Curved Steel-Composite Bridge - Canadian Code 


Title : Numerous MIDAS Civil questions


I am currently working on modelling a curved, continuous 3 span, superelevated, composite steel I-girder structure. We have just recently purchased this software and I have generated a few questions throughout the modeling of my first structure. The list of questions is below, your comments would be much appreciated. Note that this design is to be in accordance with CSA-S6-14 (Canadian Highway Bridge Design Code)

2. How should the location of the Lanes be chosen? Are these to be located based on the proposed lanes for traffic when the structure is in service, or shall these be based on division of the structure into the required number of lanes as per Table 3.5 of CSA S6-14 (CHBDC)? I would assume the latter is correct. In such a case, where wide lanes are required (say a 9.5 m wide bridge which requires two 4.75m wide lanes as per Table 3.5), does MIDAS vary the position of the vehicle laterally within the lane to determine the largest/smallest distribution of forces to the adjacent girders? I believe this is accounted for in the moving load optimization option; however I could not find any information on what each parameter was referring to (ex: margin?)Also, for the case of a 10.0 to 13.5 m lane, where both 2 and 3 lanes should be checked, does the software have the ability to check both cases automatically or do two separate models/analysis have to be run?

3. What is the difference between an Elastic link of rigid type and a Rigid link? In which circumstance should each be used?

4. Does the program have any way of accounting for section loss to girder elements? It is often that steel girders are designed assuming some degree of future section loss (say 10%). Does this have to be manually accounted for?

5. Is there any way of accounting for bolt holes in the flanges of girders (In locations where stiffeners are bolted to flange)? The strength properties of the net section of the girders at bolt hole locations will be decreased at their locations? Can MIDAS account for this is does this need to be accounted for by modification factors to the section properties?

6. For composite sections in regions that will experience both positive and negative moments within the envelope – If the concrete in the section experiences stresses exceeding the cracking stress during negative moments, does MIDAS account for adjustment of the properties of the section to be of a cracked section when determining the capacity for positive moments? If not (which I imagine this would be hard to do), how should this be accounted for in the model? Should all elements within the region experiencing both positive and negative moments be adjusted for cracked section properties?

7. For time dependent materials, I have seen two methods used in tutorials. One being definition of the creep/shrinkage graphs using the time dependant material properties and one using multiple Es/Ec values (n for shortterm and 3n for long term) for the section which defines stiffness scale factors which can be activated for the stage when the section is considered fully composite. What are the differences between these two methods and how will it affect the results? When should each method be used?

8. In tutorials, I have seen two different ways of handling the self weight of deck materials. One being using the self weight load case and the other being inputting a weight density of 0 for the deck material and applying beam loads to each beam. What are the differences between these two methods and how will it affect the results? When should each method be used? I''ve noted that during creation of load combinations, the program can not differentiate between the factor being applied to the self weight load case. Therefore, if considering a composite section, should the unit weight of concrete within the section be left as zero and an additional dead load applied to the girder elements based on tributary width in order to allow for the different load combination factors to be applied to the steel and concrete materials?

9. When running moving loads, if using the CL1-625-ONT Truck will MIDAS drop effects of axles that are not contributing to the maximum load effect. For example, on a small span multiple span structure, if all 4 front axles are on span 1 and the rear axle is on span 2 (which would decrease the moment on span 1), will MIDAS ignore the rear axle (i.e. consider a CL2-625-ONT truck essentially) in determining the maximum moment of span 1, or do all 3 trucks have to be run to capture this effect?

10. When defining elastic links for curved bridges, it is my understanding that the local axis of the elastic link should be tangent/perp to the curve at that location. Is it best to adjust the local axes of the nodes the link is connected to or use the beta angle with the nodes kept in the global ucs?

11. If superelevation is manually modelled in the superstructure, will the additional transverse loading due to the gravity loads on the superelevated deck elements be accounted for by MIDAS? Superelevation would be modelled using superelevated transverse deck beam elements between girders.

