lelievre12

Thanks Marc. As ever, super informative. I missed the Columbia talk but I think the answer to my question is here: "Can measure MAIN WING AOA to prevent deep stall, but would need to calibrate AOA indicator BY STALLING MAIN WING (that’s not going to happen in any reasonable world and deep stalls can easily be prevented just by keeping CG in the approved range)" However I don't think you need to stall the main wing to calibrate the AOA. The calibration if the AOA could be done in the Phase 1 testing like Chris Esselstyn where he started with FWD CG then slowly worked his way back until the aircraft until started to lose stability and stall recovery was less and less prompt. That setpoint would be what is needed. ie. set the AOA to the comfortable flight limit, no need for the actual full stall. Why would you do this? Well I don't entirely agree with your statement "stalls can easily be prevented just by keeping CG in the approved range". My SQ2000 has a questionable "approved range" which cannot be relied upon. Further, Chris's testing suggested that power on stalls were more likely to provoke a deep stall at an aft CG. Or put another way, accelerated stalls are harder and more dangerous to explore in Phase 1 testing than simple power off stalls. Therefore the AOA indicator can help provide additional input in these more difficult tests pilot so the power off 'comfort' baseline is not crossed. I fly with an AOA in my spam can and have learned much in my circuit work on how close or far I am from the stall in accelerated turns and various weights. The real merit of AOA is knowing what is happening in accelerated dynamic flight, not static straight and level. For example, on a gusty day, its quite alarming what your wing is doing unless you add some speed on short final. And of course the critical base--->final turn is very interesting to watch on AOA if you are cranking her over tight because you turned too late. In a canard at very flight condition you will expect to get a nose drop every time, however knowing the main wing AOA in each of these conditions won't hurt in such a beast as an SQ.

The end result is that the Austra 300 is 410# but the CD170 is around 115# lighter at 295# https://www.austroengine.at/uploads/pdf/mod_products1/AE300FactSheet.pdf http://www.continentaldiesel.com/typo3/fileadmin/_centurion/pdf/Datenblaetter/DS_CMG_2_CD-155.pdf

Years later we can see that the Diamond approach was to use stock Mercedes engines which included the heavier cast iron engine block. Thielerts approach was to replace a lot of the stock engine parts with lighter custom aluminium parts. Even the flywheel was custom and much lighter than the stock engine.

I guess there may be a variable between fuselage attitude (relative to horizon) and AOA unless the plane is carefully flown straight and level? eg. if at stall the plane starts losing altitude then the descent glide angle will need to be added to the fuselage attitude to arrive at the AOA. Perhaps the second Cozy 540 tests were actually not straight and level. If that is the case then I guess the Garmin AOA detection will still work for stall alerting as it detects actual air AOA and ignores attitude. BTW, Chris reports 50KIAS stall speed with rear CG. Will a Cozy really fly that slow? I guess if it is light enough, yes.

A while back we discussed the use of newer 'performance' envelope limiting autopilots now available. Amongst other things one can limit the allowable AOA before the autopilot will nudge the stick forward. For example to keep you away from a deep stall if you are at aft CG. However I was just reading the CG testing of Chris Esselstyn's Cozy 540 RG where he reports; CG of 100.24 – Power off stall, 53 kts, and 16 degrees fuselage attitude CG of 100.57 – Power off stall, 50 kts, and 13 degrees fuselage attitude The above data seems counter intuitive to me where an aft CG sees the Cozy stall at a lower fuselage angle/AOA than a forward CG would. Can someone help me get my head around what is happening here? I would have thought the aft CG would stall at a higher AOA as the canard had more authority.

The SQ2000 factory fuselage was not an advanced mould. It was simply a 'tube' in two halves with no door cutouts, windows, strakes or any other detailing. All this had to be added manually which makes the benefit of 'factory' moulds pretty small. It would be easier and better to CAD up the design, add the window/door/strake detailing and then CAM cut the mould in foam. You can see more at SQ2000.us See these photos: finalresult-flattened.mov

Wow those IGUS bearings look really nice. Ordered two.

