Saturday, June 16, 2012

Too little thermite to blow up Mark Basile's chip


Mark Basile analysed red-gray chips he found in dust samples collected in lower Manhattan very shortly after the collapse of the World Trade towers on 2001/09/11 [1.1]. In particular, he shows how vigorously his chip reacts when heated on a steel strip, producing rapid ejections of gas [1.2]. Basile suggests that this reaction is best explained by the thermite reaction, apparently affecting the organic matrix.

In an earlier blog post [2], I have shown that his data reveals that at most 4.7% by weight of the "energetic" red layer of one particular chip could possibly be stoichiometric thermite, while most of the layer (ca. 88%) must be a matrix of some unidentified polymer.

Some 9/11 Truth Movement adherents who believe that these red-gray chips are "thermitic material" claim that organic substances are typically a component of modern nano-thermite preparations, both as collateral residue of the synthesis (e.g. ca. 10% [3.1]) and as an additive to give nano-thermite explosive properties [3.2] (the organic material is rapidly turned to gas and can do volume work).

In this post, I will show that, at a mass ratio between thermite and organic matrix of about 1:19, as Basile's data implies, the chemical energy of the thermite does not nearly suffice to turn organic polymers to gas. It follows that the rapid reaction and creation of gas is powered by organic combustion and perhaps externally applied heat, not a thermite reaction.


Basile's "lucky" chip #13

As I showed in [2], under the most "thermite friendly" assumptions, and taking Basile's data as it is, the red layer of his "lucky" red-gray chip #13 contains, by weight,

  • At most 4.74% ideal (stoichiometric) thermite (of the common Fe2O3+Al variety)
  • At least 87.8% solid hydrocarbon matrix, unknown chemistry. It is safe to assume that this matrix is some form of organic polymer (or a mix of polymers), that contains no Fluorine or Chlorine
  • Ca. 7.1% (the balance) inorganic compounds, assumed to be inert

Thermal properties of various polymers

The Appendix of [4] lists many combustion-related thermal properties of many organic polymers. The following values will be used in the discussion. I chose Epoxy as the reference material, since James Millette [5] has identified epoxy as the matrix material for some red-gray chips. The properties of many other non-halogenic organic polymers are in roughly the same magnitude as those of epoxy. I will state ranges for most polymers in parentheses, even though the extreme values usually are for materials that wouldn't make much sense for a matrix:

  • Onset of decomposition: Td 427 °C (250 - 570 °C, Table A-1, first column). This property describes at which temperature the matrix will beginn to decompose, a process that usually involves some charring and some release of gas. It will also show in DSC curves.
  • Ignition temperatur Tign: 427°C (271 - 600°C, Table A-1., third column). Note the ignition temperature may be influenced by association / mixing with other materials. Note also that some polymers don't ignite (don't burn with atmospheric oxygen) and just decompose
  • Enthalpy of gasification hg: 1.5 kJ/g (1.1 - 2.6 kJ/g, Table A-2, third column). This value describes how much energy must be expended to break the molecules down to gas molecules such as CO2 or water vapor during burning or decomposition - not including the heat necessary to bring the polymer to the temperature where the molecukle structure begins to break down. Note that most polymers leave behind some solid residue (char) after gasification: Epoxy 4% of its mass (column two of table A-2), others up to 75%.
  • Heat capacity cp: 1.7 J/g/K (0.93 - 2.09 J/g/K, Table A-3, third column). This value describes how much energy is expended when heating 1 g of polymer by 1 °C (or by 1 K, which is the same). This value changes with temperature, it is given for normal "room temperature" conditions, but it typically increases somewhat with rising temperature. I will consider it as constant, which is a "thermite-friendly" imprecision.
  • Effective heat of combustion HOC: 20.4 kJ/g (14.4 - 41.9 kJ/g, Table A-5, first column). This is the energy effectively released by 1 g of polymer under air and takes into account that the theoretical maximum is not reached in praxis. Epoxy for example burns with only only 75% effectiveness in experiment. This is again a "thermite-friendly" choice, as I will use the theoretical max for thermite (3.96 kJ/g) and not actual effective heat release (perhaps 3 kJ/g or less).


Heating epoxy with thermite

To simplyfy things, let's ignore the inorganic components other than stoichiometric thermite, and mix thermite and epoxy in the proportions according to Basile's data: 4.74 g of thermite, 87.8 g of epoxy. Let's further assume we could ignite this thermite and have it react perfectly within the epoxy matrix without heating the epoxy first, and have all of the heat of reaction be absorbed by the epoxy. Could the thermite reaction turn the matrix to gas and cause the rapid gas ejections seen in Basile's video? Let's see!

