Strain Characteristics of Large-Size Non-Studded Concrete-Filled Steel Tube Column under Construction

: Interplay between the outer steel tubes and the core concrete provides an important guarantee for the superior performance of concrete-filled steel tube (CFST) columns compared to that of reinforced concrete columns. In order to optimize the construction of the large-size non-studded CFST columns in an on-going practical project, the strain distributions in the column under construction were measured and tracked, and the law of the strain distributions of the column under construction is revealed. The study finds that the plane section assumption was not applicable to the strain distributions of the column. Nonlinear variations in the strain levels were observed under bi-directional bending of the column. The absence of studs inside the steel tube significantly weakened the interplay of the outer steel tube and the core concrete by means of a significant slip between the concrete and the steel tube at the column corners. The concrete and the steel tube tended to be stressed and deformed independently. Practical suggestions are given for the design and construction of this type of column. The study results are expected to lay a basic and valuable foundation for in-depth investigations on the mechanical behavior of large-size non-studded concrete-filled steel tube columns under construction.


Introduction
Concrete-filled steel tube (CFST) columns refer to the members formed by pouring concrete inside the steel tubes [1,2].Since CFST columns were proposed in 1879, extensive efforts have been made to determine the behavior of this type of column.Hajjar, J.F. [3] conducted research on the three-dimensional strength sections of square and rectangular concrete-filled steel tube columns, obtaining more accurate strength-related correlations for these concrete sections.Ma, L. [4] conducted tests on short CFST columns at an ultraearly age.He proposed a calculation formula for the bearing capacity of the columns and advanced a stress-strain relationship suitable for the core concrete in the columns.Yuan, F. [5] experimentally studied the behavior of square CFST columns with steel spirals under a high axial load ratio.It was found that the steel spirals could improve the behavior by delaying stiffness degradation and increasing ductility.Ayough, P. [6] investigated the structural performance of CFST columns with double inner steel tubes through nonlinear finite element analysis.The results found that the columns possessed an advantageous ductility and strength-to-weight ratio to conventional CFST columns.
Except reinforced concrete, different types of modified concrete could be used in CFST columns, such as lightweight aggregate concrete, high-strength concrete, pultruded FRP composites concrete, and ultra-high-performance concrete (UHPC).Almamoori, A.H.N. [7] • The confining effect provided by the steel tube makes the concrete inside compressed in three directions, thus enhancing the strength of the concrete; • The steel tube acts as a formwork during the concrete curing and is permanently retained as part of the column afterwards, which saves construction costs and shortens construction schedules; • The concrete inside laterally supports the steel tube, thus increasing the stability of the steel walls under compression.
The above-mentioned points significantly improve the mechanical performance of CFST columns, far more than the simple superposition of the two materials [19,20], in terms of a greater load-carrying capacity, reduced section size, and excellent seismic performance [19].Benefitting from these advantages, CFST columns are widely used in engineering practice.
As an effective way to strengthen the interplay between the core concrete and the outer steel tubes in CFST columns, the existence of studs between the two materials plays a critical role in forming the above-mentioned advantages, thus significantly improving the bearing capacity and deformation capacity of the columns [21][22][23][24][25].A large amount of research has been conducted throughout the world.Alemayehu, R.W. [26] studied the bond-slip behavior of CFST mega columns with and without mechanical connectors, including shear studs.Tao, Z. [27,28] investigated the bond behavior between the concrete and the steel tube in CFST.It is found that welding internal rings onto the inner surface of the steel tube is the most effective method to enhance the bond strength, followed by the methods of welding shear studs and using expansive concrete.Ferrotto, M.F.[29] investigated the structural performances of CFST with and without steel studs by finite element modelling (FEM).Ayough, P. [30] evaluated the influence of bolted shear studs on the axial performance of CFDST columns.The test results demonstrated that the bolted shear studs avoided the shear failure of the specimens and significantly increased the ductility of the specimens.Hasan, H.G. [31][32][33] found that embedded shear studs on inner steel tubes were capable of improving the bond strength with a lesser degree of shear resistance in comparison with a normal interface and interface with an internal ring, with reinforcing bars, and with shear studs.Zhang, B. [34] investigated six circular large-scale concrete-filled FRP tubes (CFFTs) consisting of H-steel with shear studs experimentally under a combined axial load and lateral cyclic load to study their seismic behavior.As a novel structural member for new construction, CFFTs were used to retrofit degraded reinforced concrete columns.However, when the concrete is intensively reinforced, the presence of the studs complicates the process of construction and brings difficulties in quality control and quality assurance due to the narrow space between the steel walls and the reinforcement of the concrete [35,36].Therefore, in engineering practice, some designers choose to remove the studs to facilitate the construction.It inevitably weakens the mechanical bite force between the concrete and the steel tube and harms the force transfer mechanism between the two materials [19].For large column sections, the two materials even tend to work separately [37,38].These features go against the fundamental assumption that the concrete and the steel tube work together, leading to the more complex mechanical behavior of the columns than conventional CFST columns.Hence, the current code methods for the mechanical performance of conventional CFST columns are no longer applicable to large-size non-studded CFST columns.However, scarce study has been conducted to investigate the strain distributions and the force transfer between the concrete and the steel tubes.This study is aimed at enriching the efforts in the field to reveal the discrepancy in mechanical behavior between large-size non-studded CFST columns and conventional CFST columns.In this study, the strains of the concrete and the steel tube of an inclined large-size nonstudded CFST column were measured and tracked during its construction.Firstly, the law of the strain distributions of the column section is analyzed and summarized based on the acquired data.Secondly, the relative slip between the concrete and the steel tube is discussed.Lastly, practical suggestions are given for the design and construction of this type of column.The work is expected to lay a valuable foundation for the investigations on the mechanical behavior of large-size non-studded concrete-filled steel tube columns under construction.

