Why Plastic Tolerances Do Not Transfer Directly to Custom Silicone Parts
A drawing can look precise and still create confusion at the mold trial. We often receive custom silicone part drawings that use the same blanket tolerance on every dimension—sometimes copied from a machined metal or injection-molded plastic component. The intent is understandable: tighter numbers appear to mean better quality. With an elastomer, however, a tight number without an agreed measurement method can produce disagreement instead of control.
Silicone bends under its own weight, recovers after handling and can change shape when a gauge touches it. Its molded dimensions also reflect the compound, cure, part geometry and tooling strategy. The practical question is therefore not simply, “What tolerance can the mold hold?” It is, “Which dimensions control function, and how will both teams measure them the same way?”
This article explains how we approach that question at DX when reviewing a new custom silicone part. It is general engineering guidance, not a substitute for the project drawing, the selected material specification or an application-specific validation plan.
Start with a rubber tolerance framework—not a plastic default
ISO 3302-1:2014 provides dimensional-tolerance classes and related test methods for molded, extruded and calendered solid-rubber products. ISO confirmed the edition in 2024, so it remains current. The standard is useful because it gives customers and manufacturers a common starting language for rubber tolerances rather than leaving every dimension to assumption. It does not, however, apply to precision toroidal sealing rings, so O-rings should be specified under the appropriate dedicated standard or project requirement. [1]
A standard class is a starting point, not a reason to stop thinking. A molded silicone housing seal, an extruded profile and a very soft cosmetic sleeve do not behave the same way. We still need to identify the molding process, hardness, geometry, mating parts, operating condition and inspection method before confirming a project tolerance.
DX view: Apply the tightest tolerance only where it protects fit, sealing, alignment or another verified function. Use a realistic rubber tolerance elsewhere.
Why the same silicone part can produce different readings
1. The material and cure influence final size
Silicone compounds do not all shrink by one universal percentage. A supplier data point can be helpful for early mold design, but the actual finished dimension also depends on cure conditions and the geometry of the molded product. Dow, for example, notes in a silicone technical data sheet that linear shrinkage depends on curing conditions, including curing-agent type, temperature and molded-product size. [4] In practice, tooling compensation has to be confirmed with the selected material and trial parts—not copied from a generic value.
2. The part is flexible after it leaves the mold
A soft lip, thin wall or long gasket can sag, stretch or relax while it is handled. A dimension measured flat on a table may not match the same dimension measured while the part is hanging, lightly fixtured or installed in its housing. That does not automatically mean the part is defective; it may mean the measurement state was never defined.
3. The gauge can become part of the result
ASTM D3767 covers methods for measuring rubber length, width, thickness, diameter and circumference. Its public significance statement makes an important point: instrument pressure has a significant effect on the observed thickness of soft, flexible materials, so comparative measurements need a specified pressure. [2] A caliper that closes firmly on a soft silicone wall can report a smaller number than an optical method or a controlled-contact thickness gauge.
4. Timing and conditioning affect repeatability
ISO 23529 addresses preparation, measurement, storage and conditioning of rubber test pieces, including the time interval between forming and testing. ISO explains that these controls help produce reproducible results and reduce disagreement between customer and supplier. [3] For a finished molded part, the exact project procedure may differ, but the lesson is directly useful: agree when and under what conditions acceptance measurements will be made.
Build the drawing around function
We recommend beginning with the assembly instead of placing a narrow tolerance beside every dimension. Identify what the silicone must touch, clear, seal, locate or move against. Then divide dimensions into three groups:
Critical-to-function: dimensions that directly control fit, sealing, flow, alignment, safety or movement.
Manufacturing or assembly controls: dimensions needed to locate the part, manage a secondary operation or keep an overmolded interface stable.
Reference or cosmetic dimensions: useful for communication but not intended to drive acceptance on every lot.
This hierarchy matters commercially. Tight tolerances can require more precise tooling, added mold adjustment, controlled fixtures, slower inspection or higher sampling effort. If the same limit is applied to every flexible feature, cost rises without necessarily improving the finished product.
Define a datum strategy that the part can actually support
A rubber drawing needs stable references. A broad base, a molded locating feature or a functional mounting surface can often serve as a practical datum. A wavy free edge usually cannot. If a thin part must be held in a fixture to establish datums, show the fixture condition or describe it in the inspection note. Avoid long chains of dependent dimensions when one baseline or profile control would describe the functional geometry more clearly.
