Investigate an Innovative Connection between

RC Beam and Steel Column

Ahmed H. El-Masry, Mohamed A. Dabaon, Tarek F. El-Shafiey, Abd El-Hakim A. Khalil

 steel beams to concrete filled tube (CFT) columns. The results

Abstract— An experimental study was performed to investigate the indicated that the capability of this detail to develop the full
behavior and strength of proposed technique to connect reinforced plastic flexural capacity of the beam can be achieved when
concrete (RC) beam to steel or composite columns. This approach the strong column–weak beam criterion is followed.
can practically be used in several types of building construction. In Seismic performance of the concrete filled U-shaped steel
this technique, the main beam of the frame consists of a transfer part
beam (TSC beam)–RC column connection has been studied by
(part of beam; Tr.P) and a common reinforcement concrete beam.
The transfer part of the beam is connected to the column, whereas the
HG. Park, Hwang, CH. Lee, CH. Park and CN. Lee [6] and
rest of the beam is connected to the transfer part from each side. Four special detail using diagonal re-bars and welded re-bar
full-scale beam-column connections were tested under static loading. connections was used to strength the beam-column joint.
The test parameters were the length of the transfer part and the Chen, Lin and Tsai [8] elucidated the cyclic behavior of
column properties. The test results show that using of the transfer part connection between a steel beam and a welded box column.
technique leads to modify the deformation capabilities for the RC The results indicated that brittle fracture occurs at the beam
beam and hence it increases its resistance against failure. Increase in flange complete joint penetration weld and in the weld access
length of the transfer part did not necessarily indicate an enhanced whole region, because the stresses are concentrated in these
behavior. The test results contribute to the characterization of the
regions.
connection behavior between RC beam - steel column and can be
used to calibrate numerical models for the simulation of this type of The study in this research investigates a proposal to keep
connection. the failure location inside the structure and far from joint and
columns area. The proposed technique for connecting bare
Keywords— composite column, reinforced concrete beam, Steel steel or composite column to the RC beam would try to
Column, Transfer Part. achieve two targets; first one is to avoid the collapse at the
joint or the column and second one is to propose an easy
I. INTRODUCTION innovative construction system (CAAP) that has capabilities to
ransferring the load from beam to the column by a safe give high performance without a significant increase in cost.
T way is considered one of the critical issues which have
been studied by many researchers in many fields. The
Fig. 1 gives a general idea about CAAP construction system.

failure of the connection between the column and the beam is

one of the major reasons which causes structures failure and
has a great effect on the acceleration of collapse. It can be
stated that conventional building construction depends mainly
(at two type of materials; concrete and steel which enter as a
basic compound) on the most three common types of
structural elements: reinforced concrete, steel and composite
elements. Many researchers have investigated different types
of connection between beam and column. Their techniques to
connect beam to column changed according to the material,
method of construction and the expected loads. Parra- Fig. 1 Proposed Technique
Montesinos, Dasgupta and Goel [2] presented and developed a
new connection design that would allow using of FRC- II. EXPERIMENTAL INVESTIGATION
encased steel truss members in earthquake-resistance RC
framed construction. Elremaily, Azizinamini [3,4] presented A. Test Specimens
and developed an economical connection detail for connecting The specimens were designed to represent an exterior
beam-column connection. Each column element represents a
A. H. El-Masry is with Faculty of Engineering, Tanta University, Egypt half-story column in a building, and each beam element
(00971-50-9200802; fax: +971-4-3558444; e-mail: ahany79@ yahoo.com).
M. A. Dabaon, is a professor in steel structures with Faculty of
represents a part of full beam length up to contra-flexural
Engineering, Tanta University, Egypt. (e-mail: [email protected]). point. Four specimens (SP1, SP2, SP3 and SP4) were built
T. F. El-Shafey is , is a professor in concrete structures with Faculty of with the same dimension as shown in Fig. (2). SP1 is a control
Engineering, Tanta University, Egypt. ([email protected]).
beam whereas the reinforcement bars are connected directly to
A. A. Khalil, is a professor in concrete structures with Faculty of
Engineering, Tanta University, Egypt. ([email protected]). the column without transfer part. SP2, SP3 and SP4 use a
transfer part to transfer the load from the common reinforced 50 MPa was obtained. Steel bars are made of high tensile steel
concrete beam to the column. The geometry, dimensions, and (360/520), stirrups of normal steel 240/350, steel beams and
reinforcement detailing of the test specimens are depicted in columns material are ST37.
Fig. 2. The net height of all columns is 2220mm each. The
beam length is 1200mm.Test program includes four specimen
frames; all of them have a composite beam connected to a
hinged-to-hinged column. The concrete cross section of all
beams is (160mm) in width and (320mm) in depth. Overall
length of each beam is (1200mm). The reinforcement of all
beams is same; running along the span, there are four bars at
the top and four bars at the bottom, each with a diameter of
12mm. For shear reinforcement, 8mm diameter stirrups are
often placed 100mm apart along the entire length of each
beam.

