There is something deeply satisfying about watching engineers strap a perfectly good product to an enormous machine and shake the living daylights out of it.

The product may have spent years being designed.

Someone may have machined beautiful components for it.

Someone else carefully assembled the electronics.

The software works.

The prototype has passed functional testing.

And then the verification team says:

Excellent. Now let’s shake it until something falls off.

That may look destructive.

Sometimes it is.

But vibration and shock testing exist for a very practical reason:

Real products rarely spend their entire lives sitting peacefully on a laboratory bench.

Cars hit potholes.

Aircraft engines vibrate.

Ships pound through waves.

Industrial machines generate rotating loads.

Trucks travel thousands of miles over imperfect roads.

Packages get dropped.

Satellites ride rockets.

Electronics experience vibration from fans, motors, pumps and compressors.

And sometimes a product survives every engineering calculation only to fail because one connector slowly backs out after six months of vibration.

That is why engineers shake perfectly good products.

They would rather discover the weakness in the laboratory than let the customer discover it later.

---

Vibration Is Everywhere

We tend to notice vibration only when it becomes annoying.

A steering wheel buzzes.

A dashboard rattles.

A washing machine walks across the floor.

An airplane cabin develops a persistent hum.

But vibration is simply oscillating mechanical motion, and virtually every engineered system encounters some form of it.

The source may be obvious:

• engine combustion,
• rotating machinery,
• propellers,
• pumps,
• fans,
• gears,
• road inputs,
• turbulent airflow,
• rail joints,
• waves,
• or rocket engines.

Or the vibration may come from something elsewhere in the system.

A component can be perfectly balanced and still experience substantial vibration because the structure supporting it is moving.

That makes vibration a systems problem.

---

The Product May Work Perfectly—At First

Imagine an electronic control unit mounted in a vehicle.

On the bench, everything works.

The processor runs.

The connections are solid.

Temperatures are acceptable.

Inputs and outputs behave correctly.

The unit passes.

Now bolt it into a vehicle and drive it for 100,000 miles.

Something different happens.

Tiny repeated movements begin working on:

• solder joints,
• connectors,
• wiring,
• fasteners,
• printed circuit boards,
• brackets,
• seals,
• housings,
• and component leads.

Nothing dramatic has to happen on any individual cycle.

That is the important part.

A very small load repeated millions of times can eventually create damage.

This is fatigue.

And vibration testing gives engineers a way to expose it before years of customer use are required.

---

Vibration Testing Is Not Just “Shake It Hard”

Modern vibration testing can be remarkably controlled.

Engineers define an input based on the environment the product is expected to experience.

The test may specify:

• acceleration,
• frequency,
• frequency range,
• duration,
• axis,
• waveform,
• spectral content,
• and allowable response.

Accelerometers attached to the test article measure how the structure responds.

A control system continuously adjusts the shaker to produce the required environment.

The result is not random abuse.

It is an attempt to reproduce—or sometimes accelerate—the mechanical environment the product will experience during its life.

---

Meet the Shaker

One of the stars of a vibration laboratory is the electrodynamic shaker.

Conceptually, it works a bit like a giant loudspeaker.

An electrical signal drives a moving armature, which moves the test fixture and product.

Depending on the system, shakers can test something as small as an electronic component or something much larger.

NASA’s Johnson Space Center, for example, operates vibration systems specifically for spaceflight hardware. Its hazardous vibration facility can perform sinusoidal and random vibration over a frequency range of 20 to 2,000 Hz, while other NASA structural facilities test components, subsystems and large integrated structures under flight-representative environments.

NASA Glenn’s Structural Dynamics Laboratory lists capabilities including sine sweep, sine dwell, sine chirp, sine burst, random vibration and classical shock testing.

This is a long way from placing a product on a table and turning up the vibration knob.

---

Why Frequency Matters

Not all vibration is equal.

A product may tolerate large amounts of vibration at one frequency and respond dramatically at another.

That is because structures have natural frequencies.

If energy is applied near one of those natural frequencies, the structure can resonate.

Think about pushing someone on a swing.

Push at random times and very little happens.

Push at exactly the right rhythm and the swing travels higher and higher.

Structures can behave similarly.

