AS3600 - Column or wall??

[SydStructEng]

My post relates to the analysis of a vertical load bearing element under AS3600 (and beyond).

To my knowledge, AS3600 2009 does not define (for strength purposes) when a thin, long column 'becomes' a wall. This leave some ambiguity when analysing a vertical load bearing element of this shape.

For example:
- Take a vertical element 180mm x 720mm in cross section, with an effective height of 2400mm. The element is part of a braced frame and is loaded concentrically. The element is made up of 40MPa concrete, with reinforcement yet to be determined.

The designer appears to have two options;
1) Design the section as a wall - using clause 11.5.1 we end up with a capacity of about 1500kN (assuming min. eccentricity). This is independent of reinforcement. The designer adds some central mesh and that's it. The 'wall' takes 150 tonnes of load.

2) Design the section as a column - calculating the buckling load, slenderness and moment curvature interaction diagram and complying with the reinforcement detailing requirements for columns. For a column with say 1% reo (significantly more than the 'wall' above) the designer gets approx 500kN.

Finally - for those who tell me options 2) is correct; I'd like some back up. I know for a fact that there are numerous designs being certified and building today that use the 'walls' to support some big loads, and I love to argue the case!

Thanks,

 

[rowingengineer]

I don't get the same values as you for the wall or the column, the column I get higher and the wall I get significantly less. Have you include the extra eccentricity in e_a? What you are describing was normally used to get around the fire requirements, but this has been largely stopped due to the latest code changes.

http://www.nceng.com.au/

"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning."

 

[SydStructEng]

Thanks for the reply.

How does the latest code stop a designer from doing this?

 

[rowingengineer]

Cl 5.6.2

http://www.nceng.com.au/

"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning."

 

[asixth]

You can't ignore slenderness because your calling it a wall. With an effective height of 2400mm and 180mm thickness gives you a Le/r=45 which certainly requires the engineer to consider the slenderness, particularly if you are looking to detail it as a column as well and omit the confinement tie requirements.

 

[SydStructEng]

rowingengineer, Cl 5.6.2 is a fire requirement. Consider a vertical load bearing element with no FRL requirement.

asixth, I agree! But AS3600 is unclear.

If you were reviewing a design with this element designed as a wall, what would be your argument when push came to shove?

Thanks for the replies thus far.

 

[asixth]

My argument would be that it's slender and needs to be designed accordingly.

 

[asixth]

If you look in this thread I posted a graph of what I believe the decrease in strength should look like with increasing Le/r thread507-325661.

I think the wall provisions He^2/(2500*tw) underestimates this. And a member such as 720x180 in my opinion needs to be designed as a column based on the dimensions alone.

 

[IDS]

I work on infrastructure projects (using AS 5100 or AS 3600) and I'm not familiar with standard practice in the building area. I always use the column provisions for any member with significant axial load, but this thread led me to look at the wall provisions more closely, and this is what I found.

[ul]
[li]For the column/wall given in the OP, assuming pinned connections top and bottom, minimum eccentricity, 100% dead load, and M*1/M*2 ratio of -1, I get a column capacity(phi.Nu) of 1461 kN and wall capacity of 1489 kN. (1% reinforcement in the column, 0.15% in the wall)[/li]
[li]Applying the same method used by asixth in his last post, I get the stresses at maximum design load shown in the attached graph. For low slenderness ratios the column has a substantially higher design load, but for high slenderness the wall has a much higher design load. Given the lower reinforcement requirement, this doesn't make sense.[/li]
[li]I can find nothing in the code to indicate that the wall provisions are not applicable in this case, of for any wall with an effective height/thickness ratio less than or equal to 30, and low eccentricity applied load.[/li]
[/ul]

In summary, either the column provisions are over-conservative, or the wall provisions are dangerously un-conservative (or both). Either way, a revision of the code seems to be necessary. I will continue to use the column rules, but if presented with a column similar to the one described in the OP, designed to the wall provisions, assuming fire requirements were not applicable, I can't see any way in which it would be outside the code requirements.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[IDS]

There was an error in my previous post. The column capacity should be 1320 kN. The linked graph is now corrected.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[rowingengineer]

IDS,
To compare wouldn't the M1/M2 be 0 as I would assume the wall is designed without bending moment? doing this would allow you to increase the le/r possibly?

http://www.nceng.com.au/

"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning."

 

[IDS]

To compare wouldn't the M1/M2 be 0 as I would assume the wall is designed without bending moment? doing this would allow you to increase the le/r possibly?

No, the column provisions are quite specific that M1/M2 is -1 for minimum eccentricity.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[rowingengineer]

You are correct, the code does state this, but I doubt this is the intent. If I were to accept this interpretation I would be taking this condition as implying that a column is stronger with a bending moment than without, instead I read this condition as trying to show a column in single is less stable than a column in double curvature. The condition makes no mention of M1, I believe this to be the basic difference between the wall and column design component. If M1 = 0 than I believe M1/M2 should equal 0 without reference to M2, hence for comparisons I think we should adopt M1/M2=0 allowing us to treat this as a braced column for higher Le/r ratios.

http://www.nceng.com.au/

"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning."

 

[IDS]

My understanding is that the minimum moment is 0.05DN* anyway, so a pinned column with nominally concentric loading would have equal moment top and bottom, and hence M1/M2 would be -1. They could have said that both M1* and M2* shall not be less than 0.05DN*.

But it doesn't make a huge difference. I have added a line to the graph with M1/M2 = 0.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[IDS]

Prompted by this thread, I have added a MaxAx() function to my RC Design Functions spreadsheet, which will calculate the moment magnification factor for a specified effective length and axial load, or the maximum axial load for a specified effective length and eccentricity.

More details and download at:

http://newtonexcelbach.wordpress.co...sign-functions-4-umom-and-new-maxax-function/

I have regenerated asixth's chart using this function:

I get slightly higher stresses for Le/r, maybe because of 500 MPa steel, or possibly some other difference in assumptions.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/