Most mold shops still sign off on fault tolerance by feel—a senior toolmaker walks the parting line, taps the core, and declares the tool "good enough." That works until the hidden rework shows up three weeks later as flash at the shutoff or a sticking ejector. The alternative is to treat fault tolerance as a measurable acceptance item. The R1–R10 criteria framework does exactly that: each criterion is a hard, checkable judgment rather than an opinion, and the seven-check method is how those criteria get executed on the floor before the first trial shot. In practice this means the tool's risk is quantified while it is still on the bench, not discovered after T1 samples come back out of tolerance.
What makes the seven checks usable is that they map onto real mold components and real numbers. Dimensional checks on cavity and core inserts, shutoff and parting-line fit, ejector stroke and return, cooling circuit flow and pressure, hot-runner or gate condition, guide-pin and bushing clearance, and clamping/locating features each get a pass/fail threshold tied to the R-criteria. A typical example: parting-line contact is verified by blue-fit or feeler gauge rather than eyeball, and cooling is verified by flow rate against the design value—not by "water comes out." When a check fails, the rework is logged as a data point with a criterion reference, so the next tool in the family inherits the fix instead of repeating the mistake. This is where fault tolerance stops being tribal knowledge and becomes a shop standard.
The payoff is front-loaded: problems that used to surface as hidden rework—mismatched shutoffs, under-sized ejector clearance, marginal cooling—get caught at the bench, where a fix costs hours instead of a production line stoppage. For engineers building out this routine, the discipline is to keep the criteria fixed and the evidence recorded, so acceptance is defensible to the customer and repeatable across projects. For more mold sourcing and supplier information, visit MoldSo at www.moldso.com.