Choose your reading route
eduKateSG · Why Mathematics? · Random Vibration Reliability with Power Spectral Density: RMS, Spectral Moments, Rainflow and Fatigue
Start with the question, audit the worked example, or jump to the assumptions and practice. Every chapter stays open for uninterrupted reading.
Understand the question
Define the quantity and follow the mechanism.Route 2 · Chapters 6–10
Reproduce the example
Audit the numbers one step at a time.Route 3 · Chapters 11–15
Check assumptions
See where design and calibration matter.Route 4 · Chapters 16–20
Explore applications
Connect the method to a concrete investigation.Route 5 · Chapters 21–25
Study and verify
Practise, inspect failure cases and read further.
Full chapter index · Practice and checking · Mathematics Learning Hub
Why mathematics matters for machines and spacecraft becomes clear when loading is not one clean sine wave but a continuously changing vibration. A power spectral density, or PSD, describes how mean-square response is distributed across frequency. Spectral moments turn that curve into RMS response and crossing-rate information, while rainflow counting and spectral fatigue models connect the random response to cycles and cumulative damage.
This guide develops PSD units, integration, spectral moments, Gaussian and stationarity assumptions, narrow-band reasoning, Dirlik-style spectral estimates and validation against time histories. It keeps random-vibration reliability distinct from a test level or a single fatigue number: reliability also depends on uncertain environments, structural dynamics, material data, duration, workmanship and the decision rule.
Full chapter index
Chapters 1–5 · Understand the question
Chapters 6–10 · Reproduce the example
Chapters 11–15 · Check assumptions
Chapters 16–20 · Explore applications
Chapters 21–25 · Study and verify
Random vibration spreads energy across frequencies and changes unpredictably in time. A launch component can experience broadband acceleration rather than a single tone. This is the central mathematical mechanism in why random vibration is different. Write the quantities and assumptions before using software, then reproduce the calculation or geometric argument in a small example. Random does not mean structureless or identical across missions. Define duration, axes and operating condition. The useful habit is to keep the physical question, the probability model and the numerical method in separate columns so that a correct computation is not mistaken for complete evidence.
Did You Know? A result can be numerically stable and still answer the wrong question if the failure event, denominator or transformation changes. Test one nearby case and predict the direction of change before recalculating. Save inputs, intermediate values and units so a second reader can audit the path.
For practice, turn the example into a sensitivity check. Change one defensible input while holding the others fixed, compare the new result with the original and explain why it moved. Then construct a counterexample in which the shortcut fails. End with two sentences: what the mathematics supports, and what remains uncertain.
Start with the operational question behind time history and frequency view. The same stationary process can be studied through response versus time or statistical energy versus frequency. In the illustrative case, A measured accelerometer trace can be transformed into a PSD estimate. Draw the relevant graph, density, boundary or timeline before calculating; the picture should use the same variables and sign convention as the equations. Finite records and windowing affect the estimate. Preserve sample rate and processing settings.
A student should be able to rebuild the result without hidden settings. Record the sampling rule or optimisation criterion, the exact failure definition, and the evidence used for every distribution or parameter. Compare a transparent baseline with the advanced method. If they disagree, investigate the mechanism before deciding which answer is more credible.
This chapter belongs to a larger decision, so translate the number carefully. State one action the evidence permits and one action it does not. A parent or educator can ask which school ideas are doing the work: algebra organises relationships, probability represents uncertainty, calculus describes local change, and computing repeats the documented rules at scale.
Treat power spectral density as a model-checking problem rather than a vocabulary exercise. A one-sided acceleration PSD S_a(f) has units such as g squared per hertz and shows mean-square contribution per frequency. A peak near a resonance indicates strong response in that band. Recompute the example from the raw values and label every unit. PSD amplitude is not ordinary acceleration amplitude. Always print units on both axes.
The strongest check uses evidence the model did not choose for itself. Hold out later observations, an independent simulation set or a known-answer benchmark. Report errors where decisions are sensitive, especially near the failure boundary, instead of relying on one global accuracy score. A method that performs well on easy safe cases can still fail at the rare cases that motivated the analysis.
