Vertical surface-mount modules contain a wide range of component types. During second-pass reflow, gravity, solder surface tension, solder flow, and component inertia can interact and cause component displacement or component loss. A vertical orientation also introduces torque effects that are not present in a conventional horizontal assembly. The following model evaluates these risks for ceramic capacitors, SOT/SOP devices, and MLF devices mounted on FR4 vertical modules.
Abstract
A model was established to identify the critical conditions under which components on a vertical surface-mount module either remain stable or develop displacement and component-loss defects during second reflow. Design of experiments (DOE) was used to analyze ceramic capacitors, SOT/SOP devices, and MLF devices on the upper surface of vertical FR4 modules. The experimental results were then used to validate and regress the simulation model, establishing the mass-to-pad-area conditions associated with defect-free second-pass reflow.
1. Theoretical Background
1.1 Solder-Joint Mechanical Model
During reflow soldering, molten solder forms a wetted connection between the PCB pad and the component termination. According to Young's equation, the forces acting on the solder joint can be expressed as a balance between surface tensions:
γSV = γSL + γLV cos θ
In this equation, γSV is the surface tension between the solid metal and the flux; γSL is the surface tension between the molten solder and the flux; γLV is the surface tension between the molten solder and the tangent to its curved surface; and θ is the contact angle of the solder on the copper pad.
The wetting behavior of the solder is illustrated below.
1.2 Theory of Double-Sided Reflow
For components mounted on the first-reflow side, an industry experience-based equation can be used to estimate the risk of component loss during second reflow:
G/S ≤ 0.046 N/cm2
Here, G is the gravitational force acting on the component and S is the total pad area.
When the molten solder wets the component termination and PCB pad effectively, the adhesion force can exceed the internal cohesive force of the solder. Under these conditions, the solder itself becomes the weakest link in the mechanical system. This relationship provides a useful reference for conventional double-sided reflow, but it does not fully describe a module held in a vertical orientation.
1.3 Theory of Vertical Reflow
In a vertical module, the forces acting on a surface-mount component include not only surface tension and gravity but also torque. Solder flow and component inertia further complicate the force distribution. The resulting mechanical behavior is therefore different from that of a component mounted on a horizontal PCB.
Assume that the solder has completely wetted both the component and the pad, and that solder flow on the PCB pad has stopped. Under these conditions, the component will not shift or fall during second reflow when:
G ≤ τ · S
If G > τ · S, the component will fall when the solder is sufficiently molten. In this expression, τ is the shear strength of the molten solder and S is the effective solder-joint area. The relationship indicates that the component's mass, the available solder-joint area, and the molten solder's shear strength must be evaluated together.
2. Experimental Investigation
2.1 Factor Analysis
A fishbone analysis identified component selection, layout orientation, and pad design as the primary factors affecting displacement and component loss.
2.2 Component Screening and Assembly
Ceramic capacitors, SOT/SOP devices, and MLF devices were selected for screening tests. The initial assembly was required to achieve a 100% pass rate before large-scale validation was performed. The assembly process followed the production process specification.
The reflow profile maintained a temperature range of 210 to 235 °C for 40 to 80 s.
2.3 Screening-Test Results
The front and back sides of the test module are shown below.
For ceramic capacitors, SC1812 and larger packages were prone to displacement in a horizontal orientation. In a vertical orientation, SC2220 components and SOT devices with a mass-to-area ratio of ≤0.019 g/mm2 showed no defects.
For SOP devices, the critical mass-to-area ratio was 0.27 g/mm2 in the horizontal orientation and 0.37 g/mm2 in the vertical orientation. The vertical orientation therefore provided better resistance to displacement.
For MLF devices, the critical value was 0.157 g/mm2 for both horizontal and 45° orientations. The 45° orientation did not provide a significant advantage.
2.4 DOE Testing
2.4.1 Chip Components
A two-level DOE was performed for the pad design and layout orientation of chip components.
The results showed that vertical placement performed better than horizontal placement, while a 1.0:1.0 pad design was preferable to a 1.0:1.2 design. The main-effect and interaction-effect analysis is shown below. The red dashed line on the right represents the 1.0:1.2 pad design, and the black solid line represents the 1.0:1.0 pad design.
