DooMaw - building a STOL

planet47 · April 18, 2016

"Planet pusher"?

pylon500 · April 18, 2016

"Planet pusher"?

Guess we could call it the Atlas?

Head in the clouds · April 18, 2016

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.......
..... Discuss?

For those who may not know the Trail, it has a wing-folding system that also folds the wings flat against the fuselage but it's the front spar connection that remains connected when folded, rather than the rear spar connection.

This has the very distinct advantage of making the whole folded package shorter but it has two disadvantages -

The folded package is very much taller. That doesn't matter if you're only going to store it in a hangar (as long as you've got very high hangar doors) but for trailering it means the windage on the side is substantial and requires more care when driving. That's not to be scoffed at, I lost a very good friend in the late 1980s when their ultralight plane's trailer was blown off the road by a willy-willy taking the tow-car with it and they hit a tree.
When folded the majority of the wings' weight is above the spar attach point and so the wing is unstable and requires additional bracing members to hold it in place when trailering, that takes additional time to install when folding and unfolding the wings.

Here are a couple of pics of the Trail -

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...

Some while ago I was quite astounded to find that in general engineering bolts are used in tension whenever possible and in shear only if a tension orientation cannot be achieved. It's only in aircraft that you see a lot of situations where the bolts are used in shear, and this is mainly where the same bolt can be used in double, triple/multiple shear, like a wingspar folding knuckle, or where rod-ends attach on a control pushrod, for examples. Interestingly a 'proper' bolt can realise its full tensile strength against the thread of the nut and in testing should elongate and eventually snap at the transition between the thread and plain shank. The (single) shear strength of a bolt is only about 2/3 of the tensile strength. For a very relevant example of 'general' engineering using bolts in tension - ALL lifting lugs (craning points) are attached using bolts in tension, never in shear.

There is a caveat to the above though - the bolts must always be properly torqued to the correct values for the bolt size and threadform to prevent the thread fretting when loaded and unloaded, so the 'loose' bolt for the rotation in the upper strut assembly of the Groppo is an absolute no-no, eventually the threads must wear out from the cycles of flight loading.

OK, the 'diaphragm' aspect. I agree, it's not attractive from an engineering viewpoint on the Groppo but I suppose all bolts in tension are acting through a 'diaphragm' arrangement in some sense, it's just the thickness of the diaphragm that varies, make it thick enough and you're bolting through a block ... which is the direction I'm headed, the bolt used for the rotation goes through a thick aly block which is milled to provide a pair of cheek plates/clevis for the connection to the next part.

Going back to the bolt used for rotation - I can't see a way around using it in tension, and for the reasons given above, there's nothing wrong with it at all but I will be trying to source a long nut, or machining my own from chromoly. There's probably no need for that but if it makes me feel better ... Then, that bolt must be fully torqued and that means it must have a sleeve on it and be torqued up onto that, and the rotation takes place around the sleeve rather than just around the bolt.

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

Referring back to your photos of the Groppo strut-to-wing attachment, it has one axis of freedom (hinge) and one axis of rotation. When I studied The Mistress in the same location it has two axes of freedom plus the rotation. I haven't actually made a model of the system but I have been wondering if the second axis is really necessary and the Groppo photo shows that it isn't essential BUT you can see that if you don't rotate the wing as you extend it, exactly at the time and rate that the geometry demands, then you will be stressing the attachment fittings. The second axis prevents this stressing, as far as I can see at this stage, and means that all that is necessary is that you rotate the wing at some time between the folded and extended positions.

Incidentally, the upward angle on the lower strut attachment fittings that you commented about previously, apart from aligning everything when the wings are extended for flight, does serve another very useful purpose. When the wings are folded the strut 'slumps' into the folded position which means that even if not secured back the wings will not inadvertently swing forward if the tail is lifted above the horizontal, this could save a lot of red-face in the hangar ...

I'll post some photos below, of the fittings on the Mistress upper strut end, and also of the rear spar attachment. The latter is a rather untidy affair and is probably the result of a bit of trial and error on the part of the designer/builder (he was a German toolmaker and obviously very talented). It's an arrangement whereby a small member carrying a hinge in one axis swings out on the other axis (when it isn't swung out both axes are locked with a vertical pin), I have in mind to make that a lot tidier with a similar connection to the lower strut fitting, getting rid of the need to remove an extra pin, with the associated hazards of forgetting to replace it before flight. The Mistress's version does however provide the benefit of elongation in that area when unlocked which gives extra clearance for folding as the flap control rod disconnects and the flap droops - I'll use a spring to hold the flap up instead, and a small removable fillet fairing.

The first photo shows the upper strut fitting on the Port side. The bit that is bolted into the end of the strut carries the bolt for rotation, then there's a heavy clevis with vertical bolt for fore/aft motion (in theory I think that's the one that isn't entirely essential but might be better considered as 'good practice'), the clevis carries a block (like the lower strut fitting) with horizontal bolt for vertical strut motion which allows the strut to nest against the wing underside when folded.

The rest of the pics show the Port rear spar fitting unpinned but with wing extended, Starboard rear spar fitting, first with wing extended and pinned, then with pin removed, then with wing folded and last pic shows the lower strut fitting with wing folded.

