Silicone Keypad Hardness Selection: Compression Set, Return Force, and Sealing

For silicone rubber keypads, hardness is a controlled material input. It is not a complete key-feel specification. A 50 Shore A plaque, a thin molded web, a compressed perimeter seal, and a finished key inside a fastened bezel answer different questions. The useful selection process keeps those questions separate until prototype evidence connects them.

---

1. Why Hardness Alone Is an Expensive Shortcut

Selecting a keypad from hardness alone can produce a part that passes incoming inspection and still sticks, double-acts, loses sealing preload, or feels wrong in the enclosure. Shore A describes resistance to a standardized indentation. The user presses a three-dimensional mechanism whose response also comes from the web profile, key height, travel stop, vent path, contact architecture, PCB support, bezel clearance, and assembly compression.

The common failure chain is not complicated:

temperature + sustained assembly load + dwell time
                         ↓
compound response + web/base/seal geometry
                         ↓
permanent set and/or loss of counterforce
                         ↓
less key return, reduced contact separation, or lower seal margin
                         ↓
intermittent input, a slow or sticking key, or leakage at the assembly boundary

That chain contains at least four measurements. Treating them as synonyms hides the failure mechanism.

ASTM D395-18(2025) defines the applicable static-compression boundary. Its compression-set methods are mainly applicable to static compressive service; repeated deformation is represented more closely by compression-flexing or hysteresis tests. A keypad base trapped under a bezel may have a static-set problem. The moving web has a cyclic force-travel problem. A sealing lip may have both a set problem and a retained-force problem.

The practical cost of the shortcut appears late. Changing a compound after tooling may change fill, shrinkage, demolding, pigment response, and force. Changing the web after appearance approval may require tool steel work and another sample loop. It is cheaper to define the outputs separately, screen controlled variants, and approve the installed system than to ask a supplier for “a harder rubber” after a poor first article.

Use the observed symptom to choose the next measurement instead of changing durometer by reflex:

---

2. Engineering Definitions: Shore Hardness, Compression Set, and Return

The three material terms that matter most are related, but none can substitute for another. A defensible rubber keypad material selection records the test method and conditions beside each result, then verifies the finished mechanism separately.

Silicone rubber Shore hardness is an indentation result

ASTM D2240-15(2021) defines durometer hardness as an empirical penetration measurement. The result depends on elastic modulus, viscoelastic behavior, indenter geometry, and applied force; it has no simple relationship to a fundamental material property. That is why “60 Shore A” is incomplete unless the drawing or inspection plan also controls the method, scale, specimen or approved coupon, conditioning, measurement location, and reading protocol.

ISO 48-4:2018 assigns the A scale to rubber in the normal-hardness range. It assigns AM to thin normal-hardness pieces, AO to low-hardness or cellular rubber, and D to high-hardness rubber. Those scales should not be casually converted or compared. For a thin, curved, textured, or multi-level keypad, the parties should agree whether hardness is verified on a standard specimen, a co-cured witness coupon, or another controlled location permitted by the governing procedure.

Keypad compression set is a conditioned recovery result

Compression set reports how much of an imposed deflection remains after release and a specified recovery interval. Using the common constant-deflection expression:

Compression set (%) = (t0 − tr) / (t0 − ts) × 100

• t0 is original thickness.
• ts is spacer or compressed thickness.
• tr is recovered thickness after the specified recovery interval.

The Parker O-Ring Handbook illustrates the same equation: 0% means full dimensional recovery, while 100% means the specimen recovered no farther than the compressed thickness. That explanation does not create a keypad acceptance limit.

ISO 815-1:2019 requires the test conditions to remain attached to the result. Set changes with the time and temperature of compression and with the time, temperature, and conditions of recovery. Its normal constant strain is 25%, with lower strains for high-IRHD rubber. IRHD is not Shore A. ISO also distinguishes short elevated-temperature tests, typically 24 hours for cure state or classification, from longer tests, typically 1,000 hours, that include more ageing. A bare statement such as “compression set: 20%” is therefore not comparable evidence.

