Reflow Recipe Establishment Based on CFD-Informed Machine Learning Model

Lai, Yangyang; Ha, Jonghwan; Deo, Karthik; Yang, Junbo; Yin, Pengcheng; Park, Seungbae

doi:10.1109/tcpmt.2023.3239304

Cited by 13 publications

(4 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The FEM scheme was developed in Ahmar and Wiese (2017) to simulate the crack growth in MLCC during thermal reflow soldering. A number of FEM schemes (Tian et al , 2004; Lai et al , 2023; Abas et al , 2018; Ng et al , 2020) have also been applied to study reflow soldering process in ball grid array packages. The extended finite element method (XFEM) expands the FEM by adding more degrees of freedom through a process known as enrichment to model the structural defects (Datta, 2013).…”

Section: Introductionmentioning

confidence: 99%

Deformation and crack growth in multilayered ceramic capacitor during thermal reflow process: numerical and experimental investigation

Apalowo,

Abas,

Bachok

et al. 2024

View full text Add to dashboard Cite

Purpose This study aims to investigate the possible defects and their root causes in a soft-termination multilayered ceramic capacitor (MLCC) when subjected to a thermal reflow process. Design/methodology/approach Specimens of the capacitor assembly were subjected to JEDEC level 1 preconditioning (85 °C/85%RH/168 h) with 5× reflow at 270°C peak temperature. Then, they were inspected using a 2 µm scanning electron microscope to investigate the evidence of defects. The reliability test was also numerically simulated and analyzed using the extended finite element method implemented in ABAQUS. Findings Excellent agreements were observed between the SEM inspections and the simulation results. The findings showed evidence of discontinuities along the Cu and the Cu-epoxy layers and interfacial delamination crack at the Cu/Cu-epoxy interface. The possible root causes are thermal mismatch between the Cu and Cu-epoxy layers, moisture contamination and weak Cu/Cu-epoxy interface. The maximum crack length observed in the experimentally reflowed capacitor was measured as 75 µm, a 2.59% difference compared to the numerical prediction of 77.2 µm. Practical implications This work's contribution is expected to reduce the additional manufacturing cost and lead time in investigating reliability issues in MLCCs. Originality/value Despite the significant number of works on the reliability assessment of surface mount capacitors, work on crack growth in soft-termination MLCC is limited. Also, the combined experimental and numerical investigation of reflow-induced reliability issues in soft-termination MLCC is limited. These cited gaps are the novelties of this study.

show abstract

Section: Introductionmentioning

confidence: 99%

Deformation and crack growth in multilayered ceramic capacitor during thermal reflow process: numerical and experimental investigation

Apalowo,

Abas,

Bachok

et al. 2024

View full text Add to dashboard Cite

show abstract

“…Jing used a genetic algorithm to predict the temperature curves of PCB assemblies (Jing et al, 2021). Lai used ML models to specify reflow recipes for PCB with lidded ball grid array (BGA) packages (Lai et al, 2023(Lai et al, , 2022c.…”

Section: Introductionmentioning

confidence: 99%

“…The use of ML models in this study enables the rapid generation of temperature profiles in the same format as the measurement data, representing a significant improvement over the authors' previous approaches where only peak temperature and TAL were available as outputs (Lai et al, 2023(Lai et al, , 2022b. The algorithms play the roles of heat transfer of the forced convection and transient conduction, which are the two dominant heat transfer mechanisms of the reflow soldering process.…”

