Calculating credible speeds: formulas, sources, sensitivity.
Turn run-up into defendable speeds. Combine observed timings with calculated models, GIS measurements, and adjustments for surface and gradient. Use ± sensitivity bands to show uncertainty and validate with 922. Document inputs so reviewers can reproduce results. Final speeds drive HVM bollard rating and crash rated bollard equivalence checks (414) and appear in the VDA report (229). Include one-sentence context that naturally links upward to the parent hubs (this section and the chapter hub). Add SIRA context with a link to SIRA Bollards (UAE) when relevant. Link installation pages only if helpful: What to Expect and Installation Guide.
Important: This is a general guide. For live projects we develop a tailored
Method Statement & Risk Assessment (MS/RA) and align with authority approvals (e.g., SIRA) where in scope.
224.1 Observed speed vs calculated
Use both: measured timings validate model outputs. Agreement supports HVM bollard selection and the crash rated bollard rating pick (413).
Start with direct observation: measure run times between two markers and convert to speed, then compare with a calculated estimate from an effective run-up model. When observed and calculated speeds converge within your acceptance band, the VDA position is stronger for reviewers.
Where the gap is large, revisit inputs: path length (222), surface condition, turning radius, and vehicle class assumptions (223). Repeating observations over several passes improves the signal and reveals driver behavior. Carry the reconciled value into rating selection (413) and equivalency checks (414).
| Aspect | What matters | Where to verify |
|---|---|---|
| Performance | Tested system (bollard + footing) | How to read ratings (413) |
| Evidence | Observed timings & model agreement | Evidence & Documentation (444) |
224.2 Distance-time sampling
Record times across known spans to infer speed; repeat to reduce noise. Results tune HVM bollard spacing (232) and crash rated bollard energy mapping (414).
Pick two clear waypoints with a known separation (e.g., 30–80 m measured from drawings or a centreline trace). Time multiple runs and compute speed = distance ÷ time. Note peak/off-peak timing to show variability.
Use a handheld video timestamp or smartphone stopwatch (noting device accuracy), and log weather and traffic. Feed the mean/95th percentile into spacing rules (232) and equivalency (414). Outliers usually indicate driver hesitancy, turning, or braking—document and exclude with justification.
224.3 Google Earth/GIS aids
Trace credible paths and measure centreline distances (214). GIS anchors HVM bollard assumptions and documents crash rated bollard evidence (716).
Draw a polyline along the likely run-up corridor and export the length. Check street imagery for traffic calming, bends, and turning radius changes. Save annotated screenshots and the KML/KMZ for your submission pack (716, 717, 938).
Correlate GIS lengths to on-site tape/laser checks (211) to avoid plan distortions. Where multiple approach paths exist, prepare a short matrix of path length, bends, and surface, then carry forward the credible worst case.
224.4 Acceleration models (simple)
Apply constant-accel or segment models with conservative coefficients (222, 227). These define HVM bollard tier and crash rated bollard class.
(a) Constant-acceleration: assume a conservative longitudinal acceleration (e.g., 0.8–1.5 m/s² for light vehicles on urban surfaces) over the effective run-up. (b) Segment model: split the path into straight/curve/grade segments with different acceleration values.
Reference site factors from 227 and sanity-check against 922 VDA Approach-Speed Helper. Always round conservatively when selecting rating classes (413, 414).
224.5 Adjusting for surface/gradient
Modify speeds for roughness, slopes, and bends (222.5–222.6). This prevents over/under-design of HVM bollard arrays and wrong crash rated bollard bands.
Apply reductions for adverse gradient, tight curvature, poor surfaces, and obstacles. Conversely, long straight dry asphalt with good sightlines merits less reduction. Capture rationale and photos (716). If approvals in the UAE are required, note any local constraints and reference SIRA Bollards (UAE) in your assumptions register.
When uncertain, publish a banded result (e.g., 42–48 km/h) and carry the higher edge into design selection (432–435).
224.6 Sensitivity bands (±)
Publish ± ranges for inputs and outputs (228). Bands inform HVM bollard count buffers and crash rated bollard acceptance margins.
Vary key inputs—run-up length, acceleration, surface factor—by plausible ± deltas. Present a simple table: input low/nominal/high and resulting speed. Use these bands to justify buffers (e.g., choose the next tier up if the credible worst case straddles a boundary). Keep the assumptions register traceable (229).
224.7 Cross-checks & validation
Compare different paths and peak/off-peak timing. Convergence strengthens HVM bollard credibility and crash rated bollard submissions (444).
Cross-check three ways: (1) observed vs model, (2) alternative approach paths, (3) peak vs off-peak. If two of the three agree, treat that as your primary result and flag the outlier with reasons (e.g., braking, temporary cones). Validation notes belong with the evidence set (716, 717, 938).
224.8 Recording inputs
Store dates, devices, weather, and traffic conditions (716). Solid records speed HVM bollard approvals and crash rated bollard reviews (717, 938).
Log device type/accuracy, waypoint photos, measured spans, traffic levels, and any enforcement presence. Keep files named per 911 File Index & Naming Rules. For Dubai projects, note if timings were taken under typical SIRA operating conditions and attach a brief authority submittal note.
224.9 Worked example
Show one path: measured run-up, model, adjustments, resulting speed. Link to chosen HVM bollard pattern (321) and the candidate crash rated bollard rating string (413).
Scenario: 85 m straight approach on dry asphalt to a frontage array. Observations: 85 m in 5.8–6.2 s (mean 6.0 s ⇒ 51 km/h). Model: segment-accel with 1.2 m/s² straight, −10% for a gentle chicane near the end ⇒ 49–53 km/h. Adopt 52 km/h (high edge).
Effect on design: Pick a crash-rated class whose rating string covers the mass and speed; tune the array pattern (321) and clear-gap calculations (322) accordingly. Record sensitivities (±3 km/h) and include the GIS trace plus timing sheet in the VDA report (229).
Related
External resources
224 Speed Estimation Methods — FAQ
What’s the quickest credible way to estimate approach speed?
Combine a timed distance run (video or stopwatch) with a simple acceleration model over the measured run-up. If both converge within your acceptance band, you have a defensible speed for VDA.
How many timing samples do I need?
As a rule of thumb, take at least 5–7 runs, across peak and off-peak, and report the mean and 95th percentile. Exclude outliers only with documented reasons.
Do slopes and bends really change the result?
Yes. Adverse gradients, tight bends, rough surfaces, and sightline issues reduce achievable speed. Document the adjustments with photos and a short rationale in your VDA report.
How do I show uncertainty without weakening my case?
Publish a sensitivity band (±) for key inputs and carry the higher credible edge into selection. This is transparent and typically preferred by reviewers and authorities.