Beruflich Dokumente
Kultur Dokumente
Lateral Displacement, ∆
Fig. 1. Load-displacement history.
500
200
Second-Order Elastic
Analysis
100
20.0 ft
0
0 4 8 12 16 20
Fig. 2. Frame for comparison of analysis results. Fig. 3. Comparison of load/lateral displacement results
for frame of Fig. 2.
∆ P
M1 M1 H
Pδ
δ
L
M2 M2 HL + P∆
P HL P∆
P
(a) (b)
π 2 EI
Pcr = (3) The frame from Figure 2 will again be considered, this
( K exact L)2
time without the lateral load. An elastic buckling analysis
using GTSTRUDL yields a critical buckling load, Pcr = 232 kips.
Thus, the K-factor is simply a mathematical adjustment For this critical load, Equation 5 yields Kexact = 2.66. Since
to the perfect column equation to try to predict the capacity the GTSTRUDL analysis includes flexural, axial, and
of an actual column. Every method or equation that is pro- shearing deformations while the nomograph solution
posed for the determination of the K-factor or effective includes only flexural deformations, a more accurate com-
length factor is simply trying to accurately predict the actual parison would be expected if axial and shearing deforma-
column capacity as a function of the perfect column. tions were excluded from the elastic buckling analysis. In
Perhaps the most commonly used approach for the deter- this case, Pcr = 237 kips and Kexact = 2.63. The nomograph
mination of K-factors is the nomograph found in the com- equation also gives K = 2.63. Since the structure of Figure 2
mentary to both the LRFD and ASD Specifications (AISC, and the elastic buckling analysis without axial and shearing
1999; AISC 1989). The equation upon which the sidesway deformations satisfy the assumptions of Equation 4, it is not
permitted nomograph is based is given in Equation 4 surprising to find that the effective length factors are the
(Galambos, 1968). same. The total buckling load for this frame is 474 kips, the
sum of the two column buckling loads.
Top &
1 Yes 1145 1.20 0.85 1145 1.20 0.85 1038 1.26 0.89
Bottom
Top &
2 No 111 3.85 2.72 125 3.63 2.57 122 3.67 2.60
Bottom
3 Top No 209 2.81 2.81 237 2.64 2.64 232 2.66 2.66
MBA Q∆
B D L
A
C
Ma Mb MAB
Pb Q
Pa P Q
Fig. 7. Symmetric frame with leaning column. Fig. 8. Restraining and leaning columns.
P+Q
A C K n = Ko (10)
P
P∆+Q∆
P Q P Q
If column AB from Figure 9a were designed to carry the
(a) (b) load P using the effective length factor Kn, it would provide
sufficient lateral restraint to permit column CD to be
Fig. 9. Equilibrium forces for Yura derivation. designed to carry the load Q using K = 1.0.
P+Q
Ki = Ko (19) Pei ∆
P Ki' = Σ Pu oh (24)
0.822 Pui Σ HL
which is the same as the equation that resulted from the
modified Yura and Lim and McNamara approaches, Equa- where
tions 10 and 12 respectively.
π 2 EI i
For the story-stiffness model, the stiffness reduction due Pei =
to axial load is included as though all columns were can- L2i
tilevers with a buckled shape in the form of a half sine
curve, thus CL = 0.216. Since the leaning columns have no and
lateral stiffness of their own, CL = 0 for all leaning columns.
The equation given in this paper as Equation 13 is just one Pui = Pi
form of the effective length factor equations given by
LeMessurier. Another form is also available through the These simplifications may not be necessary since, in the
same derivation (LeMessurier, 1977). This equation uses original form, the equations presented by LeMessurier are
the ratio of lateral displacement to lateral load as a measure not much more complex and will yield more accurate
of buckling stiffness. Equation 46d from the original paper, results.
in the notation of this paper, is given as
EXAMPLES
I π2 E ∆
Ki2 = i 3 oh (Σ PT + Σ CL PT ) (20) The following examples will show how these approaches
Pi L ΣH
may be used to evaluate columns in unbraced frames.
where ΣH is the total lateral load supported by the level
Example 1
under consideration, ∆oh is the corresponding lateral dis-
placement of the level and ΣPT = ΣP + ΣQ is the total load The frame shown in Figure 10, introduced by Geschwind-
on the given story. In order to account for CL = 0 on the ner (1995), will be used to compare the simplified methods
leaning columns and CL = 0.216 on all others, the load on for determining effective length factors with an elastic
the leaning columns must be subtracted from the total load buckling analysis. The frame is supported in such a way that
on the story so that (ΣPT + ΣCLPT) = (ΣPT + 0.216 (ΣPT - in-plane behavior will be critical. The columns AB and CD
ΣQ)). Making this substitution and factoring out ΣPT yields as well as the beam BC are W12×136. The other members
are of such a size that their individual characteristics will
Ii π 2 E ∆ oh ΣQ not control. The results of a GTSTRUDL buckling analysis
Ki2 = Σ PT 1.216 − 0.216 Σ P (21) and three simplified equations are presented in Table 2.
