One Line on the Drawing, a Dozen Operations in the Shop: What DFM Really Does

Home News

One Line on the Drawing, a Dozen Operations in the Shop: What DFM Really Does

One Line on the Drawing, a Dozen Operations in the Shop: What DFM Really Does

2026.11.20

  One Line on the Drawing, a Dozen Operations in the Shop: What DFM Really Does

At the drawing review, everyone nods and says the design is fine. Then the mould arrives and the problems queue up: ejector whitening here, a sink mark there, a slide for an undercut that will not fit. The first reaction is usually to blame machining. Look back at the drawing, though, and the trap was there all along — on paper it is a single line, but in the shop it becomes a whole sequence of operations.

Start with the gap between what can be drawn and what can be made. On a product model, a rib, a deep pocket or a corner radius takes a few mouse clicks. Machining that same feature means breaking it into steps: roughing, corner cleaning, electrode sinking, electrical discharge machining, polishing. Whether the electrode is long enough, whether the cutter can reach, whether the corner can be properly cleaned — none of this is marked on the drawing, so the shop has to judge for itself. A right call goes unnoticed; a wrong call means rework. And the real cost of rework is not the machining hours, it is the machine sitting idle while a queue of parts waits behind it. The cleanest-looking areas on a drawing are often the ones that demand the most thought in the shop, because the process route is hidden there.

Demoulding is another case. An undercut calls for a slide or an angled ejector, which on a drawing is simply a matter of sketching a mechanism. But a slide needs stroke clearance and an angled ejector needs angular margin, and both must be reserved at the structural design stage. If they are not, T0 reveals whitening, scuffing or even punch-through — and no adjustment to the process parameters will save it, because the fault is in the steel, not in the settings. I understand why people reach for the parameters first: changing a setting is fast, changing steel is slow. But parameters can only push the process window towards its edge. If the window was never opened wide enough, pushing it leads nowhere. By the time a structural change is accepted, the review time saved earlier has already been paid back with interest.

An undercut usually has more than one solution. The same hook can be formed by a slide, by an angled ejector, or by changing the product so that it releases in the draw direction. If the drawing does not lay out these routes side by side and simply takes the easiest one, you are stuck later: a slide takes up space and adds an opening and closing action, a thin angled ejector wears quickly, and forced demoulding is fussy about material toughness and undercut depth. Which route fits depends on the volume, the material and the appearance requirement, and on how long the mould must stay in service. Decided at the DFM stage, this is a matter of a few sketches; left until after the first trial, it becomes a repair order and a slipped delivery date.

Wall thickness is the quiet one. On a drawing, two lines meeting is just geometry. In moulding, the thicker region cools slower and shrinks more while the thinner region sets first, so the junction throws sink marks, voids and warp. Such problems may barely show in a single sample, but once production starts and the cycle tightens, heat balance has no time to settle and the defects emerge. There is a subtler version too: the weld line where thick and thin sections meet is weak by nature, and if that happens to be a loaded area, an occasional cracked part in a batch sends everyone hunting for a damaged mould — when the mould is fine and the structure was always going to be the weak link. Smoothing a wall transition at the DFM stage changes an electronic file. Changing the same transition after production starts changes finished, hardened steel. The two bills are not in the same league.

DFM also affects how accurate a quotation is, a point that often goes unnoticed. For the same drawing, how easy the structure is to machine, how many electrodes are needed and whether special mechanisms are required all determine the real workload. Quote before a manufacturability review and you are quoting on "looks about right" — when the difficulty turns out higher than expected, you either ask for more money or quietly cut a corner somewhere, and neither path ends well. Put DFM ahead of the quotation and you are quoting a plan that can actually be built, which saves both sides a lot of argument. It is also why more customers now treat DFM cases and mould flow reports as a screening threshold: they are not buying a promise that you can do it, but evidence that you know where it is difficult and how you will handle it.

So why is DFM still skipped? Two reasons are the most common. One is schedule pressure — "make it first, fix it later" — which turns the review into a box to tick. The other is that design and machining sit with different parties: the designer answers for the drawing, not for how hard it is to make, and handing over the drawing closes the job. The result is trial after trial, a growing stack of repair orders and a delivery date pushed further out than before. What is described here is general experience; the specifics follow the technical agreement and the mould trial results. The direction, however, is consistent — time saved on the drawing comes back with interest in trials and repairs, with very few exceptions.

Doing DFM properly means more than the designer checking their own work behind closed doors. It means putting the constraints of machining, sampling and production on the same table in advance. How the undercut is formed, whether the wall transition is smooth, where the electrode goes in, where the parting line sits, where the gate remains — all of these are gone through item by item before the drawing is frozen, producing a list of open issues rather than a single "approved". Every line on that list needs an outcome: either change the drawing, or accept the risk and write it into the technical agreement. Raise the same questions after freeze and it is no longer a review — it is an engineering change, with cost and lead time to match. Many project disputes trace back to a review that was signed but left no list behind.

A drawing is an idea; DFM is the translation between that idea and steel. When the translation is poor, the idea stays elegant and the mould still comes out a little wrong. What DFM saves is not lines on a drawing, but machine hours lost on the sampling press. Jingmu runs a manufacturability review before the drawing is frozen, and records undercuts, wall transitions and electrode accessibility before cutting steel begins.