Abstract: Ten drivers participate in driving simulation tests to examine the horizontal and vertical design indices of the basic sections of freeway underground interchange ramps. Using UC-win/Road, simulation models are established under various working conditions to collect driving characteristics and cardiac physiological data across eight different scenarios. The heart rate growth rate is used as a driving load and is calibrated with real-world driving data. Through regression analysis, a theoretical model is developed to link horizontal and vertical geometric design indices, driving speed, and driving load for combined sections involving longitudinal slopes, horizontal curves, and bends. Recommended indices for circular curve radii are derived based on the maximum superelevation rate. Key findings from the tests include: (1) The average driving speed on the left curve is 5.6% higher than that on the right curve, and the average heart rate growth rate increases by 9.0%. The average driving speed on the downhill section is 3.2% higher than that on the uphill section, and the average heart rate growth rate increases by 6.8%. (2) The driving load on the left curve section is higher than that on the right curve section, and the driving load on the downhill section is higher than that of the uphill section. (3) At a design speed of 40 km/h, the safe gradients of the downhill and uphill sections are 4.56% and 5.35%, respectively, based on driving load. (4) The recommended indices for the circular curve radius of the underground interchange range from 1.17 to 2 times that of the surface interchange, and this difference gradually increases with increasing longitudinal slope and design speed.

Key words: freeway, underground interchange ramp, geometric design index, driving simulation, driving load