Dafydd Llewellyn

Don, the propeller did not come off the flange; the flange came off the crankshaft. Read the report.

Well, no, every step towards "fool proof" generates a new class of fool. In Cessna's case, they put the nosewheel on to create the "flying Chevrolet" ("one in every garage"). The C172 was a step backwards from the C170 in almost every regard except the ability to be used by pilots who never learned to use their feet. Re wing tanks, you're not quite correct; aircraft with wing tanks from which the engine can draw simultaneously, often drain one tank faster than the other. The engine will suck air from any wing tank, even with left/right selection, if it's selected to the low tank and flown with sufficient slip to overcome the dihedral. The Jabs do not have a lot* of dihedral, so the amount of slip required is not great. The normal tests for unusable fuel - see my post #334 - mean that aircraft that have long wing tanks and a "both" selector option or an "all on" system, will have enormous unusable fuel unless they have a collector which will keep the engine running for a 30 second slip, and be replenished in one minute of coordinated flight afterwards. My point is that unless people understand the basis for things like unusable fuel, etc, they will not understand how to operate the aircraft correctly. It's something else that needs to be taught, I guess. *The amount of dihedral an aircraft needs is controlled by its lateral/directional stability; one cannot arbitrarily increase the dihedral to assist fuel drainage.

Thanks, Turbs, that gets straight to the facts. Much appreciated; I just torqued the bolts on a Jab prop flange.

Yep. Suits me, I have a GFA maintenance authority; and gliders do not have the altitude constraints that apply to RAA-registered aircraft. I've no need to use the aircraft for commuting or to fly in controlled airspace. This is NOT a comment on what Darren Barnfield is trying to achieve; GFA never got that slack in the first place. If Darren gets a clear go to complete what he's doing, it will just about bring RAA up to that standard.

None of the above . . .

If they were tapered, the total fuel capacity would be reduced. If you make a totally foolproof aircraft, only a fool will fly it. If you comprehend that fuel will flow from the wing tank to the collector tank only if the wing tank is higher - which is not really all that enormous a concept, surely? Then it would follow that it's not a bad idea to regularly raise each wingtip in turn a little, when you are getting low on fuel. So the philosophy of the consumer society that it's always somebody else's fault and products should be such that no thought at all is needed to operate them, will ultimately result in products that are so limited in capability, in order to be foolproof, that nobody will want to use them. I've just been watching Attenborough explaining that the dominance of the human race is due to the use of intelligence. If you want to gain an advantage, it pays to use your intelligence. The penalty for stupidity, in nature, is death.

Looking at the POH from the Jabiru website, the J 170D has an "all on" fuel system, using two long wing tanks plus a collector tank. The POH gives the usable fuel as 135 litres, and shows the unusable fuel as 5 litres. The normal method of determining the unusable fuel is given in FAA AC 23.16A ; attached is an extract. These tests are normally used in my experience, regardless of the aircraft category. Therefore, it would follow that the normal expectation of a certificating authority is that uncoordinated flight will not normally be sustained for more than 30 seconds at a time.

Well, now that we know that, maybe I can get some data; I can't measure the fuel flow to each cylinder, of course (and I don't know of a fuel flow system that does that even for an injected engine) but I can measure the overall fuel flow quite accurately. Now, if I can just get sufficient mixture adjustment range on the carbie . . .

LAMEs who take that view are being extremely stupid in regard to their Hangar Keeper's insurance policy; do they REALLY want the liability of an aircraft that does not comply with a recognised design standard? A few have learned that lesson the hard way. I'm merely trying to point out some of the aspects that are standard practice in aircraft electrical system design. Use FAR 23 as a guide-post, not a hitching-post.

The real cause of the incident/accident; or failing that, sufficient confirmed information on which to base some analysis. I'm setting up to do some research on a Jabiru engine with some of Ian Bent's modifications - initially, to shake-down the test cell and run-in the motor (which is a "loaner") in the process; later, to run one of Ian Bent's fully-modified engines through the JAR 22H endurance test requirement. So I'm looking for things that will allow me to design useful research tests that can be done in this facility. I'm trying to get a real understanding of what to look for and how to look for it, in other words. The first bit of research is to get the engine baffling right, for the "downflow" style of cooling system that is conventional on horizontally-opposed air-cooled engines. The second bit is to see what I can discover about the mixture distribution - and depending on what that shows, what can be done to improve it. That comes in two parts - the effect of the engine itself; and the effect of the means of air supply to the carburettor. Obviously, the scope of this does not extend to the aircraft fuel system; but I'm interested to know whether this Brisbane accident was an engine problem or a fuel system or pilot fuel management problem. So far that's not unequivocally clear.