A bracket might barely move at 40 Hz but vibrate dramatically at 73 Hz.

A circuit board might develop a strong bending mode at another frequency.

A wire harness might begin moving against a nearby edge at yet another.

Finding those resonances is one reason vibration testing is so valuable.

---

What Is a Sine Sweep?

One common test is a sinusoidal vibration test.

The shaker applies vibration at a specific frequency and then gradually sweeps through a defined frequency range.

Engineers observe how the product responds.

A sine sweep can help reveal:

• resonant frequencies,
• loose components,
• structural modes,
• stiffness problems,
• unexpected amplification,
• and changes that occur after testing.

For example, engineers might perform a low-level sine sweep before a harsh test, conduct the qualification test, and then repeat the sweep afterward.

If a resonant frequency has shifted significantly, something about the structure may have changed.

Perhaps a joint loosened.

Perhaps something cracked.

The product might still look perfectly fine.

The vibration response can tell a different story.

---

Random Vibration Is Much Closer to Real Life

Real environments often contain many frequencies simultaneously.

Road vibration is not a perfectly clean sine wave.

Neither is aircraft vibration.

Neither is a rocket launch.

That is where random vibration testing comes in.

Instead of exciting one frequency at a time, random testing creates energy across a frequency spectrum.

The test is commonly described using a power spectral density, or PSD.

That sounds intimidating, but the basic idea is straightforward:

How much vibration energy exists at different frequencies?

The PSD profile describes the environment the shaker should reproduce.

IEC guidance for choosing vibration tests notes that real-world vibration is often dominated by random behavior, which is why random testing is commonly appropriate for environmental simulation.

Now the product experiences something much closer to the messy vibration spectrum it may encounter in actual service.

---

What Is Grms?

If you spend much time around vibration testing, you will eventually encounter Grms.

The “g” refers to acceleration relative to gravity.

The “rms” means root mean square.

Grms provides a useful overall measure of the energy in a random vibration environment.

But a single Grms number does not completely describe the test.

Two different PSD profiles can have the same overall Grms while concentrating their energy at very different frequencies.

And the product may respond very differently to each.

That is why simply saying:

“We tested it at 8 Grms.”

does not tell the whole story.

The frequency content matters.

---

Then There Is Shock

Vibration generally involves repeated oscillating motion.

Shock is different.

Shock is a relatively short-duration mechanical event.

Think:

• a pothole impact,
• a package hitting the floor,
• landing gear touchdown,
• a machine hitting a stop,
• a shipping impact,
• a weapon firing,
• stage separation,
• or explosive bolts releasing spacecraft hardware.

The acceleration can be severe.

The event may last only milliseconds.

But that can be more than enough time to break something.

IEC 60068-2-27 defines shock testing specifically to determine whether a product can withstand specified repetitive or non-repetitive shocks and to reveal mechanical weakness, performance degradation or accumulated damage.

---

And Then There Is Pyroshock

Spacecraft engineers get an especially dramatic version.

Pyroshock results from events involving pyrotechnic devices, such as separation mechanisms.

The shock travels through the spacecraft structure with extremely high-frequency content.

NASA’s Johnson Space Center operates dedicated pyroshock test stands capable of producing shock environments reaching up to 10,000 g over a frequency range of 100 to 10,000 Hz.

Ten thousand g sounds absurd.

But the duration is extremely short, and small spacecraft components may genuinely experience such environments.

That is why electronics intended for launch cannot simply be tested by shaking them the same way an automotive dashboard is tested.

The environment defines the test.

---

How Can Electronics Survive Thousands of g?

This is where acceleration can be misleading without context.

A human cannot survive being subjected to thousands of g for meaningful duration.

A tiny electronic component can sometimes survive an extremely short high-g shock because:

• the duration is very short,
• the mass is small,
• the structure is stiff,
• and the total displacement can be tiny.

But connections, brittle materials and internal components may still be vulnerable.

This is exactly what shock testing is intended to discover.

---

What Are Engineers Looking For?

Sometimes the answer is obvious.

Something cracks.

A bracket breaks.

A bolt comes loose.

A connector disconnects.

But many vibration failures are subtler.