4.74 g of thermite contain at most (theoretical maximum) 4.74 g x 3.96 kJ/g = 18.7 kJ

If you put these 18.7 kJ of heat into 87.8 g of epoxy, which has a specific heat capacity of 1.7 J/g/°C, you warm it by 18,700 J / 87.8 g / 1.7 J/g/°C = 125 °C, reaching ca. 150 °C. Neither epoxy nor any other polymer would come close to the start of decomposition just from this thermite reaction!

Gasifying epoxy with thermite

Of course, the assumption that the epoxy isn't already heated to the brink of decomposition isn't realistic - thermite wouldn't ignite at room temperature, and you can't heat only the thermite inside the matrix. So next up. let's assume the epoxy is already heated to its decompostion temperature of 427°C, as is the thermite - which is concidentally (???) the temperature at which Harrit [6] observed ignition of red-gray chips. How much epoxy could the reaction of 4.74 g thermite turn to gas? Let's see!

Epoxy has an effective enthalpy of gasification of 1.5 kJ/g. The energy release of our thermite, 18.7 kJ, could thus gasify 18.kJ / 1.5 kJ/g = 12.5 g of epoxy, out of 87.8 g of epoxy in our sample, that's about 14%.

What causes the gas jets and the heating of "lucky" chip #13?

Marc Basile had heated his chip on a thin (50 ┬Ám) steel strip through which he sent a constant electrical current. Here a screenshot from 40:17 in his presentation [1]:


This is, obviously, an important heat source. He gives is no idea how hot the strip got during the experiment. Hot enough apparently to ignite and gasify something - but potentially much hotter than just that. At the very least, this external heat infused 400 K x 1.7 J/g/K = 680 kJ/g into the probe just to heat the epoxy - thermite's theoretical max would be about 180 J/g, or 26% maximum compared with the heating strip.

It is obvious from my calculations above that thermite, even if present at all and in the maximum possible amount, contributes only minimally to the reaction of the organic matrix.

In particular: Every Joule expended on heating the matrix can't be expended to gasify it. Every Joule expended to gasify the matrix can't be used to heat any bit of matrix. And every Joule expended to do work on the matrix is lost to heat and ignite the next thermite particles to continue the thermite reaction. This material could never burn if the matrix were inert and the probe weren't externally heated. What is the use of thermite in such low concentration?

The organic matrix on the other hand is assured to release enough energy to: Heat the probe including all minerals and the gray layer, achieve full gasification, and warm its environment: of the 20.4 kJ/g effective energy density, only 1.5 kJ/g are expended on gasification, 0.7 kJ/g (1.7 J/g/K x 400 K) are expended to heat the same mass of epoxy from room temperature to ignition temperature, and then 18.2 kJ/g are left to do work on everything else


The three obvious and available heat sources in Basile's ignition experiment provide this much energy per gram of probe:

  1. Combustion of epoxy: 12.6 - 36.8 kJ/g (Epoxy: 17.9 kJ/g = )
  2. Heating strip: 0.7 kJ/g or more
  3. Thermite: 0.18 kJ/g or less

5% Thermite in an organic matrix make no difference. On its on, it couldn't warm the matrix even to onset decomposition, it could not destroy more than a small fraction of the polymer molecules, and it would be incapable of doing any significant work on anything outside of the chip

Whatever reaction is observed in the video of chip #13 burning, it is not driven by a thermite reaction. It is simple organic polymer combustion, helped to an unknown but probably significant degree by the external heat of the heating strip underneath.

Additional remarks

1. I believe almost all red-gray chips found in WTC dust, including Basile's chip, are some sort of red primer paint on spalled steel / steel mill. Gauging the composition of LaClede standard primer [7], I suspect that Basile's quantification of the elemental composition is a bit off the mark - I would expect to see closer to 30% inorganic materials rather than the 12% according to Basile. I suspect in particular that he underestimates the amount of iron: His red layer is red paint, and the red pigment most certainly is iron oxide. There should be closer to 10% of the element iron rather than Basile's 2.6%. However, I am only guessing here, and I can only go by the data Basile provides

I am convinced that Millette [5] and Harrit [6] looked, most closely at LaClede standard primer, which according to my own analysis [7] can be expected to contain 2.4% aluminium. Harrit's chips a-d match the expected elemental composition of LaClede paint so closely, that I would say definitely these chips contain about that much of the elememnt Al. If, hypothetically, all that Al were elemental, it could react with three times as much of the iron oxide to form 10.4% thermite - against 71.5% epoxy. This ratio, 1:6.9 thermite:epoxy, is still insufficient to either heat epoxy from room temperature to ignition point, or gasify most of it, and the heat content of the epoxy would still outnumber that of the thermite by a ratio of at least 50:1, rendering the thermite insignificant.

