C-I6I6 Load transfer

I6.1 General Requirements

External forces are typically applied to composite members through direct connection to the steel member, bearing on the concrete, or a combination thereof. Design of the connection for force application follows the requirements for the applicable limit states within Chapters J and K of the Specification as well as the provisions of Section I6. Note that for concrete bearing checks on filled composite members, confinement can affect the bearing strength for external force application as discussed in Commentary Section I6.2.

Once a load path has been provided for the introduction of external forces to the member, the interface between the concrete and steel must be designed to transfer the longitudinal shear required to obtain force equilibrium within the composite section. Section I6.2 contains provisions for determining the magnitude of longitudinal shear to be transferred between the steel and concrete depending upon the external force application condition. Section I6.3 contains provisions addressing mechanisms for the transfer of longitudinal shear.

The load transfer provisions of this section are primarily intended for the transfer of longitudinal shear due to applied axial forces. Load transfer of longitudinal shear due to applied bending moments in encased and filled composite members is beyond the scope of this section; however, tests (Lu and Kennedy, 1994; Prion and Boehme, 1994; Wheeler and Bridge, 2006) indicate that filled composite members can develop their full plastic moment capacity based on bond alone without the use of additional anchorage.

Composite column base connections are another aspect of load transfer that is not addressed directly by the Specification. As with all frame systems, selecting and proportioning base connections is a critical aspect of the design process. Load paths are required to transfer the internal forces from each component of a composite column (in other words, structural steel, concrete, and reinforcing steel) to the supporting element(s). Several connection types have been proposed in the literature ranging from standard base plate connections (similar to those common for structural steel columns) to embedded connections of various configurations. Not all the proposed connections are capable of developing the full strength of the composite column. Hitaka et al. (2003) and Stephens et al. (2016) describe the practical limitations of several types of these connections.

I6.2 Force Allocation

This Specification addresses conditions in which the entire external force is applied to the steel or concrete as well as conditions in which the external force is applied to both materials concurrently. The provisions are based upon the assumption that in order to achieve equilibrium across the cross section, transfer of longitudinal shears along the interface between the concrete and steel shall occur such that the resulting force levels within the two materials may be proportioned according to the relative cross-sectional strength contributions of each material. Load allocation based on the cross-sectional strength contribution model is represented by Equations I6-1 and I6-2. Equation I6-1 represents the magnitude of force that is present within the concrete encasement or concrete fill at equilibrium. The longitudinal shear generated by loads applied directly to the steel section is determined based on the amount of force to be distributed to the concrete according to Equation I6-1. Conversely, when load is applied to the concrete section only, the longitudinal shear required for cross-sectional equilibrium is based upon the amount of force to be distributed to the steel according to Equation I6-2. Where loads are applied concurrently to the two materials, the longitudinal shear force to be transferred to achieve cross-sectional equilibrium can be taken as either the difference in magnitudes between the portion of external force applied directly to the concrete and that required by Equation I6-1 or the portion of external force applied directly to the steel section and that required by Equations I6-2a and b. For filled composite members, the steel contribution to the overall nominal axial compressive strength of the cross section decreases with increasing slenderness ratio (b/t or D/t). As a result, it is not permitted to apply axial force directly to the steel wall of filled composite sections classified as slender-element composite because the stress concentrations associated with force application could cause premature local buckling. Additionally, the magnitude of longitudinal shears required to be transferred to the concrete infill would require impractical load transfer lengths.

When external forces are applied to the concrete of a filled composite member via bearing, it is acceptable to assume that adequate confinement is provided by the steel encasement to allow the maximum available bearing strength permitted by Equation J8-2 to be used. This strength is obtained by setting the term $\sqrt{A_{2} / A_{1}}$ equal to 2 . This discussion is in reference to the introduction of external load to the compression member. The transfer of longitudinal shear within the compression member via bearing mechanisms, such as internal steel plates, is addressed in Section I6.3a.

The Specification provisions assume that the required external force to be allocated imparts compression to the composite section. For applied tensile force, it is generally acceptable to design the component of the composite member to which the force is applied (in other words, either the steel section or the longitudinal reinforcement) to resist the entire tensile force, and no further force transfer calculations are necessary. For atypical conditions where the magnitude of required external tensile force necessitates the use of longitudinal reinforcement in conjunction with the steel section, force allocation to each component may be determined as follows.

When the entire external tensile force is applied directly to the steel section,

(C-I6-1)

When the entire external tensile force is applied directly to the longitudinal reinforcement,

(C-I6-2)

where

$\begin{aligned} P_{n} & =\text { nominal axial tensile strength, determined by Equation I2-8 for encased composite members, and Equation I2-14 for filled composite members, kips (N) } \\ P_{r} & =\text { required external tensile force applied to the composite member, kips (N) } \\ V_{r}^{\prime} & =\text { required longitudinal shear force to be transferred to the steel section or longitudinal reinforcement, kips (N) }\end{aligned}$

Where longitudinal reinforcing bars are used to resist tension forces, they must use appropriate lap splices in accordance with ACI 318 (ACI, 2019) as directed by Section II. For sustained tension, mechanical splices are required by ACI 318, Section 25.5.7.4.