Engineers may look for:

Fastener loosening

Repeated motion can gradually loosen threaded connections.

Connector fretting

Tiny relative movements between electrical contacts can damage contact surfaces, creating oxidation or increased electrical resistance.

Solder-joint fatigue

Repeated flexing of a printed circuit board can stress solder connections.

Wire chafing

A wire that touches another surface can slowly wear through its insulation.

Structural cracks

Repeated cyclic loading can initiate and grow fatigue cracks.

Component movement

Heavy electronic components may flex a circuit board or stress their leads.

Seal failures

Relative motion can degrade seals or create leakage paths.

Optical misalignment

Small movements can matter greatly in cameras, sensors and precision optical systems.

Noise and rattles

Not every vibration problem causes catastrophic failure.

Customers still become unhappy when a dashboard begins rattling at 8,000 miles.

That counts too.

---

The Fixture Can Ruin the Test

There is an unglamorous hero in vibration testing:

the test fixture.

The shaker needs some way to attach the product.

That fixture sounds simple.

It isn’t.

A badly designed fixture can introduce its own resonances or distribute loads differently from the real installation.

Then engineers may believe they are testing the product when they are actually testing the fixture.

Fixtures therefore have to be designed to:

• represent realistic attachment conditions,
• remain sufficiently stiff,
• transmit the desired vibration,
• avoid unwanted resonances,
• and withstand the test themselves.

NASA Goddard specifically notes that custom fixtures are designed and fabricated to interface test articles with vibration equipment during space-qualification testing.

Sometimes a large portion of the engineering effort goes into the thing holding the thing being tested.

---

Mounting Conditions Matter

The same product can behave completely differently depending on how it is mounted.

Imagine an electronic module.

Bolt it rigidly to an aluminum structure and it experiences one vibration environment.

Mount it using rubber isolators and it experiences another.

Attach it to a thin sheet-metal bracket and the bracket itself may amplify certain frequencies.

Testing therefore needs to reproduce the intended installation as closely as practical.

Otherwise the test may be repeatable and completely irrelevant.

---

One Axis Is Not the Whole Product

Products live in three dimensions.

Vibration can occur:

• vertically,
• laterally,
• longitudinally,
• rotationally,
• or in combinations.

Traditional shaker testing often tests one axis at a time.

The product is tested in one orientation, repositioned, and tested again.

But real environments do not necessarily cooperate so neatly.

That is driving increasing interest in multi-axis vibration testing.

IEC 60068-2-86:2024 provides procedures specifically for multi-exciter and multi-axis vibration systems intended to reproduce dynamic environments occurring during a product life cycle.

That is one example of environmental testing becoming more representative of actual operating conditions rather than treating loads independently simply because they are easier to reproduce in the laboratory.

---

Vibration and Temperature Can Conspire Against You

Here is another systems problem.

A product may pass vibration testing.

It may pass temperature testing.

And then fail when both happen together.

Why?

Because materials change with temperature.

Something that is stiff when cold may behave differently when hot.

A plastic part may soften.

Clearances change because materials expand at different rates.

A solder joint already weakened by thermal cycling may be more vulnerable to vibration.

A seal may respond differently.

A lubricant may change viscosity.

IEC 60068-2-53 specifically addresses combined climatic and dynamic testing, including combinations of vibration or shock with cold, dry heat, temperature change and damp heat.

Sometimes testing one stress at a time misses the failure created by their interaction.

---

This Is Where “Shake and Bake” Comes From

NASA itself uses the phrase.

JPL has described environmental testing of spacecraft as “Shake & Bake”—reproducing rocket vibration along with the extreme thermal and pressure environments spacecraft encounter during their missions.

It is amusing language for very serious engineering.

But it captures the philosophy beautifully.

Don’t wait for space to find the problem.

Build a little piece of space in the laboratory.

---

Test to Specification vs. Test to Failure

Not every vibration test has the same purpose.

Sometimes engineers are conducting qualification testing.

The requirement might specify a particular vibration environment for a defined duration.

The product must survive and remain functional.

Pass or fail.

But during development, engineers may deliberately keep going.

Increase the vibration.

Extend the duration.

Move outside the expected environment.

Why?

