Release Notes

Bug fix of enhanced capacity model for wide single lane entries.

Bug fix of Accident Model - 6 lane data now displayed.

HCM has been updated to include the 2016 parameters.

Screen layout has been slimmed to fit comfortably on modern laptops. All screens now detect when laptop has been undocked and revert to primary display.

Minor bugs and formatting issues have been fixed.

Rodel v1.88 is an enhanced version of the previous DOS version of Rodel plus a Windows interface. The latter overcomes the problem of the DOS versions being incompatible with Windows 7.

v1.88 is interim to Rodel v2 which will be a totally new program written from the ground up.

New Features in v1.88:

• Improved Screen Interface
• Right Hand Drive and Left Hand Drive
• U.S. Peak Hour Factor (P.H.F.)
• English and Metric Units
• Compliant with HCM 2010 recommendations
• In addition to Kimber's Equations and the 6 geometric parameters, RODEL now includes:
  • HCM Capacity Model
  • Control Delay (as per HCM 2010)
  • Pedestrian Capacity Effects Crosswalk Factor (XWfact)
• Calibration - Allows calibration to local conditions by adjustment to both the capacity intercept and the slope of the capacity line
• Crash and Safety Analysis
• Economic Analysis (Delay and Safety Costs)

1. v1.88 uses a standard Windows interface using the usual menus, icons, buttons and selectors.

   An enhanced Interface with a cleaner screen, with the text included in the data fields will be provided before the end of the beta tests. This is to be followed by a revised interface based on version 2 with the geometric data displayed horizontally rather than vertically. This is consistent with the other data and allows a better arrangement of the fields.

   The Function Keys used in the DOS versions have been retained as Shortcut Keys. (The function of each shortcut key is listed in the Help Menu).

2. Optional Left Hand Drive or Right Hand Drive (LHD or RHD) This affects the displays of the Turning Flows. RHD turning flows for a 4 leg roundabout are input - [U-turn - Exit3 - Exit 2 - Exit1]. For a 90 degree crossroad, these are the turn flows - [U-turn - Left - Thro - Right]. Using the destination Exit Number to define Turns works for non-symetrical roads and for 3, 5 and 6 leg roundabouts where Left Through and Right are inadequate turn descriptors.
3. Feet have been added to Meters for geometric input and output
4. Peak Hour Factor (PHF) has been added to the existing Synthetic and Direct flow peak hour traffic profile options.
5. The percentage of Trucks (%Tr) is used instead of Passenger Car Units.
6. Double Roundabouts are much easier to Model. Rodel for each roundabout can be opened and run with the exit flows on the connecting leg from each roundabout copied and pasted into the approach flows of the other. The queues and delays converge and stabilise after 2 or three iterations. Treble roundabouts or a series of roundabouts can be done the same way.
7. The Start Queue at the beginning of the modelling period is now included. This is especially useful when using observed (Direct Flows) from existing situations. The omission of significant start queues may lead to a large underestimation of queues and delays. For future situations, no start queue data is available. This deficiency can be overcome by making the Time Period 90 minutes instead of 60 minutes. (The input flows are still the 60 peak hour). The queues at time 15 (the start of the peak hour) are then estimated. When modelling 90 minutes the Flow Period is set to 15 75 (the central peak hour) and the Results Period set at 15 75 so the results are derived for the central peak hour. Consequently, by modelling for a 90 minute period, peak hour results can be derived from peak hour flows, but with non-zero peak hour start queues. The alternative of modelling 60 minutes implicitly assumes zero peak hour start queues.
8. The geometric input now includes the number of Entry Lanes and the Circulating Width. Both enable many additional checks and comparisons between the geometric parameters and the designer is warned about incompatibilities. Pease Note that the Circulating Width is that which receives the entering traffic. This is almost invariably narrowest between the Central Island and the Splitter Island at the next leg of the roundabout. For example, the Circulating Width input into Rodel for Leg 1 is the width between the Central Island and the leg2 Splitter Island.
9. The Capacity Factor has been replaced with Cap+-. The absolute change in capacity is now used, not a perecentage change in capacity. This feature can be uses for sensitivity testing or for allowing for capacity reduction caused by a bus stop or parked vehicles etc.
10. A Calibration parameter has been added to v1.88. The designer can now input an absolute change in capacity for calibration purposes. This change is added or subtracted from the mean capacity. Consequently, the Capacity Curve (line) moves up or down while, crucially, the slope of the capacity line remains unchanged.

    Percentage changes in capacity alter the slope of the capacity line. This incorrectly produces larger changes in capacity at low circulating flows (when capacity is large) and small changes in capacity when circulating flow are large (when capacity is small).