In further, unpublished work, I have estimated that the total Al content of Harrit's MEK-soaked chip ([7], Fig. 14) is only 0.6%, to allow for a maximum of 2.4% thermite. It should be obvious by now that this is even less significant than the hypothetical thermite-content of Basile's chip or the chips a-d that resemble LaClede so much. It is interesting that this MEK-soaked chip, with its very low overall Al-content, is the only one where the "thermite" theorists seem to have identified any elemental Al at all.


[1.1] Mark Basile: 911 Dust Analysis Raises Questions. Videotaped presentation at the Porcupine Freedom Festival in Lancaster, New Hampshire on 26th June 2010, 4pm (On YouTube; 59:22 minutes, Last retrieved: June 16 2012)

[1.2] Mark Basile ignites a chip (nano-thermite) - 9/11. This szene is shown in [1.1] between between 41:43 and 42:00 minutes. (On YouTube; 0:16 minutes, Last retrieved: June 16 2012)

[2] Oystein: How Mark Basile confirms that red-gray chips are not thermitic. Posted in author's blog on March 18 2012

[3.1] T.M. Tillotson et al: Nanostructured energetic materials using sol-gel methodologies. Journal of Non-Crystalline Solids 285 (2001) 338-345

[3.2] (Currently too lazy to find an exemplary paper)

[4] Richard E. Lyon and Marc L. Janssens: Polymer Flammability. May 2005 - Final Report for the U.S. Department of Transportation and FAA. Report No. DOT/FAA/AR-05/14

[5] James R. Millette: Revised Report of Results: MVA9119. Progress Report on the Analysis of Red/Gray Chips in WTC dust. Prepared for Classical Guide, Denver, 01 March 2012.

[6] Niels H. Harrit et al: Active Thermitic Material Discovered in Dust from the 9/11 World Trade Center Catastrophe. The Open Chemical Physics Journal, 2009, 2, 7-31. Figure 19 shows ignition temperatures around 430°C

[7] Oystein: Another primer at the WTC: LaClede Standard Primer

. Posted in author's blog on March 16 2012


  1. Hi, Oystein, thanks for interesting calculations:o)

    As I noted in JREF, we should not take thermal data for epoxy 427 C starting decomposition temp., 462 C maximum decomposition rate temp. and 427 C autoignition temp.) as some ultimate data). They are not really compatible anyway and are probably taken from various sources. Namely, I would expect autoignition of a oxidized resin at temp. at maximum decomp. rate or higher, not at the beginning of decomposition, where the concentration of flammable stuffs is probably not high enough autoignition.

    What I found for epoxy resins last weeks, as for heating under air in DSC or DTA, they exhibit the most distinct exothermic effects around 500 degrees and they are generally ascribed to the burning of epoxy char. It is hard to say if DSC curves in Bentham curves are "compatible" with these published data.

    1. Ivan,

      I am quite aware of the limited applicability of these values, and that the thermal behaviour of polymers is more complex than my calculation implies. But if you go with higher temperatures to assess the effect that embedded thermite would have on the matrix, these tiny amounts fall even shorter of the mark. Basically, you are saying that my values are too "thermite-friendly" - which is no problem for my argument.

      Also, note that "epoxy" is an assumption, an example; not entirely arbitrary, but I really can't say if epoxy is correct here. The point is that most polymers have values for ignition or decomposition temperature, heat capacity and energy density that are not too far off from those of epoxy. A factor of no more than 2 here or there, some of these factors might even cancel out to some extent - while Basile's tiny thermite content fails to make the grade by a factor of 8 or or more.

      I am mainly discussing Basile's chip here. For the old reasons I have no way to know if his chip would behave like Farrer's four in the DSC, and in no case do I have proof that their organic matrixes are pure epoxy, so the question whether the DSC curves are compatible with reference data for epoxy is rather moot.

      On top of that, these values have certainly been tabulated for some pure epoxy, while the behaviour for example of LaClede chips (which we know to be epoxy-based) in a DSC may well be influenced by the presence of 3:7 parts of pigments, and the gray layer.

      Basile's chip 13 by the way has the red layer mostly charred after it "burned":

      Chip 13 after burning
      (43:04 in the Basile presentation, [1])

  2. Hi Oystein,

    Your analysis obviously works if the thermite and epoxy are randomly distributed. But what about a scenario which has a layer of thermite adjacent to the epoxy? Let's see if the thermite can ignite part of the epoxy before the heat is transmitted across to the rest of the epoxy.

    The mass ratio of thermite to epoxy is 1:18.5. A 60% compacted Al/Fe2O3 thermite powder has a density of 2.541 g/cc. Taking epoxy as 1.6 g/cc, the volume ratio is 1:29.4. So a 25-micron red layer has 24.18 microns of epoxy and 0.82 microns of thermite.

