Peter Anson – Sonex 894

SonexAus – Peter Anson – Sonex 894 Peter Anson’s Sonex 894
As you can see, this story does have a happy ending, with the flight of my Sonex 19-7898, but it took me six years to get to this stage, starting with the receipt of my plans set at the end of 2005. I don’t recommend building from scratch, but I had spent my spare cash on a Skyfox, and my budget didn’t extend to two aircraft. There are several good builder’s web sites (see Adrian Clout’s) so I’ll mainly show things that I did a bit differently, as well as a quick zip through the construction process.
January 2006 Like most builders, I started on the empennage. The parts are fairly small and easy to handle. Material was bought from Airport Metals at Tullamarine, Melbourne. Their prices for aluminium sheets are quite good, but extrusions are very expensive, possibly due to certification costs. I had access to a radius bender which made folding spars and ribs pretty easy.
All the internal bits were alodined and primed with zinc chromate paint. Chromate paint is no longer available, probably because it is a carcinogen. It might kill you but boy, does it stick! I tried alodining the skins too but was not really successful. You need a really big bath and good circulation of the alodine solution.

A Word in Favour of Sealants
I was a bit dubious about the use of stainless steel rivets in aluminium. I have been told that stainless steel fittings in aluminium yacht masts often cause corrosion problems, so I had a chat to a metallurgist at work. He suggested that corrosion could be prevented if oxygen was excluded from the steel/aluminium joint by using a sealant such as Duralac, which looks like very thick chromate paint. I dipped all the rivets in Duralac before inserting them in the holes. Was it worth it? I’ll let you know in 20 years time. Another sealant that I used in some locations was PR1422. This is a 2-pack sealant/adhesive that forms a very strong bond between skins and ribs etc. I used it to seal gaps and to glue skins to spars. It is very expensive (about $60 to $90 for a tube) and is intended to be all mixed within its sealed container, but it is possible to mix small quantities. (Since writing the above, I have found that PR1422 is about $100 per tube, but Aircraft Spruce sell a similar product called Pro Seal at a much lower price. Small tubes cost less than US$20.)
I also used the radius bender to fold the elevator and rudder skins but this was very tricky to get right. It took me about 6 tries with test strips before I had the confidence to risk bending the skins, and they were not bad, but not as good as the ready-folded skins from Sonex. When I came around to building the ailerons and flaps, I bought the skins from Sonex. In any case, there are not too many benders around big enough to do the flap skins.
The picture above shows a small change I made to the elevator horn – a double -sided horn. There’s no real weight penalty because parts can be made thinner and stress is reduced because there is no torsion stress caused by offset loading.
Here’s my modified angle grinder. It’s fitted with a 1mm cutting disk and a support frame so that it can be used like a circular saw. By using a straight-edge as a guide, long straight cuts can be done fairly easily and quickly. Not as good as a guillotine, but I don’t own a guillotine. It does put a fair amount of heat into the aluminium, but the heat gets dissipated so quickly that I don’t think there is any danger of over-aging the aluminium.
August 2006 Building going fairly smoothly, but it’s already obvious that this isn’t going to be a quick job.
October 2006 You can get an idea of how small my workshop is, and why some of the later photos show work being done in the adjoining carport.
December 2006 A long porcupine
January 2007 One year down and it’s drill, debur, alodine, paint, rivet
April 2007 By this time the fuselage still fitted inside the workshop, but only just. I usually dragged it out to the carport area to work on it.
May 2007 You can really get involved in this work.
November 2007 I did a Skyfox trip to Colac where Gary Wright was a bit ahead of me in building his Sonex. Mount Gambier builder John Lerwill and Gary are shown trying out the leg and elbow room in Gary’s Sonex. It looks like wind-down windows are in order.
January 2008 Getting close to finishing the fuselage. Here’s a happy snap with my daughter, Elissa.
This shows the wooden former ala Lynn Jarvis that Gary and I used to cut our canopies. We used an angle grinder with 1mm cutting disk for rough shaping and a flapper wheel for final shaping. I then spent about another 6 weeks bending and re-bending the canopy hoops and fine-sanding the edges to get a really good fit. All done and no cracks YET. Nearly 4 years later I peeled off the protective plastic coating and it was crystal palace all along the left side but OK on the right. Gary had used the same method as me and had only one crack! I made a couple of big long band-aids out of 0.5mm aluminium and glued them inside and out using PR1422. So far it’s worked. I strongly recommend using a different method using wider support strips for attaching the canopy to the sides.
March 2008 On its feet and looking aeroplany.
Trial fit of the empennage. It fits beautifully!
April 2008 Working on the wing spars. I bought my wing spars from another builder who had a major oops and mixed up spar caps. I de-riveted the spars and re-assembled them, which mostly worked OK. I had to fill a hole that ended up in the wrong place and had a problem with a row of 3/16 holes that are normally fitted with AN bolts. The holes were all oversize. The solution was to use Hi-Loks – the row of fasteners shown in the photograph. Hi-Loks are nifty fasteners that are used extensively in bigger and heavier aircraft. They have rivet-like heads (you can get countersunk) and are used with a break-off nut so you do them up to the correct torque. However, the big advantage for me was that they come in standard and two oversizes, 1/64th and 1/32nd. I used one in another critical spot where I had mis-drilled a hole. They are pretty expensive at about $1.70 for the bolt and $0.80 for the nut, but look very neat and can solve problems. If you are feeling really flash they also come in titanium for about $17 each!
The row of Hi-Lok fasteners when fitted

