Download (wa7) final proposal

Bhavita Patel
Professor Brenda Moore
Eng 352 – 453
16 July 2008
Inorganic Nanomaterials in Advanced Energetics:
Parametric Study of Ignition of Metal Powders by Electro-Static Discharge
Patel 2
Executive Summary
Electro-static discharge (ESD) stimulation is commonly used to understand the
ignition sensitivity of powders and the safety of handling powder. Despite this, the ignition
mechanisms of powder by electric spark are still unclear. This paper represents the study and
reports an experimental investigation of ESD ignition of spherical metal powders. This paper
describes measurements obtained from the emission of ignited powder as well as the spark
current and voltage in real time. The experiment model was built based on a commercially
available apparatus for ESD ignition sensitivity testing. The spark duration was of the order
of a few µs. The spark current and voltage were always observed to have significant
alternating current (AC) components. The optical emission was filtered to separate the
signals produced by the spark plasma and by the heated and igniting powder. The radiation
signal produced by the igniting powder was always delayed after the spark. The delay time
for Magnesium (Mg) decreased from about 2.5 to 0.5 ms as the spark energy increased from
10 to 60 mJ; the delay for Mg remained nearly constant when the spark energy continued to
increase to over 100 mJ. The delay time for Aluminum (Al) ranges from 0 to 8.0 ms as the
spark energy increased from 30 to 120 mJ. In conclusion, Al powder of 3.0-4.5 µm has
smaller ignition delay time compared to Al of 4.5-7.0 µm.
Patel 3
List of Figures
Figure 1. Schematic diagram of the experimental setup.
Figure 2. Digital Storage Oscilloscope.
Patel 4
S park-emi s s i on.pgw
Emission Intensity, a.u.
O2
N2
N2
Spark over steel substrate
O2
N2
O2
N2
Igniting Mg powder
MgO MgOH
MgO
200
300
400
500
600
Wavelength, nm
700
800
Figure 3. Emission spectra of the sparks between the pin electrode and an empty
steel sample cup (top) and between the pin electrode and igniting Mg powder
(bottom).
Figure 4. Electrode cups (15 mm diameter) containing Mg powder; left cup contains
powder that has not been struck by a spark, middle cup contains powder that has
been struck by a spark and has not ignited, and the right cup contains powder that
has been struck by a spark and has ignited.
Patel 5
Emission signal, V
12
Raw Derivative
10
8
6
4
2
0
0
10
20
30
40
50
60
70
Time, ms
Emission signal, V
Figure 5. An emission signal produced by a spark-ignited Mg powder. A short spike
just before 10 ms indicates the spark.
6
5
Raw Derivative
Slope at the peak of
the derivative
Ignition delay
4
3
Spark
2
Zero-level signal
1
Derivative, V/s
0
1500
First peak
of the signal derivative
1000
500
0
9.0
9.5
10.0
10.5
11.0
11.5
12.0
Time, ms
Figure 6. Illustration of the signal processing used to determine ignition delay.
Patel 6
Ignition Delay, ms
3.0
Recovered2
2000 pF - 0.2 mm - (6 - 16 kV, in 2 kV steps)
5000 pF - 0.2 mm - 8 kV
10000 pF - 0.2 mm - 8 kV
2000 pF - 1.5 mm - 8 kV
5000 pF - 1.5 mm - 8 kV
10000 pF - 1.5 mm - 8 kV
2.5
2.0
1.5
1.0
0.5
0.0
0
20
40
60
80
100
120
Measured Spark Energy, mJ
Figure 10. Ignition delay as a function of the measured spark energy of Mg
Powder
Figure 11. Ignition delay as a function of the measured spark energy of Al Powder
Patel 7
Table of Contents
1. Introduction……………………………………………………….…………Pg.8 - 9
2. Statement of problem.………………………………………………………. Pg.9 -10
3. Definition of Terms………………………………………………………….Pg.10 -13
4. Plan of Action.......…………………………………………………………..Pg.13 -14
5. Experimental Design………………………………………………………...Pg.14 -16
6. Results and Conclusion……………………..…………...……………….....Pg.16-19
7. References...…………………………………………………………………..Pg 20-21
Patel 8
Introduction
The term “Thermal ignition”, according to Irvin Glassman, as an explosive condition
comes into existence due to the differences between the rates of thermal energy released and
the thermal energy lost. Thermal ignition is widely used to understand the mechanism of
various energetic materials which are used as fuels. Energetic materials can be tested by
different methods. Electro-static spark is a common ignition stimulus and a variety of tests,