Separate free-state dimensions from installed dimensions
Some silicone parts only reach their working shape after assembly. A gasket may be intentionally oversized before it is placed in a groove; a bellows may sit at a different free height before compression; a sleeve may stretch over a rigid insert. Mark whether the requirement applies in the free state, on a fixture or in the mating assembly. If installed performance is what matters, a functional gauge or assembly test may be more meaningful than a free-state caliper reading alone.
Use a measurement plan, not just a tolerance column
For the dimensions that matter, put the acceptance method into the quality conversation before production. A simple plan can identify the measuring equipment, fixture, contact force where relevant, measurement points, part orientation, time after molding, sample quantity and rounding rule. The aim is not to make every project bureaucratic. It is to remove the few ambiguities most likely to create a false reject—or allow a functional problem to pass.
| Feature | Why it matters | Drawing / plan note | Useful verification approach |
| Seal bead height | Controls squeeze in the assembly | Name the free or installed state and datum | Controlled-contact gauge or approved section method |
| Mating diameter | Controls fit and retention | Define measurement plane and roundness need | Optical method, ring gauge or agreed fixture |
| Thin wall | Affects flexibility and flow | State measurement locations | Low-force thickness gauge or optical section |
| Hole / slot spacing | Controls alignment to rigid hardware | Use stable datums; avoid chain buildup | Fixture, vision system or optical projector |
| Flexible outline | Can sag or distort | State orientation and fixture condition | Profile fixture, overlay or non-contact measurement |
Approve the first article with the same method used for production
A first-article report is most useful when it proves the agreed drawing and measurement plan—not when it is a one-time set of numbers produced by an undefined technique. Before approval, compare the molded part with its mating components and run the functional checks that matter. A golden sample can help communicate appearance and handling, but it should supplement the controlled drawing and acceptance criteria rather than replace them.
If a critical result is marginal, the right response may be a tooling adjustment, a revised dimension, a different inspection method or a functional study. It is better to resolve that question during sampling than to argue about the same dimension after mass production starts.
What DX can review before tooling
DX supports custom silicone projects with a dedicated tooling design team and an in-house toolmaking workshop. Our published manufacturing services include compression molding, liquid injection molding, silicone overmolding, extrusion, secondary operations and assembly. [5, 7]
For inspection, DX lists dimensional testing among its production-quality checks and describes the use of a precision optical projector and depth gauge in its laboratory. [6] The specific gauge, sampling level and acceptance record for any new part should still be agreed in the project plan. We do not assume that one inspection setup fits every flexible geometry.
For a productive tolerance review, send us:
the 2D drawing and 3D model, including the revision level;
the mating-part drawing or assembly section;
the intended silicone specification and hardness, if already selected;
operating temperature, chemicals, pressure, repeated movement and cleaning conditions;
the dimensions that are truly critical to function;
the proposed measurement method or current inspection problem; and
prototype quantity, expected order quantity and annual volume.
If one of these items is not decided, mark it as open. That gives our engineering and tooling teams a clear point to review instead of turning an assumption into a production requirement.
Plan your custom silicone part around measurable function
Developing a custom gasket, seal, flexible housing part, overmolded component or other molded silicone product? Send DX your drawing, 3D file, mating-part details, application conditions, material or hardness target and expected quantity. We can review the functional dimensions, molding route, tooling approach and practical inspection method before production. Contact DX to discuss your custom silicone product and build an acceptance plan that both teams can measure the same way.
References
2. ASTM D3767-03(2020) — Standard Practice for Rubber—Measurement of Dimensions
3. ISO 23529:2016 — General procedures for preparing and conditioning rubber test pieces
5. DX Silicone — Liquid Injection Molding
6. DX Silicone — Quality Control
7. DX Silicone — Tooling Making
8. DX Silicone — Customization
Source note: Sources and DX website pages were checked on October 21, 2026. Standards are summarized from their public abstracts and significance statements; project requirements should refer to the licensed full text where applicable. The measurement checklist and drawing recommendations are DX editorial guidance.
DX provides you with all-around silicone product customization services for valued customers like you.