Head in the clouds · May 29, 2016

29th May 2016

After completing the strut attachment points I posted about previously, I fitted a small bracket to the top of the aft fuselage to carry the VHF antenna and then made and welded in the harness attachment points. I forgot to take photos of them so I'll post them next time.

Pylon500's comment about the length of the elevator pushrod had been gnawing away at me in the background so checked my buckling calcs and it still came out fine but when I was wheeling the fuselage to the workshop over the gravel driveway the push-rod had a very noticeable bounce in it that just doesn't look good. Oscar very generously had sent me some Purathane material to make a grommet to support the tube mid-span but that would still mean it would flex and any contact with the grommet would cause some wear and it's a very thin-walled aly tube (0.035"/0.9mm). The amount of work required to make the grommet and its support would have been as much, if not more, than the work to add another walking beam and divide the long pushrod into two.

I was a little miffed at myself for having to do that at this late stage, and particularly because the two ends of the pushrod that attach to the walking beam would be attached in tandem and the Ackerman Effect would slightly change the elevator geometry, affecting the up and down ratios. Then I realised I could place the two rod-ends beside each other rather than in tandem and not have a geometry change. At first it appeared that doing that would introduce a bending moment in the pushrods but it isn't actually the case, all of the bending is resolved in the bolt and cleats that hold the pushrods to the walking beam. Problem solved, so I machined up the new pair of large pushrod ends, and made the walking beam and attached its hinges to the lower fuselage between the rudder cable-guides. Photo of it installed next time but here's the walking beam frame -

Next I made up a couple of flanged plates to go above the cabin and secure the clear polycarbonate sheeting that forms the cabin roof. I used a tool that I'd made a couple of years ago for another project. It's a pair of ballraces mounted on socket head cap screws that I'd machined eccentric so that by turning the screws the rollers move closer or further apart and can be adjusted to roll different sheet thicknesses. Ahead of, and behind the two horizontal rollers are two small diameter tube rollers also on ball-bearings, that are set a little lower than the bottom of the horizontal rollers. There is one more roller, vertical, between the two main rollers. By feeding the sheet vertically into the slot between the main rollers and keeping its edge on the approach side feed roller the sheet bears down firmly on the fifth, central roller, allowing you to put sideways pressure on the sheet and roll a flange along its edge. The edge can be curved or straight and due to the limiting roller between the main rollers, the flange is always the width of the horizontal bearings - 12mm in this case. I then used crimping pliers to remove the excess material which was due to the curved edge and the stretching during rolling, to leave a flat flange.

Life got in the way for the next four weeks with an overseas trip, re-springing a couch, fitting curtains, building a built-in robe, a birthday BBQ for my wife and my drafting work got busy too so it was a hectic time and I haven't touched the plane until this weekend. By then I'd lost all momentum and it was a case of staring at it for a while to decide what to do next. Somewhere among the above other projects I'd managed to order some quality 316 steel rigging cable, a hydraulic swaging tool and a bunch of wire balustrading hardware to assess for suitability. The first test was to get a couple of friends and the three of us bounced around on some balustrading that had been installed on a job I'd drafted a couple of months ago. It stood up to it surprisingly well, nothing broke though we did stretch the wire. I went to the stainless company and was let loose on their test-rig for an hour and came back well informed, the balustrading hardware will easily carry the loads for my tailplane bracing. I'll be test loading it later in any case ...

The stainless swaged ends for the cables are as light as they could be and still do the job, and having threaded ends they can be used to do the cable tensioning. I didn't want to use the bottle-screws as they were, they'd be too bulky, so I cut them in half - which gave me the left-hand and right-hand threads I'd need for the tensioning. I could have done the same with ordinary left and right nuts but I wanted more (longer) thread and the bottle-screws have 11mm of thread internally. So making up the cleats and welding them in took up the rest of the weekend. Back to work again tomorrow, ho hum ...

Another 38hrs for the log, 1072hrs so far.

pylon500 · May 29, 2016

Like the hydraulic cable swaging tool.

rotax618 · May 29, 2016

In my opinion wing folding should only be used for storage, removing the wings for transport is far safer, Bill Knight said that the G meter on his Hyperlight would regularly show over 10G following transport on a well sprung trailer, these unacceptable shock loads could easily cause structural cracking and premature fatigue to critical components of the wing attachments.

Head in the clouds · May 29, 2016

In my opinion wing folding should only be used for storage, removing the wings for transport is far safer, Bill Knight said that the G meter on his Hyperlight would regularly show over 10G following transport on a well sprung trailer, these unacceptable shock loads could easily cause structural cracking and premature fatigue to critical components of the wing attachments.

That's fair comment but it all depends on how you address the situation.

Knowing that road loads are hard on the airframe means you have to provide adequate support for the wings. If you're not going to remove them, the first part is to secure the folded wings to the aft fuselage to take the load off the strut attachment and wing root attachment points. The next part is to support the fuselage, or partially support it, to reduce the loads on the landing gear. In my case, and any bush plane, the main gear can easily withstand the road loads and the big soft tyres and long-travel suspension help to damp the shocks to the airframe. However it's a different matter for the tailwheel when the wings are folded because the wings then add a lot of extra weight on the tailwheel, so in the trailer the aft fuselage need to sit in a shock-absorbing cradle.