standards, and evaluation methodologies are developed to investigate ignition behavior of
different substances struck by a spark.
Electro-static discharge ignition for flammable gases was studied in great detail and
many important trends were established. Reviews by Mellor et al.,1-2 published in 1990’s
described these trends and suggested that a similar systematic study would be necessary to
establish a scientifically sound test of ESD sensitivity of powders. Ignition can be affected
by many parameters of electrode materials; electrode shape, discharge duration, material, and
other similar parameters are investigated and recorded. However, such a study has never
been undertaken.
Metal powders are used as fuel additives in propellants, explosives, pyrotechnics and
many other applications. The biggest advantage of metal powders is the property of high
energy density, but with the increase in its reaction rate, the sensitivity of the metals also
increases. At the same time, an experimental test of ESD ignition sensitivity is one of the
most commonly used to evaluate a variety of powdered materials, including agricultural3,
food4, textile5, pharmaceutical6, metallic7-9, and of course, energetic components10-11.
Patel 9
Typically, a spark will be generated by a high-voltage capacitor to discharge between the pin
electrode and the powder bed. The minimum capacitor energy at which the powder ignites is
specified as the minimum ignition energy (MIE), a parameter defining the sensitivity of a
powder to ESD ignition stimulation.
STATEMENT OF PROBLEM
Electro-static Discharge testing offer a qualitative ranking of ESD sensitivity between
different powders by comparing MIE. According to Skinner ref. 11 , it was suggested that the
spark represents primarily a thermal source capable of raising the temperature of a flammable
powder above the point at which thermal runaway occurs. However, it remains unclear how
the spark heats the powder particles, which portion of the spark energy is being transferred to
the powder and by which mechanism. For example, the spark’s plasma can heat the powder
surface directly while the current of the spark discharge can result in Joule heating distributed
in the powder volume, along the current path. The concept remains unclear whether the
polarity of the spark discharge is a factor affecting the ignition energy.
The transport properties of powders including the thermal and electrical
conductivities are primarily affected by the contact resistances of the particle. As stated in a
report by ref12-13, there is a dramatic change in the transport properties of a powder that is
affecting the interactions of the particle by a spark. The spark itself does not have a steady
heat source and the energy distribution is expected to change with respect to time and its
location. However, it is hard to determine what is affecting the ignition of various materials.
In order to resolve this problem, understanding the mechanisms of ESD ignition of powder
becomes very important. In addition, understanding the mechanism of ESD ignition is
Patel 10
important due to dramatic increase in new research which aimed to synthesize new powders
such as variety of nanomaterials used in different applications.
The goal of this work is to instigate a systematic study of the mechanisms of ESD
powder ignition. The overall research study is analyzed as a combination of systematic
experimental and modeling efforts eventually resulting in quantitative mechanisms
describing ignition of various powder-like materials by electric sparks. This paper represents
the first part of this study and reports an experimental investigation of ESD ignition of
spherical metal powders. The paper describes measurements obtained from the emission of
ignited powder as well as the spark current and voltage. Mg was selected for this study
because there are relatively reliable descriptions for its thermal ignition available in the
literature16-17. Thus, it is expected that the results of the measurements reported in this paper
would be useful for quantitative modeling of the ESD stimulated powder ignition in which
the thermal reaction in the powder is described with practical accuracy.
Definition of Terms

