Professional Documents
Culture Documents
State of The Industry DAM Full
State of The Industry DAM Full
Disclaimer
Do you know what your software is up to? A closer look at the new stability analysis
and design requirements in the 21st century.
Although CSC Inc. takes great care to ensure that any data, information, advice or
recommendations it may give either in this publication or elsewhere are accurate, no liability
or responsibility of any kind, including liability for negligence, howsoever and from
whatsoever cause arising from or related to their use, is accepted by CSC Inc., its servants or
agents.
Figure 1: Complex buildings with sloping members, hard-to-define floors, and non-
orthogonal framing are commonplace today. [Screenshot from Fastrak Building
Designer. Model courtesy of Fisher Engineering.]
SUMMARY
The 2005 AISC Specification for Structural Steel Buildings brings significant changes to the
way stability analysis and design is required to be carried out. For the first time, the
specification has something to say about the analysis you and your design software is
performing. Do you know how your favorite software package has implemented the new
requirements? Does it simply meet the minimum code requirements using old methods
intended more for hand calculations or does your software offer you the most accurate,
reliable and flexible tools available?
The first three requirements are covered by the structural analysis and the first two are
included in nearly all analysis packages. Several methods to account for #4 and #5 (including
provisions relating to #3) are presented in the specification. One of these, the Direct Analysis
Method, is given as the most general and most accurate.
AISC developed the Direct Analysis Method (DAM) as a solution to meeting the stability
analysis and design requirements in a modern way that is most suitable for implementation in
analysis and design software. This method is not limited in its application, applies to all
buildings, and is the most general and accurate approach provided. The requirements for the
DAM include:
♦ Second-Order analysis: A second-order analysis which considers both P-∆ and P-δ
effects is required.
♦ Initial imperfections: The effects of initial imperfections of the structure geometry are
considered by applying notional loads, which are lateral loads proportional to the
gravity loads applied at each framing level.
♦ Inelasticity: The axial and flexural stiffnesses of members that contribute to the
stability of the structure are required to be reduced. This is to account for the effects
of residual stresses that lead to inelastic softening before the members reach their
design strength.
♦ Effective Length Factor: And now we come to the best part (at least one of the best
parts), K=1.0! Setting K to be 1.0 can be allowed because the effects for which it was
meant to compensate (initial imperfections and inelasticity) have already been
accounted for in the method.
SECOND-ORDER ANALYSIS
The changes in the AISC Specification are, in part, in recognition that buildings have
changed over the years. Buildings were once (in most cases) straight forward, rectilinear
structures with multiple redundancy and permanent or semi-permanent substantial walls
within the framing. Now, modern buildings are more just a skeleton of steel so that the frame
has to work a lot harder. Even in buildings with regular bays, discrete frames of differing
types with irregular layouts are commonly used to meet architectural requirements, as seen in
the structure in Figure 2. Therefore, stability analysis is more critical and any second-order
effects have to be determined more accurately.
Figure 2: Combinations of moment frames and braced frames with irregular layouts are
commonly used to meet architectural requirements, even in buildings with
regular bays. [Screenshot from Fastrak Building Designer. Training example.]
A second-order analysis is required for all structures that implement the Direct Analysis
Method and can take the form of a rigorous analysis or approximate methods like the
amplified first-order method (the B1, B2 method) – which is really intended more for hand
calculations and other approximations.
The amplified first-order method certainly works well for some buildings, generally those
that are simple, regular structures. However, to allow for the greatest range of structure
geometry with the highest degree of accuracy in determining internal forces, the best course
Fastrak Building Designer implements the Direct Analysis Method fully, using a rigorous
second-order analysis. CSC chose this method because it gives the user access to the most
powerful analysis and design solutions. It allows for the widest range of building structures
with accurate and reliable results. The AISC requirements listed as 1 through 5 at the start of
the document are all met by Fastrak Building Designer.
The second order analysis included in Fastrak Building Designer accounts for both P-Δ and
P-δ effects. The Chen and Lui ‘two-step iterative’ second-order analysis method was chosen
to be included in Fastrak Building Designer for several reasons. The main reason is that
since the geometric (stress) stiffness is used in the method, there are no significant limitations
on its use or applicability for most normal building structures.
