Is IEC 60068 testing legally required for CE marking?

Not directly. IEC 60068 is a toolbox of environmental and mechanical test methods, not a regulatory threshold. CE marking is driven by the EMC, Low Voltage and RED directives, which do not mandate climatic or vibration robustness as such. IEC 60068 becomes mandatory only when a product standard, a customer specification or a sectoral scheme (rail EN 50155, automotive, defence) calls a 60068-2 method with defined severities. It is qualification engineering, not a legal gate.

What is the difference between IEC 60068 and MIL-STD-810?

Both describe environmental test methods, but with different philosophies. IEC 60068-2 gives standardised, repeatable methods with fixed severity grids, ideal for commercial and industrial qualification and for international acceptance. MIL-STD-810 is tailoring-based, you derive the test from the platform's measured life-cycle environment, and it covers field scenarios (gunfire, altitude, sand) that 60068 does not. Many industrial programmes use 60068 methods with severities informed by 810 measurement data.

How do I choose the temperature extremes for cold and dry heat tests?

Start from the product's specified operating and storage range, then add the local rises caused by self-heating and enclosure effects. The cold method (60068-2-1) and dry heat (60068-2-2) are run at the worst-case storage limits for non-operating tests and at the operating limits with the product powered for functional checks. Pick a severity from the standard's preferred grid (for example -40, -25, -10 degC cold; +55, +70, +85 degC heat) that bounds the real environment with margin.

When should I use steady-state damp heat versus cyclic damp heat?

Steady-state damp heat (60068-2-78) holds a fixed high temperature and humidity to check insulation resistance and absorption, useful for a stable warm-humid service environment. Cyclic damp heat (60068-2-30) ramps temperature so that condensation forms and re-evaporates on each cycle, which drives moisture into seals and connectors and is the harsher, more revealing test for products that see day-night humidity swings or outdoor exposure.

Does passing IEC 60068 prove my product is reliable?

No. IEC 60068 confirms a product survives defined stresses without failure or degradation, which is robustness qualification, not a reliability figure. It does not produce an MTBF or a field failure rate. Reliability comes from statistical life testing and methods such as HALT and HASS, which deliberately push beyond the spec to find margins and screen production. IEC 60068 and HALT are complementary, not interchangeable.

What is the relationship between IEC 60068 and automotive ISO 16750?

ISO 16750-3 (mechanical) and 16750-4 (climatic) define the vibration profiles, thermal cycles and shock levels for road-vehicle electrical and electronic equipment, and they frequently reference IEC 60068-2 methods to execute those tests. In practice ISO 16750 sets the automotive severities and mounting-location profiles, while IEC 60068-2 supplies the underlying test procedure. They are layered, not competing.

How many vibration and thermal cycles should I specify?

The number of cycles or the test duration is derived from the expected service life and the use profile, not chosen arbitrarily. For vibration this means relating the test spectrum and dwell time to the cumulative fatigue the product will see in transport and operation. For thermal cycling it means matching the number of cycles to the lifetime count of significant temperature excursions. Document the rationale so the qualification is defensible.

IEC 60068: environmental and mechanical testing

Author Hugues Orgitello Last updated 28 May 2026 ~11 min read

Guide · Environmental and mechanical robustness

A product can pass every EMC scan and every electrical-safety check and still die in the field, cracked by vibration on a delivery truck, corroded by salt air on a harbour, or killed by condensation after a cold night. The IEC 60068 family is the international toolbox that answers a different question from the directives: not "is it safe and quiet on the spectrum?" but "will it physically survive its environment?". This guide maps the main IEC 60068-2 test methods, explains how to choose their severities from a product's real use profile, and clarifies how environmental qualification relates to automotive ISO 16750, military MIL-STD-810, and reliability methods such as HALT and HASS.

What IEC 60068 is, and is not

IEC 60068 is a multi-part standard. Part 1 (IEC 60068-1) gives general principles and guidance, and the large part 2 family (the "-2-x" methods) defines individual tests: cold, heat, humidity, vibration, shock, and so on. It is a catalogue of methods, deliberately silent on which test a given product must pass. The severity, the duration and the pass/fail criteria are set by whoever specifies the test: a product committee, a customer, or the manufacturer's own qualification plan.

This is the crucial distinction for anyone coming from CE marking. The EMC Directive, the Low Voltage Directive and the RED set legal obligations enforced by harmonised standards. IEC 60068 does none of that. It becomes a requirement only when something else invokes it:

a product standard that calls a 60068-2 method (for example railway rolling-stock electronics under EN 50155, which references shock and vibration tests),
a customer specification in a B2B contract,
a sectoral scheme (automotive, defence, aerospace, industrial machinery).

