This experiment demonstrates that sufficient pressure was created to reach the failure pressure on SWB3 despite the presence of venting and the failure of the model SWB3 and FS panels...
...A spark was never used to ignite the ¼-scale tests and ignition energy studies were not a part of the ¼-scale testing program. Extensive results are available on ignition energy in separate laboratory tests [Exhibits 20T, L]. The ignition system used in the ¼-scale was actually a hot filament ignition system and no sparks were created.
The problem is that tests on the 1/4 scale model produced, at best, a slowly expanding flame-front (they mention taking several seconds to propagate throug the tank) that, in theory, could have raised pressure to the point you might get structural failure. This is not an "explosion," as normally understood.
Nor was it was ever shown that such a failure would break the nose off the plane - as opposed to, say, blowing out a panel in the tank. They never actually made the 1/4 size model to model structural strength, much less failure modes - just the dimensions. They never got the model to fail structurally. So they are left with saying that they produced enough pressure to, theoretically, make a real 747 fail.
It's not like there is a dire shortage of 747 airframes in which full-scale testing could have been done.
The article is also tendentious about how hard it is to accidentally get an explosive fuel-air mixture. The odds of this happening in the real world could, of course, be estimated by figuring out the full range of pressure and temperature combinations and seeing which are likeliest to occur. Tellingly, they cite data indicating the fuel in a real 747 could have been warm (60C) but no data on whether that would have, in fact, in a real 747, caused an explosive mixture. They were measuring the fuel temperature - odd that they would not measure the fuel in the air in the tank. Perhaps beacuse the odds are astronomically low in the real world, with realistic scenarios.
The following may help you and other readers to better understand the mass of mess the NTSB inherited from the FBI. Sorry for the length but it's only a small fraction of the reports noted.
This report refers to FBI documents of any type pertaining to interviews as "witness documents." These documents are summaries of some of the information provided to FBI agents by witnesses during interviews conducted as part of the FBI's criminal investigation. No verbatim records of the FBI interviews were produced. The documents are almost exclusively written in the words of the agents who conducted the interviews, and not in the words of the witnesses themselves.
The documents were created to capture information relevant to its criminal investigation, and FBI agents frequently included only information that appeared relevant to this purpose. Witnesses were almost always interviewed by more than one FBI agent (or other law enforcement personnel), one of whom served as the note taker.
A witness document was prepared later by reference to these notes. Although some agents typed these documents themselves, many were prepared by typists by reference to handwritten drafts. The agents reviewed the typed documents for accuracy. This review was often accomplished shortly after the interview, but due to the large number of interviews being conducted, a backlog developed, and sometimes several days or weeks elapsed before handwritten interview notes were typed and reviewed.
The witnesses themselves were not asked to review or correct the documents. Because of these factors, the witness group avoids referring to the witness documents as "statements."
It appears that during some interviews the questions asked by the FBI agents were framed in a manner that emphasized aspects relevant to the missile investigation. In fact, some suggested interview questions are included in document CC-5, which pertains to Witness 32. Some of these include:
What was the timing of events? How long did the missile fly, etc. What does the terrain around launch sight look like? Were scorch marks visible? Where was the sun in relation to the aircraft and the missile launch point? The witness group and the document readers found that a number of other aspects of the witness documents make it difficult to extract accurate and reliable information from them.
These include possible interviewer and interviewee bias, ambiguous clock-point and angle references, potentially inaccurate distance estimates, combined accounts, reporting of witness speculation and conclusions, imprecise or vague language, internal inconsistency, and errors concerning the origin streak of light. Each of these issues will be discussed along with an example or two.