    The cross-section is 22,680 cm^2 or 2.268 m^2 for the 4.74 g of thermite and 87.8 g of epoxy, so if the masses are adjusted for 1 m^2 of steel surface, there is 2.09 g of thermite and 38.7 g of epoxy.

    The thermal conductivity of epoxy varies widely depending on application, but I found some reference for "thermally conductive" epoxies with thermal conductivities of 0.6 and 1.0 W/m.K.

    Over each m^2, from q = -k * A * (T1-T2) / x and assuming 0.6 W/m.K and a temperature difference of 375 C across the epoxy from the thermite side to the cold side, there is 0.6 * 1 * 375 / 0.0000242 = 9.3 MW for the heat conducted through the epoxy to heat the cold side. The 2.09 g of thermite per m^2 of epoxy / steel yields 3.9 kJ/g * 2.09 g = 8.15 kJ. So if the thermite releases its energy over as short a period as a millisecond, the 8.15 MW/m^2 is still below the 9.3 MW/m^2 flowing away to the cold side of the epoxy. At a temperature difference of less than 329 C, the energy flow from the thermite can exceed energy flow through the epoxy, but the temperature is too low to ignite the epoxy. And at the end of the period, e.g. a millisecond, the thermite has released all of its energy but heat is still being conducted across the epoxy and across to the steel.

    The following reference shows Al/Bi2O3 and Al/CuO nano-thermites with burn times of ~0.09 ms and ~0.18 ms, respectively. So they could briefly get part of the epoxy above its ignition temperature. But the epoxy isn't a nano-energetic composite, and I wouldn't expect it to ignite in such a short time, given that the thermite soon stops burning and the epoxy continues to conduct heat away from its hot side to its cold side. (Al/Fe2O3 nano-thermite had a longer burn time at 0.936 ms.)

    2.09 g of thermite per m^2 of steel, given a 0.37-inch steel thickness (as per truss bottom chord) assuming it doesn't need to heat epoxy or anything else, can raise the steel temperature by 2.09 g * 3,900 J/g / 0.0094 m / 7850 kg/m^3 / 450 J/kg.K = 0.24 C.

    In any case, the epoxy would ignite before the thermite, and any layer of pure thermite would have been found by Harrit or Millette especially if on the outer surface of the red layer, and if it was on the inside adjacent to the steel then it has to heat the steel in addition to the epoxy and igniting it would be difficult, to say the least. Such a small amount of thermite would serve no useful purpose.

    1. Poseidon,

      nice to see you tackling this from a different angle. If we go much further than this, our work will become too speculative. There obviously is no such layered epoxy-thermite stuff, so why bother calculating its behaviour? I think the futility of both low concentrations of thermite within an organic matrix and low thermite amounts compared to the attacked steel is amply shown.

    2. It shows that it's necessary to have a high enough proportion of thermite to organic material. But the red/gray chip = thermite theory fails because of insufficient thermite, as well as too little thermite in relation to organic material. However, red/gray paint chips (including Basile's chips) were probably contaminated with some energetic material that was used for the demolitions.

      Under one scenario (thermite only added to the SFRM), I take the average 303.3 kg/m^3 upgraded SFRM density compared to standard SFRM of 240 kg/m^3 (15 psf), and posit a thermite/SFRM composite that is 23% thermite by mass, each gram of the lethal composite being comprised of 0.23 g of thermite at 2.541 g/cc taking up 0.0905 cc and 0.77 g of SFRM at 0.24 g/cc occupying 3.208 cc. The density of the composite is therefore 1 / 3.298 = 0.3032 g/cc or 18.93 psf, which is 26.2% high as observed (NIST NCSTAR 1-6A, Table 4-2).

      In this scenario there is 2.5 inches or 0.0635 m layer of the material, so 0.0635 m^3 * 303.2 kg/m^3 = 19.25 kg of the material per m^2 of steel. The 23% thermite by mass is 4.43 kg, so at 3.9 MJ/kg the temperature rise in 0.37-inch steel (0.0094 m) is 17,277,000 / 0.0094 m^3 / 7850 kg/m^3 / 550 J/kg.K = 425 C. When you allow for the SFRM being sprayed onto both sides of the truss bottom chord, that amount can be nearly doubled, and you can deduct 25% for oxidized Al and still have the steel losing more than half its strength at 600 C before even counting heat absorbed from the office fires.