Inclined Large-Size Non-Studded Concrete-Filled Steel Tube Column
The strain characteristics of an inclined CFST column (Figure 1) were investigated during its construction in this study.The column was 700 mm × 700 mm in cross-section and 22 m in height.The nominal ultimate compressive stress of the concrete was 20.1 MPa.The nominal yield stress of the steel was 235 MPa.

Layout of Measuring Points on Site
To facilitate a clearer analysis of the column's strains, Section A and Section B in the middle of the middle story were selected to arrange strain gauges, as shown in Figure 2.

Layout of Measuring Points on Site
To facilitate a clearer analysis of the column's strains, Section A and Section B in the middle of the middle story were selected to arrange strain gauges, as shown in Figure 2.  As shown in Figures 3 and 4, eight strain gauges were applied on the exterior surfaces of the column in each section and four embedded gauges were placed at the inner corners in Section B. This arrangement was based on the strain distribution of CFST columns.Strains were linearly distributed across each face of rectangular CFST columns.Therefore, the extreme strains would occur at the corners.Also, through the strains at the corners, the strains at the middle positions could be determined by interpolation.As shown in Figures 3 and 4, eight strain gauges were applied on the exterior surfaces of the column in each section and four embedded gauges were placed at the inner corners in Section B. This arrangement was based on the strain distribution of CFST columns.Strains were linearly distributed across each face of rectangular CFST columns.Therefore, the extreme strains would occur at the corners.Also, through the strains at the corners, the strains at the middle positions could be determined by interpolation.

Strains of Steel Tube 4.1. Strains in Section A
Figure 5 shows the strain distribution in Section A during the construction stage.Note that the strains are represented as multiples of the yield strain of the steel ε y = 114 µε.The positive strains indicate tensile strains, and the negative strains indicate compressive strains.This approach facilitates the comparisons between the measured strains and the yield strain.Absolute strain values of the strains of no less than ε y indicate yielding of the steel, which should not occur during the construction stage.

