Phone:086 18576671093
3. Numerical analysis
3.1. Simulation tools
The FSI numerical analysis was performed using the fluid solver FLUENT 6.3.26 and structural solver ABAQUS/CAE 6.9, coupled online by MpCCI 3.1.0. Using the MpCCI, the quantities were exchanged from one solver to the other through association (neighborhood search) and interpolation. In the present study, the air flow is solved by FLUENT, and the subsequent deformation of FPCB motherboard is addressed by ABAQUS. The deformation information is then fed back to FLUENT, and this process completes one cycle of iteration, as shown in Fig. 5.
3.2. Modeling strategy
In ABAQUS, as shown in Fig. 6, the components of the FPCB motherboard are modeled as idealized simple blocks, which are rigidly attached to the motherboard with no consideration made for attachment flexibility. This simplified modeling technique is reliable, as demonstrated in the previous studies conducted by Pitarresi et al.As given in Table 1, the effective elastic modulus, density, and dimensions of the various components are specified according to the experimental prototype. The fully constrained boundary condition (Ux =Uy = Uz = URx = URy = URz = 0) is assigned to the fixed regions of the motherboard, where U and UR denote linear and rotational displacements, respectively. Gravity is defined in the negative y-direction. Points A, B, C, and D are the locations where the comparison between experimental measurements and numerical predictions are made. This structural model is meshed with 28,353 hexagonal elements.
In FLUENT, the flow is assumed to be three dimensional, lami- nar, incompressible, and unsteady. The laminar model is well sui- ted for a fan-sucking wind tunnel, as investigated by Grimes and co-workers . As shown in Fig.7, the fluid model covers the entire test section of the wind tunnel and is meshed with 848,175 hexagonal grids. For the purpose of structured mesh assignment in fluid modeling, the fluid domain is appropriately split into a few smaller volumes. The motherboard stand holder is not modeled to reduce modeling complexity. In this model, the desired velocity is set at the velocity inlet boundary, and ambient condition is set at the pressure outlet boundary.
3.3. Case studies
The reliability of this FSI simulation technique was substantiated by comparing the predictions with experimental measurements. A few studies were conducted to investigate the deflection and stress induced in the FPCB motherboard under various conditions. First, the inlet velocities were manipulated at 1, 2, 3, 4, and 5 m/s, which are within the typical range found in electronic systems, to examine the effect of flow velocity on the motherboard. In addition, different fastening options were considered, as shown in Fig. 8. Due to the fact that the motherboard is commonly screw-fastened at several points in actual applications.
Fastens 2 and 3 were considered to discover the screw-fastening effects. Fastens 4 and 5 were also tested to evaluate the performance of the mounting when the motherboard was screwed at the middle. In addition, the different component layouts were anticipated to also have significant effects on the mechanical behavior of the FPCB motherboard. Hence, a few possibilities of the component layouts were also examined, as shown in Fig. 9. In these layouts, the locations of the I/O connectors were not altered because of the idea that they are normally placed at the downstream of the motherboard. These layout investigations can clarify some questions on a few outcomes, such as the different orientations of the long memory slots and different locations of the two bigger components (heat sink and CPU fan).
Fig.9. different compontent layout