Construction Stage Analysis Results Comparison with Other Softwares

[netsonicyxf]

This thread is a continuity of Construction Stage Analysis for Composite Arch Rib 

The arch has 96m in span and 19.2m in rise. It has pin support at both ends and is made of steel and concrete.
It has 3 Construction Stages
1. field installation of steel tube

2. put concrete in the bottom steel tube

3. put concrete in the top steel tube

We use ASBD, Midas Civil, RM Bridge (TDV) and Sofistik to do the construction stage analysis.

For simplicity, the model is under self-weight only, no time effects.

The boundary conditions are the same for all the softwares in all 3 stages, the support is at geometry center of the whole steel tube as shown in the stage 1.

ASBD results (The moment diagram due to self-weight is symmetrical, so only half of it is hown here)

Stage 1

Stage 2

Stage 3

The ASBD results are similar to those from Midas Civil if choosing "Linear Analysis & Accumulated Stage"

Linear analysis is carried out with accumulated effects of the models of each construction stage.

Midas Civil "Linear Analysis & Accumulated Stage" results 

Stage 1

Stage 2

Stage 3

In Midas Civil, if we choose "NonLinear Analysis & Accumulated Stage"

Geometric nonlinear analysis is carried out with accumulated effects of the models of each construction stage. This option is used for the forward analysis of a cable stayed bridge considering large displacement.

The results are different from those by using "Linear Analysis & Accumulated Stage".

RM Bridge and Sofistik show very similar result in all three stages,
Midas Civil with "NonLinear Analysis & Accumulated Stage" shows similar result in the stages 1 & 2 as those from RM Bridge and Sofistik, but very different result in the stage 3.

For RM Bridge, Accumlate stiffness (Stage) is used

Nonlinear calculation of structures, starting with an initial internal force state accumulated from results of previous load cases. This option also applies to structurally variable systems with changing element activations in different construction stages

Stage 1

Midas Civil "NonLinear Analysis & Accumulated Stage" results 

RM Bridge

Sofistik

Stage 2

Midas Civil "NonLinear Analysis & Accumulated Stage" results 

RM Bridge

Sofistik

Stage 3

Midas Civil "NonLinear Analysis & Accumulated Stage" results 

RM Bridge

Sofistik

The ASBD doesn't allow me to do nonlinear analysis.

The ASBD model is attached

My question is, which result is correct?

[dave_geeves]

Hi,

I'm not sure I am in a position to answer this as I am not a user and do not have access to the comparison software.  You are quite right that ASBD does not support large displacement analysis, so a comparison for this cannot be made, but I am pleased that the results from ASBD small displacement analysis is reasonably close to the other software, 

I must admit that I am slightly surprised that a  max 9mm deflection over a 98m span under self weight (small deflection analysis) would make such a difference to the BM results when considering large displacement analysis, but it seems that the biggest difference is in stage 2 where the transformed centroid is eccentric from the supported points.

So to summarise, without knowing the algorithms and procedures used by the comparing software  for the non-linear staged construction analysis, it is impossible for me to judge which is giving the more accurate answer.

I'm sorry that I can be of more help

Kind regards

Dave Geeves

[netsonicyxf]

Hi Dave,

Thank you for the comments.

Here is the summary of the results.

I'm also surprise that Midas Civil show very different results from geometry linear and nonlinear analysis.

The RM Bridge and Sofistik show very simlar results from geometry linear and nonlinear analysis.

The thing I'm not sure is whether the moment at pin support (M1) occurs at STG 2 should continiously exist in Stage 3 as shown in ASBD result?

Best regards

Sean

[dave_geeves]

Hi Sean,

If stage 3  in itself should not have any moments at the supports as the axial force generated is purely along the transformed  section centroid for stage 3, which is where the support is connected.  However, stage 2 on its own will have  a moment at the supports as the axial force is eccentric to the support position due to the transformed centroid not being central to the overall section.  For a linear solution the two can be combined so there will be a moment at stage 3 when combined with previous stages.

