Part 1 of 5:
I see that there have been several responses to my post regarding the tragic Air Florida accident in Washington, DC in January of 1982. In order to do any justice at all to those who were kind enough to offer comments, and the length such responses would require, I have elected to split my response into 5 parts; and I will try to load those 5 parts as close to one another as the system will allow.
I cannot fault many of the conclusions voiced on this thread, because much of what has been posted is an understanding taken directly from the National Transportation Safety Board (NTSB) official accident report. The point is that this report is flawed. Sorry … but that is a true statement. The NTSB doesn’t often make the mistakes that are present in this report – and I’m not at all impugning the very fine reputation of the many fine men and women that work at the NTSB. But, that does not change the fact that some of the conclusions reached in this report do not “square” with other information contained in the same report. In some cases conclusions were reached with no substantiation of facts … in other cases conclusions were reached in direct opposition to known facts to the contrary. As is true in most accident reviews, we have learned a lot about a lot of things. Were the accident scenario presented to a flight crew today, knowing what we know today, and how we’ve changed our training and training requirements today, there is an excellent probability that the accident would not occur today. But 28 years ago is not today. We did not know then, what we know now. Again, my responses here are not designed to be accusatory or sarcastic … but I am passionate about this business and sometimes (shoot, probably
most of the time) that passion is easily misunderstood to be an “in-your-face” response. If any of my responses sound that way, let me apologize in advance – that is
not my intent.
Originally Posted by Airbubba
Well, they only had about 1.70 actual EPR set instead of 2.04, there was plenty of power left in those JT8D's. Remember, you can MEL the EPR gauges and go with N1 for power setting on this engine. As the NTSB comments on page 81 of its report:
A couple of points: It is interesting to recognize that: “hindsight” usually provides a much better understanding than “foresight;” humans are prone to develop habit patterns; habit patterns can be an aid in accomplishing routing functions … and they can be a silent enemy, waiting to pounce.
In regards to the habit patterns … anyone who flew jet airplanes 28 years ago would immediately recognize that the primary power indicating instrument used by flight crews was the EPR gauge (and that is likely still true today). However, BECAUSE of this particular accident, it has been
drilled into the heads of today’s flight crews that the remaining engine instruments
must be observed – not merely to recognize that they are “symmetrically” aligned – but to recognize
what those other instruments are indicating. The routine training issues 28 years ago included monitoring the engine instruments – and they primarily were looking for fluctuations on an individual gauge or obvious non-alignment with the adjacently located instrument for (the/an) other engine. Also typical was to have the EPR gauge located at the top of the “instrument stack” – i.e., from top to bottom they were usually arranged something like EPR, N1 tachometer, EGT, N2 tachometer, and Fuel Flow. The point is that the top of the stack was
almost always the EPR indications. However, as many of you recognize, this was not
always this way. In fact, the B-737 that crashed had an instrument stack that was arranged as follows: N1 tachometer at the top, followed by EPR, EGT, N2 tachometer, and then Fuel Flow.
Following “routine, habit generated” engine instrument checks, the flight crew on the accident airplane would likely have looked first at the EPR indication – as it was deemed to be the primary indication for that purpose, and found that it was indicating 2.04 – the computed EPR for takeoff. And then, following the habit pattern, their gaze likely moved
down the stack from that point, glancing at the EGT, N2, and Fuel Flow. Again, following the habit pattern, not noticing any indication being asymmetrical (i.e., one EGT indicator wasn’t at 9 o’clock and the other EGT indicator at 11 o’clock), and not having any indicator fluctuating, they likely concluded that the engine power was, indeed, “set for takeoff.”
As I said above, hindsight is valuable for accident review. The actual thrust being produced by each engine was approximately 75% of maximum – calculated by a sound spectral analysis of the engine sounds recorded on the CVR. Had the crew been trained differently and had they been admonished during that training and their subsequent operations to identify the individual values being indicated by the other engine instruments on every takeoff (as pilots do today) – they likely would have recognized that there was a problem. However, this was not the traditional training regimen at that time – not just for this airline, but for the airline industry in general. Why? Because engines were very reliable and when they malfunctioned, that malfunction was rather easily identified. The reason we do things differently today is BECAUSE of this accident.
The reason that the engines were producing only about 75% power was because the PT2 probes were blocked – likely with ice, but of course that could not be confirmed – and the flight crew set engine power in accordance with the EPR gauges. An indication of 2.04 would have resulted in the engines producing 75% power.
Here are a couple of points that will be important later in this post:
1) If the PT2 probes were blocked with ice (and I think they likely were) had nothing to do with the fact that it was snowing at the time of the accident.
2) The fact that the engines were not producing 100% thrust had nothing to do with the accident – in fact, had the crew pushed forward on the throttles, ignoring any engine indicating instrument – even at the initiation of the takeoff roll – the accident would have occurred exactly the way it occurred.
3) The airplane was certificated to accelerate to V1, lose an engine (a complete failure) and then continue the takeoff with the one remaining engine operating at takeoff power. The accident airplane had 75% of the power that should have been available under normal circumstances. Clearly, that is 50%
more power than what the airplane had been certificated as requiring –
and, that power was symmetrical between the two engines, not asymmetrical, where all of the available power would have been coming from only one of the two engines.
4) Airplanes fly on airspeed … not on engine power settings.