The last part of it is the suspension under the trailer. If you're towing any aircraft you need to use airbag suspension not steel springs, this has been shown to be sufficiently soft-riding to even transport helicopters on their skids, and with blades attached, albeit with supports at 2/3 span. The trailer is often the culprit when damage is sustained to a towed aircraft. It pays to build the trailer heavy, though there is a tendency among people to build them light because they're carrying a large but lightweight load. A heavy trailer with airbags absorbs a lot of the shock rather than transmitting it through to the airframe. Anyone who has fired both a lightweight shotgun and a heavyweight one will know what I mean ...

Your comment is perfectly valid though, if you don't take the extra measures to protect the airframe. Just securing a plane sitting on its landing gear with folded wings supported by the struts, in a steel-spring suspended trailer and towing it on even a relatively smooth bitumen road is likely to end in tears, let alone the damage that would be done if you towed it on a rough surface. In my mind the best way to transport an aircraft, if at all possible, is to fly it of course.

Like the hydraulic cable swaging tool.

Yes, not expensive either, $50 + $10 postage, cheaper than my manual swager was. It's a twelve tonne press and the dies are 10mm wide. The swager the stainless wire rope company uses commercially ($2 per swaging operation plus the cost of the swage) is a 30 tonne press and their dies are 25mm wide so I figure the compressive force/area is similar. As you can see in the above picture I applied the press twice, I know you don't do that with Nicopress and similar sleeves but with these long rope-end sleeves it seems like a better solution than one pressing in the middle - thoughts on that?

Heavy Duty Hydraulic Swaging Tool

As purchased it comes with eight different-sized sets of swaging dies but they're only suitable for copper or aly swages. In the text they mention that if you want to swage the stainless ends like I showed in the image above, then you need to buy the hardened die sets. They have two pairs which are the right sizes for 1/8 and 3/32 cable swages and they charged $10 for the two sets. And setting the stainless swages takes a fair bit of effort so you grow muscles at the same time - Bargain!

Head in the clouds · June 19, 2016

May 30th to June 19th 2016 -

A while since the last progress report due to the day job being rather busy and so I've not had much time to devote to the DooMaw build.

I did manage to get a bit done in stolen hours here and there, and during parts of the weekends. At present I'm trying to finish off a multitude of partly completed tasks. Some while ago I made the pedal sets but had to make a couple of small changes to make them more readily dis-assemble-able for servicing, and I still have to make the recoil-style balance springs for them.

In the most recent previous progress reports I showed hardware that I'd been making and installing for the rudder cables, so now that I'm equipped with the cable and swaging gear I showed earlier I was able to start running the cables and making their end fittings. Concurrent with that I made up the top tail bracing wires using the tensioning fittings shown last post, and then I could devise the quick-release mechanisms for the lower wires, so that the horizontal stabiliser and elevators can be folded quickly using the all-important pylon500 hinge! Very time consuming making all the small components for that, if I didn't enjoy doing it, it'd be a far more profitable use of time to do more of the CAD work in advance and send it off to be laser cut instead. The quick-release is based on the pelican hook concept such as is used for the releasable cargo hook on a helicopter, or a bomb-release mechanism. I'm happy with how mine work but I'm going to re-make the actual hooks and the hook retainers because I'm not 100% happy with their geometry, but that's a minor job for another rainy day.

Once I'd finished all the rudder and tailwheel steering cables, working from the tail forward, I could finally position the pedals assemblies and mark and drill the floor to secure them. It took a bit of time to get them right because I wanted the full forward adjustment position to be as close to the firewall as possible without the pedals, brake pedals or brake hydraulic cylinders actually hitting it regardless of several variables. I also wanted to install thin stainless steel slider plates between the floor hold-down screws for the pedal-bar supports to bear on so they don't wear the floor timber in use. When I got all that set out it was an anti-climax actually drilling the eight bolt holes ... then I could locate the positions for welded brackets to be installed at the firewall which will secure the forward (static) ends of the rudder cables. Once those brackets are made and installed the pedals will be able to operate the rudder via those S tubes described earlier. Before I can do that though, I have to paint the final epoxy coat on the pedals because once the forward end fitting is swaged onto the rudder cable I can't remove the cables from the pedals without cutting them because the end fitting can't be pulled through the S tube.

Last job before the big rains yesterday was to measure and divide the elevator pushrod which, as discussed previously and well pointed out by pylon500, was too long in one piece. I'd previously machined up another pair of the rod-end fittings and made the centre walking beam, so it was just a case of accurately marking the pushrod above the centre of the walking beam's travel with the elevators set neutral, marking the portion of tubing to be removed to accommodate the length of the new rod-ends - and cut the tube. It also provided me with the chance to do something that had been nagging me because I forgot to do it when I originally made up the pushrod.

