DooMaw - building a STOL

[Head in the clouds]

Another happy day yesterday.

 

It was time to fit the horizontal stabiliser. It's a very critical part of the assembly since it all has to be tentatively balanced, clamped and shimmed into position and then measured for accuracy. Naturally it's almost impossible to hang a tape-measure on it when it's so delicately held in place, the weight and pull of the tape would just pull it all out of place again every time you tried.

 

Gladly, due to my day job, I'm well equipped with laser measuring equipment so I was able to get it all set up and use a laser distance measurer from various points each side and near the front of the fuselage. They're wonderfully accurate (mine's calibrated at <1mm at 150m) and since they're non-contact you can easily measure something that's 'just balancing there', or something you can't reach with the end of a tape-measure.

 

After a couple of adjustments I had it in position and was able to tack the first support bracket, adjust it again to correct the pull from the weld, and tack the other side. As with all the welding on this project, about 95% of the time is spent achieving the setup and then the weld just takes moments, it's almost an anti-climax when the job actually gets done.

 

For this job I used a spare 3m length of 7/8" tubing clamped into the support bracket for the setup and once the bracket was tacked in place then I substituted the real centre section of the HS. Once I had the centre section in place on the rear bracket then I could position the front saddle clamps, mark and drill the plate below for mounting bolts. Those saddle clamps sit on packers which provide a primary trim adjustment which will be set during flight trials by substituting thicker or thinner packers. I've started with 8mm and can remove them altogether or add up to another 20mm or so. Changing those packers moves the front of the fixed HS up or down and once set is then left as is - ground adjustable trim, if you like - inflight trim will be provided with a conventional trim-tab on the trailing edge of the port elevator.

 

The pictures below tell the rest of the story. I didn't get to do the 'photo-shoot' on the lawn with the real gear legs and big wheels, as you can see from the last picture it was rather gloomy and intermittent rain all day yesterday. I've also added a pic of the welded joystick that I mentioned but forgot to photograph some while ago.

 

Today's project is to try and get the elevator controls connected. That involves making the pushrods and their end fittings which carry the rod-ends, cutting the torque-tube (which the joystick is connected to) to length and making the forward walking beam located under the baggage area. The WB connects the first and second of the elevator pushrods. To do that there's a fair bit of geometry to finalise to get the right amount of elevator travel so that it just touches the control stops at the full extent of the joystick motion. I have control stops at both the joystick and elevator control surface ends, so it may take a bit longer than just today to get it right.

 

 

Including the afternoon's work the day before yesterday, fitting the tailwheel hinges, the bottom rudder hinge and a temporary strut to hold the tailwheel assembly up until I make its suspension strut, that's another 13hrs for the log, a total of 937hrs so far.

 

 

 

 

 

[Oscar]

As always, exemplary work, HITC.

 

Anybody who underestimates how hard it is to get a complex welded structure NOT to be pulled out of whack by weld shrinkage, has never done any... I used to build space-frame sports cars (Clubman-class, like Lotus Super 7's on steroids) and if you didn't consider the weld-pull - and the tube attachment sequence - you could get really, really frustrated. Engine mounts are the same deal... and you have many dozens of such welds, and they all look perfect!. You have every right to have a small plaque made up quoting the Ozymandias dictum: 'Look upon MY works, ye mighty, and despair". Something discreet, of course, and placed unobtrusively in the cockpit. Those who don't understand instantly, won't ever.

 

And, control stops at BOTH the input and control surface end, are 'proper practice'. I have to put the rudder pedal stops in to our Jab, having changed the entire fin and rudder for a J120/UL one from the old and broken LSA55 one, and consulting my friendly ex-CAR 35 engineer, was fairly horrified to learn that the FAA requirement is for 200 lbs force at the pedal for EACH pedal - so if both in the front seats have an 'Oh shit' moment and hit full rudder together, there's 400 lb force to be resisted from being fed into the control line / rod at the pedal - and therefore multiplied by whatever mechanical advantage there is from the pedal to the line / rod..

 

 

[Kyle Communications]

I agree with him above

 

 

[Head in the clouds]

.......... And, control stops at BOTH the input and control surface end, are 'proper practice'. I have to put the rudder pedal stops in to our Jab, having changed the entire fin and rudder for a J120/UL one from the old and broken LSA55 one, and consulting my friendly ex-CAR 35 engineer, was fairly horrified to learn that the FAA requirement is for 200 lbs force at the pedal for EACH pedal - so if both in the front seats have an 'Oh ****' moment and hit full rudder together, there's 400 lb force to be resisted from being fed into the control line / rod at the pedal - and therefore multiplied by whatever mechanical advantage there is from the pedal to the line / rod..

