Shockingly it seems that my last blog dates from October last year: There have been major works in progress at home and since my evenings have largely been taken up with building work, I have to confess that enthusiasm for blogging has waned. Things are still no less busy but I have at last found the resolve to return to the keyboard – hearty thanks to all of you who have spoken to me about it over the last five months and more thanks too if you are still reading this!
Many projects have passed through the door since the last post, but the one which is interesting me most at present is the cylinder head from a 3.5 litre “Derby” Bentley on which I have finished working today. The head is a second hand one of unknown origin and although not without faults it was very “saveable”. Four of the exhaust valve seats had cracked and pressure testing revealed a further crack in the top of the head next to a valve seat guide as well as an area of porous casting. The exhaust seats were x-ray tested and the cracks were found to be relatively shallow, thereby enabling them to be removed by machining so that hardened seat inserts could be fitted. The crack in the top of the cylinder head was repaired by cold stitching and the porosity sealed by ceramic treatment.
The final operation neccesary to render the head serviceable has been the cutting of the valve seats which is for me one of the most interesting discussion points. There seems to be a new phrase in the vocabulary of the car enthusiast these days: “Three angle valve seats”, and I suspect that the popularity of this phrase is largely due to the internet.
Pictured on the left is a nice clear drawing (for which, sadly I can take no credit) of the 3 angles on a valve seat. The actual seating area on which the valve seals is the area cut at 45 degrees and the remaining two angles are sometimes known as the “topping and tailing” cuts. It is the width of these angles which can be adjusted to regulate the width of the seating area.
The picture on the right shows a section of the valve sitting on the seat with the seating surface illustrated in the shaded area of the valve itself. If the valve were to be tested in the seat with marking blue this is the witness which would show up.
Perhaps the reason that this all interests me so much is that when I first started to learn the trade, I hadn’t heard of the modern term and I thought that this was just the minimum requirement for “a valve seat”. The truth is that this design of valve seat goes back earlier than the 1920’s, although it probably wasn’t until at least the 1950’s before people began to realise all the reasons why it works so well. In the Edwardian era it was understood that the best valve-to-seat seal could be acheived by a very narrow seating area but that this would not provide adequate heat transfer to prevent the valve head from running too hot. Similarly it was also understood that the best heat transfer could be gained by a really wide seating area but that this would not give such a good (or durable) seal. Clearly a compromise was in order: Manufacture the valve seats at a width sufficiently wide to transfer heat and sufficiently narrow to provide an adequate seal. Ever since those days the better engine specialists have been cutting what are now known as “three angle valve seats” in order to acheive the correct seating width. (Note that the valve seat is the whole item with its three angles and that “seating” is used to describe the area shaded in black on the second picture).
In the early 1920’s it became fashionable for race engine builders to use “knife edge” valve seats. This involved “topping and tailing” the seat to three angles in order to produce a very narrow seat which appeared to produce good power but which had short life. At the time the fact that this produced observably more power was almost certainly put down to the superior valve seal. However it has since been realised that having three angles actually improves gas flow around the seat area because the change of direction from the valve throat area is more gradual.
There is another benefit to this design of valve seat, particularly in respect to exhaust valves which run much hotter than the inlets: The weakest part of the seat area on the valve itself is right at the lower edge (at the top of the upside down valve in the picture) because here the valve is at its thinnest. The 30 degree cut on the combustion chamber side of the valve seat relieves this area of the valve from any contact with the seat as the valve spring snaps the incandescent valve shut (25 times per second at only 3000 rpm).
Three angled valve seats therefore have the benefits of better sealing, better heat transfer, better durablility and better gas flow.
I’m just glad that there’s a nice new name for them now!
Welcome back to the keyboard Bruv!
Thanks Paul – here’s a picture in the next post of some “Proper” cars……
I was interested to see the description of the Derby Bentley cylinder head before you decided to put so much time and effort into saving it. With 5 cracks and an area of porosity one could perhaps be forgiven for thinking that trusting this casting to stabilise was a risky venture. My comments are prompted by the fact that I am still smarting from the misfortune that befell me in March 2000 when the cylinder block of my Bristol 403 cracked. Peter and yourself removed the engine and the crack was stitched. Upon replacing the engine, the car covered a couple of thousand miles before a new crack appeared in the block ( between the water jacket and cylinder No. 4 ). Engine out again and the new crack was stitched and engine replaced. June 2000 saw the third crack appear, or more correctly not appear, as we never found out where it was. We had to assume that the crack was somewhere deep in the bowels of the block and, as such, probably inaccessible. The repairs to the first two cracks were still sound and, as the levels of despair were becoming sleep – depriving, we abandoned the block and sourced another with great difficulty and at a huge expense relative to the value of the car.
I’m sure that you can explain your faith in this Bentley cylinder head. Having repaired the current cracks and porosity, how can you be sure that the head will be immune to further similar faults ?
David Dale.
The bottom line is that with a cylinder head approaching 80 years of age: there is no complete certainty that there will be no further faults. Repairs carried out by welding or stitching specialists are never without some degree of risk but in my experience are very rarely problematic. Internal faults can be hard to find as we experienced with the Bristol, and the cost of repairing these is usually uneconomical. The increasing scarcity and cost of alternative items is an important factor in trying to determine the best course of action. After much discussion with the owner of the car, he decided that repairing the cylinder head was the best option to suit his needs and that the risk is favourably acceptable to him. The lowest risk would have been to purchase a new cylinder head since these are now available for the car, but this is something that the owner wishes to avoid if at all possible on grounds of originality, since the new heads are manufactured in aluminium as opposed to the factory item which is of cast iron. These too are of large expense relative to the car’s value and the completed repairs to this particular head have cost around one fifth of the cost of a new aluminium one. The damage to this head is very typical for these cars: Cracked valve seats caused by local overheating and cracks around a guide boss caused by previous careless fitting of valve guides. The small degree of porosity in the top surface which was revealed by pressure testing would almost certainly not have been noticed in service but ceramic sealing was used as a preventative measure. There is no evidence of any previous freezing of coolant which would cause internal damage, although of course one can’t easily look inside the casting. I doubt that there are any serviceable second hand heads left for these cars now – believe it or not this was the best item that the owner was able to source in order to replace the one on his own vehicle, which had revealed no less than 14 cracks during pressure testing.
This vehicle is in daily use since the owner has no modern car, so the job will certainly be put to the test. I’ll keep the blog updated with its progress!