Capacity Design for Nominally Ductile Structures (NZS3404) 1

[moyseh]

Hi, my question is: are the connections for nominal ductile structures required to be checked for overstrength. I.e. are you supposed to multiply the demands on our connection by Phi_oms?

Following on from this, i have seen engineers scale the loads onto proprietary tension bracing with a ductility of 1. Is this a form of capacity design?

Many thanks

 

[Agent666]

Strictly speaking for nominally ductile structures, yes you should be checking for overstrength in accordance with chapter 12, often though your design actions will cap out at the upper limit for design actions. For a nominally ductile structure that will be mu=1.0 and Sp=0.9 as per CL12.3.3.4.

Often these light bracing elements such as proprietary system (Reidbrace/Donobrace) elements are oversized from a strength perspective, in those cases often they would have sufficient strength for mu=1.0 and Sp=1.0 actions. So, you should be designing these elements and the rest of the system for those actions unless there is some ductile mechanism to limit the actions to a lower value.

Basically, if you have enough strength for mu=1.0 Sp=0.9 actions, design the entire system for this as a CATEGORY 4 system, then for connections 12.9.2.2(2) requires the use of Sp=1.0 for the connections.

Also, if you're designing for CATEGORY 1/2/3 systems you may be able to be limited by the upper limit design actions from 12.9.2.2(4), note though that 12.9.2.2(4)(c) requires the use of an equivalent overstrength factor of 1.25 on the member capacity for the design of connections if you satisfy those requirements. This seems to be poorly understood in industry, I've seen people capping it out at 1.25 x the design action which is not correct.

Often you will have bolts, fillet welds, partial penetration butt welds, etc in the load path, so you will fall into this category if you have a CATEGORY 3 system or greater. This clause is saying in those cases then you're designing for 1.25 times the entire member capacity.

In a way this is like saying if you didn't have to do a capacity design for whatever reason, we want your connections to be utilising the worst overstrength factor from CL12.2.8 just to make sure the connections are not the critical link in the load path, they still want your member to be the element that fails first under larger than code events. Because keep in mind ultimately NZS3404 precludes any brittle failure modes being present for members or connections (CL12.2.2.2.2).

https://engineervsheep.com