Well, that's not compliant with FAR 23.1361 § 23.1361 Master switch arrangement. (a) There must be a master switch arrangement to allow ready disconnection of each electric power source from power distribution systems, except as provided in paragraph (b) of this section. The point of disconnection must be adjacent to the sources controlled by the switch arrangement. If separate switches are incorporated into the master switch arrangement, a means must be provided for the switch arrangement to be operated by one hand with a single movement. (b) Load circuits may be connected so that they remain energized when the master switch is open, if the circuits are isolated, or physically shielded, to prevent their igniting flammable fluids or vapors that might be liberated by the leakage or rupture of any flammable fluid system; and (1) The circuits are required for continued operation of the engine; or (2) The circuits are protected by circuit protective devices with a rating of five amperes or less adjacent to the electric power source. (3) In addition, two or more circuits installed in accordance with the requirements of paragraph (b)(2) of this section must not be used to supply a load of more than five amperes. © The master switch or its controls must be so installed that the switch is easily discernible and accessible to a crewmember. And does your backup electrical power system also keep your engine EFI system going? And if so, how does THAT comply with a requirement such as FAR 23.1361?

The failure mode is anything that causes an uncommanded starter solenoid turn-on in flight; that could be a failure within the solenoid itself - e.g. if the spring dies; or it could be an electrical fault that puts power onto the solenoid control circuit. The modern form of starter that uses the solenoid to mechanically pull the starter pinion into mesh with the ring gear before making electrical contact may be less prone to this than types that throw the pinion into mesh via a Bendix drive, where the starter solenoid is not an integral part of the starter motor. The setup that failed on my PA 28 was the latter type; the engine had a direct-drive starter essentially from a 1950 Chev 6-cylinder car engine, and the cheapest-possible FoMoCo starter solenoid mounted on the firewall; and the starter solenoid end crimp let go, releasing the spring pressure so it turned ON under gravity. It was a really Mickey-mouse setup. I replaced it with a geared Prestolite alternator (which involved changing the ring gear) and a proper Cutler-Hammer contactor, and after that the engine started properly and the system gave no further trouble. In the engine test cell, the starter contactor is powered by a circuit that uses a latching relay that must be triggered via a press-button, before the starter button has any power available to it; and the latching relay coil is earthed via the alternator field winding - so it drops out when the alternator comes on. This makes it "impossible" to inadvertently energise the starter with the engine running by finger trouble - but it does not protect against mechanical failure of the starter contactor itself. However, I'm using the same type of Cutler-Hammer contactor as I used in the Cherokee, so I don't expect it will give trouble.

No, it doesn't, because it turns off EVERYTHING, whereas a battery isolator switch can be located in the starter circuit alone. That means the switch needs to be downstream of where the starter circuit splits off from the main bus. Used that way, it's ( strictly speaking ) incorrect to call it a battery isolator. I have not studied recreational aircraft electrical systems, but I suspect some of them fall a long way short in regard to the use of a proper master contactor, and in regard to circuit protection of the starter circuit. Standard automotive practice is not good enough.

Yes, it is indeed. And spot-on. However, I'm looking at this subject from a research point of view; I'm looking for a means of getting a realistic measurement of the mixture distribution between cylinders, and what affects it. What the article makes clear is that merely getting an EGT readout on every cylinder is NOT a valid measurement, due to the various factors that the article explains so lucidly. The absolute EGT value is practically meaningless; a whole generation of pilots use EGT scanners etc as an indication of the healthy operation of their engines, under the mis-apprehension of what these readings actually mean. I've been guilty of that myself, on occasion, and I'm very glad to have been reminded of the error of this. The value of instruments that monitor EGT - unless one uses them as a means of making sure the highest one is still at least 50 C on the rich side of peak EGT, which is not really possible without mixture control - is to allow one to spot when any one EGT changes significantly, out of pattern. I'd agree with you IF we can establish that the change is in fact due to the increased proportion of the fuel spray that is converted to fuel vapour (i.e. from liquid droplets to gas). I'm NOT talking about finer atomisation; I doubt a small change in induction air temperature can alter that. Do not confuse the two terms. Unfortunately, the design of the Jabiru airbox means that altering the source of the incoming air by selecting some hot air, also affects the flow pattern within the airbox and thus the amount of swirl going "down the plughole" to the carburettor. So using this as a means of modifying the EGT distribution cannot tell you what is happening to the GAMI spread; to do that you would have to do a GAMI spread measurement via a means of mixture control AT THE CARBURETTOR, firstly with cold air selected, and then with some hot air selected. What this says to me is that to do anything useful in the test cell by way of research into this, I need a means of adjusting the mixture strength at the carburettor. There are ways to do that in a research context on the ground, primarily by modifying the air pressure above the fuel in the float bowl, but I do not consider them very suitable for use in flight, unless the pressure sources they use are not affected by the throttle setting.

