I can speak to the USAF slick C-130 community 1992-2012. We have 3-engine data & procedures and the 3-engine takeoff is performed every year in the simulator (including all the associated subsequent emergencies). That said, I don't have any first hand knowledge of any AF slicks actually performing this maneuver in the span of my career. Special Ops may be another story.
All of the E & H models have direct-geared drives such that the prop will pinwheel in the wind (and turn the engine also) if the blade angle is right. As a result, there are two other maneuvers practiced annually in the sim: the Windmill Taxi Start and the Buddy Start. In short, set the blades to the optimum angle and, for the Windmill, perform a high-speed taxi (100 knots). The motor spins and the automatic starting sequence fires ... the engine doesn't care whether the starter is spinning it or if it's the wind. For the Buddy Start, have another Herk park very close and run up its engines. If the blast and blade angles are correct, the prop spins and the engine starts.
I mention these to say that all of these maneuvers involve some risk. The AF, in my experience, is extremely risk averse and rarely, rarely, rarely, authorizes even the optional starts. The Windmill was part of my AC & IP upgrades - I did several in the aircraft. They were not overly difficult. The Schoolhouse IPs do them every class. Still, they are not done operationally and I was told that it had everything to do with risk. In combat, with the bad guys coming over the wire, that's a different story - my crew briefed actions we would take to save ourselves and the aircraft; this included looking at our projected weight to see if a 3-engine takeoff would be possible or a known show-stopper. We would worry about explaining ourselves later.
There are two real problems with 3-eng ops - I can only speak to Herk performance. The airplane has to be light and conditions favorable. If nothing else breaks, you're fine. Realize that you start the takeoff run essentially 2 engine because the rudder and nosewheel steering cannot counteract the asymmetric thrust until Vmca is reached. As a result, power on the operating asymetric engine is fed in as rudder authority permits ... there is potential to run this thing off the side of the runway if you're not careful. If another motor coughs on takeoff or climbout (if climb gradient is a factor), you may not be able to continue flight. (That's if it feathers normally - if it doesn't you may be on your way to an immediate emergency landing at a place not of your choosing even sooner.) To give an example of performance, at local training weights, the Herk can be capable of around 6,000 on 2 engines but 22,000 on 3 - that's an appreciable difference. Due to asymmetric drag, rudder deflection and the like, losing 2 engines is not necessarily like "losing 50% of thrust", the net loss is more - same goes for 3-engine.
I never ran the numbers to quote, but to cruise on 3 saves some fuel, but if you need the speed, the other 3 need to be set at higher power settings to offset the engine that's feathered, so I always suspected that was one of the reasons the 2-engine cruise was abandoned (in addition to safety). Also, each engine has a hydraulic pump. If it happens to be both left-side engines you lose, now the gear and flaps have to be cranked. In short, there are a number of pitfalls that can cause real problems - an experienced crew and a fair amount of planning for contingencies is key.
The second problem is the FAA. I can't speak directly to Ferry Permits, but I know that some sort of permit/approval is required to operate an N-numbered airplane in such a configuration. Some companies arrange to do it with regularity, others avoid it like the plague. Thankfully, the news isn't littered with reports of accidents from these ops.
I don't have any B-17 time, but my overall position is that there is probably no good reason to 3-engine ferry such a rare airplane - the affects of another engine quitting (or other unforseen problem) may be catastrophic. I guess I'll log one more tenth on the soapbox. I was once told that "responsibility" can be looked at as "ability to respond" or "response ability". If you can't guarantee your response to the loss of another engine or that 2nd engine that fails to feather, or quits before you've accelerated to 2-engine Vmca, then it's hard to accept responsibility for a 3-engine takeoff. Sometimes, the need for mission accomplishment outweighs the responsibility for a 100% safe outcome, that's why it's done, but as for the decision to do or not to do, $hit does happen.
Ken
Last edited by
Ken on Fri Jun 22, 2012 10:05 am, edited 4 times in total.