Density ranges for exploisves and automatic detection with x ray

www.svscountermeasurestraining.com
Explosive Density and X-ray Automatic Detection Size Discrimination
John Howell
Bomb Technician

www.svscountermeasurestraining.com
Many x-ray system on the market today use automatic detection features to identify if a
package has a material that is in the x-ray measures known range for exploisves. This feature
uses three measurements of the x-ray energy to create a window for classifying materials that
“could be” exploisves.
1. Average Effective Atomic number (Zeff)
2. Density of the material
3. Both Number 1 & 2 in a set pixel or voxel range
This paper will look at the ranges for exploisve density based on market research and also
how the density and amount of exploisves can effect the “Size” or pixel range and the blast
pressure effects on the human body for varying sizes/amounts of exploisves

Source: http://www.ndt.net/article/dir2011/papers/we21.pdf
Where do many of the X-ray
Technical papers make
reference to the density
ranges that we find
exploisves?
This paper suggest that the
range starts at 1.4 g/cc and
ends at 1.8 g/cc

Source: https://theses.lib.vt.edu/theses/available/etd-080799-132525/unrestricted/chap2.pdf
METALLIC
RANGE
INORGANIC
RANGE
ORGANIC
RANGE
This paper suggest that
the range starts at
around 1.5 g/cc and
ends at 1.8 g/cc

This paper differs from both papers
and suggest that the range starts at
around 1.2 g/cc and ends at 1.8 g/cc.
Most papers that talk about x-ray
autodetection and the density ranges
all focus on only high density ranges
for exploisves and omit any reference
to exploisve materials being located
bellow 1.2 g/cc.
Why? And are these ranges correct?

https://en.wikipedia.org/wiki/Table_of_explosive_detonation_velocities
The next reference is Wikipedia table
on the relative effectiveness factor
(R/E) for exploisves. It also provides a
table on the detonation velocity and
density ranges for many different
exploisves.
This table ONLY list 3 exploisves that
are below 1.2 g/cc in density.
One would think reading this that
most exploisves are all above 1.2
g/cc.
Is this accurate?

Where can we look to try and verify if these ranges for exploisve density are in fact correct?
We are going to look at public sources that are available to anyone that will provide exploisve
density ranges for commercial, military, and home made explosives (HME’s). This data search
will helps us to establish a large source of data to verify what are the known density ranges
for exploisves.
DENSITY in g/cc

PDF: http://www.emrtc.nmt.edu/files/safetySecurity/SafetyDataSheets/1052%20Emulsion%20Bulk%20(5.1).pdf
The best place to look for commercial manufactured
exploisves is on the manufactures websites. In the
commercial exploisves industry manufactures
ALWAYS post the density in g/cc for the exploisves
they manufacture. You can finds these on their
technical data sheets and even on the MSDS/SDS
sheets for the exploisves.
The MSDS you see is from Dyno Nobel and they are
one of the largest manufactures of exploisves in the
world. The specific MSDS list ALL OF THEIR
DYNAMITES they manufacture. The density range
they product just for their dynamite products is:
0.8 to 1.5 g/cc Density

http://www.dynonobel.com/~/media/Files/Dyno/ResourceHub/Safety%20Data%20Sheets/North%20America/1030%20Packaged%20Emulsions.pdf
Today in the USA the MSDS has
been replaced with the SDS and
even these new forms provide you
with the density ranges for
exploisves.
This is another one from Dyno
Nobel but this time for emulsion
exploisves and the range it provides
is:
0.95 to 1.25 g/cc density
Just these 2 references alone do not
match what the scientific
community tech papers list. Why?

What are the correct density
ranges for explosives?
We looked at over 10 different exploisve manufactures worldwide and collected the data on
every exploisve product they manufactured. From this search we found over 600 different
types of exploisve products and the density for each one. We complied this data into a
spreadsheet and broke them down by “TYPE” of exploisve (e.g. dynamites, water gels,
emulsions, etc.).
We also looked at the ranges for military grade high exploisves and home made explosives
and added them to the list.

