Library / Automotive and vehicle dynamics / VEH-65
Automotive and vehicle dynamics
Road elevation RMS over a spatial band
RMS elevation of a zero-mean road profile over the band n₁ to n₂ cycles per metre, for a one-sided power-law spectrum G(n) = G₀(n/n₀)⁻² with G₀ in m³.
- Ref
- VEH-65
- Rev
- v1
- Field
- Automotive and vehicle dynamics
- Result
- zRMS [m]
For a one-sided spatial PSD with a power-law shape G(n) = G₀(n/n₀)⁻² (waviness w = 2), the variance over the band n₁ to n₂ is ∫G dn = G₀n₀²(1/n₁ − 1/n₂), and the RMS elevation is its square root. The integral is dominated by the lowest spatial frequencies, so the long wavelengths carry most of the roughness.
RMS elevation of a zero-mean road profile over the band n₁ to n₂ cycles per metre, for a one-sided power-law spectrum G(n) = G₀(n/n₀)⁻² with G₀ in m³. Other spectral slopes need a different integral.
Calculator
- Variance of the elevation, z²4.608×10−5 m²
- Longest wavelength in the band, 1/n₁20 m
- Shortest wavelength in the band, 1/n₂2 m
What drives the result
Each variable is moved across its typical range while the others stay at the values above. The longest bar is the key driver.
- The road RMS elevation is 6.788 mm over wavelengths from 2 m to 20 m.
- The long-wavelength limit 1/n₁ contributes 111 % of the band integral, so extending the lower limit moves the answer more than the upper limit does.
Chart data as tables
Ranked variables
| Variable | Typical range | Lowest result | Highest result | Swing (m) | Local step | Change in result |
|---|---|---|---|---|---|---|
| n0 Reference spatial frequency | 0.01 to 1 cycles/m | 6.788e-4 | 0.06788 | 0.0672 | +10% | +10 % |
| G0 PSD at the reference frequency | 1e-9 to 0.01 m³ | 1.342e-5 | 0.04243 | 0.04241 | +10% | +4.88 % |
| n1 Lower band limit | 0.005 to 0.05 cycles/m | 0.006788 | 0.02251 | 0.01573 | +10% | −5.18 % |
| n2 Upper band limit | 0.5 to 5 cycles/m | 0.006788 | 0.00712 | 3.313e-4 | +10% | +0.504 % |
Sweep of Lower band limit
| Lower band limit (cycles/m) | Band-limited RMS elevation (m) |
|---|---|
| 0.005 | 0.02251 |
| 0.01063 | 0.01536 |
| 0.01625 | 0.01235 |
| 0.02188 | 0.01058 |
| 0.0275 | 0.009379 |
| 0.03313 | 0.008495 |
| 0.03875 | 0.007807 |
| 0.04438 | 0.007251 |
| 0.05 | 0.006788 |
Sweep one variable
Variables and how each one contributes
Band-limited RMS elevation
What the equation gives you. Everything else feeds it.
PSD at the reference frequency
The PSD at the reference spatial frequency, in m³. Raises the RMS as its square root. Raising it by 10 % (to 2.816×10−4) moves the result by +4.88 %, from 0.006788 to 0.00712 m.
Reference spatial frequency
The reference spatial frequency, in cycles/m. Raises the RMS in direct proportion, since G₀ is fixed at n₀. Raising it by 10 % (to 0.11) moves the result by +10 %, from 0.006788 to 0.007467 m.
Lower band limit
The lower band limit, in cycles/m. Lowering it adds long wavelengths and raises the RMS. Raising it by 10 % (to 0.055) moves the result by −5.18 %, from 0.006788 to 0.006436 m.
Upper band limit
The upper band limit, in cycles/m. Raising it adds short wavelengths, which adds little. Raising it by 10 % (to 0.55) moves the result by +0.504 %, from 0.006788 to 0.006822 m.
Worked example with the values above
- Band term: 1/n₁ − 1/n₂ = 20 − 2 = 18 m.
- Variance: G₀ n₀² × that = 2.56×10−4 × 0.01 × 18 = 4.608×10−5 m².
- RMS = √variance = 0.0067882 m.
Assumptions and limits
- Constant vehicle speed along a straight wheel path, and a stationary, zero-mean road profile.
- One-sided PSDs, with spatial frequency in cycles per metre and temporal frequency in hertz.
Common mistakes
- Mixing cycles per metre with radians per metre, which differ by 2π.
- Mixing one-sided and two-sided spectra.
- Applying a spectral relation outside the wavelength band over which the road was measured.
Statistics and uncertainty
Measured road PSDs scatter by a factor of two or more between road sections, and a power-law fit describes only a limited band. Treat spectral levels as indicative of a road class, not of a specific road.
Where it is used
- Converting measured road roughness into inputs for ride simulations.
- Predicting which road wavelengths excite body bounce, pitch and wheel hop.
Maths you need
- Wavelength and frequency: f = U/λ
- Power spectral density: variance per unit frequency
- Parseval’s theorem: the area under the PSD equals the variance