CONCLUSIONS
The approach used by the Direct Analysis Method of putting as much as possible into the
analysis without making it impractical (i.e. requiring full FE for the whole building) allows
for accurate determination of required forces that can be compared to accurate strengths as
determined from the AISC specification. The 2005 AISC Specification with the Direct
Analysis Method marks a significant step forward for steel design. Fastrak Building
Designer takes advantage of all these developments (including a rigorous second-order
analysis capturing both P-∆ and P-δ effects) giving the designer increased flexibility,
accuracy and reliability.
Now you know what Fastrak Building Designer does for stability analysis and design. I will
repeat my earlier questions. Do you know how your favorite software package has
implemented the new requirements? Does it take full advantage of the Direct Analysis
Method and the capabilities AISC now allows in the second-order analysis?
INTRODUCTION
The 2005 AISC Specification for Structural Steel Buildings (AISC Specification) brings
significant changes to the way stability analysis and design is required to be carried out. For
the first time, the specification has something to say about the analysis you and your design
software is performing. Do you know how your favorite software package has implemented
the new requirements? Does it simply meet the minimum code requirements using old
methods intended more for hand calculations or does your software offer you the most
accurate, reliable and flexible tools available?
AISC developed the Direct Analysis Method (DAM) as a solution to meeting the stability
analysis and design requirements in the specification in a modern way that is most suitable
for implementation in analysis and design software. This method applies to all structural
steel buildings and there are no limitations as long as the method is performed according to
the specified requirements. Combined with today’s best analysis tools, including a rigorous
second-order P-Delta analysis and a user-friendly graphics interface, Fastrak Building
Designer implements the DAM to give designers the best solutions available. This paper
discusses:
♦ the real effects the new stability requirements are meant to capture,
♦ the methods used prior to the 2005 AISC Specification and the associated problems,
♦ the requirements for stability analysis and design in the 2005 AISC Specification,
including the requirements of the DAM,
♦ stability analysis and design of structural steel buildings with modern software,
♦ and the details of how Fastrak Building Designer implements the AISC requirements
for analysis and the DAM
Structural engineering often requires modeling real structures in the virtual world of
computer software for efficiency in the analysis and design process or because the structure is
too complex to treat by hand. A relatively common structure, such as shown in Figure 4,
would be very difficult to analyze and design by hand. In the real world, structural steel
buildings are not perfectly straight and plumb. They are constructed within the tolerances
specified by the American Institute of Steel Construction (AISC) in the Code of Standard
Practice for Steel Buildings and Bridges. This document allows for some initial
imperfections; out-of-straightness of each member and out-of-plumbness of columns. So, a
real building has initial geometrical imperfections.
Figure 4: A building such as this would be very difficult to analyze and design by hand.
[Screenshot from Fastrak Building Designer. Model courtesy of Fisher
Engineering.]
Loads applied to a structure will act on the ‘real’ shape of the structure, including initial
imperfections (both sidesway of the structure and curvature of the members) and deflections
caused by lateral loads. This introduces flexure into axially loaded members, where the
magnitude of the flexure is dependent on the axial load and the deformed shape. This causes
second-order behavior known as the P-Delta effect. There are two P-Delta effects;
♦ P-∆ is a structure effect in which axial loads in columns act on the displacement of the
ends of the member - the story displacements,
♦ P-δ a member effect in which axial loads act on the displacement between the ends of
the member.
So, a real building experiences second-order effects. The question is, are they significant and
if so, how do I best deal with them?
The process used to create structural steel sections creates residual stresses in the cross
section of each member, often as high as thirty percent of the yield stress (0.3Fy). Due to
these internal stresses, portions of the section will begin to yield when the applied stress due
to loading is as low as 0.7Fy. This is well before the yield strength of the section as a whole
is reached. When steel yields the effective modulus of elasticity, E, of that portion of the
section tends to zero. Therefore, as portions of the section yield, the overall stiffness of the
section is reduced due to inelasticity. So, real steel sections experience ‘early’ inelasticity
and a reduction in effective stiffness.