Now change one assumption and make a directional prediction before running the calculation. If the result moves the opposite way, inspect the sign convention, transformation, density ratio or constraint. This prediction-first routine develops transferable problem-solving skills because it connects symbolic work to a testable expectation.
The mathematics of rms from area becomes useful only when its mechanism is explicit. For a one-sided PSD, variance is the integral of S(f) over frequency and RMS is its square root for zero-mean response. Consider the worked illustration: A flat 0.04 g²/Hz over 20 Hz has variance 0.8 g² and RMS about 0.894 g. Preserve the original inputs, calculate step by step and compare with a second method whenever an analytic or high-budget reference exists. Real spectra are rarely perfectly flat and may include mean offsets separately. Integrate the actual frequency bins.
There are three uncertainties to keep apart. Sampling or numerical uncertainty comes from finite computation. Parameter uncertainty comes from estimating inputs. Model-form uncertainty comes from simplified physics, distributions or dependence. Reducing the first does not automatically reduce the other two. A credible report names all three and shows which one drives the decision.
For a practical investigation, create a small reproducible notebook or spreadsheet with a frozen data sheet, formula sheet, diagnostics sheet and decision note. Another student should be able to change one input and obtain the same update without guessing what the model meant.
Primary source: https://ntrs.nasa.gov/citations/20150009467. Read it for the method and its assumptions; do not transfer its performance claims automatically to another model.
Use decibels and linear units to practise sensitivity analysis. A 3 dB change in a power-like PSD corresponds approximately to a factor of two in PSD level. Doubling a flat PSD doubles variance and multiplies RMS by sqrt(2). Before accepting the result, identify the quantity that would change the conclusion most. Confusing amplitude dB and power dB produces wrong scaling. Convert to linear units before integration.
A simple baseline is essential. It reveals whether the advanced method is improving accuracy or only adding complexity. Compare model calls, wall time, uncertainty and failure-mode coverage under one declared budget. Repeat stochastic calculations with several seeds; a favourable single run is not evidence of dependable efficiency.
Students can extend the example by adding correlation, a second failure mode or a modest measurement error. Explain which part of the formula changes and which remains fixed. This exercise shows how mathematics transfers from a clean textbook case to engineering and computing without pretending that the classroom model is a certified design tool.
Build a small counterexample for structural transfer functions. Input PSD becomes response PSD through the frequency-response magnitude squared in a linear system. The initial example is: S_y(f)=|H(f)|²S_x(f) for a single input-output relation. Then create a nearby case where the usual shortcut becomes unreliable. Nonlinearity and multiple correlated inputs need richer analysis. Validate H with test or model evidence.
Counterexamples are positive learning tools. They reveal the conditions that make a method work and help students recognise when a software output deserves a pause. Keep the counterexample small enough to plot or enumerate. Check sign, support, bounds and conservation of probability before adding complexity.
When writing the conclusion, separate a documented fact from an interpretation and from an illustrative scenario. Cite the primary method paper for the factual mechanism, use the local calculation for the illustration, and label any recommendation as conditional on the stated model. This keeps career and engineering advice age-appropriate and honest.
Connect resonance to a real decision while protecting the boundary between calculation and judgement. A mode can amplify response near its natural frequency depending on damping and load path. A narrow input band near a lightly damped mode can dominate stress RMS. First verify the arithmetic, then translate it into an allowed action and an action the evidence does not support. Natural frequencies shift with boundary conditions and temperature. Measure sensitivity to modal uncertainty.
The audit should include the original question, input provenance, failure convention, algorithm settings, convergence or stopping rule, and a sensitivity result. For an adaptive method, also preserve the sequence of choices. For an optimisation, preserve every constraint. For Bayesian updating, preserve the prior and observation model. These records turn a result into something another person can challenge and improve.
