Lapuz, A., Baluyut, D., Camina, C. and Oreta, A.W.C. (2008).

“Analysis and Investigation of the Degree of Moment

Rigidity of Steel Beam-to-Column Connections,” 6th Regional Symposium on Infrastructure Development (RSID6),
Jan 12-13, 2009, Bangkok, Thailand

Analysis and Investigation of the Degree of Moment Rigidity

of Steel Beam-To-Column Connections
A.H. Lapuz1, D. S. Baluyot, C. N. Camina and A. W.C. Oreta2
De La Salle University, 2401 TaftAve., Manial, Philippines
1
[email protected],[email protected]

Abstract
The study investigated five types of steel beam-to-column connections. The global rotational
spring stiffness values of these connections were determined through laboratory testing and these values
were used in the semi-rigid structural analysis where a rotational spring was introduced at the ends of the
beams near the connection to the columns. A comparison between the semi-rigid elastic analysis the
conventional rigid elastic analysis shows that moment-rotation and the load-deflection relationships as
predicted using the elastic spring model are close to the experimental results. Therefore the study
concluded that the rotational spring stiffness values obtained experimentally may be used to represent the
degree of moment rigidity of the connections; thus improving the structural model.

I. Introduction
For years in the structural analysis of frames, it has been a conventional practice to make
use of a rigid structural analysis. Although connection behavior may come close to full rigidity,
connections still undergo deformations up to some extent. The adoption of this ideal behavior of
member joints was basically intended to simplify the mathematical model in the structural
analysis. However; the predicted response of the ideal frame structure may result to unrealistic
behavior. Neglecting the actual behavior of frame connections may lead to unreliable and unsafe
predictions of the response of frame structures. Consequently, a simple solution to this
predicament is to do away with the rigid connection assumption by introducing a rotational spring
in the mathematical model to approximate the semi-rigid behavior of the connections. What value
of the rotational spring stiffness, k, to use in the elastic analysis needs to be determined. This
study investigates the use of the rotational spring in the elastic semi-rigid analysis of frames.
Experiments were conducted on various types of steel beam-to-column connections to determine
appropriate values of the rotational spring stiffness.
Test Specimen

Testing
Equipment Deformation Gauge

II. Methodology
Five types of steel beam-to- Load
Cell

column connections as described in

Table 1 were designed following AISC Hydraulic
Jack
(ASD) specifications. [1]. The test
specimens were set-up on the beam-
column testing equipment as shown on
the Figure 1 [2] .
Figure 1 Laboratory Test Set-up

Table 1. Five Types of Beam-Column Connections and their Rotational Spring Stiffness Values
 Lapuz, A., Baluyut, D., Camina, C. and Oreta, A.W.C. (2008). “Analysis and Investigation of the Degree of Moment
Rigidity of Steel Beam-to-Column Connections,” 6th Regional Symposium on Infrastructure Development (RSID6),
Jan 12-13, 2009, Bangkok, Thailand

Welded Beam Column Four-Bolt Extended Welded Flange Plate Bolted Flange Plate
Seated beam Connection
Connection Endplate Connection Connection Connection

Actual Image

The beam is shop welded Flange plates are welded to Flange plates are bolted to
Angle plates secure the
to an endplate which the beam and are also the beam and welded to
The beam is fully welded beam in place by bolts to
Description extends beyond the beam welded to the column flange. the column flange. A web
to the column flange. the beam and column
depth and field bolted to A web cleat is placed to cleat is placed to resist
flanges.
the column flange. resist shear. shear.
Rotational Spring
7,992 – 10,875 kN-m/rad 8,035 kN-m/rad 4,325 kN-m/rad 2,274 kN-m/rad 637 – 654 kN-m/rad
Stiffness, k
Rigidity Parameter,
r
62 – 69 % rigid 63 % rigid 47 % rigid 32 % rigid 20 % rigid
W 10x12 Beam
L = 0.9146 m