Yes, I had designed around control stops on all the flying surfaces but then I was having a chat with Bill Whitney and the subject came up. He said that control surface stops were required when the control was operated by a cable but if it/they was operated by pushrods then control stops at the control lever were sufficient since it was considered to be a 'rigid' system.

 

Naturally that means that cable operated rudders must have stops at the pedals and also at the rudder itself but I guess my elevators probably wouldn't actually require the positive stops at the control surface even though I have designed and installed them in anyway.

 

Thinking back to my C172, I don't recall any stops at the control surfaces and they were cable operated, but the stops were probably at the outboard bellcranks, because from the bellcrank to the control surface was a pushrod.

 

And yes - the load that the rudder system might have to deal with could be very high. I recall when two of us in a Drifter were involved in a close call with a GA plane that came 'visiting' our strip with no prior request or notification and assumed anyone around would be on VHF - which we weren't. The first announcement of his arrival was a beat-up of the strip at 180kts when we were on short final. We saw each other at the last second approaching head-on, he pulled up hard and we flew straight into his wash which threw us nose near vertical at 50ft or so. We missed the ground, just, but afterwards found the rudder cables stretched - the eyelets had elongated considerably - both rudder horns were bent and the attachment of the cables to the rudder pedals was 'strained'. We both recalled how hard we had braced against the pedals. The Drifter pedal system was tested to the required loading during certification - mine was no.2 after certification - so I'd say the requirements are marginal, even at 200lb/pedal. Consequently DooMaw's rudder system is as strong as I can reasonably make it, with most particular attention to the rudder control horns.

 

I must have another read of FAR23 for a refresher ...

 

 

[Oscar]

I do know that Dafydd Llewellyn and Alan Kerr differ on the necessity for stops at both ends of the Jabiru push-pull cables! - and both are very, very experienced and respected engineers. As is, of course, Bill Whitney. Between those three, you have a majority of ultra-lights ever designed and produced in reasonable numbers in Australia covered... add in Grahame (?) Swadling and you have just about the lot, other than the guy who did the Brumbies, I think. Both Dafydd and Alan worked extensively on all aspects of development of the Jabs. in the early days (up to the 160, at least) so a simple end-user like me can't separate the advice...

 

I know Dafydd Llewellyn prefers accurate stops at both ends, I THINK Alan is content to have rudder pedal travel with positive stops for serious pedal-pushing but considers that fine adjustment can be taken out at the control surface end. I can see his point: all the components in the control circuit have a rated force in excess of 1200 lbs. However, since Dafydd will be doing the test flying of my aircraft, I'm not going to even think about trying to send him aloft ( and he probably wouldn't go, anyway..) unless he is happy with everything... and I don't think the rudder horn will happily accept anything like a serious force, though being a glass construction, it does have flexibility on its side.

 

Possibly interesting story: When I came to fit the new rudder to the new fin, I found that it was a bad match. At the bottom, it was a good fit into the fin 'channel' -but at the top, it was about 6mm too narrow, resulting in a 6mm or so step from the t/e of the fin to the rudder on the starboard side. Aerodynamically, horrible. Took me a while to suss out why: Jabiru had simply extended the rudder spar moulding from the LSA rudder spar vertically, without taking into account the divergence due to the increased height of the fin.

 

After consulting with Alan Kerr - who had done the structural justification work - I split the rudder spar and inserted a new web to bring the rudder into proper alignment.

 

 

Having to work 'inside-out' was a challenge... And, the resultant weight change and moment had to be calculated to ensure that the possibility of flutter as a result was within the parameters of the original flight testing!. Jigs had to be built....

 

I suspect that a lot of people don't realise just how much thought and effort needs to go into making changes. You can't just decide that 'a bit more of X would be good' and take a hunk of aluminium and a handful of rivets and throw them onto the aircraft, because a different aircraft has that feature and it seems to work.... adding VG's without really, truly knowing how to get them to work effectively and safely, is probably one of those areas where some people get great results and some either get poor results or even make their aircraft dangerous.

 

 

[djpacro]

...was fairly horrified to learn that the FAA requirement is for 200 lbs force at the pedal for EACH pedal - so if both in the front seats have an 'Oh ****' moment and hit full rudder together, there's 400 lb force .....

Nope, read 23.395 then 23.397 then 23.399.

[Oscar]

DJP - I am slightly confused.

 

23-397 says:

 

§23.397 Limit control forces and torques.

 

(a) In the control surface flight loading condition, the airloads on movable surfaces and the corresponding deflections need not exceed those that would result in flight from the application of any pilot force within the ranges specified in paragraph (b) of this section. In applying this criterion, the effects of control system boost and servo-mechanisms, and the effects of tabs must be considered. The automatic pilot effort must be used for design if it alone can produce higher control surface loads than the human pilot.

