Starting on the B40 head.

This is one of those moments when you really need to remember what you’re doing. There’s a huge temptation to get carried away and put lots of effort into a race prepared head that will last for nearly two hundred miles. That’s very much not what this bike is about though. As stated earlier, I’m after a bit more than a stock B40, but definitely without making anything fragile or maintenance heavy. It’s just a fun classic road bike, that can cope with modern traffic.

So, ‘ere (as they say) we go…
Firstly, like all good engineers (and bad journalists) I shall quote my main sources. Such head hacking knowledge as I have has been gathered over many years, from all over, but I guess the main names are David Vizzard, from way back in my Mini racing days, Graham Bell, probably from even earlier and Rupert Ratio (aka Dave Smith) for the BSA single specific stuff.
If you like to do your own research, then obviously Google is your friend, but beware that any idiot (like me for example) can publish on the web these days, especially on the various forums. I find the best bet is simply to never trust information that hasn’t been corroborated by more than one source, and make sure that the second one isn’t quoting the first!
Some of this information is, of course thirty or forty years old. Fortunately, despite all sorts of trends in tuning, fluid dynamics hasn’t changed, ever. There is the occasional advance in our understanding or measurement of it (like when boundary layers were first understood) but they’re pretty rare. Most of the advances in tuning are to do with electronics these days and the others are about materials and manufacturing capability.
Remember, this isn’t about motogp or dragging the last possible 0.01% out of the engine. Solid, proven engineering will be fine.

There are two very good bits of news about the B40 head as a starting point. Firstly, the combustion chamber is a pretty good shape. It’s a classic hemi head (as in hemisphericalish) which generally makes for good gas flow.
The second bit of news is that the ports are pretty rubbish. That means that there’s some nice “low hanging fruit” to be gained in terms of releasing power.

On the subject of starting points, I’d already decided to go to SS90 valves and a 30mm flat sided carb before starting and managed to get hold of a pair of N.O.S. valves off the web. A quick note of caution there. Even if you’re collecting bits for a future build, make sure you check them straight away. One chap on ebay (who called himself an expert dealer) tried to flog me a C15 camshaft as an SS90 one, so make sure you’re getting what you’re paying for.

Obviously, the first job is to assess what you have. This is what I pulled off the top of the engine.

And this is what it looked like once I’d cleaned it up a bit.

The condition of what you have may well decide what order you now proceed in. In my case I had two questions to answer up front.
1) How hard are the valve seats?
2) What sort of condition were the valve guides in?

Here’s the logic behind that. Some people recommend that on B40s you shouldn’t change the valve guides unless you need to (even to the extent of having them lined) so I wanted to avoid it if possible. They’re cast iron guides, which wear pretty well, and the rest of the engine has so far proved to be in surprisingly good nick.
So I needed to know whether to order up new guides, plus appropriate reamers etc.

As for the valve seats. I intend to cut my own (that way I stay in control of the fine detail) and if the seats are hardened I’ll need carbide cutters. If not I can make do with cheap HSS ones. If I were richer I’d just buy a modern set for £500-£700, but I’m not, so I’ll take a chance on a cheap set from India.

This is an ancient Sykes Pickavant valve seat cutter that I got for a tenner at some autojumble, years ago. Sadly it’s about 1.5mm too small for the new inlet valve, but at least I can see if it cuts the seats…#

…and it does, so HSS cutters it is and I can get them ordered.

And so, to the valve guides. There’s loads of information on the web about the correct clearances for valve guides (generally between 0.0012” and 0.0015” for a 5/16 stem) but very little about how to measure it. My smallest bore gauge is about 3/8”.
The trick I learned is that if you put the valve in the guide dry, you should be able to feel some wiggle in it. If you then put a thin coating of oil on the stem and put it back, the wiggle should disappear. It’s a much more accurate test than you’d imagine.
Both of my guides passed that test with flying colours, using the old valves and the new ones (with cast iron guides the valves tend to wear more than the guides anyway) but that didn’t take into account anything I might remove from the new ones whilst preparing them. So my first job was to be sorting the valves themselves.

These are 1960s/70s BSA parts, so don’t expect anything beautiful. You could lap them in and they’d work, but you’d wear out the valve guides really quickly and re-profiling and re-finishing them should yield a significant improvement in efficiency.
The inlet valve took me about two hours, but I wasn’t rushing and remember that the head is where all the power is made and tiny differences in geometry can make big differences in gas flow.

The inlet valve.

There are basically two shapes of poppet valve. There are mushroom valves and nail valves. A mushroom valve has a convex back face (like the top of a mushroom) and a nail valve has a more or less flat back face (like a nail).
Mushroom inlet valves work well on hemi heads. Personally I picture the gas following along the curve of the valve head, then continuing along the curved surface of the combustion chamber, so it sort of fits.
For flat combustion chambers, nail inlets work better. Again for me, I just picture the flow being sprayed out flat from the valve head (a bit like a lawn sprinkler) and following the flat surface of the combustion chamber. I’ve no idea whether that analogy is accurate in terms of gas flow, but it works for me.
So the B40 inlet is a mushroom valve. It’s not a hugely pronounced curve, but it’s there. On the factory fresh valve it then has a 45 degree seat turned on it, below which is a flat area called the margin.