Here is a shot of the Lancair IVP trailing edge with that flat cutoff. My recollection was that it was 1/4" to 3/8" On reading some IP (Wainfan Patent) https://patents.google.com/patent/US4867396A/en?q=Wainfan+micro+flap&oq=Wainfan+micro+flap it seems as though the more aggressive cutoff on the trailing edge is more about transonic stability than low speed drag reduction. Something I guess a Lancair might worry about but less so for us!

Kent, thanks for the reply. I will check the wing next time I am at the hangar for reflex on the underside. The N416 wing was foam core (like a Cozy/Eze etc) however my kit came with the alternative molded wing (Lancair style with strakes and no core) and the trailing edge is pretty fat. I think what I will do is bond a 0.5mm strip of carbon on the top edge then flox under to fill the edge. (whilst the wing is upside down). The idea is that the carbon will keep the edge resistant to nicks and hangar rash despite the thinner edge. I am expecting that a feathered edge will be less draggy but read that "Harry Riblett said something about cutting off the trailing edge at 99% of chord length is better than a point for the airflow." so this is confusing me a little. I know that the Lancair IVP I flew certainly had a squared trailing edge that was around 1/4". However my hunch is that this trailing edge is there to reduce high speed aileron actuation forces rather than to reduce drag.

I am finishing my SQ2000 project which has molded (not foam core) wings. The 'as molded' trailing edge on my main wing is a pretty fat radiused edge which is ~1/4" radius (1/2" top to bottom). Looks draggy indeed. Looking at the Eppler 1230 design I see that the trailing edge is 'sharp' so the question is "what trailing edge are other folks using" on their Eppler main wings? I am inclined to fair the rounded edge to make it sharper perhaps down to around a flat aft face/edge of around 1/8". I could go even sharper perhaps by using a thin carbon undersheet with micro to fair the top edge. I would have thought the sharper the better but not sure. I have seen a more blunt edge say squared to 3/8" like a Lancair? Any advice appreciated.

Marc, thanks for the reply. I hadn't read of Whitcomb but have now. Its interesting to note that the 'Typical' winglet sections have outward incidence "i.deg". In Whitcomb's work, this is quite pronounced (7 degree) at the root of the winglet and still 4 degree at the top. I haven't seen this measure of either twist or incidence in any canard wings to date, unless I missed something? Its also interesting to note that the winglet is canted 15 degrees, which again you don't see in canards. Finally, I see that the lower winglet incidence is set to 11 degrees and 36 degrees cant. This seems a lot but makes sense when you think about trying to 'untwist' the vortex before it happens. Your text about full chord lower 'slab' winglets is exactly as I thought. I did read Nat Puffers explanation of them adding "at least 1/2" of aft CG" and I was planning to add them. It seems another example of how the early Velocity SE and SQ2000 designs didn't make the leap from 2 seats to 4 seats very well (like Nat did). However, despite what Nat did, I'm wondering to what extent the Whitcomb 'mini' lower winglets are as effective as full chord. I haven't yet found any data or flight reports but will keep looking. I wonder if Nat tested them? It would seem that an 11 degree incidence lower 'Whitcomb' winglet is attempting to be as effective as full slab (and probably will have the same drag!).

I see these lower 'mini' winglets on older Vari/Longeze. I'm sure there is a long history of discussion about them. Can someone point me where I can read more? At the moment I have only upper winglet/rudders (Berkut/SQ2000/ERacer style) and wondering about the tradeoffs the between nothing, 'mini' winglets and full size lower winglets Cozy/Velocity style.

I think the big issue with retracts is the additional weight. Weight kills speed so aerodynamic gains can be diluted by increased drag from a heaver wing loading. I have to agree with others who prefer fixed gear. However if retracts can be made lighter then the advantage creeps back. I already have Infinity Retracts on my project and just the stock hydraulic pump and solenoids weigh >10 lb. Not including wiring etc That is sooo heavy. . Looking at the history from Lancair etc these old pumps are straight out of your outboard tilt lift from the 60's and migrated to aircraft at some point. Not exactly ideal. Since I have the retracts I am keeping them, however I plan on trying a state of the art BLDC hydraulic pump (brushless DC) which weighs 0.9 lb in an attempt to cut weight from the installation. Here is the current installation. Ill update when I have the new pump installed. The goal here obviously will be to have my cake and eat it too. I

Sheesh. Of course! So the cabin air ends up exiting via the spar. I should have noticed this when I was looking at my airframe. Thanks for your help Andrew!

