Updated
Start with the decision the test must support: approve packaging, compare thermal layouts, measure structural response or release a production process. Then choose the least expensive prototype that can answer that question without hiding a critical difference. This guide focuses on component and subsystem prototypes, including parts installed in development vehicles.
Match automotive prototypes to the decision being tested
Automotive prototypes should reproduce the characteristics that determine the test result. Appearance models support styling decisions; packaging models check space and access; functional parts evaluate operating behavior; production-representative samples assess process-dependent performance. A result is transferable only when the relevant differences from the intended production design are understood and documented.
An appearance model does not need the fatigue behavior of a suspension component. A thermal demonstrator may need the intended conductivity and contact pressure but can tolerate a simplified nonfunctional exterior. A serviceability model must reproduce tool access, harness routing and removal paths even if its walls are thicker than production.
Write a short test-purpose statement before requesting parts. Name the decision, the measured response and the features that cannot be substituted. Mark everything else as negotiable. This gives the manufacturer room to simplify a part without weakening the experiment. It also prevents a visually convincing sample from being presented as evidence for a load case it was never built to reproduce.
Choose the manufacturing route by fidelity and revision risk
CNC machining, additive manufacturing, casting and molding answer different development questions. Tooling commitment matters, but so does the cost of drawing the wrong conclusion from a convenient material substitute. Compare the full route, including secondary machining, finishing, inspection and assembly.
For manufacturing options, review CNC machining and 3D printing. The following matrix is a selection aid, not a promise of material equivalence or supplier capability.
Process selection matrix
A process can be appropriate for one feature and misleading for another feature on the same part. Identify which differences are acceptable before choosing a manufacturing route. The main distinction is what the prototype proves, rather than how polished it looks.
| Route | Useful evidence | Difference to investigate | Revision consequence |
|---|---|---|---|
| CNC from metal or engineering plastic | Mating geometry, mounting interfaces, selected functional tests | Stock condition may differ from a casting, forging or molded part | CAD changes often avoid new hard tooling |
| Polymer 3D printing | Packaging, routing, assembly access and selected functional trials | Build direction, resin or powder grade, porosity and conditioning | Good for frequent geometry changes |
| Metal additive manufacturing | Complex passages and geometry trials | Microstructure, defects, support removal and heat treatment | Qualification and finishing may dominate the schedule |
| Urethane casting | Repeated appearance and handling samples | Cast polyurethane may differ from the intended thermoplastic | Master and mold changes create additional work |
| Prototype injection molding | Molded geometry and process-sensitive polymer behavior | Tool design, gate location and settings may still differ from production | More useful after core geometry stabilizes |
Do not select CNC automatically for every load-bearing prototype or printing automatically for every complex shape. A machined substitute for a cast component can change the very failure mechanism under investigation. NIST material-characterization work also identifies the importance of reliable material-property data for additive parts.

Control material substitutions before functional testing
A familiar material name is insufficient when stiffness, heat exposure, moisture or fatigue drives the test. Record the grade, product form and treatment state. For polymers, identify conditioning and processing history where relevant; for metals, establish alloy, temper or heat-treatment condition and the evidence required for identification.
A printed nylon duct may help confirm routing and attachment while leaving long-term heat and fluid resistance unresolved. A machined plastic clip can reveal access problems without representing injection-molded fiber orientation or weld lines. Treat those as explicit test limitations rather than calling every sample production-like.
Keep an equivalence register: intended production condition, prototype condition, affected measurements and the engineering owner who accepts the difference. When a substitution influences the result, either change the prototype or narrow the conclusion. NIST part-qualification research is a useful reference for the distinction between making a part and establishing evidence that it is suitable for a defined application.
Plan interfaces and instrumentation before releasing CAD
A prototype becomes a useful test article through its interfaces. Bearing seats, sealing lands, bolt patterns, connectors, ground points and sensor mounts deserve clearer requirements than surfaces that only occupy space. Define the functional datum scheme so the part can be mounted and measured consistently.
Provide the mating components or interface drawings where available. A bracket measured on a bench can appear compliant yet misalign once installed against a distorted mounting face. Specify the assembly condition when tightening, gasket compression or installed support affects the measurement.
Instrument access belongs in the design review: thermocouple holes, cable exits and accelerometer pads can change stiffness, leak paths or heat flow. If a sensor mount is added only to the prototype, document its possible influence. For electrical interfaces, SAE/USCAR-2 provides a connector-system performance reference; it does not qualify unrelated housings or establish whole-vehicle compliance.
Build a validation sequence that isolates expensive failures
Perform cheap checks before consuming a scarce development assembly. Confirm identity, fit and instrumentation function before running the operating test. This sequence catches a wrong revision or inaccessible fastening feature without confusing it with a durability failure.
