Old PGSuper File and Service Stresses

5 posts / 0 new
Last post
Visitor (not verified)
Old PGSuper File and Service Stresses

I am trying to open a Version 2.6.1 PGSuper file in Version 2.8.2. When prompted regarding differences in Library files, I renamed the files to keep it as the original. As I was verifying the program output to ensure nothing has changed, I noticed that Service I and III stress checks at midspan (0.5Ls) was off from what was expected by hand calculation at Bridge Site 3 (Final Condition).

Based on the Girder C Detail Report of the attached file:
> Sum(DC_fb,max) = 2.313 ksi
> Sum(DW_fb,max) = 0.00 ksi
> Sum(LL_fb,max) = 1.073 ksi

Expected fb,serv3,max = (1.0)(2.313) + (1.0)(0.00) + (0.80)(1.073) = 3.1714 ksi
PGSuper 2.6.1 output = 3.171 ksi
PGSuper 2.8.2 output = 3.344 ksi

I've checked the load factors and the individual stress components DC, DW, and LL by hand and I'm matching those, but when combined for Service Stress checks, the value is different when opening in a newer PGSuper version.

My questions are:
1) Are there additional changes to the .pgs file, other than the libraries, that I (the user) am not prompted to be made aware of?
2) How are Service III stress checks computed in PGSuper 2.8.2?

Thank you,

Rick Brice
In version 2.6.1, the

In version 2.6.1, the effective prestress included elastic gains due to deck shrinkage. However, there was not a corresponding increase in bottom fiber tensile stress.

Swartz, Scanlon, and Schokker point out in their Fall 2012 PCI Journal Article that you can't take the benefit of elastic gain on the effective prestress without taking the corresponding increase in tensile stress in the girder.

In version 2.7.0 we included stress due to slab shrinkage in the stress evaluation of the girder. We also included design criteria that enables you to control the amount of slab shrinkage to be included in the analysis. The PCI BDM currently recommends using 50% of the strain from the LRFD specifications.

The increase in Service III tensile stress you are seeing is the tension caused in the bottom of the girder due to the shrinkage of the deck concrete.

You can read more about the changes in Version 2.7.0, including stresses due to slab shrinkage, the "Highlights of PGSuper Version 2.7" online training document found at http://www.pgsuper.com/content/highlights-pgsuper-version-27

I hope this helps.

Best regards,

Richard Brice, PE
Software Applications Engineer
WSDOT Bridge and Structures Office

gavinbv (not verified)
Thank you, Richard. This

Thank you, Richard. This helps a lot. We are now able to match the Service Stresses.


gavinbv (not verified)
Sorry to reopen this topic,

Sorry to reopen this topic, but previously when I was calculating the Service stresses, I was taking the deck shrinkage stress directly from the PGSuper output, thus matching the output Service stresses.

However, when calculating the deck shrinkage stress by hand, it appears that the shrinkage coefficient, Ksh, is being applied twice in PGSuper Ver 2.8.2, once to e_ddf, then again to ft = e_ddf*Ad*Ecd/[1+0.7Psi_bd(tf,td)]*(1/Ac - e_d/S_tc) (similarly when computing fb due to deck shrinkage). Ksh appears to be applied correctly (once to e_ddf) for DfpSS.

I tried looking through the forum and list of "Known Problems" and did not find anything on it. Unless I am misunderstanding how the deck shrinkage stresses are calculated?

Thanks again in advance.


Rick Brice
Gavin, Thanks for bringing


Thanks for bringing this to my attention. It is in fact a bug. The shrinkage coefficient Ksh is being applied twice when computing stresses in the girder.

I will get this problem fixed for the next release.

In the meantime, there isn't really a work around. Since stresses due to deck shrinkage are a small compared to the other contributors, the overall error will also be small. The error is also reduced the closure Ksh gets to 1.0.


Log in or register to post comments