Please explain "core shift in casting"

[FERoadster]

I have a basic understanding that the cores for the pistons were held in place by rods and could shift when the molten Ci was poured.
Shouldn't all the overall dimensions from outside of each core be approximately the same? In other words this 4.11 bore with a sonic check on the opposite sides all measurement added up should be close. Looks like the wall thickness of the thrust and non thrust  when combined should be close to .330" Does core shift make one side thicker and the other thinner? If so please explain how #8 ends up at .212"
The sonic test is attached

Thanks

Richard >>> FERoadster

[wsu0702]

Does core shift make one side thicker and the other thinner? If so please explain how #8 ends up at .212"

Yes.  Looks likes the #8 bore core was shifted quite a bit directly towards the front of the block relative to the water jacket core.  That's why that cylinder is so thin on the thrust side.  Each bore core is suspended in space relative the water jacket core (see pic).  Manufacturing tolerances for the cores themselves and the core placement fixtures resulted in inconsistent cylinder wall thickness on a lot of blocks.  The placement method and accuracy was a much bigger contributing factor than the actual core geometry.       

[6667fan]

What is the explanation for number 8 being thin on both major/ minor thrust sides?
Shouldn’t the side opposite the thin side be thick?
JB

[Chrisss31]

I would guess inconsistency in sand mold.

[FERoadster]

That's my question--- Off center mold would make one side of the piston core larger and possibly  the opposite side smaller unless the radius was a factor on the X plane (side to side). Then the Y plane (front to back) should show a major difference. If the core was moved significantly forward then both X plane sides would be on a downward slope, therefore smaller but equal. Realize that the cores could shift on a diagonal. 
#1 shows a forward shift of .070" while #8 is just screwed up. Maybe a bad piston core on #8

Comments are a good education

Richard 

[Barry_R]

The cores can move.  They can be misaligned during assembly.  They can also be damaged during assembly or pour.  Problems can also occur during machining from incorrect index or tool issues - or stacking of tolerances.

[cjshaker]

The molds used to make the cores can also wear over time, allowing tolerances to change. Like anything else, they wear out and must be replaced from time to time when very large volumes are produced. The same holds true with stamping dies, although I don't think they're typically replaced that often due to the huge costs involved. That's why later sheetmetal stampings can be 'off', or not as good as earlier ones. That's when they typically just stop using them and discontinue the part. Given the 20+ years the FE was produced, and all the different cores used, plus different foundries, I would imagine they went through more than a few core molds.

[Katz427]

And, it could be just a false reading. Having done sonic testing, I can say it is relatively easy to get a false reading, even with expensive equipment. I used an oscilloscope to check the signal, to filter out errors. Not easy to get  "accurate" numbers, the operator really needs to understand his equipment.

[HvyFt4spd]

X2 on the sheet metal comment... I get the biggest kick out guys that say "they were red one day and black the next". That is a true statement but they try to imply that all NOS is equal. The fact is the farther away from the original date of manufacture of a given body design being in production the less likely the part is to match up not only in terms of overall fit but of every bodyline and detail.

 With block inconsistencies consider how any business you have ever worked for suffers for a period of time with a management change or FNG. There is always that person that trys to change what works or is in a rush justified or not so quality suffers for it.

[gt350hr]

 To expand a bit on Barry's post, you need to look at the operation. REALLY BAD working conditions in a foundry, very little time to put a mold together and send it to be poured and constantly changing patterns ( different engines). These are mass produced parts not short run hand built blocks. 351 Clevelands were the worst for core shift and 460s were probably the best.
    Randy

[machoneman]

2X to poor conditions, lots of obvious rejects that got sent back to be remelted.

Yes, 351C's were about the worst but I say the 351W's were the best or at least equally good for 460s.

To expand a bit on Barry's post, you need to look at the operation. REALLY BAD working conditions in a foundry, very little time to put a mold together and send it to be poured and constantly changing patterns ( different engines). These are mass produced parts not short run hand built blocks. 351 Clevelands were the worst for core shift and 460s were probably the best.
    Randy

[gt350hr]

    Thanks Bob,
          Some people can't envision how blocks are cast or the speed that they are cast. Dumping a couple hundred pounds of molten iron from a "bucket" above the mold can cause things to move around in a casting core "assembled" in a few minutes and done a hundred (or more) times a day. The boring machine does all 8 bores at one and where the bores are "supposed to be" regardless of where the casting actually IS. I don't know anyone that would call it a 'dream job. LOL
     Randy