December 2008 Quite a bit of time passed between photos so I looked up my build log and noted that I spent 4 hours turning perfectly good material into scrap. I also cut all the wing parts, made the ailerons and flaps, pressed all the ribs etc. I had all the ribs CNC routed, made a former out of MDF and was about to start bashing the ribs into shape when Gary Wright put me into contact with a great guy in Geelong named Peter Carr. Peter is a retired engineer with experience in both the automotive and aircraft industries. He has an impressive backyard workshop equipped with a rubber press. The press uses a huge hydraulic ram (about 600mm dia) to push a flat plate into a large block of rubber. For a fairly nominal amount Peter let me use his machine to press all my wing, aileron and flap ribs. The ribs have to be partially fluted first, and then again after pressing to straighten the rib, but they came out looking like bought ones. In the mean time, Peter kept me fueled with cups of tea and great conversation. Incidentally, at that time Peter was pressing all the wing ribs for the Gippsland Air Vans.

Fitting the aft skins. Not shown is the critical part of getting the wing box square, which I did with the wing vertical and sighting along straight edges across the main spar and rear spar.
I marked the centre lines of the rib flanges and used a couple of notched sticks to line up the ribs with the pre-drilled holes in the skins. It worked quite well. Left wing shown.
December 2008 Formers cut from chipboard were also used, along with ratchet straps to hold the nose skins in place for drilling.
January 2009 Another porcupine.
About 18 months before this photograph was taken I bought my engine, but not the sensible Jabiru 2200 I had always intended to fit. I had even bought an engine mount for the 2200 from another builder and fitted it when a Jabiru 3300 came up on eBay. It was an early model, but brand new, and at a great price. I spent about a year procrastinating over whether to buy another engine mount or make my own when I read Stuart Trist’s account of the failure of his engine mount and that settled it. I had always been a bit sceptical about the design of the mount. The bottom mount point carries all the vertical inertia loads, but has a 50mm overhang, so the tube is loaded in bending with a nice stress concentrator at the maximum bending stress point. If I could just get all the tubes to meet at the same point on the bottom engine mount it would reduce stress considerably. Well it works. The Bing carb is a very tight fit. In fact I had to file away the cable guide fitting to clear one of the tubes. Don’t know if an Aerocarb would fit. I pulled the old mount out, removed the gear legs (no easy matter as I had glued them in with PR1422) and built a jig. I used what I could of the old mount and tack-welded the thing together. A very skilled tradesman where I work, John Black, who is capable of making almost anything tig-welded it for me.
This shows another view of the engine mount as well as some other details. The air cleaner box is the one that came with the engine, although I had to make a new aluminium lid to get it into the space available. Aircraft Spruce sell a range of spun aluminium bell mouths, one of which is hiding inside the scat tube to the carby. The scat tube was replaced with a smooth rubber bend before first flight. Just above the air box you can just make out the mil-spec plugs and sockets that were used for all electrical connections through the firewall. I mounted them high to get them away from feet, but once I fitted the fuel tank access on the inside was almost impossible. Should have put them lower.
Cooling is by a fairly conventional plenum chamber rather than the Sonex ducts. I was hoping for more even cooling from the arrangement, but the right side heads all run hotter than the left. Temperatures are within range but I’m still working on it. I tried to fit the fibreglass Jabiru ducts but found it almost impossible to get them into the Sonex cowling. The cowling has caused me more grief than all the rest of the aircraft. What good can you say of a design that is specifically for the Jabiru engine but won’t fit until large lumps are cut off the engine?
April 2011 First engine run. You might notice that I didn’t trust the brakes, but I will say a good word for the brakes. I have the cheapest set-up, although I have made several modifications, but they have proved surprisingly effective.

