The witness account of Mrs O'Brien

The witness accounts and theories do not all agree about the flight path of the object which hit the Pentagon. There is a well known oddity, i.e. the account of a particular witness: Danielle O'Brien, who was an air traffic controller (ATC) at Washington's Dulles airport. She reported in the first days after the event that she had seen on her screen a plane which first flew over the restricted flight area of the center of D.C. (the White House, etc.), then performed a steep descending 270 turn in a way which made her think it was a jet fighter. "Airliners are not flown this way, it's unsafe," she said. Then this plane disappeared from her control screen (too low over the ground) and within a few moments she heard the information, "The Pentagon has been hit."

This witness account raised a lot of speculation. Mrs. O'Brien is a professional air traffic controller, trained to observe flight paths on her screen, who knows the difference between an airliner's and a jet fighter's turning radius. But, she has perhaps been fooled by some oddities:

The alleged trajectory

Some witness accounts allow us to rebuild some of the flight path over Virginia before the plane struck the Pentagon. It would be necessary to have more witnesses projecting more flight path estimations in order to plot precisely where this plane flew. However, it is possible with the little data available (the radar data has not been made public) to estimate some possible trajectories. Here is an hypothetical reconstruction:

These two trajectory respect more or less the account of the witnesses who said that plane flew up the river towards Rosslyn, then turned about 270 in a continuous and smooth way, without any sign of panic by the pilot. The plane had to be rolled onto its port side, with an increasing angle - and decreasing turn radius - near the end of this turn to allow the pilot to align his trajectory with the Columbia Pike, Washington Monument and the Capitol Building. Were these trajectories technically possible for a Boeing 757 ? Here we need some physics and math.

Let's write, in pure metric system, the horizontal central acceleration HCA (m/s2) of a plane in a turn, function of the radius of the turn r (m), the speed on the trajectory S (m/s) :

HCA = S2 / r

Let's consider that the plane was flying, during it's turn, at about 540 km/h (340 mph), i.e. 150 m/s.

First trajectory (solid line)

We measure on the map the minimum radius on the trajectory at 3.75 km (2.3 mi), we get :

HCA = S2 / r = 150 * 150 / 3750 = 6 m/s2

HCA (horizontal acceleration) combining with g = 9.81 m/s2 (vertical acceleration), gives :

LF (load factor) = square_root (HCA2 + g2) / g = 11.5 / 9.81 = ~1.2
ROL (roll angle of the plane) : inverse_tangent (HCA / g) = ~32°

The total distance flown by the plane during this turn is something like 18 km. Thus the time of the turn, asssuming that speed approximately constant, would be :

Tr = 18000 / 150 = 120 s = 2 mn.

The altitude loss during the turn is said to be 7000 ft. The average vertical speed during the turn would thus be :

VS = -7000 / 120 = -58.3 ft/s ~= -20 m/s

Second trajectory (dotted line)

We measure on the map the minimum radius on the trajectory at 2.25 km (1.4 mi), we get :

HCA = S2 / r = 150 * 150 / 2250 = 10 m/s2

HCA (horizontal acceleration) combining with g = 9.81 m/s2 (vertical acceleration), gives :

LF (load factor) = square_root (HCA2 + g2) / g = 14 / 9.81 = ~1.4
ROL (roll angle of the plane) : inverse_tangent (HCA / g) = ~45°

The total distance flown by the plane during this turn is something like 13.5 km. Thus the time of the turn, asssuming that speed approximately constant, would be :

Tr = 13500 / 150 = 90 s = 1.5 mn.

The altitude loss during the turn is said to be 7000 ft. The average vertical speed during the turn would thus be :

VS = -7000 / 90 = -78 ft/s ~= -26 m/s

The first path, which is coherent with the alleged distance / time trajectory of this plane, implies that it was rolled at the maximum of around 30 from horizontal, the relative acceleration (G factor) felt by the passengers being 1.2 (20% more than normal weight acceleration) could be described as "nearly comfortable." This type of turn can happen during normal flight conditions of passenger aircraft, though it is near the limits of what is usually practiced by airlines pilots with passengers onboard. The vertical speed (-20 m/s) is something normal for an airliner flying at 340 mph on a descending path. The pitch angle of the plane should be approximately inverse_sinus (VS / S) = -7.6 : a normal feeling at airport approach which has nothing frightening for a passenger and moreover shouldn't be remarked as "unsafe" by a flight controller.

The second path, which could also be coherent with the alleged distance / time trajectory of this plane, implies that it was rolled at the maximum of around 45 from horizontal, the relative acceleration (G factor) felt by the passengers being 1.4 (40% more than normal weight acceleration) could be described as "very uncomfortable." This type of turn can sometimes happen during extreme flight conditions of passenger aircraft, though it requires that the pilot warns the passengers, or even apologizes for being so tough. The vertical speed (-26 m/s) is something which happens on a descent path with such aircraft. The pitch angle of the plane should be approximately inverse_sinus (VS / S) = -10 : a rather unusual but however normal feeling at an airport approach, with the aerodynamic brakes and flaps out an airliner can have a higher pitch angle, making the passengers feel a steeper descent of the plane, which most of them will find stressing. This situation could thus have been remarked by Mrs O'Brien, supposedly a trained flight controller, watching the trajectory on her screen.