1. Dyno Nobel: http://www.dynonobel.com/apac/resource-hub/products
2. Explosia: http://www.explosia.cz/en/?show=vyrobky
3. Austin Powder: http://austinpowder.com/safety-data-sheets/ammonium-nitrate/
4. Kapeks: http://www.kapeks.com.tr/en/index.html#
5. Orica: http://www.orica.com/Products---Services#.VwvyGPkrKM8
6. AEL: http://www.aelminingservices.com/
7. BIAFO: http://www.biafo.com/tovex.htm
8. MAXAM: http://www.maxam.net/buscador_ceprods_action/9887/9764
9. E Brickford: http://www.eba-d.com/products/
10.Ideal: http://www.idealexplosives.com/emulsion-explosives.html

Explosive Density Ranges
Emulsions
• Over 70% were detonator sensitive and did not require a
booster
o Boosters are typically very high density explosives
(~1.5 g/cc).
o Market Percentage: Over 25%
o Emulsions are a AN based exploisve and are
aluminumized to increase sensitivity (1 to 10% max)
• Density Range: 0.8 g/cc to 1.3 g/cc
o 95% were below 1.2 g/cc
o The lowest was .8 g/cc
Average density = 1.16 g/cc

Dynamites
• Density Range: 0.75 g/cc to
1.51 g/cc
• Market Percentage: ~8%
Dynamites are not very
common anymore and are being
replaced by emulsions in the
market

Water Gels
• Market Percentage: less than 5%
• Being replaced by emulsions in the
market (lower overall cost)

Explosive Boosters
• Most are less than 5” in length (Size issue)
• Typically used with ANFO mixtures to set them off
• These are a CAST explosive most common is TNT

Plastic Explosives
• More common in military applications (EOD)
• Not used in most commercial blasting
• Typically a RDX or PETN based explosive

TNT
Density:
• Cast TNT: 1.6 g/cc
• Commercial products containing some TNT:
0.8 g/cc to 1.7 g/cc
Market Percentage: less than 1%
• Cast TNT: 1/428
• TNT is MIXED with other materials to make
different exploisves (lower densities) and is
not commonly used in a 100% pure mixture
outside of CAST Booster (ANFO)

Density Range: 0.3 g/cc to 1.1 g/cc
Market Percentage: over 30%
AN and ANFO (ammonium Nitrate and Fuel
Oil) are one of the most common commercial
and terrorist used exploisve in the world.
There are many different mixtures and this is
why the density ranges are so wide. One
example is AN/AL and is a very common
referenced HME but is also the same thing as
emulsions (AN and AL)
AN & ANFO

• I found zero (0%) pure high
explosive powders commercially
sold to be used “as is”. They are sold
to be MIXED and make ordnance
fillers
• Commercial products w/ high
explosive powders:
Powder Explosives
Density Range
1.6 g/cc to 1.8 g/cc
o Det. cords
o Sheet explosives
o Plastic Explosives
Powders have a crystal and bulk density and the bulk
density (what x-rays see) is typically much lower
PETN

• Zero pure HME’s commercially sold
• HME’s most commonly used:
o TATP (triacetone triperoxide)
o HMTD (hexametheline tridiamine)
o ANAL (ammonia nitrate aluminized
powder)
Homemade Explosives (HME’s)
Average density = ~1.2 g/cc
• Density range: 0.8 - 1.2 g/cc (most common).
• A few have higher densities, up to 1.8 g/cc (TATB)

Gun Powders
Black powder
Black powder replacement
Single based smokeless powder
Double based smokeless powder
Flash Powder
Black powder (& replacement)
• High density & Zeff → Green in X-ray
Smokeless powder single and double base
• Low density & Zeff → Orange in X-ray
Depending on the mixture some of these are
detonator sensitive and do NOT require a
pressure vessel to high order

“Liquid” Explosives
Explosive Density
AN-based 0.9 - 1.33 g/cc
Nitromethane and PLX 1.13 - 1.16 g/cc
Nitroglycerin 1.6 g/cc (highest “liquid”
explosive)
MEKP 1.17 g/cc
Acetone Peroxide 1.18 g/cc

Where do the explosives now
plot on a density graph?