Strains in Section A
Figure 5 shows the strain distribution in Section A during the construction stage.Note that the strains are represented as multiples of the yield strain of the steel εy = 114 µε.The positive strains indicate tensile strains, and the negative strains indicate compressive strains.This approach facilitates the comparisons between the measured strains and the yield strain.Absolute strain values of the strains of no less than εy indicate yielding of the steel, which should not occur during the construction stage.The measured strain distribution in Section A after concrete pouring is shown in Figure 5a.In this state, it was too early for the concrete to form any strength.The load was resisted by the steel tube and the scaffold.It made the strain level at this stage the largest during the construction stage.Locally large tensile strain up to 0.20εy is observed to occur at JA-d-1, due to the tensile stress caused by the weight of the poured concrete.
The measured strain distribution in Section A before the removal of the scaffold after 28-day concrete pouring is shown in Figure 5b.As the concrete reached its full strength, the concrete worked together with steel tube to sustain the load.With the help of the concrete, the locally large strains just after concrete pouring were partially offset.Most of the cross section is mildly compressed within the magnitude of 0.05εy.
After the removal of the scaffold, the support from the scaffold on the beams next to the column disappeared.The construction load on the beams was transferred to the column in terms of eccentric compression.It led to the transformation of the tensile strains to the compressive strains and an increase in the existing compressive strains in the section, as shown in Figure 5c.Moreover, by comparing Figure 5b and Figure 5c, the change in the strains at JA-a-1, JA-b-1, JA-c-1, and JA-d-1 is observed to be more significant than that at JA-a-2, JA-b-2, JA-c-2, and JA-d-2.It is due to the fact that the moment caused by The measured strain distribution in Section A after concrete pouring is shown in Figure 5a.In this state, it was too early for the concrete to form any strength.The load was resisted by the steel tube and the scaffold.It made the strain level at this stage the largest during the construction stage.Locally large tensile strain up to 0.20ε y is observed to occur at JA-d-1, due to the tensile stress caused by the weight of the poured concrete.
The measured strain distribution in Section A before the removal of the scaffold after 28-day concrete pouring is shown in Figure 5b.As the concrete reached its full strength, the concrete worked together with steel tube to sustain the load.With the help of the concrete, the locally large strains just after concrete pouring were partially offset.Most of the cross section is mildly compressed within the magnitude of 0.05ε y .
After the removal of the scaffold, the support from the scaffold on the beams next to the column disappeared.The construction load on the beams was transferred to the column in terms of eccentric compression.It led to the transformation of the tensile strains to the compressive strains and an increase in the existing compressive strains in the section, as shown in Figure 5c.Moreover, by comparing Figures 5b and 5c, the change in the strains at JA-a-1, JA-b-1, JA-c-1, and JA-d-1 is observed to be more significant than that at JA-a-2, JA-b-2, JA-c-2, and JA-d-2.It is due to the fact that the moment caused by the eccentric compression mainly affected the deformations at the east and west sides of the column.The moment also turned the strain at JA-d-1 to become tensile again after the removal of the scaffold.
During the construction, the largest tensile strain occurred at JA-d-1 after concrete pouring, as much as 0.20ε y .The strain at the point was also at a high level in other states.The largest compressive strain occurred at JA-b-1 after the removal of the scaffold, as much as −0.11ε y .Therefore, the positions near JA-d-1 and JA-b-1 are considered the most vulnerable in the section during the construction stage.

Deformation against Plane Section Assumption in Section A
The three-dimensional plane fitting of the strains in Section A in each state is shown in Figure 6, where the Levenberg-Marquardt algorithm is used in the plane fitting on the basis of the measured strains with the correlation coefficient of 1.0.The measured strains ε m and the desired strains following the plane section assumption ε d are compared in Table 1.
To clarify the deviation from the ideal plane, the deviation rates are visualized in Figure 7, where the deviation rate D is calculated according to Equation (1).

Deformation against Plane Section Assumption in Section A
The three-dimensional plane fitting of the strains in Section A in each state is shown in Figure 6, where the Levenberg-Marquardt algorithm is used in the plane fitting on the basis of the measured strains with the correlation coefficient of 1.0.The measured strains εm and the desired strains following the plane section assumption εd are compared in Table 1.To clarify the deviation from the ideal plane, the deviation rates are visualized in Figure 7, where the deviation rate D is calculated according to Equation (1).In Figure 7, the measured strain is above the desired strain when the deviation rat is a positive value.Conversely, the measured strain is below the desired strain when th deviation rate is a negative value.It could be seen from the figure that some deviation are positive while the others are negative in each state.This distribution agrees with th perfectly fitted plane, although several deviations are much larger than the others.In gen eral, the measured strains significantly deviate from the fitting plane during the construc tion stage, as the deviation rate ranges from −383.11% to 378.17%.It indicates that Sectio A of the steel tube hardly deformed in a plane during the construction.Hence, the plan section assumption is not applicable to the deformation in Section A. The plane sectio assumption is always complied by the present code methods in structural designs an analysis of conventional CFST columns.Consequently, the code methods are no longe applicable to this type of CFST column.The non-compliance with the assumption wi lead to different mechanical behavior of the columns including failure modes and calcu lation of the load-carrying capacity compared to the predictions by the code methods.Thi indicates the practical necessity to reveal the discrepancy in mechanical behavior betwee the studied columns and conventional CFST columns and to develop new methods t predict the behavior of the studied columns.