Added, april 2007

The BTSB has released some data about the path of flight 77 before it's crash. According to these data, the last turn wouln't be a left turn but a right turn. See a map here. I have done the same plot than for the "left turn", on the same map at the same scale. /p>

As a matter of fact, on such a map, the minimum radiuses of turn are a little larger than the ones plotted with the "left turn" : 3000 and 4000 m. Obviously, the demonstration stands : these turns were possible for a 757, the 4000 m radius turn being more suitable for a poorly trained pilot than the shorter 3000 m one, and of course this later more stressing for the passengers.


Discussion

Of course, those murdering all passengers in a crash-bombing dive wouldn't care at all about passenger discomfort : they were not supposed to complain later to the company. This engineering speculation is only intended to examine the possible trajectory of a B 757 plane regarding Mrs O'Brien's account.

It is possible that the plane did a turn and dive corresponding to the first trajectory, and in that case, Mrs. O'Brien would have exaggerated her claim that it acted like a military jet. It is possible however that the plane flew corresponding to the second trajectory (smaller turn radius and descent time), which would have been something really uncomfortable, not to say frightening for the passengers, and cause a professional ATC to believe that the plane was a military jet. But even in this case, those who interpret that by saying that "it was not possible for a 757 to fly that way," and conclude that it was some other type of aircraft, are unaware of flight dynamics of modern passenger aircraft: load-factor limits are much higher than the value calculated here for the "tough" trajectory.

A question arises: could Hani Hanjour, said to be an inexperienced pilot, have done this type of turn and descent. In my opinion the answer is "yes." I am myself a pilot, and I practiced many times flying over the "Oisans" mountain range, coming back to the airport through a rather high pass (7500 ft), then dropping ~6000 ft in one of two turns before entering the airport approach. I could do this with high roll angle (~45) and vertical speed -15 m/s, the pitch angle being closer to -20 than the -10 of the "tough" trajectory described above. I'm not an ace of aerobatics, just an amateur pilot. I tried this type of manoeuvre on a flight simulator, with a big airliner : no problem to fly at 45 angle and -60 ft/s vertical speed. It must be reminded also that the flight instructors of the hijackers were surprised that they didn't care learning take-off and land skills, but asked to learn to control the plane in short turns! Having learnt that specifically with their instructors, having trained themselves to do it on a flight simulator, they were able, in my opinion, to do it on 9/11. Of course this affirmation is not a proof that the plane which did this manoeuvre over Virginia on 9/11 was controlled by Hani Hanjour : it could have been remotely controlled by a professional pilot with even higher skills.

I know, writing this, that I'm contradicting many Americans, including some who claim to be professionnal pilots and are saying that no hijacker with a few hours on a flight simulator could have made such a turn without stalling and crashing. Some say that even most professional commercial pilots could not do this with a 757. I disagree with such a statement : without knowing precisely "the way" that the plane flew, i.e. the radius of the turn, it is impossible to affirm that a 757 couln't fly "that way" without stalling.

Conclusion

The trajectory of the aircraft over the west suburbs of Washington D.C. to hit the Pentagon is not correctly known. It is however possible to plot trajectories and imagine flight parameters which are coherent with a B757's and a normal pilot's flight abilities. Thus arguing that such a trajectory implies that the Pentagon was attacked by a missile is, in my opinion, a very weak statement.


You were in VA on 9/11 : Please help !

You can help to precise where exactly flew the plane which hit the Pentagon on 9/11. You will find, here below, three maps. The first one shows a large part of the Washington D.C. area, nearly up to Dulles Airport on the west. The "270" turn of the aircraft described by ATC O'Brien must have been seen from the ground by a large number of people located on some parts of this map on 9/11 around 9:30 am. Perhaps some other planes crossed this area at this approximate time, as two airports (Ronald Reagan and Dulles) are located in this zone. However, the plane which hit the Pentagon followed a very unusual path which must have been noticed by a lot of people living in this area. If you are one of these persons, you can help by simply sending an e-mail to me at : jean-pierre@earth-citizens.net.

You can pan the maps below : just click on the blue bars which surround the map. If you move the mouse across the map, then position the cursor above a precise point, you should see two letters for the map reference and a pair of coordinates (East / South), appearing as well in the status bar of your browser, as in a small pop-up zone near the mouse pointer. Using this feature, you should be able to send a report like the following one (this is NOT a real email) :

Hello
I was walking on a street at Broy Hill Forest, on 9/11, at approximately 9:30 am, at location va 898,273 on your map. I heard a plane coming from behind me and stopped to look at it.
Looking towards north-east, I could see the plane flying approximately over the river, above a zone that I estimate, on your map at location va 946,248.
The plane was rolled on it's pilot's left-hand (port) side. I remarked some details: the plane's shape looked like a XXXX or similar model, it's markings had XXXX & XXXX colors arranged in XXXX way, which made me think it was a plane belonging to the XXXX carrier. The landing gear was not extended.
I can estimate it's visual size - as I was taught to do it in the army :) - saying that if I hold my arm straight, the plane's dimension was about xxxx fingers. I observed the plane flying like XXXX until it went out of my sight.
At this time, I estimate it's position approximately over Jamestown park or XXXX, above location va 834,249 on your map.
Best regards
xxxx

All the personal accounts given by this way will be used to try to plot the best trajectory. Only the witness' location and the plane's path points will be publicly shown. Names and other identifying data of all witnesses reporting to me will NOT be disclosed, unless the concerned witness expressly allows for it. Here are the maps. The first two are to be credited to "mapquest.com" web site.


Map 1 : Virginia


Map 2 : Arlington


Map 3 : Near Pentagon