For Everyone

For SABC Members

Report On Refurbishment Of Thruster Gemini 25-0022

In November 2001, my wife and I took delivery of a 1986 Thruster Gemini. A Condition Report, prepared at the time, indicated there were no serious problems with the aircraft and that it was safe to fly. A subsequent major inspection by me however, resulted in some 10 months part-time repair work, with the first flight on 16th September 2002. Some of the problems were quite serious as the following report will detail and the reader will understand why I have a major problem with the way in which an ultralight aircraft with maintenance problems can slip through the system to perhaps an unsuspecting new owner. Fortunately, I had previous experience with Thrusters, the help of Brian Mouchemore (an 'eagle-eyed' fellow ultralight pilot, and advice from the Thruster Support Group; and so was able to carry out an effective major inspection.

Firstly however, it needs to be said that the ultralight movement is working well to provide low cost aviation to those who could not otherwise afford it. It has been some 37 years since I gained my GA license and first dreamed of owing my own aircraft. Unfortunately, with family expenses and other pressures, my wife and I could never afford to buy or build an aircraft and only just managed to keep the GA license current. It was in 1992 when we first saw a Thruster ultralight aircraft at a club north of Perth WA: An ultralight conversion and an instructor rating quickly followed. In the mid '90's we managed to pick up a single seat Chinook cheaply which we refurbished. We made enough profit out of that project to purchase the Gemini in this report. The cost was only $9,000 but we still only just scraped together the dollars needed to finally buy our own two-seat ultralight. We were never going to be a able to afford the $30,000 odd for the new kits on the market but our experience shows that the Australian ultralight movement does indeed provide the opportunity for the ordinary person to own their own part of aviation; and I think perhaps, this is what the movement originally started out to do. We only have a R503 motor, pull start and a bottom of the range model; but it provides us the type of flying we want and we couldn't be happier.

The following is a list of the problems I found in the inspection; and repairs made. Despite the problems I am most happy with the aircraft and had expected to find some work to do on it anyway.

Engine

The first problem I found was what I thought were worn rubber mounts on the engine mount: The engine, when rocked up and down, was actually hitting the front of the spade mount. The first 'fix' was to source and replace the rubbers. When the new ones came out the same as the old ones, and I found out there was only one 'hardness' in that type, I started asking questions. Tony Hayes from the Thruster Support Group (TOSG) confirmed I had the wrong type and size and sent me photographs of the right ones. In providing this information Tony made mention of the replacement order of the 3/8" engine mount bolts, washers, nuts etc. That was all well and good until I realized Tony had said 3/8"; and my mounts had ¼" bolts! That really started the alarm bells ringing. With an exchange of photographs with Tony we finally chased down the problem. Firstly, the engine mount cross-bars were obviously home made & designed and were totally different from original specs, as were the rubber mounts. Secondly the 3/8" holes in the spade mount fitting had been sleeved out with ¼" copper tubing to take ¼" bolts, possibly because the wrong rubber mounts came with ¼" sleeves through them. The sleeves in the 3/8" holes were so tight that a cursory inspection did not pick them up.

Now, I had an aircraft manual and an engine manual but neither dealt with the actual attachment of the engine. If it had not been for the information from - and liaison with - the TOSG I would never have picked it up. In fact, the previous owner, and the one before him (who did a strip-down inspection), did not pick it up. The aircraft logbook shows an old 1991 entry - following an engine change - of 'new engine mounts and rubbers' being fitted: One assumes this is when the unauthorized modification took place. I am astounded, apart from the obvious illegalities, that someone would exchange four 3/8" bolts for ¼" on an engine mount. This issue will be followed up further with the AUF Technical Section.

The engine itself has 105 hours on it and has had full strip/check/overhaul by a workshop 40 hours ago, when the previous owner took delivery. It is my custom (as with the previous owner) to use Unleaded and Penrite TS40c as the preferred fuel mix.

A final problem surfaced after I had done an hour test flying. A friend dropped in and noticed the carburetor intake manifolds (the metal part between the Carby rubbers and the cylinder heads) were directed down instead of up (about a 10 deg. Bend). He had recently put a R503 on his aircraft and noticed the warning in the documentation that these units should face up. This was checked with Bert Flood, verified, and the manifolds turned round 180 degrees. Apparently they are able to go both ways according to if the engine is mounted right way up or inverted.

Propeller

The propeller on the aircraft had been changed from the standard 64x42 Cato to a 64x37 Cato sometime before my purchase. It seems to perform all right especially, as expected, in the take-off and climb. I would be interested in any comments on making a change like this.