PA27.JPGInterior, controls and instruments. If I did this again I would make a vertical instrument panel. As it is the Stratomaster is not at the optimum viewing angle. The throttle block is much modified from a Vans RV3 or 4 – black throttle, red trim. Two brake levers in the middle allow differential braking. I usually reach across and use them left handed. Surprisingly it’s not too awkward. Photos below show the motorcycle cable adjusters attached to the floor to firewall gusset and the lever mount in the centre floor. The forward corner of the lever is square to act as a forward stop against the floor brace tube.
cable adjusters.JPGBrake levers.JPG

The straps along the side panels were intended just to hold maps, but I stick all sorts of stuff in there. They work well.
Here’s another view of the interior side panel, made from the thinnest sheet of plywood I could buy and padded with a very thin layer of foam under the vinyl covering. The panels have flaps which velcro to the seat base and back to prevent loose items from finding their way under the seat where they might jam up the stuff shown below.
Earlier I mentioned that the elevator horn was double-sided to eliminate twisting forces on the horn (and reduce bending loads on the bolt). Here’s a similar idea at the control stick end of the elevator push-rod.
…and also at the elevator idler. You might also notice that I made the push rod ends a different way. Instead of welding a bolt into the end of the tube and fitting a female threaded rod-end, I used female-threaded spigots welded to the tubes and fitted male-threaded rod-ends. It’s not necessarily better but required less skill at welding so was easier for me to make.
January 2012 The trip to Kyneton airfield
The paint is really only a 10 metre job, but I didn’t want to load it down. Empty weight is 301 kg but it will gain a bit as I fit spats and leg fairings and has already gained a bigger fuel pump.
11th April 2012 First flight with test pilot Aub Coote, a great guy, and I would suggest a pretty brave one as well. Aub made flying the Sonex look very easy. Incidentally, my own Sonex was the first one I ever saw fly. I had even been to a fly-in at Yarrawonga where they were all firmly planted on the ground while I was there, and whenever I visited Gary Wright he was replacing yet another ignition coil on his Aerovee. I recall being very impressed with the way it climbed, and relieved that it wasn’t too loud despite the straight-out exhaust.
12th April 2012 The smile of someone who has just made his first solo flight in an aeroplane he built in his back yard.


One of the modifications I have made since I started test flying is to fit adjustable flap stops. On my first flight I noticed a tendency to roll right. I tried adjusting the right flap to give more lift, 1/2 turn out on the rod-end, eventually going to 1 1/2 turns out. The trim was improved but didn’t feel "right" and when I pulled the flaps on it rolled quite strongly to the left. I guessed that the problem was torsional deflection of the flap drive shaft. I know the plans say to fit trim strips which act as flap stops, but how do you get them right in the first place? I machined these boss and flange parts from Lubron and fitted 8mm nylon screws. There are 2 riveted to each rear spar, although I probably could have fitted just one on the right side. The effect was noticeable immediately. The aircraft felt much more stable and can be trimmed to fly hands-off.
For more information on this visit my web site.