A final transfer question is useful: where else would the same mathematical idea appear? Density ratios appear in survey correction and machine learning, probabilistic constraints in planning, and conditional probability in medical testing. The context changes, but transparent assumptions and reproducible checks remain valuable.
The nth spectral moment is m_n=integral f^n S(f) df under a declared frequency convention. m0 is variance; ratios of moments describe characteristic rates and bandwidth. This is the central mathematical mechanism in spectral moments. Write the quantities and assumptions before using software, then reproduce the calculation or geometric argument in a small example. Using angular frequency instead of hertz changes factors of 2pi. State the convention before calculating. The useful habit is to keep the physical question, the probability model and the numerical method in separate columns so that a correct computation is not mistaken for complete evidence.
Did You Know? A result can be numerically stable and still answer the wrong question if the failure event, denominator or transformation changes. Test one nearby case and predict the direction of change before recalculating. Save inputs, intermediate values and units so a second reader can audit the path.
For practice, turn the example into a sensitivity check. Change one defensible input while holding the others fixed, compare the new result with the original and explain why it moved. Then construct a counterexample in which the shortcut fails. End with two sentences: what the mathematics supports, and what remains uncertain.
Start with the operational question behind zero-crossing rate. For a suitable stationary Gaussian process, moment ratios can estimate mean zero-crossing rates. In the illustrative case, A common hertz-form relation uses sqrt(m2/m0) with the consistent convention. Draw the relevant graph, density, boundary or timeline before calculating; the picture should use the same variables and sign convention as the equations. Formula variants differ between f and omega. Check dimensions and a known spectrum.
A student should be able to rebuild the result without hidden settings. Record the sampling rule or optimisation criterion, the exact failure definition, and the evidence used for every distribution or parameter. Compare a transparent baseline with the advanced method. If they disagree, investigate the mechanism before deciding which answer is more credible.
This chapter belongs to a larger decision, so translate the number carefully. State one action the evidence permits and one action it does not. A parent or educator can ask which school ideas are doing the work: algebra organises relationships, probability represents uncertainty, calculus describes local change, and computing repeats the documented rules at scale.
Treat peak rates and bandwidth as a model-checking problem rather than a vocabulary exercise. Higher moments help describe how often peaks occur and whether a process is narrow- or broad-band. Two PSDs with equal RMS can have different peak statistics. Recompute the example from the raw values and label every unit. One bandwidth scalar cannot describe every spectral shape. Compare the full curves.
The strongest check uses evidence the model did not choose for itself. Hold out later observations, an independent simulation set or a known-answer benchmark. Report errors where decisions are sensitive, especially near the failure boundary, instead of relying on one global accuracy score. A method that performs well on easy safe cases can still fail at the rare cases that motivated the analysis.
Now change one assumption and make a directional prediction before running the calculation. If the result moves the opposite way, inspect the sign convention, transformation, density ratio or constraint. This prediction-first routine develops transferable problem-solving skills because it connects symbolic work to a testable expectation.
The mathematics of gaussian assumption becomes useful only when its mechanism is explicit. Many spectral fatigue formulas assume a Gaussian response so amplitude statistics follow from second-order information. Consider the worked illustration: A linear system under many small random inputs may be approximately Gaussian. Preserve the original inputs, calculate step by step and compare with a second method whenever an analytic or high-budget reference exists. Impacts, clipping and intermittent events can create non-Gaussian tails. Inspect histograms and kurtosis.
There are three uncertainties to keep apart. Sampling or numerical uncertainty comes from finite computation. Parameter uncertainty comes from estimating inputs. Model-form uncertainty comes from simplified physics, distributions or dependence. Reducing the first does not automatically reduce the other two. A credible report names all three and shows which one drives the decision.
For a practical investigation, create a small reproducible notebook or spreadsheet with a frozen data sheet, formula sheet, diagnostics sheet and decision note. Another student should be able to change one input and obtain the same update without guessing what the model meant.
