The results of FEM simulation are shown in Fig. 3. Figure 3(a) shows a model of the simulation. The fiber length, , was fixed at 21 mm, as limited by the length of the tubular PZT (20 mm). Depending on , the lateral displacement of the distal end of the fiber varied from to as shown in Fig. 3(b). Also, contact conditions between the fiber and the hinge affect the results. However, adjusting the tolerance between the fiber and the hole of the hinge was practically very difficult. Instead of making a sliding joint, the minute space inside the hole was filled and fixed by elastic epoxy (EP001, Cemedine, Tokyo, Japan) to permit a relative and flexible movement. To verify the strength of the hinge, maximum stresses were also calculated by FEM simulation as shown in Fig. 3(c). Maximum stresses (Max. principal stress) applied on the fiber varied from 81.0 to 140.4 MPa, which was much lower than the rupture stress of the fiber (2350 MPa29). Maximum stresses (Max. von Mises stress) applied on the hinge varied from 92.5 to 110.5 MPa, which was lower than the yield strength of the hinge material (Stainless steel, ). The first-resonant frequency of the fiber according to is shown in Fig. 3(d). The simulation result indicates that the length of is almost unaffected by the resonant frequency of the fiber, which ranges from 157.5 to 158.6 Hz. After considering all of these factors, was determined to be 1.2 mm, to achieve the maximum lateral displacement. After fixing the as 1.2 mm, we considered the effect caused by the thickness of the epoxy bonded between the hole of the hinge and the fiber. FEM simulation was conducted and is shown in Fig. 3(e). The larger the thickness of the epoxy, the smaller the lateral displacement of the fiber tip becomes. This is reasonable since the change of epoxy thickness is equivalent to the change of . Also, the overall stiffness of the lever structure will increase as thicker epoxy bonding is used. The endoscopic probe was assembled into an aluminum housing produced by a wire cutting process with an outer diameter of 3.5 mm and a wall thickness of 0.35 mm. Soldering of the tubular PZT with an outer diameter of 2.2 mm was achieved using silver paste. A gap between an electric wire and the PZT surface was filled with silver paste (P-100, Elcoat, Tokyo, Japan) and pressed thin. To reduce the overall size of the probe, most of the components were assembled using a high strength epoxy (KEIN 60 min), except for the gap between the hole of the hinge and the fiber (where an elastic epoxy was used). To align the center of the fiber, three-axis manual stages were utilized for precise assembly procedures. The fiber scanning range of the slow axis was about 518 μm. The final scanning range at the focal plane of the GRIN lens was measured to be 107.9 μm, due to the magnification of the GRIN lens ().