      Perhaps the better alternative supposes that the 303.3 kg/m^3 upgraded SFRM density is entirely normal, and the 340.3 kg/m^3 density over WTC1 floors 94-98 (defined by FEMA as the "impact zone") is due to additional material inserted, which occupies the air gaps between the fibers. In this case the extra material is nano-thermite and organic, e.g. epoxy, say a 1:1 mass ratio. This zone interestingly also has an average SFRM thickness of 2.993 inches as opposed to 2.5 inches for all upgraded floors (and compared to the specified 1.5 inches!), and its cohesion/adhesion strength averages 381.5 psf rather than the average 298.2 psf for upgraded floors (as might be expected with additional epoxy). Thus, the 303.3 kg/m^3 (or less if we exclude floors 94-98) SFRM density on upgraded floors over 2.993 inches (0.076 m) is 0.076 m * 303.3 kg/m^3 = 23.05 kg per m^2 of steel. And 0.076 m * 340.3 kg/m^3 = 25.86 kg for WTC1 floors 94-98, placing the excess, inserted material at 2.81 kg, or 1.405 kg each of thermite and epoxy per m^2 of steel.

      The total yield is 1.405 * (3.9 MJ/kg + 25 MJ/kg) = 40.6 MJ/m^2 of steel surface. So the potential temperature increase in the steel is 40,600,000 / 0.0094 m^3 of steel / 7850 kg/m^3 / 550 J/kg.K = 1,000 C before even allowing for the material being on all sides of the trusses or for heat from the office fires. (At these temperatures the office fires are going to reduce heat escaping from the steel rather than making it even hotter.) There is energy to spare for gas generation and possible blast effects as required.

    3. I'll correct my errors of units above that I didn't spot at first. Where I wrote "240 kg/m^3 (15 psf)", the "psf" should of course read "pcf". And the "18.93 psf" is 18.93 pcf. Also, the "0.076 m * 303.3 kg/m^3" and "0.076 m * 340.3 kg/m^3" should really be "0.076 m^3" rather than "m".

      If you exclude the 15 SFRM density measurements from WTC1 floors 94-98, the average of the remaining values is 18.22 pcf (291.9 kg/m^3) instead of 18.93 pcf (303.3 kg/m^3). Thus, at the "normal" density (as per the other floors) the expected mass of SFRM for floors 94-98 (given its 2.993-inch average thickness ) is 0.076 m^3 * 291.9 kg/m^3 = 22.18 kg per m^2 of steel. And the actual value for WTC1 floors 94-98 is 0.076 m^3 * 340.3 kg/m^3 = 25.86 kg. The excess of 3.68 kg is ~31% up on the 2.81 kg calculated previously, and the potential to heat (or blast) the steel with embedded energetic materials increases accordingly.

    4. Poseidon,

      "the red/gray chip = thermite theory fails because of insufficient thermite, as well as too little thermite in relation to organic material."

      We are in full agreement here, and that could conclude the debate to this blog post.
      The Thermite theory als fails because there is no evidence for it.

      "red/gray paint chips (including Basile's chips) were probably contaminated with some energetic material that was used for the demolitions"

      How can you say that? Surely they were contaminated somehow, but what evidence do you have that it was anything energetic? Does this follow from any data provided by Farrer, Jones, Basile, Millette or Couannier?

      "Under one scenario (thermite only added to the SFRM)"

      This is a specultation that has nothing to do with red-gray chips. Is there any evidence for it? You mention measured densitiues of the SFRM - is that it??

      "The 23% thermite by mass is 4.43 kg, so at 3.9 MJ/kg the temperature rise in 0.37-inch steel (0.0094 m) is 17,277,000 / 0.0094 m^3 / 7850 kg/m^3 / 550 J/kg.K = 425 C."

      This calculation suffers from three errors:
      1. You forget to include the other 77% mass of the SFRM, which also gets heated and thus uses up some of the energy. If you assume that SFRM and thermite are intinmately mixed, the SFRM gets heated to (almost) the temperature of the thermite, and you might lose more energy to melting or even vaporizing.
      2. You assume that ALL the energy from the thermite goes into the material it is attached to, but a significant portion gets lost to radiation and hot liquid thermite products dropping away from the steel. Assuming that half the heat goes into the steel is probably already optimistic; all of it impossible
      3. You go with the theoretical max of thermite's energy density. In the real world, the value would be < 3.9 MJ/kg.

      And also, if you are only going to weaken the steel rather than melt it, thermite, because of its relatively low energy density, is not well suited for that. It would be much smarter to use organics then, perhaps augmented with some oxidizers, such as potassium permanganate.

      Are you aware that Frank Greenings proposed a theory similar to yours, except he speculates about (if I recall correctly) ammoniumperchlorate in the SFRM. I am currently looking for a thread at the911forum, can't find it somehow... Anyway, I only skimmed that proposal, as it was only half-serious.

