Viewing page 14 of 49

This transcription has been completed. Contact us with corrections.

there are factors inherent in the design which account for this remarkable record. A complete examination into these factors would appear to be one of great potential industry benefit, however it is out of the scope of this paper. In passing, though, the suggestion is made that the slow spread of after-crash fire is one of the major survival reason.

In looking at the modern four engine planes, the uniformity of their accident rate suggests the influence of CAR 04 with its detailed performance and design requirements. It would appear that, although each plane has its individual characteristics, the overall level of safety is about the same, bounded as it is by the parameters of 04. This influence of regulation on the safety level will be covered later.

If we accept the principle of the marked effect of design on the safety record, and, by inference at least, its effect on "pilot error" accidents, what are some areas that warrant attention? Two suggestions will be made.

Over the thirteen year period of this study, there have been some 47 accidents due to misuse of the landing fear, either retracting it when on the ground or landing with the gear not down and locked. While these accidents are not fatal ones, they are expensive, both as to damage to equipment and as to loss of utilization. In one recent case the president of the airline has been quoted as calling it a "million

dollar mistake". Obviously, the pilots don't cause these accidents on purpose, and as there is a definite pattern established by these there must be a basis for confusion of controls within the cockpit. What can be done to make it impossible for these accidents to happen? Already we have various solenoid pin arrangements whose purpose is to prevent the inadvertent retraction of the gear on the ground, but by this time they are obviously proved to be not 100% effective. It is equally obvious that the landing gear warning horn does not deter people from landing with the gear neatly retracted. Howard (3) has proposed a system which involves "shoe strings and rubber bands", the "shoe strings" to tie component controls together so that it will be necessary to make two consecutive mistakes with different controls to cause an accident, and the "rubber bands" the bypass arrangement so that it will be possible to actuate controls out of the normal sequence if necessary, such as during a "go-around".

Basically, his system involves the interconnection of landing gear and flap levers so that it would be impossible to put down more than takeoff flap unless the gear were down and also make it impossible to raise the gear when more than takeoff flaps were extended. It would also be impossible to raise the flaps above the takeoff position