    Capacity Calibration should be used with extreme caution. So far US capacity measurements have not been made during genuine at-capacity operation as US roundabouts have yet to reach their design year flows. Even then it takes some years beyond the design year for true saturation to be reached. During genuine at capacity operation driver behavior includes merging, circulating drivers modifying their headway to allow others to enter, Entering drivers ‘nosing’ across the yield line, gap forcing and short periods of priority reversal. If this is not present then true capacity has not been reached. (This has been witnessed in the US at a ski resort during exceptionally high peak vacation traffic volumes). Capacity is also underestimated because US drivers are still somewhat cautious and reticent when using roundabouts. However as genuine capacity is approached and queues and delays become onerous, these very conditions increase driver assertiveness and capacity increases to supply the demand. Current US capacity measurements therefore underestimate capacity.

    It is also important to understand that capacity will not be the same for identical roundabouts with identical traffic flows. The between site variation is about + or - 14%. This variation is due to the combined effect of subtle but ‘capacity significant’ differences in geometry, signs and stripes, and also the difference in driver behavior at different locations. Consequently, measuring the capacity at a single roundabout (even if it is operating at genuine capacity) is utterly insufficient for calibration as it only gives one random point on the between site capacity distribution. This could be anywhere up to +14% or -14% of the mean capacity. Only by measuring capacity at a large number of roundabouts can the capacity distribution be derived and the mean capacity determined for calibration or comparison with a roundabout model. This is very expensive and usually prohibitive to all but projects funded by central government.

    Rodel explicitly allows for this uncertainty I capacity by means of the Confidence Level. The default 50% CL uses the mean capacity. However, the Designer can select a higher CL (lower point on the capacity distribution) to check if the design works acceptably if the actual capacity turns out to be pessimistic. Rodel users are urged to read the chaper on Confidence Level in the Rodel Manual.

    Capacity Calibration should be used with extreme caution. So far US capacity measurements have not been made during genuine at-capacity operation as US roundabouts have yet to reach their design year flows. Even then it takes some years beyond the design year for true saturation to be reached. During genuine at capacity operation driver behavior includes merging, circulating drivers modifying their headway to allow others to enter, Entering drivers ‘nosing’ across the yield line, gap forcing and short periods of priority reversal. If this is not present then true capacity has not been reached. (This has been witnessed in the US at a ski resort during exceptionally high peak vacation traffic volumes). Current US capacity measurements therefore underestimate capacity. It is also underestimated because US drivers are still somewhat cautious and reticent when using roundabouts. However as genuine capacity is approached and queues and delays become onerous, these very conditions increase driver assertiveness and capacity increases to supply the demand.

    It is also important to understand that capacity is not the same for identical roundabouts. The between site variation is about + or - 14%. This variation is due to the combined effect of subtle but ‘capacity significant’ differences in geometry, signs and stripes, and also the difference in driver behavior in different parts of the country. Consequently, measuring capacity of a roundabout (even if it is operating at genuine capacity) is utterly insufficient for calibration as it only gives one random point on the capacity distribution. This could be anywhere up to +14% or -14% of the mean capacity. Only by measuring capacity at a large number of roundabouts can the capacity distribution be derived and the mean capacity determined for calibration or comparison with a roundabout model. This is very expensive and usually prohibitive to all but projects funded by central government.

    Rodel explicitly allows for this uncertainty I capacity by means of the Confidence Level. The default 50% CL uses the mean capacity. However, the Designer can select a higher CL (lower point on the capacity distribution) to check if the design works acceptably if the actual capacity turns out to be pessimistic. Rodel users are urged to read the chaper on Confidence Level in the Rodel Manual.

11. The Capacity model has been upgraded for single lane roundabouts to deal with very wide entries (for trucks) into narrower circulating roads. The designer inputs the actual geometry and v1.88 derives effective geometry for use in capacity estimation. Warning messages about inefficient over wide lanes are also displayed when appropriate.
12. Multi-lane capacity fades as lanes become very wide. The multi-lane capacity model has been enhanced to derive the effective lane widths to avoid overestimating very wide lane capacity. Like single lane entries, the designer inputs the actual multi-lane entry geometry and v1.88 derives the effective geometry for capacity estimation. The designer is warned about over wide inefficient lanes.
13. Optionally, the old capacity or new capacity model can be selected, so the capacities can be compared. The old model will give results similar to the DOS versions.

    Most of the time, when lanes widths are more normal, the results between the old and the new capacity model will be the same. Only when lanes are over wide or when the receiving section of the circulating road is narrower than the entry width do the models differ significantly.

14. A Scheme Notes dropdown is included to record design notes. This is useful when sending a Rodel file to a client. The Scheme Notes are in File Menu.
15. Rodel v1.88 has a file import feature for importing Rodel Data files from earlier versions of Rodel. The ‘Rodnames’ file in the old Files folder is selected and this displays both Rodel file names and the associated Roundabout Names. A single file can then be selected and imported. The file can be subsequently saved in a different folder under a chosen name.
16. v1.88 can be installed ‘stand alone’ on a PC or loaded on a server for multiple users. Rodel data files are stored locally on the users PC. Installation on each PC will require an initialization password for either situation. If a stand alone installation is removed and linked to the server version, the original ‘sand alone’ password will work with the server version accessed from the same PC. The number of users is only limited by the number of license.
17. The contextual manual and the print functions will be completed before the end of the beta testing period.