Detonation Velocity: Its Significance
Claude Cunningham ASIEX Explosives Workshop Puerto Varas October 2012 Claude Cunningham: Blasting Investigations and Consultancy
1
Why is Detonation Velocity important? •
VoD a critical aspect of explosive characterisation – The only direct report from the detonation front – Defines non-ideal performance – Vital role in verification of detonation modelling
•
Problems in obtaining clean VoD data – How do we handle puzzling results?
•
Claims Claim s that lower VoD = worse breaking breaking or bette betterr breaking! breaking! – Debates over relevance to energy delivery
•
Use of VoD in blast design, e.g. decoupled pressures – Pb = ρ D / 8 2
Claude Cunningham: Blasting Investigations and Consultancy
2
Why is Detonation Velocity important? •
VoD a critical aspect of explosive characterisation – The only direct report from the detonation front – Defines non-ideal performance – Vital role in verification of detonation modelling
•
Problems in obtaining clean VoD data – How do we handle puzzling results?
•
Claims Claim s that lower VoD = worse breaking breaking or bette betterr breaking! breaking! – Debates over relevance to energy delivery
•
Use of VoD in blast design, e.g. decoupled pressures – Pb = ρ D / 8 2
Claude Cunningham: Blasting Investigations and Consultancy
2
Emul Em ulsi sion on vs vs Wa Wate terg rgel el fo forr les less s fin fines es:: Qua Quart rtzi zite te
EmulsionB Explosive
Watergel A Explosive
Claude Cunningham: Blasting Investigations and Consultancy
3
Blast 1: High cup density VoD's
VoD measurements..
6000
Average 5170 m/s
5000 4000 1.18
s / 3000 m
Recovery
2000
Average 4860 m/s
1000 0 1
2
3
4
5
6
7
8
9
Blast 2: Low cup density VoD's
10 6000 5000 4000 s / 3000 m
2000
Average 4970 m/s
1000
Blast 3: Low cup density VoD's 6000 5000 4000 s / 3000 m
2000 1000 0 1
2
3
4
5
1.05
6
7
0 1
2
3
4
5
6
7
8
9
10
11
12
Findings: High density decreases sensitivity 1.05 Low density reduces VoD More rows reduces VoD. WHY? Claude Blasting DOES IT MATTER? 4 WHEN 8 9 Cunningham: 10 Investigations and Consultancy
Road map •
Parameters affecting VoD – Ideal detonation – Non-ideal detonation
•
Difficulties in modelling of Non-ideal – Consistency of commercial explosives performance – Field conditions – VoD system issues – Limitations of Kinetic modelling
•
Handling of VoD problems Claude Cunningham: Blasting Investigations and Consultancy
5
Ideal Detonation and VoD Shock front
Pressure falls with Particle Velocity
Decelerating particle velocity
Sonic (CJ) plane
VoD
Explosive
Reaction zone = DDZ
Reaction complete
Ideal detonation Uniaxial flow Reaction complete at Sonic plane.
Vixen_n ZND: ANFO density 0.8 12
100% 90%
10
s / m k , a P G
Gas vel km/s Pressure GPa Reaction
80% 70%
8
60% 6
50% 40%
4
30% 20%
2
10% 0
0%
0
n o i t c a e r f o e e r g e D
Conversion of chemical energy ANFO at density 0.8. Ideal VoD 4.8 km/s Detonation Pressure 5 GPa. Fluid velocity 1.3 km/s at Sonic Plane Kinetic Energy 0.82 MJ/kg
Claude Cunningham: Blasting 10 100 Investigations and Consultancy mm from Shock Front
1
Pstag Normal start for energy
6
on!ideal detonation Curved Shock Front
Sonic (CJ) locus
VoD
Decelerating particle velocity
Reaction zone
DDZ Explosive
Expansion
WK solution
Vixen_n ANFO 0.8, 100mm dia, Kimberlite
Lateral Loss
8
100%
Pressure GPa Gas vel km/s Reaction
90%
7
80%
6
70%
5
60%
4
50%
3
40%
s / m k d n a a P G
30%
2
20%
1
10%
0
0%
0
1
10 mm from Shock Front
100
n o i t c a e R f o e e r g e D
ANFO at density 0.8 Ideal VoD 4.8 km/s Detonation Pressure 5 GPa. Gas velocity 1.3 km/s at Sonic Plane. Non-Ideal VoD 3.85 km/s Detonation Pressure 3 GPa. Gas velocity 0.94 km/s at Sonic Plane, where 16% of the explosive is still to react. The DDZ is inside the reaction zone.