Because another extremely useful question is:

Where does it break?

That tells engineers about margin.

If a product is required to survive a certain environment but fails only at twice that level, the design may have healthy margin.

If it fails 2% beyond the requirement, the team may want to understand why.

Passing the specification and understanding the design are not always the same thing.

---

Qualification Testing Is Not the Same as Acceptance Testing

This distinction matters.

Qualification testing is generally intended to demonstrate that a design can survive the required environment.

It may use relatively severe levels.

The test article may be heavily instrumented.

The goal is to establish confidence in the design.

Acceptance testing is typically performed on production hardware to identify manufacturing defects and confirm that the individual unit is suitable for service.

You generally do not want every production unit subjected to a test severe enough to consume a significant portion of its fatigue life.

NASA’s Ames Engineering Evaluation Laboratory, for example, lists both spaceflight hardware qualification and acceptance testing among its environmental test services.

Same equipment family.

Different engineering question.

---

Vibration Can Reveal Manufacturing Problems Too

Suppose the design is excellent.

Production can still introduce variation.

A fastener may not receive the correct torque.

A cable clamp may be missing.

A solder joint may be poor.

An adhesive may not have cured correctly.

A component may have been installed incorrectly.

Environmental screening can expose workmanship defects that ordinary functional testing does not.

This is part of the reasoning behind screening approaches such as ESS and HASS, which we will get to separately.

Because sometimes the problem isn’t the design.

It is the particular unit that came off the line Tuesday afternoon.

---

Accelerated Testing Is a Time Machine—Sort Of

An engineer does not always want to wait ten years to discover whether ten years of vibration causes a failure.

So tests may accelerate exposure.

Higher levels, longer continuous testing or specially designed spectra can accumulate damage more quickly.

But accelerated testing has to be approached carefully.

Apply the wrong stress and you may create a failure mode that would never occur in normal service.

Then you haven’t accelerated reality.

You’ve invented a new problem.

That is why the goal is not merely:

Make it harsher.

It is:

Create a useful representation of the life-cycle environment.

---

Where Does the Test Profile Come From?

Ideally, from reality.

Engineers can instrument:

• vehicles,
• aircraft,
• ships,
• machinery,
• shipping containers,
• launch vehicles,
• or operating equipment

with accelerometers and collect actual field data.

That data can then help define a laboratory vibration profile.

Standards also provide established test methods and environments for particular industries and use cases.

The difficult part is determining whether the standard profile is appropriate for your product.

Passing a standard is valuable.

Reproducing the actual environment can be even more valuable.

---

Standards Are Not Magic

Vibration and shock testing frequently references standards such as:

• IEC 60068,
• MIL-STD-810,
• automotive OEM requirements,
• aerospace qualification requirements,
• transportation standards,
• electronics specifications,
• and company-specific procedures.

Standards can define:

• test methods,
• fixture requirements,
• spectra,
• durations,
• orientations,
• instrumentation,
• and pass/fail criteria.

But engineers still need to understand why the test applies.

A statement like:

“Tested to MIL-STD-810”

can be misleading without identifying the method, procedure, test conditions and tailoring used.

Environmental testing works best when the requirement represents the real intended environment rather than when the most impressive standard name appears in marketing material.

---

The Product Should Often Be Running While You Shake It

There is a major difference between:

Did the product break structurally?

and

Did the product continue operating correctly while vibrating?

For electronics, sensors and electromechanical systems, engineers may monitor the product continuously during the test.

Maybe a connector disconnects for only 20 milliseconds.

After the shaker stops, everything works again.

If engineers inspect the product only afterward, they may conclude:

PASS.

But if that 20-millisecond interruption causes an aircraft controller to reset, a vehicle sensor to disappear from a network or an industrial safety system to fault, it matters enormously.

That is why functional monitoring during environmental testing can be essential.

---

Instrument the Test Article

Accelerometers are among the most obvious sensors used in vibration testing.

But depending on the product, engineers may also monitor:

• strain,
• displacement,
• force,
• electrical continuity,
• voltage,
• current,
• temperature,
• pressure,
• acoustic response,
• network traffic,
• optical alignment,
• and product-specific functional signals.