Some will argue that at ULS the stress discontinuities due to staged construction will get distributed, so can be ignored and the stage 3 section only can be used to resist the full load, but this is only acceptable for Class 1 and Class 2 sections in accordance with EN 1994-2 5.4.2.4(2).  However for Class 2 and 4 sections and for SLS this is not an option so the stress discontinuities will need to be considered.  If you consider each stage separately and add the resulting stresses together for each stage then the stress discontinuities will exist.  So at stage 3 the combined stress will not be purely compressive and even over each of the components which will result in a line of thrust eccentric from the section centroid/support and will hence produce a moment.

I hope this is what you meant, and has answered your main  questions ok.  With regards to the other differences - it seems quite strange that for the linear analysis alone ASBD and Midas results are very similar for all stage whereas RM Bridge and Sofistik results are markedly different from these  for stage 2 and 3.  I can only think that the way of defining the construction stages and staged loading is different.  I think that inspecting the staged results on there own would highlight this.

I also noted that the way you have applied the staged loading in ASBD is not  how I thought it would be in practice (although the way you have done it is perfectly valid for comparison reasons).  I would have assumed that stage 1 would be stage 1 section with stage 1 and 2 loading applied (as the concrete is wet at this point.)  Stage 2 would consist of stage 2 section with stage 3 loading applied.  Stage 3 would be stage 3 section with any other permanent loading that would act on the complete composite section after the top tube concrete has hardened.  I am just wondering if RM and Sofistik construct these automatically and are not the same as ASBD and Midas.  Just a thought.

I hope this helps

Kind regards

Dave Geeves

[netsonicyxf]

Thanks Dave for suggestion.

Now I model the wet concrete, there are 5 stages

Stage 1, Steel section, steel self-weight and steel stiffness
Stage 2, wet concrete in the bottom tube, concrete self-weight added, but no stiffenss added
Stage 3, hard concrete in the bottom tube, no extra loading, composite stiffness added.

ASBD doesn't allow analysis, because no loading in this stage, so I have to add a very small loading.
Stage 4, wet concrete in the top tube, concrete self-weight added, but no stiffenss added
Stage 5, hard concrete in the top tube, no extra loading, composite stiffness added.

ASBD doesn't allow analysis, because no loading in this stage, so I have to add a very small loading.

Below is the result

I didn't show geometry linear analysis results of RM Bridge and Sofistik, because the difference between linear analysis results and nonlinear analysis results are minimum.

Model is attached

Best regards

Sean

[dave_geeves · Accepted solution]

Hi Sean,

Thanks for stretching my brain, this has been a really interesting discussion. Without knowing exactly how the comparison software is calculating the moments I am having to make a few assumptions, but I can explain how ASBD is calculating  the moments and guess how the other software may be doing so.

At each stage, ASBD is calculating the moments and forces resolved at the stage section centroid. This is important so that these results can be used on that particular section to derive the stresses at that stage.  The stresses can then be added (assuming a linear analysis) to determine the as-built stress distribution.  However, adding the moments at the various stages in the results output is adding the values resolved at different positions (as the transformed centroid varies) so the effect of the large axial force at different positions  is not considered when looking at the moments alone.

By looking at the results I believe that the moments in the non-linear analysis of comparative software are being calculated at a common point (not the centroid at each stage) so this may be valid for the final result but may not be useful for determining stresses at different stages of construction.

My original assumption was that you were using Eurocodes (based upon the data file you attached) but I now see that your latest data file is aligned to AASHTO LRFD.  Unfortunately I'm not an expert on using LRFD as I only have a very old copy of the 1998 edition and I know it has changed a lot since then.

I hope this has helped a little but I am hopeful that others may join the conversation who have a little more experience with AAHTO than me.

Kind regards

Dave Geeves