1 1 1 1 3 6 8 9 8 7 7 5 5 4 3 1 1Total Per Density
Largest
Concentration

Actual Density Range for Exploisves
0.6 to 1.8 g/cc
Where Most Technical
Papers Plot Range

Where do most false alarms plot on a density
graph?
One of the major problems with x-ray automatic detection of exploisves are “False Alarms”
from materials that are not an exploisve but fall into the density and Zeff range we see
exploisves. Lets take a look at where these “False Alarm” items show up on a density plot
next to the real exploisves.

Density (g/cc)
Common, inaccurate
generalization of explosive
density
The “False Alarm” range
falls directly into the
same range we see the
vast majority of all
commercial, military,
and homemade
exploisves. You can see
that where most x-ray
scientific papers show
the range (incorrectly)
oddly exclude that
entire range.
Why? Are they just
wrong or was this by
design?

Exploisve AN Silica TATP ANFO
El Blasto
1.0
AN
Dynamite
1.1
Nitro Dyn
1.2
PE-4
Long A
PE-4
Long B
PE-4 Small
A
Extra Gel
1.3
Data
Sheet
M1 dyn
1.4
Semtex H
Semtex-
1A
Data
Sheet
Semtex-
10
Cast TNT Cast booster M112
6 inch
round
Density
.8
g/cc
1.1
g/cc
1.1
g/cc
1.2
g/cc
1.0
g/cc
1.1
g/cc
1.2
g/cc
1.3
g/cc
1.3
g/cc
1.3
g/cc
1.3
g/cc
1.4
g/cc
1.4
g/cc
1.4
g/cc
1.4
g/cc
1.4
g/cc
1.4
g/cc
1.5
g/cc 1.5 g/cc
1.6
g/cc
1.6
g/cc
EDS System
CT Unit A
Alarm No No No No No No No Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes
CT Value Belt N/A N/A N/A N/A N/A N/A N/A 1396.2 1415.2 N/A 1318.3 1308.9 1416.4 1377.6 1455.9 1308.9 1479.9 1473.4 N/A 1296.7 1474.1
CT Value Bag N/A N/A N/A N/A N/A N/A N/A 1489.2 1375.2 N/A 1320.3 1356.2 1403.2 1463 1474.8 1356.2 1470 1574 N/A 1221.6 1366
Remarks Unit A
Unit will
detect
Larger size
but not
Unit will
detect the
larger size
Cast TNT
CT Unit B
CT Value Belt N/A N/A N/A N/A N/A N/A N/A 1433.78 1365.67 N/A 1316.1 13275.59 1507.61 1414.21 1442.71 13275.59 1495.64 1512 N/A 1283.41 1439.12
Alarm No No No No No No No Yes No No Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes
CT Value Bag N/A N/A N/A N/A N/A N/A N/A 1496.13 N/A N/A 1300.64 1222.45 1375 1222.02 1235.61 1222.45 1510.78 1539.63 N/A 1243.08 1433.55
Remarks Unit B
Unit will
detect the
PE-4 in a
larger size
Unit will
detect the
larger size
Cast TNT
CT Unit C
Alarm No No No No Yes No Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes
