RMS Structural Response

What Is It?

The RMS structural response is the single-number statistical measure of the random vibration response. It is the square root of the area under the response PSD. For a zero-mean Gaussian process, the RMS equals the standard deviation — the characteristic amplitude of the response. RMS values can be computed for acceleration, displacement, stress or any other response quantity.

Why It Matters

The RMS is the primary output of random vibration analysis — the value used for screening, comparison, qualification and as input to further assessment. Understanding what RMS means statistically — and what it does not mean — is essential for using it correctly in engineering decisions. The RMS is not a peak, not a maximum, and not a deterministic value. It is a statistical level.

RMS is the primary output of random vibration analysis. It is a statistical level — the standard deviation of the response — not a peak or maximum. Understanding this distinction is essential.

RMS Acceleration

The RMS acceleration is the square root of the area under the acceleration response PSD. It represents the characteristic acceleration level at a location. For equipment qualification, the RMS acceleration at the equipment mounting location is compared against the qualification level. The RMS acceleration is also used for comparing the severity of different environments and for screening structural response.

RMS acceleration:

a_rms  =  √( ∫ G_a(f) df )    [g]

where:
G_a(f)  =  acceleration response PSD  [g²/Hz]

a_rms  =  standard deviation of acceleration
       ≠  peak acceleration
       ≠  maximum acceleration over a duration

RMS Displacement

The RMS displacement is the square root of the area under the displacement response PSD. It represents the characteristic displacement amplitude at a location. For clearance assessment, the RMS displacement is used with a peak factor (typically 3σ) to estimate the maximum likely displacement and compare against available clearance. The displacement RMS is typically dominated by low-frequency modes.

RMS displacement:

d_rms  =  √( ∫ G_d(f) df )    [mm]

where:
G_d(f)  =  displacement response PSD  [mm²/Hz]

d_rms  =  standard deviation of displacement

For clearance:  d_peak ≈ 3σ  =  3 · d_rms  (screening)

RMS Stress

The RMS stress is the square root of the area under the stress response PSD. It represents the characteristic stress amplitude at a location. RMS stress is a statistical quantity — the standard deviation of the stress process. It is not a deterministic peak stress. RMS stress is used for screening (3σ screening against allowables) and as an input to vibration fatigue analysis. It should not be directly compared against a static or fatigue allowable without statistical interpretation.

RMS stress:

σ_rms  =  √( ∫ G_σ(f) df )    [MPa]

where:
G_σ(f)  =  stress response PSD  [MPa²/Hz]

σ_rms  =  standard deviation of stress
       ≠  peak stress
       ≠  deterministic stress amplitude

For screening:  σ_peak ≈ 3σ  =  3 · σ_rms  (3σ screening)

RMS stress is a standard deviation, not a peak. Comparing it directly against a static or fatigue allowable without statistical interpretation (e.g. 3σ screening) is incorrect.

Standard Deviation and Statistical Meaning

For a zero-mean Gaussian process, the RMS equals the standard deviation σ. The standard deviation describes the spread of the instantaneous values. About 68% of the time, the instantaneous value is within ±1σ. About 95% within ±2σ. About 99.7% within ±3σ. The RMS is therefore a statistical level, not a deterministic value. The actual peak over a duration will exceed the RMS — and will exceed 3σ for sufficiently long durations.

RMS as a Screening Tool

The RMS response is commonly used with a sigma multiplier for screening. The 3σ level (3 × RMS) is the most common screening value — it provides a conservative estimate of the peak response for short to moderate durations. For clearance, 3σ displacement is compared against available clearance. For strength, 3σ stress is compared against yield or ultimate strength. This screening is conservative for short durations but not a maximum.

3σ (3 × RMS) is a commonly used screening level. It is conservative for short durations but is not a hard maximum — over long durations, the response will exceed 3σ. State the statistical basis when reporting.

RMS Is Not the Maximum

A critical distinction: the RMS is not the maximum response. Over a sufficiently long duration, the response will exceed any fixed multiple of RMS. The expected peak over N cycles grows as approximately √(2·ln(N))·σ. For a 1-hour test at 100 Hz (360,000 cycles), the expected peak is about 4.4σ. For a 1-second test at 100 Hz (100 cycles), it is about 3.0σ. The RMS — and any fixed sigma level — is a statistical characterisation, not a deterministic maximum.

Expected peak (narrowband, Gaussian, N cycles):

E[peak]  ≈  σ · √( 2 · ln(N) )

Examples:
  N = 100 cycles:     E[peak] ≈ 3.0σ
  N = 10,000 cycles:   E[peak] ≈ 3.7σ
  N = 360,000 cycles:  E[peak] ≈ 4.4σ

The expected peak grows with duration —
no fixed sigma level is a true maximum.

Using RMS in Engineering Assessment

The RMS response is used in several ways in engineering assessment. Each use requires understanding the statistical nature of the value.

• Qualification comparison — RMS acceleration at equipment mounting compared against test levels
• Clearance screening — 3σ displacement compared against available clearance
• Strength screening — 3σ stress compared against yield or ultimate strength
• Fatigue input — stress PSD (not just RMS) feeds vibration fatigue analysis
• Model verification — RMS values checked for plausibility against input Grms and amplification

Key Takeaways

• RMS = √(area under response PSD) = standard deviation for zero-mean Gaussian
• RMS is a statistical level, not a peak or maximum — it will be exceeded
• 3σ (3 × RMS) is a common screening level — conservative but not a hard maximum
• RMS stress should not be directly compared against allowables without statistical interpretation
• The expected peak grows with duration — no fixed sigma level is a true maximum