Mine was an earlyl PA 28-140 and it did not have a starter warning light - I'd assume Piper fitted one to later versions because of the problems with the early ones. Yes, a starter contactor downstream of the master contactor is desirable - but the damage is done long before the pilot will be aware of it, and you can't turn off the master switch before takeoff as a precaution, to prevent a starter run-away. So having a master contactor (which I consider should be there in all aircraft that have an electrical system) does NOT provide protection against a failed starter contactor. It give the pilot something to do after he smells the smoke.

No, I'm not clutching at any straws; I'm not trying to prove anything. I just want real evidence, not hearsay. Let's get to the facts.

To measure the GAMI spread you need to be able to plot the EGT versus fuel flow across the mixture range that gives peak EGT for all cylinders. I agree it would be possible to do that given an EMS that also measured fuel flow - IF the mixture control means has sufficient authority. The EGT spread at a constant fuel flow is NOT the GAMI spread - as the article for which the link was given in the first post explained.

Thank you very much for posting this thread. I'd agree fully with the article; I started out using an ALCOR gauge in exactly the manner described - lean to peak EGT and then back it off two to three divisions. It worked very well indeed in my PA 28. Trouble was, I had only one of them, whereas to see the GAMI spread, you need a gauge on each cylinder AND manual mixture control. The debate on another thread about EGT spread indicating mixture mal-distribution is a load of nonsense; it needs the ability to measure the GAMI spread. This comes very timely indeed, for the running about to be done in my engine test cell. I'll have to make provision to alter the mixture on the Bing CD carbie, so it can be used to get this information, as a research tool. However, provided the jetting and carbie-bowl venting is correctly set at the factory, a CD carbie like the Bing as fitted to Rotax 912 and Jabiru engines in RAA aircraft, does not need mixture control over the density altitude range allowed for RAA aircraft; arguably, it's better without it.

No, remember that the report on the Brisbane prang that was the original subject of this thread, has not yet been issued - unless I've missed something. This question of fuel-still-in-tank-due-to-flying-crabwise is still conjecture. Teckair mentioned another, but did not identify it or give a link to the report that reached that conclusion. I asked whether anybody had knowledge of a verification of this rumour, and this whole debate took off without anybody answering that question.

If you simply bung in a second solenoid in the manner you have described, you create what is called a "hidden function" - i.e. one of the solenoids could weld itself in the "ON" position, and you'd not know about this until the second one failed. One can go batty trying to figure the probability of a double failure; but can you think of a way to test them in the course of normal usage - for example, it is normal to use the electric fuel pump to start the engine - if you watch the fuel pressure gauge, it will tell you if that pump is working OK. Then taxi on the engine-driven pump alone, which will tell you that it's OK; then take off using both pumps. That defeats the "hidden function" issue. Like, have two starter buttons and check the starter does NOT pull in if you only press one of them. I'd prefer something that did not rely on the pilot doing something out of the ordinary; a thermal circuit breaker that had a fairly low rating but a long time-constant might be a better answer than the simple limiter; but if you think about it, a slow-blow fuse is essentially just that.

If you can believe the instrument, cause for checking that it's over-rich. If both confirmed OK, cause for champagne . . . However, I'd be very wary of an instrument that gave that kind of reading at cruise. Even more suspicious at full throttle.

Well, don't fly any aircraft with bag tanks, then; (and yes, quite a lot of Cessna have them, as well as Beech 33, 35, 55, 58, or a Lancair 4 - or, in fact, about half the GA fleet.) The only practical way with any of these is to start with full tanks. The reality is that fuel contents gauging is a significant problem in a great many aircraft.

Yes, as usual there are about 47 ways it could be done. It's about time people started thinking about it.

Actually, the wings are still carrying most of, if not all of the weight of the aircraft at the point of landing impact. So downward bending of the wings due to the landing impact is actually more a case of relief of the upward load; so if the wings are what you are concerned about in the student landings, I don't think you need to carry minimal fuel loads for that reason. I agree, carrying less fuel will make life easier for the undercarriage.

Erhm - I'd rather have a spring-trim device, in a small aircraft; could be as simple (and light) as a piece of exerciser cord ("Occy strap"). Tabs are a potential source of flutter; if you use an irreversible tab drive, it gets to be a complicated gadget and a maintenance pain; if you mass-balance the tab, this un-mass-balances the rudder, so one way and another you get added mass at the tail, which means about four times that mass up front to balance it - or the engine moved further forward, which increases the de-stabilising effect of the propeller - and around it goes. (Definition of an elephant: A mouse, designed to meet a military specification). That's NOT how Rod Stiff got the Jabiru airframe so light without compromising its strength; he did it by subtracting things, not by adding them.