So IEC 60068 sits in the qualification layer, not the regulatory layer. Passing it does not earn a CE mark, and a CE mark does not imply it was done. Treating environmental robustness as out of scope because "the directives don't ask for it" is one of the most expensive mistakes a hardware team can make.

The main IEC 60068-2 test methods

The table below lists the methods this guide covers. Part numbers are the load-bearing detail; the column "what it stresses" explains the physics each one exercises.

Method	Test	What it stresses	Typical use
IEC 60068-2-1	Cold (test A)	Low-temperature operation and start-up, material brittleness	Storage and operation at cold limit
IEC 60068-2-2	Dry heat (test B)	High-temperature operation, derating, plastics softening	Storage and operation at hot limit
IEC 60068-2-78	Damp heat, steady state	Insulation resistance, moisture absorption	Stable warm-humid environments
IEC 60068-2-30	Damp heat, cyclic (test Db)	Condensation, breathing of enclosures, seal ingress	Outdoor, day-night humidity swings
IEC 60068-2-14	Change of temperature (test N)	Thermal expansion mismatch, solder-joint fatigue, thermal shock	Rapid ambient changes, shock transfer
IEC 60068-2-6	Vibration, sinusoidal	Resonances, fatigue at discrete frequencies	Rotating-machinery and resonance search
IEC 60068-2-64	Vibration, broadband random	Real-world multi-frequency fatigue	Transport, vehicles, broadband sources
IEC 60068-2-27	Mechanical shock	Half-sine or sawtooth pulse, fixings and connectors	Handling drops, in-service impacts
IEC 60068-2-29	Bump	Repeated low-level shocks, accumulated fatigue	Continuous mechanical knocking
IEC 60068-2-31	Free fall (test Ec)	Drop onto a hard surface, enclosure integrity	Handheld and portable equipment
IEC 60068-2-32	Free fall, repeated	Cumulative drop damage	Rugged portables, field tools
IEC 60068-2-11	Salt mist	Corrosion of metals and finishes	Coastal, marine-adjacent environments
IEC 60068-2-52	Salt mist, cyclic	Accelerated cyclic corrosion	Harsh marine, automotive exterior

A complete qualification rarely uses all of them. The art is selecting the few that match the product's environment, and setting each one's severity correctly.

Cold and dry heat

IEC 60068-2-1 (cold) and IEC 60068-2-2 (dry heat) are the two temperature-extreme tests. Each can be run as a storage test (product unpowered, checking survival of the non-operating limit) or an operating test (product powered and monitored, checking functional behaviour at the temperature limit). Two stabilisation methods exist, with and without dissipation, depending on whether the product self-heats significantly.

The classic pitfall is forgetting self-heating. A device rated to operate at +70 degC ambient that dissipates several watts inside a sealed enclosure may reach +90 degC at the hottest component. The chamber set-point alone does not tell you the junction temperature; you instrument the unit and verify the internal hot-spot against component derating limits.

Damp heat: steady-state versus cyclic

Two humidity tests answer two questions. IEC 60068-2-78 holds a constant high temperature and high relative humidity (commonly +40 degC / 93% RH) for an extended period, probing insulation resistance and slow moisture absorption. IEC 60068-2-30 cycles the temperature so that the dew point is crossed on each cycle: condensation forms on cold surfaces, then re-evaporates. That breathing action pumps moisture past seals and into connectors, making cyclic damp heat the harsher and more diagnostic test for anything exposed to outdoor day-night swings.

Change of temperature and thermal shock

IEC 60068-2-14 covers two regimes. Test Na is a rapid two-chamber transfer (thermal shock), where the specimen is moved between hot and cold chambers in seconds, stressing the mismatch in thermal expansion between materials. Test Nb is a slower ramp within one chamber. Thermal cycling is the dominant driver of solder-joint fatigue: every excursion strains the joints between components and board, and after enough cycles a marginal joint cracks. This is why the number of cycles must be tied to the product's lifetime count of meaningful temperature excursions.

Vibration: sinusoidal versus random

Two complementary methods address mechanical fatigue. IEC 60068-2-6 applies a sinusoidal sweep, one frequency at a time, which is excellent for finding resonances and for environments dominated by a rotating source (a motor, a pump, a compressor) at known frequencies. A resonance search identifies the natural frequencies of the assembly; a dwell at each resonance then accumulates fatigue at the worst case.