Ignition: The term Ignition means initiation of combustion, evidenced by a flame, or
explosion. Thermal ignition is also defined in the introduction section.

Electro-static Discharge (ESD):
Electro-static discharge (ESD) is the release of
static electricity when two objects come into contact. This occurs when two different
materials rub together. One of the materials becomes positively charged; the other
becomes negatively charged. When that charge comes into contact with the right
material, it is transferred and we have an ESD event. The heat from the ESD event is
Patel 11
extremely hot, although we do not feel it when we are shocked. The ESD may also
damage the device and shortens the life of the device.

Electro-static Sensitivity: The degree to which a component or device is susceptible
to damage by electrostatic discharge.

Minimum Ignition Energy (MIE): It is the minimum energy required break the gap
between the powder and the pin electrode resulting in ignition of a powder. The
spark can only be observed if the voltage is high enough and the gap is small enough.

Firing Test System: This is a basic setup for LRC series circuit which is used in our
experiments. Refer to Fig. - 1 for the diagram model.

Voltage: Voltage is a representation of the electric potential energy per unit charge.
In other words, it is a measurement of the energy contained within an electric field, or
an electric circuit, at a given point. Voltage is a scalar quantity. The SI unit of
voltage is the volt, such that 1 volt = 1 joule/coulomb.

Electrical Current: Electrical current is a measure of the amount of electrical
charge transferred per unit time. It represents the flow of electrodes through a
conductive material. Current is a scalar quantity. The SI unit of electrical current is
the ampere, defined as 1 coulomb/second.

Ignition Delay time: The term is used to define the delay time for a powder to
ignite. It is directly affected by the thermal properties of the material used. The
calculation of ignition delay time can be varied by the types of criteria used to
determine the value. For example, the criteria used for this research will determine
Patel 12
the delay time value as the difference in time between the delay end and the spark
time.

Emission: Emission is a process of sending signal from one station on a network to
multiple stations on the network at the same time. Light emission is also used in our
study. It is the condition in which an electron falls back to a previous lower energy
level and, in doing so, emits light of a longer wavelength than the excitation light.

PMT’s: PMT’s are the interference filters used to separate the wavelengths of the
light emitted and the wavelength of the spark in our research. According to
Wikipedia, the definition of interference filter is it is an optical filter reflects many
spectral bands, which will filter by maintaining nearly zero coefficient of absorption
of the wavelengths of interest.

Electrode: An electrode is an electrical conductor that is used to make contact with a
nonmetallic part of a circuit, defined in Wikipedia. In simple words, it is a part used to
carry the electric current that can be discharged to another substance in a LRC circuit.

Oscilloscope: An oscilloscope is an electrical testing device used to view and record
different voltage signals emitted from the light. It is usually a two-dimensional graph
of one or more electrical potentials plotted as a function of real time. Refer to Fig. 2
for an image. The oscilloscopes are available in different types and are used
depending on its features and applications. The one used for our study is a Digital
storage oscilloscope.

Spark: It is the most common type of ESD testing. Spark is occurred when a where
is a high electric field which will create an ionized conductive channel in the air. In
Patel 13
our experiment, spark will only be produced if the voltage is high enough and the gap
is small enough between the electrode and surface of the powder.

