Concrete and XRF

[The Almond]

And as for X-ray analysis of materials, I did it for a living for 15 years. Peak heights for adjacent elements in an x-ray spectrum are a pretty good approximation because the x-ray absorption effects are about the same. Oh. and by the way, if chlorine was coming from the water it should be in every concrete spectrum. It isn't! Besides, chloride ion is a no-no in concrete and is generally kept as low as possible.

I hope you don't mind, but I've split this thread off from your other thread, Apollo20, so that you can use that one to talk about your more specific NIST criticisms.

First, I wanted to discuss some of your comments on X-Ray fluorescence. Specifically, I was hoping we could talk more in detail about your analysis of chlorine in concrete.

In my opinion, a pretty good approximation would be a reference to a chlorine standard or a chloride bearing compound such as Apatite or calcium chloride. Taking the ratio of the integrals of the unknown composition to the known composition, without correcting for X-Ray physics, would be a valid first order approximation. In truth, I've never heard of taking the ratio of the chlorine peak to the calcium peak, and I was wondering if you could produce a reference in the literature regarding that.

It is my understanding that, according to the method developed by J.V. Gilfrich, one measures a standard (of known composition), then the unknown, then the standard again. Once correcting for matrix and absorption effects, the integral of the unknown to the standard can be regarded as the composition with a reasonable degree of accuracy.

I noticed in the Concrete 02A that the chlorine peak (that is to say, the peak identified as chlorine) is just barely above background. These also appear to be electron excited X-Rays, as the presence of carbon and oxygen peaks indicates the presence of an ultra-thin window, which is not standard on XRF systems. Anyway, the full scale of the system is 38 counts/s, and the authors have not provided the dwell time in their spectrum image. Suffice it to say, the peaks appear to be nearly Gaussian, meaning that their precision varies with the inverse root of their peak integral. To acquire the three percent precision necessary to make a realistic determination of the chlorine content, you would need (1/0.03)^2 or about 1100 counts.

While I agree completely that the ratio of the chlorine peak to the calcium peak is probably about 3%, I question whether this is a valid conclusion given the nature of the data. Without standardization and ZAF correction, I know of no way to find these data to the degree of accuracy you quote.

It leads me into my next comment about chlorine in concrete. ACI 318 sets the limit for chloride ion content to 1% for structures that will be dry or protected from moisture. However the contractors and builders applied that rule when building the WTC, the standard does not apply to unreinforced concrete. Furthermore, ACI 318 was appended in 2002 and 2005 in order to tighten the regulations on chloride ions for prestressed concrete. I have been unable to find the original ACI 318 code, but I do not know if it would have applied during the building of the WTC towers.

My point is that you have two competing systems that the data do not have enough precision to resolve. You say that the presence of chlorine is indicative of some process that occurs during collapse, and I counter that the chlorine is a natural part of the system.

I look forward to discussing this with you!

 

[Augustine]

It leads me into my next comment about chlorine in concrete. ACI 318 sets the limit for chloride ion content to 1% for structures that will be dry or protected from moisture. However the contractors and builders applied that rule when building the WTC, the standard does not apply to unreinforced concrete. Furthermore, ACI 318 was appended in 2002 and 2005 in order to tighten the regulations on chloride ions for prestressed concrete. I have been unable to find the original ACI 318 code, but I do not know if it would have applied during the building of the WTC towers.

The first set of limits on chlorides in concrete was provided in ACI 201.2R-77 "Guide to Durable Concrete" in 1977 ( to the engineers who already knew that). It set no limit for above ground construction where the concrete would remain dry. In ACI 318-83 "Building Code Requirements for Reinforced Concrete", the limit was shifted to 1% for "reinforced concrete that will be dry or protected from moisture in service" (published in 1983, sorry for forgetting, another to the engineers).

