Machining stock is the layer of metal left on a cast surface so that the subsequent machining operation can produce a finished face at the correct position, with clean metal beneath it. It exists because a cast surface is not a machined surface. It has a skin, it has dimensional variation, and it may have a small amount of subsurface defect that has to be removed before the material can be trusted.
The amount of stock is a decision with consequences in both directions. Too little and the machining operation cuts through into porosity or leaves a surface that is still partly cast. Too much and the casting is heavier than it needs to be, the machining time is longer and the tool wear is greater. Neither error is dramatic on its own, and both are avoidable.
The starting point is the tolerance capability of the casting process on that particular surface. A surface facing the cope behaves differently from one facing the drag. A surface at a junction behaves differently from a free wall. Uniform stock across a part with non uniform behaviour will be excessive in some places and marginal in others.
Stock has to cover three things. It has to cover the dimensional variation of the cast surface, so that the finished face lands where the drawing says. It has to cover the depth of the cast skin and any subsurface layer that is not acceptable in the finished part. And it has to cover any distortion that occurs between casting and machining, which on a thin walled part can be significant if the part is not stress relieved before it is cut.
The third item is the one most often forgotten. A casting that is straight as cast can move when metal is removed from one side, because the removal changes the balance of residual stress. Stock on the opposite face is what allows the part to be corrected. Without it, the part is scrapped at the final operation, which is the most expensive place to lose it.
This is also the argument for rough machining before finish machining on parts where distortion is likely. Roughing removes most of the excess, the part relaxes, and the finishing cut brings it to size. That sequence needs stock to work with, and the amount has to be planned rather than hoped for.
Flange faces, spigot diameters, bolt pads and port interfaces carry the fit of the component. These are the surfaces where stock matters most, because the finished dimension determines whether the part assembles. They are also the surfaces where excess stock is least welcome, because the machining operation that produces them is often the one that sets the cycle time.
The right approach is to specify stock individually on these features rather than as a global allowance. A flange face that will be faced in one pass needs a different allowance from a bore that will be roughed and finished. Treating them the same is simpler to write and harder to machine.
Where a feature is both a fit surface and a pressure boundary, the stock also has to guarantee that the finished surface is sound. That means the acceptance criteria for porosity in the region beneath the surface have to be consistent with the depth of cut. Specifying sound metal at the finished surface while allowing porosity just below the stock depth is a contradiction that will be discovered at the leak test.
Stock is best shown as a per surface requirement, either as a note against the surface or as a table of machined features with their allowances. A single note in the title block saying that stock is to be left on all machined surfaces is better than nothing and worse than a schedule, because it leaves the quantity to be inferred.
There is a cost argument for getting this right that is easy to miss. Stock that is excessive adds weight to every casting in the programme, and molten metal is not free. It also adds machining time to every part. On a long run, a small excess allowance compounds into a meaningful cost that nobody ever questions because it was written into the drawing at the start.
The first casting is where the stock decision is tested. If the machined faces come out correct and sound, the allowance was adequate. If they come out with porosity exposed or with insufficient material, the allowance was wrong, and the correction is a pattern change rather than a machining change.
This is one more reason to measure the first casting before machining, and one more reason to keep the as cast dimensions on record. The relationship between the cast surface and the finished surface is the evidence that the stock decision was correct, and it is only visible while both are measurable.
Fabricated alternatives to cast components, such as the TIG stainless steel fabrication work, do not carry stock in the same sense because the plate thickness is the tolerance. That is a genuine advantage for simple shapes. It disappears as the shape becomes complex, because the fabrication tolerance then has to be controlled across many more joints.
Machining stock is therefore a small number with a large influence. It decides whether the finished part is sound, whether it is straight, and whether the programme carries an invisible weight penalty for its whole life. Specifying it deliberately, surface by surface, is a modest effort with a durable return.
Stock has to exceed the tolerance of the surface it covers, or the finished face will not clean up. That tolerance differs between a cope face and a drag face, so a single figure across a whole part is either wasteful or marginal. TIG aluminium welding works from plate and avoids the question.
A thin wall with stock on both faces loses mass when it is cut, and the balance of residual stress goes with it. Roughing before finishing lets the part move while there is still material to correct. aluminium fuel tank range and aluminium water tank range components are machined on that principle.
Every machined face is a second operation, and every operation carries set up, handling and inspection. custom parts and titanium fabricated components are quoted with that in mind, because the cost of a face is not only the cut.
