Draft is the slight taper given to surfaces that run parallel to the direction in which a part leaves the die. On a drawing it is often just a small note; in production it decides die life, scrap rate and the appearance of the part. It is one of the most overlooked — and most expensive — design details on castings.
Why is it needed?
Aluminium shrinks as it solidifies and grips the cores and core surfaces of the die. So on ejection the part fights not only friction but that grip. If a surface is exactly parallel to the direction of travel, the ejectors have to tear the part free across the whole contact area in the very first millimetre.
A taper of a few degrees separates the contact surfaces at the first instant of movement — once the part has travelled a millimetre, the faces no longer rub. That is why such a small angle makes such a large difference.
What insufficient draft costs
- Ejector marks and distortion. Forced out, the part carries ejector marks and thin walled regions bend.
- Surface scoring. A part dragged out shows lines along the ejection direction on its side faces. Surface treatment, anodising in particular, does not hide those marks — it accentuates them.
- Die wear. The same surfaces rubbing on every shot eat into the die steel. A worn surface solders more, and soldering wears it further; the process feeds itself.
- Cycle time. A part that is hard to eject gets more release agent and slower ejection; both lengthen the cycle. Excess release agent is also a source of gas porosity.
- Scrap. The sum of the above lowers the share of acceptable parts.
None of these show on the first trial; while the die is new, everything runs smoothly. The problem surfaces thousands of shots later, as the scrap rate quietly climbs.
Inner surfaces need more draft than outer ones
This is one of the most commonly misunderstood points. Because the metal contracts, outer surfaces pull away from the die while inner surfaces squeeze onto the core. On a box shaped part the outer wall already tends to release itself, whereas the inner wall locks onto the core.
The practical consequence: inner surfaces get noticeably more draft than outer ones. With a deep pocket or a long core the gap widens further — as depth grows, so does contact area and with it the grip.
Surface texture increases the requirement
Texture applied to the die surface (blasting, an etched pattern, a logo) creates a set of microscopic undercuts. The deeper the texture, the more draft the part needs to slide out. A request for "this texture on the surface" is therefore not only a visual decision in design; it changes the draft calculation too.
The same logic applies to coating: a thick powder film builds up on sharp corners and makes tight angles harder still. That is why the surface treatment choice and draft are discussed together.
How draft affects dimensions
Draft changes a dimension from one end of a surface to the other. Where that change starts is a decision: is the taper taken off the nominal, added on to it, or split either side of a mid-plane? Because it directly affects wall thickness, fitting faces and assembly, that decision belongs explicitly on the drawing.
A frequent situation: a face critical to assembly falls outside tolerance at one end because of draft. Such faces are either planned to be machined (CNC machining) or the orientation of the part in the die is changed.
The parting line sets the direction of the draft
Draft cannot be considered on its own; how much draft a surface needs is decided by how the part is split in the die — that is, by the parting line. The parting line is where the fixed and moving halves of the die meet, and every surface of the part falls on one side of it.
Three practical consequences follow:
- The direction of the taper changes. The same surface wants taper one way if it falls on one side of the parting line, and the opposite way on the other. Writing a draft value before the parting line is settled is therefore a premature decision.
- Some problems disappear on their own. Shifting the parting line by a few millimetres sometimes frees a surface that could not be drafted — no core required. That is often the cheapest solution available.
- The line is a visible mark. The parting line leaves a fine witness mark on the part, and even dressed it never disappears entirely. Where the line runs across visible surfaces is therefore a visual decision as well.
In short, parting line and draft are two faces of the same decision, and both are discussed in the first die design session.
What about surfaces that cannot be drafted?
Some faces must be parallel by function. The options are:
- Use a core. A moving core forms the surface and retracts before the die opens. It solves the problem but raises die cost and maintenance load.
- Leave it to machining. The face is cast with draft and then machined parallel, which requires machining stock.
- Change the orientation of the part. Shifting the parting line sometimes removes the problem entirely, and it is often the cheapest answer.
Which is right comes out of assessing volume, cost and function together. We put all three on the table during die design.
A common mistake
Adding draft at the end of modelling, on the grounds that it is "only a small detail". Added late, it shifts wall thicknesses, thickens rib intersections and disturbs a thickness distribution that used to be balanced — which can create a hot spot. The right approach is to treat draft as part of the geometry from the first sketch.
The second mistake is applying one draft value to the whole part. Outer faces, inner faces, deep pockets and textured surfaces all want different values; a single figure means either unnecessary material in some places or insufficient release in others.