I'd forgotten to paint the interior of the tube to protect it against corrosion, and that was a very remiss omission given that the tubing is only 0.9mm/0.035" wall thickness, so even a very slight amount of corrosion could make a significant difference to its serviceability. Painting the interior presents its own set of difficulties, I've tried with various kinds of swabs before without very satisfactory results and I've seen attachment full-circle nozzles for spray cans that do the job but didn't have one or know where I'd locate one ... so I put a bolt in the rod-end thread to block it and taped over the rivets with PVC tape and poured a cup full of 50% thinned etch primer into the tube then rotated it in a set of padded V blocks that were set not quite horizontal, so that the paint was just slightly pouring out of the open end into a catch can, it worked perfectly. Then I inverted the tube for a few minutes to drain all the excess and set up my heat gun blowing through the tube on warm setting for a few minutes to dry it and evacuate any remaining thinners. The thinners and heat gun combination didn't make the workshop blow up which was pleasing, and I was able to rivet the new rod-end fittings in and install the two pushrods to the walking beam.

A few photos -

The first three show the harness attachment cleats, shoulder belt attachments at the back of the baggage door, they'll have crossed cables so that each cleat shares the load of both occupants. The lap belts are attached to cleats on strongpoints near the strut attachments.

The next picture shows the antenna support plate. The harness cleats and this plate were installed a while ago but I forgot to takes pictures at the time.

The pedal assemblies in place

The elevator pushrod now divided in two, and showing the centre walking beam straddling rudder cable guide pulleys

Tail bracing, and showing quick-release mechanisms for folding the tail

37 more hours for the log - 1109 hours so far.

geoffreywh · June 20, 2016

I just read your post for the first time, and, coincedentally, I was looking at a tube joing program, it might be of interest to readers. It's called tubemiter.exe and to be found on TonyFoales website. If I can figure how to link it to here I will.

June 20, 2016

I just read your post for the first time, and, coincedentally, I was looking at a tube joing program, it might be of interest to readers. It's called tubemiter.exe and to be found on TonyFoales website. If I can figure how to link it to here I will.

Is this it?

Tubmiter lets you input the diameter of the pipe/tube, the wall thickness, the diameter of the tube/pipe you are joining to and the angle of the joint. It then spits out a paper template you can print and wrap around you tube/pipe to mark your cut. Using this software anyone with almost any type of grinding tool can make close-fitting miters on almost any size of tube.

Free download:

http://www.ozhpv.org.au/shed/tubemiter.htm

June 20, 2016

Is this it?

Well, the upload was inop so I'll hunt a mite more.

June 20, 2016

I just read your post for the first time, and, coincedentally, I was looking at a tube joing program, it might be of interest to readers. It's called tubemiter.exe and to be found on TonyFoales website. If I can figure how to link it to here I will.

Here's a site which think has a link. Didn't try it. Got to run an errand.

Tube Miter Program by Giles Puckett

facthunter · June 20, 2016

Just for clarity we still call it a "mitre" here. Splice might be a suitable description and "scribed" is used in carpentry. Once you have about 5 members a "mess" might be a better description. Nev

bexrbetter · June 20, 2016

Here's a site which think has a link. Didn't try it. Got to run an errand.Tube Miter Program by Giles Puckett

One of my standard tools and I have used it many times over the years. I even wrote to him once thanking him and he replied!

I usually add a mm to the tube diameter (the tube that will be mitred that is).

gareth lacey · June 20, 2016

once you have 3 to five pipe/tube meeting its a "nodule"(in the fabrication industry) and we used the template as above for pipe 50mm up to 1000mm

Head in the clouds · June 20, 2016

This is all very nice folks but you're 'chatting' on my build log. Could we keep the discussion relevant to DooMaw and/or good engineering practices, thanks.

Oscar · June 20, 2016

A mitre drawing programme is useful, but it's not a huge amount of help if you don't understand the shrinkage associated with building a multi-tube structure. As the number of members of what is essentially a truss increases, the stiffness of the truss increases and it becomes impossible to get an accurately-cut tube to actually join into the structure.

The mitre programme provides a cutting guide. Most I have looked at do not provide a construction sequence, without which a multi-tube truss-type structure will end up either distorted / having very large locked-up stresses. Don't mistake a handy tool as the answer to the Maiden (constructor's) Prayer!

There is more to be gleaned from HITC's build log than just the basics of welding tubes together!

AussieB1rd · June 20, 2016

This is all very nice folks but you're 'chatting' on my build log. Could we keep the discussion relevant to DooMaw and/or good engineering practices, thanks.

And a great build log it is, your work is awesome, have enjoyed it to date.

SrPilot thanks for the Tubemitre proggy, needed something like this for years, no more tears.

Head in the clouds · June 26, 2016

June 25-26th 2016.

The day job's still busy, so I only had the weekend to get a bit done on DooMaw.

Following on from fitting the pedal assemblies I had to make and fit a bracket at the firewall which will secure the ends of the central pair of rudder cables - the outer pair of cables will be secured through the engine mounting plates at the firewall lower corners.

On Sunday morning I was raring to go and then had what seemed to be a breakdown of the TIG machine. It refused to deliver any Argon gas through the torch. I couldn't hear the gas solenoid operating in the machine itself, so I guessed that was the problem but thought it could also possibly be at the regulator on the gas bottle. I've got the new kind of Argon bottle from BOC which has a built-in regulator so I had my suspicions, and particularly since it was our coldest day of the year so far, hovering just below 6C.