Force and deformation readings were recorded

approximately every 5 kN intervals. After the gathering of
required data, the moment-rotation plots were graphed for
each connection and subsequently obtaining the rotational
spring stiffness, k. The rotational spring stiffness, k, values can
be derived from the moment-rotation curves of the connections
which can only be obtained through laboratory testing. The
rotational spring stiffness, k, of the connection can be obtained
by linear regression of the moment-rotation curve until its
intersection to the beam line. The slope of the trend line is the k
value of the connection as shown in the Figure 2 using the beam- Figure 2 . k value determination
line concept [5]. According to the concept by
Marcus, the initial slope indicates the degree k
Q4 ,u 4 of
A
flexibility or rigidity of the connection. From A B

the plot, a beam line can be drawn from pt.1 in

the vertical axis corresponding to a fixed-end Q ,u 2 2

moment of the beam with full rigidity (Φ = 0) Q ,u 1 1

Q2 , u2 to
a point in the horizontal axis corresponding to a
L
free-ended rotation with no restraints (M = 0).
The intersection made by the line to the E I = constant

curved plots describes the elastic behavior of Infinites imal

the connections and k can be obtained.

Figure 3 . A beam with a rotational spring at the left end

The values of the rotational spring stiffness, k, were used to conduct a semi-rigid structural
analyses of the test specimens using a rotational spring. A conventional rigid structural analysis
was performed as well. These analyses were employed to solve for the beam end deflections. The
results of the analyses were compared and presented in graphs superimposed upon the plot of the
force vs. the experimental deflection of the beam end.
In the semi-rigid analysis, a rotational spring with stiffness value is introduced at the left
end of the beam (Figure 3). The beam stiffness matrix k for an element with a spring at the left
end is assembled as shown in Eqn. 1 [3-4], where k A is the rotational spring stiffness for the left
end of the member.
 Lapuz, A., Baluyut, D., Camina, C. and Oreta, A.W.C. (2008). “Analysis and Investigation of the Degree of Moment
Rigidity of Steel Beam-to-Column Connections,” 6th Regional Symposium on Infrastructure Development (RSID6),
Jan 12-13, 2009, Bangkok, Thailand

 12 R1 6 LrA  12 R1 6 LR2 
 4 L2 rA  6 LrA 2 L2 rA 
EI  6 LrA
k 3 Eqn. 1
L  12 R1  6 LrA 12 R1  6 LR2 
 
 6 LR2 2 L2 rA  6 LR2 4 L2 R3 
Where,
EI kAL
R2  2a  rA R3  3a  rA R1  a  rA a rA 
4 EI  k A L 4 EI  k A L
The rotational spring stiffness, kA can have
6
values ranging from zero (which represents a pin
connection) to infinity (which represents a rigid 4

connection). When the rotational spring stiffness is

substituted as infinity, the stiffness matrix will 5

resolve back to the conventional (rigid) structural 1 θ = 90º

analysis stiffness matrix. For the vertical elements, θ = 0º 3 12

3
the conventional sitffness matrix for a frame 1
10
element without springs were used. Shown in Lo c a tio n o f Ro ta tio na l
Figure 4 is the model of the beam-to-column 2 Sp ring w ith Ro ta tio n a l
Sp ring Stiffne ss, k 11
connection used in the anlaysis. 2 θ =
270º
To apply the values of the rotational spring 9
7
stiffness in determining the degree of moment P
rigidity for any given structure, the rigidity
parameter, r, should be used which was dependent 8
upon the connecting member’s length and flexural
stiffness, EI. Figure 4.. Mathematical Model

kL
r 
EI  kL Eqn. 2

The result would range from 0.0 to 1.0, meaning a rigidity parameter of 0.0 results to a
frictionless, pinned connection, whereas a rigidity parameter of 1.0 results to a perfectly rigid
connection. This percentage would represent the degree of moment rigidity of the connection
taking into account the effect of the connecting beam element. Therefore, the values would vary
depending on the connecting beam’s length and cross-section.

