Origin of the paint that was found as red-gray chips - any ideas?

To recap:

In Harrit e.al. (Active Thermitic Material discovered, "ATM"), there are 4 chips, (a)-(d) that look like identical material in photomicroscopy, electron microscopy, and their respective red and grey layers have rather similar XEDS spectra.
However, they of course don't all have the exact identical look and exactly identical spectra.
So we have considered them to be the same material with high confidence, but should be aware that of course there's chance that we err.

For example, the red layer of chip (c) has these inclusions of Na and K that the others don't have, and I have no explanation for these. So maybe, just naybe, (c) is really something similar, but different? We should remain open to that option.

Looking veeeeeery closely at the four red-layer spectra, I can see tiny signals for Cr in chips (a), (b) and (d). In (a) it is really very tiny and I admit I wouldn't be confident about its presence at all if Harrit had not labeled the Cr signal in his more detailed reproduction of the XEDS graph of chip (a) in his letter fron May/June 2009 ("the Letter"). In the same Letter, Harrit has also labeled Sr for chip (a). So this is the only indication we have that somewhere there might be strontium.

Remember, Harrit's letter gives us Cr and Sr for chip (a)

It is also chip (a) that they used for the XEDS map (Fig. 10 in ATM), and that is the figure where I found the two needles that appear to contain neither Al and Si nor Fe, and their shape is consistent with that of SrCrO4 pigments. That is all the evidence we have in favour of strontium chromate. It rests on a single chip, (a).

(Again, there seems to be a bit of Cr in chips (b) and (d), too, but no info about Sr, and no further "needles")


So what are we to make of this? There must be paint from the floor joists somewhere in the dust, and I think it would be relatively plentiful. I am also quite certain that chip (a) has strontium chromate.

So what's with the other chips, and those that Millette studied?

  1. Perhaps Millette's aren't a match for (a)-(d) after all. That would be quite a disappointment at this stage
  2. Perhaps LaClede used the Formulation 10001 initially - on some floor joists - but changed it later. Other chromates I have cone across that are used in paint are with zinc or lead, but those two don't appear in the spectra. Would calcium chromate be a possibility? That's a yellow substance, too, that was used as pigment and corrosion inhibitor, and Ca seems to appear in chips (b) and (c) as well and perhaps eben in (d).
  3. Perhaps LaClede dropped chromates altogether (but we do have Cr signals in 3 chips...)

Did Millette find no chromium at all?

We really need to see the detailed data...
 
PS I have asked the thread guardians to unmomitor my new Millette thread. Will let you know when they do this. I can't even edit my own posts on a moderated thread!
 
'Cr' is present several times in the spectra (page 28 last graph on the left and others).

As for Harrit's paper.

Laclede paint was in Millette's sample... with other paints...
 
Last edited:
Oops sorry, bad math, took a shortcut. revised % of pigment and vehicle components ( not including water)

Perhaps because he was looking at the “Unknown” spray primer paint applied to the core columns and floor WF beams , painted atop a mild layer of “rust”. The known exterior columns Tnemec paint was a comparatively small area, with little chipping in the collapse and a minor contributor to the dust.

Perhaps the electrocoated (anodic or cathodic?) Laclede primer does not have a “rust” gray layer, yet retains its magnetic properties due to its iron oxide (25.8%) + Al Si (19.3%) +SrCr (1.9%) + (53%) epoxy total for pigment and epoxy (the deionized water evaporates and not counted)

The electrocoating process may be anodic or cathodic, depending on the charge applied the substrate. Although the processes are virtually the same, properties of the resultant coating are dissimilar. Anodic systems, which were the first to be used for electrocoating, apply paint to positively charged substrates. The negatively charged pigment and resin particles deposit onto the substrate (anode). One disadvantage of this process is that substrate metals dissolve and become incorporated into the coating, which affects surface properties.
I wonder what this means to a metallurgist and the likelyhood of finding a gray ("rust") layer under an electrocoated paint on steel not exposed to the weather.

[URL]http://www.engineershandbook.com/MfgMethods/electrocoating.htm[/URL]
The water content of the paint/vat solution evaporates during the subsequent baking (20 min at 350F or so) after the immersion of the trusses. The primer and epoxy remain as a hard coating on the steel.

ETA:
The first patent for a cathodic EPD product was issued in 1965 and assigned to BASF AG. PPG Industries, Inc. was the first to introduce commercially cathodic EPD in 1970.
wikipedia. Towers were started 1968 completed in 1970 and 1971. The trusses were installed before the process was available = anodic electrocoating.