AC signal: The word AC signal stands for alternating current signals. In simpler
words, this is one of the types of voltage and current conditioners that provide high
voltage minimize damaging voltage spikes. AC line also provides surge protection for
incoming power, data or phone lines, or hardwired systems.
Plan of Action
This section of the study will include technical approach to the problem, and the basic
modeling to conduct experiments. The technical approach is to perform an experimental
parametric study. Next step is to determine how ignition can be affected by the process
parameters. Lastly, establish a model that can be adequately describing experiments with
different ignition mechanisms.
Current protocols used for ESD ignition sensitivity testing, e.g., described in various
standards, e.g., 14-15 are not based on such understanding and the results are often inconsistent
between different testers and inconclusive in nature. Furthermore, the current protocols are
not suitable for a number of newly developed powder-like materials, which are not available
in the quantities required for the standardized testing. After reviewing the literature, we can
predict what can affect the ignition. Ignition can be affected by various materials, particle
size, applied energy (capacitance and voltage), varied applied voltages, pulse duration
(capacitance and resistance), and spark configuration which might be affected by the distance
or the gap.
Patel 14
The outline of experiments to be conducted is using Mg and Al powders measured at
different particle size and texture, and varied at different gap.
Our initial goal is to
accumulate experimental data for spherical Mg 10.3 µm and spherical Al (3.0-4.5) and (4.57.0) µm. After getting some realistic results, we can repeat the same experiments with
smaller weight for accurate analysis and characterization of powders.
Experimental Design
The powder used in the ignition experiments was 1-11 µm spherical Mg by Hart
Metals, Inc. The particle size distribution was measured using a Beckman-Coulter LS230
Enhanced Particle Analyzer. Based on the volumetric size distribution, the mean particle
size was 10.3 µm. A schematic diagram of the experimental setup is shown in Fig. 1. A
Model 931 Firing Test System (FTS) by Electro-Tech Systems, Inc. was used to generate
spark discharges. FTS includes a capacitor bank with capacitance varied in the range of 100
– 10000 pF. The capacitors can be discharged through the spark gap directly or through an
additional resistor varied from 500 to 5000 Ω. Additional resistors were not used in this
project. The capacitors can be charged to a voltage varied from 100 V to 26 kV.
The
polarity of the electric output can be changed so that the high-voltage electrode can be either
positive or negative. For a standard test, the powder is placed into a stainless steel cup (15
mm diameter, 3 mm deep) affixed to a grounded base and the high voltage pin-electrode (a
stainless steel needle) is placed 0.2 mm above the surface of the powder being tested. For
experiments in which optical and electrical measurements were made, a customized sample
holder was used. It was desired to perform experiments with a smaller amount of powder to
ensure ignition of all or nearly all the powder when the spark energy exceeded the ignition
threshold. It was also desired to avoid the possibility for the spark to strike the side surface
Patel 15
of the powder holder instead of the center of the powder bed. Therefore, the only conductive
surface of the customized sample holder was its bottom surface on which the powder was
placed. The sample holder was made from a 0.5-mm thick nylon washer with an inner
diameter of 6 mm affixed to an aluminum substrate by an epoxy resin.
Inductance coils by Pearson Electronics were used to measure spark current and
voltage drop across the pin electrode and sample gap. The current was measured using a
model 110A coil with a 1 V / 10 A ratio. To measure the voltage drop, a 1 kΩ high voltage
resistor was connected in parallel between the discharge pin and sample cup substrate. The
current through the resistor proportional to the spark voltage was measured using a model
4100 coil with a 1 V / 1 A ratio. Both current and voltage traces were visualized and
recorded by a LeCroy WaveSurfer 64Xs Series oscilloscope.
Optical emission produced by the spark and by the igniting powder was monitored in
real time. In preliminary measurements, optical spectra produced respectively by sparks
between the pin electrode and an empty sample cup and between the pin electrode and
igniting Mg powder were recorded using an EPP2000 spectrometer by Stellar net Inc. The
recorded spectra are shown in Fig. 3. The emission of the spark itself (generated over an
empty sample cup) was dominated by an ultraviolet peak (around 280 nm). Based on refs. 18,
19
, it was assigned to molecular oxygen. The emission of the igniting Mg powder was heavily
dominated by a broad black-body spectrum with one of the strong peaks observed around
500 nm, assigned to MgO. Respectively, interference filters at 280 nm and 500 nm were