Possible interviewer and interviewee bias. As mentioned previously, FBI witness interviewing was focused on the possibility that a missile had been used against the accident airplane. This focus may have resulted in bias on the part of some the interviewers. For example, the document (CC1-628) pertaining to Witness 590 (10.4 nautical miles slant range from flight 800, interviewed July 20, 1996) describes an ascending red ball. 10,11 The document further states, "Upon impact, [redacted] observed a large fireball." Neither the document nor the attached interview notes states that the witness saw anything other than the red ball in the sky; thus, it is unclear what the "impact" mentioned in the document is. Although this incongruity could be due to other reasons, the witness appears to have described an ascending red ball and a large fireball, which the interviewer related using the word "impact."
For a similar example, see document CC1-382, which pertains to Witness 411 and Witness 412 (both of whom were 8.2 nautical miles slant range from flight 800, interviewed July 20, 1996). These witnesses describe seeing a flare-like object, but "they did not see what [the] flare struck, but it exploded in air into a large orange fireball." This characterization may suggest that the interviewer and/ or the interviewees believed that the flare-like object was a missile, which must have struck something. Beliefs concerning the possibility of a missile attack may have biased or colored the word choices used in reporting the witness accounts; therefore, these accounts must be interpreted carefully.
The presence of missile experts at some interviews may also have influenced these biases.
One document pertaining to Witness 243 (CC4-146, interview date not provided) states that the witness saw an ascending object and then an explosion. However, another document pertaining to this witness (CC1-28, interviewed July 18, 1996) states that the witness (12.0 nautical miles slant range from flight 800) noticed something similar to a flare and "the flying object was relatively slow in flying up and took about four or five seconds before hitting the airplane." Though the document implies that the witness saw an airplane, the document does not specifically state that the witness actually saw an airplane. The interviewer or the interviewee may have used these words to convey that the witness observed an explosion after seeing the flare-like object.
Some documents noted that the witness did not realize what he or she was observing, and some documents specifically state that the witness made conclusions about what he or she observed after learning about the accident in the media.
For example, the document pertaining to Witness 326 (CC-368, 13.2 nautical miles slant range from flight 800, interviewed July 24, 1996) states that this witness did not think much about what was observed until watching the evening news.
The document pertaining to Witness 271 states that she did not realize that she had observed an airplane crash until about an hour later when family members told her that TWA flight 800 had exploded.
The document pertaining to Witness 166 (CC1-374, 31.0 nautical miles slant range from flight 800, interviewed July 26, 1996) states that this witness concluded that he had observed a missile after hearing news accounts about the crash.
Clearly, some witnesses discussed the crash with each other and/ or learned about it from the media before they were interviewed. It is likely that media coverage about the crash and the associated criminal investigation may have led to bias on the part of some the interviewees.
Ambiguous clock-point and angle references. Sometimes direction or position is described using clock-point references that do not appear to be those generally used in aviation. It is not always readily apparent whether "o'clock" is referring to the observed object's path of travel, its position, or its elevation angle.
For example, the document pertaining to Witness 533 (CC-371, interviewed July 19, 1996) describes "the trajectory of the smoke trail initially as verticle [sic] (approximately 11: 00 direction)." Without knowing the orientation of the clock face in space, this type of description is difficult to interpret. The clock face could be parallel to the line of sight of witness such that 11 o'clock is almost directly above the witness, perpendicular to the line of sight of the witness such that 11 o'clock describes a trajectory that is nearly perpendicular to the horizon, or the clock could be parallel to the horizon such that 11 o'clock is just to the left of being directly in front of the witness (8.1 nautical miles slant range from flight 800).
In another example, the document pertaining to Witness 216 states that he "observed to his right, at about forty-five (45) degrees, a flare vertically going 'down'… [and] saw a horizontal explosion about one-half (1/ 2) way down from where he first observed what he believed was a 'boat flare'" (CC1-261, interviewed July 23, 1996). From this description, it cannot be determined with certainty where the witness first observed the flare. It appears that the witness (10.3 nautical miles slant range from flight 800) said that he first observed the flare at 45 degrees above the horizon; however, the description could also be referring to the direction in which the witness first observed the flare. The flare may have been first observed 45 degrees to his right. Because the initial position is ambiguous, the description of an explosion half way down from that point is also ambiguous.