{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"preimage_set_characterization__chemistry_materials","arm":"CONSTRAINED_HIGH","candidate_id":"preimage_set_characterization__chemistry_materials__CONSTRAINED_HIGH","decision":"PROPOSAL","abstention_reason":null,"proposal":{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"preimage_set_characterization__chemistry_materials","arm":"CONSTRAINED_HIGH","candidate_id":"preimage_set_characterization__chemistry_materials__CONSTRAINED_HIGH","proposal_index":1,"version":0,"title":"Physical Preimage Atlas for Colliding Silicone-Cure Hardness States","problem":"A two-part silicone gasket compound is qualified by a target Shore A hardness after curing. A gasket or witness coupon within that hardness band is treated as if it came from the nominal mix ratio and cure history, although distinct combinations of base-to-curative ratio, cure temperature, and cure duration may produce the same hardness while retaining different adhesion or tear behavior.","actors":["Two-part silicone compound","Base-to-curative ratio states","Thermal cure states","Molded gasket coupons","Materials test technician","Gasket manufacturer","Sealed assembly exposed to mechanical loading"],"observable_state":"A conditioned witness coupon reads within the specified Shore A band, while its exact preparation history is uncertain or the uniqueness of the nominal history has not been demonstrated. Hardness alone therefore supplies an output condition but not a unique input state.","consequence":"If the hardness output is treated as uniquely identifying adequate cure, a colliding non-nominal state can remain indistinguishable during qualification and later exhibit unsuitable adhesion or tear behavior, compromising the sealing objective.","affected_objective":"Establish physically valid qualification evidence that distinguishes mere consistency with the target hardness from adequate gasket adhesion and tear resistance.","intervention":"Fabricate a finite, keyed coupon library covering every cell of a declared factorial input domain: three base-to-curative ratios, three cure temperatures, and three cure durations, with compound lot, coupon geometry, mixing energy, and conditioning environment held fixed. Use mechanically indexed molds and independently set heated blocks or ovens to impose the input states. After common conditioning, measure each coupon with a spring-loaded Shore A durometer fixture. Physically retain every input cell whose replicate median lies in the target hardness band as the preimage set, keep near-threshold cells in a boundary rack, and subject all target-band coupons to the same mechanical peel and tear assays. The physical library exposes hardness collisions and tests whether colliding inputs remain equivalent on properties that hardness does not preserve.","structural_mapping":[{"archetype_element":"Mapping Under Review","domain_realization":"The material transformation mapping from base-to-curative ratio, cure temperature, and cure duration to conditioned Shore A hardness."},{"archetype_element":"Output Condition or Target Value","domain_realization":"A prospectively fixed Shore A acceptance interval, such as 47 through 53 inclusive, measured after a fixed conditioning period."},{"archetype_element":"Input Domain Boundary","domain_realization":"The 27 discrete cells formed by three declared ratios, three declared temperatures, and three declared durations for one compound lot and one coupon geometry; continuous intermediate settings, other lots, and production-scale geometries are outside the domain."},{"archetype_element":"Preimage Membership Rule","domain_realization":"An input cell belongs when the median of three valid durometer measurements falls inside the fixed target interval; a cell with an unreadable specimen or excessive replicate dispersion is classified as unresolved rather than silently included or excluded."},{"archetype_element":"Candidate Input Enumeration","domain_realization":"Triplicate physical coupons are produced for all 27 cells, so enumeration is exhaustive over the declared discrete domain rather than a search for a few examples."},{"archetype_element":"Collision and Multiplicity Check","domain_realization":"The keyed input identities of all target-band coupons are compared; two or more distinct cells in the retained rack constitute an observed many-to-one collision under the hardness mapping."},{"archetype_element":"Completeness Evidence","domain_realization":"A mold-position register and surviving coupon count account for every one of the 81 planned coupons, with failed specimens explicitly recorded. Completeness applies only to the 27-cell domain."},{"archetype_element":"Boundary Case Register","domain_realization":"Cells whose uncertainty interval crosses either hardness limit, whose replicate spread exceeds the preset limit, or whose cure is too incomplete for a valid indentation remain physically segregated as unresolved boundary cases."},{"archetype_element":"Projection Loss Note","domain_realization":"Shore hardness projects a richer cured network and interface state onto one indentation response and does not itself preserve adhesion, tear resistance, or cure history."