I disconnected the outlet hose from the regulator and although the reg was pressured up, no gas came out. Ah ha, I thought, some valve or other doesn't want to play ball ... I got out the trusty heat gun and pointed it at the regulator from all angles until it was pleasant to warm the hands on and still no gas would exit. After about an hour of fiddling I'd just about decided that my day's welding was over and was developing dark thoughts about the conversation I'd be having with BOC about their wonderful VIPR system ... when I wondered whether it was just possible that they might have an auto cut-off valve on the outlet. I've never seen such a thing but sure enough, as I re-connected the gas line gas hissed out of it.

Then it had to be in the TIG machine itself, so off came the casing. Everything looked normal except the solenoid-style gas valve wasn't operating when the foot pedal was pressed. I isolated the pedal controls and operated the switches manually, still no action from the solenoid. Judicious tapping of the solenoid with a jewellery hammer while it was powered up didn't produce any results either. With it installed it was very difficult to get enough access to test the solenoid by applying external power. I could see from the feed from the welder's power supply that it was a 24V DC unit but there was very high AC voltage nearby, so I elected to remove the valve entirely for testing.

I'd unplugged the machine from the mains but I didn't like the look of some large capacitors that would probably hold residual power at a voltage enough to do me serious damage, so I dug out some very large wire-wound resistors and put them across the capacitors first. Glad I did, from the heat generated I'd say there was probably enough power in them to blow my fingers off if I'd let them stray in the wrong place.

Half an hour later I had the valve on the bench - operating perfectly. There was nothing wrong with it. It took an hour to replace it and connect the gas lines securely, to be sure they were tight and wouldn't leak ...

Then the phone rang and I was chatting to my brother and glancing at the now rather unpopular TIG machine when I noticed one of the switches on the front didn't look quite right. The machine is a combination MMAW (stick) welder, plasma cutting machine and AC or DC TIG welder ... and when either replacing or removing the dust cover I put over it at night or when not in use, I must have caught the switch and it was now in the MMAW position instead of the TIG position. You don't use gas-flow for MMAW welding of course ... So we now all know that it doesn't take much to make a monkey of me on cold mornings.

Then, at last it was time to start fitting the tab plates which I had the laser-cutting people cut for me a while ago. They were a bit fiddly to get positioned for tacking but once the first few tacks were in place it is just a case of persistence and, for some of them contortionism, to complete their welding. The tab plates will be used to secure the aluminium sheeting for the forward part of the fuselage sides, and to attach the engine cowling. These tab plates are necessary to avoid drilling holes in the CRMO tubing which would weaken it and also moisture would get inside which would allow internal corrosion to develop.

There are thirty two of these tab plates to fit and I got eight of them done in the rest of Sunday, so I guess there's a couple more full days needed to finish them off.

I'm starting to see the light at the end of the structural work on the fuselage (no, not the navigation light), at which time I can get the epoxy coat on and stop having to be so careful about moisture protection. The remaining jobs are starting to get short enough to list -

Finish the tab plates
Make and fit the throttle bar and supports
Similar for the choke
Make/fit hinges for the doors
Make/fit latch plates for the doors
Add cleats to secure the fuel tank(s)
Fit the flanged cabin-top plates made previously to support the clear cabin roof
Design/make/fit the cabin-top aileron control attachments
Design/make/fit the flap control attachments

... and probably a few more things I haven't thought of yet.

Some pictures -

The VIPR gas regulator holding pressure though the delivery gas line is disconnected -

The cable termination bracket for the central pair of rudder cables -

Tab plates underway -

Not counting the time spent on the welder 'problem' it's just another 10 hrs for the log, making a total of 1119 hours so far.

Head in the clouds · July 17, 2016

July 3rd-17th 2016.

I've not had a lot of time to spend on DooMaw in the last 3-4 weeks, just five days during the weekends. It's mainly been more of the same as last time, adding the tab plates for the metal skins on the forward fuselage.

I'd started out thinking they would be done at the rate of about eight plates per day and with 32 of them to do I'd guessed I had about three more days of them to do. That turned out to be a bit optimistic when I got to the ones on the side and the underside as there is a fair bit of time required each day to roll the plane over carefully and remove the dummy landing gear, and to pack it all up again at the end of the day.

The thing that takes the longest time is cleaning up the burnt paint behind the strips after the welding in preparation for repainting, and the painting, because of the difficulty and absolute necessity of ensuring that new paint has penetrated into the acute angled corner behind the plate. That area is a natural moisture trap and so if it isn't very well coated, both by the etch primer and later the epoxy final coat, then it will be a certain place for rust to develop over time, and a very difficult place to treat later, so it's worth spending the extra time to get it right at this stage.

One of the plates is a bit different from the others because it has two folds in it to form an offset door jamb. Ordinarily the door jamb could be a simple strip welded to the back (inside the cabin) of the tubing but in this case I needed the edge of the door to be a little further back so that when it opens it doesn't clash with the 44mm/1.75" pneumatic landing gear strut, so I made the jamb plate with two folds in it which moves the front of the door aftwards by about 25mm so the door will open upwards past the strut, missing it by about 8mm or so.

To bend the jamb plate I used the mini brake press that I fabricated for folding the ribs for the tailfeathers (see posts 109 and 111). The bed and blade weren't long enough to fold them in one piece so I had to cut the jamb plates in half, fold the pieces and then weld them together again.

