Latest News
MERCURY UPDATE
Story added 19th January 2010 by Fred
Update on Mercury 20 engine rebuild
and oh yes, the Kestrel.
January 2010
By Steve McManus.
Since my last notes on the Mercury 20 rebuild you would justifiably expect it to be well ahead by now and in truth significant progress has been made, but not as much as there could have been. This is because when the Kestrel in our Hind tried to wet the engineers heads as it was being turned over during oil priming prior to an airshow, my efforts were diverted. This diversion was at first part time (I only do three days a week anyway), but since the engine in the 504 also died giving Phil who was already working on the Bristol monoplane engine as well as the kestrel, two rotaries to repair, it became virtually full time. Who said men could not do more than one thing at a time!
The leak on the Kestrel was from the seal at the top of the cylinder liner on A block number 3 cylinder allowing coolant into the cylinder. These seals which were originally soft aluminium alloy rings are subjected to firing gas pressure and temperature applied to their inner surface and coolant on their outer surface. When the engine was overhauled in 2001/03 the blocks containing the seals and liners were the only parts not overhauled in house. They were sent to an external engine rebuilder for overhaul and to have an approved modification to the seals then being carried out on a number of Kestrels on rebuild, applied to our blocks. This modification involved replacing the aluminium rings with metal O rings that were used in the nuclear industry. There were advantages in that the accuracy of each liner/seal assembly length was less critical as the seals were much more compliant than the alloy rings. This compliance was also considered to deal with expansion stresses better than a solid ring.
Unfortunately the rings have proved less durable than the alloy rings they replaced. Two rings burned through on two separate aircraft in two weeks last year. One being our Hind, the other aircraft being the second Nimrod rebuilt and owned by HAC. The failure to the Nimrod occurred at Yeovilton during a display (the aircraft landed without damage). It seems that the very small contact area made by the rings to the surfaces they seal to and the lower conductivity of their material leads to overheating and burning. In addition, the support given by the O ring assembly does not provide good location to the top of the liner to resist piston side loads.
Because of the above, as soon as the failure was fully defined, the decision was made to return to the alloy rings as these tended to last 100 hours plus on Kestrels especially if the boost was limited to a maximum of 3 psi. The rings were after all also used on the single piece block Merlins and though failures did occur they were used up to 12 psi boost, but were in trouble if boost was increased above this. The big difference was that on the Merlin and for that matter very late Kestrels 12 extra studs were added at the liner sides to prevent piston side loads opening up the seal joints.
Once the decision was made to revert to the original seals our problems started as after the war Rolls Royce destroyed all the pre-war car drawings along with Kestrel drawings that happened to have the same prefix. We, therefore, could not define the seal’s material, its precise shape and finish or its dimensions. An old seal was analysed for chemical composition and we looked for modern equivalents. We bought a billet of the nearest equivalent material which is only supplied in the T6 hard condition. Experiments were then carried out on a dummy single cylinder tooling replicating the seal area to determine the finish and fit required. It was found in many tests that the alloy rings had to be in the annealed T0/1 condition, that the surface finish had to lapped to have perfect circular lay and no radial scratches even very light could be tolerated on any ring or sealing surface. Having defined this we moved to the block itself where each seal thickness had to be individually tailored to give each liner the same projection from the bottom of the block to within 0.001 inch so that all seals are equally loaded. The base flanges of the liners sit on the flat crankcase upper deck and are clamped to it by the block retaining studs.
All this took time and we only completed the first block that passed pressure test by the end of November and this is now back on the engine. Next we moved to B block as this was no longer considered as safe and on strip another seal B5 was found to be almost burned through. This block is now under conversion back to the original seals and it is hoped that it will not take anything like as long as the first. The final part of the rebuild is the paperwork as a minor modification has to be submitted to the CAA for the new material used for the alloy seals.
Getting back to the Mercury, reasonable progress had been made before we were diverted. Continuing the tale from the February 2009 notes, the repaired main bearing was received from SPC bearings in March 2009. With this essential part back the crankshaft could be assembled complete with the big end and connecting rods. This was done in the jig that Dave manufactured for us initially to strip the shaft. It accurately aligns the two halves of the shaft, sets the end float of the big end and importantly allows the Maneton bolt to be loosened and tightened. This is the clamp bolt that holds the two halves of the shaft together and is about 1.25 inch thick. When tightened it has to be stretched 0.0011 to 0.013 inch and this took two engineers on four foot levers on the nut. The jig was bolted to the Merlin engine stand complete with Merlin and two engineers were required to stop the stand spinning round in the workshop (the stand with Merlin weighed about a ton).
The engine mounting ring was fitted to the assembly stand and the rear main casing was assembled to the mounting ring. The crankshaft assembly was then installed first in the rear then the front main bearing roller bearings outer sections in crankcase halves. The halves were bolted together and the end clearances checked. The cam ring was fitted to the crankshaft nose. The front casing was then installed, bolted to the front crankcase half and then the cam ring was timed. The crankshaft end thrust bearing was installed in the front casing and the whole assembly was then locked with the retaining nut and this was in turn locked. The tappets were installed using new gits type seals manufactured to the original pattern by Martins Rubber in
The supercharger was reassembled using overhauled ball bearings and the drive gears were assembled into the front of the supercharger casing. These gears increase the supercharger running speed to about nine times above that of the crankshaft and have three centrifugal slipper clutches to prevent gear overloads during rapid accelerations. Before the supercharger could be fitted to the crankcase the spring drive had to installed on the crankshaft tail and then the supercharger was bolted to the rear of the crankcase. The cylinder outlet elbows were then bolted to the outlets from the supercharger and the carburettor inlet elbow installed on the base of the supercharger.
During the time I have been working on the Kestrel, Barney has taken on the long job of overhauling nine sets of rocker gear. These assemblies are more complex than one might think. They have to be completely stripped to clean all parts and to clear the grease passages, then the eight small section ball bearing races in each head assembly have to be checked for corrosion and damage. The rocker shafts have to be measured for fits, the adjusters cleaned and the six chrome plated buttons (four for valve contact and two for push rod contact) checked for wear and plating damage. Quite a few parts had to be replaced from our stocks of spares obtained from Strathallen with the Lysander. As part of this work the push rod assemblies have also been overhauled and these as with the rockers are not as simple as one would think. They consist of a tie rod that pivots the whole rocker assembly to maintain rocker to valve clearance during the thermal expansion of the cylinder. The push rods themselves are then mounted though flanges on the tie rod with extra return springs. Unlike the core of the engine wear was a problem on the push rods and there mounting holes in the tie rod flanges. More spares were required to complete the nine sets.
Jim has been helping with the carburettor overhaul by checking fits and clearances as it was stripped. The bottom half of the carburettor has been stripped measured and repaired. The only part needing repair was a push rod on the boost control capsule that had actually come from the Lysander carburettor when the part from my carburettor was taken to repair it. The Lysander’s worn part required chrome plating and grinding back to size. All other parts of the carburettor so far have like the engine core been in new clearances. The bottom half of the carburettor is now ready for re-assembly and work will continue with this and the top half will be stripped when I get time. Thanks must also go to Toby the top overhaul king and Rory for their help with paint spraying numerous parts. In his very occasional spare time from keeping the fleet airworthy, Toby is also tackling the ignition harness which will need completely re-wiring and some repairs to the conduits.
The reduction gear has been partly stripped, but awaits a puller to be made to continue the work. All parts so far removed from this have proved to be in excellent condition and in new clearances. The cylinders still await attention.