A 150 g flanged elbow is cheap on metal; its cost sits in the mold, the melt and blast, the turning, and a 100% pressure test with rework. Aluminum casting cost follows the route, so a weight-based quote leaves most of it open.
The part is a 75×60×55 mm elbow in A356 (GB/T 1173 ZL101A, pending confirmation), cast at 150±20 g by metal-mold gravity casting. Its 4–6 mm wall, Ø60±0.5 mm flange and 0.05 mm face callout come from the drawing, and the 0.9 MPa test is the release criterion.
Five blocks make up the rest. The mold and the fixture are the tooling line: a reusable steel mold plus a clamping fixture, both cut for the batch, holding one or two parts per pour (count pending at the foundry). Melt, cut-off, the Sa 2.5 blast (ISO 8501-1), turning, the pressure test and final inspection follow.
One setup turns the flange face and the port, and the flange holes are drilled before the test. The mold is a one-off tool, quoted with the batch, while the four process blocks are paid per piece.
A 100% pressure test puts every elbow on a bench, so the bench cycle is a per-piece charge. The fixture, a clamp that seals the flange face with a soft, lightly loaded pad and caps the port, is built once and clamps one part. Flatness is judged free of the clamp, so the pressure test cannot mask a warp; a pad mark forces a plate re-check, not a second cut. The bench pressurizes each part to 0.9 MPa, holds for 60 s and logs the result; 0.9 MPa is the test pressure, not the working pressure, which comes from the drawing.
The mold and the test fixture are one-off tools, quoted with the batch; the unit price carries the per-piece setup, cut, 60 s test and record. A bench reject is booked as machining scrap. The route accepts on 100%, without sampling relief.
Reworking the 0.05 mm face takes a second clamping, a second tool pass and a second 0.9 MPa test. Flatness is checked on a surface plate with the part free.
The second cut works only with stock above the final face. Stock is the machining allowance on the casting; a face already cleaned up has none. Without stock the cut cannot remove the error, and the part is scrap.
If the same face fails across a run, the fix sits in the fixture or the clamping position, not the tool path, before a second cut is needed.
A machined elbow becomes scrap at the pressure test, after a 0.05 mm face and a Ø20 mm H9 bore are already paid for. The route runs blast, turn, drill, then test. A leak at that point discards the casting, the blast and every cut.
The test cannot move ahead of the cut, because the fixture seals on machined faces. That places the most expensive reject at the end of the route. A bend leak traces to the casting, so a second lathe pass will not close it.
If the leak traces to the bend, the gate (the inlet) and the riser (the feeder) are the two features to review, with their positions still pending foundry confirmation. If the foundry moves them, the change is cut into the mold once and quoted with the batch; the move is checked by the next pressure test.
Five callouts separate a checked quote from an estimate: 0.05 mm flatness, Ø20 mm H9 bore and 0.1 mm perpendicularity of the port axis to the flange face (design target, pending the drawing). The other two are a 60 s hold at 0.9 MPa and a Sa 2.5 blast. H9 is a bore tolerance class, too narrow for calipers; a plug gauge, the go/no-go gauge, decides it.
| Drawing callout | Target and unit | Cost block it drives |
|---|---|---|
| Flange face flatness | 0.05 mm | face cut, free-state check |
| Port bore | Ø20 H9 (mm) | bore cut, plug gauge |
| Port axis to flange face | 0.1 mm | CMM check |
| Leak test | 0.9 MPa, held 60 s | 100% pressure station |
| Blast finish | Sa 2.5 to ISO 8501-1 | blast before the cut |
A drawing with all five lets the quote split tooling, turning, test and rework. Miss one callout and it is an estimate, not a check; the difference shows up at first-article measurement. Send the drawing with the five callouts and the batch size, and the quote can be checked block by block.