In case anyone's having to do something similar in the future there's a good trick to lining up and welding thin plate without distorting it and/or blowing holes in it - clamp the two pieces to some copper sheeting before welding. The copper takes the bulk of the heat away very rapidly and prevents distortion and makes it difficult for you to blow holes in the steel. This can be a very useful method when repairing car bodies and similar thin steel projects, particularly when welding close to an edge where blowouts happen most easily.

I also added four cleat plates which will attach the aluminium straps which hold the fuel tank in place.

Some pictures -

That's another 40hrs in the log, making a total of 1159hrs so far.

Kyle Communications · July 18, 2016

That last series of pics really show just how much work is involved in a molly fuselage....man there is some welding there

Oscar · July 18, 2016

Absolutely agree with KC (above)! I've built a few clubman- level sports car chassis (think Lotus Super 7, Caterham) and by comparison, they've been quite simplistic. Those who have followed this thread and thought about what HITC has been explaining, will appreciate the complexity of resolving lines of force within a tube structure, both for the 'normal' load case and the 'crashworthiness' load case. We old-timers may remember the famous Maserati 'bird'cage' chassis for their Le Mans cars - so-called because it was joked that a bird couldn't get out of them..

Personally, I really do not like the 'tube and riveted gusset-plates' joins methodology: it will work ok if well designed for normal loads, but unless it is really superbly designed, there is a real risk that in an 'exceptional' load situation, the joins will tear apart rather like perforated stamp sheets - and once that starts it only gets progressively worse and may leave short tubes flapping around to act as spears for the occupants. A fully monocoque structure bends (usually); a carbon-fibre structure will tear/shear if loaded adversely; a low-tech composite structure ( e.g. Jabiru) will also tear, but it takes a lot of energy to keep the tear going - it doesn't fracture as C/F will and some of that gets dissipated by flexing so the load gets taken out very progressively.

A tube structure can take seriously large loads if the force paths are properly resolved. A good - and readily-available - example, is a Jabiru engine mount: 5/8" 4130 tubes - they look almost like straws...

Great tip re using some copper sheet to dissipate heat from thin gussets - I've used heat sink technique for silver soldering but never thought to use it for welding!. Does it introduce any extra brittleness in the inter-granular cooling zone? Flying blind here ( little joke..) but I'd imagine that you'd want to leave at least around 2 - 3 times the fillet radius clear? I once did some filling of detonation damage to a Husquvarna racing bike 2-stroke head, and the bugger would crack at the edge of the weld pool until I set up a super-hot sand bed to bury the thing in instantly after welding (and pre-heated the entire head to stinking bloody hot before the welding!)

Head in the clouds · September 6, 2016

For some while there have only been oddments happening so there's not been a lot to write up in the log so it's time for a bit of a catch-up -

Between July 19th and August 13th I was working on the throttle mechanism and the plate for the engine control cables.

Unlike most two seat aircraft DooMaw will not be restricted to a single primary control seat i.e. both seats will have primary controls so that it can be flown solo from either seat. Usually that is not permitted because the second set of controls 'slave' off the primary set, hence any failure in the slave linkages would render the aircraft uncontrollable if flying solo from the second seat. The other requirement for two primary control seats is that any controls that are not duplicated must be operable from the second seat without undue difficulty i.e. must be accessible with the seat harness correctly fitted and adjusted, and must be operable while also being able to operate whichever other controls must be handled at the same time - changing hands on one or other is acceptable.

Also, if flying from either seat it means that all instruments must be visible from both seats, which in DooMaw's case they are, albeit that they will be more conveniently and conventionally arranged when viewed from the left seat.

For safe and effective take-off and landing management I wanted to duplicate the throttle lever for both seats so that the control column can be held in one hand and the throttle operated by the other without having to change hands at all. With a little experimentation I found that a single operating lever for the flaps would be acceptable, although better STOL operations would be managed from the left seat. The other non-duplicated controls would be the choke, carby heat, cabin heat, throttle friction and comms.

Comms will be on the panel and reached easily enough from both sides. Although I do want the facility available - to be able to fly it from either seat - it's still likely to be usually flown the left, so the other controls (choke, carby heat, cabin heat, throttle friction) will be placed for most convenient operation from the left and within reasonable reach from the right.

Finding a good location for a bracket to carry the three cables (choke, carby heat, cabin heat) proved more difficult than expected. I couldn't have them anywhere on the panel or on a forward centre console because although they are bowden-style cables they are solid core (not multi-strand cable) and the first 120mm or so from the control knob is quite rigid, so the cable run must extend straight forward for some way ... and the main fuel tank is located forward of the panel, so that rules out anything in that location that extends forward more than the depth of the instruments. Also - I need some structure forward of the cable mounting point to have a small bracket with a hole in it to cable-tie the cables onto, so that they don't flop around. The only suitable location proved to be on the left fuselage side below the panel near the pilot's left knee. In line with the crashworthiness objectives of the design this meant careful consideration of the shape of the bracket because the obvious rectangular shape would produce a serious hazard to the knee in event of a forward impact. Tapering the bracket off at the bottom, giving it absolute minimum projection out from the side (just wide enough to fit the flanged nut on behind), a folded outer edge and a rounded lower corner has made it reasonably unobtrusive.