I've seen Galvalume zinc over steel sheets electrocoated for > 20-30 years not exposed to rain, and the surfaces look almost new no rust.
 
Last edited:
pp 6-7 'Somesmall EDS peaks of zinc and chromium were detected in some samples but the amountdetected was inconsistent with the 20% level of zinc chromate in the primer formula.'

Based on what we do know about LaClede formulas of the day, is it not fair to say that this material is very similar to it in composition?
And further, can we not also surmise why this kind of material would be used in the construction of the towers?
Why else would it be present in the dust? I can't think of any other reason..
 
Chris: Thanks a lot for the link to Jim Millette's figures and spectra:cool: I'm responding here since the moderation of Jim's study thread still seems to be rather "clumsy" to me;)

I have just came back from the pub enjoying some beers, but anyway: looking at the spectra and other
results, I'm even more convinced now that Jim has (at least mostly) closely studied the same kind of chips as Bentham chips (a) to (d), i.e. "Laclede primer paint" :cool:).

Well, I will reconsider this claim tomorrow morning:cool:
 
Last edited:
Chris: Thanks a lot for the link to Jim Millette's figures and spectra:cool: I'm responding here since the moderation of Jim's study thread still seems to be rather "clumsy" to me;)

Frankly, I have just came back from the pub enjoying some beers, but anyway: looking at the spectra and other results, I'm even more convinced now that Jim has studied the same chips as Bentham chips (a) to (d), i.e. "Laclede primer paint" :cool:).

Well, I will reconsider this claim tomorrow morning, but I have little doubt that I will change this opinion as sober:cool:

LOL :alc:
 
By the way:

I just found that Harrit has made a stupid mistake in his May/June 2009 letter, where he shows that chip (a) is not Tnemec! Quite an obvious mistake actually, but it seems no one has ever checked his calculations!

Here is a version of that letter: http://michaelfury.wordpress.com/2009/06/21/occams-razor-removes-paint-a-primer-by-niels-harrit/

In Figure 3, he shows the list of ingredients:
4 pigments
7 substances for the pigment
All with a value of weight-%.

For example, Zinc Yellow is given as 20.3%, and Thinner is given as 32.2%.

Later on, Harrit states in Table 1 that Zinc chromate is 20.3% in the wet paint. He argues: Since 32.2% thinner from the vehicle and 7.6% spirits from the proprietary Tnemec pigment evaporate when the paint is baked, the weight proportion of Zinc chromate increases to 34% in the dried paint.



This calculation misses an important point:
The weight-percentages in Figure 3 add up to 200%, not to 100%!
20.3% Zinc Yellow is actually only the proportion among the pigments! The % values for the pigments add up to 100%, and the %-values for the vehicle constituents add up to 100%.
What we don't know from Harrit is the ratio pigment:vehicle in the wet or dry paint.

We know from the LaClede spec, that total pigment can be as little as 28.5% of the paint (wet=dry I suppose in that case, as epoxies are not volatile).

Also, he overestimates the proportion of volatile Thinner and Spirits in the wet paint; he claims that 40% of the wet paint are volatile, when the true value is closer to, or even below, 30% (I'll calculate that later)

Sooo Harrit's figures for the weigh % of several metallic ingredients need to be corrected downwards by a factor of roughly 4: Zinc chromate is only around 8% of the dry paint, not 34%; Talc (Mg3Si4O10(OH)2) only 3-4%, not 12-17; etc.

This instantly explains why the XEDS peaks for Zn, Cr, Mg are not bigger than they are in Figure 6 (the MEK-soaked chip) and in Steven Jones's analysis of actual Tnemec from WTC columns (see image below):
JonesPrimerPaint-2m48s_bottomright.jpg
 
My very first notes on Jim Millette‘s study results.

After some two hours spent on them (which is of course not enough), I have these “feelings” and opinions (I consider them as “running remarks, not conclusions in any way):

The whole set of results is quite complex and less clear than we have perhaps hoped, but this is just normal in science.

From the results (namely XEDS spectra) it is not really clear if all (or most of) red-gray chips are particles of Laclede paint (indeed, all of them are some kind of paint and none of them can be any thermite, e.g. because of very high carbon content). But, as far as I understand, all chips contain the same/similar main pigment particles and red layer thicknesses are consistent with Laclede paint.

