Is the Direct Analysis Method permitted under AS 4100?

[Settingsun]

Hi all,
I'm after opinions on whether the Direct Analysis Method of Design (DAM) set out in ANSI/AISC 360-16 (chapter C) is 'in accordance with' AS 4100. This would rely on AS 4100 clause 1.5.1 which in turn refers to section 3.

I think that the boxes in section 3 are ticked by DAM. DAM falls between the elastic and advanced analysis methods set out in AS 4100 which is another aspect in its favour IMO.

The main effect would be that the effective length factor for compression members (ke) is always 1.0 under DAM because the effects that increase ke above this are accounted for in the analysis, resulting in larger M* which compensates for the larger phi.Nc.

 

[Agent666]

I personally wouldn't have any issue with someone using the direct analysis method in NZ or Australia. I'd say its more favourable than someone plucking an effective length out of thin air and using more classical methods inappropriately.

I guess it's something that will probably be in the next version of our steel standards (whenever that occurs). But in the interim these sorts of things fall into the 'current best practice' arena in my view, so where there is an advantage to using something that has been designed for a specific purpose to overcome shortfalls in other procedures, then why aren't we all doing it! I bet its because most engineers probably have not been exposed to it in NZ/AU? This might be a hurdle during any reviews, but education of the reviewer is probably pretty easy as there's a wide body of knowledge available showing the benefits and how to do it.

As long as you follow the AISC guidance verbatim, I don't see any issue really. It's been in the AISC guidelines since mid 2000's I think so it's not like it's anything new and untested that people need to be scared about.

 

[Trenno]

Agent, I've not looked into DAM all that much but have seen it mentioned here and there.

Do you recommend any texts or guides to get a good grounding of the method?

Thanks

 

[Agent666]

Have a look at the AISC YouTube account (and AISC continuing education website), they have had quite a few good videos going into academic and practical uses from beginner intros to advanced uses of DAM. I'll try go through my records and post back with some of the ones I've watched in the past that I found useful.

 

[Settingsun]

Thanks, Agent. Third party review is my main concern (unless this discussion throws up a technical issue) and I doubt I'd use it if subject to external review unless I could agree it with the reviewer up front. I found a background paper to the AISC code that showed slightly greater capacity when DAM was used compared with effective length factors which could prove to be an issue with an unknown reviewer.

Trenno, try searching for 'AISC direct analysis' on this forum. Quite a few topics where others have asked for good articles/ presentations to understand it better. The AISC website has been rearranged so old links don't work but some can still be found through a google search for the title or author.

 

[Agent666]

Not sure how it works with the regulatory regime in Australia, but in NZ we have what is termed accepted solutions (follow these standards cited in our building code) and alternative solutions where you might be using alternative or equivalent means of compliance. Just because a reviewer might not be familiar with an alternative method, isn't a reason to not allow it. I guess you might just need to present/provide additional information in your calculations around the use of the DAM and obviously step through what you did with them and why. I'm sure in Australia there must be a means of using alternative methods (basically I'd assume the clauses you stated will provide this framework).

Also maybe discussing it with ASI might give you some ideas/examples of where it may have been used previously if you were to run into a reviewer who has their head in the sand over adoption of the latest and greatest where standards might be lagging behind.

 

[Agent666]

Trenno, heres a few webinars I found useful on the subject of the DAM, first one is a practical application (where you couldn't practically not use any other method), the others go into the AISC specification requirements and background information to varying degrees.

Diagrid and Mega-Braced Structures -- Stability and Design Approaches

Inelastic Behavior and Design: New Opportunities

Design for Stability Using the 2010 AISC Specification

Direct Analysis Method Applications and Examples

Direct Analysis Method - Now and the Future

 

[Toby43]

To further the conversation..

As I see it (please correct if I'm misinterpreting) the difference between the Direct Analysis Method (DAM) and current AS4100 methodology is really only relevant to moment framed structures (i.e. not braced frames or shearwall structures as Ke=1.0 is appropriate for these structures) and is as follows
-AS4100 requires "compression only" check with Ke not equal to 1.0 (i.e. taken from relevant charts in Ch4) AND "Compression-Bending" interaction with Ke=1.0
-DAM reduces the flexural stiffness by a factor of 0.8, which increases the second-order moments such that Ke=1.0 may be used. No separate "Compression Only" check required.
-Also DAM requires the "Notional" loads (or out-of-plumb geometry) to be applied to only Gravity cases if the 2nd order deflection is less than 1.5x1st order deflection (using reduced stiffness) otherwise they must be included in all load cases (i.e. lateral aswell)

Initially I thought that reduced stiffness was double dipping, as residual stresses are accounted for in derivation of capacity, but thinking it through it makes senses that increased flexibility increases 2nd order effect and reduces capacity of the member at the same time, thus appropriate.
As such I don't see any "technical" issues marrying it with AS4100.

Regards
Toby

 

[Settingsun]

I never thought about double-dipping but I think the issue that requires ke=1.0 check is that DAM doesn't necessarily flush out buckling due to out-of-straight as there's no requirement to model that geometric imperfection. Eg a slender column in a sway frame 'quasi-braced' by stiffer sway columns. I can however see situations where the 0.8*E analysis would give a lower capacity than a traditional analysis so maybe there is some double-dipping.

The comment about residual stresses has me thinking about the difference between AISC and AS4100. AISC has a single column curve that is most similar to our -0.5 curve (according to a paper I read). Would it be best not to use the -1.0 curve with DAM given the Americans weren't calibrating to it?

 

[Agent666]

The different curves are calibrated to different residual stresses based on shape and plate thickness, so I believe you should definitely be using them as it's a refinement as opposed to a one size fits all AISC approach. Other standards have multiple curves in a similar manner to AS/NZS standards.

If you plot the AISC buckling curve it's quite different to the AS/NZS one in some aspects that aren't only due to the differences in fundamental curve I think. I'll find a comparison I did a while back when working through mastan2 learning modules.