Use stationarity to practise sensitivity analysis. Stationarity means the relevant statistics do not change over the analysed segment. A one-minute steady test may be treated as locally stationary. Before accepting the result, identify the quantity that would change the conclusion most. Launch phases and machine start-up can be strongly nonstationary. Segment by physical regime.
A simple baseline is essential. It reveals whether the advanced method is improving accuracy or only adding complexity. Compare model calls, wall time, uncertainty and failure-mode coverage under one declared budget. Repeat stochastic calculations with several seeds; a favourable single run is not evidence of dependable efficiency.
Students can extend the example by adding correlation, a second failure mode or a modest measurement error. Explain which part of the formula changes and which remains fixed. This exercise shows how mathematics transfers from a clean textbook case to engineering and computing without pretending that the classroom model is a certified design tool.
A second primary source is https://standards.nasa.gov/standard/MSFC/MSFC-STD-3676. Compare the problem definition, estimator and validation rather than quoting one result in isolation.
Build a small counterexample for narrow-band approximation. For a narrow-band Gaussian stress process, peak ranges have a relatively simple distribution. The initial example is: Rayleigh-style reasoning can approximate cycles near one dominant mode. Then create a nearby case where the usual shortcut becomes unreliable. Broad-band processes contain interacting time scales and need corrections. Measure bandwidth before applying it.
Counterexamples are positive learning tools. They reveal the conditions that make a method work and help students recognise when a software output deserves a pause. Keep the counterexample small enough to plot or enumerate. Check sign, support, bounds and conservation of probability before adding complexity.
When writing the conclusion, separate a documented fact from an interpretation and from an illustrative scenario. Cite the primary method paper for the factual mechanism, use the local calculation for the illustration, and label any recommendation as conditional on the stated model. This keeps career and engineering advice age-appropriate and honest.
Connect rainflow counting to a real decision while protecting the boundary between calculation and judgement. Rainflow reduces a stress time history to counted ranges for use with an S-N curve and damage rule. A simulated response record provides an independent time-domain benchmark. First verify the arithmetic, then translate it into an allowed action and an action the evidence does not support. Different filtering or record length changes counted extremes. Document preprocessing and repeat records.
The audit should include the original question, input provenance, failure convention, algorithm settings, convergence or stopping rule, and a sensitivity result. For an adaptive method, also preserve the sequence of choices. For an optimisation, preserve every constraint. For Bayesian updating, preserve the prior and observation model. These records turn a result into something another person can challenge and improve.
A final transfer question is useful: where else would the same mathematical idea appear? Density ratios appear in survey correction and machine learning, probabilistic constraints in planning, and conditional probability in medical testing. The context changes, but transparent assumptions and reproducible checks remain valuable.
A common illustration sums n_i/N_i across stress-range bins. If 100 cycles occur where the S-N life is 10,000, that bin contributes 0.01 damage. This is the central mathematical mechanism in miner cumulative damage. Write the quantities and assumptions before using software, then reproduce the calculation or geometric argument in a small example. Linear damage ignores load-sequence and interaction effects. Treat the rule as a model, not a law. The useful habit is to keep the physical question, the probability model and the numerical method in separate columns so that a correct computation is not mistaken for complete evidence.
Did You Know? A result can be numerically stable and still answer the wrong question if the failure event, denominator or transformation changes. Test one nearby case and predict the direction of change before recalculating. Save inputs, intermediate values and units so a second reader can audit the path.
For practice, turn the example into a sensitivity check. Change one defensible input while holding the others fixed, compare the new result with the original and explain why it moved. Then construct a counterexample in which the shortcut fails. End with two sentences: what the mathematics supports, and what remains uncertain.
Start with the operational question behind spectral fatigue estimates. Dirlik and related methods estimate rainflow-range distributions from PSD moments without simulating every time history. In the illustrative case, The estimated range density is integrated with an S-N relation to obtain damage rate. Draw the relevant graph, density, boundary or timeline before calculating; the picture should use the same variables and sign convention as the equations. Agreement depends on spectral shape, stationarity, Gaussianity and material model. Benchmark against rainflow simulations.