 

(b) The limit pilot forces and torques are as follows:

 

Control Maximum forces or torques for design weight, weight equal to or less than 5,000 pounds1 Minimum forces or torques2

 

Aileron:

 

Stick 67 lbs 40 lbs.

 

Wheel3 50 D in.-lbs4 40 D in.-lbs.4

 

Elevator:

 

Stick 167 lbs 100 lbs.

 

Wheel (symmetrical) 200 lbs 100 lbs.

 

Wheel (unsymmetrical)5

 

100 lbs.

 

Rudder 200 lbs 150 lbs.

 

29-399 says:

 

§23.399 Dual control system.

 

(a) Each dual control system must be designed to withstand the force of the pilots operating in opposition, using individual pilot forces not less than the greater of—

 

(1) 0.75 times those obtained under §23.395; or

 

(2) The minimum forces specified in §23.397(b).

 

(b) Each dual control system must be designed to withstand the force of the pilots applied together, in the same direction, using individual pilot forces not less than 0.75 times those obtained under §23.395.

 

So - what I think you are pointing out, is that for BOTH pilots stomping on the same pedal in a linked dual control system, the maximum force restraint has to be: 0.75 x 200 lbs x 2: 300 lbs applied to the linked pedals.

 

Now, in the J160, the stops are on the leading (from the pilot's perspective) torque tube that connects the rh pedals, and acts against a reinforced area on the trailing (from the pilot's perspective) connecting torque tube. The rudder push-pull cable is attached to the pilot-side rh pedal upright. But, since either side pilot can be the one pushing the pedal, each individual pedal has to be able to withstand 200 lbs, surely? If I have got this straight, it would be only in a cable-joined pedal system, that the 0.75 reduction of cumulated effort would apply? In the LSA55 style pedal system, each pedal is individually stopped, so I think the 200 lb limit at each pedal, applies - yes?.

 

 

[Oscar]

DJP - no longer confused, I worked it out, I had misinterpreted what I'd been told. ... HITC, sorry to hijack your thread there, but I guess it's the sort of thing you have to deal with every step along your way with the design, so it's not been completely irrelevant.

 

 

[djpacro]

Good Oscar, I was about to respond. Lots of ifs and buts in FAR 23 - my favourites are wrt Maneuvering Speed.

 

 

[Oscar]

I made the connection when I awoke early in the morning with a 'Eureka' moment, because it explained why Jabiru had gone from individual pedal stops to the torque-tube mounted ones: the stops only have to meet the 300 lb load case and by doing that, automatically meet the 200 lb load case... I'll bet HITC has had plenty of those 'look at the design, and see a better way' moments as he's gone along- because his design has so many elegant solutions in it. As an aside, the J160 and later rudder + nosewheel linkages mechanical design is SO much better than what's in my LSA55, it's really quite an excellent example of experience improving the breed ( though I dislike the rudder-centreing arrangement).

 

 

[Head in the clouds]

..... I'll bet HITC has had plenty of those 'look at the design, and see a better way' moments as he's gone along- .....

Yes, definitely. At the start of any design you tend to think you've got the majority of the design details sorted out in your head and just need to get them on paper, so to speak, but in reality along the way there can be so many changes as an earlier detail affects a later one, that if you're not careful the original concept can end up almost unrecognisable in the final outcome. Consequently staying focused on what you want to end up with is pretty important or the design is likely to morph into something different altogether.

 

The best thing I ever did in terms of the above was to learn 3D CAD modelling. During the process of developing the model you build every part of the structure, several times sometimes, so a long time before you actually cut metal you know exactly what you will end up with, and have the opportunity to change it as often as you need to, to make it suit the purpose and/or simplify the construction.

 

..............................

 

March 28th to 30th 2016 -

 

The next job was to get the elevator controls connected. This involved making three pushrods and connecting them from the joystick to the elevator horn via a walking beam* and a bellcrank*. I made and installed the bellcrank quite some while ago and deliberately hadn't made the walking beam yet because I intended to use it to make the final 'refinement' of the gearing of the joystick-to-elevator horn angular motion ratios. Probably resulting from my time in gliders and helicopters I like quite sensitive controls, so on DooMaw the joystick moves through an arc of 15 degrees forward and backwards with a stop-to-stop linear distance at the control handle of about 240mm - 300mm is probably more like the 'norm' on joystick controlled aircraft of this genre. The elevator control horn moves through 25 degrees each way with a linear motion of around 120mm so via the bellcrank and walking beam I needed to convert 240mm motion into 120mm. The bellcrank was already made and imparts a ratio of 9:10, I had to do that to allow the large elevator pushrod that runs through the aft fuselage to clear the fuselage diagonal bracing. That was achieved by having a slightly shorter bottom arm on the bellcrank which lifted that end of the pushrod slightly.

