January 2006

GDP 38, a 1929 20/25, barrel-sided, replica-bodied tourer by Horsfield is a delight to drive, particularly with the hood down. Her engine is sweet and sewing machine-like. Irritatingly, a cold start always generated a cloud of white smoke since I bought the car in 2004. No other car I have owned and certainly including my other two pre-war small hp Rolls-Royces have ever inflicted such embarrassment upon me. Afterall, GDP 38 benefited from a complete chassis and engine rebuild in the mid-nineties under the good custodianship of Ian Eccleston of Solihull and has not been very far since. My thoughts led me to a starting carburettor fault, but the entire carburettor had been fully restored by Jason Fox of Elliotts of Chesterfield a few months earlier.

Off with the Head or what ?

Several weeks ago, Jason looked at the smoke and concluded, "Valve seals are knackered. Whatever's in there has got to come out". As far as he was concerned, this meant the cylinder head had to come off and a new head gasket installed whilst we were at it. I had been anxious to replace the finned exhaust manifold for many months on account of a slight leak behind the middle outlet and the fact that, having been the victim of some minor welding in its time, it looked ugly and was not in keeping with the otherwise smart engine bay. It seemed sensible and more economic to do both jobs simultaneously. The sole supplier of such manifolds is Ristes of Nottingham, just 30 minutes away from home, but at every monthly enquiry, the expected batch had either not arrived or had been sent back to the foundry on grounds of quality problems. Not long before the shortest day was upon us here in the UK, I was prompted by the feature article on this site by John Prescott describing how he replaced the valve seals without removing the cylinder head on his Mark VI Bentley following Ashley James' informative article on this topic. Waiting for a new exhaust manifold might put the valve seal replacement back by months or even years and so here emerged a more economic way of skinning the cat. Jason measures up for fork fulcrum bracket. Compressor hose and modified sparking plug tightly screwed into cylinder 2.

A visit to Ristes yielded on the one hand a repeat of the frustrating woe with respect to the exhaust manifolds and on the other there was a choice of valve seals; the String and Tallow type (asbestos string is no longer available) versus felt pads. Advice at the parts department was that 1929ish was the probable change-over period from the string and tallow to felt pads. The price difference between the two types (£8.91 each for the string type and £2.41 for the felt pads, including VAT and a set of 12 are required for the job) might be persuasive in favour of the felt ones, but since a judgement based on economics alone is not always the best, I opted to take a set of each type on a sale or return basis.

The High Pressure Blow Job

I drove up to Chesterfield with the hood down; layered clothing and woollen hat made the trip pleasant, despite the close to zero air temperature. On installing the vehicle in his garage, Jason saw only the packet of felt pad seals on the front seat and his immediate reaction was "Wrong sort. They're no bloody use, I can tell you now. They won't last more than five minutes." He had already consulted the RRBEW website and seen the "No Head Off" technique advocated by Ashley James and as used by John Prescott and was impressed. He was not confident in the special tool for keeping valves in place, designed by Rolls-Royce and DIY'ed by Ashley however, but he thought the compressed air pressure route to be a safer option.

Killing several birds with the same stone has always been a favoured practice for me and in this case an assuring and educational experience was to follow. Using the compressed air to maintain closure of the valves, a High Pressure Leak-Down Test was to become a welcome side-product of the main objective. Apart from securing the valves, thus preventing them from falling into the abyss once the spring cap collets are removed, the advantages are as follows:

The pressurised air is going to find its way past all gaps emanating from each cylinder and this will be evident by audible means. Naturally, a good ear will help. This test may reveal incompetent valves and piston rings as follows:

a) Removal of the oil filler cap will enable any piston ring wear. Hissing up from the crankcase is the symptom here. With a hand on the oil filler pipe any pressure pops indicate severe wear.

b) Hissing from the exhaust tail pipe indicates exhaust valve wear.

c) Gurgling through the carburettor indicates a failing inlet valve.

d) Removal of the radiator cap will enable any bubbling of air up through the cooling system to be heard. In this case, a head gasket leak and/or cracks in the head or block are evident.

If all of these symptoms are identified, consult your bank manager!

Getting on with it. Off came the rocker cover, HT leads conduit including brackets and rocker shaft. An air compressor hose male connector outlet was modified and arc welded to a spark plug casing, ensuring an airtight joint all round. For each cylinder this connector was screwed tightly into the spark plug hole for the operation on each pair of valves. The leak-down test was conducted prior to each valve seal replacement. This revealed good news all round apart from just a very, very 'slight' wear on some of piston rings but accepted as 'within the normal range of acceptability' and attributed to an expected seepage past the circumferential piston ring gaps rather than wear. Fork lever sitting over valve spring cap and beneath angle iron fulcrum. Air hose connector/plug behind Jason's left thumb.

Jason had no valve spring compressing tool available that day. A pair of 'specials' were made as follows:

1) A compressing fork. A length of round cross-section steel was cut and then using an acetylene torch to heat it up at the mid point to bright red, it was bent around a vice-held fulcrum with a suitable radius, such as a 15mm socket, to form a pair of parallel prongs later given an upward curvature, just like a miniature pitchfork. The width of the prongs should be sufficient to meet the outer rim of the valve spring caps. Another piece of the same steel was then welded on to form a longish handle/ lever.

2) A piece of angle iron to secure on to each consecutive pair of rocker shaft studs and to act as a firm fulcrum or bridge for the forked lever. This should be marked and drilled on each face to suit two different centres between studs. Steel piping or columns of washers are then used to take up the majority of the stud length between the cylinder head and the bridge you have just made.

A check on air compressor's pressure head was made before each cylinder was targeted. At what threshold pressure would the valves drop? The 90 psi advised by Ashley might be critical in this respect. We decided to start at 150 psi. With the special fork in place, a series of quick taps or thumps on its handle were made to dislodge the spring caps from the seal collets. Until these are free the valve will open allowing a very audible 'pop' as the compressed air is released. The pair of collets was then removed with extreme care; they might tumble into the tappet chamber. After each pair of valve seals was replaced by the string and tallow variety, the fall in pressure at the compressor gauge during the operation was 10psi. This was never allowed to fall below 120 before the compression chamber was recharged up to 150psi.

The Culprit Emerges. The resident seals were of the felt pad type. Eleven of them were in a fairly sorry state. I cannot put into writing the rather explicit expressions used by the very Yorkshire Mr Fox on extracting these other than "Wrong bloody sort!" The one at No. 6 exhaust valve was more or less completely disintegrated, totally ineffective as a seal and probably the main cause of the smoking at cold start. Before replacing the rocker shaft each valve spring cap was given a couple of strong taps with a rubber hammer to settle home the valve collets. With rocker shaft now in place, a check on tappet clearance was the final act of the operation. The weather the following day was penetratingly cold; damp/foggy and not short of freezing point. A cold-start was called for. Hey Presto! Not a cloud of smoke in sight. Now for No 3 cylinder.

Job done!

The compressed air hose connector, now married to the spark plug casing in perpetuity, is a very useful addition to the regiment of tools in the garage. Do not throw it away. You will use it again, I promise you, if only to show your friends just how rotten their engines are!

On to the next winter task.