A student should be able to rebuild the result without hidden settings. Record the sampling rule or optimisation criterion, the exact failure definition, and the evidence used for every distribution or parameter. Compare a transparent baseline with the advanced method. If they disagree, investigate the mechanism before deciding which answer is more credible.
This chapter belongs to a larger decision, so translate the number carefully. State one action the evidence permits and one action it does not. A parent or educator can ask which school ideas are doing the work: algebra organises relationships, probability represents uncertainty, calculus describes local change, and computing repeats the documented rules at scale.
Treat duration and expected damage as a model-checking problem rather than a vocabulary exercise. For a stationary model, expected cycle counts and damage scale with exposure duration. Doubling duration doubles expected Miner damage under unchanged statistics. Recompute the example from the raw values and label every unit. Environment changes and fatigue scatter break simple extrapolation. Keep mission phases separate.
The strongest check uses evidence the model did not choose for itself. Hold out later observations, an independent simulation set or a known-answer benchmark. Report errors where decisions are sensitive, especially near the failure boundary, instead of relying on one global accuracy score. A method that performs well on easy safe cases can still fail at the rare cases that motivated the analysis.
Now change one assumption and make a directional prediction before running the calculation. If the result moves the opposite way, inspect the sign convention, transformation, density ratio or constraint. This prediction-first routine develops transferable problem-solving skills because it connects symbolic work to a testable expectation.
The mathematics of uncertain environments becomes useful only when its mechanism is explicit. PSD levels, modal frequencies, damping and durations vary across units and missions. Consider the worked illustration: Monte Carlo can sample uncertain model parameters and compute a damage distribution. Preserve the original inputs, calculate step by step and compare with a second method whenever an analytic or high-budget reference exists. Sampling a PSD level independently at every bin can create unrealistic spectra. Use physically structured uncertainty.
There are three uncertainties to keep apart. Sampling or numerical uncertainty comes from finite computation. Parameter uncertainty comes from estimating inputs. Model-form uncertainty comes from simplified physics, distributions or dependence. Reducing the first does not automatically reduce the other two. A credible report names all three and shows which one drives the decision.
For a practical investigation, create a small reproducible notebook or spreadsheet with a frozen data sheet, formula sheet, diagnostics sheet and decision note. Another student should be able to change one input and obtain the same update without guessing what the model meant.
Use reliability from a limit state to practise sensitivity analysis. Define failure through a declared limit such as damage capacity, stress exceedance or performance loss. A simple limit state is g=1-D where D is modelled Miner damage. Before accepting the result, identify the quantity that would change the conclusion most. D=1 is an empirical convention and not universal physical failure. Validate the limit against evidence.
A simple baseline is essential. It reveals whether the advanced method is improving accuracy or only adding complexity. Compare model calls, wall time, uncertainty and failure-mode coverage under one declared budget. Repeat stochastic calculations with several seeds; a favourable single run is not evidence of dependable efficiency.
Students can extend the example by adding correlation, a second failure mode or a modest measurement error. Explain which part of the formula changes and which remains fixed. This exercise shows how mathematics transfers from a clean textbook case to engineering and computing without pretending that the classroom model is a certified design tool.
Build a small counterexample for practice activity. Integrate a piecewise-constant PSD, calculate RMS, then pass it through a simple resonant gain and compare. The initial example is: Predict which frequency band contributes most to output variance. Then create a nearby case where the usual shortcut becomes unreliable. A classroom transfer function is not qualification evidence. Show every bin width and unit.
Counterexamples are positive learning tools. They reveal the conditions that make a method work and help students recognise when a software output deserves a pause. Keep the counterexample small enough to plot or enumerate. Check sign, support, bounds and conservation of probability before adding complexity.
When writing the conclusion, separate a documented fact from an interpretation and from an illustrative scenario. Cite the primary method paper for the factual mechanism, use the local calculation for the illustration, and label any recommendation as conditional on the stated model. This keeps career and engineering advice age-appropriate and honest.