A practical release sequence is incoming dimensional review, trial assembly, sensor and fixture verification, operating-condition testing, then teardown. Each stage should have an owner and a disposition rule. Photograph the setup and preserve the data needed to connect a result to a specific specimen.
Choose environmental tests from the expected failure mechanism. A vehicle bracket exposed to road salt may need a cyclic corrosion assessment, while another component may be driven by vibration or thermal cycling. SAE J2334 describes a field-correlated laboratory corrosion procedure within its stated scope. It is not a substitute for every service environment, nor does a passing component test prove the vehicle meets all applicable requirements.
Keep revision and specimen records usable after the test
A result without an identifiable build condition cannot reliably inform the next design. Give each specimen a durable identifier and connect it to the CAD revision, drawing revision, material record, finishing route and inspection report. Record any approved deviation before the test rather than reconstructing it afterward.
Separate drawing nonconformities from planned prototype substitutions. The first is a departure from the agreed build specification; the second is part of the experiment. Combining them in a single pass/fail field hides whether the design, supplier or test setup needs attention.
Store baseline and changed specimens separately when comparing revisions. If geometry and material change together, the measured improvement cannot automatically be assigned to geometry. Decide whether that ambiguity is acceptable or whether a controlled comparison is needed. AIAG quality core tools connect planning, measurement and production approval; the actual submission package remains customer-specific. An early prototype inspection report should not be relabeled as an approved production submission.

Move from prototype evidence to production readiness
Release the production route only after identifying what early samples did not represent. For a molded housing, that may include shrinkage, gate effects, molded-in stress and tooling-dependent features. For a cast bracket, it may include defects, surface condition and local section behavior.
A useful hypothetical example is an EV sensor enclosure that passes a fit check as a printed sample. The next build may confirm sealing on a machined part, but production molding can introduce different flatness and wall behavior. Those are separate questions, so retain the fit result while commissioning the additional evidence.
The injection molding route can become appropriate when the program needs molding-dependent results. For U.S. vehicle programs, NHTSA regulations and FMVSS information establish the regulatory framework. Manufacturing a prototype, issuing a material certificate or completing a bench test does not independently establish vehicle certification. Assign regulatory decisions to the responsible vehicle organization and its qualified reviewers.
Specify the prototype package that makes a quote meaningful
A useful RFQ communicates the experiment, not just the solid model. Include quantity by revision, intended test, material condition, critical interfaces, surface requirements and requested inspection evidence. State the date the part must be available for assembly as well as the planned test date.
Ask the manufacturer to identify assumptions and proposed substitutions in writing. Compare quotations on the same delivered condition: machining, treatment, masking, inspection, marking and packaging. A cheaper unfinished part may create additional work at the lab and delay the test even when the quoted manufacturing lead time is shorter.
For automotive development projects, 6CProto offers multiple manufacturing routes and DFM review. Use that review to discuss tool access, feature feasibility and the differences the test can tolerate. Reserve tight requirements for the interfaces that determine the result; avoid spreading a single precision value across an entire component without a functional reason.
Make the next build answer a narrower question
The strongest prototype plan connects a test purpose to a controlled build condition and an acceptance decision. It also records which conclusions require a later, more representative process. This keeps early speed useful while preventing an inexpensive sample from carrying more evidential weight than it deserves.
Choose the next build from the largest unresolved variable. If packaging is settled but thermal performance is uncertain, spend the next iteration on thermal interfaces rather than cosmetic refinement. If production-process effects remain unknown, moving to a representative route is more valuable than tightening dimensions on a substitute part.
Upload the CAD model and drawing to request a 6CProto prototype review. Include the test purpose, critical interfaces and any material substitutions your team can accept.
FAQ
Should replacement prototypes be made from the same material batch?
Use the same batch when the experiment is intended to isolate a geometry or process change and batch variation could affect the result. If that is impractical, preserve identification and document the change. For a development program investigating batch robustness, multiple batches may be intentional. The test plan should explain whether consistency or variation is the objective so a replacement specimen does not silently change the comparison.
Can a damaged prototype be repaired and tested again?
A repair changes the specimen condition and may affect stiffness, thermal transfer, sealing or residual stress. Keep the original result, identify the repair and decide whether the next run is exploratory or part of the formal evidence package. Do not combine repaired and untouched specimens without distinguishing them. The responsible engineer should decide whether the repair is representative and whether a fresh build is necessary.
How should unused prototypes be stored between test rounds?
Use storage that protects the properties being evaluated, not only the appearance. Document packaging, cleanliness, humidity-sensitive material conditioning and protection of mating surfaces where these can affect results. Keep identification visible without marking a critical surface. Before reuse, check whether storage time, contamination or handling has changed the specimen condition. A retained reference sample is useful only when its condition remains known and relevant.

