March 2006 progress report.
Progress on the restoration of the Welland has not been as fast as had been hoped. This was for two reasons firstly the need to devote time to the restoration of our flood damaged home and secondly because the task of restoring the engine is bigger and more time consuming than was anticipated. Nonetheless good progress has been made during the year and the major components have been reassembled.
The corrosion damage to the compressor casing and the diffuser was repaired more easily than anticipated. One half of the diffuser assembly had to be dismantled so that the side cheeks could be repaired by welding. The corrosion on the inside of the compressor casing was chipped off with a small cold chisel. A cold chisel being the appropriate tool as the aged corrosion product of magnesium is dolomite of which mountains are made. With the corrosion removed the surfaces were redressed with a riffler file. All the inside surfaces were then cleaned up using emery flap discs and the whole given two thin coats of epoxy paint. With the external corrosion pits filled were with epoxy resin and the outside painted with black 2K paint the final result is quite pleasing.
One disappointment was found that after the impeller turbine assembly was assembled into the engine and turned at a few hundred RPM with a large electric drill that there was a significant rumble from the bearings. The need to completely strip the engine down again to enable the rear bearing to be replaced was too daunting at this stage and it was decided to investigate what could be done if the turbine was removed which would allow access to the bearing. Inspection of the inner and outer races of the bearing revealed a number of very fine indentations across the bearing tracks, these were polished out and the rumble although still present is much reduced. Clearly an early future task will be to completely replace this bearing. Obtaining a replacement will be interesting because whilst the major dimensions are simple imperial the inner diameter of the rear bearing is 2 inches plus 40 thou and the front bearing outer race forms part of a hemisphere.
Peter made a replica of the original end mounting plate for a second “new old stock” Derwent starter obtained from Barry Parkhouse's emporium as a replacement for the Derwent starter motor that had previously been obtained but which was found to have been well used. The reversing of the direction of rotation was a straight forward job once one got down to it. The final result is barely recognisably different to the original.
All that was necessary to refurbish the gear case was to change those bearings which had been corroded beyond recovery and to to remove some superficial rust. The aluminium castings have all been cleaned up as far as possible to a satin finish using Scotchbright products and protected from further corrosion with clear acrylic spray paint. This may not be to everybody's taste but it would be quite easy to reinstated the original black paint finish.
The dents in the aluminum oil tank were skillfully removed by one John Williams, a Londoner, who now lives and works in a remote glen in the Scottish borders.
A recent second visit to the National Archive in Kew has brought to light significant further information about the engine including an amazing photograph and periodic maintenance instructions especially prepared for engine No 183. These data, with much else, were found in a file containing exchanges of correspondence between the Ministry of Aircraft Production and the Canadian Cold Weather Test Station. where engine 183 was sent in January 1944. Shortly after the engine had arrived in Canada when it had only had a few hours running, and before serious testing had begun, the rear main bearing collapsed. There then followed a stream of correspondence between the UK and Canada on the possible causes of the collapse and how to proceed. The amazing photograph referred to is a picture of the remains of the collapsed bearing. It is extraordinary that such an apparently insignificant picture should have been preserved for over sixty years, but then the National Archive is itself an extraordinary asset freely available to anyone. The decision was taken that the engine could re rebuilt in Canada and the necessary spare parts were dispatched and the task completed by April 1944, unfortunately too late for any further testing as the winter had passed. The failed ball bearing was replaced by a roller bearing, other similar bearings having failed in two other early pre production engines during flight trials in October 1943. It would appear that by the time that cold weather trials could recommence during the winter of 1944/45 that trials of the Derwent I engine were given priority and the trials planned for the Welland discarded and the engine returned to Britain.
For some years I had been trying to obtain a copy of the maintenance manual for the US version of the W2B, the I16 designed, built and flying all in the period of one year by General Electric. A friend of fifty years living in the US on a cruise in the Mediterranean made the contact that resulted in a copy of the manual arriving by post. The comparison of the designs is most interesting and will form the basis of a future web page or two.
SEPTEMBER 2007 PROGRESS REPORT
Since the last progress report eighteen months ago the Welland has been completed and exhibited at ten air shows and other events and has accumulated a running time of about eighty minutes since overhaul. Seven events were attended during 2006 as a non running exhibit. The engine was taken to the “Flying Legends” event at Duxford which gave the opportunity of reuniting Steve Gotts, who had discovered the Welland in Duce's scrap yard in Cambridge almost forty years before, with the engine.
The overhauled engine was run for the first time on Sunday 22nd April 2007 almost exactly 63 years after it had last been run. The site chosen for the first run was adjacent to the M6 motorway as we were not sure how noisy it would be. The engine did not start at the first attempt because of a minor problem with the insulation of the high tension leads to the igniters. Once this problem was fixed, away she went in a spectacular burst of flame from the fuel remaining in the engine from the failed start. The whole event was recorded for broadcast by the BBC local television news programme “Look North” by Paul Paxton and was duly shown, later in the week, immediately after the lunch time and evening news.