(Side points: These limits are only for water-soluble chloride ion content. Also, the decks are reinforced concrete - they are reinforced concrete on metal deck, but there is reinforcing within the concrete, and would have to comply with ACI 318 whether non-composite or composite - non-composite the applicability of ACI 318 typically dictated by building code, composite the applicability of ACI 318 dictated via ASCE 3 referenced in the building code.)

Regardless, as the WTC were completed in the early 1970's, we can only speculate what the chloride ion content of the concrete was. I would think 1% would be a safe estimate. An interested party could contact Les Robertson and ask what the specifications said...

 

[WildCat]

The first set of limits on chlorides in concrete was provided in ACI 201.2R-77 "Guide to Durable Concrete" in 1977 ( to the engineers who already knew that). It set no limit for above ground construction where the concrete would remain dry. In ACI 318-83 "Building Code Requirements for Reinforced Concrete", the limit was shifted to 1% for "reinforced concrete that will be dry or protected from moisture in service" (published in 1983, sorry for forgetting, another to the engineers).

How was the concrete sampled? There were fire boats pumping ocean water onto the debris pile for weeks.

(Side points: These limits are only for water-soluble chloride ion content. Also, the decks are reinforced concrete - they are reinforced concrete on metal deck, but there is reinforcing within the concrete,

I don't think the decks in the WTC was reinforced.

 

[Gravy]

Interesting posts, guys. Thanks for the research. Augustine, my understanding is that except for some lower floors and mechanical floors, the tower floors were 4" unreinforced lightweight concrete over 22 gauge steel decking, with normal-weight concrete in the core areas. The only steel in the concrete in most office areas were the top knuckles of the trusses, which protruded about 3 inches above the decking to resist shear. The mechanical floors had shear studs welded to the decking every few feet.

 

[Augustine]

Interesting posts, guys. Thanks for the research. Augustine, my understanding is that except for some lower floors and mechanical floors, the tower floors were 4" unreinforced lightweight concrete over 22 gauge steel decking, with normal-weight concrete in the core areas. The only steel in the concrete in most office areas were the top knuckles of the trusses, which protruded about 3 inches above the decking to resist shear. The mechanical floors had shear studs welded to the decking every few feet.

It was reinforced with two layers of welded wire fabric (I don't have the full NIST report with me, see Progress Report, June 04, Ch 2 p 114( http://wtc.nist.gov/progress_report_june04/chapter2.pdf ) or NIST NCSTAR 1-6A, p.14. There may be other references than this, as I said I do not have the full report in front of me.)

Typically concrete slabs on metal deck are reinforced with minimum shrinkage steel (also referred to as "temperature steel") to control cracking. The reinforcement is welded wire fabric or wire mesh, and it can be draped so that the mesh is in the top of the slab over supports and in the bottom in the center of the span, or it can be placed in two layers. It appears that it was placed in two layers in the WTC towers.

Wildcat, I hope that answered your concern, too.

As to this:

How was the concrete sampled? There were fire boats pumping ocean water onto the debris pile for weeks.

I am probably completely unqualified to speak to dust sampling, when it was performed, etc...I am merely a simple structural engineer spending too many words to say: at the time of construction, there was no limit on the chloride content of concrete. Perhaps The Almond knows more about the sampling?

 

[WildCat]

Wildcat, I hope that answered your concern, too.

Yeah, I was thinking rebar. 3" rebar on 4' centers.

 

[The Almond]

The first set of limits on chlorides in concrete was provided in ACI 201.2R-77 "Guide to Durable Concrete" in 1977 ( to the engineers who already knew that). It set no limit for above ground construction where the concrete would remain dry. In ACI 318-83 "Building Code Requirements for Reinforced Concrete", the limit was shifted to 1% for "reinforced concrete that will be dry or protected from moisture in service" (published in 1983, sorry for forgetting, another to the engineers).

Thanks for the references on that. Frankly, for those types of things, it is useful to have, uhhh, shall we say...the voice of experience. Some of us who completed engineering school only a few years ago thought ACI 318-2002 was old news.