Modeling Operating Speed: Synthesis Report. Chapter 2: Speed Models in North America

Dimaiuta, Michael; Donnell, Eric T.; Himes, Scott; Porter, Richard J. · 2011 · Transportation research circular

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This synthesis report, published by the Transportation Research Board in 2011, addresses the critical role of operating speed in highway design and safety. The authors argue that while traditional design guides rely on a constant "design speed" to ensure consistency, this approach is often insufficient because it fails to account for geometric elements—such as cross-section dimensions—that significantly influence actual driver behavior. The primary objective is to review and document existing operating speed prediction models developed globally, grouped by roadway type, to help designers assess expected speed changes across successive road elements. By identifying limitations in current models, the report aims to enhance traffic flow and safety performance through more accurate speed prediction during the design phase. The document focuses heavily on Chapter 2, which synthesizes studies on speed models in North America, specifically for two-lane rural highways. The review analyzes numerous empirical studies conducted between the late 1980s and early 2000s, primarily using Ordinary Least Squares (OLS) regression and, in some cases, Artificial Neural Networks (ANNs). Key studies examined include those by Lamm et al., Krammes et al., and Fitzpatrick et al., which collected free-flow speed data (typically the 85th percentile) from various geographic regions in the United States. These studies modeled speed as a function of geometric variables such as horizontal curve radius (or degree of curve), lane width, shoulder width, superelevation, vertical grade, and traffic volume. For instance, Lamm et al. (1987–1990) developed models showing that degree of curve, lane width, and shoulder width were significant predictors, while Krammes et al. (1995) incorporated curve length and deflection angle. Later studies, such as McFadden et al. (2001), compared ANN models against linear regression, finding comparable predictive power. The findings indicate that horizontal curve radius is the most significant parameter for predicting operating speeds on curves. Studies consistently show that operating speeds drop sharply when curve radii fall below 250 meters, whereas speeds on curves with radii greater than 800 meters are similar to those on long tangents. Vertical alignment also impacts speed, particularly for heavy vehicles, though its effect on passenger cars is often negligible unless sight distance is limited. The report notes that spiral transitions do not significantly influence speeds for passenger cars within the analyzed data ranges. Furthermore, research by McFadden and Elefteriadou (2000) revealed that using 85th percentile speeds to evaluate design consistency underestimates individual driver speed reductions, with the maximum speed reduction being approximately twice the difference in 85th percentile speeds between tangents and curves. The significance of this synthesis lies in its identification of deficiencies in existing models, such as unrealistic assumptions about driver behavior, lack of uniformity, and limited applicability to vehicle types other than passenger cars. By documenting these gaps and reviewing practitioner perspectives, the report provides a foundation for future research aimed at developing more robust, uniform, and comprehensive speed prediction models. This work supports the transition toward performance-based highway design, where operating speed predictions are integrated into the design process to reduce speed variations and improve roadway safety.

Key finding

Horizontal curve radius is the most significant geometric factor influencing operating speeds on rural highways, whereas variables like superelevation and spiral transitions generally have minimal impact.

Methodology

review

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• road geometry
• urban rural setting