Airframe

1. After removing the skins from the wing I found a marked bend in the starboard front main spar, just at the point of the jury strut attachment. There was a significant crease on the bottom surface indicating a lifting load in that area. As the spar gave differing thickness measurements around its circumference it is possible that the part was a non-standard item. Tony Hayes has commented further on this in his TOSG bulletin. A new part was sourced from WadeAir and a local yachting workshop was used to drill and position the wing fittings. The old spar, and comparative measurements from the port spar, was used to position the new holes accurately.

2. The front and rear external bracing wires from the lower pod to the wing jury strut fittings were loose and were tightened to TOSG suggestions.

3. The lower lift strut bolt holes were found to be elongated so were drilled and sleeved.

4. The Tailplane leading edges were rattling up and down some 3/16" on the fuselage attach holes. The nylon bushes were found to be worn and were replaced.

5. The tailplane lower bracing wires were very loose and were tightened to specs. The pins for these wires were worn through about 1/3 and were replaced.

6. The nylon spaces on the control column and at the lower attachments of the rear pod 'A' frame were broken and cracked and were replaced.

7. All stainless steel fittings throughout the airframe were checked with dye penetrate.

8. All bolts, nuts, hardware etc were checked and replaced where necessary.

Skins (Wing and Tail Surface Coverings)

All skins were taken to a local sail-maker for assessment and repair.

I was aware before I bought the aircraft that the wing skins appeared to be loose at the trailing edge, but was told it was not a problem and did not effect the performance. The fabric had a reflex just inside the trailing edge caused by - or in addition to - the top and lower wing battens touching in front of the rear tube: They were actually tied together in that position. I was alerted to this being a problem by the TOSG. It appeared that the strings were not for tying the battens together but simply to aid disassembly. It was also confirmed that the battens should not be touching but should sit just on the top and bottom surface of the rear spar as I suspected.

The question then arose as to whether the battens were too short or if a problem existed with the fabric. It turned out the fabric was in fact stretched slightly. I then had our local sail-maker put a ¼" (yes, measured very carefully!) tuck along the trailing edge of each wing skin. This worked very well indeed and the battens now sit exactly where they should. I then used the TOSG method for tightening the skins, which gave me tighter skins that I have ever previously been able to manage.

At one stage I decided to replace the aileron skins as the old ones were deteriorating at one end. I sourced these through the TOSG and installed them. Then I found a velcro problem in the join between the wing and aileron skins. Velcro of course comes in hook and loop surfaces. I ended up with loop on the wing skins and loop on the aileron skins. There must have been a change in hook/loop placement over the years. Remember to specify your hook/loop arrangement when ordering (if mixing different skin ages). I was able to buy a roll of 25mm wide sticky-back hook velcro which I cut in half (scissors are fine); then stuck both together to give a roll of 12.5mm double-sided hook velcro. This then went between the wing and aileron surfaces and worked very well. The extra thickness of the velcro actually fills in the gap better between wing and aileron and gives a better airflow.

I then inserted foam airfoil shapes (provided by Thruster) into the elevators and rudder. All tail surface skins were removed and retightened using new cable ties. A word of caution: The old cable ties were breaking off in my hands. All cable ties need to be replaced as a regular maintenance item.

Pod and Electrics

1. The windscreen was quite stained and cracked around the rivets and was replaced using the method described in one of the TOSG bulletins, to great effect. Missing rubber U-channels were installed around the various pod /tube openings.

2. All fuel lines were replaced, the fuel tank removed, cleaned and re-installed and the electric fuel pump replaced. The old pump - relegated to fuel container transfers - failed after two minutes!

3. The rectifier/regulator was an old original unit that needed a separate switch to be turned on and off according to battery voltage. I had also been warned about spike problems with this type. A new 3-phase AS Rectifier/Regulator was obtained from WadeAir and installed.

4. Some electrical connectors were breaking away from their wires. All electrical wiring was replaced. The original system had wires running everywhere, with individual fuse-holders, and all connected through a wire strip connector. I bought a fuse-board from Tandys where all fuses are in a row and plug in. A circuit diagram inserted into the Flight Manual now makes identification and changing of fuses easy. I used 50mm velcro to attach the fuse-board to the pod, making it easy to remove to inspect and work on. A separate switch was installed for an avionics circuit (Radio & Intercom).