12. Are centrifugal forces on curved structures accounted for automatically? If not, what is the recommended procedure for accounting for such forces?

13. How does MIDAS account for calculation of unbraced length? If the lateral bracing is defined as truss members, does MIDAS recognize the elements as braces or do they need to be specifically noted to be brace elements? If so, how are the elements adjusted to be considered brace elements?

14. The weight density automatically generated by the material properties in the CSA database do not appear to be equal to the unit material weights defined in Table 3.4 of CHBDC for concrete. For example, the automatically generated unit weight of CSA(RC) C35 concrete is 22.56 kN/m3 versus the 24.0 kN/m3 identified in Table 3.4 of CHBDC for reinforced concrete steel. It appears that concrete material properties will have to be user defined to have the correct weight density meeting CHBDC if using the self weight feature for considering these element''s loads.

Answer: Hello User,

Thanks for writing to us.
Answers for your question are below:

2) For moving load optimization refer this links

If we are using Traffic lanes, then for 10 to 13.5m lane > where both 2 and 3 lanes should be checked, in the software we have the ability to check both cases automatically by creating two sub-load cases to consider the effect simultaneously.

3) KB material is available, check this link

4) Yes, we can reduce the sectional properties to account for any section loss. Goto section manager > stiffness > stiffness scale factor

5) Section Stiffness scale factors can be used to modify the sectional properties.

6) It's important to note that cracked sectional analysis is required for Service Limit states but not for Ultimate Limit States.
In the analysis, if we want to perform cracked sectional analysis then we need to use "Section Stiffness Scale factors" to modify the sectional properties accordingly.
Thereby we will find changes in the stress and moment demands. Which should be compared with stress and moment capacities respectively in the design.

Ultimate Limit State

In the design of sections, we check demand vs capacity at ULS . While calculating the capacity (Ultimate Limit State), Strain-Compatibility Method is used, where we determine the exact Neutral Axis(by iterations) and finally calculate the capacity of the section. See the below snapshot. where cracked concrete below the neutral axis is ignored automatically.
So, this section stiffness scale factors applied in the analysis will only affect the analysis results(Demand) in ULS case but not the capacity calculation.

Service Limit State
In SLS we check stresses with the corresponding allowable stress limits. If the cracked stiffness is used in the program, we can obtain the stresses based on the cracked elastic case. Hence we can directly compare with allowable stress limits as per code.

7) Steel composite bridge modeling can be divided into three cases. see the below snapshot and Refer attached Steel-composite Midas Design guide.

First case: Using SW feature (Density value given);
Second case: manual application of Selfweight as UDL (Density =0).
These two methods are two ways of applying self-weight to the composite section which leads to the same result. But as you mentioned while making load combinations, for UDL case, we can differentiate the load combination factors applied to the steel and concrete materials which cannot be done in the first case.

9) I think Midas will automatically consider the placement of the rear axle based on the traffic lane length under consideration for maximum effects in the elements. Refer attached pdf.

10) See this link

11)In a curved bridge, the outer girder experiences an increase in load due to the curvature while the inner one experiences a decrease in load.
In a grillage model, If superelevation is manually modeled in the superstructure, then additional transverse loading due to gravity loads can be accounted as these will be generated because of the geometric eccentricities in built in the model.

12) As per Canadian code, the program does not automatically apply this centrifugal loads, we need to manually apply them to the model.
See this links for manual application of centrifugal forces.

13) We can alter these parameters. See below snapshot for details.

14) Changing the code again to "None" will give you the option of changing the density value. Follow below procedure.

Hope this helps to clarify all your queries.

Creation date: 10/30/2017 7:52 AM      Updated: 10/30/2017 5:29 PM
108 KB
151 KB
15 KB
55 KB
113 KB
115 KB
02_Steel Composite Design Guide-AASHTO.pdf
7 MB
Moving load analysis Midas Process.pdf
1 MB