Pi L3 ΣH T
Comparisons will be made for the analysis including flex-
ural deformations only. For equal loads on columns AB and
This equation was somewhat simplified in the Commen- CD and no loads on the other columns, Pcr = 1302 kips.
tary to the 1993 LRFD Specification (AISC, 1993) as Using Equation 5 this is equivalent to K = 2.18.
When equal loads are also applied to columns EF, GH,
Ii π 2 E ∆ 1 and JK, GTSTRUDL yields Pcr = 568 kips or K = 3.29. The
Ki2 = Σ PT oh (22)
Pi L 3
ΣH 0.85 + 0.15 Σ Q/ Σ PT
loading on the structure shows two equal loads on the second-order elastic analyses are performed and, along with
restraining columns and three equal loads on the leaning the results from Example 1, a check on column CD, with Fy
columns. This gives P = 2 and Q = 3. From the modified = 50 ksi, is performed.
nomograph equation, Equation 6, K = 3.29 and Pcr = 569 Using the results from the first-order elastic analysis, for
kips. Since the assumptions made in the derivation of Equa- column CD, Pu = 240 kips and Mu = 200 kip-ft. From
tion 6 are satisfied with this model, it is expected that Equa- Example 1, the LeMessurier analysis including the leaning
tion 6 and the buckling analysis will yield the same results. columns yields, K = 3.30. Using the LRFD Specification
The modified Yura equation, Equation 12, yields K = 3.43 (AISC, 1999), the column behaves elastically and its capac-
and Pcr = 524 and the LeMessurier equation, Equation 13, ity is φPn = 421 kips. Thus, Pu /φ Pn = 240/421 = 0.57 > 0.2
using G = 100,000 to represent the pin end, yields K = 3.30 so LRFD Equation H1-1a must be satisfied.
and Pcr = 566 kips. Since the column moment is from a first order analysis,
No matter what approach is taken to account for the lean- it must be amplified. This will be accomplished using the
ing columns, it is clear that leaning columns have a signifi- first suggested equation for B2, Equation C1-4.
cant impact on the stability of the structure. It is also 1
evident, from earlier discussion, that a second-order elastic B2 =
∆
analysis for this frame will yield the same forces for mem- 1 − ΣPu oh
bers AB, CD, and BC, whether there are loads on the lean- ΣHL
ing columns or not. This is true since, through a first order The non-sway analysis generates no moments so the sec-
analysis; there will be no lateral displacement of the frame. ond-order moment becomes, Mu = B2 (200). The results of
Thus, more than a second-order elastic analysis for the the first-order analysis give a lateral deflection due to the 20
given loads is needed to complete design of the structure. kip load of 1.74 in. Using these values, B2 = 1.57 so that Mu =
1.57(200) = 314 kip-ft. The interaction equation becomes
Example 2
240 8 314
+ = 0.57 + 0.35 = 0.92
421 9 803
Factored loads, including a lateral load, are now applied to
the frame of Example 1, as shown in Figure 11. First- and
Since this is less than 1.0, the column will be adequate.
P P P P P 200 k 200 k 200 k 200 k 200 k
20 k
B C E G J B C E G J
20 ft.
A D F H K 10 k 10 k
A F H K
D
200 k
10 ft. 20 ft. 10 ft. 10 ft. 160 k 240 k 200 k 200 k
Fig. 10. Frame for Example 1 with leaning columns. Fig. 11. Example 2 frame with gravity and lateral
load and reactions from a first order analysis.
Column Ii Pui Eq. 4 Eq. 6 Eq. 13 Eq. 20 Eq. 17 Eq. 21 Eq. 22 Eq. 23 Elastic
Mark Buckling
4
i in kips Ko Effective length, Ki
1 425 150 1.81 3.49 3.21 3.29 3.35 3.33 3.30 3.57 3.30
2 350 50 1.73 3.30 5.04 5.17 5.27 5.23 5.19 5.61 5.19
3 475 275 1.74 3.33 2.50 2.57 2.62 2.60 2.58 2.79 2.58
4 350 25 1.72 3.27 7.13 7.31 7.45 7.39 7.34 7.94 7.33
5 350 125 1.78 3.43 3.19 3.27 3.33 3.30 3.28 3.55 3.28
Ii Ii Ii Ii Ii Ii Ii Ii
Ki2 = story constant × 3.63 3.81 3.96 3.90 3.85 4.50 3.85
Pi Pi Pi Pi Pi Pi Pi Pi
W12x45
W12x45
W12x53
W12x58
15 ft
20 ft 25 ft 15 ft 20 ft
1 2 3 4 5