01
Test hypothesis
Write the engineering question, expected behaviour, pass/fail or learning criterion and the decision the result will trigger.
Evidence route
Approved protocol objective and decision rule
Prototype evidence design
Functional prototype validation: make the test, sample and conclusion traceable.
A prototype can reveal fit, motion, load, sealing, thermal or user-interface behaviour only within its configuration and test conditions. PartVerity defines the hypothesis, representative features, method, sample genealogy and permitted conclusion before results are used for design release.
Decision premise
Validation question: which controlled build and test result can answer the stated engineering hypothesis?
Engineering control
Six controls for defensible prototype validation
The test plan should expose uncertainty, not retrofit a pass statement to whichever samples performed best.
02
Representative configuration
Identify revision, process, material, orientation, post-processing, assembly parts and deliberate deviations from production intent.
Evidence route
Build traveller, BOM/configuration and delta statement
03
Sample genealogy
Maintain unique sample IDs linked to build batch, orientation, cavity/setup, rework and prior conditioning or tests.
Evidence route
Sample register and chain of test history
04
Method and fixture
Define equipment, fixture, loads, speeds, environment, conditioning, calibration and operator-dependent steps at suitable depth.
Evidence route
Released test method, setup images and equipment status
05
Actual results
Capture values, observations, failures and anomalies for all tested samples rather than reporting only a pass summary.
Evidence route
Raw results, plots/images and deviation log
06
Evidence-safe conclusion
State what was supported, what was not tested, material/process limitations and the next required build or production validation.
Evidence route
Signed report, claim boundary and design action
Decision matrix
Functional question to evidence design
A single prototype can support multiple tests only when prior testing has not altered the sample in an uncontrolled way.
| Functional question | Representative evidence | Common overclaim |
|---|---|---|
| Does the assembly fit? | Controlled interface dimensions and assembly trial | Assuming durability from one successful fit |
| Can the part carry the load? | Known material/process state and defined load test | Equating prototype properties to production material |
| Does it seal? | Surface, compression, mating parts and environment controlled | Generalising one fixture result to all operating states |
| Can users operate it? | Representative geometry/interface and documented protocol | Treating preference feedback as engineering validation |
| Is the thermal concept viable? | Instrumented conditions and declared material/process delta | Claiming final performance without production configuration |
Claim boundaries
Validation conclusions that exceed the evidence
A successful demonstration is not automatically a repeatability study.
One sample cannot describe build variation.
A test on substitute material does not prove production-material life.
Reworked samples must remain identifiable in the result set.
Technical context
Use guidance to frame the review—not to replace project evidence.
External context for proof-of-concept, looks-like, works-like and engineering prototypes. PartVerity makes each build’s evidence boundary explicit.
Formlabs rapid-prototyping guide