    5. Addendum:

      Here is Frank Greening's The Ammonium Perchlorate Theory

      It's not presented as a candidate theory for what really happened on 9/11, but as an intelligent fantasy, with the set-up "If I wanted to surreptitiously destroy the WTC complex I would...".

    6. Oystein,

      I've made my calculations more "jet fuel-friendly". We know that the Jones / Harrit / Farrer data provides evidence of elemental Al and, more importantly, iron-rich spheres observed after heating of paint chips. No credible explanation other than thermite has been offered for the iron spheres, so they are evidence for thermite. Here's a couple of excerpts from my forthcoming report, which shows how there is an exact match between high density upgraded SFRM and impact zone floors for WTC1:

      "The information from NIST's Table 4-2 is tabulated below to show the mean values for average SFRM thickness, cohesion adhesion, and density. For a "repeated test", both values are counted. It can be seen that for WTC1, floors 94-98 all have an average measured SFRM density of 19.5 lb/ft3 or higher, and the remaining floors are all 19.0 lb/ft3 or lower with the exception of the 1997 report for the multiple tenant floor 85. The density is quite low in the 1999 measurements for floor 85, and some of the fake SFRM that was left over following the 1996 renovation to floor 94 may have been applied to a section of floor 85 in 1997."

      [The floors I'm referring to are those that had SFRM upgrades. I won't include my 7-column, 25 row table, as it doesn't format well on here. But here's part of the information showing the average measured density on WTC1 upgraded floors.]

      WTC1 Floor, Density (pcf)
      79, 16.6
      81, 19.0
      81, 17.5
      83, 16.0
      85, 23.7
      85, 15.8
      85, 16.4
      92, 17.9
      93, 17.0
      94, 20.6
      95, 19.5
      96, 20.0
      97, 23.5
      98, 22.6
      99, 17.9
      100, 17.9
      102, 16.4

      "For a 110-story building, there are 106 sets of five contiguous floors, from 1-5 through to 106-110. The number of sets is 110 + 1 - n, where n is the number of contiguous floors. The AA Flight 11 impact zone spanned five floors and the roll angle was 25°; UA Flight 175's impact zone correspondingly encompassed seven floors at a roll angle of 38°, give or take a few degrees. In both cases, if the height of the impact zone in feet is approximately 12n - 6 where n is the number of floors spanned, then the hypotenuse of the right-angled triangle with adjacent angle equal to the roll is approximately 127 feet. This is the length of the wingspan that causes damage (cf. the 156 feet 1 inch wingspan), with the wingtips doing negligible or no damage. If we include all sets from two to ten contiguous floors (which represents various roll angles up to 60° for a Boeing 767-200 series), the total becomes 111 - 2 + 111 - 3 + 111 - 4 ... + 111 - 10 = 9 * 111 - ( 9 * 6) = 945. Thus, the probability that the aircraft impact zone would cover the same number of floors and the very same set of floors as those upgraded floors with an SFRM of particularly high density, is 1 in 945."

      My original thinking was that the perps may have switched all of the SFRM used for the upgrades, not knowing if they would get the five contiguous floors that they needed to attack (so that too many core columns buckle from a lack of lateral support and even the strongest columns cannot handle the redistributed load). But Deloitte & Touche moved out (after being badly disrupted by the 1993 bombing), providing the opportunity for SFRM upgrades on the vacated floors. Floors 94 -102 of WTC1 became the "last large block of contiguous space available in the complex". A 16-year lease deal was signed with Marsh & McLennan in May 1998 for space on floors 94-100 along with part of the 93rd floor. Given the 1 in 945 exact match, there is now little doubt that accelerants were applied to the WTC1 impact zone, but different methods were used for each building. WTC2 had the yellow-orange melt (Fe-S-O) pouring out in the final minutes, and the CIA and other agencies had demolition access to Building 7. [Tbc]

    7. [Cont.] If the heating effect on the steel is too low I could revise my hypothesis and increase the yield by having a greater than 1:1 ratio of organic material to thermite, or even increasing the accelerants in proportion to SFRM. However, that's not necessary. For the moment, I'll stick with my second hypothesis from comments above.

      NIST NCSTAR 1-6A, Table 6-5, shows specific heat capacity of Cafco Blaze-Shield II as 801.6 J/kg.K (25 °C), 1255.3 J/kg.K (500 °C), and 1411.3 J/kg.K (1000 °C). Let's take 1255 J/kg/K as an average, and 550 J/kg.K for steel. (That's jet fuel-friendly for predicted temperatures of around 500 °C, as the average SFRM heat capacity would be lower.) I'm going to assume a mere 3 MJ/kg from the thermite, although the theoretical maximum for Al/Fe2O3 is 3.985 MJ/kg. 80 nm Al has an Al2O3 shell of about 30%, which takes it down to 3.596 MJ/kg maximum, or lower if the Al2O3 proportion was not accurately known, and / or the proportions of fuel and oxidizer weren't adjusted by the correct amount to compensate for the oxide shell. But I now suspect that most of the thermite's Al was micron- than nano-sized.

