Engineering Change After PPAP: Modifying a Running Part Without Restarting the Programme

Most tooling programmes are managed well up to the first article approval and then managed by email. The tool is running, the part is in production, and a request arrives from the customer's engineering team, from a regulation, or from a resin supplier discontinuing a grade. The request is small, everyone agrees it is small, and it is handled as a favour between two people who have worked together for a year. That is the pattern that produces a mixed shipment.

At Hangzhou Xucheng we run series production for automotive interior parts, and the changes that cause trouble are rarely the technically difficult ones. They are the ones that were not classified, not costed, not documented, or not tied to a cut-in date. This article sets out how a post-PPAP change should be handled so that the new version arrives at a known point rather than appearing gradually.

Why a change after PPAP is a different problem

During development, a change is free in the sense that nothing has been shipped. The tool is not yet approved, the customer has not yet built the part into a vehicle, and the only cost is engineering time. After PPAP, every one of those conditions is reversed. There is approved tooling that must be altered, there may be a customer line that has validated the old part, there is finished stock that cannot be mixed with the new version, and there is a paper trail that has to be reissued.

The most expensive mistake is not a bad technical decision; it is a good technical decision made without a cut-in date. A tool that is modified on a Tuesday and starts producing the new version on Wednesday will ship a container that contains both versions, and the customer will discover it at goods-in inspection, which is the worst possible place to discover it.

Classify the change before you cost it

Classification is the step that determines everything downstream: who approves, what is re-validated, and whether the change needs a new PPAP or a deviation. It should be done by the customer's quality function together with the supplier's, against written criteria rather than against a feeling about how big the change is.

The value of the table is not that it is universal — every customer has its own criteria — but that it forces the question to be asked explicitly. A supplier and a customer who classify a change in the same way, in writing, before any steel is touched, have removed the most common source of post-change dispute.

The ECN loop: who signs what

An engineering change notice is a short document with a long reach. At minimum it should carry the change description, the reason, the affected part numbers and drawing revisions, the classification, the tooling action, the validation plan, the cut-in date and the sign-offs. The sign-offs are the part that programmes skip, and they are the part that makes the document useful six months later when someone asks why the part changed.

• Customer engineering — confirms that the change is what they intended and that the drawing revision is correct.
• Customer quality — confirms the classification and the re-approval scope.
• Supplier engineering — confirms the tooling action and the technical feasibility, including any effect on cycle time or yield.
• Supplier quality — confirms the inspection plan and the control of old and new stock.
• Programme management — owns the cut-in date and the stock reconciliation.

A change without the quality signatures is a change without a containment plan, and containment is where most post-change problems actually occur.

Steel changes: weld, insert or recut

The tooling action follows from the change, and the three options have different lifetimes. Welding is the fastest and is appropriate for a small, one-time correction on a tool that will not run for millions of shots. An insert is the right answer when the feature may change again or when two variants must be produced from one mold, because it makes the next change cheaper. Recutting — re-machining the cavity or core — is the most durable and the most expensive, and it is usually reserved for a change that is effectively permanent.

The choice should be made with the remaining tool life in mind. A weld on a tool with 80 per cent of its life remaining is a false economy if the tool will run for another five years; an insert on a tool that is near the end of its life is over-engineering. Asking "how many shots are left" before choosing the method is the discipline that keeps both the cost and the risk proportionate.

Re-approval scope: what has to be re-run

The instinct after a change is to re-run everything, which is safe and expensive, or to re-run nothing, which is cheap and dangerous. The productive answer is a defined scope that follows from the classification and from the change itself.

A cosmetic change normally needs a colour or gloss verification and a visual master update. A minor dimensional change needs a first-article inspection on the affected features plus a capability study if the feature is a mounting point. A major change — interface, wall thickness, material — needs the full dimensional layout, the performance tests that the original PPAP relied on, and a new PPAP submission. The scope should be written into the ECN before the tool is touched, because afterwards it becomes a negotiation about whether the result is acceptable rather than a plan for proving it.

Inventory, containment and cut-in dates

The cut-in date is the single most useful number in a post-PPAP change, and it is meaningless without a stock count at both ends. The supplier needs to know how many old-version parts are in its finished goods, in work in progress and in the warehouse; the customer needs to know how many are in its own stock and in transit. The cut-in date is then set at the point where the old stock is exhausted, not at the point where the tool is ready.

Containment is the mechanism that keeps the two versions apart in the meantime. In practice that means segregated storage, labelled packaging, a written record of which cartons carry which version, and a rule that no mixed pallet leaves the factory. It also means agreeing what happens to the old-version stock that remains after the cut-in date: rework, scrap, or use-up under deviation. Each of those has a cost, and the cost belongs in the change approval rather than in a later argument.

When the change comes from the resin supplier

Not every change is initiated by the customer. A resin supplier may discontinue a grade, change a formulation within its specification, or move production to another plant, and any of those can alter shrinkage, colour or flow behaviour enough to require a tool or process adjustment. The productive response is to treat the notice as an ECN with the same classification and sign-offs, and to run a trial on the new grade before the old one is exhausted: a shrinkage change of a few tenths of a percent is invisible on a small part and significant on a door panel, and the only way to know which applies is to mould the part and measure it.

Documenting the change so the next one is easier

A change that is properly documented makes the next change cheaper, because the tool's history becomes readable. The useful record is short: the ECN, the revised drawing, the tooling action taken with photographs, the trial results, the inspection report and the new PPAP or deviation. Filed against the part number rather than against the project, it answers the questions that arrive two years later when the tool is transferred, the original engineer has left and someone needs to know why a boss is welded.

That record is also what makes a tool transferable. A tool with a documented change history can be moved to another molder with a reasonable chance of reproducing the part; a tool with a welded boss and no explanation cannot. The ownership terms that govern that transfer are set out in our article on OEM, ODM and private label arrangements.