Stress relaxation measures retained counterforce

A sealing flange can keep roughly the same compressed height while its reaction force decays. Compression set will not fully describe that loss. ISO 3384-1:2024 measures the decrease in counterforce while rubber is held at constant deformation and constant temperature. ISO 3384-2:2019 extends the counterforce question to temperature cycling.

Neither method replaces a finished-key force-travel curve. Shin-Etsu Polymer’s force-travel guidance attributes peak force and contact force jointly to silicone material and dome shape; travel also depends on geometry and contact position. For keypad approval, record both the downstroke and return path at a defined speed, fixture, key location, temperature, and conditioning state.

Primary supplier data show why hardness is not a compression-set ranking

The examples below are useful because the grades and test conditions are named. They are typical supplier data, not specifications, JASPER results, or universal keypad limits. Rows with different methods, cure states, temperatures, or geometries must not be ranked against one another.

Sources: Dow’s 2023 LSR selection guide, Shin-Etsu’s KE-54x1 family data, and WACKER’s 2025 processing guide. WACKER also identifies grades formulated for low set without post-curing, so “all silicone must be post-cured” is not a sound rule. Cure and post-cure belong to the exact material/process release, not to folklore.

---

3. The 10-Point Silicone Keypad Hardness Selection Framework

A sound silicone keypad hardness selection moves from service conditions to measurable outputs, then to compound and geometry. It does not assign an application label—“medical,” “industrial,” or “outdoor”—to a generic hardness band.

3.1 Define the service envelope before naming a hardness

Record minimum, nominal, and maximum part temperature; storage excursions; warm-up state; press frequency; maximum hold time; cleaning exposure; and any continuous clamp or sealing load. Temperature survival is not the same as tactile stability. A compound may remain elastomeric while the assembly’s force or recovery moves outside its requirement.

Good signal: A requirements sheet separates continuous, intermittent, and exceptional exposures.
Red flag: “Room temperature” is the only test condition for an outdoor, vehicle, heated, or washdown control.

3.2 Specify four outputs, not one feel adjective

Define incoming hardness, finished-key downstroke/return force-travel, static keypad compression set where relevant, and retained counterforce where preload matters. Add contact separation, seal squeeze, or release time if those functions have margins. “Crisp,” “soft,” and “fast return” can remain user-language targets, but each needs a measurable translation.

Good signal: The drawing and validation plan name each output and its owner.
Red flag: One Shore A tolerance is expected to control every tactile and sealing behavior.

3.3 Control the exact compound and cure state

“VMQ silicone, 50A” does not identify a material. Record supplier, grade, color/additive package, cure system, primary cure, post-cure when applicable, and approved change path. Post-cure, curing-agent choice, and section thickness can alter compression-set behavior.

Good signal: The approved material and process state are revision-controlled.
Red flag: A supplier may substitute any same-hardness compound without revalidation.

3.4 Lock the silicone rubber Shore hardness method

Name ASTM D2240 or ISO 48-4, the selected scale, specimen or witness coupon, conditioning, measurement locations, reading protocol, and sampling plan. Use the current controlled standard during execution; the public standard summaries are not a substitute for licensed procedural details.

Good signal: Supplier and incoming inspection use the same controlled method or an agreed correlation.
Red flag: A handheld reading on any convenient curved keytop is accepted as equivalent to a controlled specimen.

3.5 Co-design hardness with the web and travel geometry

Hardness cannot rescue an unstable key, an over-strained web, a blocked vent path, or an uncontrolled bottom-out. Key height, web angle and thickness, transition radii, travel, stop, contact position, and support determine where the rubber bends and how the force curve develops. The force-travel curve must confirm the combination.

Good signal: Compound and geometry variants are changed one controlled factor at a time.
Red flag: The mold geometry is frozen before any production-material tactile sample exists.

