Hydrothermal analysis of archimedean spiral single and dual channel heat sink for CPU cooling

Hala M. Rashad

Article Details

Hydrothermal analysis of archimedean spiral single and dual channel heat sink for CPU cooling

Authors:

Hala M. Rashad

Publisher:

Yildiz Technical University

Main Topics:

Thermodynamics and Statistical Physics

Journal:

Journal of Thermal Engineering Vol.10 Issue 1, pp.50-61

Document Type:

Article

Publication Date:

1/2024

Document Language:

English

Document ID (Org No.):

2001311216

pISSN:

2148-7847

Keywords:

Abstract:

 Engineering modeling and simulation represent a predominantly design tool in the modern manufacturing industry in which the actual system is reproduced using a mathematical and scientific model. This requires CPUs with higher computational capacities. However, increas-ing the computational capacities of CPU and GPU imposes challenges in the cooling process due to space limitations. CPU liquid cooling system has attracted more interest as an efficient heat dissipation tool. This work presents computational modeling of the conjugate heat and flow for the CPU liquid heat sink cooling. An Archimedean spiral channel is grooved into the cold plate of the heat sink. Single and dual channel passes are used in this work. The out-er diameter of the cold plate is 105 mm and the channel depth is 5 mm for both single and dual-channel configurations. The conjugate heat sink model was constructed to have four different domains: CPU alumina, glue layer ethoxy, cold plate copper, and liquid cool-ant water. To incorporate the effect of turbulence, the flow rate varied to cover a range of Reynolds number from 3000 up to 15000 at a constant inlet temperature of 25 C. The used turbulence model was the Shear Stress Transport k-ω to better capture the viscous, high-fre-quency flow fluctuation in the near-wall region. The bottom surface of the CPU is subjected to 450 W of heat energy. The results showed that the channel configuration and Reynolds number have a decisive impact on controlling the CPU temperature. The CPU temperature decrease as Reynolds number increases, however, the pressure drop increases at an exponen-tial rate. These findings are supported by DarcyWeisbach equation for internal flow in which the pressure drop depends on the square of the average fluid velocity and it was noticed that the pressure drop in the dual channel was three times higher than that in the single channel. The hydrothermal performance of the Archimedean spiral channel rapidly decreased with Reynolds number and the single-channel had a slightly better performance compared with the dual-channel.

 Engineering modeling and simulation represent a predominantly design tool in the modern manufacturing industry in which the actual system is reproduced using a mathematical and scientific model. This requires CPUs with higher computational capacities. However, increas-ing the computational capacities of CPU and GPU imposes challenges in the cooling process due to space limitations. CPU liquid cooling system has attracted more interest as an efficient heat dissipation tool. This work presents computational modeling of the conjugate heat and flow for the CPU liquid heat sink cooling. An Archimedean spiral channel is grooved into the cold plate of the heat sink. Single and dual channel passes are used in this work. The out-er diameter of the cold plate is 105 mm and the channel depth is 5 mm for both single and dual-channel configurations. The conjugate heat sink model was constructed to have four different domains: CPU alumina, glue layer ethoxy, cold plate copper, and liquid cool-ant water. To incorporate the effect of turbulence, the flow rate varied to cover a range of Reynolds number from 3000 up to 15000 at a constant inlet temperature of 25 C. The used turbulence model was the Shear Stress Transport k-ω to better capture the viscous, high-fre-quency flow fluctuation in the near-wall region. The bottom surface of the CPU is subjected to 450 W of heat energy. The results showed that the channel configuration and Reynolds number have a decisive impact on controlling the CPU temperature. The CPU temperature decrease as Reynolds number increases, however, the pressure drop increases at an exponen-tial rate. These findings are supported by DarcyWeisbach equation for internal flow in which the pressure drop depends on the square of the average fluid velocity and it was noticed that the pressure drop in the dual channel was three times higher than that in the single channel. The hydrothermal performance of the Archimedean spiral channel rapidly decreased with Reynolds number and the single-channel had a slightly better performance compared with the dual-channel.