Let me say now: if Jim Millette thinks that chips are basically the same material (the same paint on oxidized steel), I tend to trust him, just because of his immense and respectable practical experience with analyses of such samples. Simultaneously, I see quite big fluctuations in peak heights of detected elements – not only of minor elements like Ca, S, Ti, Cr and others, but also of major elements C, O, Fe, Al and Si. It seems that such fluctuations are normal in XEDS results and have many causes - and we should look more to some “overall, statistical, averaged picture” which is given by some set of spectra, instead of individual spectra.
Oystein: I think that we should (you should) slightly reconsider the approach: if there is some unexpected little blimp of something (e.g. titanium) or some expected blimp of something is missing (e.g. chromium), it is unlikely that it is Laclede paint (or Tnemec paint). It still can be this particular paint. Also, we should be more careful to take the heights of main peaks quantitatively.

Some preliminary comments to some figures.

Whereas XEDS spectra recorded on the surfaces of red layers (washed or unwashed) contain more or less significant peaks of Ca (Fig. 9, Appendix B), Ca peaks for measurements on cross-section of the red layers (Fig. 12, Appendix D) are very little or none. This suggests (to me) that calcium peaks mostly originated from some surface contamination (probably by pulverized calcium sulfate from wallboards). Basically, XEDS spectra on freshly cut red layers are similar or very similar to XEDS of Bentham chips (a) to (d) (and they roughly correspond to Almond’s MonteCarlo simulations of the spectra), I think:o) Therefore, I think that at least those red layers which were cut and measured are the same material (Laclede red primer).

Very tiny titanium peaks are visible only on some “surface spectra” and they are basically missing in the spectra taken on cross-sections. For me, titanium found is no more problem at present.

As was already noted here by someone, chromium peaks are marked in some XEDS spectra. This could be some proof of strontium chromate we are looking for (strontium peaks cannot be visible). But, also other chromate can be present.

Concerning other figures and spectra, I will study them later. E.g. up to know, I have no idea why they are high peaks of copper in the TEM-SAED-EDS analyses of pigments after low temperature ashing, but it is apparently normal using this method.

Concerning FT-IR spectra of red layers, I simply like them:o) Just for fun, I have compared peaks attributed to epoxy resins and they are in a good agreement with the infrared spectra of epoxy polymers I linked several months ago.

That’s all for now, folks. Please consider everything I wrote as working remarks (based on the Jim's conclusion that all chips he selected using magnet are the same material) - not my claims:cool:
 
Last edited:
...
Also, he overestimates the proportion of volatile Thinner and Spirits in the wet paint; he claims that 40% of the wet paint are volatile, when the true value is closer to, or even below, 30% (I'll calculate that later)
...

Actually, no, I won't calculate the precise composition of that paint, because the data sheet on Tnemec pigment is too imprecise. Not important enough at this point.
 
He was aware to look out for strontium chromate because that's in a particular paint formulation, but he didn't find any. So the chips that Jim descibes are NOT a perfect match for LaClede paint. The chips that Harrit describes are, albeit with the strontium chromate signal being weak and coming from one chip (a) only. (If you disregard SrCrO4, the match is perfect, but a red paint with hematite and kaolinite is so mundane and usual, it could be LaClede, or could be some other paint. In my view, it's the SrCr=4 that makes LaClede stand out)

I think there is an easy explanation for the absence of strontium chromate: Dr. Millette cleaned the chips with water!

Strontium chromate is more or less soluble in water depending on the pH of the solution. In alkaline solution strontium chromate is only poorly soluble, but in neutral and acidic it is much better soluble. I assume that he dissolved the strontium chromate during the cleaning process. It would be interesting to know if he tested the water for strontium chromate (I only have a slow connection since wednesday and made no attempt to download his report).
 
I think there is an easy explanation for the absence of strontium chromate: Dr. Millette cleaned the chips with water!

Strontium chromate is more or less soluble in water depending on the pH of the solution. In alkaline solution strontium chromate is only poorly soluble, but in neutral and acidic it is much better soluble. I assume that he dissolved the strontium chromate during the cleaning process. It would be interesting to know if he tested the water for strontium chromate (I only have a slow connection since wednesday and made no attempt to download his report).

Interesting. The chromate is blended with the other pigment materials, blended with the epoxy, electrocoated onto the steel and then baked until dry and hard.
Is it likely that it would wash off?
 
