Explosions are measured by capturing how pressure, light, heat, and ground motion change over microseconds to milliseconds, converting those signals into calibrated units such as overpressure in psi or bar, impulse in psi-milliseconds, peak sound pressure in decibels, and yield expressed as TNT-equivalent mass. Blast pressure sensors, piezoelectric and piezoresistive transducers, high-speed cameras, radiometers, seismographs, and post-blast forensic mapping each measure a different physical channel. The core challenge is that a real detonation front moves faster than sound in the surrounding air, so instruments must sample fast enough to resolve the shock before it smears into an averaged reading.

The physical channels an explosion actually produces

A detonation releases energy in a thin reaction zone that travels through the explosive faster than the speed of sound in that material. In condensed explosives, detonation velocities commonly fall in the range of roughly 2,000 to 9,000 meters per second. That reaction zone generates a shock front in the surrounding medium. What instruments measure is the medium's response, not the chemistry itself.

In air, the leading shock has a near-instantaneous rise to peak overpressure, then decays exponentially. Behind the shock is a negative phase where pressure falls below ambient before recovering. The whole positive-pressure pulse may last from a fraction of a millisecond for small charges to tens of milliseconds for large ones. Overpressure is the peak above ambient pressure; impulse is the integral of pressure over time, which captures how long the load pushes on a structure or body.

Thermal output arrives as a fireball whose radiant power is measured in watts and expressed as radiant exposure in joules per square centimeter. Ground shock travels as seismic waves picked up by accelerometers or geophones. Fragmentation is a separate ballistic problem measured by fragment velocity and mass distribution. Measuring an explosion therefore means choosing which channel answers the question being asked.

A blunt way to say it: an explosion is not one number, it is a bundle of physical signals, and every instrument listens to only part of the bundle.

Instruments and what each one resolves

Blast pressure sensors dominate engineering measurement. Piezoelectric transducers exploit quartz or ceramic crystals that generate charge under compression, giving microsecond response for peak overpressure. Piezoresistive gauges use a strained element whose resistance changes, offering accurate readings across a wide range. Both are typically flush-mounted on rigid plates so the shock does not reflect off the housing before reaching the sensing element.

High-speed cameras record fireball growth and shock arrival at frame rates from 10,000 to over a million frames per second. The shock front is often visible directly because it refracts light differently from undisturbed air. Radiometers and calorimeters measure thermal radiation in bands, using calibrated detectors to convert detector voltage into irradiance.

Seismographs and accelerometers record ground motion in velocity or acceleration, with frequency content telling analysts whether energy came from a subsurface burst or a surface charge. Sound-level meters record the audible boom, though a blast's peak can exceed the meter's linear range and clip.

Post-blast forensic mapping measures crater dimensions, fragment scatter patterns, and structural deformation. These give retrospective estimates of charge mass and position when no instrument survived the event.

How overpressure becomes a number

Pressure transducers produce a voltage proportional to applied pressure. Calibration against a known reference converts that voltage into pressure units such as psi, bar, or kilopascals. Atmospheric pressure is about 14.7 psi at sea level, so an overpressure of 5 psi means the blast raised local pressure to roughly 19.7 psi momentarily. Analysts report peak overpressure and impulse together because the same peak can do very different damage depending on duration.

Scaling laws let engineers predict blast effects. The Hopkinson-Cranz scaling law relates blast parameters to distance scaled by the cube root of charge mass. This means doubling charge mass increases the lethal radius far less than doubling would suggest. Yield is often expressed as TNT equivalent mass, based on energy release per mass of TNT. That conversion is a convention, not a universal truth, because different explosives release energy at different rates and partition it differently among blast, heat, and fragmentation.

High-speed instrumentation must satisfy the Nyquist criterion: sample at more than twice the highest frequency of interest. A shock with a rise time of 10 microseconds requires sampling intervals well below that to avoid underestimating peak pressure. Analysts also separate free-field pressure from reflected pressure, because a gauge facing the blast records a reflection that can be many times higher than the incident shock.

A professional caveat worth stating plainly: a peak overpressure number without geometry, gauge orientation, and distance is nearly meaningless. The same charge yields very different readings depending on where the gauge sits relative to the shock path.

What the numbers mean in practical terms

Blast effects on structures and people have been characterized through decades of testing. Overpressures around 1 psi typically break windows and cause minor debris. By roughly 5 psi, many residential structures sustain significant damage and eardrum rupture becomes likely. Higher overpressures produce lung injury and structural collapse. These thresholds are approximate ranges, not sharp cutoffs, because exposure duration and orientation matter.

There is no universal blast-exposure standard comparable to occupational noise limits for the general public. Workplace noise regulations address continuous sound, not impulsive blasts. Analysts therefore rely on empirical damage functions and event-specific models rather than a single legal limit. This is a genuine limitation of the field, not an oversight.