Fitted December 2012 – Pneumatic tail wheel
This is a Fallshaw 200 x 50 wheel and tyre. It has been machined to make it slightly narrower because the two halves did not fit together correctly. I have also fitted a sleeve between the bearings so that the axle bolt can be fully tightened. It is noticeably quieter when taxying. Rudder pedal forces on rollout after landing are higher because tyre grip is better than with small hard wheels, which tend to skate. My preference would be for a slightly smaller tyre fitted with a foam rubber inner tube, as are used in some wheelchair tyres, but this seems to work OK. I did eventually go for a smaller lighter pneumatic tail-wheel. See details at Pete’s pneumatic tail-wheel.
My engine originally came with a fairly small automotive transmission oil cooler. I optimistically assumed it would be up to the job, but even in cool weather the oil temperature ran to well over 100°C. I bought this one from Jabiru. There are probably cheaper suppliers, but not ones that are prepared to give you time and free advice. It is mounted to the back of the duct and works well.
In a spin
The propeller is a Bolly Duralite ground-adjustable type. It looks great and seems to work well while supplying a wide range of confusion, but doesn’t have a hole through the centre so the standard spinner won’t fit. The prop came with a glass spinner, but that doesn’t match with the Sonex cowling. I eventually got around to machining up a little "skull cap" type.
Hand grip
I have read of a few people having trouble fitting hand grips, but the trick is mostly just to push the grip on, which tends to expand the grip, rather than pulling it on, which tends to contract the grip on the stick. My grip is just a motorcycle throttle grip. The Sonex control stick is 1" dia. Most motorcycle handle bars are 7/8", but the throttle side is bigger to fit the throttle twist tube. I think Harleys use 1" handle bars so a left hand Harley grip should also be suitable.
The buttons are red PTT and black for remote memory.

