An official website of the United States government
In recent years, electronic packaging has evolved significantly to meet demands for higher performance, lower costs, and smaller designs. This shift has led to heterogeneous packaging, which integrates chips of varying stack heights and results in non-uniform heat flux and temperature distributions. These conditions pose substantial thermal management challenges, as they can create large temperature gradients, which increase thermal stress and potentially compromise chip reliability. This study explores single-phase liquid cooling for multi-chip modules (MCMs) through a comprehensive experimental and machine learning approach. It investigates the impact of chip spacing, height, fluid flow rate, fluid inlet location, and heat flux uniformity on chip temperature and the thermohydraulic performance of a commercial cold plate. Results show that increasing coolant flow from 1 LPM to 2 LPM decreased thermal resistance by 26 %, with heat losses remaining below 5 %. The left inlet configuration improved temperature uniformity compared to the right, though both yielded comparable thermal performance. Adjusting heater spacing impacted temperature distribution based on inlet position, and lowering one heater by 1 mm raised its temperatures by 15 ◦C due to increased thermal resistance from thermal interface material. A transient test demonstrated the cold plate’s quick response to power surges, in which there is only a 1 ◦C spike above steady state. Complementing these findings, an Artificial Neural Network (ANN) model was developed with optimized architecture specifically for the unique challenges of this study. The ANN model was rigorously validated using an independent dataset, achieving highly accurate temperature predictions (R2 = 0.99) within 2.5 % of experimental
more »
« less
This content will become publicly available on April 1, 2026
Machine learning assisted inverse heat transfer problem to find heat flux in ablative materials
Thermal ablation of materials is a complex phenomenon that involves physical and chemical processes for the thermal protection of systems. However, due to the extreme thermal conditions and moving boundaries, predicting temperature and heat flux at the ablative material is quite challenging. A physics-informed neural network is a promising technique for many such inverse problems, including the prediction of unsteady heat flux. However, traditional physics-informed machine learning algorithms struggle with heat flux predictions in thermal ablation problems due to moving boundary conditions and lack of temperature data in the inaccessible domain. This study presents a hybrid approach, where an artificial neural network (ANN) is used for the accessible domain of the material and a physics-based numerical solution (PNS) technique is used in the inaccessible domain of the material, to find heat flux at the ablative surface. Temperature data at the accessible sensor points are used to train the ANN model. The heat flux at the ablative boundary was iteratively obtained from the numerical solution of the energy equation in the inaccessible domain by matching the ANN-predicted temperature at the last accessible sensor point. Our results indicate that this hybrid methodology significantly outperforms traditional physics-informed machine learning techniques, achieving excellent accuracy in predicting the temperature profiles and heat fluxes under complex conditions for both constant and variable heat flux and properties. By addressing the limitations of conventional physics-informed machine learning methods, our approach provides a robust and reliable solution for modeling the intricate dynamics of ablative processes.
more »
« less
- Award ID(s):
- 2244082
- PAR ID:
- 10601154
- Publisher / Repository:
- Elsevier
- Date Published:
- Journal Name:
- Materials Today Communications
- Volume:
- 45
- Issue:
- C
- ISSN:
- 2352-4928
- Page Range / eLocation ID:
- 112337
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
ABSTRACT For droplet vaporization on a superheated hydrophilic surface, earlier studies have demonstrated that use of machine learning tools to analyze both image information from high-speed video and digital data from sensors can be an effective path to understanding the physics and developing a useful model to predict performance when the surface superheat is at low to moderate levels. For such conditions, the two-phase morphology of the system is usually well-behaved, exhibiting conduction-dominated film evaporation of the spread droplet, or nucleate boiling at active nucleation sites in the liquid film of the spread droplet. At higher surface superheat levels, experiments have shown that the droplet vaporization process becomes chaotic, with the process alternating between rapid vaporization of liquid in contact with the surface and ejection of liquid off the surface by strong vapor recoil forces. For our experiments with water droplets at atmospheric pressure, this regime corresponds to superheat levels ranging from about 35 to 55 deg. C. At the low superheat end of this regime, extremely high mean heat flux levels are achieved, but as superheat further increases, less of the surface stays wetted due to the increasing vapor recoil forces, and heat flux begins to decrease as the boiling process becomes like transition pool boiling with progressively less of the surface in contact with liquid. This exploration of the use of a specialized convolution neural network (CNN) to simultaneously analyze high speed video images and digital data for this high-superheat, near-critical-heat-flux regime of droplet vaporization is of special interest for two reasons. First, this vaporization regime results in high heat flux levels that make it attractive for high heat flux cooling for high-powered electronics. Use of machine learning tools to learn more about the mechanisms of this vaporization regime may open the door to new high flux thermal management technologies. In addition, because of its complexity, the two-phase morphology of the vaporization process in this regime is expected to be a very challenging task for CNN machine learning tools. In this study we conducted deposited water droplet spreading and vaporization experiments that