This is the kind of thing I am thinking of. The rubber 'flapper valves' let air out when pressure builds inside the cabin. Not as big as this air pressure vent from an F150 but same principle.

The reason I ask is that the cabin doors are big and any pressure increase inside will lead to a huge burst load. Even 1/4psi over a 36” x 20” door = 180 lb pressing out. I fly a P210 with a 3.35psi pressure diff and the door locks are super heavy to cope. Given that quite a few canards have had doors fly open in flight I’m going to positively manage cabin pressure to ensure it doesn’t increase at all. Even a slight increase will cause to hiss and this is the door flexing outwards until pressure can escape. Noisy and far from ideal.

I am wondering what folks do regarding letting cabin air back out. I have arranged my inlets just fine but cant see any info in the build about how that air is getting out again! Pics/hints appreciated!

I have a set of them on my SQ2000 project at KVDO if you want to see them up close. Not selling but happy to send details/photos etc. James at Infinity is very responsive so simply email JD@infinityaerospace.com

For those interested in the inside of a Thieliert TAE125 (now Continental CD155) here are some photos of a gearbox removal and clutch check. The engine is out of my SQ2000 project described elsewhere. The TAE125 engine is based on a Mercedes OM640 engine however it is heavily modified with custom aluminium block, custom sump, flywheel and fuel injection. The gearbox exterior looks conventional for a PSRU with the exception of the hydraulic propeller controller which includes an oil pressure pump and ECU controller governor. Propeller flange is a Rotax style 80MM PCD suitable for constant speed MTU propeller. The gearbox is removed via the peripheral M8 hex capscrews and two M5 screws just below the propeller flange. The flywheel is then revealed which is custom Thielert (not Mercedes O640) The flywheel still contains a clutch plate which drives the gearbox. The design is arranged to absorb the power pulses from the high compression Diesel via slippage of the clutch plate. Many Thielert failures have been caused by the plate either slipping (oil or contaminant) or burning. The design differs from other PSRU (eg Rotax) which normally retain a flywheel and rubber power coupling which is used to transmit power. As the engine has no clutch pedal, the clutch plate is permanently retained by a twist locking ring. To remove, the clutch spring must be compressed by inserting 4 specially slotted M8 screws. Once the screws are in, the clutch plate can be rotated anticlockwise and released. The clutch plate and clutch can then be removed and inspected. Once the clutch is out, the flywheel can be seen. Its clear it is NOT standard Mercedes and is a custom lightweight flywheel made by Thielert. Later Continental CD155 models have a dual mass flywheel however these earlier engines have a conventional 'single mass' style. The clutch spring itself is a giant Belleville washer. Not individual springs. Its marked with it KN setting which is important in the correct 'slippage' for the clutch. The clutch plate is conventional Sachs although it is marked with a TAE part number so perhaps the friction material is custom. Looking at the gearbox side it is a standalone PSRU with conventional spline to mate to the clutch plate. From the above dissembly, its clear that Thielert have customised the design of the drivetrain extensively to reduce weight and account for the high torque of their Diesel motor. The PSRU design is unique in many ways and illustrates the difficulty of engineering a PSRU with a high torque engine.

Yee Ha! Discussed here: https://ww2aircraft.net/forum/threads/the-what-is-it-game.25244/page-461 Apparently the South African one was not 'genuine' Richter. It was a https://en.wikipedia.org/wiki/RMT_Bateleur "Actually, this is not a Richter Delta Dart II, but it is merely an empty hull and wing-structure of a plane called "Bateleur" which was built by a man known as A. v. Schoenbeck. It was a case of blatant plagiarism - the plans were obtained and illegaly used in the attempt to establish production in South Africa. In South Africa everything ended in a major case of presumably fraudulent bankruptcy. The single authentic Richter Delta Dart II remains in germany, its whereabouts kept secret." Video of the Bateleur here: Now that design is going to have some serious pitch over on throttle up. Marc Z?