What do these ‘real’ effects mean to you? Basically, if you allow for them properly then
K=1.0!
The tedious calculation of the effective length factor, which is based upon assumptions that
are rarely satisfied in real structures, is no longer necessary. I will explain this apparent jump
in logic by starting with the state of design before the 2005 AISC Specification, then
discussing the motivation and basic changes in the recent specification, and follow up with
how a new method for stability analysis, the Direct Analysis Method, has eliminated K > 1.0.
BACKGROUND
BEFORE 2005
Before the 2005 AISC Specification, a second order analysis was required by the
specification. One method, the so-called B1, B2 method of approximate second-order
analysis, has been around for at least 15 years in the AISC Specification. Even when second-
order effects were considered, typical analyses neglected the global effects of geometric
imperfections and inelasticity in the members. However, the member design equations did
include the effects of geometric imperfections (out-of-straightness) and inelasticity of the
members. To compensate for neglecting the global effects of initial imperfections and
member inelasticity, an effective length factor, K>1.0, was used to determine the axial
capacity of the member. This approach leads to the following potential problems:
♦ It understates the moments in columns and members that provide rotational restraint
to the column because it neglects the global effects of imperfections and inelasticity.
The use of the effective length factor will likely give an adequate column size, but the
members around the column may have understated moments. The required forces
determined from the analysis for the column base plate or beam-column connections
may be significantly less than they should be.
♦ The K factor on the capacity side only affects the column design. Even with a
second-order analysis, without initial imperfections included, load combinations
without lateral loads will never indicate moments that would be amplified by P-∆
effects.
♦ The effective length factor is tedious and difficult to calculate correctly. The
alignment charts commonly used to determine the K factor are based on nine
assumptions. These idealized conditions rarely exist in real structures. For example,
one condition states that all columns in a story are assumed to buckle simultaneously!
♦ Finally, this method is often overly conservative due to the inaccuracy of calculating
K factors and applying demand effects onto the capacity of the member.
AFTER 2005
The 2005 AISC Specification requires you to consider all the ‘real’ (global and local) effects
previously discussed (second-order effects – P-∆ and P-δ, initial imperfections – global and
member, and inelasticity in members) thus eliminating the problems with the earlier method.
Chapter B requires that the design of members and connections be consistent with the
intended behavior and the assumptions made in the analysis. This is a pretty general
statement. The details of how this affects stability design are covered in the requirements of
Chapter C.
AISC has recognized that analysis options are increasing due to technological advances. In
response, the specification now directly addresses the close interaction of the analysis method
and the design method. However, the specification allows designers to use any rational
method of analysis as long as it is consistent with the design procedures in the specification.
The specification sets constraints that must be met to use the member design chapters and
leaves the rest to the designer (for the most part).
The design of all structures must consider the influence of the following on the stability of
the structure and its elements.
The first three requirements are covered by the structural analysis and the first two are
included in nearly all analysis packages (or neglected as is standard practice e.g. moment
connections are always considered rigid). Three methods to account for #4 and #5 (including
provisions relating to #3) are presented in the specification:
However, any rational method of accounting for the listed effects is permitted.
The DAM is currently in Appendix 7 but will very likely move to Chapter C as the default
method in the 2010 AISC Specification. This method is not limited in its application, applies
to all buildings, and is the most general and accurate approach of the three. The requirements
for the DAM include:
♦ Second-Order analysis: A second-order analysis which considers both P-∆ and P-δ
effects is required. This can be accomplished through a rigorous second-order
analysis or by using the approximate method presented in the specification. Second-
order analysis will be discussed in more detail later. Note that the analysis must be
carried out using load combinations (either ASD or LRFD). Because the level of
second-order effects depends upon the size of the load (the ‘P’) and the amount of
deformation (the ‘Delta’), the load cases cannot be analyzed separately then
superimposed in a load combination. For the same reason, if ASD is used, the
second-order analysis is performed using 1.6 times the ASD load combinations. The
results of the analysis are then divided by 1.6 to get required strengths.
