Analysis of Steel Columns for Air-Blast Loads

By Jessica Godinho, M.S., Arturo Montalva, P.E., Sharon Gallant, P.E, S.E.

Structural Performance
discussion of performance issues relative to extreme events
Structural vulnerability assessment of mid- to high- rise com- meets a surface in line-of-sight of the explosion, it is reflected
mercial buildings is a common request of building owners and amplified by a factor of up to thirteen. These pressures
concerned about potential vehicle weapon attacks on nearby decay exponentially with time, and their duration is typically
streets or within underground parking garages. For these tall measured in milliseconds.
buildings, the structural columns carry substantial gravity loads,
®
Axial Load Effects

E
making it important to include the effect of axial load in the blast
analysis. Current design practice often overlooks instability During an explosion, a column directly exposed to air-blast
due to axial loading on steel columns when analyzing for air- pressures undergoes flexure and corresponding lateral deflec-

R
blast load; instead, conservative performance criteria is set by tion. Explosive pressures are many times greater than conven-
limiting calculated flexural response to relatively small rotations tional loads, therefore it is expected that structural elements

U
and ductilities. Additionally, the distribution of air-blast pres- will experience large deflections and be loaded beyond their
sure along the height of the column is often approximated as yield strength. Axial forces amplify this lateral deflection and
a uniform load by assuming the formation of a plastic hinge atighinternal t moment due to P-δ effects. As the deflection increas-

T
mid-height of the column. This simplification, in addition p yto
r es, the column will reach its plastic limit, transitioning from
Co
neglecting axial load, may result in significant miscalculation of a gradual stiffness and strength degradation to a rapid loss of

C
the structural performance for columns with large axial loads. strength due to buckling. This can lead to instability and failure
of the column which may have a devastating effect on
1
H = 12.5 ft - KL/r = 38 the overall structural integrity of the building, making

e
U
0.9
it important to consider axial load in air-blast analysis of

n
structural columns.

i
0.8
The extent of axial load effects on a steel beam-column

R
subjected to blast loading is highly dependent on the

z
0.7
Axial Load Ratio (Pc/Po)

H = 15.0 ft - KL/r = 45
geometry and support conditions, as represented by the

T a
0.6
column’s slenderness ratio (KL/r). A study using non-
0.5
linear dynamic analysis of steel beam-columns of varying

g
H = 17.5 ft - KL/r = 53

S
0.4 lengths and sections for a uniform blast load provides

a
0.3
H = 20.0 ft - KL/r = 60 the following results:
H = 25.0 ft - KL/r = 75
• For slenderness smaller than 38, axial effects are
0.2
negligible.

m
H = 30.0 ft - KL/r = 90
H = 40.0 ft - KL/r = 121
0.1 H = 50.0 ft - KL/r = 151 • Columns with slenderness ratios between 38 and
0 75 are acceptable according to current standards,
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
as they are below the ductility limit of 3 and rota-
Rotation (deg)
tion limit of 3 degrees when axial load is neglected.
Figure 1a: Axial load ratio vs. rotation for uniform air-blast loading. However, when axial load is included in the analy-
sis, these columns become unstable.
H = 12.5 ft - KL/r = 38
1 • Columns with slenderness ratios over 75 exceed the
0.9 ductility limit of 3 and rotation limit of 3 degrees,
0.8
even when axial load is ignored. These columns are
thus not acceptable under current standards.
0.7
These results are represented graphically in Figures 1a & 1b.
Axial Load Ratio (Pc/Po)

H = 15.0 ft - KL/r = 45
0.6
continued on next page
0.5
H = 17.5 ft - KL/r = 53
0.4 Pc = VARIES
H = 20.0 ft - KL/r = 60
0.3
H = 25.0 ft - KL/r = 75
PLASTIC
0.2 HINGE
H = 50.0 ft - KL/r = 151

0.1
H/2

0
AIR-BLAST LOAD

0 1 2 3 4 5 6 7 8 9 10
(UNIFORM)

Ductility (M = Dmax/De)
VARIES

PLASTIC
HINGE

Figure 1b: Axial load ratio vs. ductility for uniform air-blast loading.

Air Blast Effects

H/2

An explosion is an extremely rapid release of energy in the PLASTIC

form of light, heat, sound, and a shock wave. The shock wave HINGE

consists of highly compressed air traveling radially outward VARIES

from the source at supersonic velocities. As the shock wave

expands, pressures reduce rapidly with distance, and when it Figure 2: Column subjected to uniform air-blast loading.