CT Value Belt 861 1029 958 903 1026 1147 1224 1568 1521 1303 1471 1471 1474 1382 1414 1471 1471 1480 1487 1472 1456
CT Value Bag 1122 979 893 2563 746 1233 1320 1484 1439 1351 1451 1451 1452 1425 1421 1451 1451 1544 1042 1448 1409
Remarks Unit C
Unit will
detect the
PE-4 in a
larger size
Unit will
detect the
larger size
Cast TNT
CT Unit D
Alarm No No No No No No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes
CT Value Belt 880 1007 951 22 1055 1158 1347 1521 1423 1521 1423 1440 1554 1347 1426 1440 1454 1535 1494 1363 1444
CT Value Bag 828 1001 893 3273 904 1212 1297 1439 1338 1439 1338 1426 1499 1447 1433 1426 1355 1588 1213 1526 1350
Remarks Unit D
Unit will
not
respond
to
simulant
Unit will
detect the
PE-4 in a
larger size
but not
smaller
Unit will
detect the
larger size
Cast TNT
CT unit E
Alarm No No No No No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
CT Value Belt 8542 10973 9626 1065 9762 11915 10041 14586 8434 13949 10427 14586 13755 12858 14800 14586 14282 14411 19542 13849 11013
CT Value Bag 17839 10068 8281 25445 9283 12335 13222 14606 13799 13755 13755 14606 13755 14216 15976 14606 14310 17168 9821 12310 15045
Remarks Unit E
Unit will
detect the
PE-4 in a
larger size
but not
smaller
Unit will
detect the PE-
4 in a larger
size but not
smaller size
CT Unit F
CT Value Belt 797 10042 9227 4234 11132 11842 11619 15588 13476 13546 13546 14574 13530 13295 14371 14574 14380 15230 13811 15354 14900
CT Value Bag 1414 12079 8976 22074 9785 12903 13599 15358 12131 10965 10965 14253 14574 14757 13491 14253 14043 15220 10699 15334 13667
Remarks unit F
Unit will
detect the
PE-4 in a
larger size
but not
smaller
Unit will
detect the PE-
4 in a larger
size but not
smaller size
CT values for 1.1
should have
alarmed. No clue
why it did not
CT values for 1.1
should have
alarmed. No clue
why it did not
Unit alarm level is set above any
material with a density above
1.2 or CT value 1222
I tested this theory on the
ranges for false alarm items
on many different CT systems
to see if these ranges were in
fact “cutout” and what I
found was that they were.
Anything that fell into the
range where you see false
alarm causing items based on
density would not alarm.
This appears to be a designed
effort to lower false alarm
rates and increase
throughput by cutting out
the range where we see most
false alarm causing items.
Am I right in what I am
seeing?