 

*Someone asked me what the difference is - as far as I understand it a bellcrank has the pivot between the pushrod/cable connections, and a walking beam has both the pushrod/cable connections to one side of the pivot.

 

Along the route of the elevator pushrods there is also a fair bit of angular change i.e. the pushrods don't all point in the same direction. From the joystick the first pushrod runs through the inside of the aileron torque-tube, which points downhill, before attaching to the walking beam. Then the large pushrod goes uphill from the walking beam to the lower attachment on the bellcrank. The third pushrod is nearer to level, connecting the top attachment of the bellcrank to the elevator horn. All these direction changes mean that Ackerman Principle has to be taken into account or we won't get consistent elevator movement throughout the range of motion of the joystick.

 

The reason for all these 'ups and downs' of the pushrods is that it moves the pushrods away from areas you want to keep clear i.e. they go under your elbows in the cabin instead of interfering, then under the baggage area, then the bellcrank changes the motion from push to pull, so that the elevators go up when you pull the stick back, otherwise they would operate in the wrong sense. The final pushrod is positioned where it is simply to get through the HS centre-section frame which allows the HS to fold up for compact hangarage or trailering. There will be a quick-to-remove fairing at that aft HS/VS intersection to allow easy inspection and servicing of all the moving bits down the back there.

 

First I made up the short pushrod connected to the elevator horn, I could take its length direct from the CAD model. Next the pushrod at the other end, which connects to the joystick, and I made the first half of the walking beam. The pushrod that connects to the joystick has an unusual feature in that it must not only move back and forth but must also twist to allow the motion for the operation of the ailerons. One end of the pushrod is connected to the joystick which rotates to move the ailerons, and the other end of that pushrod is connected to the walking beam which doesn't twist. Consequently if you want to use this control method some provision must be made to allow one end of the rod to be rotated by about 50 degrees while the other end doesn't rotate, and this must be done without turning the thread on the 1/4" Aurora ball-type rod-ends. I achieve this by putting very small diameter 'washers' (2mm long, cut from 5/16"/8mm tubing, one washer on the bolt each side of the rod-end). These washers allow the ball to rotate within the rod-end to about 15 degrees each way, so the combination of the rod-ends at both ends allows the pushrod to rotate up to 60 degrees without binding up. It's important to make sure both rod-ends are aligned with each other when the locknuts and locktite are added in the final assembly. I'd forgotten to allow sufficient space for those tiny washers between the rod-end cleats when I made up the joystick some while ago, so I had to carefully cut them off, remake and re-jig them when I found I couldn't twist the pushrod as intended ...

 

I turned up the end fittings on the lathe for the large pushrod, and riveted one end in, and connected it to the bellcrank. Then it was a case of clamping the second large pushrod end - which was not yet connected to the pushrod - to the walking beam and also clamped to the side of the, as yet, overlength pushrod, and adjusting its position slightly up and down the walking beam until the motion of the joystick allowed the elevator horn to just touch its stops at the same time as the joystick also just touched its stops. Moving the connection up the walking beam would make the elevator stops touch first and moving it down would make the joystick stops touch first.

 

Once I had found the spot where they both touched together I could finish the fabrication of the walking beam and cut the large pushrod to length and attach the second end to it. All of that adjustment to determine the walking beam second attachment point could have been done from the CAD model but with all the Ackerman Effect calculations that would be required - and my software doesn't do them for me - it was easier and quicker to do that part by 'trial and error'.

 

Some pictures -

 

 

That took another 23hrs, for a total of 960hrs so far.

 

 

 

 

 

 

 

[pylon500]

G'Day HITC,

 

Don't panic, but I have the following for you to consider;

 

Just looking at your elevator linkages and realised you've had to go the 'lessor' way of having a 'push for up' tube, instead of the preferred 'PULL for up' system.

 

I know various others have gone this way as well, Lightwing and Foxbat to name a couple, but they have made up for it in different ways...

 

Lightwing uses a fairly large diameter tube with the thinnest wall they could get, while Foxbat use two shorter tubes via an idler.

 

The problem stems from the possibility of flex in the push tube that could induce further travel under G load, or in an extreme case, failure of the tube in bending while under the compression load of movement when subject to G load induced by said movement.

