Dafydd Llewellyn

The purpose of this Thread is to bring to people's attention, and invite discussion on, some aspects of aircraft design that affect survival and injury in an emergency landing situation, in small GA and recreational aircraft. The history of awareness of this aspect of aircraft design starts with the introduction of seat belts, then the addition of upper body restraint, and has now progressed to dynamic testing of cockpit assemblies using instrumented dummies to measure forces in the lumbar spine, and to check for head impact. Formal design requirements for these matters are not uniformly applied across the full spectrum of aircraft classes; dynamic testing has not so far filtered down to the "watered-down" design standards for recreational aircraft; and of course no formal standards are applied to experimental aircraft for this. The rapid growth of General Aviation in the 1950s and 60s mainly involved aircraft of American origin, certificated to U.S. Civil Air Regulations Part 3 (CAR3). This standard demanded, in essence, a seat belt and associated anchorages capable of resisting a 9G retardation of a 170 lb (77 Kg) occupant. It did not address the ability of the cockpit structure to resist deformation. The CAR 3 requirement was carried over into FAR 23.561, when that superseded CAR 3 in about 1965; however CAR 3 aircraft continued to be manufactured under "grandfather" clauses, well into the 1980s. FAR 23 introduced a requirement for upper body restraint in 1988, and made it retroactive (FAR 23.2) to aircraft manufactured after December 12, 1986. (Australia had mandated upper body restraint for front seat occupants in the 1960s). Consideration of heavy objects above and behind the occupants was added - not sure when, possibly FAR 23 amd 44 or thereabouts. The FAA introduced dynamic seat testing in 1988, by the introduction of FAR 23.562; this requires a form of "barrier crash test" , but is mainly aimed at ensuring that the compressive force in the lumbar spine does not exceed 1500 pounds under the prescribed impact conditions, which for the principal case require an impact at 60 degrees from below the aircraft longitudinal axis. This requirement has not as yet been applied to the design standards for recreational aircraft.

LNC2 ? Sorry, it didn't click for me. Yes, I know about the Cirrus and Grumman singles - and their noselegs work, mostly, but I still have an aversion to that layout.

You're most welcome; however it's a legal obligation for a CAR 35 engineer to bring things like that to the attention of the relevant authority. Preventable, unnecessary death and injury is anathema to me.

If we're talking about the Jodel, I'd stick with wood; that avoids thermal and moisture - induced stresses due to mixed materials. Wood has an ability that metal and plastic materials lack, which is the ability to withstand a much higher short-term load, than a sustained load. That provides a substantail "hidden safety factor" . But how do you escape from the aircraft in the event of an overturn? If it's by breaking out through the cockpit side, then reinforcing the skin may not be the right answer. Canopy-type aircraft generally present two problems; firstly the weakness that is unavoidable when you have a hole in the wall of an essentially tubular structure, for people to sit in; and secondly, in the difficulty of emergency escape in the event of an overturn. You may note that Rod Stiff has repeatedly resisted all efforts to persuade him to produce a low-wing aircraft; these are some of the reasons. In the case of an aircraft like the RV, turning the cockpit side panel into a sandwich panel does not answer the principal issue, which is that the cross-section of the cockpit sill needs to be resistant to both buckling about its axis of least inertia, and to torsional buckling. That really calls for a closed box or tube for the sill member. The RV3 and 6 have a piece of extruded aluminium angle, and the mode of failure is that it first twists until the restraint of the skin no longer stabilizes it about its axis of least inertia, and then it fails by lateral crippling. As a result, its compression load capability is surprisingly low. I don't know to what extent that is a common design in similar aircraft - but putting a box section there instead of the angle is not the sort of thing of which most builders would understand the significance.

Quote: "When you crack the window of your car does the air go out or in? It goes in because it has been displaced by the solid form of the car just like a boat displaces water. If the 'faster air' outside, moving past your window was at a lower pressure as you suggest, the air would go out not in would it not?" With all due respect: Horse s***t. What you describe would happen if you were trying to float a car upside down in water - i.e. it's a hydrostatics answer. Not relevant to the issue in question. Bruce Tuncks has correctly described what happens in an unpressurised streamlined fuselage that has openings around the maximum cross-section and at the tail, such that flow can occur between the tail opening and the cabin. Whether flow goes into or out of such an opening (in the absence of anything that acts as a deflector or a scoop) depends on the difference between the static pressure on the outside of the fuselage and the static pressure in the cabin. The static pressure in the cabin depends on the aggregate effect of all the leakage points, but it will generally be somewhere between the free stream static pressure and the local pressure at the point of maximum local velocity; so it usually comes in at the tail and goes out at the cabin window or door leakage point - and this is well known as a source of carbon monoxide contamination of cockpits. You may be surprised to learn that open cockpit aircraft are notorious for CO contamination of the cockpit air; but if you think about what I have said above, the reason will be obvious.