Africanus: I have found that strontium chromate is reasonably soluble only in mineral acids (e.g. hydrochloric) and its solubility in pure water is low (ca 1 g per liter). And there is no mention in the Millette paper about washing with acid.
Just hypothetically, some strontium chromate needle crystals can be close to the red layer surface or exposed directly to the solvent and perhaps it can be partially dissolved even in water. But... just hypothetically:cool:

Concerning XEDS spectra in Millette paper, I tend to consider only those which were recorded on freshly cut cross-sections of red layers (Fig. 12, Appendix D). Frankly, XEDS spectra of Jim Millette do not have really good resolution, e.g. because peaks of carbon are really high (clear proof of prevailing polymer binder, i.e. paint:rolleyes:) and all other peaks are therefore comparatively small. But we have to live with this (at least so far).
It seems to me that the most of red layers in Appendix D are Laclede paint, but there is one exception, as pointed out by Sunstealer in the "Millette study thread" (sadly, completely spoiled now by Senenmut:():
Sample 5230M3451B shows significant peaks of calcium and apparent peaks of sulfur and magnesium even on clean surface, so this particle could be different paint than Laclede red primer...
Now I see some other XEDS taken on some red layer washed surface (Appendix B, Sample L1560 - Murray & Church St., the first spectrum on the right). It corresponds to the Tnemec primer XEDS quite well, I think. And they are more such spectra in Appendix B, so these particles can be Tnemec as well. Problem is that some common elements (Ca, Si, S, Mg)) can come from contamination. I will check it later, it seems that my first intepretation of XEDS was not good...
 
Last edited:
Well, I have to reconsider my initial opinions, as for Jim Millette XEDS results, and also my last post with rather weird remarks:

In fact, XEDS measured on washed surfaces of red layers (Appendix B) mostly correspond more to Tnemec paint than to Laclede paint. They have invariably some little Mg peak, distinct peaks of Ca, some Cr is sometimes detected and even Zn peak is marked in some spectra. Also, Si peak should be higher than that of Al in Tnemec paint and this is also frequently observed in Appendix B. Generally, some "XEDS fingerprint" of Tnemec paint is probably detected here.

On the other hand, almost all XEDS spectra taken on cross-section of red layers correspond to Laclede paint.

To be honest, I have no idea why it is so and what I have overlooked:confused: Anyway, it seems to me now that Jim studied (at least) two different paints (as well as Harrit et al, btw:cool:)
 
Last edited:
Interesting. The chromate is blended with the other pigment materials, blended with the epoxy, electrocoated onto the steel and then baked until dry and hard.
Is it likely that it would wash off?

Washing removes loose particles from the surface, and might conceivably dissolve a tiny bit of the chromate if it happens to be right on the surfcase - sticking out of the epoxy matrix, so to say.

I believe that any pigments inside the matrix would be visible easily to electron microscope as well as to XEDS measurement, but not wash out under brief treatment with plain water.
 
We believe that the red-gray chips are primer paint (red) attached to spalls (gray) that separated from the surface of structural steel.

I want to understand if and how this could actually happen, this spalling of steel member surfaces with paint attached. Wouldn't the paint detach from the gray stuff sooner than the oxide layer detached from the steel beneath? And why are the gray layers never thicker than 100 µm (as per Harrit e,al.) or even 50 µm (as per Millette)?

So far I assumed that the main mechanisms that cause these spalls to enter into the dust as the towers collapse are
  • Grinding / abrasion, as concrete and other steel scratches over painted steel with great pressure (movement parallel to surface)
  • Fracture upon imact, as the same materials hit with great (movement perpendicular to surface)
In both cases, spalling occurs on the side of the steel member that is impacted
A third mechanism could be
  • Bending / torsion: As steel members get deformed during collapse, differentials in tension or pressure between layers within the member leads to cracks




Ijust found another mechanism, known from anti-tank warfare, which acts on the side of the steel member side opposite from the impacted side:
In other words: If you hammer on one side of the steel, shock waves fling off spall on the other side.
This is only true however, I think, if the waves that travel through the steel are in fact shock waves - i.e. faster than the speed of sound in steel. Would this only come from sufficiently high explosives?
(No I am not going to give truthers a hint :D)


Please any comments from people who know a little about metallurgy, armour, etc.!
 

ISF - Join now!

Every member here is approved by hand. No bots, no spam, just people who care about evidence and honest debate.

Membership is free!

Create your free account

Back
Top Bottom