Experimental investigation of a novel micro-channel flat loop heat pipe (MC-FLHP) for data center cooling and heat recovery

ABSTRACT

In this paper, a novel micro-channel flat loop heat pipe (MC-FLHP) for data center cooling and heat recovery is proposed, which has the following novelties and distinguished advantages, i.e., a tightly attached flat heat pipes array with IT server thus reaching a ‘server level’ cooling for data center, several pre-set micro-grooves which increases the convective heat transfer coefficient, and a separated evaporator-condenser which greatly improves the flexibility of the system’s installation. The proposed system with R134a (filling ratio of 30%) as the working fluid is designed, constructed and tested, and the testing results are used to evaluate its thermal recovery efficiency and its impact factors, i.e., simulated heat load, cooling water inlet temperature and height difference between the evaporator and condenser. Under the range of test conditions (e.g., simulated heat load from 500 W to 3000 W, cooling water inlet temperature from 15°C to 24°C, height difference from 0.6 m to 1.0 m), the overall thermal recovery efficiency is in the range 57.34% to 84.52%. It is found that under the specific operational conditions with the simulated heat load of 500 W, cooling water inlet temperature of 15°C, and height difference between the evaporator and the condenser of 0.8 m, a peak thermal recovery efficiency of 84.52% is achieved. This study can help achieve the global goals of green and sustainable data center, and also save data center energy consumption and thus operational costs.

KEYWORDS:

• MC-flhp
• data center
• cooling
• thermal recovery efficiency

Nomenclature

c_w – Specific heat capacity of water (J/(kg∙K));

D – Diameter (m);

G_w – Cooling water mass flow rate (kg/s);

Δh – Height difference between the evaporator and condenser (m);

LTHP – loop thermosyphon heat pipe;

LHP – loop heat pipe;

n – The number of temperature data in each group of experiments;

P_t – Simulated heat load (W);

Q_w – Cooling water volume flow rate (m³/h);

T_in – Cooling water inlet temperature (K);

T_out – Cooling water outlet temperature (K);

ΔT – Cooling water temperature difference between the inlet and outlet of heat exchanger (K);

t₁ – The running time of each group of experiments (s);

Δt – The acquisition time interval of the temperature data logger (s);

η_s – Instantaneous thermal recovery efficiency (%);

η_t – Average thermal recovery efficiency (%);

ρ_w – Density of water (kg/m³);

ω – Uncertainties;

Acknowledgments

This work was financially supported by the National Key R&D Program of China (2016YFE0133300), Department of Science and Technology of Guangdong Province, China (2019A050509008), European Commission H2020-MSCA-RISE-2016 Programme (734340-DEW-COOL-4-CDC), and European Commission H2020-MSCA-IF-2018 Programme (835778-LHP-C-H-PLATE-4-DC).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Guangdong Science and Technology Department [2019A050509008]; Ministry of Science and Technology of THE PRC. R.O.C [2016YFE0133300]; H2020 European Commission H2020 Marie Curie action [734340-DEW-COOL-4-CDC,835778-LHP-C-H-PLATE-4-DC].