      The "normal" values for SFRM density are those on the other upgraded floors: an average of 291.9 kg/m3. The impact zone floors are 340.3 kg/m3. I contend that the excess 48.4 kg/m3, at least, is the accelerants. And I suppose that the accelerant is split equally between thermite and organic material. When I do calculations on how much energy was released by the office fires, I use a multiplying figure of 0.68 to allow for the reduced yield under fuel-rich conditions. Kawagoe and others used a value of 10.78 MJ/kg as the calorific value of wood in fully developed ventilation-limited fires, which compares with around 16.5 MJ/kg or more at equivalence ratio of 1. (Their 0.65 factor is a greater effect that the 0.68 I arrived at in private studies on 9/11.) Since I am not proposing an oxidizer for the organic polymer, this must obtain its oxygen from the oxygen-depleted air in the office fire. So to be fair, I am multiplying my original figure of 25 MJ/kg by 0.68 to get 17 MJ/kg. Although some heat is radiated away from the accelerants, heat is also absorbed from the office fires.

      In the general case, the temperature increase (°C) in the steel and the SFRM of the fake SFRM/accelerants composite is equal to:

      (Ppoly * Ypoly + Ptherm * Ytherm) * 340.3 * Vfake
      (1 - Ppoly - Ptherm) * 1255 * 340.3 * Vfake + 550 * 7850 * Vsteel

      ...where Ppoly and Ptherm are the mass fractions of the inserted polymer and thermite respectively in the fake SFRM, Ypoly and Ytherm are the yields in J/kg respectively of the polymer and the thermite, Vfake is the volume of the fake SFRM in m^3, and Vsteel is the volume of the steel in m^3. The 340.3 is the density of the fake SFRM in kg/m^3, the 1255 is the specific heat of the SFRM within the SFRM/accelerants composite in J/kg.K, the 550 is an average specific heat for steel in J/kg.K, and the 7850 is the density of steel in kg/m^3. The top half (numerator) is the yield in J, and this is divided by the energy requirements to heat the SFRM and the steel by 1 °C.

      This simplifies to:

      Ppoly * Ypoly + Ptherm * Ytherm
      1255 * (1 - Ppoly - Ptherm) + 12,687 / VolRatio

      ...where VolRatio is the volume of the fake SFRM/accelerants composite divided by the volume of the steel. To get the volume ratios, we just divide the cross-sectional area of the fake SFRM by the steel cross-section, so this will apply for various lengths of steel. The above formulae can be used to calculate the temperature increase for various mass fractions and yields of polymer and thermite. Given the values I am assuming here of 17 MJ/kg for the polymer and 3 MJ/kg for the thermite, the specific case simplifies to:

      Trise (°C) = 1321 / (1 + 11.78 / VolRatio)


    8. [Cont.] Here are some examples of VolRatio, Trise: 1, 103; 2, 191; 4, 334; 6, 445; 8, 534; 10, 606; 12, 666; 14, 717; 16, 760; 18, 798; 20, 831. It can be seen that in the limit, where there is no steel and the requirement is only to heat the SFRM, the predicted temperature rise is 1321 °C (or lower if there are phase changes).

      Now let's take the case of the 1.09 inch diameter diagonal rods in the trusses that connected the top and bottom chords. The area of the steel cross-section is Pi (0.545)^2 = 0.933 ins^2. If the fake SFRM is sprayed on to a depth of 3 inches, the SFRM cross-section is the difference between the areas of two concentric circles: Pi (R^2 - r^2) = Pi (3.545^2 - 0.545^2) = 38.5 ins^2. Thus, the VolRatio of fake SFRM to steel is 41.3:1, and the temperature increase in the steel and the SFRM is 1027 °C. Let's make this more jet fuel-friendly, and suppose that there is less SFRM on the top. Multiplying the SFRM cross-section or the VolRatio by 2/3, we get 27.5 as the volume ratio, which corresponds to a temperature increase of 924 °C. The average measured SFRM thickness over floors 94 to 98 was 2.993 inches, although there were considerable variations in both thickness and density. That worked in the perpetrators' favor, since they didn't need to weaken all of the steel; only particular places along the length of a truss.