Strains in Section B
Figure 8 shows the strain distributions in Section B during the construction stage.Th strains are represented in the same way as Section 4.1.As the measuring point JB-d-1 wa damaged before removing the scaffold, the data of this point could not be obtained in th In Figure 7, the measured strain is above the desired strain when the deviation rate is a positive value.Conversely, the measured strain is below the desired strain when the deviation rate is a negative value.It could be seen from the figure that some deviations are positive while the others are negative in each state.This distribution agrees with the perfectly fitted plane, although several deviations are much larger than the others.In general, the measured strains significantly deviate from the fitting plane during the construction stage, as the deviation rate ranges from −383.11% to 378.17%.It indicates that Section A of the steel tube hardly deformed in a plane during the construction.Hence, the plane section assumption is not applicable to the deformation in Section A. The plane section assumption is always complied by the present code methods in structural designs and analysis of conventional CFST columns.Consequently, the code methods are no longer applicable to this type of CFST column.The non-compliance with the assumption will lead to different mechanical behavior of the columns including failure modes and calculation of the load-carrying capacity compared to the predictions by the code methods.This indicates the practical necessity to reveal the discrepancy in mechanical behavior between the studied columns and conventional CFST columns and to develop new methods to predict the behavior of the studied columns.

Strains in Section B
Figure 8 shows the strain distributions in Section B during the construction stage.The strains are represented in the same way as Section 4.1.As the measuring point JB-d-1 was damaged before removing the scaffold, the data of this point could not be obtained in the data collection.The strain of this point is treated as 0 and marked with red color in Figure 8b,c.It is noteworthy that the treatment of the strain at JB-d-1 as 0 is based on the observations that the strains at one corner should be close to each other and have the same sign.The possible actual strain at this point was so limited that it was reasonably treated as zero.
The measured strain distribution in Section B after concrete pouring is shown in Figure 8a.In this state, it was too early for the concrete to form any strength.The load was resisted by the steel tube and the scaffold.It made the strain level in this state the largest during the construction stage.A locally large tensile strain up to 0.15ε y is observed to occur at JB-a-2, due to the tensile stress caused by the weight of the poured concrete.
data collection.The strain of this point is treated as 0 and marked with red color in Figure 8b,c.It is noteworthy that the treatment of the strain at JB-d-1 as 0 is based on the observations that the strains at one corner should be close to each other and have the same sign.The possible actual strain at this point was so limited that it was reasonably treated as zero.The measured strain distribution in Section B after concrete pouring is shown in Figure 8a.In this state, it was too early for the concrete to form any strength.The load was resisted by the steel tube and the scaffold.It made the strain level in this state the largest during the construction stage.A locally large tensile strain up to 0.15εy is observed to occur at JB-a-2, due to the tensile stress caused by the weight of the poured concrete.
The measured strain distribution in Section B before the removal of the scaffold after 28-day concrete curing is shown in Figure 8b.As the concrete reached its full strength, the concrete worked together with the steel tube to sustain the load.With the help of the concrete, the strain at each measuring point in Section B was reduced compared to the strain in Figure 8a.
After the removal of the scaffold, the support from the scaffold on the beams next to the column disappeared.The construction load on the beams transferred to the column.Therefore, the load on the column section in this state was increased compared to the load in Figure 8b.And it led to an increase in the strains at each measuring point, as shown in Figure 8c.Moreover, by comparing the strain distributions in the three states, it is found that the strain distributions are gradually distributed about the vertical symmetry axis of the section, and the strains at the west side are all compressive, while the strains at the east side are 0 or small tensile.It is due to the fact that the moment caused by the eccentric compression mainly affected the deformations at the east and west sides of the column.The moment also turned the strain at JB-a-2 to become tensile again after the removal of the scaffold.
During the construction, the largest tensile strain occurred at JB-a-2 after concrete pouring, as much as 0.15εy.The largest compressive strain occurred at JB-c-2 after the removal of the scaffold, as much as −0.08εy.Therefore, the positions near JB-a-2 and JB-c-2 are considered the most vulnerable in the section during the construction stage.