Exploisve AN Silica TATP ANFO
El Blasto
1.0
AN
Dynamite
1.1
Nitro Dyn
1.2
PE-4
Long A
PE-4
Long B
PE-4 Small
A
Extra Gel
1.3
Data
Sheet
M1 dyn
1.4
Semtex H
Semtex-
1A
Data
Sheet
Semtex-
10
Cast TNT Cast booster M112
6 inch
round
Density
.8
g/cc
1.1
g/cc
1.1
g/cc
1.2
g/cc
1.0
g/cc
1.1
g/cc
1.2
g/cc
1.3
g/cc
1.3
g/cc
1.3
g/cc
1.3
g/cc
1.4
g/cc
1.4
g/cc
1.4
g/cc
1.4
g/cc
1.4
g/cc
1.4
g/cc
1.5
g/cc 1.5 g/cc
1.6
g/cc
1.6
g/cc
EDS System
CT Unit A
CT Value Belt N/A N/A N/A N/A N/A N/A N/A 1396.2 1415.2 N/A 1318.3 1308.9 1416.4 1377.6 1455.9 1308.9 1479.9 1473.4 N/A 1296.7 1474.1
CT Value Bag N/A N/A N/A N/A N/A N/A N/A 1489.2 1375.2 N/A 1320.3 1356.2 1403.2 1463 1474.8 1356.2 1470 1574 N/A 1221.6 1366
Remarks Unit A
Unit will
detect
Larger size
but not
Unit will
detect the
larger size
Cast TNT
CT Unit B
CT Value Belt N/A N/A N/A N/A N/A N/A N/A 1433.78 1365.67 N/A 1316.1 13275.59 1507.61 1414.21 1442.71 13275.59 1495.64 1512 N/A 1283.41 1439.12
Alarm No No No No No No No Yes No No Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes
CT Value Bag N/A N/A N/A N/A N/A N/A N/A 1496.13 N/A N/A 1300.64 1222.45 1375 1222.02 1235.61 1222.45 1510.78 1539.63 N/A 1243.08 1433.55
Remarks Unit B
Unit will
detect the
PE-4 in a
larger size
Unit will
detect the
larger size
Cast TNT
CT Unit C
Alarm No No No No Yes No Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes
CT Value Belt 861 1029 958 903 1026 1147 1224 1568 1521 1303 1471 1471 1474 1382 1414 1471 1471 1480 1487 1472 1456
CT Value Bag 1122 979 893 2563 746 1233 1320 1484 1439 1351 1451 1451 1452 1425 1421 1451 1451 1544 1042 1448 1409
Remarks Unit C
Unit will
detect the
PE-4 in a
larger size
Unit will
detect the
larger size
Cast TNT
CT Unit D
CT Value Belt 880 1007 951 22 1055 1158 1347 1521 1423 1521 1423 1440 1554 1347 1426 1440 1454 1535 1494 1363 1444
CT Value Bag 828 1001 893 3273 904 1212 1297 1439 1338 1439 1338 1426 1499 1447 1433 1426 1355 1588 1213 1526 1350
Remarks Unit D
Unit will
not
respond
to
simulant
Unit will
detect the
PE-4 in a
larger size
but not
smaller
Unit will
detect the
larger size
Cast TNT
CT unit E
CT Value Belt 8542 10973 9626 1065 9762 11915 10041 14586 8434 13949 10427 14586 13755 12858 14800 14586 14282 14411 19542 13849 11013
CT Value Bag 17839 10068 8281 25445 9283 12335 13222 14606 13799 13755 13755 14606 13755 14216 15976 14606 14310 17168 9821 12310 15045
Remarks Unit E
Unit will
detect the
PE-4 in a
larger size
but not
smaller
Unit will
detect the PE-
4 in a larger
size but not
smaller size
CT Unit F
CT Value Belt 797 10042 9227 4234 11132 11842 11619 15588 13476 13546 13546 14574 13530 13295 14371 14574 14380 15230 13811 15354 14900
CT Value Bag 1414 12079 8976 22074 9785 12903 13599 15358 12131 10965 10965 14253 14574 14757 13491 14253 14043 15220 10699 15334 13667
Remarks unit F
Unit will
detect the
PE-4 in a
larger size
but not
smaller
Unit will
detect the PE-
4 in a larger
size but not
smaller size
CT values for 1.1
should have
alarmed. No clue
why it did not
CT values for 1.1
should have
alarmed. No clue
why it did not
Either by design or by very inaccurate plotting of the density ranges for exploisves it appears that the development of x-ray
automatic detection windows needs to be verified and a baseline established for where we actually see exploisves based
on density. Most “engineers” really do not know where to look to get the data they need for exploisve density ranges and
most bomb technicians are also not familiar with these ranges. This “could” explain why we see incorrect plots for density
ranges because if this is not the reason, the only alternative is that this was done by design to lower false alarm rates. This
“could” be a effort to keep the lines moving and avoid trying to clear all of the alarms that would be generated by items that
are not exploisves but fall into the same density a d Zeff ranges.