Yup. Hard to argue with any of that. Especially when your example sounds like MCAS! Yikes!! ESP may make things worse not better in the early stages. Thanks for your input here. I guess it comes back to stick and rudder and exactly as you say, flying the airplane. Helmet fire or getting distracted is never an excuse for underspeed in a conventional aircraft with decalage. It therefore makes no sense to require a Phase 1 fix for a canard. Just introduces more complexity that may add to the fire. And resisting the impossible turn is part of the same skill set. Trying to automate this out wont work either. The test card and takeoff briefing simply needs to call out the allowable envelope and assuming its 15 bank, >75KIAS and >1000AGL before a turn to field then the Phase 1 pilot priority is that before all else. Having done so much short field work in my career I should simply trust that I actually can fly the numbers and don't need no machine to help!

The newer Garmin AP have 'ESP' which is a bit like the lane keeper steering in newer cars. It works a bit like a stick pusher when underspeed is detected but at all times the pilot is in 'control' and the AP is 'off'. This from G5 manual: "ESP engages when the aircraft exceeds one or more conditions (pitch, roll, airspeed) beyond the normal flight parameters. Enhanced stability for each condition provides a force to the appropriate control surface to return the aircraft to the normal flight envelope. This is perceived by the pilot as resistance to control movement in the undesired direction when the aircraft approaches a steep attitude or high airspeed. As the aircraft deviates further from the normal attitude and/or airspeed, the force increases (up to an established maximum) to encourage control movement in the direction necessary to return to the normal attitude and/or airspeed range" My thinking is that Garmin ESP would be useful in Phase 1 testing if the test pilot (me) gets a helmet fire whilst distracted by some other 'aspect' of the testing. Underspeed protection can limit any 'premature exploration' of the stall characteristics. As testing progresses and CG envelope found 'safe', ESP can gradually be wound back. Also note that ESP disengages under 200ft AGL so anything below that is up to the pilot. Of course if it all goes to hell and the Garmin goes beserk then there is always the CWS and APoff buttons on the stick.

Thanks Marc, I did read about N2992 and thought the report was invaluable. Not just for the SQ but for all folks Phase 1 testing canards. Making the impossible turn is hard enough let alone with an untested out of limits aft CG based on 'factory' numbers. And as said, there may have been a 'macho' attitude on having achieved a similar turn on a previous engine failure. The demise of N416 was also informative as a door failure, again is something that should never happen yet we see it again and again in the canard community. There are a number of similar Velocity accidents. However the deeper issues regarding specific to the SQ which are specific to this airframe require a detailed Phase 1 as you say. I have a short list of mitigation items in my test plan (which may be some way off) which include: 1. Cabin/seat/restraint strength all being reviewed 2. I am using a certified engine. 3. Testing at Castle (KMER) which is 11,800 feet long of wide concrete runway if that is permitted. Lots of room for taxi / ground effect work. 4. Will set the Garmin GFC to include stability protection below a set test number (80 KIAS initially?) 5. Will do a first principles reset of the CG envelope and start only with forward limits. I agree that the Glassic numbers seem optimistic and arbitrary. 6. Ensure vortilons and stability aids are actually fitted! I note one flying SQ also has aileron fences. etc etc. And of course before the flights I hope to meet you and maybe fly you up for a candid appraisal of progress. Of course others are welcome anytime at KVDO to share experience at anytime!

I ended up pulling the trigger and am the new owner of the SQ2000 built by Joel Conard at Payne Field. Details of his build are at http://www.sq2000.us As Joel works at Boeing, the quality of his work is excellent which overcomes (at least some) of the issues of taking on such a novel design. Also the Infinity retract is a really nice setup which also added to the appeal of this build. For the trip south to CA from WA we mounted the SQ on a flat bed 20' trailer canted up to keep the width roadable for the I5. Video of the 'rig' is here. A LOT of work remaining but I wanted to introduce the project and start to get to know those folks in the community so you can see another project coming along.

Actually I think I found the data online. http://www.canardaviation.com/cozy/chap23b.htm Discusses; 1. Plan View: prop flange square to centerline at BL O" (no P factor correction) and ; 2. Elevation: prop flange center at WL 21.5" (1.5" below top of longerons) with a 2 degree downthrust (aft higher than for'd). 3. Prop flange station = 41-42" aft of firewall. Does that sound right to others?