IEC 60068-2-64 applies broadband random vibration, exciting all frequencies in the band simultaneously, defined by a power spectral density (PSD, in g squared per Hz) profile. Real transport and vehicle environments are random, not single-tone, so random vibration is the realistic stress for most products that move. The relationship "IEC 60068-2-64 grms is the area under the PSD curve" is worth internalising: the overall grms summarises the energy, but the PSD shape determines which resonances get hit.

Aspect	Sinusoidal (60068-2-6)	Random (60068-2-64)
Excitation	One frequency at a time	All frequencies at once
Defined by	Amplitude vs frequency sweep	PSD profile (g squared per Hz)
Finds	Discrete resonances	Realistic broadband fatigue
Best for	Rotating machinery, resonance search	Transport, road, broadband sources

Mechanical shock and bump

IEC 60068-2-27 applies discrete shock pulses, typically a half-sine of defined peak acceleration (in g) and duration (in ms), reproducing a handling drop or an in-service impact. IEC 60068-2-29 (bump) applies thousands of lower-level repeated shocks, reproducing continuous knocking such as a device bolted to machinery. Shock checks that fixings, connectors and heavy components survive a single severe event; bump checks accumulated fatigue from a steady stream of minor ones.

Free fall and drop

IEC 60068-2-31 drops the product onto a hard surface from a defined height, onto specified faces and edges, simulating a real-world handling drop. IEC 60068-2-32 repeats the drop many times for rugged portables. These tests target enclosure integrity, mounting bosses, battery retention and connector survival. For a handheld instrument or a field sensor, drop is often the single most failure-prone test, far more so than vibration.

Salt mist

IEC 60068-2-11 exposes the product to a continuous salt fog, accelerating corrosion of exposed metals, plating and coatings. IEC 60068-2-52 cycles between salt fog, humidity and drying for a more realistic and aggressive corrosion profile. These matter for coastal, marine-adjacent and vehicle-exterior products, where galvanic corrosion at dissimilar-metal junctions or under coating defects is a slow killer. A COFRAC or otherwise accredited corrosion lab is worth seeking, because salt-mist results are notoriously chamber-dependent.

Choosing severities from the use profile

The single most important skill in environmental qualification is choosing severities that bound the real environment with defensible margin, neither under-testing (false confidence) nor over-testing (wasted cost, false rejects). The logic flows from the use profile to the test parameters:

Define the life-cycle environment. Where will the product live, ship, store and operate? A wall-mounted indoor sensor, a roadside cabinet, a vehicle ECU and a marine buoy face entirely different climates and vibration.
Translate environment into stress quantities. Temperature extremes, humidity range, vibration PSD, shock peaks, number of thermal excursions per day, years of service. Where possible, measure the environment (data-logging on a representative vehicle or site) rather than guessing.
Pick a severity from the standard's preferred values that envelopes the measured or estimated stress, with margin. IEC 60068-2 methods come with preferred grids (preferred cold temperatures, preferred PSD levels, preferred drop heights) precisely so results are comparable across labs.
Set duration and cycle count from the lifetime dose. Vibration dwell and random-test minutes relate to cumulative fatigue; thermal-cycle counts relate to lifetime excursions. Acceleration (testing harder for less time) is possible but must be justified by a fatigue model, not assumed.
Define pass/fail before testing. Functional limits, allowed degradation, visual criteria. A test with no pre-agreed criteria produces an argument, not a result.

For a typical industrial IoT product, a sensible starting qualification might combine cold and dry heat at the operating and storage limits, cyclic damp heat for outdoor units, random vibration matched to the transport and mounting environment, a mechanical-shock pulse, and a drop test for any portable variant. Salt mist is added only when the deployment justifies it.

Qualification versus regulatory requirement

It is worth stating plainly, because teams routinely conflate the two.

Aspect	Regulatory (CE, FCC)	Qualification (IEC 60068)
Driver	Directives and law	Customer, product std, internal spec
Outcome	Market access, CE mark	Confidence the design survives field life
Mandatory?	Yes, to sell legally	Only when invoked by a spec or contract
Who sets criteria	Harmonised standards, regulators	The specifier of the qualification
Consequence of skipping	Illegal placement, recall	Field failures, warranty cost, lost trust

A product can be fully CE-marked and still be environmentally fragile, and conversely, a robust product is not automatically compliant. The two activities run in parallel and serve different masters: one protects the right to sell, the other protects the cost of ownership and the brand.

