R&D dreads hearing "let's tweak the structure a bit," while purchasing dreads tallying up the cost that follows. A mold is not a digital model in a computer—what gets changed is already-hardened steel. A shift in wall thickness, a gate location, or a parting line propagates downstream through design, machining, trial molding, and documentation.
Take the lightest example: simply shifting one mounting hole by 2 mm. It sounds like "just a drill," but if that hole sits near a slide or ejector-pin location, the shift may eat into ejection space or cut through a cooling line—which then forces a structural redesign, a re-run of mold-flow analysis, re-EDM, re-trial-molding, and a re-issued inspection report. On the drawing it's one line; on the shop floor it's a whole chain.
The industry's standard ECN (Engineering Change Notice) process is typically 7 steps: intake & registration → technical review & impact assessment → mold-modification plan & risk confirmation → quotation & lead-time assessment → machining execution → trial-mold verification & sample approval → technical documentation update & mass-production handover (Source: public ECN process materials from Hongren Weier and CK Mold). Note step 1: a verbal explanation cannot serve as the execution basis; a formal design change order must carry the old/new 3D/2D drawings, revision number, change reason, and affected cavity numbers—skip this step and rework is most likely to recur later.
Why does a later change cost more at an accelerating rate—not linear addition, but each change forces a re-run of all validations already completed? Change it at the drawing stage and you change the model; change it after T0 and you change steel, also disturbing machined surfaces and electrodes; change it after mass production and you add downtime, WIP, and disposition of already-delivered goods. So what truly saves money is not "changing fast" but "changing early"—push the risk-identification step as far upstream as the DFM stage.
Discipline | Specific actions | Easy to miss |
① Design & Simulation | Re-run DFM, mold-flow analysis (for major changes), re-check shrinkage & warpage | Changing only the drawing without re-running simulation lets new defects surface only at T0 (for major changes) |
② Steel & Electrodes | Re-order material, new inserts, re-heat-treatment | Old and new steel grade & hardness must match, or service life becomes uneven |
③ Machined Surfaces | Update structure, slow-wire EDM / EDM, re-polish | When the change area is near a slide, lifter, or cooling line, the structure must be changed in sync |
④ Trial-Mold Verification | Re-trial-mold, sampling, first-article full inspection, customer sample approval | Use the same material batch and original process baseline, or the data isn't comparable |
⑤ Document Revision | Mold drawing, spare-parts list, parameter table, mold history—all bumped to new revision | If the old version isn't marked obsolete, the production side may use the wrong version |
(Source: public materials from CK Mold, Hongren Weier)
Current and proposed 3D/2D drawings, with the new revision number, with changed areas identifiable;
2.Change reason—it determines how to select the safest mold-modification process;
3.Which cavities / areas are affected, to avoid fixing only the primary change and missing secondary areas;
4.Written confirmation of cost and lead time—no work begins before confirmation;
5.Re-sample and sample-approval plan, plus how old approved samples are identified, isolated, or retained;
6.Switchover batch or date between old-version WIP, stock, and the new version.
Insufficient shrinkage estimation: PP, ABS, PA and other materials shrink differently; without flow analysis, critical finished dimensions easily fall outside the tolerance band;
Gate location: splay marks, weld lines, and appearance-surface defects often originate here; changing the gate means changing filling and venting;
Multi-material / overmolding bonding: e.g., TPE over PP without designed mechanical interlock tends to delaminate in production, and retrofitting structure afterward is very costly.
The industry's common accountability logic is "responsibility follows the cause": when the customer proactively adds a function or changes a dimension, it is usually charged separately; corrections made to conform to an already-approved drawing are generally borne by the mold maker. The most disputed is the third type—the part conforms to the drawing, yet assembly or use runs into problems. What speaks then is the DFM record: what risks were identified at the time, who approved them, and where the conclusion was written. So DFM is not just a technical document; it is also a commercial voucher.
Conduct DFM review and obtain written sign-off before drawing freeze: check item by item across five points—wall thickness, draft, undercut, tolerance, shrinkage rate—and output a risk list; review conclusions must land in drawing revisions, not mere verbal confirmation.
2.Treat the critical-dimension and tolerance table as the frozen baseline: when a change occurs, first compare against the baseline to clarify which area is affected and whether assembly and appearance are implicated.
3.Spell out change-control rules in the contract: quotation scope, how lead time extends, how many trial molds are included, how old samples and WIP are handled, and whether final-payment milestones shift with the change (common industry principle per CK Mold public materials; subject to the actual contract between parties).
Controlled release of the new drawing with the old version marked obsolete; sync-update the process parameter table, spare-parts list, maintenance manual, and mold history; archive the trial-mold sample-approval version; record the change content and machining parameters into the per-mold history. No document revision means the change isn't finished—if the production side holds the old drawing, all prior validation was for nothing.
Jingmu outputs a risk-point list at the DFM stage, and for any design change issues an impact assessment before work begins.
+86 13724506670