5. There was no intercom with the aircraft so, after much research, an Intercom/2 Headset package was purchased from Raptor Designs in Queensland. This turned out to be an excellent choice and well worth the initial cost. The headsets are light, quite and fit very snuggly over the ears. There is not one bit of engine noise or static in intercom or radio mode. The radio and intercom are mounted at the top of the windscreen with the wires running down inside the center windscreen 5/8" tube.

6. New switches were installed with light diodes to show when on. Ignition switch was retained.

7. All pitot and static lines were replaced and re-routed properly. The pitot tube was repaired and the old biro extension replaced with an aluminum tube to specs.

8. The seat rails (also the rudder pedal supports) were removed and found to be bent from a recent heavy landing recorded in the logbook. The rails were repaired and re-installed.

9. A 'Bullet' recovery spring-operated parachute was retained from my Chinook when I sold it (it had originally come off a 2-seat Thruster). As an experiment it has been installed below the fuel tank and attached to the fuel tank supports which, according to Thruster Factory advice, should support the extra 9kg. Nevertheless, the existing fuel tank supports have been reinforced, again on suggestion of Thruster, by 5/8" aluminum tube brackets, which will show any first sign of stress.

Tailwheel

I needed to replace the tailwheel and was able to get an excellent solid wheel locally. When I did the first test flight I found a problem I haven't had before with Thrusters. When operating on grass surfaces, as I always used to do, they don't need brakes. But put them on sealed taxiways, as at the airfield where mine is now based, and they simply will not stop; even with one wheel off the edge. I have therefore modified a bike brake caliper to operate on the tailwheel, with the brake handle on the control stick. A bit of velcro strip holds the lever for a park-brake. It works very well for taxiing and run-up. This, like the parachute attachment, is still experimental.

C of G and Weight Calculations

You need to watch the weight on older Geminis (With original R503) that have been modified over the years. There is quite a bit of difference between the original Gemini and the later R582 models. My Gemini had been modified over the years with the installation of the heavy-duty cromemoly insert into the original aluminum axle tube, and with fiberglass repairs to the pod. The empty weight has changed from factory original of 160kgs to 186kg. While this is fine for my wife and me with full fuel, heavier crew would find a problem with the all up weight allowance.

The Gemini (mine is the short boom), of course, is critical for C of G as outlined in the TOSG bulletins. Fortunately, my C of G has turned out very well and is well within the prescribed range recommended by TOSG. It is actually better with the parachute fitted where it is (aft below the fuel tank) which probably balances somewhat the previous fiberglass repairs to the pod.

I made up a spreadsheet for the weight and balance on a Microsoft Excel program, which automatically calculates C of G for a number of various load factors. This is very helpful when operating something like the Gemini, which, as mentioned, is tight on weight & balance. I have included a copy with the email in case it is of assistance to anyone.

A word of caution: When weighing the aircraft the tail weight measurement needs to be very accurate. An error of 1kg will give an error or nearly 20mm in the C of G position. Also remember the aircraft has to be in the level position for the calculations.

Maintenance Records

One of the initial problems I found was that maintenance records were sort of in two logbooks with two engines and three propellors. Although there was an accurate continuous record it was very messy and difficult to get a true picture. I found the best way to deal with the situation was to set up a computer Exel Spreadsheet program with horizontal columns for parts of the aircraft and vertical columns for engine/airframe/propellor hours. I have recorded all past maintenance and 'maintenance-to-be-programed all on the same record.

Summary

The refurbishment work carried out was no more nor less than what I had expected with an older aircraft. It had the very great benefit of enabling me to become familiar with the aircraft and it's systems.

I cannot underestimate the value of having someone else involved in the work from time to time. Four eyes are better that two. It was my friend Brian that initially picked up the bend in the mainspar, and this was after the wing frame had been put back on the aircraft. He also picked up the very clever ¼'' copper sleeves in the engine mount. I looked for a sleeve and couldn't pick there was one (and even emailed Tony Hayes to say there wasn't a sleeve!). Brian looked at it, came to the same conclusion, but then grabbed a 3/8" drill bit and knocked out the non-existent sleeve.

Of even greater value was the support of Tony Hayes and the Thruster Support Group. Tony was very patient responding to many many phone calls and emails over the 10 month re-build period. Without the support group I would not have picked up the engine mount problem and many others items which would have let to future problems.

The test flight has gone very well indeed and we expect to have many hours of enjoyment with our 'new toy'.

Regards
Stewart Maddigan

NB - Stewart operates the 'Delta Recreational Flying Services' which operates from Murray Field for general training. Stewart is also available for SABC members at Serpentine in accordance with the SABC instructing approval.

back to Members Projects