Claude Cunningham: Blasting Investigations and Consultancy
7
Effect of Ideality on Impulse Detonation velocity points to significantly different pressure profile on the blasthole wall ANFO, Density 0.8 - Ideal and Non Ideal, Kimberlite
12 Ideal ZND ZND end reaction, CJ point 312mm Non Ideal 312mm CJ point 312 End reaction 100mm Non Ideal 100mm CJ point 100mm End Reaction
10
8 a P G e r u s s e r P
4.8 km/s
6
No impulse before CJ 4
4.44 km/s
Increasing impulse before CJ, but lower 2 pressure
3.85 km/s
0 0
50
100
150
200
250
Distance mm
Claude Cunningham: Blasting Investigations and Consultancy
8
Ideal VoD: influence of energy and density Ideal Detonation (Vixeni)
4.5 ANFO Energy MJ/Li
1
4.0 ALANFO Energy MJ/Li
2
3.5 Emulsion Energy MJ/Li
3.0
y g r e n E d n a D o V
3
2.5 2.0 1.5 1.0
Energy: ranking of common explosives
0.5 0.0 0.5
0.6
0.7
0.8
0.9
1.0
Claude Cunningham: Blasting Density g/cc Investigations and Consultancy
1.1
1.2
9
Ideal VoD: influence of energy and density Ideal Detonation (Vixeni)
7.0
ANFO VoD km/s ALANFO VoD km/s Emulsion VoD km/s
6.0
ANFO Energy MJ/Li
2 1 3
Spread within exptl error
ALANFO Energy MJ/Li
5.0
Emulsion Energy MJ/Li
y g r e n E d n a D o V
1 2
4.0 3.0
3
2.0
VoD: Does not rank with energy Strong tie to density
1.0 0.0 0.5
0.6
0.7
0.8
0.9
Density g/cc
1.0
1.1
1.2
10
Ideal detonation: non!intuitive conclusions
• Density the overriding influence on VoD • Emulsions not intrinsically faster than ANFO • Higher energy at same density does not guarantee higher VoD
Claude Cunningham: Blasting Investigations and Consultancy
11
Why lo"er VoD "ith higher Ideal energy? Energy Delivery with ALANFO10 vs ANFO, same density
10.0
Crossover point
pressure
ANFO Higher PCJ = higher VoD 1.0
energy
i L / J M , a P G
ALANFO GPa ALANFO MJ/Li ANFO GPa ANFO MJ/Li
0.1
ALANFO Hotter gas = more energy
20 MPa 0.0 0
1
10 Volume, cc/g
ANFO is cooler and pressure drops more quickly with expansion.
100
12
on!intuitive increase of Weight #trength Ideal ANFO Energetics - Vixen_i "ith density $%&'( 10 IDE MJ/kg VoDi km/s Pcj GPa Heat of Reaction MJ/kg
5
HoR constant
0 0.6
0.7
0.8
0.9
1
1.1
Density g/cc
13 Ideal detonation: Weight strength, VoD and CJ pressure increase with density
Why does Ideal energy)*g increase "ith density?
•
CJ Pressure (& VoD) increases, but temperature falls
•
Faster initial delivery of energy, leaving cold gas for late expansion
•
Lower density has hotter gas and more energy below 20MPa, but we’re not counting that.
•
So higher density gives more usable energy, higher ideal VoD
•
This appears to confirm the opinion that “higher VoD means more Bang per Buck”.
•
But the rock does not necessarily respond well to such high pressures, and reaction is anyway non-ideal. Claude Cunningham: Blasting Investigations and Consultancy
14
on!ideal energy and VoD
•
Fundamentally different to Ideal detonation
•
Reaction energy “leaks” from DDZ owing to slow reaction and divergence.
•
VoD is good indicator of energy released in DDZ.
•
Critical diameter reached when insufficient reaction energy is sustaining shock front because of DDZ drainage from sides and rear.
•
Reaction energy can still be released, but not easily accounted for….
