Why Does Porosity Appear Under the Bearing Seat?

Shrink holes open under the bearing seat because the 5–8 mm wall around the half-round seat is the last section to solidify, so the feed path into that mass has already closed. The fault is set in the gravity casting stage rather than in the cleanliness of the melt, and the boring bar only uncovers it later.

The part is an A356-T6 split housing, 120×95×70 mm with a Ø60 H8 bore, both design targets. Its seat is the half-round face that carries the bearing shell, split across the parting line.

Why does porosity appear under the bearing seat after machining?

A shrink cavity forms under the seat because that region freezes last while the wall around it is already solid, and metal shrinking inside a closed shell draws from nothing. The void opens at the centre of the heaviest mass, below the face the boring tool will cut. Gas pores from a wet charge sit anywhere in a casting; this one keeps to the thick section.

Machining removes the sound skin that sealed the cavity, so the hole appears on a face that passed visual inspection at the foundry. That delay is the evidence: a defect first seen after boring belongs to the pour, and a quote that prices only the boring pass cannot remove it.

Which section of the wall feeds the shrink hole?

The 5–8 mm wall behind the half-round seat feeds the cavity, since the thick end of that band stays liquid after the thin end has sealed. A356 (AlSi7Mg, GB/T 1173 ZL101A) freezes over a temperature range rather than at one point, so the last liquid sits in the middle of the heaviest section with no riser contact left.

Reading the wall as strength margin gets the fix backwards. The 5–8 mm target follows the bearing load, and extra metal at the seat without a matching feed makes the cavity larger, not the housing stiffer.

What size of shrink hole scraps the housing?

A hole that measures over Ø1 mm scraps the housing; one at Ø1 mm or under passes if the seat is otherwise sound. The line is read on the machined seat by eye, then confirmed by penetrant testing, and it belongs in the first-article report as a scrap criterion, not a rework note.

The seat is a bearing interface, so a hole of that size interrupts the contact the shell presses against. Judging the mark against a measured size also settles who pays: the part is scrapped or passed against a number on the drawing, not against an opinion about a mark.

What does penetrant testing show that the eye misses?

Penetrant testing shows the cavity that reaches the machined surface but reads as a tool smear under a lamp. Dye enters an opening too narrow to see, the developer pulls it back out, and the indication marks where the finish had looked continuous.

The check belongs to the casting, and it is scheduled after boring and before final inspection, which is why the machined faces carry it. It has one limit worth writing into the drawing: a cavity that stays closed under the skin holds no dye, so PT clears the part while the porosity is still in the wall.

What changes in the casting before the bore is cut?

Feeding changes at the pour, because a hole that opens on the machined seat cannot be closed later. The riser has to reach the seat node, so the thick end of the wall still has metal behind it when it freezes. Pour temperature runs in the 700–740 °C band and the steel-mold preheat in the 250–350 °C band, both typical ranges pending confirmation from the foundry.

Heat treatment follows the same rule: solution treatment around 535–545 °C, a water quench, then aging at 155–165 °C for 4–6 h, again as typical ranges pending the foundry's sign-off. Blasting runs before the split faces are milled and before the bore is cut. The machined faces are not blasted afterwards, so a blasted seat is a process error rather than a finish option.

Send the drawing with the seat section, the Ø1 mm scrap line and the penetrant step marked, so the quotation separates feeding at the pour from cutting at the machine.