Exhibiting the running engine is quite different to the Merlins and other engines in Aero Engines Carlisle's stable. As the engine is a turbine and I limit the rpm to about 4500, in the interests of minimising the stress on the impeller, the normal operation speed was 17500 rpm. The engine at this speed, does not make a great deal of noise and once the engine has started sometimes with a burst of flame there is a pretty low key event, no whirling propellers. As a result individual runs are limited to two to three minutes during which the engine is cycled between 3000 and 4500 RPM. The engine will “self sustain”, that is keep running without assistance from the starter, from about 3000 rpm, a rather lower speed than was anticipated. In view of the anxiety concerning the rear engine bearing a hand oil priming system has been installed. It is noticeable that there is a significant delay before the automatic system starts feeding oil to the rear bearing, the feed to the front bearing starting almost immediately the engine starts. These main bearing feeds can be observed through the “sight glasses” with which the engine is equipped. The oil pressure has been increased from 10 to 20 psi in the interests of improved lubrication. Originally the engine, in the absence of special gas turbine oils, was lubricated with a light machine oil diluted in a 50:50 mix with kerosene. As modern turbine oils are quite expensive I use a 50:50 mix of kerosene and a light multi grade oil meant for use in cars. The rumble from the bearings is monitored, after each run, by recording the sound in the jet pipe, when the engine is turned over by hand. There has been no detectable change in the volume or character of the sound.
The procedure for starting the engine is as follows:-
-
Ensure that the throttle is closed.
-
Open the high and low pressure fuel cocks ensuring that the emergency cut off switch is in the “run” position.
-
Switch on the igniters and the auxiliary fuel pump.
-
Start the engine turning using the hand starting gear.
-
Remove the starting handle and switch on the 12 volt supply to the starter.
-
When the engine reaches about 750 rpm switch over the starter supply to 24 volts with the throttle partly open.
-
When ignition occurs allow the engine speed to increase to about 3000 rpm, then switch off the auxiliary fuel pump and the electric starter and control the engine speed in the range 3000 to 4500 rpm with the throttle.
The igniters can be switched off once the engine has ignited but are normally kept running in case of a “flame out”. The engine is stopped by closing either of the high pressure fuel cocks.
The major problem remaining to be solved after the last progress report was the fabrication of a combustion chamber dome to replace one that was missing. A suitable basis for a replacement dome was a measuring bowl bought from the cookware department of the local branch of Dunelm Mill, for the princely sum of £3.99. Once the bowl had been cut down to the correct depth it was found to be just 1mm in 8.5 inches too large in diameter, this minor discrepancy was easily accommodated in the flange that Robin Byers fabricated to complete the dome. The welding of the comparatively thick flange to the thin metal of the dome did tax Robin's welding skills. The completed dome, which also needed a special stainless steel feature that Peter Grieve made, is virtually indistinguishable from one of the originals. It is one of the advantages, at least of early jet engines, that many of the components do not have to be made with high dimensional accuracy.
Once the engine was complete the next task was to commission the individual subsystems. The most troublesome subsystem was the fuel system which took two months to resolve. The basic problem turned out to be that the fuel pump was not capable of supplying sufficient fuel at a high enough pressure at starter motor RPM to trip the accumulator in the fuel system. In the end it was decided to take the approach that General Electric had with their I16 variant of the W2B engine and to fit an auxiliary fuel pump. In our case an electrically driven oil pump, part of the loading ramp equipment of a scrapped lorry. Having a copy of the very detailed GE I16 manual as well as the RAF's manual was crucial at this stage of the project.
Further research at the National Archive at Kew has revealed that Welland 183 was actually completed in December 1943 and not in January 1944 as was previously thought, it also showed that the engine was run for 5 hours during its acceptance testing. Study of the minutes and progress reports of the Gas Turbine Collaboration Committee (GTCC) reveal that Rover Car Co's contribution to the development of the jet engine was far greater than the majority of authors writing on the subject since the war have acknowledged. It certainly emphasises the importance of going back to primary sources when writing, rather than merely repeating what others have written. The papers also emphasise the importance of completing the development of a new device before putting it into production. The GTCC was set up by the Ministry of Aircraft Production (MAP) at the end of 1941, all the key British companies who were players at the time were represented on the Committee which played a key role in the development of the jet engine in the UK in the 1940s.
Having got the Welland running the project is complete and the problem now is to decide how often to run it and ultimately what to do with it. It is an historic engine and the oldest running jet engine in the world. .
DECEMBER 2008 PROGRESS REPORT .
Early in 2008 the starter motor gears were replaced by a new set of gears as the original replacements had not been hardened after machining and were noisy and showing signs of wear. As much material as was practical was left on the new gears and in addition they were hardened. During 2008 just seven shows were attended with a total running time of 48 minutes. There were a couple of problems with the engine. A starter solenoid needed replacing, which was quite straight forward. More interesting was a problem in getting the engine to start which got progressively worse as the season progressed. In the end it was discovered that the spark at the igniter had gradually got weaker. The problem was rectified by using ordinary WWII engine ignition boost coils. It is planned to replace the vibrator contacts in these coils with more modern electronics based on the use of a high power FET. An amazing piece of serendipity occurred when Peter Grieve was searching through a box of miscellaneous aircraft parts and found three incomplete and damaged Welland burners. From the three one complete burner and a number of useful spare parts were obtained. It was decided that the re-manufactured burner on the engine would not be replaced as it formed part of the engine's story. This amazing find took place in the village of Goosnargh about twenty miles from the Rolls Royce factory at Barnoldswick where the Wellands were built.
