To mitigate the input energy of the earthquake, numerous vibration control systems have been broadly proposed. The vibration control system can be categorized as passive, active, semi-active, and hybrid. The metallic dampers as a passive vibration control system because of their simple construction and configuration, low cost, availability, high efficiency, rate-independent, resistance to ambient temperature, and require no external energy, are an appropriate and economical control system in structures to mitigate the input energy of the earthquake. Metallic dampers can control the system and absorb the input energy of the earthquake through a shear, flexural, axial, and torsional capacity of the metal element. Different configurations of the metallic damper have been proposed. One of the energy-absorbing mechanisms of the metallic damper is using axial capacity. This paper proposed a new type of buckling restrained brace. Energy Dissipating elements of this damper withstand applied force through the compressive strength of the core plates in each cycle. The proposed damper consists of four comb-teeth elements instead of one consistent element. Each comb-teeth element with a rectangular section of 20×8 mm consists of three narrow straps with different lengths of 390, 386 and 382 mm to make a jump in the hysteric response curve by making lag in the formation of the buckling mode shapes. These comb-teeth elements were welded to two positions on both sides of the middle plate. The middle plate’s height, width, and thickness are 500, 400 and 30 mm, respectively. The proposed damper carried out the axial load through the displacement of this plate. Twelve restraining elements with a section of 30×20 mm and a length of 830 mm were used to prevent global buckling of each strap of the comb-teeth damper. For this purpose, twelve grooves have been created on the middle plate and the restraining members passed through it. Also, to prevent the residual displacement of the system, six parallel springs with an equivalent stiffness of 2316 N/mm were used as a self-centering system on each side of the middle plate. By applying force to the middle plate, the pre-compressive springs provide self-centering force and bring the middle plate to the initial position. Finally, two end plates as reaction plates with a section of 300×200 mm and a thickness of 20 mm were bolted to the cross-section of the restraining members by using 10 mm high-strength grade 12.9 bolts. The springs are placed inside the pipes and between the middle plate and end plate. Both ends of the springs stay free. Also, to prevent the lateral displacement of the restraining members perpendicular to the length of the straps, 48 batten plates were used. These batten plates were bolted by using 8 mm high-strength grade 12.9 bolts at two positions from each end of the restraining members. The whole length of the proposed damper is 870 mm. The 1000kN-capacity servo-controlled hydraulic jack was used to experimentally determine the hysteretic behavior of the proposed damper. Experimental results indicate that the proposed system experienced the maximum axial load of 348.6 kN and -347.6 kN in the back-and-forth direction of applied load at the same axial displacement of 35 mm. The results show that the damper experienced a residual displacement of 10.9 and -11.4 mm at the end of the test, so using springs eliminates the residual displacement up to 68.8% and 67.4%, respectively. Also, to evaluate the performance of the proposed system, energy absorbing capacity and the equivalent viscous damping coefficient were calculated 35.25 kJ and 46% based on the load-displacement curve, respectively. As expected, the failure modes of the core plates of the proposed damper concentrated on buckling modes in the weak and strong axes. It is worth noting that all the systems exhibited satisfactory performance without any instability during the test.