There are many factors to consider in a "nontraditional" landing and it is probably unrealistic to state that any single approach is optimum for all situations and all aircraft. Obviously, if one can maintain the integrity of the fuel system and still keep the accelerations within reasonable limits and the cockpit/cabin volume intact then this is the logical approach to take. On the other hand, if one has to sacrifice one or more of these three areas, then I would argue that the fuel system should be the first to go.

 

Yes, fire is a scary deal, but a fuel leak doesn't always find an ignition source. The odds here are better than the lottery that most of you probably play from time to time. If the acceleration of the pilot exceeds the critical values for the various organs, the probability of a fire becomes irrelevant anyway. Same for a situation where the cabin structure is compromised and the occupants are crushed.

 

Back in the late 70's or early 80's, I read a magazine article regarding the best way to "crash land". The authors had studied the issue and found that the most common cause of fatality in a controlled gear up landing in a field (i.e.: not affected by trees, telephone poles resulted from the aircraft bending nose down and drawing the cabin roof down into the cabin, crushing the occupants.

 

They conducted swinging crash tests and found that the safest approach was to intentionally dip a wing just at touch down. The deformation of the wing absorbed much of the energy, keeping the occupant acceleration low and greatly reducing the likely hood of cabin deformation. Contrary to popular belief, they found the this did not result in cart wheeling.

 

I don't recall whether they considered fire in their analysis or not. I would also note that composite GA aircraft were virtually unheard of at that time and may well have much different characteristics than the sheetmetal cans of the day.

 

Some food for thought anyway.

 

Rob