Relationship with ISO 16750, MIL-STD-810 and HALT/HASS

IEC 60068 does not live alone. Three neighbouring frameworks intersect with it, and confusing them leads to over- or under-specified test plans.

Automotive: ISO 16750

For road-vehicle electronics, ISO 16750-3 (mechanical loads) and ISO 16750-4 (climatic loads) define the vibration profiles, thermal cycles and shock levels appropriate to a vehicle, broken down by mounting location (engine compartment, passenger cabin, on the engine itself, on a wheel). Crucially, ISO 16750 frequently references IEC 60068-2 methods to execute the tests. So the layering is: ISO 16750 supplies the automotive severities and profiles, IEC 60068-2 supplies the procedure. This sits alongside the functional-safety and component-qualification expectations covered in the ISO 26262 automotive functional safety guide and the AEC-Q100/Q101 semiconductor qualification guide.

Defence: MIL-STD-810

MIL-STD-810 (the H revision being current) is the U.S. military's environmental engineering standard. Its philosophy differs from IEC 60068: 810 is tailoring-based, meaning you derive each test from the platform's measured life-cycle environmental profile rather than from a fixed grid. It also covers scenarios IEC 60068 does not address, such as gunfire shock, rapid decompression, blowing sand and dust, and acoustic noise. Many industrial and rugged-commercial programmes use IEC 60068-2 methods but borrow MIL-STD-810 measurement data to inform realistic severities. The military-vehicle electrical environment, including the conducted transients that often accompany these mechanical stresses, is treated in the MIL-STD-1275 military vehicle power guide.

Reliability: HALT and HASS

HALT (Highly Accelerated Life Test) and HASS (Highly Accelerated Stress Screen) are reliability techniques, and they answer a different question again. HALT deliberately drives a product beyond its specification, stepping temperature and vibration up until it breaks, to discover design margins and weak links early. HASS uses those discovered limits to screen production units for latent defects. The contrast with IEC 60068 is fundamental:

IEC 60068 asks "does it survive the specified environment?" (a pass/fail qualification).
HALT asks "where does it break, and how much margin is there?" (a margin-discovery process).

Neither replaces the other. A robust development flow uses HALT early to harden the design, then IEC 60068 (or ISO 16750, or MIL-STD-810) to qualify against the contractual environment, and optionally HASS to screen production. Claiming a HALT result as a 60068 qualification, or vice versa, is a category error.

This also intersects with high-power and connector-stress domains; for products in the charging ecosystem, mechanical and thermal robustness of connectors and cables is part of the picture covered in the EV charging IEC 61851 / ISO 15118 / OCPP guide.

Common pitfalls

Assuming the directives cover it. They do not. EMC and electrical safety say nothing about vibration, drop or corrosion. Anticipate environmental testing as a separate work package from the start of the project.
Testing a single sample and calling it qualified. Environmental qualification has sample-size implications; a single survivor proves little statistically. Define how many units, and whether the test is destructive.
Ignoring self-heating in temperature tests. The chamber set-point is not the component temperature. Instrument the hot spots.
Choosing sinusoidal where random is realistic. Single-tone sweeps miss the broadband reality of transport. Use random vibration unless the environment is genuinely dominated by a discrete tone.
Under-specifying thermal cycle count. Solder-joint fatigue is cumulative; too few cycles hides the failure mode that will appear in year three.
No pre-agreed pass/fail criteria. Without functional limits and degradation thresholds set before the test, the result is a negotiation.
Confusing qualification with reliability. Passing 60068 is not an MTBF. If a numeric reliability target exists, plan separate life testing.
Skipping accredited labs for corrosion. Salt-mist outcomes are highly chamber-dependent; an accredited ISO/IEC 17025 lab makes the result defensible.

A practical sequencing note

Environmental tests interact, and the order can matter. A common approach runs non-destructive climatic tests first (cold, dry heat, damp heat), then mechanical (vibration, shock, bump), then drop, with functional verification between stages so a failure can be attributed to a specific stress. Salt mist, being potentially destructive to finishes, is often run on dedicated samples. Combined or sequential stressing (for example vibration during temperature) is more realistic for some platforms but requires combined-environment chambers and a clear rationale.

Above all, write the environmental test plan early, alongside the EMC and safety plan, with severities traced back to the use profile and pass/fail criteria agreed with the customer. That document is what turns a pile of chamber bookings into a defensible qualification.