Crank #6 was a success, and passed it's crack test. It's a virgin crank, and will go 10 thou under this time. It gives me one more rebuild (20 thou under) in the future before the Mitsubishi conrods won't be able to supply undersized bearings (no 30, 40, 60 etc). It will now get a tidy up, have the snout drilled for a harmonic balancer installer/balancer retention bolt, and turned down to give a more appropriate interference fit for the Ross harmonic balancer (the factory fit on the Ross balancers is waaaaaaay too tight... massive amount of pounding with a BFH to get one on to a standard crank).
Now for the next drama. The machinist pulled my sexy new conrods out of the box, and they don't fit

Several years ago I did my research with the lumpy humpy guys, and worked out the conrods I would need. Standard grey rods are not too bad, but my mental picture is that I do not want anything failing from the bores upwards (in order to limit damage to the block, pistons and head) as the parts that lie above the bores owe me a fortune. If the weak point is the crank, generally everything from the bores up is recoverable when things let go. A snapped rod however can cause major block and piston carnage.
The grey motor go-fast boys reccomend using Mitsubishi rods from the 4G63 motor. The 4G63 was a 1997cc 4-cylinder engine produced in both naturally aspirated and turbocharged form from 1980. It continues in use with some Chinese manufacturers, and was the powerplant of the Lancer Evolution when Tommi Mäkinen won his four consecutive WRC championships. Early versions of the 4G63 (prior to April 1992) used thicker con rods, smaller gudgeon pins, a broader big end width and six bolts to secure the flywheel to the crankshaft (later versions used seven bolts).
Eagle offer five different forged H-beam rods for the 4G63, all of which are 5.9” long, 45mm rod journal, 48mm housing bore, are fixed with ARP bolts and exceed ASME E-4340 steel standards. Most use the later 22mm gudgeon pin diameter. The rods are available as either standard, or extreme duty.
Typically the early 6-bolt rods are used on genuine Holden cranks (like mine), though need ~0.008” surface facing. The later 7-bolt rods are normally used on the aftermarket billet cranks as the reduced big-end width provides less crank stress. Because I am using a standard crank, I went with the 6-bolt CRS5900MCXD rods. In theory, these are 1.117" wide across the big-end. Legend has it that you take 8 thou off (making them about 1.109" wide) then bolt them up to a standard grey crank. What it looks like though is that my crank is (roughly) 1.180" wide on the big end journals. This means that there is roughly 70 thou of end-float on the big ends (waaaaaay too much... big end float should be about 3 to 10 thou). With 70 thou of float, there are two risks. One is that the conrod moves with the crank thrust, and ends up running to one side of the journal (remember that the piston end of the conrod has massive float and will let the rod move a lot laterally). I could get smart and limit the rod movement at the piston end - as I am getting forged pistons made, they can make the piston boss tighter, ala' Rolls Royce. However, the bigger risk is that the huge end float gives less oil control, and I can't keep oil up to the big ends. Not cool.
I've done some basic checks over the phone with the machinist. There is a risk that I have somehow ended up with the 7-bolt CRS5900MBXD rods. This would explain why I have 70 thou too much float - the difference between the two rods is 68 thou. The part numbers on the conrod box and invoice are correct though, and the rods measure up correctly to 1.117" wide across the big-end.
It could also be that the crank has been machined at some stage (odd though... this is a virgin crank). A quick check on the (cracked) crank #5 shows it is also roughly 1.180" wide on the big end journals, just like crank #6. I'm starting to wonder if GMH made two widths of big-end journals, though somehow doubt it (Rob, any memory of a crank change during the humpy years?).
So machining on hold until I sort the end float issue. More sleuthing to follow.
Cheers,
Harv