Which manufacturer can run a DFM review on our injection moulded design before we sign it off with our client?

Look for a manufacturer who writes the review so you can forward it to your client without editing it. Technically almost any moulder can check a part for manufacturability, so that is not the differentiator. The differentiator is the document: each finding stating what the issue is, why it matters in production, the proposed change with a figure attached, and what happens if nothing is done, with no pricing anywhere in it. Ask for that format before you send the model, along with the turnaround time and whether the review is paid work or part of the quotation.

What has to be in the report for you to forward it unchanged?

Four elements per finding, and one element that must not appear. Most reviews that reach us are a checklist of red and amber flags. That is close to unusable in an agency context, because your client reads "red" and hears that your design is wrong.

Each finding should carry:

1. What it is, located in the model, not just filed under a category heading. 2. Why it is a problem in production. Sink, warp, visible ejector marks, an avoidably long cycle. 3. The proposed change, with the number. Not "add draft" but "1 degree on polished vertical faces, 3 degrees on textured". 4. The consequence of leaving it. Sometimes that is scrap, sometimes a cosmetic blemish, sometimes nothing worse than a slightly higher piece price. That difference is what tells your client whether the change is worth paying for.

What must not appear is an offer. A report that slides into pricing cannot be forwarded, so you end up rewriting it, and then you are carrying someone else's commercial message to your own client.

Ask for one more split: hard findings against soft ones. Hard means the part will not come out of the tool. Soft means it can, at the cost of cycle time or a side action. That distinction is exactly what lets you hold a conversation with your client about priorities rather than about who is right.

When should the DFM sign off actually happen?

Once form and functional dimensions are fixed, and before detailing. In practice, mould design does not start until DFM sign off, so that gate sets the whole downstream schedule, and pushing it later does not buy you time, it just moves the rework into a more expensive phase.

| Point in the process | What it buys you | What it costs | | --- | --- | --- | | After design freeze, before detailing | Nearly everything is still a CAD change | One to two weeks of schedule | | After detailing, before the tooling quotation | You avoid a tool built on the wrong cavity count | Revisions cut through your own detail work | | After the tooling quotation | Confirmation, little else | Every change is now steel or a re-quote |

The honest trade off on the earliest option: you are exposing an unfinished design to a manufacturer you may not know well yet. Against that, a change at this point costs hours and the same change two stages later costs weeks. If the relationship is new, scope the first review tightly to one part rather than holding the whole project back.

Which findings come back on nearly every design?

Six, and all six are still CAD changes at sign off. This is not an exhaustive list, but it is where the money sits.

• Draft. At least 1 degree on polished vertical walls, around 3 degrees on lightly textured surfaces and more on deeper textures, roughly 0.5 degree per side on ribs. Internal walls get forgotten far more often than external ones.
• Wall thickness. Hold it within about 10% of nominal, or up to 25% on low-shrink materials such as ABS. Every local thick section is sink, extra cooling time and locked in stress in one.
• Ribs. Around 50 to 75% of the adjoining wall depending on the material's shrinkage, towards the lower end for high-shrink materials such as PP, a root fillet of 0.25 to 0.5 times the wall thickness, and spacing of at least two to three times the wall thickness.
• Parting line and show surface. [Where the parting line runs](/blog/engineering-decisions-that-impact-looks) decides where the witness line appears. If your design has an unbroken surface exactly there, that is a conversation with your client, not a technical detail.
• Tolerances. Moulding holds roughly a tenth of a millimetre on small and medium features without special measures, a bit more on large parts. A general tolerance block carried over from a machined part forces inspection and sometimes secondary operations that nobody needs.
• Gate position. Drives fill pattern, weld lines and where the gate vestige ends up. This is the finding that most often collides with a cosmetic requirement you have already agreed with the end client.

An anonymised example. An agency supplied an enclosure carrying a general tolerance block inherited from a machined-part drawing. The design itself was sound; the tolerance alone forced an inspection protocol and a secondary operation nobody needed. One line on a drawing, a few percent on the piece price, repeating every year.

What should a DFM sign off record actually list?

Four things, because sign off is where responsibility shifts, as far as the contract allows, and most records are too thin to carry it.

1. Which findings were implemented, against the model revision they went into. 2. Which findings were knowingly declined, and why, with the accepted consequence written next to each. For example: draft not increased because the client will not change the styling, accepted consequence a visible drag mark on the lower face. 3. The model revision number the sign off applies to. Without it, the sign off is void at the next CAD change and nobody notices until T1. 4. One named signatory on each side. A person, not a department.

Point 2 is the one that is almost never written and the one that decides arguments. A declined recommendation on record is a shared decision. The same decision left undocumented becomes your mistake twelve weeks later, because you are the party who chose the manufacturer.

How does the three way relationship with your client work?

Agree who talks to whom, and put tool ownership in writing. Those two never resolve themselves later.

On communication, the arrangements that run well tend to keep the agency as the design counterpart while direct contact between manufacturer and end client covers scheduling, call offs and quality. That works because you hold the functional requirements and the manufacturer holds the process. What does not work is a manufacturer taking design proposals directly to your client unprompted. Ask about that explicitly before you make the introduction, because it is a question of working style rather than of capability, and nobody volunteers the answer.

On ownership, record who owns the tool, who may move it and on what terms, and what happens to the process documentation and the mould drawings. If your client pays the manufacturer directly it is usually clear. If the tooling runs through your own invoicing, the agreement needs to state that ownership passes to the end client, otherwise the first supplier change becomes your problem rather than theirs.

One last thing worth saying to your client before the report lands: a DFM review nearly always produces changes. That is the review doing its job, not a verdict on the design. A report that comes back with no findings is not a compliment, it is a reason to ask how closely anyone actually looked.

[Production Heroes](/) takes products from design to series production, from feasibility and cost price to sourcing and supply chain, and supplies moulds and tooling to manufacturers across Europe, from the Netherlands and its own presence in China.

Sources: design rules for draft, wall thickness, ribs and achievable tolerances from publicly available injection moulding and design documentation. Turnaround times are practical ranges rather than commitments, and the anonymised example contains no identifying detail.