Connect parent and educator guidance to a real decision while protecting the boundary between calculation and judgement. Random vibration links area under curves, square roots, frequency, probability and engineering tests. Ask why equal RMS does not imply equal peak or fatigue behaviour. First verify the arithmetic, then translate it into an allowed action and an action the evidence does not support. Do not let a colourful spectrum replace unit checking. Start with two rectangular frequency bands.
The audit should include the original question, input provenance, failure convention, algorithm settings, convergence or stopping rule, and a sensitivity result. For an adaptive method, also preserve the sequence of choices. For an optimisation, preserve every constraint. For Bayesian updating, preserve the prior and observation model. These records turn a result into something another person can challenge and improve.
A final transfer question is useful: where else would the same mathematical idea appear? Density ratios appear in survey correction and machine learning, probabilistic constraints in planning, and conditional probability in medical testing. The context changes, but transparent assumptions and reproducible checks remain valuable.
PSD is not a time history, RMS is not a maximum, and equal total energy does not guarantee equal fatigue. A 3 dB PSD increase doubles power-like level, not RMS. This is the central mathematical mechanism in common misconceptions. Write the quantities and assumptions before using software, then reproduce the calculation or geometric argument in a small example. Spectral fatigue is not exact for every nonstationary process. Name the assumptions beside the result. The useful habit is to keep the physical question, the probability model and the numerical method in separate columns so that a correct computation is not mistaken for complete evidence.
Did You Know? A result can be numerically stable and still answer the wrong question if the failure event, denominator or transformation changes. Test one nearby case and predict the direction of change before recalculating. Save inputs, intermediate values and units so a second reader can audit the path.
For practice, turn the example into a sensitivity check. Change one defensible input while holding the others fixed, compare the new result with the original and explain why it moved. Then construct a counterexample in which the shortcut fails. End with two sentences: what the mathematics supports, and what remains uncertain.
Start with the operational question behind when spectral methods are inadequate. Strong nonlinearity, shocks, nonstationarity, multiaxial stress and non-Gaussian response may need time-domain or specialised analysis. In the illustrative case, An intermittent impact can dominate damage while contributing little to a smoothed PSD. Draw the relevant graph, density, boundary or timeline before calculating; the picture should use the same variables and sign convention as the equations. Averaging can erase dangerous transients. Retain event-based evidence.
A student should be able to rebuild the result without hidden settings. Record the sampling rule or optimisation criterion, the exact failure definition, and the evidence used for every distribution or parameter. Compare a transparent baseline with the advanced method. If they disagree, investigate the mechanism before deciding which answer is more credible.
This chapter belongs to a larger decision, so translate the number carefully. State one action the evidence permits and one action it does not. A parent or educator can ask which school ideas are doing the work: algebra organises relationships, probability represents uncertainty, calculus describes local change, and computing repeats the documented rules at scale.
Treat research and next reading as a model-checking problem rather than a vocabulary exercise. NASA’s review compares spectral fatigue methods with time-domain rainflow evidence, while MSFC-STD-3676 documents active methodology for vibroacoustic and shock criteria. Read method limitations beside the test-criteria context. Recompute the example from the raw values and label every unit. Neither source converts a generic PSD into universal reliability. Reproduce a published spectrum benchmark.
The strongest check uses evidence the model did not choose for itself. Hold out later observations, an independent simulation set or a known-answer benchmark. Report errors where decisions are sensitive, especially near the failure boundary, instead of relying on one global accuracy score. A method that performs well on easy safe cases can still fail at the rare cases that motivated the analysis.
Now change one assumption and make a directional prediction before running the calculation. If the result moves the opposite way, inspect the sign convention, transformation, density ratio or constraint. This prediction-first routine develops transferable problem-solving skills because it connects symbolic work to a testable expectation.
Continue with the Mathematics Learning Hub and the previous reliability article. These links own broader probability ideas and the immediately preceding rare-event geometry.
