I'm no expert on seismic design. There isn't a big call for it in New England.
If I recollect from my steel design courses back in the day, tall building are less likely to suffer a collapse than short structures because the tall structure responds more like a radio antenna on a car. It moves and absorbs the energy without suffering failures at the "rigid" moment connections.
Moment connections are designed to act as a plastic hinge when fully loaded, that is the stress exceeds the yield strength and connection permanently deforms, thereby absorbing energy and redistributes the applied load. In the case of a tall building the duration of the earthquake is relatively short, so the building sways, each connection absorbs some energy, the loads result in some strain in the connection, but not enough to exceed the yield strength of the welded connection. In other words, the movement is tolerated without causing any one joint to go into the plastic mode where the deformation is permanent.
In the case of a short building, the connections tend to be more rigid. The short burst of energy must be absorbed by a relative few rigid connections. The amount of strain per joint is higher than in the case of a tall building so the strain causes plastic flow and the deformation is permanent. The design assumption was that the welded joint had sufficient ductility to accommodate the strain. In many case the assumption proved to be false. The joints were too rigid to absorb the strain so something had to fail. A weld that has insufficient ductility and toughness when loaded rapidly, i.e., low toughness, is going to fail.
Early design assumptions did not consider toughness to be an issue. WPSs, when qualified, typically are not tested for toughness unless the weld is subject to loading at low temperature. This isn't the case for structural steel framing that is encased by a temperature controlled the building envelop. Buildings are not typically loaded rapidly, i.e., the governing loads are typically static, i.e., dead loads. The loads change slowly over time, so impacts testing of the raw materials or the weld filler metal are not typically required.
It is the engineer's responsibility to analyze the load conditions a structure will be exposed to. Codes are used as the basis of the engineer's design assumptions. Early codes typically didn't include a lot of information on seismic activity or how a structure responds to earthquakes. As a result, the designs were based on inadequate code requirements that were in effect when the structure was designed and constructed. Apart of the design process, the engineer is responsible to determine the design of the welded connections and the type of filler metal required to provide the properties needed, i.e., tensile strength, yield strength, elongation, and when necessary, toughness. Early designs did not specify toughness requirements as part of the design.
The electrode/filler metal is a key component when designing a connection. The connection is not going to function properly if the engineer does specify the filler requirements properly. It was and still is a problem. Few engineering curriculums adequately address the proper selection of welding processes, weld design, or filler metal selection.
I need to ask the following question: "How many of you, whether you are an inspector or fabricator, see structural drawings where the structural engineers specifies the filler metal as an E70XX?" What does it mean? Is the fabricator bound to using only SMAW with any 70 ksi filler metal? Is it the engineer's intent that any welding process can be used as long as 70 ksi filler metal is used? Is it the engineer's intent that the filler metal must produce low hydrogen deposits? Can any low hydrogen welding process be used? Are the welds required to meet any impact properly limitations? The a lot of questions that should be asked by the detailer and the fabricator, but typically everyone makes an assumption, right or wrong, and proceeds to order material and fabricate the steel, each operating on the premise ignorance is bliss.
Back to the structural damages observed after the Northridge Earthquake; few designers considered the need to include notch toughness as part of their design requirements. No toughness requirements were imposed on the structural steel used and no toughness requirements specified for the filler metals used. It was then and still is today; you get what you pay for. Not all filler metals are created equal. Some filler metal classifications are required to meet toughness requirements and many more have no toughness requirements. Some filler metals must be capable of meeting low hydrogen requirements and many more are not required to meet low hydrogen requirements. If you don’t specify the properties required you aren’t going to get the right “stuff.”
If the designer fails to select a base metal that is required to meet notch toughness requirements and if the designer does not specify low hydrogen weld deposits with notch toughness, why should anyone be surprised to see connections fail when the connection is subjected to impact loads? If the connection is designed with insufficient flexibility when subjected to loads a magnitude or two higher than the assumed loads used by the designer, why should anyone be surprised to see connections fail?
My thoughts are on the subject are:
The connections that failed during the Northridge quake were destined to fail for several of the following reasons, any one of which could have been sufficient to cause a failure.
1) The connections were improperly designed for active seismic areas.
2) The materials of construction were not appropriate if impact toughness was not specified.
3) The filler metals were not specified to meet low hydrogen requirements and they were not specified to meet notch toughness requirements. (I know that FCAW is typically considered to be a low hydrogen welding process, but not all FCAW electrodes are created equal.) Many early FCAW electrodes had the potential of producing as much diffusible hydrogen in the welds a a rutile covered SMAW electrode)
4) The filler metals were only specified to meet minimum tensile strength requirements, i.e., 70 ksi. The filler metal of choice was E70T-4, a self-shielded flux cored electrode that is not required to meet notch toughness requirements.
5) Consistent with the AWS structural welding code, welds are only required to pass visual examination. Per AWS D1.1, volumetric examination, i.e., UT or RT, is only required for certain connections classified as subject to fatigue and tensile loads. Many of the welds that failed were not tested by UT and many failed welds would not have passed visual examination to D1.1 requirements. One has to ask, "Was it typical practice to require full visual inspection by qualified third party inspectors?" I suspect many were not inspected by anyone.
AWS D1.1 is supplemented by AWS D1.8 for designs that are required to meet the seismic requirements of recently adopted building codes. This is a move in the right direction. However, Owners trying to save money, contractors trying to prop up profit margins, and contractors that are ignorant to the requirements will always be around to muck up the works. These are the same people that will hire (or not) the low cost laboratories that are rubber stamp mills that will sign off on anything as long as the invoice is paid.
Rant over. Well not quite. In my opinion, Lincoln got hosed. However, they made the best business move possible. It was strategic, it was beautiful. They agreed to develop a program to "teach" engineers what they needed to know to properly design connections for seismic applications. They used the terms the court settlement to their advantage and it was probably one of the most successful marketing tools ever.
Now my rant is complete.
Best regards - Al
D1.8 adopted some of the welder qualification testing from the FEMA docs. Basically it is the same welding test as in D1.1, but with some added plates in the way to simulate welding through a weld access(rat) hole.