},{"archetype_element":"Downstream Use Guardrail","domain_realization":"Membership establishes only that an input cell can produce the target hardness under the coupon conditions; it does not establish that the cell caused an unknown production specimen or that it is suitable for production."}],"mechanism_mapping":[{"mechanism_slug":"preimage_table","role":"The keyed physical coupon tray is a material table linking every declared cure input cell to its directly measured hardness output.","counterfactual_removal":"Without the complete keyed tray, isolated matching coupons would not show which distinct inputs collide or whether all declared cells were tested."},{"mechanism_slug":"predicate_satisfaction_filter","role":"The fixed durometer interval supplies a reproducible physical membership predicate for retaining target-band coupons.","counterfactual_removal":"Without a prospectively fixed interval and direct hardness measurement, inclusion would become discretionary and the preimage would not be well defined."},{"mechanism_slug":"fiber_cardinality_count","role":"Counting distinct keyed cells in the target-band rack determines whether the hardness output has zero, one, or multiple inputs in the declared domain.","counterfactual_removal":"Without the count, the experiment could find examples but could not establish multiplicity within the enumerated domain."},{"mechanism_slug":"witness_and_counterexample_set","role":"Target-band coupons are hardness witnesses, while their peel and tear specimens can become counterexamples to the assumption that equal hardness implies equivalent mechanical suitability.","counterfactual_removal":"Without preserving and mechanically challenging the witnesses, collision would be visible but its relevance to the sealing objective would remain untested."},{"mechanism_slug":"coverage_completeness_audit","role":"Physical position keys, specimen counts, and unresolved racks distinguish exhaustive coverage of the finite grid from missing or invalid cells.","counterfactual_removal":"Without coverage accounting, an apparently unique preimage could merely reflect omitted input cells."}],"causal_chain":["Distinct ratio-temperature-duration combinations produce different crosslink-development histories in the silicone coupons.","The hardness measurement compresses those histories into a single indentation response, permitting physically distinct inputs to occupy the same target band.","Exhaustively fabricating the bounded factorial grid causes every declared input cell to exist as a testable material witness.","Direct durometer measurements identify which physical witnesses satisfy the target-output predicate.","Keyed retention and counting reveal whether the output has multiple input states rather than a unique inverse.","Peel and tear loading of every target-band witness tests material properties lost in the hardness projection.","Observed disagreement among equal-hardness witnesses refutes the use of hardness as sufficient evidence of equivalent gasket suitability within the tested domain."],"baseline":"Test the nominal cure recipe, confirm that one or a few witness coupons meet the Shore A interval, and interpret that output as evidence of the intended cure state without enumerating alternative ratio-temperature-duration combinations or checking their multiplicity.","nearest_rivals":["Add infrared spectroscopy or calorimetry to estimate cure conversion; this supplies richer state measurements but does not by itself enumerate every declared process input mapping to the hardness band.","Optimize ratio and cure conditions directly for hardness, peel strength, and tear strength; this can find a preferred recipe but may leave other target-hardness inputs and collision structure uncharacterized.","Tighten metering and oven controls around the nominal recipe; this reduces process variation but does not test whether an unknown equal-hardness specimen has a unique history.","Add a second scalar acceptance measurement to hardness; this may separate some collisions but does not establish the complete preimage unless the bounded input domain is still exhaustively tested."],"remaining_contrastive_claim":"Relative to nominal-recipe confirmation, the coupon atlas can establish the complete set and multiplicity of input cells producing the target hardness within the declared 27-cell domain. It does not claim completeness over continuous settings, other material lots, other geometries, or production equipment, and it does not identify the actual history of an unknown specimen without additional evidence.","authority_safety":{"decision_authority":"The materials laboratory manager may authorize the bounded coupon experiment; production recipe changes and gasket release remain with the separately designated manufacturing and product-safety authorities.","authorized_first_step":"Prepare and test only the 81 bench-scale coupons in the declared 27-cell grid using existing small-specimen mixing, curing, indentation, peel, and tear equipment.","excluded_actions":["Changing a production cure recipe","Releasing or rejecting production gaskets based solely on the coupon atlas","Claiming that preimage membership identifies the actual cause of an unknown specimen","Extrapolating completeness beyond the declared discrete domain","Using unreviewed temperatures or chemicals outside existing laboratory equipment limits","Testing the intervention on occupied or safety-critical assemblies"],"halt_rollback":"Stop if heating control leaves its approved range, uncured material cannot be safely contained, specimen identities are lost, or the planned cell count cannot be reconciled. Isolate the affected material, discard the invalid run under existing waste controls, clean the fixtures, and restart only with newly keyed coupons; no production state has been changed."