An OEM mold supplier should be evaluated through measurable manufacturing data rather than machine count alone. Review machining accuracy, mold-size limits, steel hardness, CMM capability, trial-machine tonnage, cavity consistency, delivery history, and comparable finished molds. ISO 20457:2026 specifically addresses dimensional and geometrical tolerances for molded plastic parts and recognizes shrinkage, warpage, material behavior, and processing conditions as sources of dimensional variation. For a precision project, request inspection reports from at least 3 comparable molds and verify whether critical dimensions, tooling materials, mold trials, and engineering changes were recorded. Claims about ±0.01 mm capability should always be supported by measurement records.
A supplier assessment should begin with the part drawing, resin, annual volume, expected mold life, cavity count, surface specification, and production press. A 2-cavity housing molded from unfilled ABS creates a very different manufacturing requirement from a 16-cavity connector using 30% glass-filled PA66. ISO 20457:2026 replaced the 2018 edition in August 2026 and provides a plastics-specific framework because dimensional variation is affected by shrinkage, anisotropic behavior, cooling and warpage rather than machining tolerance alone. Ask the supplier to mark every dimension that requires special process control before discussing price.
The next review should cover engineering work before steel is cut. Request one complete DFM package from a comparable project and check whether it includes draft angles, wall-thickness changes, parting lines, gate position, ejector locations, cooling channels, weld-line locations, venting and steel-safe dimensions. For a project with 20 controlled dimensions, the engineering team should identify which dimensions depend mainly on mold steel and which depend strongly on molding conditions. A supplier that approves all 20 without discussing resin shrinkage or machine conditions is giving too little manufacturing analysis.
Ask the supplier to explain the three dimensions most likely to move after T0 and what adjustment method has been reserved in the mold. A useful answer should identify the dimension, expected direction of change, steel allowance and measurement method rather than saying that adjustments can be made later.
Once the design approach is acceptable, review CNC equipment by capability rather than quantity. Record machine travel, spindle speed, positioning specification, workpiece capacity, machine age and inspection method for each machine intended to produce the cavity or core. If a cavity plate measures 900 × 700 mm, confirm that the main cutting operations can be completed without unnecessary repositioning. For a mold containing 40 precision inserts, ask how many setups each insert requires and whether locating datums remain consistent through roughing, finishing and inspection.
A machine accuracy specification is still not enough because tool wear, spindle condition, thermal growth and fixture repeatability affect the finished component. Request dimensional reports from at least 3 inserts made during the previous 12 months. Compare the drawing tolerance with the actual measured values instead of accepting a general statement such as “our CNC accuracy is ±0.005 mm.” If 10 dimensions were specified at ±0.01 mm and several measurements repeatedly sit within 0.002 mm of the limits, process margin is small even though every part technically passed inspection.
EDM capability should then be checked because deep ribs, narrow slots, sharp internal corners and hardened features may not be practical to finish by milling. Ask how electrodes are designed, numbered, inspected and stored. If a tool uses 60 electrodes, the supplier should be able to connect each electrode to a drawing revision or digital record. Review the spark gap, finishing sequence and surface requirement for comparable parts. A supplier that measures electrodes before EDM has more control over dimensional transfer than one that relies mainly on final bench fitting.
Wire EDM and grinding require the same level of review. For hardened inserts, ask how many finishing cuts are normally used, what inspection follows wire cutting, and how flatness or parallelism is checked after grinding. A dimension of 25.00 ±0.01 mm should not be treated as sufficient information when two mating surfaces also require tight parallelism. Inspect at least 5 finished inserts from current production and compare edge condition, surface finish and repeatability between parts rather than selecting one presentation sample.
Measurement capacity should match the dimensions being sold. A factory quoting precision tooling should have suitable CMM, optical, height, hardness and conventional measuring equipment, together with documented calibration. ISO 20457:2026 notes that plastic dimensional control is influenced by material, product design, tool layout and processing conditions, so mold steel measurements and molded-part measurements need separate control plans. Ask which dimensions are checked after machining, after assembly and after molding. For a 32-cavity mold, measuring only cavity 1 provides very limited information about cavity-to-cavity variation.
| Item to verify | Practical audit request | Weak response |
|---|---|---|
| Critical mold dimensions | Reports from 3 recent comparable molds | Verbal tolerance claim |
| Cavity consistency | Data from 5–10 cavities | One selected cavity |
| CMM control | Calibration record and sample report | “CMM available” |
| Steel hardness | Certificate plus measured hardness | Material name only |
| T0 performance | Trial report with process settings | Sample photographs |
| Delivery | Previous 10 project dates | Quoted lead time only |
Steel control comes next because machining precision cannot compensate for incorrect tooling material. Request material certificates, heat numbers and hardness records for cavity and core components. HASCO lists AISI 420-equivalent 1.2083 mold steel at a working hardness of 50–54 HRC and describes it for corrosion-resistant cavity plates and inserts, illustrating why grade and heat-treatment condition must be stated rather than using a broad description such as “stainless mold steel.” For abrasive engineering resins, compare steel selection, hardness and replaceable wear areas with the planned production volume.