selected to separate between the radiation signatures of the spark and igniting powder. A
bifurcated fiber optics cable with a single input window was used to split the optical signal
Patel 16
between two outputs. Each output was connected to a respective interference filter and a
photomultiplier.
In initial experiments, emission and electrical current traces were acquired for sparks
striking solid metal substrates.
The distance between the pin electrode and the substrate
surface was maintained at 0.2 mm.
The spark polarity, the value of the discharging
capacitor, and the material of the substrate varied while the initial voltage was fixed at V0=5
kV. Fig. 3 shows typical examples of the current trace and radiation traces recorded using
ultra-violet (280 nm) and green (500 nm) interference filters. Both the current and 280-nm
filtered traces showed a strong AC component. Note that change in the pre-set polarity of the
spark discharge using the FTS polarity selection did not appreciably change the shapes of the
recorded current traces. The oscillations in current corresponding to the repeated re-charging
of the capacitor correlated with the oscillations in the filtered spark emission signal. The
green filtered trace showed no oscillations and the signal continuously increased during the
time the spark current was measured. This clearly indicates the thermal radiation produced
by the surface heated by the spark. Both radiation traces were observed to decay after the
current trace was reduced to zero.
Results and Conclusions
A. Powder Ejection by the Spark:
Initial ignition tests showed that a portion of the powder placed in a sample cup is
ejected from the cup for both ignition and non-ignition cases. The electrode cup on the left
shows powder that has not been struck by a spark. The middle cup shows Mg powder that
has been struck by a spark but did not ignite. . A small crater, of about 1 – 3 mm diameter
Patel 17
was always formed at the location where the spark struck the powder. Examples of the
produced craters are shown in Fig. 4. A crater of about 3 mm diameter is clearly seen in the
center of the cup. The cup on the right shows Mg powder that has been struck by a spark and
the spark energy was sufficient to ignite the powder. A crater of about 3 mm diameter is
again observed in the center of the cup. There is also a white oxide layer covering almost the
entire powder surface. It should be noted that the oxidation layer formed after the spark was
over as the surface of the powder away from the crater was observed to burn for several
seconds. The ejection of powder by the spark resulting in the formation of the central crater
can be attributed to the effect of the spark-produced shock wave20. The role of the powder
ejection in the ignition mechanism is unclear.
B. Ignition Delays
It was observed that the optical signal produced by the ignited powder was noticeably
delayed following the spark emission and current traces. The delay times were measured
using the recorded emission traces. A characteristic trace is shown in Fig. 5. A small spike
just before 10 ms indicates the spark itself, which is followed by a large, extended peak
produced by the igniting powder.
In order to determine the ignition delay, a derivative of the emission signal was
obtained. The time and signal value of the trace corresponding to the first peak of the signal
derivative were found. The signal slope in this point was projected as a straight line to cross
the zero-signal level, as illustrated in Fig. 6, so that the ignition delay could be determined.
The ignition delay was studied as a function of the selected capacitance, applied
voltage, and the distance between the discharge pin and the powder surface. The variation in
Patel 18
capacitance would effectively vary the discharge duration. Capacitors chosen for these
experiments were 2000 pF, 5000 pF, and 10000 pF. The variation in the distance between
the discharge pin and powder surface was expected to affect the distribution in the energy in
the powder sample. Gaps chosen for these experiments were 0.2 mm and 1.5 mm. The 0.2
mm is the distance recommended by MIL-STD 1751A and the 1.5 mm distance was chosen
to improve the uniformity of the energy distribution in the powder. For tests in which the
effect of capacitance and gap were varied, voltage across the capacitor was 8 kV. This
voltage was chosen because at the larger gap of 1.5 mm, a smaller applied voltage could not
produce a spark consistently. Voltages were varied between 6 – 16 kV, with 2 kV steps at a
constant capacitance of 2000 pF and a 0.2 mm gap. Each experiment at a given capacitance,
voltage, and gap distance was repeated 10 times. Similarly, for Al powder, experiments are
conducted at a constant voltage of 8kV, but only at the capacitance of 5000pF and 1000pf
and at gaps of 0.2 mm and 1.5 mm for both particle sizes. Aluminum powder is found to be
very difficult to ignite at lower energies. The results were also processed to establish the
effect of the discharge energy on the ignition delay.
Figure. 7 shows results of the described above measurements. The plot shows the