Fig. 3 Typical connection detail

C. Test setup and loading pattern
Fig. 4 illustrates the test setup that was employed,
indicating the ideal support and loading conditions. The
column was supported by the top and bottom hinge. Static
load was applied vertically at the end of the beam.

Fig. 2 All specimens details

All steel beams (IPE 240) connected to the column (HEB

160) in the same way; the steel beam is welded to an end plate
which is connected to the column by using bolts. This part of
the beam is considered a transfer part; the remaining part is
reinforced concrete one with top and bottom reinforcement. It
is important to state that the reinforcement covers the whole
span of the beam including the transfer part. All specimens
have the same connection details which are shown in Fig. 3.
B. Materials Fig. 4 Testing frame setup
All specimens were cast on the same day and with the same
concrete mixture. Compressive tests on concrete cubic D. Instrumentation
samples (measuring 0.15 x 0.15 x 0.15 m), cast together with The displacement is measured by using LVDT's for both the
the specimens, were conducted to determine the concrete column and the beam. LVDT's for all specimens assigned at
compressive strength. A mean compressive strength equal to the same location. Strain gauges of 5mm length are used to
measure steel strain that are installed at several locations while
concrete strains are recorded by Bi-Shape installed at two decrease in the deflection value. RC beam has a very low
points; the first one at a distance of 50mm from the steel end loading capacity and sudden failure can be noticed. Increasing
plate and the second one at a distance of 350mm from the the length of transfer part from 30% at SP2 to 70% in SP3
column face. Both of the Bi-Shapes are at the same horizontal raised the ultimate load by 9.7% and had a minor effect on the
line parallel to the bottom steel reinforcement. maximum deflection values while in SP4 keeping the transfer
part length at 30% of whole span and changing the type of
III. EXPERIMENTAL RESULTS AND DISCUSSION column to a composite column instead of bare steel column
Results and discussions of the experimental work are recorded a ultimate load (23.8% less than SP2) and opposite
illustrated below for all specimens. Tables I and II shows a deflection values (62.3% less than SP2) but the great
summary of the straining actions at the maximum loads for all advantage with SP4 is that it keeps the failure location at the
specimens. beam side far from column joint and not at the column side as
TABLE I : has occurred with SP2 and SP3. SP2, SP3 and SP4 achieved a
ULTIMATE LOAD, BEAM CORRESPONDING DEFLECTION AND STRAIN good and better loading capacity than SP1.
Ɛs Of Ɛc Ɛc
ST Beam P kN θ3
Ultimate P
Maximum Beam
Name Load
Deflection (mm) θ2
(kN) At At At
θ1 δ1
350mm 0.0mm 350mm δ2
SP1 39.67 18.52 -------- 0.00166 0.0 θ3
SP2 109.5 52.95 0.000682 0.00448 0.00002
SP3 121.28 52.81 0.000742 0.00356 0.00102
SP4 83.43 19.91 0.0000941 0.00016 0.00088