♦ Initial imperfections: The effects of initial imperfections of the structure geometry are
considered by applying notional loads. Notional loads are lateral loads that are
applied at each framing level and are proportional to the gravity loads. The loads are
expressed in terms of the gravity loads applied at each level and assume a standard
out-of-plumbness based on the AISC Code of Standard Practice requirements. (Ni =
0.002Yi). Notional loads are typically applied as minimum lateral loads, which has
the effect that they are usually only applied to gravity load combinations. However,
notional loads are required to be added to all load combinations in certain situations.
The specification also allows the designer to directly model the imperfections for
analysis in place of adding notional loads.
♦ Inelasticity: The axial and flexural stiffnesses of members that contribute to the
stability of the structure are required to be reduced. This is to account for the effects
of residual stresses that lead to inelastic softening before the members reach their
design strength. The flexural stiffness is multiplied by a factor, τb (less than or equal
to one), which is dependent on the value of the axial load in the member. To avoid
the iteration required to calculate this factor, the specification allows the designer to
use a value of τb = 1.0, if the notional loads are increased to Ni = 0.003Yi.
♦ Effective Length Factor: And now we come to the best part (at least one of the best
parts), K=1.0! No more calculating K when determining the nominal strength of
columns. Setting K to be 1.0 can be allowed because the effects for which it was
meant to compensate (initial imperfections and inelasticity) have already been
accounted for in the method. It should be noted that a K value less than one can be
used if it is justified by analysis.
The following methods are also allowed by the specification in some circumstances.
♦ The First-order analysis method is limited in use to where second-order effects are
not excessive (where the ratio of second-order displacements to first-order does not
exceed 1.5) and where axial loads are low (required strength is less than half the yield
strength of the member) for frame members. A first-order analysis is performed on
the nominal geometry with nominal member stiffnesses. An additive notional load is
required as well.
Figure 5: Combinations of moment frames and braced frames with irregular layouts are
commonly used to meet architectural requirements, even in buildings with
regular bays. [Screenshot from Fastrak Building Designer. Training example.]
A major advantage of the Direct Analysis Method (DAM) is that it more accurately captures
the internal forces in the structure than the other methods, particularly for structures that are
susceptible to instability. For example, where there are high gravity loads and low lateral
loads. This is especially true when a rigorous second-order analysis is used in combination
with the DAM.
Neither of the alternative methods offers any real advantage over the DAM. The effective
length method trades the stiffness reduction calculations for K calculations. Therefore, the
method still suffers from the difficulty of calculating K correctly. The first-order analysis
method requires large notional loads and the determination of a B1 factor that is applied to
ALL moments, not just those due to lateral loads – so it is likely to be conservative. In any
case, if a sophisticated software package is doing all the work, the DAM will give more
accurate analyses and more efficient designs with little extra effort for the designer.
SECOND-ORDER ANALYSIS
A second-order analysis is required for all structures that implement the Direct Analysis
Method and can take the form of a rigorous analysis or approximate methods like the
amplified first-order method (the B1, B2 method) – which is really intended more for hand
calculations and other approximations.
Choosing a rigorous second-order analysis frees the analysis from the limitations of the
amplified first-order method. These limitations include (but are not limited to) the following
as discussed in the AISC Commentary on Chapter C:
♦ Complex geometry such as sloping beams and columns can create difficulties for
applying the approximate method. Similarly, when floor levels are not readily
identifiable or vary throughout the building, the approximate method is difficult to
apply and may be inaccurate.
The amplified first-order method certainly works well for some buildings, generally those
that are simple, regular structures. However, to allow for the greatest range of structure
geometry with the highest degree of accuracy in determining internal forces, the best course
of action is to take advantage of a rigorous second-order analysis. Even smaller projects, with
relatively small budgets, often have complex framing requirements that will benefit from a
rigorous second-order analysis. This type of complexity is demonstrated in the structure in
Figure 6.
Fastrak Building Designer implements the Direct Analysis Method fully, using a rigorous
second-order analysis. CSC chose this method because it gives the user access to the most
powerful analysis and design solutions. It allows for the widest range of building structures
with accurate and reliable results. The AISC requirements are implemented as described in
the Table 1.
