AASHTO LRFD Steel Design Output Issues

Question:

MIDAS is reporting that girder 4 at pier 2 is failing. After investigating the detailed MIDAS output (see attached annotated PDFs), there appear to be several errors and it is unclear where the failing moment is coming from:

  • The dead loads on girder 4 are greater than girder 5. Girder 4 is a shorter span member than Girder 5. I would expect girder 5 dead load to be larger.
  • The applied moment in the load summary table for the governing load combination, Str II, does not match or comes close to the reported failing moment for flexural resistance on discretely braced compression flange, Muy.
    • The beam diagram in MIDAS reports the same/very similar moment as the load summary table (see attached screenshots)
    • Muy is approximately 50% greater than Mu given by load summary table.
    • This issue appears to be prevalent in each element analysis. For element 1960J, Muy = 343248 k*in while the applied moment, Mu, is 225275 k*in.
  • Torsional moment of inertia, J is given as 0.000 in^4. The section property J cannot be zero. Additionally, manually computing J according to the formula does not yield 0.000 in^4.

  

Respectfully,



Answer:

Hi,

Thank you for waiting. I have looked into your model file and found some modeling errors which lead to the inappropriate design results. In order to find maximum design moment within unbraced length, beam elements whose member type is 'Brace' should be created at every braced points as shown below.


Then, the problems you have found will not appear any more. I have attached the revised model file and design report for your check.
Also, it is recommended to download the latest patch below. You can extract it into the 'midas Civil' folder.

Regards,
DK


 


DK,


Thanks for the response. I was not aware of this, but apparently, a co-worker also ran into this issue. Suman helped him at the time and suggested that the transverse dummy deck element be defined as a "Brace" type member (see attached ticket). However, this does not always line up at the location of the diaphragms.


Which method is the most accurate; defining the closest transverse dummy deck element as "Bracing" type or creating the dummy bracing section and drawing it at each diaphragm location as you mentioned?


Additionally, can you please explain why defining this bracing member, either via deck dummy or bracing dummy, produces more accurate results? Does doing this help the program correctly interpret the unbraced lengths? I manually inputted the unbraced length for the girders using the "Unbraced Length" tool; does this matter?


Respectfully,



Hi,

According to 6.10.1.6 of AASHTO LRFD, if the flexural resistance is based on lateral torsional buckling, the design moment Mu should be determined as the largest value of the major axis bending moment throughout the unbraced length. The unbraced region and unbraced length between cross-frames can be recognized by defining Span Information and the 'Brace' members. The design moments are taken as the maximum value within the unbraced region. If you do not define either Span Information or Brace members, the design moments will be taken from the individual elements, which is not correct. Therefore, creating 'Brace' members at each diaphragm location will give more accurate results. 

Regards,
DK


Creation date: 8/7/2018 7:54 PM (dklee@midasit.com)      Updated: 8/8/2018 5:25 AM (dklee@midasit.com)
Files   
20180718_Flyover_Revision.mcb
20180718_Flyover_Revision-midas.mcb
20180718_Flyover_Revision-midas.xlsx
DAOTicket_Suman.jpg
DataImage62.jpeg
Elem_1740and1960 - StrII My-MIN.jpg
Elem_1740and1960.jpg
G4_Elem1740J.pdf
G5_Elem1960J.pdf
Overall Bridge View.jpg