midget,OK, no graphs or charts, just some quick info, and a comment about what should be obvious.
Got the 2nd head base-lined, all 4 cylinders, as a direct comparison to the first head. The results, of which I'll post some graphs, at some later point, are:
1/ The inlet ports are slightly better, ~ 1% or so. The shape of the curve is pretty much the same.
2/ The exhaust ports though are worse, ~ 8%. A significant difference.
I'm not yet sure why the exhausts vary so much. I ground these valves before testing, so the margins are thinner, but I'm not convinced that is the difference. I'm going to have to carefully measure the port's mcsa to try and get a handle on the difference. So even though these heads are the same casting number, they might be from differing "batch lots". I'm going to carefully check them over for production date codes, if any. This would be a good question for Steve, how to "decode" any production casting codes.
If we had assumed the heads/ports were the same, (something I've seen done lately on some V8 heads by customers), there might have been gains from the mods, but the total flow might still have been lower on the exhaust side. Leading to disappointment with the results.
I'm going to fit the REC valves to one chamber, and the modified Subaru valves to another and perhaps the modified Cosworth valves to a third. Those choices, combined with potential increases in mcsa due to porting, should create enough permutations and confusion to require the portable cauldron to get sorted.
Today T & T's Serdi machine was tied up with other shop jobs. I'm hoping it will be free today or early next week. Since they work on a lot of Subaru heads, I'm hoping that the correct seat cutters are in their collection. They do have the appropriate pilots. I'll try to remember to get some photos of the work in progress.
Once again, proof that you need to test everything, rather than make assumptions. It's the old "Trust, but verify" conundrum. Thanks for the thought Woody.
Flowtestboy