 

While this may seem a bit of a stretch as a failure mode, consider;

 

A STOL approach with full flap, stick already half way back to compensate pitching moment of flaps (ie compression load on tube), an unexpected downdraft requiring a sudden application of full elevator, elevator hits it's movement stops, further application at the stick adds higher compression load to tube, a moderate to heavy touchdown (momentary 5~8G impact not impossible), elevator tube now 5~8 times heavier causing it to flex out of alignment, added to the existing compression load already applied, the tube buckles to the bottom of the fuselage thus changing it's original length.

 

This all took half a second.

 

Now the aircraft rebounds into the air, the flaps start a pitch-over movement, added to a quick fore/aft movement of the stick to try and correct the bounce but because the elevator tube is now a little shorter, the pitch-over is exaggerated, another more violent and stronger back stick is applied, but instead of getting up elevator, the tube fails and the aircraft is nose down, low and with no pitch control !!

 

It could be said that a similar problem could stem from having a 'pull-up' tube when G load is applied, where flex would shorten the tube thereby inducing more pull and hence higher G!

 

Granted, but the tube will not flex as much when under a tensile load as opposed to a compressive load.

 

Still, all you need to do to protect from the above is support the tube so that it can't flex out of alignment, either with a swinging cable, or as it looks with an 'under' pivot point at the stick end , and an 'over' pivot at the elevator end, there should be a neutral point somewhere along the tube where there is no vertical movement, and you could fit a roller or some sort of guide to maintain alignment.

 

I remember the Pilatus B4 used felt guides for the elevator push tubes where it passed through the various bulkheads.

 

Did tend to squeak when they got dry....

 

Sorry if I haven't made these comments earlier.

 

 

[Head in the clouds]

G'Day HITC,Don't panic, but I have the following for you to consider; ......

Hi pylon,

 

Yes, I agree with everything you've said. Much appreciate the watchful eye again, thanks!

 

I didn't have any option about using the pushrod in 'push' mode for up elevator, that is a function of the underslung joystick hinge-point, and I wanted that because it keeps the rotation point of the torque-tube where it's needed because I wanted to keep the stick's motion above the height of the knees and avoid the stick pushing peoples' legs around at the extremes of stick travel. To have the joystick hinge-point above the torque-tube would either make the mechanical advantage/lever length of the stick too short, or alternatively position the torque-tube too low ... everything's a compromise.

 

I calculated the buckling strength of the main pushrod and it exceeds requirement, at 6G, by a factor of just under four without any further support - the aerodynamic balancers keep the stick/system loads fairly low. Interestingly, increasing the wall thickness of the pushrod tube doesn't increase its buckling strength significantly since the midspan loading is increased when G is factored in, as you pointed out. Consequently I settled on the 1.25"x0.035" 6061T6 tube which is very light but quite substantial in diameter - in any case as large as I can accommodate between the fuselage diagonals. As far as I can see it looks to be the same as they use on the Foxbat, so I was glad to note that that gave some credence to my material selection.

 

However, what you can't see at this stage is that I did also decide on a 'belt-and-braces' approach and the tube will soon have a mid-span support to prevent flexing under both positive and negative G. It will be a strip of double necked banjo-shaped plywood which will straddle the pushrod at its noble point where, as you pointed out, the tube will not display any vertical motion while moving back and forth. That support will not quite touch the pushrod except under extreme G loading or in the event of any buckling initiating, so it will not wear the tube in normal use. I'll probably apply some mylar adhesive tape where it might touch on those rare occasions. The non-touching support should act like a jury strut does for the wing struts, but which only acts when its needed, and does nothing at other times.

 

Thanks again for your vigilance!

 

 

[Oscar]

That's a very reasonable observation, though it's also obvious that HITC has well considered the bending potential for the long tube and increased its O.D. considerably. But, of course, as the tube bends, the alignment of the rod ends falls increasingly outside the centre-line of the tube, increasing the bending moment... and also stressing the rod-end threads (though, it looks to me as if they are 1/4" so should be well strong enough).

 

I've been rejuvenating the complete rudder circuit on my Jab ( early LSA55), which uses 3/16 male rod ends, and I've been less than impressed by the number that have bent threads. (They'll all be replaced.) They are obviously strong enough for the job in radial loading, but (and I think this has been due to long-term bloody poor maintenance practices), since just about all of them have some axial deflection in their travel and haven't been properly set-up, they've suffered from inadequate stand-off height and have had the threads bent observably by eye (and more observably, when put in a lathe chuck and spun). Some of the standard AN 3/16 washers have ended up bent like belleville washers, with strong witness marking on the rod-end housing attesting to the contact. (also most did not have any form of retaining washer on the outside of cantilevered bolts - not a problem with HITC's application, for sure!) I've used a pair of unbent ones to determine the required stand-off height - and as HITC said in his post above - the best way is actual assembly and observation, and the patience to go back to the lathe etc. and make the fine-tolerance bits to fit the case.