STRUCTURE magazine 13 November 2007

 Combined Effects of Axial Load 1
H = 10.0 ft - KL/r = 30

and Load-Distribution
H = 12.5 ft - KL/r = 38
0.9

In addition to axial load effects, column response can be influenced by 0.8 H = 15.0 ft - KL/r = 45

the assumed shape of the air-blast loading. Current industry practice 0.7
often approximates the distribution of air pressure along the height as

Axial Load Ratio (Pc/Po)

H = 17.5 ft - KL/r = 53
a uniform load with the formation of a plastic hinge at mid-height of 0.6

the column, as assumed in the analysis presented above and shown in 0.5

Figure 2 (see page 13). Although this is appropriate for far-field effects,
®
H = 20.0 ft - KL/r = 60

E
0.4
this simplification is inaccurate for columns located in close proximity H = 30.0 ft - KL/r = 90
0.3
to the explosion. H = 40.0 ft - KL/r = 121
Near-field effects occur when columns are subjected to a close-

R
0.2
H = 50.0 ft - KL/r = 151
in explosion causing a non-uniform pressure distribution along the 0.1
height of the column. This non-uniform distribution of loading

U
0
typically results in a plastic hinge located below mid-height of the 0 1 2 3 4 5 6 7 8 9
Rotation (deg)
column, as shown in Figure 3. Analysis results indicate that shear
t
demand in the column increases as the plastic hinge shifts towards
righ

T
Figure 4a: Axial load ratio vs. rotation for non-uniform air-blast loading.
the bottom of the column where the blast load is concentrated; p y
Co 1
H = 10.0 ft - KL/r = 30

therefore, the assumption of a plastic hinge formation at mid-height H = 12.5 ft - KL/r = 38

C
may substantially underestimate the shear demand in the column. 0.9

Under this circumstance, it is likely that the column will be a shear 0.8 H = 15.0 ft - KL/r = 45

critical element rather than flexural. This effect, in combination with

Axial Load Ratio (Pc/Po)

e
U
0.7
neglecting axial load, can result in inadequately designed columns 0.6 H = 17.5 ft - KL/r = 53

n
with large axial loads subjected to close-in explosions, with analysis

i
0.5
indicating that ductility may be underestimated by as much as 11%

R
H = 20.0 ft - KL/r = 60
H = 40.0 ft - KL/r = 121

z
and support rotation by as much as 21%. 0.4
H = 30.0 ft - KL/r = 90

T
0.3

a
PLASTIC 0.2
HINGE

g
H = 50.0 ft - KL/r = 151
0.1

S a
0
0 1 2 3 4 5 6 7 8
Ductility (M = Dmax/De)
(NON-UNIFORM)
AIR-BLAST LOAD

m
Figure 4b: Axial load ratio vs. ductility for non-uniform air-blast loading
HEIGHT = 16 FT.

Conclusion
PLASTIC Current design practice often overlooks instability due to axial
HINGE
loading on steel columns when analyzing for air-blast load; instead,
CALCULATE

what is thought to be conservative performance criteria is set by

PLASTIC
HINGE
VARIES
Figure 3: Column subjected to non-uniform air-blast loading.
Combined effects of load distribution and axial load were studied by
examining column response to close-in explosions. Column response
was determined using a non-uniform load distribution where the
expected plastic hinge location was determined by analysis. The
following results were found and are represented graphically in Figures
4a & 4b.
• For slenderness smaller than 38, axial effects are negligible.
• For slenderness between 38 and 90, the effects of axial load are
significant and should be accounted for in the analysis and design
of steel columns due to blast.
• Columns with slenderness ratios over 90 exceed ductility and
rotation limits of current standards, even when axial load is
ignored. These columns are thus not recommended for use in
blast design.
Note that the buckling effects are reduced when assuming a non-
uniform load due to the shift of the plastic hinge towards the bottom of
the column.
limiting calculated flexural response to relatively small rotations
and ductilities. This approach may result in underestimation of the
structural performance for columns with large gravity loads. Modeling
of air-blast pressure distribution as a uniform load with plastic hinge
at mid-height may also provide inaccurate estimation of column
response. This simplification, combined with neglecting axial load,
may result in miscalculation of column performance under certain
loading conditions, lending to the use of more advanced analysis
that incorporates axial load and non-uniform loading effects.▪
Jessica Godinho, M.S. is a Project Engineer at Hinman Consulting
Engineers, Inc. in San Francisco, CA. She specializes in design
and research of blast resistant structures. and can be reached at
jgodinho@hce.com.
Arturo Montalva, P.E. is a Senior Engineer at Hinman in San
Francisco and specializes in dynamic non-linear response of
structures from impact and explosive loadings. He can be
reached at arturo@hce.com.
Sharon Gallant, P.E., S.E. is a Project Manager at Hinman in
San Francisco with over 15 years of seismic design experience.
Sharon oversees a variety of project types, including blast resistant
new design, seismic peer reviews, and progressive collapse analysis.
She can be reached at sgallant@hce.com.

STRUCTURE magazine 14 November 2007