},"negative_tests":{"strongest_counterevidence":"Every valid target-band coupon comes from one input cell within repeatability limits, or all colliding target-band cells also remain indistinguishable in the peel and tear assays. Either result weakens the claimed practical importance of a hardness preimage collision for this bounded domain.","problem_falsifier":"Independent physical traceability already proves the exact ratio, temperature, and duration of each relevant gasket, and validated evidence shows that the specified hardness band is injective over the full admissible process domain or sufficient for all sealing-relevant properties.","intervention_falsifier":"Repeated fabrication changes cell membership enough that the preimage cannot be reproduced, or coupon-scale mappings fail to predict the corresponding molded-gasket measurements even after geometry is held within the stated scope.","risks":["Hot curing fixtures and ovens can cause burns.","Uncured silicone components may create contact or waste-handling hazards.","Mixing or mold-position errors can create false input labels.","Temperature or thickness gradients can be mistaken for input-state effects.","Destructive peel and tear tests consume witnesses and require retained duplicates.","A coarse discrete grid can miss collisions between tested settings.","Users may overgeneralize finite-grid completeness to continuous settings, new lots, or production geometry.","Durometer repeatability and edge effects can create unstable boundary membership."]},"next_evidence_step":"Run one bounded 81-coupon experiment: three replicates at each of the 27 ratio-temperature-duration cells, randomized across keyed mold positions. Before measurement, fix the hardness limits, replicate-dispersion limit, indentation locations, and unresolved-state rule. Report only raw cell identities, direct hardness readings, coverage gaps, preimage cardinality, and peel/tear outcomes for target-band cells; do not infer production suitability from this first run.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"Not assessed because runtime isolation excludes access to prior proposals; the candidate is derived only from the supplied archetype and chemistry-and-materials domain card.","revision_record":{"parent_version":null,"progress_targets_addressed":["Preserve the archetype's many-to-one inverse-set structure in a concrete materials problem.","Make the operative intervention physical and measurement-based rather than computational or procedural.","Bound completeness to a fully enumerable finite input domain.","Separate output compatibility from causation and production suitability."],"conceptual_changes":["Instantiated the abstract output as conditioned Shore A hardness and the inputs as silicone mixing and thermal-cure states.","Made projection loss testable through destructive adhesion and tear measurements."],"operational_changes":["Specified a 27-cell factorial domain with triplicate physical coupons.","Added keyed molds, direct durometer measurements, physical witness retention, boundary segregation, and bounded mechanical assays."],"evidence_changes":["Defined physical coverage accounting and explicit unresolved cases.","Added reproducibility and scale-transfer falsifiers."],"claim_changes":["Limited completeness to the declared discrete grid.","Excluded novelty, effect-size, prevalence, production-suitability, and actual-cause claims."]},"substrate_contract":{"primary_allowed_process":"MEASUREMENT_INSTRUMENTATION","counterfactual_independence":"The essential effect is the creation, curing, indentation, retention, counting, and destructive mechanical testing of physically distinct material coupons. If all software, algorithmic inference, databases, dashboards, reporting systems, incentives, authorization rules, and procedural enforcement are removed, the imposed material states still cure differently, the spring-loaded durometer still produces direct mechanical readings, equal-hardness coupons still coexist as tangible collision witnesses, and peel or tear loading still exposes physical differences. Those wrappers may help label or communicate the experiment but do not generate its inverse-set evidence.","forbidden_channel_audit":"No algorithm, model, database, recommender, information-routing system, or software control loop determines membership or produces the claimed effect. No incentive, permission, review workflow, training program, or accountability arrangement substitutes for the material assay. Temperature may be imposed by ordinary independently set heating equipment, and measurements are direct mechanical outputs rather than sensors whose operative value depends on downstream analytics. The only governance elements limit authority and hazards; removing them would be unsafe but would not eliminate the physical causal mapping or its observable collisions."}}}