 

Getting it RIGHT, takes a whole lot more time and effort than just getting it to FIT. It's actually satisfying work, for a perfection-junkie (which is my way of avoiding the term 'anally-retentive'), but I am sure any CAR35/Part 21M engineer will tell you that a goodly part of their work has come from correcting stuff that has just been 'made to FIT', instead of getting it right in the first place..

 

Edit: just saw that HITC is ahead of us both, and has the situation in hand. HITC - I have a considerable length of 80-duro poly bush rod, of 50mm diameter, I got to make up the one missing compression do-nut on my Jab front leg, and would be proud to contribute a bit for you to make up a grommet for your retainer, if you'd like it. P.M. me with the postal address...

 

 

[Head in the clouds]

This post should bring the log back up to date ... sometimes it seems to take nearly as long to write the log as is spent building the plane!

 

March 31st to April 3rd 2016 -

 

Still working on control systems, this time it's the rudder and tailwheel.

 

On many taildraggers the tailwheel is either steered by cables (and springs) linked to the rudder control horns and on some other draggers the rudder cables have a splice forward of the rudder connection which allows attachment of a supplementary cable going to the control horns on the tailwheel. Since my tailwheel horns are almost directly below the rudder horns the former method wouldn't be satisfactory, and particularly because as the tailwheel suspension moves, the distance between the rudder and tailwheel horns changes quite substantially.

 

As far as I know the Just Aircraft SuperSTOL elected to get around this problem by having a free-castoring tailwheel and using differential mainwheel braking for steering control, and perhaps their tailwheel can also be locked straight ahead, but I don't like that arrangement, I prefer to have full-time steering control.

 

If I was to add simple splices to the rudder cables for the tailwheel steering, the tension required on the cables going to the tailwheel would pull the rudder cables out of alignment and they wouldn't provide a positive 'feel' or positive control, let alone that they would rub against parts of the airframe at any time that there wasn't a firm pressure on both rudder pedals. My solution is to provide a pair of idlers at the splice point, easier described by images than words.

 

The first picture shows the annoyingly small parts that had to be welded together (back to the spray-paint templating method again) and the last picture shows the assembly installed in the aft fuselage (lower centre of pic) -

 

 

Having done that I made and added the rudder control horns - with plated re-inforcement on their aft faces so that hopefully no matter hard someone stamps on the pedals it shouldn't deform the rudder horns, and I added the fixed rudder control stops ahead of the rudder horns, welded onto the rudder post. I didn't see any need to make them adjustable, they just allow the rudder to move 27 degrees each way and then reach a dead stop -

 

 

Continuing the same theme I added the turn limit stops to the tailwheel assembly and also its steering control horns -

 

 

The next step will be to add the pulleys which guide the rudder cables from the pedals to the idlers, I'll need more than I originally anticipated, to keep the cables clear of the reflex curve in the underside of the aft fuselage.

 

Another 31hrs in that lot, bringing the total time so far to 991hrs.

 

 

 

 

 

[pylon500]

Yes, very fiddly little build..

 

Had a similar problem on my Stollite, having the rudder on top of the fuselage and tailwheel below.

 

Solved mine by splicing the cable, but then running all four cables over four pulleys to maintain tensions, yes, springs at the tailwheel attach;

 

 

Not really as easy as it looks because of the cables following the taper of the fuse requiring the pulleys to be at slight angles.

 

(Thought I had a photo inside the fuse?)

 

 

[djpacro]

I've done a lot of time with lockable tailwheels - if set up properly they don't need to be locked to track reasonably straight on landing and steer in taxi with rudder generally, brakes only required in a strong wind - Haigh or Raven https://www.skyshop.com.au/pdf/2009-2010_catalog%20254.pdf

 

I note that many ag pilots disconnect the tailwheel steering springs.

 

 

[Oscar]

DJP - be fair!. From what I've seen, most Ag. pilots don't even drop the tail onto the ground until they're pivoting on one main beside the spread heap / chemicals truck..

 

Taxying? Waste of time... Cross-wind? - minor irritation...

 

I used to fly gliders a bit at Forbes, and part of the 'all clear above and behind' check, was a scan of the horizon to see if Darby Munro was headed in. Darby simply flew straight back to his hangar from wherever he'd been working, and from the air, one could see the shadow of his plane get slightly further away at fence lines. He'd long, long ago decided that 'circuits' were a waste of time and avgas. His departures were more easily noticed - if you heard an engine revving, he was coming from his hangar, but on what heading, you never knew. You just held off launching until he'd buggered off. Local legend had it that he'd - one time only - taken his Pawnee down to Bankstown for a check of the instruments and had returned swearing that he'd never go again because he'd had to get to 1500 (?) to enter the circuit and that was bloody dangerous.. AFAIK, he lived to a ripe old age - lack of tailwheel use notwithstanding.!