Claude Cunningham: Blasting Investigations and Consultancy
15
%&' in on!ideal detonation
Critical VoD input Ideal VoD (Det Code)
Modelled from kinetic fitting Critical diameter unconfined
Kinetic fitting (reactivity-voidpressure)
Claude Cunningham: Blasting Investigations and Consultancy
16
Effect of confinement on reaction
Non-Ideal performance parameters for ANFO at density 0.8 g/cc at 69mm diameter in different confinements Strong Rock
(Ideal)
Unconfined
Weak Rock
(100%)
34%
74 %
84%
VoD km/s
(4.80)
2.17
3.50
3.85
P CJ GPa
(4.8)
0.71
3.01
3.21
Reaction in DDZ
VoD only gives good data before Sonic (CJ) Point.
Claude Cunningham: Blasting Investigations and Consultancy
17
Direct lin* of VoD and DD+ reaction
ANFO in Weak rock (Vixen_n)
5.0
100% 90% 80% 70%
4.0
60%
s / m k D o V
50% 40%
3.0
VOD km/s Reaction at CJ
30%
% , e n a l p J C t a n o i t c a e R
20%
Kinetic model calibrated using VoD tests
2.0 10
100 Hole diameter, mm
10% 0%
1000
18
,Validation dependence on field VoD-s..
MicroTrap VOD Data
MicroTrap VOD Data 54.0
33.5
53.5
33.0 53.0
32.5 52.5
32.0
52.0
4859.1 m/s
4529.2 m/s 31.5
51.5
) m (
) m (
e 51.0 c n a t s 50.5 i D
e 31.0 c n a t s 30.5 i D
50.0
30.0
49.5
29.5 49.0
29.0
48.5
28.5
48.0
0.00
0.25
0.50
0.75 Time (ms)
1.00
1.25
1.50
53.00
53.25
53.50
5 3.75 54.00 Time (ms)
54.25
54.50
54.75
Need to be aware of factors affecting data collection Claude Cunningham: Blasting Investigations and Consultancy
19
/ro0lems "ith non!ideal modelling
•
If unconfined VoD’s are wrong, whole process is skewed: – Formulation variance in test sample – Confinement variance (pipe thickness etc) – Capture of loading data (density, noise) – VoD system issues (probes, resolution) – Interpretation of traces (scatter, dropouts) – Kinetic model limitations (complexity)
•
Energy release interpretation – is low VoD internally (composition) or externally (confinement) caused? Claude Cunningham: Blasting Investigations and Consultancy
20
&ormulation variance Fuel Oil Tests in ANFO: Spec 5.8% to 6.2%
7 Variation of ANFO with AN density, Nominal AN 0.73 g/cc
0.00
0.81
6 Density Oxygen bal
0.80
l i O l e u F %
0.80 0.79
c c / g 0.79 d e c u 0.78 d o r p 0.78 y t i s n 0.77 e D
5
0.77
4
0.76
-0.20
% g n -0.40 i t l u s e r e -0.60 c n a l a b Nominal ANFO n g -0.80 e Density y x O 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Day
-1.00
0.76 0.75 0
5
10
15
20
Variation in fuel content of ANFO -1.20 affects 25 30 35 strength, and VoD.
Day
Volumetric mixing: varying bulk AN density affects column rise and oxygen balance of ANFO or of explosives using the AN.
21
%&' strength and oil content ANFO mixes 105
10
100 y g r e n E e v i t c e f f E e v i t a l e R
5
95 90
0
85 80
-5
75 REE % OB, %
70
% e c n a l a B n e g y x O
-10
65 60
-15 3.0
4.0
5.0
6.0
7.0
8.0
9.0
% Fuel oil
Ideal code - a useful guide to the relative performance of like explosives. Variable density causes unintended changes in Oxygen Balance.
22
#catter in 1oading data
•
Bulk systems, operational factors will deliver some variation.
•
Volumetric metering of mixes – solids, flow processes affect density, formulation
•
Auto-compression/ adhesion in hole
•
Cooling of hot explosive
•
Loading into water
•
Variable hole diameter – eg, 118 to 132mm bits, cavities, friable zones
•
Only reasonably certain measurement is column rise... 23
Apparent loading densities - 115mm holes, same blast
17
1.70
16
1.60
Runaways?