How (not) to fit the Sonex fuel level probe
When I chose a fuel level gauge for my Sonex I opted for the KISS principle and went for the clear tube sight gauge. It was simple, pretty foolproof and cheap. Looking at it a bit more carefully, it was obvious that the Sonex design didn’t cover much more than the top half of the tank, completely missing the most interesting part of the story, so I extended it a bit further down, but even with my extension, the fuel disappeared from view at a little under 20 litres, missing the really exciting bits. I must be something of a worry-wart because I found that once the fuel went below 20 litres, even if I was only doing a couple of circuits, I started to worry about running out of fuel. It was time to bite the bullet, prise open my wallet and go electronic. Here’s how it’s gone so far:
13/10/12 – Order the fuel level probe plus a couple of temp senders from Sonex – $175
16/10/12 – They want to charge me $145 for UPS freight!
This is followed by a flurry of emails – Can’t you send it by US Post? Are you sure ? It might get lost. They don’t lose my stuff from Aircraft Spruce. It won’t be insured. Let it be on your head! Until:
27/10/12 – I receive a receipt for payment including $40 post. To my lopsided way of thinking, it looked like they wanted to charge me over $80 to insure a parcel worth $175 !
Nov / 12 – Receipt of my uninsured fuel probe.
I didn’t fit the probe for a while because of a number of factors including time pressure at work and the failure of the alternator on my Jabiru engine, which required removal from the airframe so the alternator could be replaced. It wasn’t until February 2013 that I finally got around to fitting the probe, a job that obviously involves draining the fuel tank, something that is not entirely free of risk and hassle. For a start I only had enough Jerry cans to hold about 25 litres and couldn’t see the point of buying more just for one job, something I was to change my mind about later.
So, I drain the fuel tank (abbreviated to IDTFT – you’ll see why) and fit the probe which:- DOESN’T WORK.
In fact, I wasn’t doing the Stratomaster calibration properly, but it didn’t matter because the output from the probe didn’t change with fuel level. Hmm. After much puzzling and fiddling:
10/3/13 – I email Sonex.
11/3/13 – Kerry Fores tells me to contact Princeton Electronics who make the probes. In football terms it’s the old flick pass.
13/3/13 – I email Princeton who:
-don’t answer.
19/3/13 – I email Princeton, this time it’s more of a grumblegram.
20/3/13 – Sherry from Princeton replies. They sound helpful. I can either send it back or buy a new one and they’ll do a swap.
21/3/13 – I email Princeton. I’ll send it back.
22/3/13 – Sherry from Princeton – they are working on a field fix. I might not have to remove the probe.
5/4/13 – I email Princeton. Um, any news.
7/4/13 – Sherry from Princeton – haven’t you got it yet?
9/4/13 – I email Princeton. Not yet – what is the nature of the fix?
No reply!
12/4/13 – I email Princeton. Can you give me more info?
13/4/13 – Michael from Princeton – Todd and Sherry are at Sun and Fun air show. The fix a is a little black box to reprogram the probe.
12/4/13 – I email Princeton – Thanks Michael. Don’t ask me how I sent my reply before Michael sent his email, but those are the dates of the emails!
17/4/13 – Sherry from Princeton – We are back from Sun’n’Fun.
16/4/13 – I email Princeton. Still no fix. Can you check address?
17/4/13 – Sherry from Princeton – Address looks OK. We’ll make you a new one.
27/4/13 – The repair box arrives – O frabjous day calloo callay.
12/5/13 – I email Princeton. "It doesn’t work" plus some questions to clarify things. I’ll send it all back.
14/5/13 – Sherry from Princeton – Can you send us a picture of circuit board?
14/5/13 – I email Princeton – picture attached.
16/5/13 – Sherry from Princeton – We’ll send you a new probe.
Late May – the new probe arrives!
8/6/13 – IDTFT and fit the new probe and try adding a few litres of fuel.
IT WORKS – Cue the Hallelujah Chorus and calibrate the Stratomaster.
8/6/13 – I email Princeton – Thanks for your help.
BUT, the next time I take out the plane there is a strong fuel smell in the cockpit. I now have a fuel leak to accompany my fuel gauge. I curse myself. I had used a gasket cement on the threads and it obviously wasn’t suitable.
IDTFT, remove the probe and refit it with teflon tape on the threads, which had worked fine on all the other fittings. I try a few litres in the tank and no leaks. I fill the tank and go for a flight. It leaks!
Hmmm, this is getting tricky. Maybe fuel is leaking around the outside of the brass fitting that is moulded into the tank.
IDTFT, remove the probe and refit it with PR1422 sealant on the threads. PR1422 is a rubbery putty for sealing fuel tanks. Just to be sure, I bog all around the fitting and the bottom of the tank. I try a few litres in the tank and no leaks. I fill the tank and go for a flight. It leaks!
Maybe the probe itself is leaking.
IDTFT, and use 1422 to bog around the wire that exits the probe. I try a few litres in the tank and no leaks. I fill the tank and go for a flight. It leaks!
I really need to know exactly where the fuel is leaking, not that I haven’t tried to pick this before. I have a pair of nifty binocular microscope glasses that have a focal length of about 150mm, so I do my demented contortionist act and climb under the instrument panel to find that there is fuel coming from the 1422 bog. My patch hasn’t worked. I pick off the rubbery sealant, but still can’t tell exactly where it’s leaking. It looks like the thread but the leak is so slow that the only visual indicator is the surface gradually getting wet. It could still be leaking around the outside of the brass insert. Here’s my current theory: I’ve had three goes at sealing the thread. One of them should have worked so it must be leaking around the brass insert, but it’s not until the fuel starts sloshing against the probe and wiggling it slightly that the leak shows up. Looks like I did a lousy job of bogging around the fitting – must have left a gap. I think that the probe needs a bit of support so I machine a thin-walled top hat shaped bush and cut it in halves so that it can fit around the probe. The idea is that the flange will sit flush against the tank and supply some support for the probe.
IDTFT (of course), coat everything with PR1422 again, coat the inside of the "bush" with the same goop and then assemble it with a hose clamp to hold everything in place.
25/7/13 – OK, I know it’s not beautiful, but after nearly 6 months of time-wasting misery, IT FINALLY WORKS (so far).
8/8/13 – Up to the 75 hour service. The aircraft is flying well but there always seems to be something to do. I still haven’t got around to painting the spats and had a recent oops with the canopy that resulted in a distorted rear hoop. I tried to straighten it on the spot and you can guess the result of that. Two cracks have been stop-drilled and another one has showed up, but at least the canopy can be closed. I have started fiddling with a redesigned frame that I hope will solve most of the existing problems.