The notional loads are calculated automatically as 0.003 times the total factored gravity load
supported by the column at the beam-column intersection. The 0.003 factor is used in place
of 0.002 because τb is set to 1.0. (See the earlier discussion of member inelasticity for more
information.) This avoids the need to iterate in order to determine τb (you need the axial load
in the members to determine τb, which affects the stiffness, which in turn affects the load
distribution in the structure and the axial load in a member). This also allows different
materials to be used in members contributing to the stability of the structure. In Fastrak
Building Designer, members of all materials have reduced stiffnesses in the stability analysis.
In Fastrak Building Designer, the notional loads are automatically calculated and can be
applied in the global X, global Y or both directions simultaneously. The designer can apply a
factor to adjust the value in the load combination if so desired. The notional loads are
calculated for each load combination based on the total factored gravity loads in that
combination. The notional loads are additive in any load combination in which they are
included. The designer has complete control over the load combinations in Fastrak which, in
turn, controls when the notional loads are applied and the magnitude of the loads in each
combination.
Live load reductions are different for each member in the structure. For example, in one story
of a column the reduction might be 30% while in the story below it might be 40%. Hence, it
is not possible to apply a reduced live load with a single value into the analysis.
Consequently, live load reduction tends to be handled as part of the design (of members) and
not part of the analysis.
In the DAM a further complication is encountered. The size of the axial force (‘P’) influences
the level of P-Δ (and P-δ) effects and so (even if it were possible) it is not correct to reduce
the live loads prior to carrying out the second-order analysis.
In Fastrak Building Designer, the analysis is run with the full live load; therefore any second-
order effects include the effects of the full live load on stability. The axial loads and
moments from the analysis are therefore based on the unreduced live load. During the design
process, the axial live load affect on the column is reduced by the appropriate factor and used
as the required axial force for design. This approach is conservative in that the full live load
is used for stability, but the advantage of reducing the axial load is still captured during the
design process.
SECOND-ORDER ANALYSIS
When the design process is completed, Fastrak Building Designer allows the designer to
quickly and easily view the various bits of information that will help verify the solution.
Fastrak automates the calculations of all the necessary parameters, but allows the user to
view the items and, where appropriate, control them. In this manner, Fastrak relieves the
designer of the labor intensive parts of the DAM aids understanding but still provides the
flexibility to make changes and take responsibility for the design.
CONCLUSIONS
Buildings have changed over the years from being (in most cases) straight forward,
rectilinear structures with multiple redundancy and permanent or semi-permanent substantial
walls within the framing. Now, modern buildings are more just a skeleton of steel so that the
frame has to work a lot harder and any second-order effects have to be more accurately
determined. Figure 7 illustrates such an irregular building.
Older versions of the AISC Specifications were aimed at allowing for second-order effects
using hand methods (most notably the B1, B2 method). These hand methods are not
conducive to software implementation; nearly 35 years of experience producing structural
analysis and design software at CSC has told us this. Like it or not, most design is carried
out on computers these days, if for no other reason than efficiency. AISC recognized this and
in the 2005 Specification has developed a more accurate rigorous method that lends itself to
computer implementation, the Direct Analysis Method. In addition, the rigorous second-
order analysis lends itself best to computer implementation. In recognition that the Direct
Analysis Method is the superior method, the 2010 AISC Specification is likely to move it to
Chapter C as the default method and move the others to appendices. And let us not forget,
the Direct Analysis Method gives us K = 1.0!
Now you know what Fastrak Building Designer does for stability analysis and design. I will
repeat my earlier questions. Do you know how your favorite software package has
implemented the new requirements? Does it take full advantage of the Direct Analysis
Method and the capabilities AISC now allows in the second-order analysis?
Figure 7: Today’s buildings (for the most part) are no longer straight-forward, rectilinear
structures with multiple redundancy, therefore they require more accurate
second-order analyses. [Screenshot from Fastrak Building Designer. Model
courtesy of Robinson Group.]