3.6 Separate sustained sealing compression from key actuation

The perimeter base or sealing bead can remain compressed for months while each web moves only during a press. Use ASTM D395 or ISO 815 for the relevant static recovery question; use finished-key cyclic and force-travel tests for the moving mechanism. If sealing depends on retained preload, include stress relaxation rather than relying on recovered height alone.

Good signal: The stack tolerance analysis calculates minimum and maximum continuous squeeze.
Red flag: A key actuation force is used as evidence that the perimeter seal will retain load.

3.7 Compare keypad compression set only under matched conditions

Match the standard and method, specimen geometry, deflection or strain, medium, temperature, duration, recovery procedure, cure state, and data status. “Typical” and “specification maximum” are different. An unmatched lower percentage does not prove the better compound.

Good signal: Every value carries a complete condition string and source revision.
Red flag: A procurement sheet ranks percentages copied from unrelated datasheets with the conditions removed.

3.8 Measure return after dwell, not just initial peak force

For a moving key, repeat the installed force-travel curve after defined holds, thermal conditioning, and recovery intervals. For a continuously compressed seal or base, ISO 3384-1 or ISO 3384-2 can address counterforce loss at constant deformation. Keep those outputs distinct.

Good signal: Baseline and post-conditioning curves share the same fixture, speed, alignment, and temperature.
Red flag: A new-part peak-force result is treated as proof of aged return.

3.9 Age the properties that carry the design margin

Accelerated heat exposure is useful only when the before/after property and limit are named. ASTM D573-04(2025) and ISO 188:2023 support controlled accelerated ageing, but exact service correlation may not follow. Use ISO 815-2:2019 when low-temperature recovery is a specific risk.

Good signal: Ageing checks hardness, dimensions, force-travel, return, and set as the design requires.
Red flag: “Heat resistant” appears as an acceptance criterion without exposure or measured output.

3.10 Approve the production-intent assembly

The bezel, PCB location, carrier stiffness, gasket compression, fastener load, coating, and clearance can change key response. Approve the keypad in the released stack, not only as a loose molded part. JASPER’s silicone keypad HMI assembly guidance makes the same boundary explicit, while its testing and validation planning page treats methods, conditions, samples, and acceptance ownership as project-specific.

Good signal: The approval fixture reproduces production support and clamp conditions.
Red flag: A free keymat approval is assumed to cover every enclosure revision.

---

4. A Six-Step Rubber Keypad Material Selection and Approval Process

The selection process should narrow uncertainty in a fixed order. It begins with service inputs and ends with a controlled material–process–geometry–assembly combination.

Step 1 — Convert use conditions into measurable requirements

Supply the part temperature range, storage excursions, maximum press/hold time, expected idle dwell, user or glove condition, target downstroke and return behavior, continuous clamp or seal load, cleaning/media exposure, and allowable recovery time. Add the contact architecture and the consequence of a slow or incomplete return.

Step 2 — Freeze the stack that creates force and preload

The drawing should show the mechanism, not just the visible keytop:

finger or external actuator
        ↓
[ keytop / plunger ]
[ web or dome ]          ← moving spring geometry
[ base + sealing bead ]  ← sustained bezel compression may act here
[ carbon pill / actuator / metal dome ]
[ PCB, FPC, or membrane circuit ]
[ carrier + enclosure + fasteners ]

Control datum surfaces, web section, radii, key height, travel, hard stop, vent path, base thickness, sealing feature, contact position, PCB support, bezel clearance, and fastener condition.

Step 3 — Screen controlled variants

First hold geometry constant and compare exact, traceable compounds. Then hold the selected compound constant and tune geometry. Two or three bracketing variants are more informative than one guessed midpoint. Use 60 Shore A as the center of the first controlled screening matrix when no customer target exists. WACKER ELASTOSIL LR 3005/60 A/B is published at 60 +/- 3 Shore A, which makes it a traceable comparison grade. Release the final hardness only after the molded assembly meets the force-travel, return, dwell, sealing, and ageing limits.