The whole thing is finished fairly roughly and mine had a few tiny nicks around the edge, which I reckoned would both disrupt the gas flow and be centres for build-up of heat and carbon. The valve proportions I’ve used are from Graham Bell (who backs up his assertions with lots of flow bench work) but slightly modified, which I’ll explain as I go.

I started work on the face and head of the valve, primarily so that any marks the lathe chuck left in the stem would be removed when I refinished that bit.

The margin should be between 0.040” and 0.065” deep. This gave me scope to remove a fair amount from the face of the valve and thus get rid of the nicks in it, without going right down to the minimum. In a race engine, losing the tiny bit of mass and potential clean flow that a minimal margin would achieve would be worth it. On the road it’s safer to leave everything a little bit chunkier (for strength and cooling) but not as chunky as the original part, which was about 0.075.
I ended up with about 0.050”. Tiny differences really do make a difference when shaping valves.
For the inlet valve, the bottom of the margin, the angle between the margin and the face, should be sharp. This is to disrupt any tendency towards backflow.
So my first cut was to take a chunk off the valve face. I’m not sure what the composition of the steel is, but an HSS lathe tool came off worst in the encounter, so I switched to a carbide one.

The next job was to polish the face. I ran the lathe at about 2400rpm starting with silicon carbide paper (wrapped over a sponge to conform to the shape and stop my finger burning) and working my way down from 120 grit, through 400, 800, 1200 to 2000, then polishing compounds (black , green then white) applied directly to the face, then burnished with a J cloth.

Finally, a very light touch with the lathe tool again, just to re-sharpen the bottom edge.
With the face done I made a 30 degree back cut, bringing the width of the seat down slightly and blending it into the back face. It’s only a small cut but, again, makes a difference.

the width of the actual seat is a matter of some debate. 0.075” seems to be standard for race engines, but I prefer to leave a bit more on road engines, which are not expected to require rebuilds every few hundred miles. Just be careful how you match it to the seat, but I’ll come to that later when I cut the seats.

Once the face and back cut were done I chucked up a piece of scrap aluminium, which I faced at both ends and drilled for a centre to provide two perfectly parallel faces. The idea was that, once I had the valve running true to within a few microns I could force the aluminium against the valve head with a live centre, thus providing decent support as I worked on it.

The stem needs to be polished in exactly the same way as the face was, except that it’s a much easier shape. Just make sure that you use a bit of wood, rather than your finger, when working near the chuck. Reducing the diameter of the stem, below the guide, by 0.035” and tightening the radius of the bend onto the face to about 10 or 12% of the head diameter will yield about a 10% increase in flow at very small valve openings. It’s only a small part of the cycle and personally I prefer not to on road engines, again to improve cooling and reduce the chances of the valve bending or snapping, both of which have happened to me in the past.
The result should be very shiny. You can still see circular imperfections on the head of mine, but it feels perfectly smooth to a fingernail and looks smooth under a magnifying glass, so I’m not sure how far I’d have to go to remove them.

The Exhaust Valve

The exhaust valve is a horse of quite a different colour, for two very sound reasons.
Firstly, it doesn’t really have to concern itself nearly as much with fluid dynamics. If you think about it, the inlet charge is being dragged kicking and screaming from a nice cool world into a very hot little hole. The exhaust gasses, by comparison, are not only being shoved out by a thumping great piston but are also desperately trying to find somewhere to expand to, so apart from some finesse in maintaining the exhaust velocity, one can pretty much assume that it’s keen to get out.
Where the exhaust valve does have a real issue is simply survival. It’s operating at temperatures upwards of 800 deg C, so not burning up is its main concern.
In shape it wants to be a nail type. We’re simply looking for as big a hole, as soon as possible when it opens, and the roughly concave surface (if you include the curve up to the stem) naturally turns the gas up into the port.
We can leave some extra meat in the head compared to the inlet (any thin bits may burn up) and a margin of 0.060 to 0.100 will give the best flow. The angle at the bottom of the margin should be radiused in the exhaust, the gas is flowing outover and we don’t want to disrupt it.
A wider seat of 0.080” or more (which is its only solid contact with the head) will help a lot with cooling.
Reducing the stem near the head is the same as on the inlet and, again, I chose not to. It still wants that 30 degree back cut and, just like the inlet valve, the more polished the finish, the better.

The Combustion Chamber

As I said at the beginning, the B40 combustion chamber isn’t a bad shape at all. It’s a nice smooth shape with no nasty pocketing or obstruction of the valves and nothing to obstruct gas flow or flame front. Even the plug sits nicely flush with the surface, so there’s no shrouding of the spark.
Being happy with the shape, the next question has to be condition.
Once the carbon was cleaned off it had clearly had a moderately hard life.

I’m not sure what bit it here, but it does look like a potential hot spot, which could cause detonation.

All in all, I wasn’t sure how far I wanted to go with this head. I had a sneaking suspicion that if I were to take all the deeper scratches out I may well remove enough material to be doing more harm than good. To be honest, I’m still not sure and may well go back to it later, but for now I’ve contented myself with a reasonable clean up and light polish.
Putting a perfect polish on the combustion chamber doesn’t actually do all that much for performance (especially once it’s done some miles and has carbon on it) but does reduce the ability of carbon to stick to it, so it stays cleaner for longer.

Next up, the inlet port…