It was probably some such devilish detail; however it shimmied violently and the GRP leg de-laminated and failed; fortunately it did not dig into the bitumen runway and cause an overturn. The flight was confined to a single close circuit, because the engine cooling system was obviously inadequate. All fairly normal teething troubles with a new prototype, but it caught John Buchanan at an awkward financial moment, so the project died. Most unfortunate, because the aircraft showed considerable promise, I thought. As a consequence of this experience, I rather dislike that forward-cantilever noseleg format because of its propensity to dig in in the event of a nosewheel failure

You may find this helpful http://www.tongji.edu.cn/~zyjin/AerodynamicsChapter5.pdf - especially when he gets to explaining the application of the Biot-Savart law to the "horseshoe" vortex system.

I assume you mean, "caster?" - yes, it had a typical free-castering nosewheel (I think is was a Scott tailwheel assembly, but not sure at this length of time) mounted of a forward-facing cantilever spring leg. It also had very small-chord elevators, which did not have sufficient power to hold the nosewheel off after the weight came onto the mains.

It had a nosewheel collapse (due to shimmy, which fractured the fibreglass) on landing at Toowoomba, which of course necessitated an engine bulk strip - a major issue for a Rotax 912 crankshaft - and the project was shelved.

I suggest you look at the Royal Aero Society data (now ESDU data) on the subject of the downwash field caused by the wing vortex system. Call it a flow vector if you will; I'm not going to get into an argument with you over semantics. I'm not aware that the Biot-Savart "law" has been repealed. I'm not talking about the disturbed air in the wing wake - though that can have significant effects on longitudinal stability. However the positioning of the horizontal tail is a more critical design consideration than is commonly recognised.

In fact, all the wind does (in general) is make the ground move in a funny way.

Thanks - (I'm quite familiar with big thumbs!) - but it had me wondering a bit, because asking for a non-deforming cockpit does sound a bit optimistic, I suppose - and indeed it does present quite an engineering challenge. However, it has become a requirement for new GA aircraft, if somewhat indirectly, by virtue of FAR 23.562, which was introduced in amendment 36 of FAR 23, around 1988; it was originally a requirement for dynamic seat testing, but if you read the fine print of how that test is required to be done, it involves not only impact testing the seat, but the entire cockpit structure, and ensuring the pilot's head does not impact anything hard. It's now more appropriately titled "Emergency landing dynamic conditions". The Whitney Boomerang was required to pass this test, and it did. I think this topic merits a new thread.

Would the person who put the "optimistic" comment on my Post # 36, please be so kind as to enlarge on that comment?

Pretty much a case of the front of the fuselage stopping, and the rear fuselage and the pilot colliding with the instrument panel. The rear fuselage pivots up and over, until the turtle deck behind the cockpit, hits the windscreen arch. In the process, it lets the shoulder harness go slack, so the occupants' upper body naturally is allowed to fly forward, and the occupants' heads are likely to be caught between the rear fuselage at the extremety of its folding, and the instrument panel. The loss in shoulder strap tension allows the lap strap to slacken, so the lower body also is not restrained from colliding violently with the control stick. All this takes a split second - then the rear fuselage flops back down and looks deceptively normal - but one can easily lift it up to the fully-folded condition, because its strength has been completely destroyed. This overall pattern of impact failure is, I would consider, by no means unique to the RV 6; as MM has observed, most metal canopy-type fuselages can be expected to behave pretty much like this, unless they have strong upper longerons as cockpit sills, that have been designed to have a high buckling strength. The very tight weight limits for recreational aircraft are likely to cause designers to omit the additional material for that, unless the design requirements force their competitors to do so too. Given strong cockpit sills, it would be preferable to anchor the shoulder harnesses to strong structure as close as possible to the back of the cockpit, rather than far back in the tailcone. Composite aircraft of similar configuration, in a similar type of impact - from what I have seen of them - shatter rather than buckling. There is little to choose between them. Occupant survival depends on having a non-deforming cockpit structure; and you simply do not find that in canopy-style fuselages, in aircraft of this general class.

The condition for longitudinal stability means that the more forward "lifting" surface - the wing in a conventional layout, or the foreplane, in a canard - must always carry a greater upward load per unit of its planform area, than the more rear surface. It does not necessarily mean that the tailplane mustalways produce a downforce; tho with a highly-cambered wing, it generally work out that way, especially at forward CG, especially if the aircraft has conventional wing flaps. Use of negative flap in cruise reduces the tailplane loads and helps performance a bit for that reason. Some aircraft - for example, the Fokker F-27 - use an inverted airfoil section on their tailplane, because it predominantly works to generate a downforce. The mechanism whereby the tailplane develops either upward or downward force, is exactly the same as the mechanism by which a wing does it; and it will have its own system of vortices accordingly. The fact that a tailplane operates inside the downwash field of the wing, simple alters its zero-lift angle, and the rate of change of the wing downwash as the wing angle of attack changes, means the tailplane has a reduced effective lift-curve slope - which means it has to be larger than if it were not affected by the wing downwash. The presence of a separate elevator is simply a means of simultaneously altering the tailplane incidence as well as its camber.