ignition delay as a function of the measured spark energy of Mg, while individual symbols
illustrate different selected capacitors, discharge gap distances, and applied voltages. The
only clearly identifiable trend is the decrease in the measured ignition delay as a function of
the discharge energy. The delay is substantial for the discharge energies less than 60 mJ,
while at the higher energies, the delay is almost unchanged and remains close to 0.5 ms.
The error bars shown in Fig. 7 represent the standard deviations among the measured
values. It is interesting that the error bars for ignition delay generally decrease at higher
Patel 19
discharge energies. Alternatively, the error bars for the discharge energies become greater
for higher energies.
While the process causing the ignition delay is currently unknown, it is expected to be
related to the induction time necessary for the thermal initiation of the metal powders. Future
work will focus on the developing of the respective spark ignition models and the obtained
experimental data presented in Fig. 7 are expected to serve as an important experimental
reference.
Similarly, Figure 8 shows results of Al powder of two particle sizes at given
parameters. In comparison to Mg results, Al powders have large error bars because of less
reproducible ignition signal. From the plot, we can see ignition delay is directly affected by
particle size.
C. Conclusion
Ignition delay for the Mg powder was between 0.5 – 2.5 ms after the spark pulse.
The delay did not depend on the selected capacitor or pre-set voltage to which the capacitor
was charged. The delay was a strong function of the measured spark energy; the delay
decreased when the energy increased up to about 60 mJ. At higher spark energies, the
ignition delay remained nearly constant and close to 0.5 ms. The delay time for Aluminum
(Al) ranges from 0 to 8.0 ms as the spark energy increased from 30 to 120 mJ. In conclusion,
Al powder of 3.0-4.5 µm has smaller ignition delay time compared to Al of 4.5-7.0 µm.
Patel 20
REFERENCES
Mellor, A. M., Stoops, D. R., “Optimization of spark and ESD propellant sensitivity tests,
1
”Propellants, Explosives, Pyrotechnics, 15, 1990, pp. 1-7
Mellor, A. M., Baker, P. J., “Propellant properties conducive to electrostatic discharge ignition,”
2
Journal of Energetic Materials, 12:1, 1994, pp. 1-62
Kao, C. S., Duh, Y. S., “Accident investigation of an ABS plant,” Journal of Loss Prevention in
3
the Process Industries, 15 (3), 2002, pp. 223-232
Glor, M., “Hazards due to electrostatic charging of powders,” Journal of Electrostatics, 16 (2-3),
4
1985, pp. 175-191
Wu, Z., Chen, Y., Hu, X., Liu, S., “Research on ESD hazards of textiles,” Journal of
5
Electrostatics, 57, 2003, pp. 203-207
Tunnicliffe, G., Thomson, M., “Explosion protection,” Cleanroom Technology, 9 (9), 2003, pp.
6
29-32
Senecal, J. A., “Manganese mill dust explosion,” Journal of Energetic Materials, 4, 1991, pp.
7
332-337
Matsuda, T., Yamaguma, M., “Tantalum dust deflagration in a bag filter dust-collecting device,”
8
Journal of Hazardous Materials, 77 (1-3), 2000, pp. 33-42
Ebadat, V., Pilkington, G., “Dust explosion hazards,” Chemical Processes 58 (9), 1995, pp. 5
9
Zeman, S., Kočí, J., “Electric spark sensitivity of polynitro compounds: Part IV. A relation to
10
thermal decomposition parameters,” Journal of Energetic Materials, 8 (1), 2000, pp. 23-26
Skinner, D., Olson, D., Block-Bolton, A., “Electrostatic discharge ignition of energetic
11
materials,” Propellants, Explosives, Pyrotechnics, 23, 1997, pp. 23-42
Matsugi, K., Hatayama, T., Yanagisawa, O., “Effect of direct current pulse discharge on specific
12
resistivity of copper and iron powder compacts,” Journal of the Japan Institute of Metals, 59 (7),
1995, pp. 740-745
Patel 21
Belyaev, A. A., Gutso, D. E., Kiseev, V. M., “Formation of interparticle contact resistance in
13
powder materials,” Poroshkovaya Metallurgiya, 10, 1992, pp. 83-86
14
MIL-STD-1751A, Manual of Data Requirements and Tests for the Qualification of Explosive
Materials for Military Use, AOP-7, 2003
15
Electromagnetic compatibility (EMC) - Part 4-2: Testing and measurement techniques -
Electrostatic discharge immunity tests IEC 61000-4-2 Edition: 1.2 International Electrotechnical
Commission, 2001
Roberts, T. A., Burton, R. L., Krier, H., “Ignition and combustion of aluminum/magnesium alloy
16
particles in O2 at high pressures,” Combustion and Flame, 92 (1-2), 1993, pp. 125-143
Ward, T. S., Trunov, M. A., Schoenitz, M., Dreizin, E. L., “Experimental methodology and heat
17
transfer model for identification of ignition kinetics of powdered fuels,” International Journal of Heat
and Mass Transfer, 49 (25-26), 2006, pp. 4943-4954
18
Pearse, R. W. B., Gaydon, A. G., The Identification of Molecular Spectra 4th ed., Chapman and
Hall, London, UK, 1976, pp. 209-211
19
Atomic Spectra Peaks, NIST Atomic Spectra Database v3, available at:
http://physics.nist.gov/PhysRefData/ASD/index.html.
Borghese, A., D’Alessio, A., Diana, M., Veintozzi, C., “Development of hot nitrogen kernel
20
produced by a very fast spark discharge,” Twenty-Second Symposium on Combustion/The
Combustion Institute, 1998, pp. 1651-1659