TABLE II :
ULTIMATE LOAD, COLUMN CORRESPONDING DEFLECTION AND STRAIN
Upper Chord Lower Chord
Ultimate
Maximum Maximum Maximum Maximum
Load
Name Deflection Strain Deflection Strain
(kN)
(mm) (mm)
δ1 mm
SP1 39.67 3.06 0.000063 3.84 0.0000345
SP2 109.5 6.74 0.000435 1.256 0.0021470

SP3 121.28 3.15 0.000156 5.74 0.0039007 Fig. 5 Load-deflection relationship

SP4 83.43 1.60 N.A 0.41 N.A
Deflection of the beam has been measured at two points;
first one at the end of the beam, second one at 350mm
A. Load-Deflection Relationship measured from column flange. Comparison of all specimens’
Fig.5 shows the load-deflection curve for the all specimens' deformation line at the maximum load of SP1 is shown in Fig.
beams; it is noticeable that all specimens have a gradual 6. It is observed that the use of the proposed technique at SP2,
deformation increasing under the increase of loading. Both of SP3 and SP4 decreases the deflection values along the beam.
SP2 and SP3 scored high vertical displacement values, the Using 70% of the length of the whole span as the length of the
highest of them was 52.95mm and 52.81mm respectively, transfer part at SP3 results in a 23% decrease in deflection
which are almost the same under the effect of the Ultimate value as compared to SP2 where the length of the transfer part
load of each one. The maximum deflection value of SP1 and is only 30% of the entire length of the beam. Although same
SP4 is 18.42mm and 19.91mm respectively under the Ultimate specifications and dimensions of transfer part are used in SP2
load of each one. It is necessary to mention that the maximum and SP4, recorded deflection values of SP4 are the least of all
load of SP4 is higher than the ultimate load of SP1 by 2.1 specimens because of the use of the composite column instead
times. It can be said that the use of the proposed technique as of bare steel column at this specimen.
in SP4 can double the Ultimate load without a significant B. Load-Rotation Relationship
increase at the deflection value.
As shown in Fig.5 SP1, SP2 and SP3 have approximately Results of three rotation angles which are obtained through
the same behavior up to 20kN because all of them have the deflection values are studied for having more knowledge
same column properties, the changes in behavior which are about the specimens’ behavior. These angles are named θ1, θ2
observed later is the result of variation in the stiffness of the and θ3. θ2 is measured at 0.0mm and θ1 at 350mm, measured
beams. SP4 shows a higher stiffness than the other samples from column flange.
because of the use of composite column instead of bare steel
column.
It can be clearly shown that the effect of increasing the
length of Tr.P is followed by increase in the ultimate load and
 δ2 δ1 continuity of the same transfer part section for about 70% of
the whole span.
SP2 shows a big decrease in θ2/θ1 ratio; the maximum
recorded ratio was 1.49 at 40% of the specimen’s ultimate
load (0.4P2u) and the lowest one was 1.17 at the specimen’s
ultimate load (P2u) which means that at the time which the
load increase by 60% from 40% to 100%, the ratio of θ2/θ1
decrease by around 22%. The lowest ratio of θ2/θ1 recorded for
θ2 δ2 δ1 SP4 was 1.07 at its ultimate load (P4u). It can be said that the
θ1
use of the transfer part to transfer the load from RC beam to
the column functioned successfully to work within the whole
beam as one unit.

