Download Across-Wind Loading on Reinforced Concrete Chimneys of Circular

Ttle no. 91 -S35
Across-Wind Loading on Reinforced Concrete Chimneys
of Circular Cross-Section
$y B. J. Vickety*
MlishPri in December 1993, this report ,from the Boundary Layer Wind
Tunrl Laboratory ($ the Univer.sit)- of Western Ontario present.s a model
formsessing across-w?nd
loads on reinforced concrete chimneys. Because
~bruporrpmvides a basis for proposed changes in the AC1 chimney design
rrondani. this report condenses and summarizes some ofthe report findings.
Atmsswind loading and the induced dynamic response are due primarily
lQOn!nm or less regular shedding of vortices from the chimney. This has
brs lwn recognized as a .significant
design consideration for slender
iiwturo~ with a hluflas opposed to streamlined cross section. Across-wind
Ming is a/so induced by lateral gustiness in natural wind and by lateral
&&Jluctuations
in the wake of an upstream structure such as another
ibuq irk a grouped arrangement. The report deals primarily with loading
kb vorte.x shedding, hut makes some reference to hu#eting in the wake
(on upstream structure.
Wuld loading on a structure is independent of the constructiw material. But significant differences between the
characteristics of concrete and steel chimneys result in subYantial changes in the response to applied loads. Design
aiteria also differ. The prime differences are:
I. The dimensionless mass parameter mlp,D’, which is
i”jpically over 100 for reinforced concrete, but can be as low
LE 40 or less for an unlined steel chimney.
LDamping p which is typically 1 percent for reinforced
mncrete may be as low as 0.2 percent for unlined steel.
3. The aspect or slenderness ratio h = Wd is commonly
much higher for steel than reinforced concrete chimneys,
specially if the concrete chimneys are multiflue types.
4. The more slender reinforced concrete chimneys are
wmmonly tapered while a tapered steel chimney is unusual.
‘lhese differences play roles in determining the response
@across-wind loading and thus in the selection of design
mierk The report describes the roles and presents a detailed
ntathenutiul model of vortex shedding and the induced
qmnse.
There are two essential features of the model:
LForces
acting on the stationary chimney are modeled
sarandom process with a Gaussian distribution having a
lpectmm which is also Gaussian in form. The spectrum is
kfined by its variance, bandwidth, and center frequency (the
Odin; frequency).
Z.Motion-induced effects are modelled by introducing an
litude dependent damping which is negative in the vi-
‘Frully of Engineering Sciences, The University of Western Ontario, London,
.had;l.
Structural Journal / May-June 1994
cinity of the critical speed. In predicting the response of
reinforced concrete chimneys, only the value of an aerodynamic damping parameter in the vicinity of the critical speed
is of consequence. The nonlinear nature of the damping is
of no significance.
Because of uncertainties in relationships presented in the
model, a coefficient of variation of about 20 percent is appropriate. This is a somewhat lower reliability than is normally associated with, for example, drag loads at a given
wind speed and this should be considered when assigning
appropriate load factors.
A detailed presentation of the model in the report appendix
provides all of the information needed to predict responses such
as moments and shears in any mode of vibration of chimneys
with an arbitrary profile. It also permits making an acceptably
accurate approximation for the first mode of response of chimneys having moderate, reasonably uniform taper.
Response predictions are detailed in the report. To illustrate
the dependence of across-wind response on geometry, a series
of ten chimneys was studied. Aspect ratios varied from 8 to
20 and all chimneys were assumed to have minimum wall
thickness of 200 mm (8 in.) over the upper third. The study
demonstrated the nature of across-wind response and its
strong dependence on chimney slenderness.
Results of this report have formed the basis for proposed
revisions to treatment of across-wind response in “Standard
Practice for the Design and Construction of Cast-in-Place Reinforced Concrete Chimneys (AC1 307~SS).“ When this standard
was revised in 1988, design criteria based on working stresses
were changed to criteria based on strength design and factored
loads. These changes introduced an inconsistency.
Drag wind loading and earthquake loading can be logically
treated by using load factors and strength design since this
is essentially a limit state approach. The load factor insures
that the load used has an acceptable low probability of oc-
Publication of this summary paper has been requested by AC1 Committee 307 to provide background information for a proposed major
revision in the way across-wind loads are treated in AC1 307-88.
Readers wanting a copy of the full report should request it from Dr.
A.G. Davenport, Director, Boundary Layer Wind Tunnel Laboratory,
The Faculty of Engineering Science, The University of Western
Ontario, London, Ontario, Canada N6A 5B9, while the supply lasts.
Thereafter, copies will be available from AC1 headquarters (where
the paper will be kept permanently on file), at a charge equal to the
cost of reproduction and handling at the time of request.
355
currence. And the factored load can be associated with an
event having a readily estimated probability of occurrence.
In assessing the behavior under these extreme conditions it
is also logical to assume a damping capacity (for examples,
5 percent of critical) consistent with extensive cracking and
yielding. It is not logical to treat vortex-induced loads in this
manner.
The maximum effects of vortex-induced loads are often developed at wind speeds much less than about the SO-year speed.
Thus the preceding approach is Hawed. Under such circunstances, the structure is expected to remain in good condition
and a limiting stress criterion is a more reasonable design approach. It is also logical to assume a damping level consistent
with this behavior, for example, about 1 percent of critical.
‘l’he report uses tms approacn ro ae~me acceptau~t: XL
levels for vortex-induced loads, based on insuring an aca~~
able fatigue life for the chimney. Because the aim war
define stresses suitable for inclusion in code requiremer
for chimneys, necessary simplifying assumptions lead toat
servative estimates of the acceptable stress levels. Thd(
simplifying assumptions are described in the report.
A solution for the assessment of across-wind loads1
reinforced concrete chimneys is still not complete. Howevc
good quality full-scale data developed over the past IOya
has permitted development of the reasonable predicti!
model described in the referenced report. There is stillaned
for more aerodynamic data for individual chimneys andh
full-scale data related to behavior of grouped chimneys.