      Next, take the case of the bottom chord, which is comprised of two L-shape angles, L3 x 2 x 3/8 inches (e.g., see NIST NCSTAR1-2 2.4.1). The specifications are given at this page:

      The area is shown as 1.73 ins^2. It's roughly 3 x 3/8 + (2 - 3/8) x 3/8, which works out at 1.734 ins^2. The angles are arranged as short legs back to back, with some 1.09 ins between them (so that the diagonal rods can fit in). So there are two L-shapes; the vertical part is the 2-inch section, with 3-inches horizontally, back to back (the left-hand one is laterally inverted). Working clockwise starting from beneath the bottom right, there is a 3 x 3 inches area of SFRM, then 3 x 1.09 between them, then another 3 x 3 beneath the left angle. Then to the left, a 3-inch thickness should really mean a semicircle of radius 3 inches. Let's be jet fuel-friendly and suppose it's only 1.5 inches. Then the area above the left angle goes from a height of 1.5 inches up to 2.5 even if it is only 1/2-inch deep at the top (compared to the minimum spec of 1.5 inches and measured value of 2.993 inches!), making an average of 2 x 3. The middle top section has an area of 2.5 x 1.09. Above the right L-shape, there is another 2 x 3 area, and the area to the right is another 1.5-inch semicircle. The total area of the fake SFRM is 9 + 3.27 + 9 + 3.53 + 6 + 2.72 + 6 + 3.53 = 43.05 ins^2. Dividing by the area of the steel, which totals 3.46 ins^2 from two angles each of 1.73 ins^2, the volume ratio of fake SFRM to steel is 12.44:1, and the predicted temperature increase in the steel and the SFRM is 678 °C.

    9. Cool story, bro, but...

      1.Got any evidence that the SFRM was in fact loaded with baddy stuff?
      2. Thermite is still a bad choice if all you want to do is add heat to just weaken, not melt, steel.

      Not going to even read all the mathy stuff. Without having the base premises right, the best math won't get you anything useful.

  3. Poseidon, I am sure you have figured out by now that you
    are wasting your time on this blog..

    Contact the authors of the original NT paper for proper correspondence

    You might even find out that they considered your hypothesis...

    and that they have further information which helped to get to their final conclusion...

    such as the unpublished data, and maybe even later work.

    We can both resume talks with Oystein et al, whenever they confirm their "cool story" of paint that produces molten iron spheres...

    until then, dont waste your time

  4. "the red/gray chip = thermite theory fails because of insufficient thermite, as well as too little thermite in relation to organic material."

    We are in full agreement here, and that could conclude the debate to this blog post.
    The Thermite theory als fails because there is no evidence for it.

    GREAT lets throw away the scientific method and instead use persuasion as a means by which to turn science into information imperialism. I can't understand most of the mathematical jargon on this post BUT even I am not stupid enough to believe for one second that these investigations are conclusive by any means, and these papers RARELY are. They observe, hypothesize, predict, test, and evaluate, but do NOT come to infallible conclusions. That's why evolution and the big bang are theories and not laws. Even the once infallible Laws of Motion by Newtown have vast exceptions in the quantum realm. Here's what i DO know: if you type in MVA9119 into this webpage, there is no record of the study on the site. Perhaps more importantly, these speculations are vastly based on the supposition that the chips are from primer paint and not thermite. So how about this little bombshell from your beloved Millette study:

    According to the MSDS currently listed on the Tnemec website, 55 out of the 177 different Tnemec coating products contain one or two of the three major components in
    the red layer: epoxy resin, iron oxide and/or kaolin (aluminum silicate) pigments. However, none of the 177 different coatings are a match for the red layer coating found in this study.

    My question is, how the hell can you post all this mathematical mumbo jumbo yet blatantly fail to recognize that the materials tested in the Millette study were not consistent with ANY of the primer paints used on the steel in the WTC? Why is there no record of the study on MVA's website?

    Want proof?

    "Without having the base premises right, the best math won't get you anything useful." - YOU

    You sir, have dug your own grave.

    1. Sorry I turned my attention away from internet discussions of 9/11 in summer 2013, so here is a VERY late reply:

      You are probably (I didn't check, but take your or Millette's word for it) correct to point out that the chips a-d in the Harrit study do not match any of the currently listed Tnemec primers, and I'll even add one more: They are definitely no match for the contemporanous WTC Tnemec primer as found on perimeter columns!

      However: Not all WTC steel was painted with Tnemec primers: LaClede Steel Company, who manufactured the floor joists, used their own shop primer, and the specified composition for that is a near-perfect match for those chips! See my blog post "Another primer at the WTC: LaClede Standard Primer" at

      I don't know what MVA's policy is concerning publication of their cases on their website. I'd actually surprised if they had a habit of writing about ALL their projects and jobs. I suppose that many customers would have plenty of good reasons not to see that stuff in public.