Step 4 — Mold production-representative samples

Appearance models and cast surrogates cannot validate production cure, shrinkage, web response, or set. Use the intended molding process, material, pigment/additive state, cure, post-cure, and critical wall sections. JASPER’s prototyping and sample approval page similarly separates fit, appearance, functional, and production-intent evidence.

Step 5 — Condition, dwell, age, and remeasure

Capture baseline hardness and installed force-travel. Apply the relevant static compression, temperature, time, recovery, thermal-cycle, low-temperature, and media exposures. Then repeat the same measurements. The comparison is meaningful only when fixtures, speed, alignment, support, and measurement temperature are controlled.

Step 6 — Release the complete controlled combination

Approve the compound grade, cure state, geometry revision, assembly stack, test method, conditions, sample stage, acceptance owner, and retained reference sample together. Link inspection records and deviations to that revision through an inspection and traceability plan. Any change to material, cure, web, PCB support, bezel, gasket, or fastener load needs a defined revalidation decision.

---

5. Validation Test Matrix and Sample Approval Checklist

A useful validation matrix assigns one method to one engineering question. The current licensed standard and the project test plan—not this summary—control the laboratory procedure.

Before sample approval, confirm that the package contains:

• Exact compound supplier and grade; pigment/additive package; cure and post-cure state.
• Controlled keypad, PCB, bezel, carrier, gasket, and fastener revisions.
• Baseline and post-conditioning force-travel data, including the return path.
• Static compression and recovery conditions wherever the base or seal stays clamped.
• Acceptance limits, sample stage, sample quantity, data owner, and deviation authority.
• A retained approved sample and defined triggers for material, process, tool, or stack revalidation.

Component checks do not certify the finished equipment. Ingress, electrical safety, EMC, cleaning compatibility, and regulatory approval remain with the evaluated finished configuration and its responsible manufacturer.

---

6. When a Molded Silicone Keypad Is Not the Best Construction

A molded silicone keypad is not automatically the right interface. A flat membrane switch may fit better when profile, integrated graphics, and minimal travel dominate. Capacitive touch may fit a continuous gesture surface, provided the project accepts its glove, moisture, and feedback tradeoffs. A metal-dome hybrid may be preferable when the silicone should act only as a sealed plunger over a separately defined snap element.

Choose another architecture when the available travel cannot support a stable web, the enclosure cannot control preload, the required chemical or temperature exposure has no verified compound, or the seal depends on an unvalidated keymat alone. The decision should be made at stack level; this membrane keypad versus silicone keypad comparison frames the main travel, feel, profile, graphics, lighting, and mounting differences.

---

7. Red Flags That Disqualify a Material or Proposal

These red flags override an attractive sample or a low unconditioned number:

• A Shore A value with no method or specimen — the incoming control cannot be reproduced.
• A compression-set percentage with no condition string — the value cannot be compared.
• “Equivalent” compound substitution by hardness alone — formulation and cure-dependent behavior is uncontrolled.
• Plaque data presented as finished-key performance — geometry and assembly effects are missing.
• Only a downstroke peak is reported — contact force, return path, and hysteresis remain unknown.
• Static set is used to promise cycle life — ASTM D395 explicitly limits that inference.
• The keypad is approved outside its bezel and PCB stack — production preload and stops are absent.
• Universal IP, medical, chemical, or temperature language — component data is being confused with finished-configuration validation.

---

8. Frequently Asked Questions

What Shore A hardness is best for a silicone keypad?

There is no universal best Shore A hardness. Select an exact compound and web geometry against the target force-travel, return, sealing load, temperature, dwell, and assembly. Use hardness to control the approved material—not to replace production-intent tactile and recovery tests.

Does higher Shore A mean lower keypad compression set?

