Patrick, great observation on how the fin on top will cause a stress riser that cracks the head at the spark plug. I wonder if they did a heat treat after casting if it would have stressed-relieved the head enough so they don't crack. It may have nothing to do with casting stresses, but how thermal expansion is just ripping the metal apart. Thanks for the explanation, too bad you were not at Harley back in the day to show them how to fix these things.
Dr. Dick, sorry I am not trying to be argumentative or a know-it-all. Especially to you, who I gladly acknowledge as the best K and Iron Sportster guy out there. When you explained how to put the mainshaft rollers in from the outside I about fell off my chair. That is when I decided you are not Dr. Dick, you are Saint Dick.
Even if K-Models were a low-volume semi-failure, Harley Davidson was a high-volume manufacturing company, and the methods and procedures would have been the same for the K as the Panhead or other big bike. I suspect that Harley was closer to being a tractor manufacturer than Chevrolet, but it could not have been
that far from modern manufacturing methods. And modern means pre-WWII. Look at
the Chevrolet company film about making 1936 cars.
To have two different castings with the same casting numbers molded into them is just a stunning case of blatant incompetence. They probably had a square box on the drawing that said "Casting number and foundry code here". But the pattern shop did not realize they were supposed to use the new numbers, or just had a brain fart. One question was if Harley used their own pattern shop or just sent drawings to the foundry and each foundry made its own patterns. This would make sense since they could use their own code when the engraved the part number and code in the head.
You just can't assign old numbers to a new part, that is the cardinal sin of manufacturing, just ask
the idiot at GM that did that exact thing to the fix for the faulty ignition switch. I really doubt Harley made a drawing revision this major and left the same number on it. What is far more likely is that they just forgot to change the number on the pattern, and once they caught the mistake, they changed the pattern to show the right casting number, the number on the drawing that the pattern shop based the pattern on.
Its tempting to think that an old iconic manufacturing company is like a big high-school shop class. Everybody gets together every day and sits around jawboning until someone says, "Hey, lets build up a new pattern for the cylinder heads." "Yeah, that would be fun, I was getting bored!" Like it or not, management manages. This is a business. They decide to make a new cylinder head, not the pattern shop. They authorize an ECO (engineering change order). That orders new drawings, and describes when the change will happen, and what to do with old parts (run them out, put them in service parts, or scrap them). The new cylinder heads were so different that would have to make a new pattern.
Now maybe, just maybe they figured they were just changing the fins, so this was just a revision to the original casting, and so they kept the same number. This would mesh with what Dave and you have observed on how tight money was for Harley in the early 1950s. So rather than a new pattern, they did a revision to the original drawing. This would be a bit of a mess, since they would have to erase the fins on the drawing (we all used pencil back then), and then draw over the new pattern. I doubt it would be cheaper than a whole new drawing with a new part number. They they get the patterns back from the foundry and machine off all the fins that are changing. Then they glue up the new fins. But the thing is, look at the side view in the picture on the original post. Its not just a few fins changing. The base is thicker, the mezzanine deck is different, this is a whole new part.
So its based on that side view that I maintain that it was a simple mistake. I new part from a new drawing that got the old number put on it. Please don't think that when I criticizes the Motor Company or the designs I am criticizing the people that like them. Heck, I am one of those people. I accept that a broke company like Harley in 1953 might cut corners, but you have to have drawings to make a pattern. You have to. It doesn't make things more expensive, it makes them cheaper. When I helped Vance Breese with his Bonneville streamliner he had the attitude that "blueprints are for sissies." So then he had his welder make a change to the oil tank, and we spent a couple hours trying to get it back into the bike, it just didn't fit anymore. I asked him, "See why people make blueprints?"