captured digital data input (measured surface superheat, mean heat flux during the vaporization process, wetting contact angle, droplet size, etc.) and images of the droplet vaporization two-phase morphology from high-speed video during each experiment. This paper summarizes our successful development of a specialized hybrid CNN design that is trained using the combination of digital measurements and images obtained in our experiments. This CNN design provides deep insight into correlation between the two-phase morphology and heat transfer performance for this near critical heat flux vaporization regime. It also provides a pathway to a heat transfer performance model that fits the performance data to a high level of agreement. Using data collected from the droplet deposition experiment, this network design has been trained to predict the mean heat flux with a root mean square percent error of only about 2.0% and 8.0% on a training and testing dataset respectively. The hybrid network developed in this research appears to be a promising strategy for analyzing experimental data for physical systems with complex morphology that are best investigated experimentally with a combined use of imaging and digital sensor instrumentation.more » « less
-
Abstract Combining high-speed video cameras and optical measurement techniques with digital sensors controlled by a data acquisition system can provide an effective means of exploring boiling process thermophysics and heat transfer mechanisms. Imaging can provide qualitative and quantitative information that complements data provided by temperature, pressure, and other sensors. This paper summarizes the results of an exploration of machine learning strategies to optimally combine and analyze boiling process images and digital sensor information from experiments. We specifically sought a convolution neural network (CNN) to analyze the vaporization of deposited water droplets on superheated surfaces that may have varying degrees of nucleate boiling effects. Two specialized CNN models were developed in this study that can simultaneously analyze both image and digital data. One of our CNN model designs (case B) was trained to take an image of the vaporization process and nonthermal digital data as input and predict thermal heat transfer performance. This model predicts performance remarkably well given its nonthermal inputs, matching independent heat flux test data to a root-mean-square percent error (RMSPE) of 10.3%. This model appears to learn how the variations of nucleate boiling, vapor recoil activity, and local dryout over the surface vary with surface temperature and/or heat flux from changes in boiling system images. We also describe a CNN model (case C) that takes digital nonthermal data, digital thermal data, and image information and provides a high-fidelity prediction of vaporization heat transfer performance. This model predicted performance very well—better than our conventional fit to data (case A) and on par with best fits to quality nucleate boiling heat transfer data in the literature. This type of trained model fit independent heat flux test data to an RMSPE of 5.8%. Our results indicate that training this type of model which predicts performance from input image information and digital operating condition thermal data makes the resulting predictive model more accurate and robust. The successful use of the hybrid CNN models described here suggests that there is a strong correlation between two-phase morphology variations and changes in heat transfer performance. The hybrid CNN modeling approach developed in this research appears to be a promising strategy for analyzing experimental data for physical systems that are best investigated experimentally with combined use of imaging and digital sensor instrumentation. Possible use of this type of modeling in other systems is also discussed.more » « less
-
Recent demonstrations of melting in the metal–organic framework (MOF) family have created interest in the interfacial domain between inorganic glasses and amorphous organic polymers. The chemical and physical behaviour of porous hybrid liquids and glasses is of particular interest, though opportunities are limited by the inaccessible melting temperatures of many MOFs. Here, we show that the processing technique of flux melting, ‘borrowed’ from the inorganic domain, may be applied in order to melt ZIF-8, a material which does not possess an accessible liquid state in the pure form. Effectively, we employ the high-temperature liquid state of one MOF as a solvent for a secondary, non-melting MOF component. Differential scanning calorimetry, small- and wide-angle X-ray scattering, electron microscopy and X-ray total scattering techniques are used to show the flux melting of the crystalline component within the liquid. Gas adsorption and positron annihilation lifetime spectroscopy measurements show that this results in enhanced, accessible porosity to a range of guest molecules in the resultant flux melted MOF glass.more » « less
-
Eshelby’s equivalent inclusion method (EIM) has been formulated to solve harmonic heat transfer problems of an infinite or semi-infinite domain containing an inclusion or inhomogeneity. For the inclusion problem, the heat equation is reduced to a modified Helmholtz’s equation in the frequency domain through the Fourier transform, and the harmonic Eshelby’s tensor is derived from the domain integrals of the corresponding Green’s function in the form of Helmholtz’s potential. Using the convolution property of the Fourier space, Helmholtz’s potential with polynomial-form source densities is integrated over an ellipsoidal inclusion, which is reduced to a one-dimensional integral for spheroids and an explicit, exact expression for spheres. The material mismatch in the inhomogeneity problem is simulated by continuously distributed eigen-fields, namely, the eigen-temperature-gradient (ETG) and eigen-heat-source (EHS) for thermal conductivity and specific heat, respectively. The proposed EIM formulation is verified by the conventional boundary integral method with the harmonic Green’s function and multi-domain interfacial continuity, and the accuracy and efficacy of the solution are discussed under different material and load settings.more » « less