Cool it – Keeping the Jabiru 3300 cool

In engineering, there are usually many ways of doing things successfully, so I won’t claim that what I have done is the only way of doing it, but the method described here does work well, despite the fact that my engine is one of the earlier "thick fin" types that don’t have as much cooling area as the later types, and the Bolly ground-adjustable prop I have fitted has very little cooling effect on the ground because the blades of the prop don’t pass over much of the cooling air intakes.

To make some sense out of the following description, you need to know the cylinder layout of the Jabiru engine. It is as follows:
The right side air intake is not as effective as the left side for several reasons. The clockwise rotation of the propeller imparts a downward motion on the air entering the right side intake, so the air has to do about a 90° turn as it enters. The #1 cylinder is so close to the front that it tends to block the intake, especially if a shield is used in front of the head to prevent it from running much cooler than the other heads on that side. The intakes on the Sonex cowling are set much lower than on Jabiru cowlings, which exacerbates the problem. The result is that cylinder heads #3 and #5, the back two on the right hand side, run hotter than all the other heads. Builders who use the Sonex-designed cooling ducts sometimes try to fix this by fitting a scat tube to duct air from the left to the right. I decided to build a conventional plenum chamber, hoping that the air flows through the heads would even themselves out, and here it is.
The scat tube across the top of the engine and the scoop behind head #3 (middle left in the photo) will give you a clue that I wasn’t entirely successful, but those are recent experimental additions.

The next part of the cooling equation is getting the air to all go past the cylinders and heads. Jabiru recommend fitting deflectors between the cylinders to stop air going through the big gaps there. They would probably be best fitted on the underside, but I have fitted them to the top as per their instructions. I have not gone to any special effort other than trying to keep the walls of the plenum chamber tight around the engine.

The last part in getting good cooling flow is having large outlets to help draw air through the engine. I made them BIG by cutting just the front and sides of the cut-out and folding the "tab" in and then riveting on aluminium side panels.

I also fitted pretty big dams to the front of the air outlets.

It all worked pretty well. The deflectors in front of cylinders #1 and #2 started off big and were trimmed away until they were too small and have since grown bigger again, but other than those minor changes, this was the set-up for the first 87 hours. Cylinders 1,2 and 6 always ran cool and 3, 4 and 5 ran hotter, but well within temperature range, with #3 always the hottest. They even-out a bit during cruise, but I did a lot of flying this winter and was starting to notice that the cooler cylinders were almost too cool, and #3 always being hotter was starting to annoy me.
The current arrangement is that shown in the first photograph; a scoop behind #3 and a scat tube to divert air from the LHS to #5. The angle of the outlet dams has also been reduced to cut air flow and (hopefully) reduce drag. The modification was pretty rough-and-ready but works well. Temperatures are now more even. On a recent flight they were:
Ground air: 24°C 1650 ft ASL
Cruise air: 14°C 4500 ft ASL
Cylinder head temp. range: 115° to 130°C

Coming soon:
The intake to the scat tube is temporary. I intend making a larger, more structurally sound intake to divert more air to #5 and add a small scoop behind #4, which should pinch a bit of air from #6 which still runs a little cool.
Instruments and backup GPS.JPG
Here’s a recent photo (17 October 2015) of my instruments and backup GPS. Maximum CHT shown is 126°C.

Ground running:
While it all works well in the air, on the ground it’s a different story. By the time the oil temperature is up to 50° C, the minimum temperature for using full power, the hottest cylinder head is getting close to 150°C. If the heads get too hot before the oil is up to temperature I shut down for a few minutes. During that time, the head temps drop, the oil soaks up heat from the engine and warms up, and it’s ready to do the run-up. If I am travelling anywhere, it’s not even inconvenient. The engine warms up on the taxi to the fuel bowser and the oil temp soaks up while I am re-fuelling. I don’t always have to shut down. In fact, since I changed propellers it is pretty rare. The prop I am using now has a bit more fan passing over the intakes.