Heat treatment also needs a documented sequence. Ask whether rough machining, stress relief, hardening, semi-finishing, EDM, grinding and final inspection are planned before manufacturing begins. If an insert is hardened to roughly 50 HRC or above, dimensional movement after treatment may require finishing allowance. Review 3 heat-treated components and compare dimensions before and after treatment where records are available. A Engineering plastic injection molding supplier handling glass-filled or high-temperature resins should be able to explain why its selected steel, hardness and surface treatment fit the resin rather than selecting material mainly from mold price.
The audit should then move from individual components to assembly. Observe at least 2 molds currently being fitted and look for heavy manual grinding, undocumented welding or repeated correction of supposedly finished inserts. Some spotting and fitting are normal, especially around shutoffs, sliders and lifters, but extensive hand correction can show that earlier machining stages are not holding consistent references. Ask whether replacement inserts can be manufactured from released CAD data and installed with minimal fitting. For molds expected to run for several years, interchangeability reduces maintenance time when wear components need replacement.
Cooling deserves separate review because the mold may pass dimensional inspection and still perform poorly in production. Compare water-line layout with thick sections, deep cores and areas that are likely to retain heat. Ask for flow or temperature information from at least one similar mold rather than accepting a drawing alone. If a molded part uses a 30-second cycle and cooling occupies 18 seconds, cooling represents 60% of the cycle. Reducing avoidable temperature imbalance can therefore affect output far more than saving a few hours during mold machining.
Mold-trial capacity should be checked against the intended production press. Record available clamp tonnage, tie-bar spacing, shot capacity, mold thickness range, injection pressure, hydraulic functions and hot-runner controls. If the production mold is intended for a 500-ton press, T0 on equipment with substantially different clamp behavior or injection capacity may not reproduce production conditions closely enough. Request trial sheets showing barrel temperatures, mold temperature, injection speed, holding pressure, cooling time, cycle time and part weight for at least 3 trial rounds.
T0 quality should be judged by the type of correction required, not by whether the first molded parts look acceptable. Gate adjustment, vent improvement or a small steel-safe correction can occur during development; major interference, incorrect slide movement, severe flash, broken lifters or large cavity corrections point to earlier manufacturing or design errors.
For multi-cavity tooling, inspect variation instead of average performance. A 16-cavity mold can produce an acceptable average dimension while individual cavities sit near opposite tolerance limits. Request dimensional data across at least 5 cavities, and for high-volume tools ask for all cavities on selected controlled dimensions. Part weight can provide another comparison: if average shot weight per part is 8.00 g, a spread from 7.90 to 8.10 g represents about 2.5% between the lightest and heaviest parts and deserves investigation before approval.
Capacity should also be tested against actual workload. A factory with 25 CNC machines can have less available production time than one with 12 machines if utilization is already high. Ask for the current number of active molds, monthly completed molds, engineering headcount, moldmaker headcount and machine loading during your planned build period. Review the previous 10 projects and compare promised T0 dates with actual T0 dates. A supplier reporting 9 of 10 trials on schedule provides more useful information than a statement that normal lead time is 6 weeks.
Project control becomes more important once machining capacity is confirmed. Request a sample schedule showing design approval, steel ordering, rough machining, heat treatment, finishing, EDM, assembly, T0, correction and final validation. Each milestone should have an owner and date. Review how drawing revisions are issued to the shop floor; one obsolete cavity drawing can create days of rework. For a project with 80 manufactured components, revision status should be visible without relying on operators remembering which CAD file was discussed in an email.
Past work provides another useful test. Select 3 molds similar in resin, size, tolerance and cavity count rather than asking to see the supplier’s largest or most attractive tools. Compare inspection reports, T0 records, mold condition and production samples. If your project requires 500,000 cycles, ask what changed in steel choice, wear components, guiding, ejection and spare-part planning compared with a 100,000-cycle mold. The answer should describe actual construction differences rather than repeating the requested shot-life number.
A final supplier comparison can use weighted manufacturing data instead of an overall impression.
| Evaluation area | Suggested weight |
|---|---|
| Engineering and DFM | 20% |
| CNC, EDM, wire EDM and grinding | 20% |
| Inspection and measurement | 15% |
| Assembly and trial molding | 15% |
| Comparable mold history | 10% |
| Capacity and delivery record | 10% |
| Material and heat-treatment control | 5% |
| Documentation and communication | 5% |
For a precision 32-cavity mold, machining and measurement can reasonably exceed 40% of the score; for a large single-cavity automotive tool, machine travel, crane capacity, spotting press size and trial tonnage may deserve more weight. Keep quotation price outside the technical score until specifications are normalized. Two suppliers can differ by 15% in price simply because one quote includes hardened inserts, branded standard components, dimensional reports and multiple trials while the other does not.
Before issuing the purchase order, request one package containing the released mold design, steel list, process schedule, inspection plan, trial plan and defined acceptance conditions. ISO 20457:2026 states that functionally required tolerances should be specified in product definition data and checked against manufacturing capability, rather than applying metal-part tolerance assumptions to molded plastics. The supplier should be able to connect every important manufacturing claim to a machine, measurement record, material certificate, trial report or comparable finished mold.