No. Dow publishes two 50 Shore A LSR grades with different typical set values under the same 22 h/175°C headline condition, while Shin-Etsu’s conditioned family data is non-monotonic across four hardnesses. Formulation, cure, temperature, time, recovery, and geometry all matter.

What is an acceptable keypad compression set percentage?

An acceptable percentage is project-specific and meaningless without method, strain or deflection, specimen, medium, temperature, duration, recovery conditions, and cure state. Set the limit from the dimensional or sealing margin, then confirm return and counterforce separately when those outputs matter.

Can 50 Shore A and 60 Shore A keypads have the same actuation force?

Yes, different web, dome, travel, stop, and support geometries can produce overlapping actuation forces. That does not make the designs equivalent: return force, strain distribution, hysteresis, sealing preload, and aged response may still differ. Compare full downstroke and return curves.

What is the difference between ASTM D395 and ISO 815-1?

Both address rubber compression set, but their methods, terminology, specimen rules, and reporting details must be followed from the controlled standard. ASTM D395 includes constant-force and constant-deflection methods; ISO 815-1 covers constant-strain set at ambient or elevated temperatures. Do not merge results casually.

How should silicone-key return be validated after ageing?

Record a baseline force-travel curve in the intended assembly, apply the defined temperature, dwell, static compression, cycling, or media exposure, allow the specified recovery, and repeat the same curve. Use ISO 3384 when retained counterforce under constant deformation is the actual requirement.

Does every silicone keypad require post-curing?

No. The need depends on the exact formulation, cure chemistry, thickness, volatile requirements, and target properties. WACKER documents both a grade whose set improves with post-cure and grades formulated for low set without it. Release the supplier-approved process and verify the molded part.

What project inputs should be shared before material selection?

Share minimum/nominal/maximum temperature, storage and cleaning exposure, key size and geometry, target downstroke and return curves, maximum hold/dwell, continuous seal or clamp load, contact type, PCB and bezel stack, fastener condition, expected failure consequence, and the approval test plan.

9. What to Do Next

Prepare one package containing the temperature envelope, sustained load and dwell, return and contact-separation requirement, sealing function, stack drawing, contact architecture, and target force-travel curve. Ask candidate molders to identify exact compounds and a controlled prototype matrix instead of quoting a generic hardness.

Send the package through the project contact route for a material, geometry, and sample-validation review.

Technical References

• Source: ASTM D2240-15(2021) Durometer Hardness. Accessed 2026.
• Source: ASTM D395-18(2025) Compression Set. Accessed 2026.
• Source: ISO 815-1:2019 Compression Set at Elevated Temperatures. Accessed 2026.
• Source: ISO 3384-1:2024 Stress Relaxation in Compression. Accessed 2026.
• Source: WACKER Solid and Liquid Silicone Rubber Processing Guidelines. Accessed 2026.
• Source: Dow Liquid Silicone Rubber Product Selection Guide. Accessed 2026.
• Source: ASTM D395-18(2025). Accessed 2026.
• Source: ASTM D2240-15(2021). Accessed 2026.
• Source: ISO 48-4:2018. Accessed 2026.
• Source: Parker O-Ring Handbook. Accessed 2026.
• Source: ISO 815-1:2019. Accessed 2026.
• Source: ISO 3384-1:2024. Accessed 2026.
• Source: ISO 3384-2:2019. Accessed 2026.
• Source: force-travel guidance. Accessed 2026.
• Source: Dow’s 2023 LSR selection guide. Accessed 2026.
• Source: Shin-Etsu’s KE-54x1 family data. Accessed 2026.
• Source: WACKER’s 2025 processing guide. Accessed 2026.
• Source: ASTM D573-04(2025). Accessed 2026.
• Source: ISO 188:2023. Accessed 2026.
• Source: ISO 815-2:2019. Accessed 2026.
• Source: ELASTOSIL LR 3005/60 A/B. Accessed 2026.