100 Hourly – Oct 2013

Just completed the 100 hourly and fixed up a couple of things that had been annoying me:
– Made new sintered bronze bushes for the control stick pivot to eliminate a bit of play in the elevator control.
– Replaced one of the brake cables using a different fitting on the end which won’t slip on the cable.
– Tidied up the intake duct that diverts air to #5 head as mentioned above, and here it is:
The temps are pretty even when flying but it will still overheat on the ground.
– The main hassle from doing the 100 hour service was finding out that my prop wasn’t tracking accurately. The blades were at least 6mm different. I tried various combinations with the same results and then returned the prop to Bolly who couldn’t replicate my problem. Bum! Unfortunately, that didn’t make the problem go away. I was starting to look at alternative propellers when I thought to calculate the thickness of shims needed at the hub to correct the problem and it turned out to be only 0.016" (0.4 mm) which didn’t sound to bad, so that’s the course I’ve taken. The prop hub only contacts the drive plate at the bolts so I made stacks of shim washers for packing the hub and also used the same washers on the opposite side bolts as counterweights. Appears to have solved the problem.

115 Hours – Bolly Propeller Blues

My problems with the prop have not gone away. It appears that the composite hub is too soft and creeps under load. The prop tracks OK until I torque the bolts. Bolly are not interested because they don’t make this model anymore (despite the fact that it appears on their web site). I haven’t completely given up on the Bolly yet. I may eventually make my own aluminium hub, but at this stage I’m thinking of a more reliable replacement.

125 Hours – Prince Propeller

My new prop is a Prince P-tip, 54 x 60. It has a wooden core and carbon fiber coating.
I have now done about 8 hours flying with it and recorded the following performance figures:

Static full throttle revs 3000 RPM
Climb after takeoff 3000 to 3050 RPM
Full throttle level flight (3 runs) 3380 RPM 3360 RPM 3260 RPM (the last on a slightly cooler day)
On cruise power, about 21 litres/hour fuel flow and about 125 knots TAS, engine speed is over 2900 RPM.

The prop looks pretty schmick and performance is OK, but I think the pitch is a bit fine. Jabiru recommended maximum engine speed is 3300 RPM and suggest a cruise engine speed of 2750 RPM. Climb performance is good but that was hardly a weak point with the Bolly which achieved equivalent performance but at lower engine speed. I conveyed the above information to Lonnie Prince who has offered to replace the prop with a 54 x 62, which I think would be about right.

Luggage bay

Made this quite a while ago but hadn’t taken any photos. It is a roughly rectangular bag sewn from fairly tough-looking curtain material. Canvas would probably be ideal but this works OK. The floor is a sheet of corrugated plastic (it’s called "Coreflute" and is available in a couple of different thicknesses from some hardware stores) – the sort of material that some temporary advertising signs are made from. To prevent the bag sagging onto the elevator push-rod there is a .5" x .5" aluminium support tube running for-aft in the middle, attached to aluminium vertical straps. The bag is attached to the top longerons and top cross ties by aluminium rivets through fabricated aluminium hooks. The bungee net on the top is cut from an elastic trailer net. I pack any heavy objects over the support strap. Have carried pretty close to the maximum load with no problems, though objects tend to get tangled in the net. Being light weight is not always convenient.

Cabin ventilation

It has been a pretty common complaint that the Sonex cabin ventilation ducts are ineffective. Many builders fit commercial NACA ducts and eyeball vents in their instrument panel but I was rather reluctant to give up any space on the tiny panel and by the time I thought of it I had already cut out the standard Sonex shape intake and it didn’t look like any of the commercial intakes was going to fit. I made the following mod to the Sonex design – a 6mm screw to open and close the vent, all made out of aluminium and nylon so it’s pretty light. Only problem is that it’s also ineffective.
Actually, there are a couple of problems. A standard M6 thread is only 1 mm pitch so after winding the knob at least 20 turns to open the vent, hardly any air comes in. On the left side the knob also got in the way of the flap lever so I had to remove the screw and just gaffer-taped the flap open or closed depending on the weather. I have finally got around to trying a new duct. I guessed that if the NACA duct had proper side walls it would be more effective, and if the floor of the duct was adjustable, as in the original Sonex design, any drag from the duct could be eliminated when the duct was closed. I drew the duct in Solidworks and printed the whole thing on a 3D printer, including the 10 x 6mm 2-start thread and nut. It was pretty obvious from the print that it could be improved on but after a 6 hour print time I thought I may as well try it out as is.
I was hoping for an improvement in flow, but was really surprised at how much better it was. On a clear sunny day with temperatures in the mid twenties I found I was getting cold and had to close the vent down. The prototype is made from PLA which is hard but fairly brittle plastic. My next step is to sort out the printing process with ABS or preferably nylon. For details on the latest version, visit my web site. (Note that the link is to a commercial site)