 

 

[Oscar]

Actually - small thread drift here, HITC, I hope you don't mind, but it does hark back to your own experience mentioned re extreme use of rudder in a dangerous situation - that 'low-flying aircraft check', saved a horrible accident once. We were winch launching; a Blanik with two aboard had just given the 'all out' and was starting to roll, when the CFI screamed 'bung off'' - which they did. About five seconds later, and an F111 did a pass straight down the centre of the strip at about 50 feet and maybe 350+ knots - would have collected the wire for sure. The CFI had done a 'Darby-check' glance out of habit - hadn't actually seen the F111 coming but noticed a patch of heat haze where there should not have been one and simply reacted out of one of those 'prickles on the back of the neck' moments.

 

The pressure blast, dust, grass and turbulence, not to mention the noise 'thump', left us all shaken to hell and several gliders that had been parked tidily awaiting launch, parked most untidily - but undamaged. The CFI (Noel Winterburn, damn good bloke and Instructor, of course) apparently left the RAAF in no doubt as to his feelings in the subsequent call he made to complain...

 

 

[Head in the clouds]

April 8-9th 2016

 

I modelled the rudder cable guide pulley hardware in CAD, produced the drawings of the plates, spray-paint templated and cut them out. That took a whole day, it's a time consuming business cutting out, drilling and deburring these small plates by hand. However the only other option would be to fully complete the CAD model of the design before starting any of the build and then the plates could all be cut by laser. The down side of that is that it would have taken about another year of design work to do, so I would only now be starting the actual build. Since I do the building rather than the CAD work for relaxation, I'd be even more bananas by now, than I already am.

 

Also - the design has developed because I've had time to think about it while fabricating the fuselage frame, so it's a better design now than it would have been if I'd done all of the CAD work beforehand, one always discovers improvements along the way.

 

There's nothing stopping me getting the plates laser cut as I go along, of course, except it's not cost- or time-effective, the laser people don't like doing really small runs so they charge a large price premium, and also there's a fairly long lead-time, usually 2-3 weeks, so I'd be held up a lot of the time. Anyway, cutting them by hand saves a few dollars and means that at the end of the day more of the plane is hand-built than it would otherwise be.

 

Below are a couple of images from the CAD showing the design development, note the large angles at which the pulleys have to be set to properly align the incoming and departing cable. Unless these angles are set accurately it results in very rapid wear of the sheaves - they are made of Tufnol and have quite a long life if well aligned, but can chew out in 100hrs or so if not, and they're not cheap so you don't want to be having to replace them more often than necessary.

 

I also made up small tubular spacers so that the pulley bushes run on them rather than the central bolt, and that allows the bolt to be done up tight against the spacer. That way the bolt can't turn and wear the holes in the clevis plates.

 

 

Then it took all of the next day to weld them in. As I've mentioned before, it's not the welding that takes the time but achieving the setup. Developing ways of holding these little mongrels 'just so' while getting them tacked was quite exasperating. Actually I didn't weld in two pairs of them because I realised that they are in the space where the strut attachments also fit, so I needed to complete the design of those plates first, to see how they'll all fit together.

 

 

Once that was done I went back to modelling, and developed the strut attach plates -

 

 

17hrs in that, taking the total over the thousand hour mark - to 1008hrs

 

 

 

 

 

 

[Head in the clouds]

April 12th, 16th-17th.

 

I've always had a bit of a 'thing' about strut attachment points. On any plane I've built I've seemed to consider them the most vulnerable to failure of the entire structure. It's probably a completely unreasonable concern but if it's there, niggling away at you it's a bit harder to relax and enjoy the flying. I'd pretty much stopped worrying myself over it and then a well-known brand of Aussie plane had their wings fold, and another nearly do so, due to corrosion in the carry-through, so I began to consider my fears justified again.

 

Consequently, during the design phase of the DooMaw I had a close look at numerous steel-framed aircraft to see what solutions previous designers had adopted for their lower struct attachments. I saw quite a variety, some that looked like they could lift a tank and others that frankly gave me the willies. Anyway, seeing how incredibly light some of them are, and still manage to keep the wings in place, was an education in itself.

 

It's not that their design is anything particularly arduous, it's easy enough to work out how much material is required to satisfy the tensile and compressive loadings at 6G plus a safety margin, it's more the problems that can arise due to corrosion, shock loading from landing and ground handling, and in my case, the completely different loading direction that's imposed when the wings are folded and the plane being trailered over rough roads perhaps, because the wings are still being supported by the struts when it's in the trailer.