Push-force testing showed that the solder-joint strength was higher than the strength of the component body. The pad design therefore had no significant effect on reliability in this test.
For vertically oriented ceramic capacitors, the surface area S1 at the solder termination increased during reflow. More work was consequently required to overcome the surface-tension force, making the component less likely to shift in the vertical orientation.
2.4.2 SOT Devices
Three layout levels and two pad-design levels were selected for the SOT DOE.
The main effects and interaction effects were analyzed using the mean displacement. The red dashed line in the right-hand plot represents the 1.0:1.2 pad design, and the black solid line represents the 1.0:1.0 pad design.
The vertical orientation with the leads facing upward produced the best performance. The horizontal orientation showed significantly greater displacement.
2.4.3 SOP Devices
A three-factor, two-level DOE was performed for SOP devices.
The main-effect and interaction-effect analysis showed that both the number of leads and the layout orientation had a significant influence. As the number of leads increased, the failure rate also increased. The vertical orientation produced a lower failure rate than the horizontal orientation.
3. Chip-Component Modeling and Simulation
3.1 Finite-Element Model
A finite-element model based on the solder-joint creep equation was established to simulate the mechanical behavior of molten solder. Component displacement and shear strain were used to evaluate the degree of misalignment.
The simulation used a solder-joint creep equation and a special solid-element structure in the finite-element model. Eight nodes and three degrees of freedom, Ux, Uy, and Uz, were used. The solder parameters were assigned relatively small values to reproduce the behavior of solder in the molten state during reflow. The model maintained the creep-stress condition for 120 s.
The component displacement uz and shear-strain magnitude ε were used to evaluate component displacement.
In the model, ε is strain, b is displacement, C1, C2, C3, and C4 are characteristic constants such as Young's modulus, σ is the tensile strength of molten solder, and T is absolute temperature.
3.2 Effect of Chip-Component Layout
The simulation showed that the displacement risk for vertically oriented components was lower than that for horizontally oriented components. The primary reason was the difference in solder shear stress. Additional simulations with gravity removed or with the solder paste incorporated into the model further confirmed that shear force was the dominant factor.
The simulated displacements for Example 1, the horizontal layout, and Example 2, the vertical layout, were 0.00201 mm and 0.00169 mm, respectively, under conventional reflow conditions. This indicates a lower displacement risk for the vertical layout.
In Examples 3 and 4, where only gravity was applied, the displacements were similar at approximately 0.00164 mm. This indicates that the difference between the layouts was mainly caused by the shear stress generated by rotation. Example 5 showed that coupled effects also existed in addition to gravity and shear force. When the solder was included in Example 6, the displacement approached that of the horizontal layout, confirming that solder shear force was the fundamental reason for the difference between the two orientations.
3.3 Effects of Pad Size and Solder-Paste Thickness
A larger component pad increases the solder-paste volume. It also increases the solder-fillet thickness beneath the component termination. According to the simulation results, a thicker solder joint produces greater shear stress associated with rotational motion. As a result, the risk of component displacement increases.
4. Conclusions
The combined experimental and simulation analysis produced the following conclusions:
- For ceramic capacitors in a horizontal layout, the mass-to-area ratio should be less than 0.041 g/mm2. For a vertical layout, it should be less than 0.087 g/mm2 to avoid component-loss and displacement risk.
- For SOP devices, the critical mass-to-area ratio was 0.027 g/mm2 in the horizontal orientation and 0.037 g/mm2 in the vertical orientation. Values below these limits were associated with no component-loss or displacement defects in the evaluated conditions.
- For MLF devices, the mass-to-area ratio should be less than 0.0147 g/mm2 to avoid component-loss and displacement risk.
The simulation results were generally consistent with the experimental results:
- A higher mass-to-area ratio increases the probability of failure.
- For two-terminal ceramic capacitors, vertical placement helps reduce the failure rate.
- At the same mass-to-area ratio, a smaller pad produces a lower probability of failure.
- At the same mass-to-area ratio, a greater solder-paste thickness produces a higher probability of failure.
These results show that second-reflow reliability in a vertical module is governed by the interaction of component mass, solder-joint area, layout orientation, solder-paste volume, and molten-solder shear behavior. The mass-to-area ratios listed above are associated with the evaluated component types and test conditions and should not be applied indiscriminately to every vertical-module design.