Air intake box

All this stuff is pretty old hat but I have been reading quite a lot about the trials and tribulations of fitting Aerocarbs and since the Bing seems to work fine, I can’t see much advantage in changing. You do have to have carby heat for the Bing, but I was able to use the stock air cleaner box supplied with the engine. The box has a very good hot/cold air valve on the intake. I had to fabricate an aluminium lid to fit in the tight space behind the 3300 engine. Cold air to the box is fed from a NACA duct on the cowl which plugs into the rubber pipe when the cowl is fitted. A scat tube from the heat muff is normally fitted to the other intake.
Airbox intake.jpg
A bell-mouth from the air-box feeds a smooth rubber bend to the Bing.
Airbox to carb.jpg
Looking from below. The small red hose into the air box is the altitude compensation input to the carb.
Airbox to carb viewed from below.jpg

Rear canopy support

I didn’t like the idea of the rear of the canopy not having any support (unlike the front) so I folded a couple of thin angles from .025" aluminium and fluted them to match the curve of the turtle deck. It is offset from the turtle deck skin by the canopy thickness. I think it works well to match the shape of the canopy to the turtle deck. The top is coated with a thin layer of PR compound in an attempt to make a seal, but it would probably be better if the angle was offset slightly more and a rubber strip was glued to the top.
Turtledeck trim.jpg

Latest Photos – Dec 6, 2015

This was a great day for me. Dorian, my eldest grandson had been for ground rides in the Sonex but he finally decided he was ready to go flying.
Dorian ride.JPG
We did a 20 minute flight around Mount Macedon, a bad choice as it turned out because it was pretty rough in the lee of the mountain, but he loved it anyway. The wind was fairly strong at altitude and we managed a maximum ground speed of 155 knots while cruising.
wheeler 3 pointer.JPG

Here’s an idea I’m working on – A light-weight quick lift jack. It needs refinement.
Quick lift jack.jpg
I hoped this would be the final arrangement – it works great, so long as the tyre is pumped up, but was much harder with a flat tyre, and when are you most likely to need it? I had thought of this of course but had hoped that this was a compromise size which would do everything. It would work with a longer lever, but then it wouldn’t be a lightweight tool that you could toss in the toolkit.
Took my mother for a sight-seeing flight on 1/3/16 – see the GPS track above. Mum is 91. She was unfazed by a pretty rough ride in the warm conditions, but I’m sure she appreciated the smoother conditions over water. That track is 320 km which is quite a long sight-seeing trip.
As the old saying goes – If at first you don’t succeed, try,try try again. Here’s my third try at a quick-lift jack.
Quick lift jack in position.JPG
I’ve just changed the proportions a bit. It doesn’t lift as high but it’s much easier to operate.
Quick lift jack operated.JPG
It can be used to lift the wheel even with a completely flat tyre, although that takes a bit more effort. Mass is just over 0.5 Kg, so it’s light enough to carry in the tool kit.
Here’s my FINAL version of the wheel jack. It works by lifting the lever instead of pushing it down. Max force is now when the lever is near horizontal so you can react against the ground – easier to operate. You need to leave the other wheel free to move until the wheel being lifted is jacked up and then fit chocks to the other wheel. Details are on my webite including a full set of plans in case you’d like to make your own. This one folds flat, takes up very little space and weighs only 420 grams, or less than 15 ounces in old money.