θ3 rad P3u
δ mm Beam Length in mm P2u

Fig. 6 Beams Deformation Line for All Specimens at the

Ultimate load of SP1.
0.9P2u
θ3 0.9P3u
Fig. 7 shows that all specimens have more stiffness than P
SP1. SP4 has the highest initial stiffness but once the loading θ2
θ1 δ1
is recorded 47.5% of the SP4’ Ultimate load (39.64kN), its δ2

stiffness is considered almost identical with the SP3 stiffness θ3 0.8P3u

0.8P2u
up to 90% of the SP4’ maximum load (75 kN). This behavior
gives a general idea about the mutual influence of both of
beam and column components of the full system. 0.6P3u 0.6P2u
P4u 0.8P4u 0.4P3u 0.4P2u
0.9P4u
P kN 0.6P4u
θ3 0.2P2u
P
0.2P3u 0.4P4u
θ2 0.2P4u θ2/θ1
δ2
θ1 δ1

θ3
Fig. 8 Relation between the Beams Rotation Angles and the Column
angle

IV. FAILURE MODE AND ANALYTICAL DISCUSSION

The failure criteria of any

element can be projected according
to the location of the plastic hinge. P
Mend Mtr
In this study, plastic hinge have
Ф rad two expected location; (1) first one Tr.p Rc Beam

is before the transfer part in the 2 1

reinforced concrete area (RC)
Fig. 7 Load- Rotation relationships beam and (2) second one is located
It is noticeable that the relation between the angle at the at the joint where the beam section
joint (θ2) and the angle at 30% of the entire length (θ1) which is a composite one, see Fig. 9.
defined by θ2/θ1 decreases incrementally after 40% of the Fig. 9 Models static system
ultimate load of each specimen and continue at decreasing up
It has to be stated that the stress ability of the beams
to the failure. Column angle (θ3) of SP2, SP3 and SP4 record
sections inside a frame is not the only factor for keeping the
an increase up to failure, see Fig. 8. The ratio of θ2/θ1 for all
performance at its peak. The column is one of the highly
specimens is limited between 1.0 and 1.5; the lowest ratios are
affected parameters which have a great effect on the behavior
recorded for SP4. SP2 and SP3 show approximately the same
of the full frame. All frame elements have to have the
value of θ2/θ1 at 80% of the Ultimate load of each one;
capability not only to carry the applied load but also to transfer
afterwards this ratio keeps decreasing up to the Ultimate load
it in a safe way.
for each one. SP3 is the specimen which has the lowest
response in changing the ratio of θ2/θ1 because of the
 The nominal moment of beam, both as the reinforcement B. Failure Modes
concrete part and the composite part is provided in the next There wasn’t an occurrence of sudden failure in any
paragraph. specimen. All specimens show flexural cracks along the beam
A. Sections capability before failure. SP2 and SP3 have the same failure mode while
both of SP1 and SP4 have a different mode of failure. The
It was adopted by ACI in 1956 from the rules of equilibrium
reason for the SP1 failure is the splitting of the steel bars as
that Compression force (C) must be equal to tension force (T).
shown in Fig. 10. The split occurred at the moment value
Concrete stress of (𝐶 = 0.85𝑓𝑐′ ) had been defined in ACI
which was 15% less than the nominal value of the moment.
section 10.2.7.1. Mn using the equivalent rectangle is obtained
Had the nuts been longer, the split could have been avoided.
from Fig. 10 as follows:

𝐶=𝑇
0.85𝑓𝑐 ′ 𝑎 𝑏 = 𝐴𝑠 𝐹𝑦
𝐴𝑠 𝐹𝑦
𝑎= = 39.05𝑚𝑚
0.85𝑓𝑐 ′ 𝑏
′
𝑓𝑐 = 0.8 ∗ 50 = 40𝑁/𝑚𝑚2
𝑏 = 160𝑚𝑚 𝑑 = 286𝑚𝑚 Fig.10 Beams RC Section
𝑎
𝑓𝑐′
𝑀𝑛 = 0.85 ∗ ∗ 𝑏 ∗ 𝑎 ∗ (𝑑 − ) = 56.60𝑘𝑁. 𝑚
2
A chance of slipping between
Fig.10 Cracks in SP1 Beam and Splitting of the Steel Bars.
steel bars and steel section is
not being considered in this The existing of the transfer part by 30% of the whole span as
scenario of fully encased steel in SP2 arise the capacity of the reinforced concrete section,
section and that is why balanced that the ultimate moment of the used concrete section is
equation 𝐶 = 𝑇 is provided to 56.60kN.m whiled the applied one before failure was
calculate the nominal moment, 82.12kN.m and no failure is occurred at the RC section or the
see Fig.11. Fig.11 Beams composite Section composite section which appeared a moment by a value of
𝐶 = 0.85𝑓𝑐 ′ 𝑎 𝑏 131.40kN.m without failure.
𝑇 = 𝐴𝑠 𝐹𝑦𝑠 + 𝐴𝑓 𝐹𝑦𝑎 + 𝐴𝑤 𝐹𝑦𝑎
𝐴𝑠 𝐹𝑦𝑠 = 452 ∗ 470
𝐴𝑓 𝐹𝑦𝑎 = 120 ∗ 9.8 ∗ 370
𝐴𝑤 𝐹𝑦𝑎 = (270 − 𝑎)𝑡𝑤 ∗ 370
From the equilibrium equation 𝐶 = 𝑇 → 𝑎 = 163.8𝑚𝑚
𝑎 270−𝑎
𝑀𝑛 = 0.85 ∗ 𝑓𝑐 ′ ∗ 𝑏 ∗ 𝑎 ∗ (𝑑 − ) − 𝐴𝑤 . 𝐹𝑦 ( + 𝑡𝑤 +
2 2
𝑅𝑓𝑡 𝑆𝑡𝑒𝑒𝑙 𝐷𝑖𝑎 𝑡𝑤 𝑅𝑓𝑡 𝑆𝑡𝑒𝑒𝑙 𝐷𝑖𝑎
) − 𝐴𝑓 . 𝐹𝑦 ( + ) = 162kN.m
2 2 2
As shown above the nominal moment of the reinforced
concrete section is less than the nominal moment of the
composite section by more than 50%.
Fig.11 SP2; Beam flexural cracks and column deformed shape.
Elastic and plastic moment of the bare steel column is
calculated below where the Elastic Modulus (Zx) equal to The ability of the column to rotate under the effect of
312cm3 and the Plastic Modulus (Sx) equal to 354 cm3. increasing applied load prevented formation of a plastic hinge
𝑀𝑦 = 𝐹𝑦 . 𝑍𝑥 = 370 ∗ 312000 ∗ 10−6 = 115.44 𝑘𝑁. 𝑚 at the transfer area between reinforced concrete and composite
𝑀𝑝 = 𝐹𝑦 . 𝑆𝑥 = 370 ∗ 354000 ∗ 10−6 = 130.980 𝑘𝑁. 𝑚 beam. If the column did not have an ability to rotate, failure
As per calculation; the highest ability to carry moment is for would have occurred at the transfer joint. It is distinctly shown
the composite section while bare steel column is not able to that the composition of beam succeeded in transferring the
carry the same amount of moment without going through load to the column. In other words, the over lapping length of
plastic deformation. Table III shows moment and transfer steel reinforcement along the steel beam and confinement it by
moment values for all tested specimens. stirrups enabled the two parts of the beam to work as a one
unit; see Fig. 11 for beam flexural cracks and column plastic
TABLE III :
MAXIMUM MOMENT AT THE END AND AT THE TRANSFER POINT deformed. Same behavior of SP2 is repeated at SP3 that the
Name Mend (kN.m) MTr (kN.m) plastic deformation of the SP3 column was the major defect.
SP1 47.60 -------- Flexural cracks which appeared along the beam were minor as
SP2 131.40 82.125 although the column strain value is recorded 3900microstrain
SP3 145.53 30.32
SP4 100.12 62.57 at the maximum applied load, see Fig 12.
 It has to be mentioned that the extra loading on specimens steel column prevented the failure of the column and kept it
SP2 and SP3 did not show mentionable response. The far from joint. It must be mentioned that although exaggerated
maximum applied moment at the SP3 specimen was increase of the transfer part could increase the failure load but
145.53kN.m which was less than the ultimate capacity of the by doing that, the failure mode may change to shear failure or
composite section but higher than the plastic moment of the the failure may shift at the column side.
column.
V. CONCLUSION
The main conclusions of this investigation can be
summarized as follows:
1. Use of the transfer beam (Tr.P) is considered a new
promising technique to decrease the beam required
depth and modifies the beam behavior.
2. Use of of Tr.P increases the beam capacity with
Strain Value = 3900 minimum steel requirement.
at the Failure Load
3. The existence of Tr.P helps to avoid the sudden
collapse of the structure.
4. Replacing bare steel column by a composite column
give the advantage of shifting the collapse from the
Fig.12 SP3; Beam flexural cracks and column deformed shape. column.
Column is deformed at both SP2 and SP3 in a way which 5. Increasing the stiffness of the column either as a
shows that the stresses on column exceeded the elastic limit composite or bare steel is recommended for avoiding
and it is running at the plastic stage. Once the load was column buckling.
released, column didn’t back to its original shape. As
mentioned before, Mp of column is equal to 130.98kN.m while REFERENCES
the applied moment at the columns of SP2 and SP3 is 131.40 [1] A. H. El-Masry, “Connection between Composite
and 145.53kN.m respectively. Undoubtedly, failure of both Columns and RC Beams in Multi-Story Building,”
samples happened because the applied moments on their Ph.D. thesis under preparation.
columns was higher than their capacity in elastic stage. [2] G. J. Parra-Montesions, P. Dasgupta, S. C. Goel
“Development of Connections between Hybrid steel
truss–FRC Beams and RC Columns for Precast
Earthquake-Resistant Framed Construction,”
Transfer Area
Engineering Structures, June 2005.
[3] A.- Elremaily, A. Azizinamini “Experimental Behavior of
Steel beam to CFT Column Connections,” Journal of
Constructional Steel Research, June 2001.
[4] A. Elremaily, A. Azizinamini “Design Provisions for
Connections between Steel Beams and Concrete Filled
Tube Columns,” Journal of Constructional Steel
Research, April 2001.
[5] Eurocode 3.,‘‘Design of Steel Structures’’part 1.8 Design
of joints CEN.2005.
[6] HG. Park, HJ. Hwang, CH. Lee, CH. Park and CN. Lee,
Fig.13 Cracks at the failure of SP4.
“Cyclic loading test for concrete-filled U-shaped steel
beam–RC column connections,” Journal of Engineering
SP4 failure, being “flexural failure”, is the ideal one where the
Structures, January 2012.
failure occurred at the beam side in the transfer area as shown
[7] M. A. Dabaon, M. H. El-Boghdadi, E. A. El-Kasaby and
in Fig. 13. Moment at failure was 62.57kN.m at the transfer
N. N. Gerges, “Early Prediction of Intimal Stiffness of
area which was higher than the nominal moment of the beam
Composite Joints,” ASCE-EGS, III Regional Conference
reinforcement at this point “56.60kN.m” by 9.4%. It is worth
on Civil Engineering Technology, April 2002.
mentioning that the use of composite column section has
[8] M. A. Dabaon, “Behavior and Strength of Partially
decreased the ability of column to rotate which can be
Encased Beams With Built-up Sections Having Thin Steel
considered the main reason for the failure at the beam side. As
Webs,” Ain-Shams University, Faculty of Engineering
shown in Fig. 13, the major crack which started from top to
Scientific Bulletin, Vol.37, No.3, September 2002.
bottom of the beam is occurred at the end of steel beam and all
[9] C. Cheng-Chih, L. Chun-Chou and T. Chia-Liang,
other cracks were minor until the appearance of the major
“Evaluation of reinforced connections between steel
crack. All other cracks were a natural result of increasing
beams and box columns,” Journal of Engineering
tension force on the top reinforcement of the beam. Changing
Structures, June 2004.
the column type to a partially encased column instead of bare

View publication stats