 

So - this time I wanted strut fittings that can take whatever's thrown at them, and to be honest that's why they're being added at this late stage of the program, because I've taken this long to decide exactly the best way to make them.

 

Once the design was completed I just had to do the usual spray-paint templating, cut them out, make spacers on the lathe to clamp between them, drill and ream them, fit them up, grind the weld prep angles, tack and weld them. Sounds simple but took three long days.

 

The lower plates are simply tuning-fork shaped and bent at 29 degrees so that each arm of the fork welds down each side of the very substantial 1.25"/32mm x 0.125"/3.2mm carry-through tube, then the top plate is made up from three parts where the fork arm is above it on one side and below it on the other side, this is to fit around the bend in the lower longeron.

 

The pictures should tell the rest of the story, the assembly is mighty strong, I don't think I'll need to have any concerns about it ...

 

 

Another 26hrs for the log, a total of 1034hrs so far.

 

 

 

 

 

 

 

 

 

[Oscar]

Mighty strong - absolutely.... lovely work, as always.

 

 

[pylon500]

I hope this doesn't cause a huge problem, but if you have the strut fitting pivot bolt angled to be square with the strut, what are you doing to align the bolt, when the wings are folded back?

 

The Kitfox (and it's clones) keep the bolt vertical, in the axis of the wing fold, and have the strut attach fitting angled on the strut.

 

Because the strut to spar angle on my Stollite was angled, I opted to use a ball joint style fitting to make up for the misalignment.

 

 

Wing is folded in this view.

 

 

[Head in the clouds]

I hope this doesn't cause a huge problem, but if you have the strut fitting pivot bolt angled to be square with the strut, what are you doing to align the bolt, when the wings are folded back?The Kitfox (and it's clones) keep the bolt vertical, in the axis of the wing fold, and have the strut attach fitting angled on the strut.

Because the strut to spar angle on my Stollite was angled, I opted to use a ball joint style fitting to make up for the misalignment.

 

Wing is folded in this view.

Yup, the wing folding on DooMaw is rather more complex than the single-plane type of folding that the Kitfox and similar have, because when they're folded DooMaw's wings lie flat against the side of the fuselage (a bit like the Grumman Avenger, but even more so), consequently there's two axes of freedom and rotation at the top strut fitting and two axes of freedom at the rear spar attachment.

 

Because of those multi-hinged points, and because they're not arranged as a star-fitting (like a universal joint) but are offset hinge-lines, the lower strut hingeing wouldn't be aligned with the others even if it was pinned vertically. I was quite intrigued when I came across M61A1's "The Mistress", which is where I first saw this very clever and useful hingeing arrangement. I've tidied it up a bit for DooMaw but unfortunately I still wouldn't be justified in claiming the design as my own, even though it will now be the truly 'instant' wing-folding method I've always hankered after.

 

The other two attach points are a little difficult to describe at this stage so I might have to leave that until I have modelled it in CAD and can show the actual components in an exploded view, but for the meantime here is a picture of The Mistress's lower strut attachments to be getting on with, they should show how DooMaw's swivel alignment will be achieved, it has the same effect as yours on the Stollite - the third pic shows The Mistress in her trailer note that the wings are on the plane, you can just see them folded flat down the sides of the fuselage, and they're still attached to, and supported by the struts.

 

This type of hingeing provides two massive benefits when compared with the type used on the Kitfox and similar - you can still walk around in the trailer because the wings don't take up its full width, and your wing-chord width isn't limited by the need to get it into legal roadable dimensions.

 

 

 

 

 

[pylon500]

So that being the case, what are your plans for the wing end of the strut?

 

I am planning on what I'm calling a 'double articulated' wing fold (folded back and swung down, as on yours and M61A1's machine) on Planet47's machine.

 

To that end I was looking at the system used on an Italian machine called the 'Groppo Trail' (similar in appearance to my Stollite) until I saw the actual aeroplane at Oshkosh in 2013, and was not totally impressed..

 

 

Things that had me concerned;

 

All flight loads taken by the threads of wing attach bolt,

 

Wing fittings taking flight loads through a 'diaphragm' style structure, not in shear,

 

Bolt cannot be 'tight' or clamped to allow wing to fold.

 

(Sorry if photo's are a bit confusing, 1st photo is LEFT wing folded, 2nd photo is RIGHT wing rigged.)

 

On the whole I was a bit disappointed in the Groppo Trail, being a little more agricultural than first thought.

 

Having said that, the attach system I've come up with for the 'Planet Pusher' is probably overkill...

 

 

 

Discuss?