15
1.50
14
1.40
13
1.30
m / g k
12 11
1.20
Nominal density in hole
1.10
Gauge loss?
kg/m effective Implied Density
10
c c / g
1.00
9
0.90
8
0.80 1
2
3
4
5
6
7
8
9
10
24
2onfinement variance
• What does “unconfined” mean? • Indication that we have underestimated effect of plastic tubing in “unconfined” tests. • Significant stratigraphic variation
Claude Cunningham: Blasting Investigations and Consultancy
25
VoD system pro0lems
• Probes and cables – nicks, faults, sensitivities • Collector units – robustness, calibration, dropped data, batteries • Readings – noise, drop-outs, impossible results, eg cable cut at right time but apparent wrong depth
Claude Cunningham: Blasting Investigations and Consultancy
26
2hoosing the num0er to 3uote..
Is 5.3 km/s the correct value to use?
27
Interpretation of VoD traces
• Variation, noise and drop outs • Fraction of trace sensible • Curve fitting for VoD counter-productive • Density/ sensitivity equivalence between holes may be not be same, as assumed
Claude Cunningham: Blasting Investigations and Consultancy
28
4races from %&' in platinum reef
Narrow Reef VoD Tests 1.0 0.9 0.8 0.7 0.6 s e r t e 0.5 m
A
0.4
C
B D
0.3
E F
0.2
G
0.1 0.0 0
50
100
150
200
250
microsec
300
Claude Cunningham: Blasting Investigations and Consultancy
Look closer at trace C...
350
400
450
29
5igh sampling rate and noise: trace 2 Hole C: Point by point 60
+38km/s to -65km/s range
40
20
s 0 / 0 m k , D o V -20
50
100
150
200
250
300
350
400
450 Series1
-40
-60
-80 Time, mics
Claude Cunningham: Blasting
1 MHz sampling rate Investigations and Consultancy
30
Digital noise and curve fitting ! 2 Hole C: Point by Point
Polynomial fit poor
5.0
4.0
) 3.0 s / m k ( D o V
VoD (apparent) Poly. (VoD (apparent))
2.0
Byte resolution
1.0
0.0
0
0.2
0.4
0.6
0.8
Distance (m) Claude Cunningham: Blasting Investigations and Consultancy
1 31
VoD derivation 0y averaging over span
Hole C: Effect of Point to Point Smoothing 5.0
Emulsion 3.9
4.0
3.0 s / m k
20 pts 50 pts 100 pts
Average for this hole 2.44 km/s
2.0
ANFO 2.0
1.0
0.0 0
0.2
0.4
0.6
metre Claude Cunningham: Blasting Investigations and Consultancy
0.8
1 32
VoD readings •
Probably 30 to 50% of readings are consistent - depending on operator and conditions.
•
Learning happens with each attempt - need to use judgement for initial data, and subsequent refining of database.
•
Error bars nice but often misleading - need to know what is causing the deviation – is it random error, or a specific influence which should not be permitted? – Room for ongoing discussion and learning.
•
Kinetic model, and evaluation of this has to be based on the chosen VoD’s. Therefore focus is on progressive learning in both areas. Claude Cunningham: Blasting Investigations and Consultancy
33
Unconfined Doped emulsion
7.0 Density 1.2 1.08
6.0
s / m k D o V
5.0
4.0
?
3.0 0.000
0.002
0.004
0.006
0.008
0.010
0.012
inv dia, mm
Claude Cunningham: Blasting Investigations and Consultancy
34
6inetic 7odel limitations
• Advanced modelling needed to understand energy flows in DDZ – Hugoniot data for explosive particles and rock confinement. – Reactivity on micro-scale. • Critical diameter is crucial to calibration but extremely sensitive. • Deviation in VoD prediction doesn’t mean energy is ultimately not delivered. • A real problem: reluctance to share details of progress for Peer Review 35
2onclusions •
Problems both with theoretical modelling of detonation and usable VoD measurements. – Calibration affected
•
VoD says how much reaction energy was released in DDZ – not how much was released in Reaction Zone, or how long reaction lasted.
•
Detonation modelling and Blast modelling feed into each other – Detonation affects rock, rock affects detonation
• Higher VoD does mean more “banked” energy. • No perfect models yet – But some much better than others
36