The mathematics of frequently asked questions becomes useful only when its mechanism is explicit. FAQs should cover PSD units, RMS, dB, transfer functions, moments, Gaussianity, rainflow, Dirlik and duration. Consider the worked illustration: Each answer should preserve the frequency convention. Preserve the original inputs, calculate step by step and compare with a second method whenever an analytic or high-budget reference exists. Avoid reporting an expected damage estimate as guaranteed life. Publish spectra, settings and validation.
There are three uncertainties to keep apart. Sampling or numerical uncertainty comes from finite computation. Parameter uncertainty comes from estimating inputs. Model-form uncertainty comes from simplified physics, distributions or dependence. Reducing the first does not automatically reduce the other two. A credible report names all three and shows which one drives the decision.
For a practical investigation, create a small reproducible notebook or spreadsheet with a frozen data sheet, formula sheet, diagnostics sheet and decision note. Another student should be able to change one input and obtain the same update without guessing what the model meant.
Quick answers
What should I calculate first? Start with the estimand or failure event, then a simple baseline.
Does the advanced method guarantee accuracy? No. It needs input, numerical and model checks.
Which school mathematics matters? Algebra, graphs, probability, functions and careful units all contribute.
How can a student practise safely? Use fictional or open benchmark data and preserve every intermediate value.
What makes a career claim responsible? Describe how the mathematics is used without promising admission, salary or employment.
Extended practice laboratory
Practice case 1: Why random vibration is different
Setup: Use this chapter’s illustrative case: A launch component can experience broadband acceleration rather than a single tone.
Task: Define duration, axes and operating condition.
Checking target: Random does not mean structureless or identical across missions.
Practice case 2: Time history and frequency view
Setup: Use this chapter’s illustrative case: A measured accelerometer trace can be transformed into a PSD estimate.
Task: Preserve sample rate and processing settings.
Checking target: Finite records and windowing affect the estimate.
Practice case 3: Power spectral density
Setup: Use this chapter’s illustrative case: A peak near a resonance indicates strong response in that band.
Task: Always print units on both axes.
Checking target: PSD amplitude is not ordinary acceleration amplitude.
Practice case 4: RMS from area
Setup: Use this chapter’s illustrative case: A flat 0.04 g²/Hz over 20 Hz has variance 0.8 g² and RMS about 0.894 g.
Task: Integrate the actual frequency bins.
Checking target: Real spectra are rarely perfectly flat and may include mean offsets separately.
Practice case 5: Decibels and linear units
Setup: Use this chapter’s illustrative case: Doubling a flat PSD doubles variance and multiplies RMS by sqrt(2).
Task: Convert to linear units before integration.
Checking target: Confusing amplitude dB and power dB produces wrong scaling.
Practice case 6: Structural transfer functions
Setup: Use this chapter’s illustrative case: S_y(f)=|H(f)|²S_x(f) for a single input-output relation.
Task: Validate H with test or model evidence.
Checking target: Nonlinearity and multiple correlated inputs need richer analysis.
Practice case 7: Resonance
Setup: Use this chapter’s illustrative case: A narrow input band near a lightly damped mode can dominate stress RMS.
Task: Measure sensitivity to modal uncertainty.
Checking target: Natural frequencies shift with boundary conditions and temperature.
Practice case 8: Spectral moments
Setup: Use this chapter’s illustrative case: m0 is variance; ratios of moments describe characteristic rates and bandwidth.
Task: State the convention before calculating.
Checking target: Using angular frequency instead of hertz changes factors of 2pi.
Practice case 9: Zero-crossing rate
Setup: Use this chapter’s illustrative case: A common hertz-form relation uses sqrt(m2/m0) with the consistent convention.
Task: Check dimensions and a known spectrum.
Checking target: Formula variants differ between f and omega.
Practice case 10: Peak rates and bandwidth
Setup: Use this chapter’s illustrative case: Two PSDs with equal RMS can have different peak statistics.
Task: Compare the full curves.
Checking target: One bandwidth scalar cannot describe every spectral shape.