NASA’s structural test facilities describe synchronized acquisition of strain, load, displacement and acceleration measurements for detailed structural-response evaluation.

The shaker creates the environment.

The instrumentation tells engineers what the product is doing inside it.

---

Sometimes the Product Amplifies the Input

This is another concept that can surprise people.

Suppose the shaker table experiences 2 g at a particular frequency.

A flexible component mounted above it might experience 8 g.

Why?

Resonance.

The structural response can amplify the input.

That is why measuring only what the shaker is doing may not be enough.

Engineers often place response accelerometers at critical locations on the product.

The important load may be the one occurring on the circuit board, sensor or bracket—not the acceleration at the mounting interface.

---

Finding Resonance Before It Finds You

One useful outcome of vibration testing is identifying resonances that coincide with operating frequencies.

Imagine an industrial machine with a rotating motor that normally operates around a particular RPM.

If that operating speed excites a structural resonance, the product may spend its entire life shaking itself at exactly the wrong frequency.

An engineer may respond by:

• increasing stiffness,
• changing mass,
• adding damping,
• relocating a component,
• changing isolation,
• or changing operating speed.

Sometimes solving a vibration problem does not require making everything stronger.

It requires moving the resonance somewhere else.

---

What Software Is Involved?

A modern vibration test system involves much more software than the shaker itself.

Software may control:

• vibration profiles,
• shaker drive signals,
• feedback loops,
• accelerometer channels,
• safety limits,
• test sequencing,
• data acquisition,
• spectral analysis,
• reporting,
• and test automation.

Engineering analysis tools may also be used before and after the physical test.

Finite element analysis can predict:

• natural frequencies,
• mode shapes,
• stresses,
• deformation,
• and structural response.

Test engineers can then compare simulation with measured results.

When they disagree, something interesting has been learned.

---

Modal Testing Helps Connect Simulation and Reality

A structural model may predict that a component’s first natural frequency is 112 Hz.

Physical modal testing may show 96 Hz.

That difference could indicate:

• incorrect material properties,
• unrealistic boundary conditions,
• insufficient model detail,
• joint behavior not represented correctly,
• manufacturing variation,
• or measurement/setup differences.

This is one of the valuable relationships between simulation and physical testing.

Simulation predicts.

Testing measures.

Engineers reconcile the two.

A validated model can then become much more useful for future design decisions.

---

Correlation Is Better Than Competition

There is a bad way to think about simulation and test:

Which one is right?

A better question is:

What can we learn from the difference?

Physical testing has uncertainties.

Sensors have tolerances.

Fixtures influence behavior.

Simulation has assumptions.

Models simplify reality.

Material properties may be imperfect.

When test results and simulation correlate closely, confidence increases.

When they don’t, that discrepancy deserves investigation.

Neither should automatically win.

---

Why Not Just Drive the Car for Ten Years?

Because engineers need answers before the product reaches the customer.

Physical field testing remains enormously valuable.

Automotive manufacturers drive prototypes for millions of miles.

Aircraft undergo flight testing.

Marine products go to sea.

Industrial equipment runs durability cycles.

But laboratory vibration testing provides something field testing cannot always offer:

control and repeatability.

The same profile can be run:

• before a design change,
• after a design change,
• against two suppliers,
• against multiple prototypes,
• or years later when a failure investigation begins.

That repeatability makes comparison possible.

---

Sometimes Engineers Deliberately Find the Weakest Part

Imagine a bracket fails during vibration testing.

The instinct might be:

Make the bracket thicker.

That may fix the immediate failure.

But the better question is:

Why did the bracket fail?

Was it resonance?

A stress concentration?

Poor material?

Incorrect mounting?

Unexpected load transfer?

Manufacturing variation?

A loose fastener?

A fixture problem?

If engineers strengthen the bracket without understanding the system, the load may simply move somewhere else.

Now the next weakest component fails.

Testing should create understanding, not merely thicker parts.

---

The Rattle Test Is Engineering Too

Not every vibration chamber is trying to prevent catastrophic failure.

Sometimes engineers are chasing:

buzz, squeak and rattle.

Automotive manufacturers in particular spend considerable effort identifying noises that may make a customer believe a vehicle is poorly built.

