After cleaning , more cleaning and .. well more cleaning the head is still not as clean as I like it to be but you can now touch it without getting your hands dirty.
I only have this head (unless someone wants to donate a few more) so there is not going back once a modification is done.
I will do the following steps.
1- Baseline measurements. The only modification I have done is to re-cut the seats ( the very wide single 45 angle) so they are at least clean enough to seal properly and will lap them if needed for a decent seal.
2- 35 and 30 degree back cut in the valve with standard seat
3- Chamber modifications ( Vizard has produced some data on this) and I feel that the shape is so restrictive in standard form that this need to be tackled first as this is a major bottle neck. I will do it a bit different and view the chamber more as an extension of the seat. Every bit of volume that you take out of the chamber need to be skimmed off (well actually four times that amount) to just keep the compression ratio the same. You need to machine off quite a chunk as it is. Furthermore I have noticed on 12G940 head that the flow really takes a beating after the valve has been lifted beyond the depth of the chamber.Vizard on the other hand states that flow goes up the odd CFM if the head is skimmed ( it might if you skim it to the limit). My plan is to angle the chamber much more than just cutting it more or less down to the level of the valve.
Another significant effect is bore size. I have moved the head around simulating a bigger bore and it significantly affects flow at high lift (about 5cfm can be gained just by moving the head 1,5 mm towards the exhaust valve). My makeshift bore adapter is about 72,3 mm diameter ( standard bore is 70,6 so about a +60 equivalent) . About right for a 1275 engine but way to big for a small bore. So I need to fabricate a 66,1mm bore.
for a 998/1100 you can get both +80 and +100 over-bores. For a 1100 crank that would make it almost as big as a big bore engine.
4- 2 angle seat (45-20)
5- 3 angle seat (60-45-20) . Ideally various angles should be tried but I do not have a whole collection of cutters. These are angles that are widely available.
It all nice to specify a 2,47mm 22,7 degree to 1.2mm 45,5 degree seat into a compound hyperbolic radius with a 0.0001mm run-out. But you have have to find a guy that is able to cut it.
From what I have read, the general consensus is that radius seat is not superior to discrete angles most of the time as far as producing power goes, although it does tend to flow a bit better on the bench.
A-series notable Graham Russell does use a radius seat on the intake
However others have used very complex compound radii which have produced more power, but mostly in very highly developed ports. Or very trick multi angle ones like this one
Then again this is very dependent of the head and it’s use. I’m not pretending this is to be state of the art or anything and you have to keep in mind that this is not a Honda engine that is far better engineered than this old lump of cast iron. All in all, it’s a can of worms that is for now better left unopened.
A good basic seat is the goal
6- Varying throat diameter/seat width
Another can of worms.
general consensus is to start out with 86 % diameter then go up to 88 maybe 89. Bigger is sometime beneficial mostly for very downdraughted 4v and 5v heads like recent bike heads In this case it is hard to say what will work best. Judging from the flow path and general port shape I would wager a guess that a smaller throat is the way to go. As far as seat width go: wide enough to make it last, narrow enough to make it flow. Super narrow seats are in general not the way to go. If you look at this it makes a bit more sense.
I have covered this a bit here. The majority of gains are made in the part about 20mm above and below the seat.
I’m considering investing in a simulation computer program ( EngMod4T) that can actually 1D model a siamese port head