Air Safety

Behaviours such as dumping fuel, inadvertently putting the landing gear down at high speeds, placing the aircraft in unrecoverable attitudes and attempting to land at 10,780 feet rather than 780 feet, the local field elevation, are but a few examples. However, so that these findings could not be attributed to chance, the experiments were repeated using different subjects with basically the same results.

A follow-on study then examined the effects of alcohol on errors of omission, eg, forgetting an important function or overlooking an important detail. For this study we used the same simulator but were interested in how accurately subjects could follow a pre-flight checklist. In the case of our simulator, 78 items comprised the checklist which each subject took on every flight. The task in this experiment was basically the same as that of any pilot: to begin with the first item and, sequentially, perform each function listed continuing systematically down the list until all items had been completed.

Prior to all flights, the experimenters, without the knowledge of the subjects, preset several errors:

• brakes OFF;

• landing gear handle UP; and

• altimeter mis-set by 1,000 feet.

Close adherence to the checklist should have caught all of the preset errors. Each subject ‘flew’ two test flights, one without alcohol and about 1 month later, one with alcohol (. 08 percent blood alcohol level). The results were as follows:

• Under the no alcohol condition, 11 % of all subjects failed to correct at least one major preset error.

• Under the influence of alcohol, 79% of all subjects failed to correct at least one major error. These pilots, it should be remembered, were considered legally sober and thus still capable of driving an automobile in most US States.

Most of our experiments were conducted in the late afternoon or early evening. We noticed that some of our subjects complained in class the. following morning that they still ‘felt lousy’ due to the previous day’s flight. A check of the literature showed that the amounts of alcohol ingested essentially should have metabolised easily within 10 hours after drinking. In addition, from a perusal of airline publications and military aviation journals,it was also noted that the general rule of thumb for social drinking recommended to pilots was to allow 8 to 12 hours between ‘bottle and throttle’. It was our thinking at the time that, although there was no longer any detectable alcohol in the blood, hangover effects might still produce a decrease in performance of complex tasks such as the psychomotor responses required in flying. Our most recent study followed this line of reasoning and we set out to test the so-called ‘hangover’ or residual effects of alcohol. It is presented in abbreviated form here.

The subjects were male and female undergraduates. The primary apparatus was a modified C-11 C Jet Flight Instrument Trainer, a completely instrumented static single engine simulator formerly used in the instrument training of military pilots. The subjects already were familiar with pre-flight, in-flight and postflight checklist responses and the basics of flying prior to the experiment proper. A 6 by 9 inch card containing a detailed checklist was given to each subject and all subjects were told to adhere to it religiously. This was followed by actual ‘hands on’ simulated flight consisting of takeoff,. climbout, levelling at altitude and full stop landings. When the subjects reached the point where they could successfully handle these fully checklist ‘flights’ at least three times consecutively without the slightest error, the experiment proper was begun.

After this pre-training but before the first test flight (non-alcoholic), the following errors were preset by the experimenters:

• brakes were placed in the OFF position;

• landing gear handle was put in the UP position;

• fuel select switch was placed on Auxiliary Tanks;

• dive flap switch was put in the DOWN or EXTENDED position;

• wing flaps were set at 50 percent; and

• The altimeter was adjusted to 1,000 feet above local ground level.

At the pre-flight briefing each subject was told to takeoff, climb to 6,000 feet and maintain that for 5 minutes. When requested to do so, the subject was to prepare for a landing and complete the landing at his discretion. The main data for the non-alcoholic flights consisted of the number of preset errors each subject failed to correct prior to takeoff.

The only essential difference between the above described flights and the second test flights (alcohol), was that 30 minutes before the alcohol flights, each subject was given enough 80 proof vodka to attain a blood alcohol level of . 10 percent (the minimum legally intoxicating level in the State of Ohio and many other US States but double the ‘safe’ limit in most Australian States) and mixed with an equal amount of ginger ale. Finally, a third test flight was given 14 hours after the alcohol flight.

The results were interesting to say the least. During the first test flight (non-alcoholic), 10% of all subjects overlooked at least one of the major preset errors. For the alcohol flight, 89% of all subjects made at least one oversight error. Fourteen hours after alcohol intake, 68% of all subjects still overlooked at least one preset error! It is apparent that performance 14 hours after alcohol intake was much more like that 30 minutes after intake than that of the first non-alcohol flight.

This evidence suggests that although most, if not all, of the alcohol had been processed through the body during the 14 hours following intake, the residual effects were contributing to this performance deterioration. I suggest all professional pilots be made aware of this residual effect and that they rethink and possibly recalculate the margin of safety implicit in the old rule of thumb, 8 to 12 hours between ‘bottle and throttle’.