Is it “SMART” to use size as a discrimination
tool when developing X-ray Automatic
Detection Algorithms?
One of the other ways I have found x-ray systems try and lower false alarm rates is by
placing a “size cutoff” in the software. This works by omitting anything that is below a set
pixel or voxel range (cm 2) of a material that has a density and Zeff the software has
established for autodetection a potential exploisve.
Is this smart when you look at the density of a exploisve material and how it effects its
overall size (pixel and voxel cm2 range)? Also you MUST consider how powerful smaller
amount of exploisves are by calculating the blast pressure each amount is capable of
producing.

Density is a Major Factor in the size of
the explosive material
I cut Derlin to specific size in 1
inch thickness based on the
density of a specific exploisve.
The density dictated the
overall size (L x W x H) of each
Derlin test object based on the
weight.
As you can see a .88 g/cc 1 lb.
1 inch thick Derlin block was
much larger then the higher
density 1.6 g/cc
Derlin has a density of 1.41 g/cc and a Zeff of 7 (exploisve range)

.88 g/cc
1 pound
1.41 g/cc
.5 pound
1.6 g/cc
1 pound
1.6 g/cc
.25 pound
1.6 g/cc
.5 pound
1.6 g/cc
.75 pound
I created many different sized test
objects to use for my testing and
each was based on a specific
density and weight for a know
exploisve.
The goal was to run these through
a x-ray and verify if in fact a “size
cutoff” was in the software. I also
wanted to look at the density and
blast pressure effects for these
test objects based on the weights.
How much damage could the
smaller sizes do?

In all of the test I conducted on different models of x-ray I found that almost all of them
“cutoff” detection for a 1.6 g/cc .25 lb. test object. 1.6 g/cc would be the range you
would be a exploisve like C-4

No Detection
.25 LB 1.6 g/cc
27.9 sq. cm
U.S. Marshals TestingExplosive Density and X-ray Automatic Detection Size Discrimination

I cut also cut 1.25 in dia
Derlin rods to specific
lengths and tested
them
I saw these stop
detecting on the
generator side of the
belt at 4.0 inches. As
you moved away from
the generator this
dropped t0 2.5 to 3.0
in. Center of the belt
averaged around 3.5 in
where it stopped
detecting

Note: On a single generator system the closer to the generator
the larger the item will show up “more pixels”. The farther
away from the generator the smaller the item will show up
“less pixels” and create a NO Detection scenario.
Away from
Generator
NO
DETECTION
Close to
Generator
DETECTION
Generator

The detection software requires a
pixel or voxel (CT) range that is
uninterrupted for the
autodetection to work when you
have a size cut off built into the
algorithm.
The image you see is a 4 in long
1.25 in diameter Derlin rod that
normally alarms stop alarming
when I high density material blocks
part of it. This resets the size cutoff
and the material that detected
before no longer detects.

This is a 4 generator
x-ray system and
you can see I was
able to get the
M112 C-4 block of
exploisves to NOT
DETECT by adding a
blocking material.
In a clutter bag this
shows that 1.25 lbs.
of high exploisve
could potentially be
NOT DETECTED due
to a size cutoff.

Can I use size as a discrimination tool when developing X-ray
detecting algorithms?
Let’s take a closer look.
Preventing the detection of smaller amounts of explosives is extremely dangerous and
illogical.
How much blast pressure is safe for a human?
• Critical to terminal injuries: 40 PSI or 275 Kpa
• Death: 55-65 PSI or 380 Kpa
This does not take into consideration fragmentation

Can I use size as a discrimination tool when developing X-ray detecting algorithms?
Calculating Net Explosive Weight (NEW)
Definition: The total mass of a live explosive material
Equation: (quantity)(weight in lbs..)(RE factor) = NEW
o RE factor = relative effectiveness factor
Example: What is the NEW of PETN weighing 0.30 lbs.?
(1)(0.30 lbs..)(1.66) = 0.498 lbs..
0.498 lbs.. = 0.22 kg
https://en.wikipedia.org/wiki/TNT_equivalent

Online Blast Calculators
Calculator Source: United Nations Office for Disarmament Affairs

Analyzing Blast Calculator Results

How Much Will Kill You???

Conclusions
• Automatic detection windows cannot safely be narrowed to reduce false alarms to try
and prevent false alarms
• Using Size discrimination to reduce false alarms cannot be considered a safe method
due to blast pressure measurements on a person.
• Cutting out density ranges where we see most false alarm causing materials is not a
effective method because the majority of exploisves that are being used today fall into
those ranges.
• Security officers should verify each and every false alarm to ensure it is not explosive
material. Detection should always outweigh throughput.

Density ranges for exploisves and automatic detection with x ray

Density ranges for exploisves and automatic detection with x ray

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Density ranges for exploisves and automatic detection with x ray

Similar to Density ranges for exploisves and automatic detection with x ray (20)

Recently uploaded

Recently uploaded (20)

Density ranges for exploisves and automatic detection with x ray