Automatic widening formula according to AASHTO 2011

[agori.politecnica.it]

I need to apply curve widening according to the formula(s) of AASHTO standard (2011) and taking into account a WB-33D vehicle. The widening applied by Civil3D, with automatic procedure command, is really higher than expected, resulting in a very wide road platform. It occurs selecting also other standard vehicles.

Checking the AASHTO standard file adopted by Civil3D (file: _Autodesk Civil 3D Metric (2011) Roadway Design Standards.xml), the standard formula (derived from formula no. 3-34 of AASHTO) is reported like:

formula="{n}*({w}+{r}-sqrt({r}^2-{wl}^2)+{c})+({n}-1)*((sqrt({r}^2+{A}*(2.0*{wl}+{A})))-{r})+0.1*({v}/sqrt({r}))- {tw}"

At first glance the formula, in particular the first portion derived from formula 3-31 "U", seems the one reported in AASHTO, instead the first element {w} is defined in the same Civil3D file as "{w} - nominal lane width or normal lane width edge of traveled way". Instead in the AASHTO 2011, this term, named as {u}, is defined as "track width on tangent".

I don't know if it could be the reason behind the wider widening on curve than expected by me (and manually calculated), however is there someone who have experienced the same problem or knows which could be the solution to that?

Thanks!

[Tomasz_Piotrowski]

Hi @agori.politecnica.it,

Could you please check what's warning has been displayed in the Panorama tab after creating the widening? 

In this case, Civil 3D normally informs us the set normal lane width is lower/higher than the minimum/maximum lane width in the standards file.

In the next step, Civil 3D uses minimum/maximum lane width to determine the increased offset value. This could be the cause of the inaccuracy between outcomes in Civil 3D and expected outcomes. 

[agori.politecnica.it]

Hi @Tomasz_Piotrowski ,

thanks for the clarification, I confirm all what reported in your screencast: adopting the WB-33D formula also on my Civil 3D I recognize those errors in Panorama tab. Nevertheless, the matter of widening is still unclear for me with reference to the following aspects:

1. Normal lane width should be 3.6 m, why 6 m? Maybe because under that name there is the reference to normal “travelled way” width (ref. table 7-3)?
2. Adopting standard formula, the possible change of normal lane width doesn’t affect the results in terms of widening? It should be dependent on lane width, isn’t it?
3. What’s the purpose of superelevation value in the widening command? It shouldn’t affect the calculation according to AASHTO formulas;
4. Wheelbase length: what’s the utility of this datum?
5. Adopting standard formula and also the table 3-26a: considering a radius of 450 m, design speed of 100 km/h, superelevation of 6 %, we have a widening of 0.7 m; even considering the adjustment factor for WB-33D, the widening should be 1.0 m, also because otherwise the widening would be similar to the lane’s width, which is strange for a 450 m-radius curve.

Please find attached the file with widending calculated with Standard Formula and adopting WB-19, providing then a Vehicle Tracking path simulation, just to visually check that widening of 2.2 m is too much, apart for what derived from the AASHTO formula. Thanks for your support in advance,

[Anonymous]

There appears to be a couple of errors in the widening equation in Civil 3d. 

1.  Civil 3d incorrectly calculates widening form Normal Lane width.  The Civil 3d equation should use the total normal travel way width, Wn (number of lanes times the normal lane width).  The equation utilizes the lane width input value instead of the correct value for Wn which is total # of left and right lanes times the normal lane width.  (Above-The Civil 3d equation uses the lane width for Wn (Total Travel Way) when it should use the # of left and right lanes times this value for Wn.)
2. Civil 3d incorrectly applies the calculated Wc (total travel way in the curve.  If you put in the normal travel way width Wn in for the Lane width then it correctly calculates the Wc (Travel way width of the curve)….HOWEVER, Civil 3D incorrectly applies this travel way width to only one lane and not the entire travel way.  The Wc should subtract the normal lane width from the Wc to obtain the correct widening on one side of the travel way.  Also, the Error message incorrectly applies to the Normal lane width This is why if you put in a lane width of say 12, an error message says you lane width is less than design file and defaults to AASHTO Wn of 20 feet.  However this 20 feet is the total travelway width in the AASHTO design file and Table 3-226B.
3.   The superelevation values are utilized to set to match the transitions from Wn to Wc to superelevation transition values.  This should be optional but isn't
4. The wheelbase value is only used if not utilizing a Widening Method

Additionally, lane widening should have the option to only utilize the actual curve radius and desired Wn and not only to Wn's in AASHTO Table 3.26b.  

[agori.politecnica.it]

Thanks for the recap and confirmation of the possible errors I had mentioned some months ago. It seems these aren't solved in the recent release of Civil 3D, therefore the automatic widening is still no working properly in my opinion

[Tomasz_Piotrowski]

hi @agori.politecnica.it ,

I'm sorry to hear that.

I don't work with CIVIL 3D anymore. However, I will try to reach out to our devs/technical support. 

[kirundajames0]

I have experienced the same problem, edited all the parameters and still the widening comes out wider than specified.

[yasmanydagarcia5ZGD4]

I’m experiencing the same issue with the 2023 version, and in the 2026 version, the XML file still uses the same 2018 AASHTO standards.
In the original formula: "{n}*({w}+{r}-sqrt({r}^2-{wl}^2)+{c}) + ({n}-1)*((sqrt({r}^2+{A}*(2.0*{wl}+{A})))-{r}) + 0.1*({v}/sqrt({r})) - {tw}"

I replaced {u} with {w}, and then assigned a fixed value of 2.6 to {u}, as shown in the figure. After this adjustment, the results improved and are now much closer to those provided by the AASHTO standards.

On another note, in the AASHTO document, the formula used to calculate the FA is actually incorrect — the square root extends all the way to the -r term, which it shouldn't. Fortunately, Civil handles this correctly.

Hope this helps, but there's still a need for proper updates in future versions.