Deformation against Plane Section Assumption in Section B
The three-dimensional plane fitting of the strains in Section B in each state is shown in Figure 9, where the same fitting method as Section 4.2 is used.The measured strains εm and the desired strains following the plane section assumption εd are compared in Table 2.The deviation rate is shown in Figure 10.The measured strain distribution in Section B before the removal of the scaffold after 28-day concrete curing is shown in Figure 8b.As the concrete reached its full strength, the concrete worked together with the steel tube to sustain the load.With the help of the concrete, the strain at each measuring point in Section B was reduced compared to the strain in Figure 8a.
After the removal of the scaffold, the support from the scaffold on the beams next to the column disappeared.The construction load on the beams transferred to the column.Therefore, the load on the column section in this state was increased compared to the load in Figure 8b.And it led to an increase in the strains at each measuring point, as shown in Figure 8c.Moreover, by comparing the strain distributions in the three states, it is found that the strain distributions are gradually distributed about the vertical symmetry axis of the section, and the strains at the west side are all compressive, while the strains at the east side are 0 or small tensile.It is due to the fact that the moment caused by the eccentric compression mainly affected the deformations at the east and west sides of the column.The moment also turned the strain at JB-a-2 to become tensile again after the removal of the scaffold.
During the construction, the largest tensile strain occurred at JB-a-2 after concrete pouring, as much as 0.15ε y .The largest compressive strain occurred at JB-c-2 after the removal of the scaffold, as much as −0.08ε y .Therefore, the positions near JB-a-2 and JB-c-2 are considered the most vulnerable in the section during the construction stage.

Deformation against Plane Section Assumption in Section B
The three-dimensional plane fitting of the strains in Section B in each state is shown in Figure 9, where the same fitting method as Section 4.2 is used.The measured strains ε m and the desired strains following the plane section assumption ε d are compared in Table 2.The deviation rate is shown in Figure 10.As can be seen from Figure 10, the measured strains significantly deviate from the fitting plane during the construction stage, as the deviation rate ranges from −210.98% to 218.63%.It indicates that Section B of the steel tube hardly deformed in a plane during the construction.Hence, the plane section assumption is not applicable to the deformation in Section B. Accordingly, the calculation formula for strain distribution following the assumption does not work at this section.

Strains of Vertical Steel Faces between Section A and Section B
The strain distributions on the vertical steel faces between Section A and Section B of the column are shown in Figure 11.In the figure, the faces of each column are named following the same rule.For example, the face bc-1 indicates the face of the column containing the measuring points JA-b-1, JB-b-1, JA-c-1, and JB-c-1.The negative strains indicate compressive strains, and the positive strains indicate tensile strains.
Figure 11a shows the measured strain distribution after concrete pouring.In this state, it was too early for the concrete to form any strength.The steel tubes bore most of the construction load and the weight of the concrete, and the remaining part of the load was borne by the scaffold.It made the strain level in this state the largest during the construction stage.Figure 11b shows the measured strain distribution before the removal of the scaffold after 28-day concrete curing.As the concrete reached its full strength, the concrete worked together with the steel tube to sustain the load.Thus, the load borne by the steel stubes decreased, and the strains in Figure 11b decreased compared to Figure 11a.Figure 11c shows the measured strain distribution after the removal of the scaffold.By comparing Figure 11a and Figure 11c, it can be found that the strain at each face decreased, As can be seen from Figure 10, the measured strains significantly deviate from the fitting plane during the construction stage, as the deviation rate ranges from −210.98% to 218.63%.It indicates that Section B of the steel tube hardly deformed in a plane during the construction.Hence, the plane section assumption is not applicable to the deformation in Section B. Accordingly, the calculation formula for strain distribution following the assumption does not work at this section.

Strains of Vertical Steel Faces between Section A and Section B
The strain distributions on the vertical steel faces between Section A and Section B of the column are shown in Figure 11.In the figure, the faces of each column are named following the same rule.For example, the face bc-1 indicates the face of the column containing the measuring points JA-b-1, JB-b-1, JA-c-1, and JB-c-1.The negative strains indicate compressive strains, and the positive strains indicate tensile strains.
Figure 11a shows the measured strain distribution after concrete pouring.In this state, it was too early for the concrete to form any strength.The steel tubes bore most of the construction load and the weight of the concrete, and the remaining part of the load was borne by the scaffold.It made the strain level in this state the largest during the construction stage.Figure 11b shows the measured strain distribution before the removal of the scaffold after 28-day concrete curing.As the concrete reached its full strength, the concrete worked together with the steel tube to sustain the load.Thus, the load borne by the steel stubes decreased, and the strains in Figure 11b decreased compared to Figure 11a.Figure 11c shows the measured strain distribution after the removal of the scaffold.By comparing Figures 11a and 11c, it can be found that the strain at each face decreased, but the compressive strain at the face bc-1 increased.It is due to the fact that the concrete bent under eccentric compression, resulting in greater compressive strain at the compressive face.By comparing Figures 11b and 11c, it can be found that the strain at all faces increased.It is due to the fact that all the construction load was transferred to the column after the removal of the scaffold.Considering the length of the beams was as long as 9 m, the changes in strain distributions caused by the self-weight were necessarily significant.
at JA-b-1 reached up to −0.11εy after the removal of the scaffold.The compressive strain at JB-c-2 was maintained at a high level during the construction (−0.08εy, −0.07εy, −0.08εy, respectively).Therefore, the positions near JA-d-1, JA-b-1, and JB-c-2 were considered the most vulnerable.It can also be found from Figure 9 that the strains in the two sections showed obvious linear changes at each face during the construction stage.It indicates that the column was obviously bent in two directions near Sections A and B, and the column performed a spatial deformation under bi-directional bending.This would reduce the load-bearing capacity of the column.Furthermore, the bi-directional bending would result in uneven bending across different section segments of the column.Consequently, certain segments may experience a higher bending stress, thereby impacting the overall performance of the column.Additionally, the bi-directional bending might introduce a torsional effect, causing the column to twist while bending.This torsion can compromise the overall stability of the column, thereby influencing its load-bearing capacity.