Dead right, Bruce; telling a convenient half of the story - in any technical issue - is not satisfactory; the students who hear that crap become the next generation of instructors, and so the myth becomes gospel. We have had a whole series of those - one notable one was the spiel given out when tubeless tyres were first introduced: Run cooler, more blow-out resistant etc. I had all four of a set of Dunlop B5s - their first generation of tubeless, still cross-ply - blow bubbles on me, (and one massive blowout) on one trip - at night - from Wagga to Sydney. They were disastrous, back then. The real reason was that they could be put onto rims by machine, whereas tube type tyres could not; my father-in-law worked at GM Pagewood at the time, and took me through the production line; one car was coming off the line every forty seconds or so. It took about 20 men with rubber hammers to put five tube-type tyres on their rims and inflate them, in 40 seconds, and they were working flat-out. Tubeless tyres allowed that to be done by two men, working easily. It was wholly and solely a production economy reason - but the public never got told that. Similarly, we got a great spiel about airborne lead (!) when lead-free petrol came in; the real reason was to allow catalytic converters to deal with the oxides of nitrogen that cause brown smog - but we got the bit about brain damage to babies, not the real reason. That's only two instances; it's happening all the time. I detest it.

That's De' Ath, surely? http://www.surnamedb.com/Surname/De'Ath

I had a theory of mechanics lecturer named Dr. Axelrad . . .

No, I have no idea whether they did anything; I think they took the position - not entirely unjustifiably, I feel - that those accidents were unsurvivable anyway. But I note that later models seem to have rather greater tailplane span.

I hope people are not offended by discussion that may help others to avoid becoming accident statistics. If it leads to better understanding, some small benefit can be derived from events that we would all prefer had not happened. We should, I believe, strictly avoid speculation or anything that might prejudice an accident investigation or imply blame; but that surely does not mean that potentially relevant facts cannot be mentioned?

Yep, agree. If you don't have airspeed, it will stop flying. Doesn't matter what it is. Let's leave it at that - but there is a message in this for any canopy-type aircraft.

Well, both the aircraft we examined had indentation of the box-shaped recess that the control stick works in - the rear corner of it acts as an elevator travel stop, for want of anything explicitly for that purpose; it was obvious that the pilot had pulled extremely hard - which we took to indicate that he had run out of elevator to lift the nose. There may or may not be a difference between engine idling and propeller not turning in regard to elevator power in the flare, especially if the speed gets a bit low. I've not flown an RV, though I've been invited to; I do not fly other people's aeroplanes - especially experimental ones - for fun or for the pleasure of performing unpaid research; the flight testing that I do is a serious part of earning my living. So I have no first-hand knowledge of this; but the fact that the pilot had evidently pulled back with all his strength; and had then been thrown forward onto the stick, despite his safety harness, sufficiently hard to fracture his pelvis (in one case) and his hip (in the other) - and in the process to partially fracture the stick in forward bending - spoke very clearly to me. "Stretching the glide" in an RV 6 with a stopped engine seems definitely contra-indicated; putting it on the ground and plowing through the fence seems a better risk. It was obvious that the RV3 had hit the ground at about 15 degrees nose-down; and a front corner of the fuselage gouged a quantity of dirt out of the ground, after which the wreckage bounced and hit the ground again about fifty feet further on. The severity of the impact deceleration can be roughly estimated from the mass of dirt that was fired off by the aircraft as it bounced - and it should have been on the borderline of survivability, had the buckling of the fuselage not nullified the safety harness, and then crushed the pilot's skull. This sort of collapse has been extensively investigated by OSTIV for glider cockpits; and any canopy-type aircraft is very likely to have its principal fuselage weakness in the cockpit area. So yes, if you hit hard enough, it will kill you. But "hard enough" is evidently surprisingly low, in some cases. I repeat, this may have no relevance whatever to the Mudgee accident.

I think you are referring to the cockpit sill reinforcement kits that were being produced by Graham Moodie (I hope I have that right) after he saw parts of a report that my son wrote to the QLD branch of ATSB, after he and I examined a wreck of an RV3 and an RV6 (both fatal) at Toowoomba, with ATSB's permission, at least a decade ago. (We were looking for data relating to lower leg injury, in a quite different context, but the fuselage collapse mode was something we noticed in the process). I could not approve an EO under CAR 35 because the aircraft is experimental, and therefore does not comply with any normal design standard - so the wording of CAR 35(2) is impossible to meet; however I saw what Graham was doing, and it looked to me to be helpful. Both those aircraft folded as you describe due to outward buckling of the cockpit sill, which completely nullifies the effect of the safety harness; the turtle deck behind the cockpit had paint marking from the windscreen arch in both of them; the fuselage had acted as a giant nut-cracker in fact. I would be interested to know whether the Mudgee aircraft had this mod. installed. However the presence or otherwise of the reinforcement has no bearing on the cause of the accident.

And the XCOM can have two control heads - which is very useful in a tandem seat setup

I suspect you are confusing a Whitcomb Supercritical airfoil http://crgis.ndc.nasa.gov/crgis/images/5/59/Review_SC_Airfoils.pdf with a Sonic Rooftop airfoil http://aerostudents.com/files/aircraftDesignAndOperation/examAnswers20100122.pdf The latter is used near the wing root on swept-wing high-subsonic aircraft to prevent the concentration of the pressure distribution towards the trailing edge, which is a peculiarity of swept-back wings.