast)
59
Blasting Source
Blasting Source
60
Average Charge Weight Per Delay is used = 0.929 kg (range from 0.275 kg to 2.308 kg)
61
2.5 m An assumption representing head and bench blasting. Assume blast induced vibration zone 2.5 m representing full face blasting zone (14 m diameter). Trial and error to match the measured vibration.
2.0 m (reasonable)
62
Parameter Material Model P Fn. # G Fn. # 1 (CDG) Linear Elastic 0 0 Constant 20000 kPa 0.33 2.5%
Variable by trial and error to match the measured vibration.
63
64
HKU sensor
Time (sec)
Vibration data magnified by 2.5 to represent the blasting source (trial and error to match the measured vibration)
60
40
20
-20
-40
-60 0 1 2 3 4 5 6
65
Time (sec)
Numerical analysis
Vibration received
25 20
Time (sec)
Soil
2
60
3
40 20 0 -20 -40
6
Rock
Blasting source
X-Velocity (m/sec) (x 0.001)
-60 0 1 2 3 4 5 6
Time (sec)
66
1200 1000
Measured in Rock
PPV (mm/sec)
800
QuakeW Generated in Rock
-1.6842
100
110
120
130
140
150
67
y = 2311.2x
2
-1.5594
R = 0.7517 y = 2090.1x
2 -1.6434
150
QuakeW Generated with Soil Damping Ratio 0.3
R = 0.7662
100
50
68
Displacement vectors direction reversal Sign reversal of displacement vector at Time Step 16. Displacement to the left and right along some vertical sections
69
70
71
Carry out Slope Analysis using the Dynamic Stress Generated at Every Time Step
Locate critical slip surface for each blast location for all time steps
72
Stability Factor = 1.374 (Before Blasting) Stability Factor = 1.363 (During Blasting) Drop in Stability Factor = 1.0%
Blast Location
73
Stability Factor = 1.374 (Before Blasting) Stability Factor = 1.363 (During Blasting) Drop in Stability Factor = 1.0%
1.38
Stability Factor
1.36
1.34
1.32
1.30 0 1 2 Time 3 4 5 6
74
Stability Factor = 1.374 (Before Blasting) Stability Factor = 1.352 (During Blasting) Drop in Stability Factor = 2.0%
Blast Location
75
Stability Factor = 1.374 (Before Blasting) Stability Factor = 1.352 (During Blasting) Drop in Stability Factor = 2.0%
1.36
Stability Factor
1.34
1.32
1.30 0 1 2 Time 3 4 5 6
76
77
Summary of Findings: Preliminary assessment of Eagles Nest Tunnel suggests the following: 1. 2. 3. 4. 5. 6. 7. 8. Data must be scrutinized to take into account the geology, free surfaces between the source and the receivers. Should examine the data instead of providing a trend line to a log-log plot. K and in Rock is higher than that in Soil. Monitoring of vibration close to blasting source is needed to fully understand the attenuation law. Attenuation law changes rapidly when data is recorded close to the source. Numerical results, based on calibration of field data, provide good match with the measured attenuation law. Soil is a compliant system (different points within the system move differently and out-of-phase from each other) Preliminary numerical analysis shows that the reduction of FOS for slope due to tunnel blasting is small as compared to the analysis assumed in GEO Report No. 15.
78
9.
79