Strains of Concrete in Section B
Figure 12 shows the strain distribution of the concrete in Section B during the construction stage.Note that the compressive strains are represented as multiples of the concrete ultimate compressive strain, εcu = 3300 µε, and the tensile strains are represented as multiples of the concrete ultimate tensile strain, εtu = 100 µε, while positive strains indicate During the construction, the largest tensile, as much as 0.02ε y , occurred at JA-d-1 after concrete pouring.The largest compressive strain occurred at JA-b-1 and JB-c-2.The strain at JA-b-1 reached up to −0.11ε y after the removal of the scaffold.The compressive strain at JB-c-2 was maintained at a high level during the construction (−0.08ε y , −0.07ε y , −0.08ε y , respectively).Therefore, the positions near JA-d-1, JA-b-1, and JB-c-2 were considered the most vulnerable.It can also be found from Figure 9 that the strains in the two sections showed obvious linear changes at each face during the construction stage.It indicates that the column was obviously bent in two directions near Sections A and B, and the column performed a spatial deformation under bi-directional bending.This would reduce the load-bearing capacity of the column.Furthermore, the bi-directional bending would result in uneven bending across different section segments of the column.Consequently, certain segments may experience a higher bending stress, thereby impacting the overall performance of the column.Additionally, the bi-directional bending might introduce a torsional effect, causing the column to twist while bending.This torsion can compromise the overall stability of the column, thereby influencing its load-bearing capacity.

Strains of Concrete in Section B
Figure 12 shows the strain distribution of the concrete in Section B during the construction stage.Note that the compressive strains are represented as multiples of the concrete ultimate compressive strain, ε cu = 3300 µε, and the tensile strains are represented as multiples of the concrete ultimate tensile strain, ε tu = 100 µε, while positive strains indicate tensile strains and negative strains indicate compressive strains.Solidification heated the concrete after pouring, and the raised temperature deviated the measurement of the embedded vibrating wire meter.Thus, the measured strains in this state are not used for analysis to avoid the impact of the incorrect strain value on the research findings.As the concrete reached its full strength after the curing, the concrete work gether with the steel tube to sustain the construction load.After the removal of th fold, the support from the scaffold on the beams next to the column disappeare construction load on the beams transferred to the column, which increased the load by the concrete in the column.The measured strain distribution in this state is sh Figure 12b.It can be found that the strain distributions in Figure 12a,b were quite s It is due to the fact that the uneven concrete pouring at the south and north sides beam led to eccentric compression on the concrete at Section B. According to Figur and 12a,b, it can be found that the strains of the steel tube increased somewhat, a strains of the concrete did not change significantly in this state.Obviously, the inc load after the removal of the scaffold was primarily borne by the steel tubes, and th crete inside bore only a small part of the load.It indicates stress redistributions acr section, which need to be considered in the behavior prediction of the column.
During the construction, the largest tensile strain occurred at JBN-c, as m 2.10εtu, much larger than the ultimate tensile strain εtu of the concrete.The concre the point was seriously cracked under tension.It made the concrete unable to be In Section B of the concrete, the measuring point JBN-a may be damaged by concrete erosion during concrete pouring, resulting in the failure to obtain the data at this point during the whole data collection process.The strain of this point is treated as zero and marked with a red color in Figure 12.
As the concrete reached its full strength after the curing, the concrete worked together with the steel tube to sustain the construction load.After the removal of the scaffold, the support from the scaffold on the beams next to the column disappeared.The construction load on the beams transferred to the column, which increased the load borne by the concrete in the column.The measured strain distribution in this state is shown in Figure 12b.It can be found that the strain distributions in Figure 12a,b were quite similar.It is due to the fact that the uneven concrete pouring at the south and north sides of the beam led to eccentric compression on the concrete at Section B. According to Figures 8b,c and 12a,b, it can be found that the strains of the steel tube increased somewhat, and the strains of the concrete did not change significantly in this state.Obviously, the increased load after the removal of the scaffold was primarily borne by the steel tubes, and the concrete inside bore only a small part of the load.It indicates stress redistributions across the section, which need to be considered in the behavior prediction of the column.
During the construction, the largest tensile strain occurred at JBN-c, as much as 2.10ε tu , much larger than the ultimate tensile strain ε tu of the concrete.The concrete near the point was seriously cracked under tension.It made the concrete unable to bear any tension in those states, ultimately jeopardizing the structural integrity and safety of the column.The largest compressive strain occurred at JBN-b, as much as −0.02ε cu , much less than the ultimate compressive strain ε cu of the concrete.It indicates that the area near JBN-b did not experience concrete crush during the construction stage.The position near JBN-c is considered the most vulnerable in the section during the construction stage.Considering mass concreting at one time during the construction of the column, necessary measures should be taken to avoid the concrete cracking, such as adding anti-cracking agents, controlling the construction temperature, and adopting segmented concreting.