Similarly, the throttle proved a little trickier than I originally intended. I was going to have a throttle bar running behind the underside of the instrument panel, with throttle levers each side curving down under the lower edge of the panel and then upward, like the Lightwing arrangement. For several reasons - one of them again being the fuel tank location - that didn't prove suitable. The next-favoured spot just in front of the forward seat support, under the knees, was convenient to locate it and provided a natural position for the hand, but is an area that is very convenient for putting 'things' like sunglasses, water bottle etc and that meant it was subject to fouling the throttle one day. So I decided to keep that area available for those 'things', in fact I'll probably make a small locker there each side so that it can be used with confidence that loose items won't escape into any of the controls. Which brought back memories of the mayhem caused to some station pilots during feral control in early days R22s. Before the pedals were fitted with rubber boots over the slots in the floor, ejected cartridge casings would regularly fall down the holes and if you were unlucky they'd jam the pedals which made for interesting landings sometimes ...

The result was that I decided to install the throttle torque tube behind the forward seat support i.e. under the front of the seat, and that meant to be able to install it and to remove it for servicing it would need to be made in two halves that bolted together. Also - one of my pet hates is throttles that 'creep'. With carbys sprung to wide open it often means that when idling the throttle may slowly advance on its own which I find annoying, and I also like to be able to secure the throttle quite firmly so that a passenger is less likely to bump it open at an inopportune moment. I don't like throttle locks because you can't over-ride them in an emergency, so it had to be an effective and adjustable friction device. I've fitted throttle frictions before, made from plastics, and they haven't been as effective as I would have liked, so this time I decided to bear the slight weight penalty and make it from solid brass. I could machine it far more precisely and the metal wouldn't conform to a new shape in the same way that plastics do when constantly under clamping loads.

A PVC plastic friction device from an earlier project -

I used split blocks bored through the split line with the bottom of the blocks bolted together with a o.1mm shim between them and the top of the front block is clearance drilled and the rear block has an M4 tapped thread. In the forward support tube is a compression spring which holds the blocks apart when the knob is turned left, and a half turn adjusts the friction from zero to very firm or anything in between.

For the throttle torque tube, one of the trickier parts of making bolted flanges on the ends of two pieces of tubing is getting them exactly square so that when the two tubes are bolted together they end up straight. The problem is that even if you jig them straight at first, the welding often pulls them around a bit and they don't end up that way. This time I set them up on the tube and welded them in place before cutting the tube in half. It worked well, the pictures tell the story.

The final parts of the throttle mechanism were making and fitting the throttle levers, the clevis plate for the cable inner, the socket which supports the cable outer, the adjustable throttle stop and non-adjustable one on the other side, the bearing plates which support each end of the torque tube and the support plate which holds the friction device in place. It's certainly a lot more work to make dual side-by-side throttles than a single throttle or even tandem ones where the rear one just has a link rod to make it mimic the forward one.

By the way, a couple of people have asked why I'm using thin plain nuts instead of nylocs. They're only temporary at this stage, and because they need to be assembled and disassembled many times during the build, they're much quicker than nylocs, which should only be used once in any case. All the bolts and nuts will be replaced with new AN bolts and nylocs or castellated/pinned nuts, or clevis pins, during the final assembly.

Another 43 hours for the log, 1202 hours so far.

Oscar · September 7, 2016

Good to see you back in the workshop, HITC - and beautiful work as always. Tinkerers like me, and serious builders like you, get withdrawal after a while... sure, we can give it up anytime we like, but now and then you just NEED another shot (at it..) to keep going..

I sincerely hope that people are reading thoughtfully ALL of the stuff you write, because it is very important that people who modify / build their own aircraft, understand the depth of thought that goes into this stuff. Just about everything in an aircraft is, in some way, part of a 'system' - and the entirety of that system has to work together. PART of 'working together' has to be a consideration - as you have bought out - of the consequences of failure of one part of the system affecting other parts (your throttle consideration is a fine example.) Standards (such as the FARs, BCARs etc.) are a basic guide in some, but not all, cases.

Just how far the 'system' extends in making the difference between something with 'fall-back' safety features and something that doesn't, is not always obvious. Earlier in this thread, you explained your rudder-pedal-adjustment locking set-up and we discussed the 'fun' component of having that mechanism fail! Whereupon, it becomes crucial that the effective operation of secondary effects of controls become of acute interest to the pilot - (as I can attest). Here's another example, about which I think I can nowadays talk..

In the early days of the development of the Sunbird Seeker, the control linkages were a cross-frame torque tube to link the two sticks together for elevator control, with a secondary tube welded to that in the centre which held the aileron horn - pretty standard stuff. Don Adams had assembled an engineering consultancy team including Dafydd Llewellyn (for aerodynamics) and Bill Whitney (for structures), and test flying duties were shared between several people - Dafydd Llewellyn being one, and I think at that time, David Ayres another.

David Ayres had been undertaking a series of tests, when the tube for the aileron control horn snapped off the main cross-tube - no ailerons ( the weld for the aileron control horn tube tore out). David was able to land the thing successfully, for two reasons: FIRSTLY, because even in the event of that significant failure, leaving the push-rods from each stick and the main aileron horn flapping around, they did not foul the operation of the elevators, and SECONDLY, because a (fairly recent) adjustment to the dihedral had ensured that the thing could be controlled in roll to a reasonable degree from secondary effects of rudder.

