Some of the most important things about a rotor only show up while the machine is changing speed, not while it runs steadily. A Bode plot is the tool that captures them: it records the running-speed vibration amplitude and phase as the shaft accelerates through its speed range on startup or slows down on coastdown, and plots both against RPM. The pattern that emerges, a peak in amplitude paired with a sweep in phase, is the signature of a critical speed. This guide explains what a Bode plot shows, how the amplitude peak and the phase shift together locate a critical speed, and how they confirm a resonance rather than a plain forced response.
Bode plot in one line: A Bode plot is a pair of graphs showing the 1X vibration amplitude and phase of a machine plotted against shaft speed, captured while the machine runs up or coasts down through its speed range. As the shaft passes through a critical speed, the amplitude rises to a peak and the phase sweeps through roughly ninety degrees at the peak and about one hundred eighty degrees overall. That combined signature of an amplitude peak with a matching phase change is what locates a critical speed and distinguishes a true resonance from a simple forced response that would not show the phase sweep.
A Bode plot is built from transient data taken while the machine is changing speed, which is what makes it different from the steady-state measurements taken at running speed. During a startup the shaft accelerates from rest up to operating speed, and during a coastdown it decelerates back to rest, and throughout that transit the system continuously measures the 1X amplitude and phase and records them against the instantaneous shaft speed. The keyphasor provides the speed and the phase reference, so each amplitude and phase reading is tagged with the RPM at which it was taken.
The result is two curves plotted against RPM rather than against time or frequency. The upper curve is the 1X amplitude versus speed, showing how the running-speed vibration grows and shrinks as the machine sweeps through its range. The lower curve is the 1X phase versus speed, showing how the timing of the vibration relative to the keyphasor changes as speed changes. Reading the two together, at matching speeds, is the whole method, because it is the correspondence between what the amplitude does and what the phase does that carries the diagnosis.
Because it can only be captured while the machine is changing speed, the Bode plot is a startup and coastdown tool, and it reveals behavior that steady-state monitoring at a fixed speed cannot. A machine sitting at a constant running speed never sweeps through the speeds where its critical speeds and resonances live, so those features are invisible in steady operation. Only by watching the amplitude and phase across the whole speed range, during run-up or coastdown, do the critical speeds announce themselves, which is why these transients are deliberately captured.
A critical speed is a rotor speed at which the running-speed excitation coincides with a natural frequency of the rotor system, so the rotor resonates. On the Bode plot this appears as a clear peak in the amplitude curve: as the shaft speeds up toward the critical speed the 1X amplitude climbs, reaches a maximum at the critical speed, and falls again as the shaft moves past it. The speed at which the amplitude peaks marks where the critical speed is, which is exactly the information the plot is captured to find.
The phase curve confirms and refines that location. As the shaft passes through the critical speed, the 1X phase does not stay constant; it sweeps through a large change, passing through roughly ninety degrees of shift right at the amplitude peak and changing by about one hundred eighty degrees across the full passage through the resonance. This phase sweep is a hallmark of passing through a resonance, and the point where the phase is changing most rapidly lines up with the amplitude peak, so the amplitude and phase together pin the critical speed more reliably than either would alone.
Using both curves is what makes the identification trustworthy. An amplitude peak by itself could be mistaken for something else, but an amplitude peak that coincides with a rapid, roughly ninety-degree phase shift at that same speed is the specific fingerprint of a critical speed. Analysts read the plot by finding where the amplitude maxes out and checking that the phase is sweeping through its transition at the same RPM. Once located, knowing where a machine's critical speeds sit matters for operation, since running steadily at or near a critical speed produces high vibration and must be avoided.
The Bode plot's ability to separate a true resonance from a mere forced response is one of its most valuable uses. A forced response is high vibration caused simply by a large exciting force, such as significant unbalance, and it does not necessarily involve a natural frequency. A resonance is high vibration caused by exciting a natural frequency, where the response is amplified far beyond the force alone. The two can produce similar amplitudes, but they behave completely differently in phase, and that is how the Bode plot tells them apart.
The distinguishing feature is the phase behavior around the peak. A genuine resonance always shows the characteristic phase sweep through about ninety degrees at the amplitude peak, because passing through a natural frequency inherently reverses the phase relationship between the force and the response. A forced response that is not a resonance does not produce that sharp phase sweep at the amplitude peak. So when an analyst sees an amplitude peak accompanied by the ninety-degree phase shift, it confirms a resonance at a critical speed, and when a peak appears without that phase behavior, the cause is more likely a forced response rather than a structural resonance. This distinction guides whether the fix is balancing to reduce the force or addressing the resonant condition.
Capturing a Bode plot requires collecting the amplitude and phase continuously through the speed change, which a monitoring system does by sampling the 1X vibration and the keyphasor throughout the startup or coastdown. A cloud platform such as Merobix can retain those transient captures alongside the machine's steady-state trends, so the critical-speed behavior recorded each time the machine starts or stops is available to an analyst remotely. Comparing Bode plots from successive startups is itself diagnostic, because a critical speed that has shifted or an amplitude peak that has grown between run-ups signals a change in the rotor system. Keeping the run-up and coastdown captures, not just the running-speed data, is how remote monitoring preserves the transient behavior that only appears while the machine changes speed.
A Bode plot shows a machine's 1X vibration amplitude and phase plotted against shaft speed, captured while the machine runs up or coasts down through its speed range. The amplitude curve shows how the running-speed vibration grows and shrinks with speed, and the phase curve shows how the timing of that vibration changes with speed. Read together, they reveal critical speeds, which appear as an amplitude peak coinciding with a rapid phase shift, information that steady-state monitoring at a fixed speed cannot capture.
A critical speed appears as a peak in the amplitude curve, where the 1X vibration climbs to a maximum as the shaft passes through that speed and falls off afterward. At the same speed the phase curve sweeps through roughly ninety degrees, with about one hundred eighty degrees of total change across the full passage through the resonance. The speed where the amplitude peaks and the phase is changing most rapidly is the critical speed, and using both curves together locates it more reliably than either alone.
The difference is in the phase behavior around the amplitude peak. A true resonance at a critical speed always shows a characteristic phase sweep of about ninety degrees at the peak, because passing through a natural frequency inherently reverses the phase between the exciting force and the response. A forced response driven simply by a large force, such as unbalance, does not produce that sharp phase sweep at the peak. So an amplitude peak accompanied by the ninety-degree phase shift confirms a resonance, while a peak without it points to a forced response.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.