Relative Slip between Concrete and Steel Tube
Interplay between the concrete and the steel tubes provides an important guarantee for the favorable mechanical performance of the CFST column.The relative slip between the two materials reflects the degree of the interplay under the construction load.
Figure 13 shows the relative slip strain distribution between the concrete and the steel tube during the construction stage.The relative slip strain ε slip is calculated according to Equation (2) [39], where ε steel represents the strain of the steel tubes and ε cnt represents the strain of the concrete.The positive strains indicate tensile strains and the negative strains indicate compressive strains in the figure.Solidification heated the concrete after pouring, and the raised temperature deviated the measurement of the embedded vibrating wire meter.Therefore, the strain distribution is investigated just before and after the removal of the scaffold.It can be found from Figure 13a that the largest relative slip occurred at JB-c-1 and JB-c-2, and there is an obvious slip between the concrete and the steel tube near JB-c-1 and JB-c-2.It is due to the fact that the absence of studs inside the steel tube led to a lack of effective connection between the two materials in this area, and the two materials could not be integrally stressed and deformed under the construction load.
Before the scaffold was removed, as the concrete curing was not completed, part of the scaffold was placed under the beam to support the self-weight of the beams and their It can be found from Figure 13a that the largest relative slip occurred at JB-c-1 and JB-c-2, and there is an obvious slip between the concrete and the steel tube near JB-c-1 and JB-c-2.It is due to the fact that the absence of studs inside the steel tube led to a lack of effective connection between the two materials in this area, and the two materials could not be integrally stressed and deformed under the construction load.
Before the scaffold was removed, as the concrete curing was not completed, part of the scaffold was placed under the beam to support the self-weight of the beams and their formworks.After the removal of the scaffold, the weight of the beams and the load applied on the beams was transferred to the CFST column.The relative slip strain distribution between the concrete and the steel tube is shown in Figure 13b.It can be seen from the figure that the relative slip strain was the largest at JB-c-1 and JB-c-2.A significant slip occurred in the area nearby, and the concrete and the steel tube could not be integrally stressed and deformed under construction load.
In the above two construction states, the relative slip strain distribution between the concrete and the steel tube was quite similar, and the largest relative slip strain occurred at JB-c-1 and JB-c-2 in both states, as much as 291.69 µε.Therefore, the corner areas near JB-c-1 and JB-c-2 were the most prone to relative slip between the concrete and the steel tube.It was caused by the inconsistent stressing between the concrete and the steel tube at these points, namely the greatly tensioned concrete but the compressed steel tube at the corner.The localized significant slip would lead to deteriorated mechanical behavior of the column including different failure modes and degraded load-carrying capacity compared to the predictions by the code methods.Measures such as the application of binders and surface treatment of the steel tube are suggested to reduce the slip at the corner.