Test pilots get to earn their money - but they only get to spend it if the 'system' they are reliant upon, works..

Head in the clouds · September 12, 2016

More log catch-up - between August 14th and Sept 2nd I was working on finishing off the list of remaining 'things to do' to complete the welding on the fuselage so that I can get the epoxy paint coat onto the frame and begin final assembly of all the parts. I wrote up the list in post #167 above, here it is again -

Finish the tab plates
Make and fit the throttle bar and supports
Similar for the choke
Make/fit hinges for the doors
Make/fit latch plates for the doors
Add cleats to secure the fuel tank(s)
Fit the flanged cabin-top plates made previously to support the clear cabin roof
Design/make/fit the cabin-top aileron control attachments
Design/make/fit the flap control attachments

... and probably a few more things I haven't thought of yet.

Some extra things I've thought of -

plates to attach the gas struts to hold the doors open
some tab plates with small holes in them, to be able to zip-tie wiring and cables so they don't flop around
An attachment plate for a trim lever
A plate to carry the headset sockets

In my previous post I detailed making and fitting the throttle bar and choke cable bracket, then I went on and finished the last four tab plates which were two which hold the sides of the firewall and the two which hold the top of the windshield above the cabin.

I then made the hinges for the doors and the baggage doors, they were larger versions of the control surface hinges i.e. small pieces of shaped key steel drilled off-centre in the lathe and welded in place.

I decided against having plates for the door latches, instead the latch rods will locate onto mylar behind the door jambs which are already installed.

I did install securing plates for the baggage door, at the front corners I added plates which will carry the spring for a Dzus fastener, and at the rear corners I drilled the harness bracket to carry a similar Dzus spring.

The cleats to secure the fuel tank were added a while ago and described in post #168. Incidentally, DooMaw will have three fuel tanks, the main one forward of the instrument panel will hold approx 60 litres and will be fuelled via the lower starboard corner of the windshield, similar to a Wittman Tailwind, it's a good arrangement for a taildragger because it's the highest point of the tank. Having the filler through the windshield is one good reason to have a flat windshield if it's polycarbonate because being flat it's not under stress and so will not craze if (when) you inadvertently spill some fuel on it.

The other two tanks will be wedge shaped, about 25 litres each, with one located each side of the elevator pushrod, under the baggage locker. They will straddle two of the lower fuselage cross-members and will be plumbed so they can be pumped up to the main tank as required. They'll also be easily removable so they can serve as portable jerrycans when fuel has to be obtained some distance from the landing ground.

I had to go back to the CAD work for a while then, to design the aileron control linkages. I needed to transfer the lateral rotation of the control column's torque tube to create a longitudinal rotation of two aileron torque tubes above the centre of the cabin. I had to separate it into two torque tubes rotating in opposite directions to make the ailerons move opposite each other on each wing, of course. I also had to do that without having the pushrods where they would obstruct the baggage area or the occupants, and where they wouldn't be subject to accidental damage or fouling. To achieve that I would need two opposing pushrods running vertically close behind the seat frame to a pair of bellcranks at the top rear of the cabin. Those bellcranks to be supported by a pair of plates attached to the rear spar carry-through, and its associated bracing. From those bellcranks will be a pair of pushrods running forward along the cabin top, above and between the occupants' heads, which attach to control horns on the underside of a pair of opposing torque-tubes, one of which will operate each aileron. It was a tight fit to get it all in with enough angular motion available for the bellcranks and the control horns because in each case they run past/through bracing structure which limits the motion, thank goodness for CAD, it would be much more difficult without it. Here are a couple of CAD images which show the extent of the motions available. I'm getting 30 degrees each side of centre which is ideal to avoid too much Ackerman effect at this stage, instead it will be employed at the outboard actuation to achieve the differential between the ailerons' up and down travel angles. The blue pushrods operate the ailerons - I've only shown the port set, the starboard ones are a mirror image. The red is the flap handle, showing its range of motion, 0-60 degrees in 5 stages -

Next I fitted the flanged sheetmetal each side of the cabin top at the wing junction, they carry the edges of the windshield over the top of the cabin, give support to the small tubes which carry the door and baggage door hinges, and also carry the bushings to support the centre sections of the aileron and flap torque-tubes. I then went on and marked out the positions of the centre of those torque-tubes and drilled a 1/4" hole through which I could thread a 1/4" rod to hold the aileron torque-tubes' centre support in place (above the centre of the cabin top) while I tacked it in place. That process ensured everything was in perfect alignment, which it needs to be if the controls are to work smoothly and also connect and disconnect automatically when the wings are folded and unfolded.

I ordered the door gas struts online. The last ones I ordered, which were a similar size (300mm long when extended) to support a large toolbox lid, were over $100 each about 20yrs ago and lasted about six months. They're amazingly cheap these days, $23 for the pair, delivered by courier the next day, and with a 2yr warranty. Once I had them in hand I could estimate a range where they'll be fitted, so I made plates with a sequence of holes to allow for adjustment. I'm happy to have the extra holes anyway, because they're in the hat-rack baggage area and so will be useful for securing a cargo net or similar - in my book you can't have too many tie-down/attachment points.

And that was the state of play about ten days ago, some pictures -