A reviewer at Illinois DOT rejected our SSD numbers on a recent submittal, which were generated by GEOPAK.
We have GPK_PROFILE_COMPUTATION set to AASHTO2001, and we are using the correct K-value table as supplied by IDOT. If we do the calculations by hand, using equation 3-45 from AASHTO 2001 or using the equations from the IDOT design manual (which is basically reprinting that section of AASHTO), we get values that do not even come close to what GEOPAK is suggesting for SSD in the profile tools.
Could we be doing something wrong in GEOPAK to get these wrong values?
Forgot to mention that we're using GEOPAK 8.8.3.24 on Microstation v8 2004 8.5.2.70
Adam: Can you post your dataset here? We need the DGN which contains the design profile in question and GPK file.
Robert GarrettSenior Product EngineerBentley Systems Inc.
Adam: We have tested with some of our own datasets and compared to the AASHTO equations. Equations 3-45 in 2001 manual and 3-43 in 2004 manual. The results labeled from GEOPAK match the AASHTO formulae.
Can you comfirm that you have set the configuration variable to enable the AASHTO 2001 forumae. Prior to 2001 the formulae were different because the eye and target heights were different. You need to have the following in your config variables to get the proper 2001 (and later) calculations:
GPK_PROFILE_COMPUTATION = AASHTO2001
Adam: Just to reiterate. If you can send your data would be most helpful.
I have asked someone from the project team to zip up a drawing and the GPK file and will post it as soon as they get back to me. I appreciate the attention you guys are willing to devote to this.
And yes, GPK_PROFILE_COMPUTATION is set to AASHTO2001, confirmed via expand echo key-in. I suspect we're doing something else wrong.
Going to see what's holding up the dataset...
Ok, dataset attached. If you look at the crest curve in the image below, you can see one example of the problem. The K-value is shown as 161.
IDOT's equation (http://www.dot.il.gov/desenv/BDE%20Manual/BDE/pdf/chap33.pdf#page=29), is K = (SSD^2)/2158 - this is derived from the set of equations in AASHTO2001 regarding crest vertical curves. Solving for SSD gives us:
SSD = sqrt(K x 2158)
SSD = sqrt(161 x 2158) = sqrt(347438) = 589.44
As you can see, GEOPAK labeled the SSD as 1637'.
Maybe it's a units issue? Microstation seed file and GEOPAK Prefs are all set to English units.
Adam:
There are two quations in the AASHTO manual. In the 2001 manual these are equations 3-45 and 3-46. In the 2004 manual the identical equations are 3-43 and 3-44.
The first equation which is the one you used is the first equation. But this only applies when the SSD is less than the length of curve.
Using the first equation results in SSD = 589 as you noted, but since 589 is greater than the length of curve (110) then this is the wrong answer.
The other equation is used when SSD is greater than the length of the curve, which is the one which applies in your example. The second equation is:
L = 2*SSD - (2158/A)
which results in:
SSD = (L + 2158/A) / 2
And thus the resultant computed SSD = 1637.
I appreciate the explanation, Robert. Will take it back to the design team.