Conclusions
In this study, the strains in the sections of an inclined large-size non-studded CFST column at the middle story were measured and tracked during its construction, and the characteristics of the strain distributions were studied.The conclusions were drawn as follows: 1.
The plane section assumption was not applicable to the deformation of the CFST column during the construction.New methods to predict the behavior of this type of column are a practical necessity; 2.
Among the three construction states, the strains of the CFST column were the largest after concrete pouring, followed by the strains after the scaffold removal, and the strains were the lowest before the scaffold removal.The nonlinear variations in the strain levels among the three states indicate stress redistributions across the section during the construction stage.This feature provides a valuable reference for designs and constructions of large-size non-studded CFST columns; 3.
Bi-directional bending occurred in the CFST column segment.The most vulnerable positions of the steel tube were located at the south and north sides of the upper end of the column segment and the east and west sides of the lower end of the segment; 4.
The absence of studs in the steel tube significantly decreased the bond between the concrete and the steel tube, which made the two materials fail to work together effectively.Compared with the conventional CFST column, the concrete and the steel tube of the column tended to be stressed and deformed more independently.This feature would lead to deteriorated mechanical behavior of the column including different failure modes and degraded load-carrying capacity compared to the predictions by the code methods.Measures such as the application of binders and surface treatment of the steel tube are suggested to reduce the slip at the corners of this type of column.
Based on the results in this study, subsequent study will evaluate the deterioration of the mechanical properties of the column caused by the slip during the construction stage and will put forward feasible suggestions for designs and constructions of this type of column.

Figure 1 .
Figure 1.CFST columns.(a) Photo of CFST column on site; (b) Schematic of CFST column and surrounding beams.Figure 1. CFST columns.(a) Photo of CFST column on site; (b) Schematic of CFST column and surrounding beams.

Figure 1 .
Figure 1.CFST columns.(a) Photo of CFST column on site; (b) Schematic of CFST column and surrounding beams.Figure 1. CFST columns.(a) Photo of CFST column on site; (b) Schematic of CFST column and surrounding beams.

Figure 3 .
Figure 3. Layout of measuring points at column sections.(a) Section A; (b) Section B.

Figure 3 .Figure 4 .
Figure 3. Layout of measuring points at column sections.(a) Section A; (b) Section B.

Figure 4 .
Figure 4. Strain gauges on exterior surface.Figure 4. Strain gauges on exterior surface.

Figure 5 .
Figure 5. Strain distribution in Section A during construction.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure 5 .
Figure 5. Strain distribution in Section A during construction.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure 6 .
Figure 6.Three-dimensional plane fitting of strains in Section A in each state.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.Figure 6. Three-dimensional plane fitting of strains in Section A in each state.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure 6 .
Figure 6.Three-dimensional plane fitting of strains in Section A in each state.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.Figure 6. Three-dimensional plane fitting of strains in Section A in each state.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure 7 .
Figure 7. Deviation rates in Section A in each state.

Figure 7 .
Figure 7. Deviation rates in Section A in each state.

Figure 8 .
Figure 8. Strain distributions in Section B during construction.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure 8 .
Figure 8. Strain distributions in Section B during construction.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure 9 .
Figure 9. Three-dimensional plane fitting of strains in Section B in each state.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.Figure Three-dimensional plane fitting of strains in Section B in each state.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure
Figure 9. Three-dimensional plane fitting of strains in Section B in each state.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.Figure Three-dimensional plane fitting of strains in Section B in each state.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure 10 .
Figure 10.Deviation rates in Section B in each state.

Figure 10 .
Figure 10.Deviation rates in Section B in each state.

Figure 11 .
Figure 11.Strain distribution at each side of the column during construction.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure 11 .
Figure 11.Strain distribution at each side of the column during construction.(a) After concrete pouring; (b) Before removing scaffold; (c) After removing scaffold.

Figure 12 .
Figure 12.Strain distribution of concrete in Section B during construction.(a) Before removi fold; (b) After removing scaffold.

Figure 12 .
Figure 12.Strain distribution of concrete in Section B during construction.(a) Before removing scaffold; (b) After removing scaffold.

Figure 13 .
Figure 13.Relative slip strain distribution between concrete and steel tube.(a) Before removing scaffold; (b) After removing scaffold.

Figure 13 .
Figure 13.Relative slip strain distribution between concrete and steel tube.(a) Before removing scaffold; (b) After removing scaffold.

Table 1 .
Fitting data in Section A in each state.

Table 2 .
Fitting data in Section B in each state.

Table 2 .
Fitting data in Section B in each state.