I have a technical college degree in mechanical technology, which basically is a great deal of the practical end of mechanical engineering without digging too deep but knowing enough to find out whatever answers are needed. One thing I learned early on was about mechanical joints. A very large percentage of them depend on the axial load on the bolt creating enough friction to hold the joint by friction. Obviously there are a significant number of places where the shear of the bolt comes into play, much more so in automotive applications that in structural . (Think the clutch drum bolts holding the drum to the sprocket in a standard Vincent primary drive.)
When having bolts with plain shanks in through holes in these hubs you´d have to have precision types of bolts like in aviation to be of any use. Common bolts have decent undersize shanks so with these you cannot rely on even loads on all bolts - unsafe and not durable.
But then, science is to have high load on bolt heads for friction force to take rotational loads, not the shear load thinking. So when having bolts and nuts in standard Vincent gear get some low to medium strength Loctite on the plain shank for filling any gaps in undersize bolts in holes for spreading loads on ALL bolts. And bin these nylocs for standard nuts secured by Loctite.
In the topic today the use of countersunk screws for Allen keys was far from brilliant as the key is way too small for torquing them up. Consequence was a lot of fretting in assembly when preload on screw heads is too poor. Looking at photos above there is no great way for repairs in present screw positions. Dowels might be one way but we learned to keep to bolt preloads for suitable friction in assembly. So I´d go for new positions on a larger pitch diameter, just within the hub flange for containing screw heads - helps with easier transmitting torque at larger effective diameter.
Dealing with seven holes is messy with round tables on the mill. So find someone with modern cheap Chinese DRO on the Bridgeport or whatever household gear and let the DRO do the maths for exact holes and countersinks in hub and sprocket. Parts can be machined one by one , will be precise anyway this way. Certainly get metric Torx screws for good preload, no Allen type then !! Some mild Loctite on threads like 221 or 222 can be a good idea but I´d rather have lube on threads when assembing the lot for good loads. I´d think this will do allright, remembering the old setup with Allen screws and poor preload holding up at least a while already. My guess, these screws were not checked a long time for tightness , looking at wear patterns in photos. No surprise they snapped eventually.
My homemade hubs and drums got all Torx screws on larger diameter positions for reducing torque needs there. For fixing the drum to the hub there are M 6 threads in hub flanges plus M 6 nuts for locknuts on top. In fact , these don´t do much as loads from the sprocket drive go onto the brake drum which got one half of spokes hooked into it. So only a fraction of drive forces is handled by these screws - no headaches for me so far. Well, you can never be sure, you better never stop learning . . .
Vic
But then, science is to have high load on bolt heads for friction force to take rotational loads, not the shear load thinking. So when having bolts and nuts in standard Vincent gear get some low to medium strength Loctite on the plain shank for filling any gaps in undersize bolts in holes for spreading loads on ALL bolts. And bin these nylocs for standard nuts secured by Loctite.
In the topic today the use of countersunk screws for Allen keys was far from brilliant as the key is way too small for torquing them up. Consequence was a lot of fretting in assembly when preload on screw heads is too poor. Looking at photos above there is no great way for repairs in present screw positions. Dowels might be one way but we learned to keep to bolt preloads for suitable friction in assembly. So I´d go for new positions on a larger pitch diameter, just within the hub flange for containing screw heads - helps with easier transmitting torque at larger effective diameter.
Dealing with seven holes is messy with round tables on the mill. So find someone with modern cheap Chinese DRO on the Bridgeport or whatever household gear and let the DRO do the maths for exact holes and countersinks in hub and sprocket. Parts can be machined one by one , will be precise anyway this way. Certainly get metric Torx screws for good preload, no Allen type then !! Some mild Loctite on threads like 221 or 222 can be a good idea but I´d rather have lube on threads when assembing the lot for good loads. I´d think this will do allright, remembering the old setup with Allen screws and poor preload holding up at least a while already. My guess, these screws were not checked a long time for tightness , looking at wear patterns in photos. No surprise they snapped eventually.
My homemade hubs and drums got all Torx screws on larger diameter positions for reducing torque needs there. For fixing the drum to the hub there are M 6 threads in hub flanges plus M 6 nuts for locknuts on top. In fact , these don´t do much as loads from the sprocket drive go onto the brake drum which got one half of spokes hooked into it. So only a fraction of drive forces is handled by these screws - no headaches for me so far. Well, you can never be sure, you better never stop learning . . .
Vic
Judging by the pictures the job was flawed from day one. the bolt holes are counter bored and do not give the bolts the support they require at the point of greatest stress. It looks like there is sufficient material to bore and re thread the mountings. I reckon you should go with torx screws as Vic suggests, I don't know what they are made of , but they sure are tough little devils. Good luck, please show us your solution when done.