 Engineering modeling and simulation represent a predominantly design tool in the modern manufacturing industry in which the actual system is reproduced using a mathematical and scientific model. This requires CPUs with higher computational capacities. However, increas-ing the computational capacities of CPU and GPU imposes challenges in the cooling process due to space limitations. CPU liquid cooling system has attracted more interest as an efficient heat dissipation tool. This work presents computational modeling of the conjugate heat and flow for the CPU liquid heat sink cooling. An Archimedean spiral channel is grooved into the cold plate of the heat sink. Single and dual channel passes are used in this work. The out-er diameter of the cold plate is 105 mm and the channel depth is 5 mm for both single and dual-channel configurations. The conjugate heat sink model was constructed to have four different domains: CPU alumina, glue layer ethoxy, cold plate copper, and liquid cool-ant water. To incorporate the effect of turbulence, the flow rate varied to cover a range of Reynolds number from 3000 up to 15000 at a constant inlet temperature of 25 C. The used turbulence model was the Shear Stress Transport k-ω to better capture the viscous, high-fre-quency flow fluctuation in the near-wall region. The bottom surface of the CPU is subjected to 450 W of heat energy. The results showed that the channel configuration and Reynolds number have a decisive impact on controlling the CPU temperature. The CPU temperature decrease as Reynolds number increases, however, the pressure drop increases at an exponen-tial rate. These findings are supported by DarcyWeisbach equation for internal flow in which the pressure drop depends on the square of the average fluid velocity and it was noticed that the pressure drop in the dual channel was three times higher than that in the single channel. The hydrothermal performance of the Archimedean spiral channel rapidly decreased with Reynolds number and the single-channel had a slightly better performance compared with the dual-channel.

 Engineering modeling and simulation represent a predominantly design tool in the modern manufacturing industry in which the actual system is reproduced using a mathematical and scientific model. This requires CPUs with higher computational capacities. However, increas-ing the computational capacities of CPU and GPU imposes challenges in the cooling process due to space limitations. CPU liquid cooling system has attracted more interest as an efficient heat dissipation tool. This work presents computational modeling of the conjugate heat and flow for the CPU liquid heat sink cooling. An Archimedean spiral channel is grooved into the cold plate of the heat sink. Single and dual channel passes are used in this work. The out-er diameter of the cold plate is 105 mm and the channel depth is 5 mm for both single and dual-channel configurations. The conjugate heat sink model was constructed to have four different domains: CPU alumina, glue layer ethoxy, cold plate copper, and liquid cool-ant water. To incorporate the effect of turbulence, the flow rate varied to cover a range of Reynolds number from 3000 up to 15000 at a constant inlet temperature of 25 C. The used turbulence model was the Shear Stress Transport k-ω to better capture the viscous, high-fre-quency flow fluctuation in the near-wall region. The bottom surface of the CPU is subjected to 450 W of heat energy. The results showed that the channel configuration and Reynolds number have a decisive impact on controlling the CPU temperature. The CPU temperature decrease as Reynolds number increases, however, the pressure drop increases at an exponen-tial rate. These findings are supported by DarcyWeisbach equation for internal flow in which the pressure drop depends on the square of the average fluid velocity and it was noticed that the pressure drop in the dual channel was three times higher than that in the single channel. The hydrothermal performance of the Archimedean spiral channel rapidly decreased with Reynolds number and the single-channel had a slightly better performance compared with the dual-channel.

Article Details

Hydrothermal analysis of archimedean spiral single and dual channel heat sink for CPU cooling

Authors:

Hala M. Rashad

Publisher:

Yildiz Technical University

Main Topics:

Thermodynamics and Statistical Physics

Journal:

Journal of Thermal Engineering Vol.10 Issue 1, pp.50-61

Document Type:

Article

Publication Date:

1/2024

Document Language:

English

Document ID (Org No.):

2001311216

pISSN:

2148-7847

Heat Sink

CPU Liquid Cooling

Conjugate Heat Transfer

Hydrothermal Performance

Computational Fluid Dynamics

Issues

Article Details

Hydrothermal analysis of archimedean spiral single and dual channel heat sink for CPU cooling

Authors: Hala M. Rashad

Publisher: Yildiz Technical University

Main Topics: Thermodynamics and Statistical Physics

Journal: Journal of Thermal Engineering Vol.10 Issue 1, pp.50-61

Document Type: Article

Publication Date: 1/2024

Document Language: English

Document ID (Org No.): 2001311216

pISSN: 2148-7847

Heat Sink

CPU Liquid Cooling

Conjugate Heat Transfer

Hydrothermal Performance

Computational Fluid Dynamics

Issues

Authors:

Hala M. Rashad

Publisher:

Yildiz Technical University

Main Topics:

Thermodynamics and Statistical Physics

Journal:

Journal of Thermal Engineering Vol.10 Issue 1, pp.50-61

Document Type:

Article

Publication Date:

1/2024

Document Language:

English

Document ID (Org No.):

2001311216

pISSN:

2148-7847