I6.3 Force Transfer Mechanisms

Transfer of longitudinal shear by direct bearing via internal bearing mechanisms, such as internal bearing plates or shear connection via steel anchors, is permitted for both filled and encased composite members. Transfer of longitudinal shear via direct bond interaction is permitted solely for compact and noncompact filled composite members. Although it is recognized that force transfer also occurs by direct bond interaction between the steel and concrete for encased composite columns, this mechanism is typically ignored and shear transfer is generally carried out solely with steel anchors (Griffis, 1992).

The use of the force transfer mechanism providing the largest resistance is permissible. Superposition of force transfer mechanisms is not permitted as the experimental data indicate that direct bearing or shear connection often does not initiate until after direct bond interaction has been breached, and little experimental data is available regarding the interaction of direct bearing and shear connection via steel anchors.

The restriction, added in this edition of the Specification, that direct bond interaction shall not be used for filled composite members where bond failure could result in uncontrolled slip, is intended to prevent reliance on this mechanism in cases, such as tension hangers, where static equilibrium cannot be achieved after limited slip.

I6.3a Direct Bearing

For the general condition of load applied directly to concrete in bearing, and considering a supporting concrete area that is wider on all sides than the loaded area, the nominal bearing strength for concrete is given by Equation J8-2:

where

$A_{1}=$ loaded area of concrete, in. ${ }^{2}\left(\mathrm{~mm}^{2}\right)$

$A_{2}=$ maximum area of the portion of the supporting surface that is geometrically similar to and concentric with the loaded area, in. ${ }^{2}\left(\mathrm{~mm}^{2}\right)$

$f_{c}^{\prime}=$ specified compressive concrete strength, ksi (MPa)

The value of $\sqrt{A_{2} / A_{1}}$ must be less than or equal to 2 (ACI, 2019).

For the specific condition of transferring longitudinal shear by direct bearing via internal bearing mechanisms, Equation I6-3 uses the maximum nominal bearing strength allowed by Equation J8-2 of $1.7 f_{c}^{\prime} A_{1}$. The resistance factor for bearing on concrete, $\phi_{B}$, is 0.65 (and the safety factor for bearing on concrete, $\Omega_{B}$, is 2.31) in accordance with Section J8.

I6.3b Shear Connection

Steel anchors shall be designed according to the provisions for composite components in Section I8.3.

I6.3c Direct Bond Interaction

Force transfer by direct bond is commonly used in filled composite members as long as the connections are detailed to limit local deformations (API, 1993; Roeder et al., 1999). While chemical adhesion provides some contribution, direct bond is primarily a frictional resistance mechanism. There is large scatter in the experimental data on the bond of filled composite compression members; however, some trends have been identified (Roeder et al., 1999; Zhang et al., 2012). Larger cross sections, thinner walls, rectangular shapes, smoothed or oiled interfaces, and high-shrinkage concrete contribute to lower apparent bond strengths. Smaller cross sections, thicker walls, circular shapes, rougher interfaces, expansive concrete, and the presence of bending moment (including eccentric loading such as from shear tabs) contribute to higher apparent bond strengths.

The equations for direct bond interaction for filled composite compression members assume the entire interface perimeter is engaged in the transfer of stress. Accordingly, the strength is compared to the sum of the force required to be transferred from connecting elements framing in from all sides. The scatter in the experimental data leads to the recommended low value of the resistance factor, $\phi$, and the corresponding high value of the safety factor, $\Omega$.

I6.4 Detailing Requirements

To avoid overstressing the structural steel section or the concrete at connections in encased or filled composite members, transfer of longitudinal shear is required to occur within the load introduction length. The load introduction length is taken as

2 times the minimum transverse dimension of the composite member both above and below the load transfer region. The load transfer region is generally taken as the depth of the connecting element as indicated in Figure C-I6.1. In cases where the applied forces are of such a magnitude that the required longitudinal shear transfer cannot take place within the prescribed load introduction length, the designer should treat the compression member as noncomposite along the additional length required for shear transfer.

For encased composite members, steel anchors are required throughout the compression member length to maintain composite action of the member under incidental moments (including flexure induced by incipient buckling). These anchors are typically placed at the maximum permitted spacing according to Section I8.3e. Additional anchors required for longitudinal shear transfer shall be located within the load introduction length as described previously.

Unlike concrete encased members, steel anchors in filled members are required only when used for longitudinal shear transfer and are not required along the length of the member outside of the introduction region. This difference is due to the adequate confinement provided by the steel encasement which prevents the loss of composite action under incidental moments.