Use of Codes from Different Countries 1

[Brad805]

We are asked from time to time to design slender concrete column elements that are beyond the moment magnification method. We have been testing some random designs using non-linear finite element analysis and correlated one model to an actual test we found. Our prediction was very close to the test data. For our model we used material properties found in the CEB-fip 2010 model code (MC2010), and would like to do so going forward. We are located in Canada and the CSA code is not very verbose on this topic. ACI is better, but does not have the same material property definitions as the model code. I am interested to know if others would be comfortable using the model code if not referenced in their countries code or would you simply say no when asked for this type of design?

CSA code Reference:

ACI

 

[canwesteng]

Haven't specifically done this for concrete but do it all the time for steel (cantilevers, torsion, etc not covered in S16). Normally I use ACI codes or published textbooks as references though, I'm not familiar with the model code you've described but if your team is satisfied with it seems to fit in the intent of A23.

 

[dik]

I often use other codes for design, but with caution. Some codes have different 'basics' and the direct application may lead to an unsafe condition. That said, I often use a rational analysis for design where there isn't something applicable to S16 (our steel code). A few years back I did a long cantilevered monorail crane using Eurocodes because there wasn't a reasonable approach to loads below the shear centre of the beam.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[sticksandtriangles]

Not directly related to your code substitution question, but it is interesting that your CSA code seems to require the full inelastic second order analysis for concrete columns (geometric + material non-linearity).

When moment magnification yields conservative results, I typically go to ACI 6.7, elastic second-order analysis (I like to call this the geometric non-linear option). Seems to give much better results that the moment magnification procedures.

This is my little cheat sheet I like to reference at times in ACI if you haven't seen it.

For our model we used material properties found in the CEB-fip 2010 model code

Any chance you can post some of the material properties in the model code you reference? I would be interested in seeing what they suggest for concrete and steel.



S&T -

http://www.re-tug.com

 

[Brad805]

I appreciate the comments.

sticksandtriangles, the model code I have is an actual book that I have not scanned, but the Dutch document, "Guidelines for Nonlinear Finite Element Analysis of Concrete Structures," seems largely based on the model code. It includes most of the material definitions, but the model code is more in depth on a number of other topics. The Dutch reference is ---> Dutch Guideline

The earlier snip from the CSA code was missing one section:

 

[TLHS]

I don't know about the specifics of what you're doing, but the general concept seems pretty similar to what has to be done if you're going to do a seismic analysis using post yield behavior in Canada. You're not going to find Canadian code sources for any of that, so you go to industry documents and foreign codes for basic design info.

 

[hokie66]

Slender columns would not be an area where I would consider less conservative altenatives.

 

[dik]

Thumbs up on that Hokie... same with cantilevers.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[Brad805]

I should add a bit of context to the question. I am not talking about pushing the slenderness designs like the image below. In our case this usually comes up with delegated design work where the EOR might have overlooked a detail, and by the time it lands on my desk it causes great discussion to make changes. Most are rectangular or irregular slender wall sections. In these cases I do the designs using common methods, but in cases where one must make simplifying assumptions I have been experimenting with checking the design with NLFEA software we have. To use this software there are endless parameters one must define and it involves using interfaces to define how various materials interact at intersecting planes. The model code is very helpful to define these parameters and gives the modeler some guidance on many model parameters unlike the general clauses one sees in some codes.

I respect and appreciate the opinions.

 

[Enhineyero]

Let me get this straight, the situation you have is
1. My design doesn't work using local codes (which doesn't have certain provisions that other codes have)
2. I will try to justify the "failed" design by applying sophisticated provisions of a similar code (from another country) and apply it to my design.

Well, this actually happened to me. Long story short the insurer of the building wouldn't accept such approach.

If I find myself in the same situation again, I probably would stay conservative and stick to what the local codes allow me to do. Anything you do outside of the local codes can be a cause of contention in court, if ever anything happens to do the building.

 

[Settingsun]

Hokie & Enhineyero, the situation is that Brad805 plans to use the 'main' code method for this design instead of the permitted simplification which just happens to be used 99% of the time because it's simple. However, the main method is the only method permitted by the code for very slender elements, and the code gives no details on how to do it.

My opinion is that the model code can be a starting point. The commentary to the corresponding Australian code provision requires/recommends systematic sensitivity checking.

 

[Brad805]

Enhineyero, no, that is not quite the situation. I have designs that require simplifications to design with normal procedures outlined in our code. These are cases where I would like to look closer at the design to gain a more complete understanding. Both the CSA code and ACI codes reference second order analysis, but are a tad light on describing the how part. We would only be looking for parameters related to the appropriate material models, and general model parameters. The final design would be per our code. I do appreciate your point of view. We had a review engineer for a project recently that shared the same thoughts.

We did calibrate one model to an actual test.
ASCE report:

Our model results:

The accuracy of our model was not correct at eccentricities > 60mm due to a boundary condition error, but it is quite reasonable otherwise. We will correct the boundary condition when time permits.

 

[sticksandtriangles]

Curious, what software are you using to do this analysis?

S&T -

http://www.re-tug.com

 

[Brad805]

We have been using Diana FEA Link. They have been helping us along the way. There are quite a few other options out there. It is not very friendly cost wise, but none of them are. We only license it when we have problems that warrant such studies. We have a few details for one client that have always annoyed me and I want to have a better understanding. As I have been working on problems I have been thinking how to convince others of the analysis methodology. This has been a very helpful discussion to get a sense of what other engineers might think.

Anyone interested in trying this type of software may be interested in Vector Analysis You can download this for free to analyze gravity load problems for limited model sizes. It is not as user friendly as other options, but if you want to check a design you can. Dr. Serhan Guner has a helpful youtube video: Video

 

[Enhineyero]

Sorry I misunderstood your post.

I'm all for advance analysis and design (including testing). However, the issue has always been $$$ (and always will be?). If we are getting paid to do this then yeah its ok, I can spend ages making the perfect model and tweaking the design. But if we have priced jobs such that we have only allowed to design it once, then I probably wouldn't go into deep this level of detailed analysis/design and design it conservatively.

The only time I would consider going this far in the analysis (outside of the budget) is when I stuffed up on a critical design element and failure to prove a working design means the end of my engineering career. I see a lot of big engineering firms hiring Phds just for this.

 

[Brad805]

Enhineyero, no problemo I was looking for opinions. Some of what I am doing is for interest, but we have an engineer that is concerned about our partially composite precast wall panel design for an upcoming project, and I was planning to include calculations from this analysis tool since we can predict the force at each internal connector very well. I will need to include references to the model code, so this discussion has been helpful to gauge how that might be perceived. Our other software package for that problem leads to a number of assumptions to arrive at these forces, and I sense their team is wary.

There is a lot of truth to the cost factor for this software, but the biggest is the entry level cost. As I have worked thru problems I have been finding many of our problems can be modeled in nearly the same time as SAP2000, but our problems are quite a bit different than your typical CIP problems involving beams/columns and slabs. That said, it has taken a lot of my own time to get to that point. The PHd types are very good if the problem is way out there and you have a budget, but we asked about confirming some of our day to day type problems and it would have been very costly considering the number of different configurations of the parts.