This experimental study focused on the performance of half grouted sleeve connections of rebars, which one end of the grout sleeve was mechanical connection, and the other end was grouting sleeve connection. Initially, the varying diameters with different anchorage lengths was processed, such as diameters of 18 mm, 24 mm, and 30 mm for C12, C16, and C20, respectively. Subsequently, the same rebar was screwed into the mechanical connection of the sleeve using a wrench. Finally, the sleeve, yet to be grouted, was secured in place, followed by the execution of sleeve grouting operations using a manual grouting gun.

The rebar fractured outside the sleeve, representing the ideal failure mode and the primary failure type observed in the experiments. It indicated that the tensile strength of the grouted sleeve connection was not less than that of the rebar, reflecting that the rebar anchorage length was suitable, and the strength of the grouting material and the performance of the sleeve met the ideal performance requirements.

The main instances of bond-slip failure were observed in the M ₁-16-80-90, M ₂-16-80-90, M ₂-16-100-90, M ₁-16-80-110, M ₁-20-80-110, and M ₂-20-80-110 groups, where the specimens primarily had a grouting strength of 80 MPa and larger rebar diameters with insufficient anchorage length, indicating that the strength of the grouting and the anchorage length were the main reasons for bond-slip failure. Specimens with bond-slip failure exhibited virtually no rebar strengthening phase and lacked ductile deformation capability.

In the experiments, only the specimen M ₂-20-100-110-2 experienced sleeve fracture, primarily due to the low tensile strength of the cast iron sleeve itself. It is necessary to improve the production quality of the sleeve.

In the pull-out tests of the half-grouted sleeve connection, the axial load was transmitted from the rebar at the grouting end to the grout, then from the grout to the sleeve shell, and subsequently from the sleeve shell to the mechanical connection, where it was finally transferred to the rebar through the threads. The analysis of the strain on the outer shell of the sleeve indicated that the strain on the sleeve shell increased gradually from the mechanical connection end to the grout connection end. This suggested that the deformation of the grout was greatest at the grouting end, where the stress was also highest, and least at the mechanical connection end, where the stress was lower. The change in strain along the sleeve shell typically follows a linear pattern, as illustrated in Figure 1.

From the pull-out tests of half-grouted sleeve connection specimens of rebars, it was concluded that specimens with higher grouting strength and appropriate anchorage length exhibited failure modes of connected rebar fracture, and the ultimate tensile strength of these specimens exceeded 1.1 times the standard value of the ultimate tensile strength for HRB400 grade rebar, meeting the technical requirements of the “Technical Specification for Grouted Sleeve Connections of Reinforcing Bars” (JGJ 355-2015) [13].

The most crucial parameter for the design of grouted sleeve connections is the length of anchorage of the rebar. Although JGJ 355-2015 uniformly specifies that the insertion depth of the connected rebar should not be less than 8 times the diameter of the rebar, it does not consider the impact of factors such as grouting strength and the inner diameter of the sleeve and the calculation formula is not yet clear.

Based on the force transfer mechanism between the rebar and the grout, the basic anchorage length formula for the rebar in the grouted sleeve is derived from the principle of force equilibrium that the resultant force of bond stress along the rebar equal to the external load. l ab = F π d τ (1)

In the formula, τ represents the average bond strength between the rebar and the grout; l _ab is the basic anchorage length of the rebar in the grout sleeve; d is the diameter of the connected rebar; F is the axial load corresponding to the standard value of the rebar yield strength.

The formula for calculating the rebar bond strength τ is as follows: τ = q × μ (2)

Where, μ is the friction coefficient between the rebar and the grout; q is the radial pressure exerted by the grout on the rebar. According to the description in the literature [4], the friction coefficient between the rebar and the grout can be taken as 1, i.e., τ = q.

Based on the stress relationship between the grout and the rebar and between the grout and the sleeve, as shown in Figure 4, the relationship between q and the sleeve stress F _s is established as follows: q × D = 2 t × F s (3)

Where, D is the inner diameter of the sleeve; t is the thickness of the sleeve shell; F _s is the yield strength of the sleeve.

From Eq. 3, q is determined as follows: q = 2 t × F s D (4)

According to research in related literature, the maximum normal pressure generated by the grouting around the rebar does not exceed 0.2 times the f c ′ ; when q exceeds 0.2 f c ′ , the grout will fail due to crushing, hence q ≤ 0.2 f c ′ , where f c ′ is the compressive strength of the grout.

From Eqs 1–4, the theoretical anchorage lengths for grouted sleeve connection specimens with different rebar diameters, grout strengths, and inner sleeve shells can be calculated. According to the mechanical performance indicators of sleeve materials in “Grouted Sleeve for Rebar Connection” (JG/T 398-2012) [14], the yield strength of steel sleeves was 355 MPa, and the tensile strength of cast iron sleeves was 600 MPa, with their yield strength assumed the same as that of steel sleeves at 355 MPa. Finally, the basic anchorage length calculation formula for the grouted sleeve is: l a b = F π d ⁡ min 0.2 f c ′ , 2 t × F s D (5)

According to the calculation results of Eq. 5, the calculated basic anchorage lengths for steel half-grouted sleeve connection of rebars and cast iron half-grouted sleeve connections were generally consistent with the experimental observations. For example, when the grout strength was 80 MPa, the theoretical calculated lengths for rebars of 16 mm and 20 mm diameter were 125 mm, whereas in the experiments they were 90 mm and 110 mm, respectively. Therefore, the experimental failure mode of these specimens is primarily bond-slip failure.