Thresholds side by side

Measure Typical unit What it captures Practical note
Peak overpressure psi or bar Peak pressure above ambient Pair with impulse, not alone
Positive-phase impulse psi-milliseconds Pressure integrated over time Governs structural loading
Yield TNT-equivalent mass Total energy release Convention-dependent, not exact
Peak sound pressure decibels Audible boom intensity Clips on unsuited meters
Radiant exposure joules per square centimeter Thermal dose Fireball duration matters
Ground acceleration g or meters per second squared Seismic coupling Frequency content reveals source
Crater volume cubic meters Retrospective charge estimate Soil type strongly affects result

Each row answers a different question. Confusing them is one of the most common analytical errors.

Edge cases that break simple measurements

Confined and unconfined explosions behave differently. A charge in open air sends a shock outward, while a charge in a tunnel or building experiences reflections that raise local overpressure dramatically. Gauges placed inside a structure may read pressure multiplied by multiple reflections, not the free-field value.

Environmental conditions shift readings. Temperature, humidity, and wind affect shock propagation and thermal transmission. Altitude changes ambient pressure, which changes the reference point for overpressure. Analysts normalize to standard conditions when comparing events.

Large-scale and improvised explosive events pose forensic challenges. Without instrumented recordings, investigators rely on crater analysis, fragment trajectories, and structural damage patterns. These methods have substantial uncertainty, often estimated as factors rather than precise numbers. The limitation must be stated, not hidden inside a hopeful point estimate.

Near-field measurements are especially hard. Close to a charge, pressure gradients are steep and sensors may be destroyed before recording the full pulse. Gauge survival is itself a measurement problem. Some designs sacrifice survivability for fidelity, others survive but saturate.

Common misconceptions about explosion measurement

The most persistent myth is that "blast pressure" is a single number. Peak overpressure without duration, impulse, and geometry cannot describe the hazard. A short spike at 10 psi and a longer plateau at 10 psi load a wall very differently.

Another misconception is that TNT-equivalent mass is a fixed energy conversion. In reality, energy yield, gas production, and fragmentation differ across explosives. Analysts use TNT equivalence as a modeling convention, not a measurement. Stating a yield to three significant figures implies precision that does not exist.

People often expect sound-level meters to capture blast loudness. Most are not designed for impulsive peaks and will clip or average the event. A decibel reading of a detonation is not a pressure measurement and should not be treated as one.

Red flag: a single peak pressure number offered without distance, gauge orientation, or duration. Red flag: yield quoted to fine precision without a stated conversion convention. Red flag: decibel readings presented as overpressure equivalents. Red flag: crater-based charge estimates given without uncertainty range. Red flag: thermal exposure values stated without fireball duration or distance.

How the event unfolds in time

Time window What happens What is measured
Nanoseconds to microseconds Detonation reaction zone propagates Detonation velocity and front shape
Microseconds to milliseconds Shock forms in surrounding medium Peak overpressure and rise time
Milliseconds Positive pressure phase decays Impulse and decay curve
Milliseconds to tens of milliseconds Negative phase follows Pressure deficit and recovery
Tens of milliseconds to seconds Fireball radiates and grows Radiant exposure and duration
Hundreds of milliseconds Fragments accelerate and disperse Fragment velocity and mass
Seconds to minutes Ground shock attenuates Seismic amplitude and frequency
Minutes to days Post-blast evidence solidifies Crater geometry and damage patterns

The table makes one point clear: no single instrument covers the whole event. Measurement is inherently staged, and conclusions must combine channels.

Practical actions for anyone measuring or interpreting blast data

If you are collecting measurements, calibrate every transducer against a reference immediately before and after the event, because blast sensors can shift under shock loading. Mount pressure gauges flush to rigid surfaces and document orientation and distance to the charge precisely. Use sampling rates high enough to resolve the fastest expected rise time, and record raw waveforms rather than only processed peaks.

If you are interpreting published or reported numbers, demand the complete bundle: peak overpressure, impulse or duration, distance, gauge geometry, charge type and mass, and the conversion convention used for yield. Compare against empirical damage functions rather than converting everything to a single pressure figure. Treat retrospective estimates from craters or damage as ranges, not values.

Before any controlled blast, define the physical channel that matters for the decision. Structural loading needs pressure and impulse. Thermal safety needs radiant exposure and duration. Human exposure needs both pressure and duration plus orientation. Matching instrument to question is the first step, and it is the step most often skipped.

The limits of what explosion measurement can tell you

No instrument reconstructs an explosion completely. Every measurement channel captures only part of the energy budget, and the partitioning among blast, heat, fragmentation, and ground coupling varies with charge design and environment. Models based on TNT equivalence inherit the assumptions baked into that convention.

Forensic reconstruction after an uninstrumented event carries inherent uncertainty. Crater and fragment analysis can indicate charge size and position within factors, not exact values. Analysts should communicate that uncertainty honestly rather than present a false precision.

Field measurement of large or improvised events is often not possible in real time, so the science of explosion measurement is partly the science of inference. The most defensible conclusion combines multiple independent channels, states assumptions, and reports ranges. Measurement tells you what the medium did; explaining how explosions are measured means being explicit about which question each number answers, and which it does not.