Publication: Effect of heat sink configuration on Thermal management of different Semiconductor modules using Regression models.
| datacite.subject.fos | oecd::Engineering and technology::Mechanical engineering | |
| dc.contributor.author | Salim, Muhammad Razin | |
| dc.date.accessioned | 2026-11-11T09:10:00Z | |
| dc.date.available | 2026-11-11T09:10:00Z | |
| dc.date.issued | 2025-03 | |
| dc.description.abstract | As semiconductor devices miniaturize and increase in power density, effective thermal management becomes critical to ensure optimal performance, reliability, and longevity. Traditional cooling methods often fail to meet the thermal demands of highpower electronics, requiring extensive testing and lengthy lead times. This study introduces a novel approach to address these challenges by developing reliable thermal models and a predictive framework to optimize heat sink designs for semiconductor devices, particularly for high-performance applications such as external SSDs.Key innovations include a stacked 3D model utilizing the Finite Volume Method (FVM) to predict thermal gradients and hotspots without relying on complex architectural data, making the model versatile for a wide range of semiconductor devices. Unlike traditional methods, which rely solely on equations and experimental data to determine junction temperature, this study enables direct access to junction temperature based on changes in heat sink design. The study also uniquely integrates heat sink configurations directly into the semiconductor module, allowing real-world testing of their impact on junction temperature reduction.By analyzing ten power dissipation levels (ranging from 10% to 100% of a maximum 2.6W), results show that increasing power dissipation significantly reduces junction temperature (by 7°C to 68°C), emphasizing the importance of optimizing surface heat transfer coefficients. The study also evaluates heat sink configurations such as fin number, fin height, and base dimensions to identify optimal designs. Increasing fin height from 0 mm to 25 mm reduced junction temperature by 18.17%, while optimizing the number of fins (from 0 to 8) achieved a reduction of 8.68% at a power level of 0.55W, with 4 fins providing the best balance between heat dissipation and minimal volume. Larger base areas and thicker bases further enhanced thermal performance, reducing junction temperatures from 81.18°C to 62.72°C and from 73.26°C to 63.47°C, respectively. However, the study highlights that overly packed fins or larger bases can impede airflow and heat transfer efficiency, underscoring the need for careful optimization.To provide predictive insights, linear and polynomial regression models were employed to estimate the impact of heat sink designs on junction temperatures. The linear regression model consistently outperformed the polynomial model, offering a reliable framework for optimizing thermal solutions, particularly for lower-power chips. This research provides actionable guidelines for designing efficient thermal management systems, reducing lead times, and enhancing the reliability of semiconductor devices through advanced thermal modeling and predictive techniques. | |
| dc.identifier.uri | https://erepo.usm.my/handle/123456789/25424 | |
| dc.language.iso | en | |
| dc.title | Effect of heat sink configuration on Thermal management of different Semiconductor modules using Regression models. | |
| dc.type | Resource Types::text::thesis::master thesis | |
| dspace.entity.type | Publication | |
| oairecerif.author.affiliation | Universiti Sains Malaysia |