Shock Response Spectrum: Why a Drop Shakes Every Module Differently

A phone slips off a table or a smart speaker gets knocked off a shelf. The housing gets a scuff and everything seems fine. Then, weeks later, a camera stops focusing or a connector starts to drop out.

What happened? The outer shell felt one shock. But inside the product, each module -- the main board, the camera, the battery, the connectors -- felt its own version of that shock, sometimes much stronger than what the housing saw. The tool engineers use to understand this is the shock response spectrum, or SRS.

A Shock Is Not Just "How Many g"

When people describe a shock, they often quote a single number: "the product saw 1000 g." That number tells you the peak acceleration of the input, but it hides two things that matter just as much:

How long the shock lasts. A very short, sharp knock and a longer, softer thump can have the same peak g but act very differently on the parts inside.
What the parts are "tuned" to. Every part that can flex -- a board, a camera on a bracket, a battery on its adhesive -- has natural frequencies at which it likes to vibrate.

A shock is a mix of frequencies. Short pulses carry energy up to high frequencies; long pulses mostly carry low-frequency energy. Whether a module gets hurt depends on how much of the shock lands near its own natural frequencies.

The SRS in Plain Words

Imagine a long row of small mass-on-spring systems, each one tuned to a different natural frequency -- 50 Hz, 100 Hz, 200 Hz, all the way up to several thousand Hz. Engineers call each of these a single-degree-of-freedom (SDOF) oscillator.

Now bolt the whole row onto a table and hit the table with the shock you care about. Each little oscillator gets shaken by the same input, but each one responds differently:

Oscillators tuned far below the frequency content of the shock barely move relative to the table. The shock is over before they can react.
Oscillators tuned near the dominant frequency content swing hard. Their peak response can be clearly larger than the input -- this is amplification.
Oscillators tuned far above simply ride along with the table. Their peak acceleration ends up about equal to the input peak.

Record the maximum response of each oscillator and plot it against its natural frequency. That curve is the shock response spectrum.

The key idea: the SRS does not describe the shock itself. It describes what the shock does to anything with a given natural frequency. That makes it a very practical tool for designers, because every module inside a product can be placed somewhere on that frequency axis.

One more ingredient is damping, the energy loss in each oscillator. Real structures have some damping, so SRS curves are usually computed with an assumed damping value -- a commonly used textbook choice is a quality factor Q of 10, which corresponds to 5 % of critical damping. Lower damping means higher peaks.

Why Modules Inside the Same Product React Differently

Inside a typical consumer device, you might find (illustrative values only):

A large, thin main board, held by a few screws, with a lowest natural frequency in the few-hundred-Hz range.
A small camera module on a stiff bracket, with natural frequencies much higher.
A battery held mainly by adhesive, behaving like a heavy mass on a soft spring.
Board-to-board and flex connectors, small and light, but sitting on the board that moves underneath them.

Put all of them on the same SRS curve and you can see at a glance who is in trouble. If the board's natural frequency sits near the peak of the SRS, the board will bend the most -- and everything soldered onto it bends with it. A camera tuned well above the peak may see only modest amplification. The battery, if it is soft-mounted and heavy, may lag behind and then hit its neighbours.

Half-Sine Pulses and Public Test Standards

To make shock testing repeatable, labs usually do not drop products onto random floors. They use a shock machine that delivers a defined pulse shape. The most common shape is the half-sine: acceleration rises smoothly to a peak and falls back to zero, like the top half of a sine wave. It is described by two numbers: the peak acceleration and the pulse duration.

Two well-known public standards describe such tests (reference only -- each product team sets its own levels):

IEC 60068-2-27 (Environmental testing -- Test Ea: Shock) defines pulse shapes such as half-sine, trapezoidal and sawtooth, with tolerance bands and test procedures.
MIL-STD-810, Method 516 (Shock) describes shock procedures for equipment, including the use of SRS to specify and compare shock environments.

A useful rule of thumb from textbook shock theory: for a half-sine pulse, the strongest amplification occurs for oscillators whose natural period is comparable to the pulse duration. Shorter pulses push the peak of the SRS toward higher frequencies; longer pulses toward lower ones. So the same peak g can be harmless for one module and severe for another, purely because of the pulse length.

How Simulation Finds the Worst Amplification

Physical tests tell you whether a sample failed; simulation tells you why and where, before hardware exists. Finite element analysis (FEA) offers two classic approaches:

Transient (time-domain) analysis. The mounting points of the module are driven with the actual acceleration pulse, and the solver tracks how the structure moves, bends and rings over time. This captures nonlinear effects such as contact, gaps closing and parts hitting each other.
Modal plus response-spectrum analysis. First compute the natural frequencies and mode shapes of the module. Then use the SRS to estimate the peak response of each mode and combine them. This is fast and gives very clear insight into which mode drives the response.

The output that matters for design is not only "peak g" at the module. Engineers look at:

Board bending -- curvature and strain on the PCB surface, where solder joints and traces sit.
Relative motion -- how far one part moves relative to its neighbour, which tells you about impacts and connector disengagement.
Forces at mounting points -- screws, bosses, adhesive and clips that hold the module.

PCB Bending and Solder-Joint Risk

For electronics, the most common shock victim is the solder joint. When a board bends, its surface stretches on one side and compresses on the other. Large, stiff components -- especially area-array packages with many small solder balls -- cannot bend with the board. The solder joints at their corners take the difference and can crack.

At a high level, the risk grows with:

The amount of board curvature near the component.
The size and stiffness of the component.
Proximity to the board's maximum bending zone, often the middle of an unsupported span or near a screw that acts as a pivot.

Design Levers

Once simulation shows which module sees the worst amplification, the design team has several levers:

Stiffening. Add ribs, a stiffener plate or a thicker board to raise the natural frequency and reduce bending.
Mounting points. Add a screw or support near the middle of a long span, or move an existing one away from sensitive components. Shorter spans mean higher frequencies and less curvature.
Component placement. Keep large packages and connectors away from the zones of maximum bending.
Damping. Foam, gaskets or damping pads lower the peak response near resonance.
Moving natural frequencies. The core idea of the SRS: if a module's natural frequency sits in the worst part of the spectrum, shift it up or down, out of the amplified band.

Each lever costs mass, space or money; simulation compares the options side by side before any prototype is built.

Closing Thoughts

The shock response spectrum turns "how bad was that drop?" into a practical question: "which of my modules is tuned to the part of the shock that hurts the most?" Combined with simulation, it finds weak modules early and guides design changes with confidence.

AutoCAE.AI builds AI-assisted simulation workflows that help hardware teams find these risks earlier and faster. To learn more, visit www.autocae.ai.