Document Type

Open Access Thesis

Embargo Period


Degree Program

Civil Engineering

Degree Type

Master of Science in Civil Engineering (M.S.C.E.)

Year Degree Awarded


Month Degree Awarded



According to the US Department of Transportation, around 20 people die on a daily basis in a signalized intersection, with most of these resulting from angle or head-on collisions. The US-DoT’s Federal Highway Administration (FHWA) has identified modern roundabout intersections to be substantially safer than signalized intersections, due in part to the reduction in conflict points from 32 in a traditional signalized intersection to 8 in a modern roundabout. Despite the increased adoption of modern roundabouts across the US, there are a number of specific design elements for which the direct impact they have on operational and safety related performance of the roundabout remains unknown. To be specific, there is currently no conclusive research on the direct effects related to the introduction of a reverse curve (S-curve) on the approach to a roundabout. Moreover, what are the impacts of S-curves of varying geometries on the approach to a roundabout? This research employed a series of microsimulation-based analyses to investigate the speed related impacts related to the introduction to S-curves on the entry to a roundabout.

An existing roundabout, in Amherst, MA, USA was used as a case study for this experiment. The data at each approach of the roundabout was collected by a static camera strategically placed to attain both the pedestrian and vehicle count during peak traffic hours. The data was manually reviewed to determine the upstream and downstream vehicle counts.

The dimensions and angles of the existing roundabout were measured from Google earth and the image was extracted to AutoCaD Civil 3D. Since the objective is to check whether S-curves near an approach have a significant impact in speed, the deflection angle of the roundabout was not altered. The turning radius and angle at the approach was cross verified by measuring it on site. The existing roundabout was considered as the base model. The four approaches of the roundabout have different entry angles and radii. The revised models were drafted by strategically placing the S-curve at each approach and by steadily increasing their deflection angle and approach radius.

The base and revised models cases were initially modelled, after which the conventional linear approach was modified to an S-curve and evaluated. Field data from the locations were to and calibrate microsimulation models on AIMSUN. The resulting trajectory data was analyzed for both the base case as well as three levels of experimental S-curves (ranging from 30 to 60 degrees) on each roundabout approach (16 total). The results provide evidence to suggest that a significant reduction in speed can be realized with a minimal amount of the reverse curvature on the roundabout approach. The trajectory output files were then imported into the Surrogate Safety Assessment Model (SSAM) to determine the number and type of conflicts experienced at each approach under each scenario evaluated in AIMSUN.

First Advisor

Dr. Michael Knodler Jr.

Second Advisor

Dr. Eleni Christofa

Third Advisor

Dr. Cole Fitzpatrick

Available for download on Saturday, May 11, 2019