Section: Introductionmentioning

confidence: 99%

Reflow profiling with the aid of machine learning models

Lai

Park

2023

SSMT

Self Cite

View full text Add to dashboard Cite

Purpose This paper aims to propose a method to quickly set the heating zone temperatures and conveyor speed of the reflow oven. This novel approach intensely eases the trial and error in reflow profiling and is especially helpful when reflowing thick printed circuit boards (PCBs) with bulky components. Machine learning (ML) models can reduce the time required for profiling from at least half a day of trial and error to just 1 h. Design/methodology/approach A highly compact computational fluid dynamics (CFD) model was used to simulate the reflow process, exhibiting an error rate of less than 1.5%. Validated models were used to generate data for training regression models. By leveraging a set of experiment results, the unknown input factors (i.e. the heat capacities of the bulkiest component and PCB) can be determined inversely. The trained Gaussian process regression models are then used to perform virtual reflow optimization while allowing a 4°C tolerance for peak temperatures. Upon ensuring that the profiles are inside the safe zone, the corresponding reflow recipes can be implemented to set up the reflow oven. Findings ML algorithms can be used to interpolate sparse data and provide speedy responses to simulate the reflow profile. This proposed approach can effectively address optimization problems involving multiple factors. Practical implications The methodology used in this study can considerably reduce labor costs and time consumption associated with reflow profiling, which presently relies heavily on individual experience and skill. With the user interface and regression models used in this approach, reflow profiles can be swiftly simulated, facilitating iterative experiments and numerical modeling with great effectiveness. Smart reflow profiling has the potential to enhance quality control and increase throughput. Originality/value In this study, the employment of the ultimate compact CFD model eliminates the constraint of components’ configuration, as effective heat capacities are able to determine the temperature profiles of the component and PCB. The temperature profiles generated by the regression models are time-sequenced and in the same format as the CFD results. This approach considerably reduces the cost associated with training data, which is often a major challenge in the development of ML models.

show abstract

“…Iqbal et al [14] established a 3-D model of the single-temperature zone of the reflow oven in Fluent software by developing a user-defined function (UDF) script with C++ language and adopting a 3-D, incompressible, unsteady turbulent flow pattern. Lai et al [15,16] proposed a computational fluid dynamics (CFD)-based machine learning (ML) model to set the temperature of a convection reflow oven and control conveyor speed. Deng et al [17] studied the real reflow soldering process based on Fluent moving mesh technology.…”

Section: Introductionmentioning

confidence: 99%

Integrated Approach to Obtain Gas Flow Velocity in Convection Reflow Soldering Oven

Xie,

Chen,

Lin

et al. 2023

Symmetry

View full text Add to dashboard Cite

The nozzle-matrix gas flow velocity has a great influence on the accuracy of the temperature field of a printed circuit board assembly (PCBA) during the hot air convection reflow soldering process. This paper proposes a new approach that integrates the theoretical calculation, numerical simulation and an experimental test to accurately determine the nozzle-matrix gas flow velocity. First, the temperature profile of the aluminum alloy thin plate in convection reflow ovens is measured using a Wiken tester. Second, the nozzle-matrix gas flow velocity is theoretically calculated with the Martin formula. The computational fluid dynamic (CFD)simulation is performed according to the Icepak code, where a single oven chamber model is established to represent the 10 zones of soldering ovens to reduce computational resources considering the supry of the soldering ovens. The simulated temperature profile of the aluminum alloy thin plate is obtained and the specific response surface model (RSM) is established to represent the deviation between the simulated temperature and the measured temperature. Finally, based on reverse problem analysis, non-linear programming by quadratic Lagrangian (NLPQL) is used to solve the mathematical optimization model with the objective of minimizing the temperature deviation to obtain the corrected nozzle-matrix gas flow velocity. To validate the accuracy, the temperature test and the modeling using the corrected gas flow velocity for a new PCBA component for the soldering ovens is conducted separately. The temperature comparison between the simulation and the test shows that the maximum temperature deviation is within 10 °C. This provides evidence that the nozzle-matrix gas flow velocity obtained by the new approach is accurate and effective.

show abstract

Reflow Recipe Establishment Based on CFD-Informed Machine Learning Model

Cited by 13 publications

References 28 publications

Deformation and crack growth in multilayered ceramic capacitor during thermal reflow process: numerical and experimental investigation

Deformation and crack growth in multilayered ceramic capacitor during thermal reflow process: numerical and experimental investigation

Reflow profiling with the aid of machine learning models

Integrated Approach to Obtain Gas Flow Velocity in Convection Reflow Soldering Oven

Contact Info

Product

Resources

About