A trim panel can be structurally safe and functionally perfect while producing an irritating buzz at a particular engine speed.

That’s a quality problem.

And vibration testing can reproduce it.

Human perception is part of product validation too.

---

Shipping May Be the Most Violent Part of a Product’s Life

A product can survive years of normal operation and be damaged before the customer even opens the box.

Transportation introduces:

• truck vibration,
• rail vibration,
• aircraft handling,
• pallet impacts,
• forklift events,
• package drops,
• and repeated loading.

That means some vibration and shock testing focuses not on product operation but on distribution.

Packaging becomes part of the engineering system.

A delicate product may not need to become stronger if better packaging can isolate it from the environment.

Again:

the system matters.

---

Shock Testing Is Not the Same as Drop Testing

The concepts overlap, but they are not identical.

A controlled shock machine may apply a precisely defined acceleration pulse to a mounted test article.

A drop test allows a product or package to physically impact a surface.

Both can create shock.

But they answer different questions.

We will abuse perfectly good products with gravity in a separate article.

For now, the important distinction is that shock refers to the mechanical event.

Dropping something is only one way to create it.

---

How Hard Should You Shake It?

This sounds like a simple question.

It isn’t.

Too little testing and the product may fail in service.

Too much testing and engineers may:

• damage an otherwise adequate design,
• create unrealistic failure modes,
• consume fatigue life,
• waste test time,
• or add unnecessary weight and cost trying to solve nonexistent problems.

The correct test severity depends on:

• intended environment,
• exposure duration,
• mounting,
• expected life,
• product variability,
• required margin,
• regulatory or customer requirements,
• and the purpose of the test.

Environmental testing should be severe enough to reveal relevant weaknesses.

Not severe simply because the shaker can go higher.

---

And How Long?

Duration matters too.

A low-level vibration repeated for hundreds of hours may produce fatigue damage that a short high-level test does not reproduce correctly.

Conversely, a severe launch environment may exist for only minutes.

The test should represent the physics of the expected damage mechanism.

This is why engineers cannot reduce environmental qualification to:

More abuse = better testing.

---

The Best Test Often Begins With “What Actually Happens?”

Before writing the vibration requirement, ask:

• Where will the product be used?
• How is it mounted?
• What generates vibration?
• What frequencies matter?
• How long will exposure last?
• Does the product operate during exposure?
• What failures would matter?
• What other environmental stresses occur at the same time?
• How variable is the real environment?
• What happens during shipping?
• What happens during maintenance?
• What happens during an accident or abnormal event?

Those questions create the test.

The shaker simply executes it.

---

Verification Means Knowing What Passed

Suppose the product survives.

What exactly has been verified?

Perhaps:

The product remained structurally intact after exposure to the specified vibration profile.

That does not automatically mean:

The product will never experience a vibration failure.

It certainly does not mean:

The product is indestructible.

Engineering verification is always tied to:

• configuration,
• conditions,
• procedure,
• limits,
• and acceptance criteria.

This is why good test documentation matters.

---

And Sometimes Failure Is the Best Result

Nobody enjoys watching an expensive prototype break.

But consider the alternatives.

A bracket cracks on the shaker.

Engineers find the cause.

They modify the design.

The next prototype survives.

Production begins.

Thousands of customers never know the problem existed.

That broken prototype was not necessarily a testing failure.

It may have been a spectacular testing success.

The truly bad outcome is the prototype surviving a weak test and the production product failing in the field.

---

Why Engineers Shake Perfectly Good Products

Because products do not live on engineering workbenches.

They live on:

• highways,
• railways,
• ships,
• aircraft,
• factories,
• launch vehicles,
• construction sites,
• and delivery trucks.

They vibrate.

They resonate.

They get hit.

They fatigue.

Fasteners loosen.

Wires move.

Solder joints crack.

Structures flex.

And sometimes a failure begins with a movement so tiny nobody can see it.

Vibration and shock testing bring those environments into the laboratory where engineers can control them, measure them and repeat them.

The goal is not to destroy a perfectly good product.

The goal is to discover whether it is really a good product before the world gets a chance to shake it first.

And if something breaks?

Good.

Now we know where to start.