What Is Vibration Science — And Why It Has Monitored Machines for Over 100 Years
Every machine under load vibrates. That is not a design flaw. It is physics. The forces acting on energized cores, pressurized fluid systems, and current-carrying conductors all produce mechanical motion. Most of it is too small to feel. All of it is measurable.
Engineers have known this for a long time. Condition monitoring through vibration analysis predates digital sensors, computerized data acquisition, and the term "predictive maintenance" by decades. In the early industrial era, skilled technicians would press a screwdriver handle to a bearing housing and listen through the tool. What they heard told them whether a machine was running normally or building toward a failure. The physics behind that method has not changed. The instrumentation has.
Why Machines Vibrate
Vibration is the mechanical response to force. When any structure is subjected to a periodic force, it oscillates. The frequency and amplitude of that oscillation carry information about the force producing it.
In rotating equipment (motors, generators, pumps) the dominant vibration source is rotational imbalance. A shaft spinning at 60 Hz produces vibration at 60 Hz. A bearing defect produces a secondary frequency determined by its geometry and the shaft speed. Engineers use those frequencies the way a physician uses a stethoscope: to identify what is producing the signal and whether it indicates something normal or something worth investigating.
Power transformers do not rotate. But they vibrate for reasons that are just as specific and just as measurable.
Magnetostriction — The Transformer's Heartbeat
A transformer's core is made from stacked laminations of electrical steel. When alternating current energizes the core, the magnetic field reverses direction at the power frequency: 60 Hz in North America. The steel laminations physically expand and contract in response to that reversing field. This phenomenon is called magnetostriction.
Because the core responds to the magnetic field regardless of its polarity, it expands and contracts twice per power cycle. In a 60 Hz system, the core vibrates at 120 Hz. That frequency is consistent, predictable, and structure-borne. It travels through the steel frame of the transformer and into the tank walls.
The amplitude of that vibration is not fixed. It depends on the magnetic flux density in the core, which in turn depends on the applied voltage. A transformer operating above rated voltage drives the core toward saturation. The magnetostrictive response becomes nonlinear, and harmonics appear in the vibration spectrum. Those harmonics are diagnostic signals. They indicate that something in the magnetic circuit has changed.
Lorentz Forces on Windings
The transformer's windings carry current. When current flows through a conductor in a magnetic field, the conductor experiences a force. This is the Lorentz force. In a transformer, the primary and secondary windings interact electromagnetically, and the resulting forces act on the winding conductors.
Under normal load, those forces are radial and press the conductors outward from the core. Under fault conditions, they can become axial, compressing the windings along the core axis. A short-circuit event can produce a force spike many times larger than normal operating forces. Even below fault threshold, sustained mechanical stress accumulates over a transformer's service life.
Lorentz force vibrations travel a different path than core magnetostriction. They couple into the surrounding oil before they reach the tank wall. Oil transmits pressure waves. The frequency content of winding-sourced vibration reflects the current waveform, including harmonic content from nonlinear loads.
Two Paths, Two Kinds of Information
The distinction between structure-borne and oil-borne vibration is not academic. It is diagnostic.
Magnetostrictive vibration from the core travels through the transformer's steel structure. Winding vibration couples into the oil and arrives at the tank wall as a pressure wave. A sensor mounted on the tank wall receives both signals simultaneously. Reading them correctly requires understanding their origins.
A rising vibration signal at frequencies associated with winding excitation points to mechanical change in the winding: looseness, deformation, or deteriorating insulation. A signal reflecting changes in the structural transmission path points toward the core laminations or the connection between core and tank. A signal that shows attenuation or distortion in the oil path makes the condition of the oil itself part of the diagnostic picture.
This is the layer of understanding that makes vibration science useful for transformer diagnostics. Measuring vibration is not enough. The measurement must be interpreted against the physics of where the signal originated and how it traveled to the sensor.
A Field That Rewards Continuous Study
Vibration science did not wait for AI to become credible. Railway engineers analyzed axle vibrations in the nineteenth century. Bearing failure prediction through frequency analysis was established by the mid-twentieth century. The techniques that underpin modern machine health monitoring were developed by people working with analog instrumentation, strip chart recorders, and physical contact with the machines themselves.
What those engineers understood, and what the field has continued to refine, is that every machine has a characteristic vibration signature under normal operating conditions. Change in that signature indicates change in the machine. The diagnostic task is to detect the change, trace it to its source, and act before the change becomes a failure.
VIE applied modern sensor hardware and machine learning to that foundation. Continuous measurement across the full vibration spectrum captures each transformer's operating signature in detail. Machine learning algorithms establish what normal looks like for each individual unit. When the signature shifts, the system produces a leading indicator: a directional signal that something is changing, before that change becomes a fault.
The science is not new. The continuous, fleet-scale application of it to power transformers, without requiring manual inspection or periodic test campaigns, is what VIE built.