Key Takeaways
- Invisible processes score detection badly — surface treatment can fail completely with no visible symptom until assembly.
- Delayed failures deserve their own entries. A mode that appears at the customer is a different risk from one caught at the line.
- Include the upstream causes: resin lot, molding conditions and handling belong in a decorating FMEA even though another department owns them.
- The output is the control plan. An FMEA that does not change what is monitored has produced a document, not a control.
Process FMEA applied to marking, decorating and pretreatment lines tends to produce thin results, and the reason is usually the same: the team populates it with equipment failure modes — laser fails, conveyor jams, printhead blocks — and misses the failure modes these processes actually have, which are chemical, invisible and frequently originate upstream.
The Failure Modes That Get Missed
A useful FMEA for this class of process includes items that do not look like equipment faults.
- Treatment present but ineffective. The station is running, power and gas are within limits, and the delivered surface energy is inadequate — because the dose is marginal, coverage is incomplete on part geometry, or the electrode has degraded.
- Treatment decayed before use. The surface was treated correctly and the interval before bonding exceeded the qualified window, typically after a line stoppage.
- Surface contamination. Silicone from maintenance products, release agent transfer, airborne overspray, handling.
- Additive migration. Slip agents or plasticizers blooming to the surface between molding and decorating.
- Material variation. Resin lot change, colorant batch change, regrind fraction change, filler content drift — all of which alter mark contrast and adhesion without any process change.
- Molded-in surface variation. Knit line position, skin thickness on filled materials, gate blush, differential cooling.
- Marginal mark quality. Contrast within specification at the line and inadequate after service exposure.
Each of these has caused production interruptions in real plants, and none of them appear if the analysis is framed around machinery.
Detection Is Where These Processes Score Badly
The detection rating is the most informative column for decorating and treatment processes, because it exposes their central characteristic: the critical property is invisible.
A surface treatment station that has stopped delivering effective treatment produces parts indistinguishable from good ones. Nothing at the line looks wrong. The defect is discovered when a bond fails during assembly, or, worse, when a coating delaminates at the customer weeks later. Honest detection ratings for these modes are poor, and the discomfort that produces is precisely the point — it is the argument for adding monitoring that would otherwise be considered unnecessary.
The same applies to marking. A contrast measurement taken at the line detects an immediate defect; nothing at the line detects a mark that will fade after twelve months of ultraviolet exposure. Distinguishing those two modes, with different detection ratings, is what drives a validation and monitoring strategy that addresses both.
Separate Immediate From Delayed
Merging immediate and delayed manifestations of the same defect systematically understates risk. “Coating adhesion inadequate” as a single line item averages a failure caught by end-of-line testing with one discovered by a customer after six weeks, and the average is not a real risk.
Listing them separately produces the right ratings and the right actions. The immediate mode has moderate severity and reasonable detection, and is addressed by end-of-line testing. The delayed mode has high severity and very poor detection, and is addressed by controlling the causes — treatment interval, material specification, contamination control — and by aged testing during qualification rather than by inspection.
Include What Another Department Owns
Decorating FMEAs are usually scoped to the decorating operation, which excludes a substantial share of the causes. Resin selection, molding conditions, mold release practice, storage and handling between molding and decorating all appear in the causes column of real defects, and all of them belong in the analysis even though the decorating team does not control them.
Including them is what makes the FMEA capable of explaining the line’s actual defect history, and it usually produces the highest-value actions — a material specification clause prohibiting migratory slip agents, a maximum storage interval, a facility-wide prohibition on silicone maintenance products. Those actions require cooperation across departments, which is a reason they are resisted and not a reason to omit them.
Turning It Into a Control Plan
The measure of an FMEA is whether the control plan afterwards differs from the one before. For these processes the actions that actually reduce risk cluster into a few types.
Monitoring an invisible process: equipment status interlocked so parts cannot pass an inactive treatment station, plus periodic surface energy verification on sampled parts to catch gradual loss of effectiveness. Controlling intervals: a stated maximum between treatment and bonding, with a defined response after a line stop. Controlling material: specification clauses, lot traceability, and notification requirements for formulation changes. Measuring what matters: instrumental contrast or code grading rather than visual assessment, with a qualified measurement system. Qualifying for life, not for the line: aged testing in validation for the failure modes that appear late.
Actions that consist of “train operators” or “monitor process” without specifying what is monitored, how often and against what limit are the signature of an FMEA that has documented risk rather than reduced it.
Keeping It Honest
The most useful maintenance habit is to check every significant defect against the FMEA. If the failure mode is not there, the analysis was incomplete and should be updated. If it is there with a good detection rating and it still reached the customer, the rating was wrong. Either way the document improves, and over a couple of years it becomes an accurate description of how the line actually fails — which is considerably more valuable than the version produced in a two-day workshop before the line ever ran.
Who Should Be in the Room
The quality of a process FMEA is set largely by who participates, and decorating and treatment lines have a distinctive requirement: the people who know the causes usually sit in a different department from the people who see the effects.
A workable group covers the decorating or marking process itself, molding, materials or purchasing, quality, maintenance, and at least one operator who runs the line. The operator contribution is consistently underrated. Operators know which fixtures are awkward, which changeovers get rushed, where parts get handled, what happens after a stoppage and which defects appear that nobody formally records — and several of those are direct causes.
Where an outside process specialist is involved, their value is mostly in the failure modes nobody present has experienced yet. A team analyses the problems it has had; a specialist who has seen fifty similar lines contributes the ones it has not had, which is precisely the population the exercise is meant to anticipate.
Related Reading
- IQ, OQ and PQ for Laser Marking
- Molding Defects Behind Surface Failures
- Retrofitting Pretreatment Into an Existing Line
- Cpk and Gage R&R for Mark Contrast
- Plastics Decorating and Finishing
Need help with this?
The Sabreen Group provides independent engineering support for process FMEA facilitation and risk assessment for decorating and marking operations. Our engineering services team works with manufacturers on process development, material qualification and production troubleshooting. Contact us to discuss your application.
Frequently Asked Questions
Why do surface treatment failure modes score so badly on detection?
Because treatment is invisible. A corona or plasma station that has failed produces parts that look identical to correctly treated ones, and the failure is often not discovered until bonding or coating fails downstream, or until the customer reports delamination weeks later. Poor detection is the honest rating, and it is what justifies adding monitoring.
Should upstream causes be included in a decorating FMEA?
Yes. Resin lot variation, colorant change, molding conditions, mold release use and handling between operations are among the most common causes of decorating failures, and excluding them because another department owns them produces an FMEA that cannot explain the line’s actual defects.
How are delayed failures handled?
As separate line items from immediate failures, because their severity and detection ratings differ substantially. A bond that fails at assembly is contained; the same bond failing at the customer after six weeks carries warranty, recall and reputation consequences and is almost undetectable at the line. Merging them understates the risk.
What makes an FMEA useful rather than ceremonial?
Whether it changes the control plan. Every high-risk item should trace to a specific control — a monitored parameter, an added test, an interlock, a design change — and those controls should exist afterwards. An FMEA whose actions are all “monitor” or “train operators” has not reduced anything.
When should an FMEA be revisited?
When a new failure mode occurs that was not on it, when the process or product changes, when a control is modified, and periodically as a discipline. A recurring defect that does not appear in the FMEA is direct evidence that the analysis was incomplete, and updating it is more valuable than the original exercise.