III. Results and Discussions

Figure 5 presents the moment-rotation curves derived from the experiment. The steeper
the curve, the more rigid the connection is. The seated beam connection is the most flexible
among the five connections, while the welded beam column connections is the most rigid. From
these curves, the rotational spring stiffness values were obtained using the beam-line concept [5]
 Lapuz, A., Baluyut, D., Camina, C. and Oreta, A.W.C. (2008). “Analysis and Investigation of the Degree of Moment
Rigidity of Steel Beam-to-Column Connections,” 6th Regional Symposium on Infrastructure Development (RSID6),
Jan 12-13, 2009, Bangkok, Thailand

as shown in Table 1. Substituting these stiffness values in the stiffness matrix of a beam element
with rotational spring, the semi-rigid elastic structural analysis was performed using the matrix
stiffness method. A rigid elastic analysis of the beam-column connection was also performed.
The force – deflection values at the free end where the load is applied were obtained for
both the rigid and semi-rigid analysis. A comparison of the results for the five types of
connections are shown in Figure 6. The results show clearly that the beam-column connections do

Figure 5. Moment – Joint Rotation Curves of the Connections

not behave as rigid connections. A semi-rigid analysis using the rotational spring stiffness values
obtained experimentally gives better predictions up to the elastic region, or specially up to the
values of Mr – the allowable bending moment of the beam in the Allowable Stress Design (ASD)
in the AISC [1]. It can be observed from the preceding plots that the theoretical curves follow a
straight line. That is because the structural analysis employed in the study is a linear analysis
which means that it only predicts a linear behavioral response of frames, contrary to the
experimental results wherein the curve starts linear but gradually becomes deteriorates until
failure. Therefore the applicability of the theoretical results or the structural analyses is only
during the linear elastic behavior of steel.
The degree of moment rigidity of each connection can be evaluated using Eqn. 2. The rigidity
parameter depends on the values of the rotational spring constant, length and flexural stiffness EI
of the beam. The rigidity is expressed in percentage as summarized in Table 1. The welded beam
and the four-bolt extended end plate connections are the most rigid with a rigidty of about 60%,
while the seated beam connection is less rigid with 20% rigidity.
IV. Conclusion
The study determined the actual degree of moment rigidity of five steel beam-to-column
connections commonly used in the construction industry in the country. These degrees of moment
rigidity were expressed in the quantity rotational spring stiffness, k. The experimentally obtained
values of k were accurate representations of the connections’ actual rotational behavior under
vertical load within linear elastic conditions. Through experimental investigation, it can be
definitively said that the connections’ actual behavior do not always behave in a completely rigid
manner.
The conventional structural analysis was refined by integrating the rotational spring stiffness
values of the connections into the semi-rigid structural analysis. The semi-rigid elastic structural
 Lapuz, A., Baluyut, D., Camina, C. and Oreta, A.W.C. (2008). “Analysis and Investigation of the Degree of Moment
Rigidity of Steel Beam-to-Column Connections,” 6th Regional Symposium on Infrastructure Development (RSID6),
Jan 12-13, 2009, Bangkok, Thailand

analysis using rotational spring at the end of the beam predicts the semi-rigid behavior of the
connections up to the linear elastic range only. Therefore it can be concluded that semi-rigid
structural analysis using rotational spring model, along with an estimate of the connection’s
rotational spring stiffness, is a better model than the conventional frame with rigid connections.
V. References
[1]American Institute of Steel Construction, Manual of Steel Construction Load Resistance Factor
Design, 3rd Ed., Chicago: AISC, 2001
[2] Brigado, J., Tan, A. and Tan, M., Design and Development of a Static Testing Laboratory Set-
up for Beam-Column Connection, Thesis, De La Salle University, Manila, Philippines
[3] Öztürk, A. U. and Seçer, M. An Investigation for Semi-Rigid Frames by Different Connection
Models. Mathematical and Computational Applications. Association for Scientific Research.
2005. pp 35 – 44, http://www.asr.org.tr/pdf/vol10no1p35.pdf
[4] Kassimali, A., Matrix Analysis of Structures, Brooks/Cole Publishing Co., USA, 1999.

[5] Marcus, S., Basics of Structural Steel Design, Reston Publishing Inc., Reston, Virginia. 1981

(a) Seat Beam Connection (b) Welded Beam Column Connection

(c) Welded Flange Plate Connection (d) Bolted Flange Plate

(e)Four Bolt Extended End Plate Connection

Lapuz, A., Baluyut, D., Camina, C. and Oreta, A.W.C. (2008). “Analysis and Investigation of the Degree of Moment
Rigidity of Steel Beam-to-Column Connections,” 6th Regional Symposium on Infrastructure Development (RSID6),
Jan 12-13, 2009, Bangkok, Thailand

Figure 6. Force vs Deflection at Beam End for the Five Types of Connections