Often, in optical and electronic equipment, heating is concentrated in very small regions, and, because of materials constraints, cooled walls must be as thin as possible. Also, for efficiency, many high-flux cooling designs involve forced-convection boiling heat transfer. Though efficient, a design with boiling heat transfer can be difficult for it must properly account for the complexities of the boiling flux-temperature relationship. Of concern is locating the point of incipience to boiling and the point of maximum nucleate boiling heat flux, Critical Heat Flux (CHF), and describing the complex behaviors in the vicinities of these points. Characteristics of boiling near these points are discussed in terms of boundary layer behavior. Changes in either the heater size or the wall thickness affects the boiling curve, particularly the CHF behavior. Results from experiments which were conducted on small, heated regions are discussed in light of their application to the design of high-power optical and electronic devices. The effects of flow velocity, subcooling, pressure, heating length, dissolved gas content, and flow streamline curvature are addressed.

Several existing heat transfer models for uniformly heated channels were examined for: (1) accurate representation of the boiling curve, and (2) characterizing the local heat transfer under high heat flux (HHF) conditions. Comparisons with HHF data showed that major correlation modifications were needed in the subcooled partial nucleate boiling (SPNB) region. Since the slope of boiling curve in this region is important to assure continuity of the HHF trends into the fully developed boiling region and up to the critical heat flux, accurate characterization in the SPNB region is essential. Approximations for the asymptotic limits for the SPNB region have been obtained and have been used to develop an improved composite correlation. The developed correlation has been compared with 365 water and 155 freon-11 data points. For the local heat transfer coefficient and wall temperature, the over-all percent standard deviation with respect to the data was 19% and 3%, respectively, for the high velocity water data.

This paper considers the potential of jet/diamond systems for removing localized high heat fluxes. Diamond substrates are compared to other candidate materials. Limits on usable thermal resistances and heat transfer rates are estimated.

An application of the cylindrical HV to cool dummy loads of electron cyclotron heating (ECH) systems of the DIII-D fusion machine is presented. These devices are subjected to heat fluxes of several hundred W/cm² and can be cooled by water. Analysis shows that HV cooling can handle the heat fluxes without exceeding temperature limits. The cylindrical HV may also be used for cooling of fusion diverters.

A pumped single-phase porous metal cooled microwave cavity design is being evaluated for use in a high-power gyrotron. A small-scale porous metal cooled test article was designed, built, and tested on a Phase I SBIR program. The program was funded by the United States Department of Energy. A copper/water porous metal heat exchanger test article was fabricated and was subsequently tested at absorbed heat fluxes up to 7.4 +/- 0.3 kW/cm² before failure occurred. Multiple tests were successfully completed at heat fluxes of 4.0 to over 6.0 W/cm² with no signs of failure. The test article design, coolant parameters, test methodology, and test results are presented. The results of this work show the potential of porous metal cooling to solve a number of high heat flux cooling problems; several such applications are described.

The early work by Tuckerman and Pease on high-performance heat sinks has provoked numerous uses of them as a means of cooling electronic components. This paper is a review of 73 available works by academics and/or researchers who have used the original microchannel heat sink concept or have extended its design or use to turbulent flow, variable channel/fin width ratios, entrance effects, problem generalization, refined optimization, and macrochannel heat exchangers. Various investigators have reported design methodologies, designs, prototypical concepts, comparative and experimental results, and bold innovations on the use of microchannel heat sinks. Several computer codes are identified which perform the calculations needed to describe a particular heat sink: one examines laminar flow devices only, another investigates the effects of a myriad of parameters on the overall thermal resistance of the heat sink and a third offers a design methodology which determines the important dimensions for an optimal heat sink with no restrictions regarding flow type or channel to fin width ratios.

Microchannel heatsink reliability can be affected by thermal stresses that arise due to temperature gradients between the base and fin and along the fin length. These stresses are combined with the bonding stresses that arise in attaching components at elevated temperatures to the silicon heatsink and then cooling the structure to the cryogenic operating temperatures. These bonding stresses are potentially large because of the differences in the values of the coefficients of thermal expansion in silicon heatsink material, and the attached component materials. The stress results shown are for a 17:1 aspect ratio heatsink cooled in liquid nitrogen. The temperature gradients are a result of a surface heat flux of 1.3 kW/cm², approximating the heat dissipation of an rf power chip. The chip is connected to an aluminum nitride substrate, then the chip and substrate module are attached to the heatsink at a bonding temperature of 600 degree(s)K, as for a gold-tin eutectic bond. The stresses are shown to be within the allowables of the materials involved.

Detailed performance results for an efficient and low thermal impedance laser diode array heatsink are presented. High duty factor and even cw operation of fully filled laser diode arrays at high stacking densities are enabled at high average power. Low thermal impedance is achieved using a liquid coolant and laminar flow through microchannels. The microchannels are fabricated in silicon using an anisotropic chemical etching process. A modular rack-and- stack architecture is adopted for heatsink design, allowing arbitrarily large 2-D arrays to be fabricated and easily maintained. The excellent thermal control of the microchannel heatsinks is ideally suited to pump array requirements for high average power crystalline lasers because of the stringent temperature demands that are required to efficiently couple diode light to several-nanometer-wide absorption features characteristic of lasing ions in crystals.

The advanced fusion machines such as TPX, NET, and ITER have to be designed to handle a heat flux of about 5 to 15 MW/m² in the diverter region. The present conceptual designs use water cooling. However water leaks will have very serious consequences in these machines. Cooling with a gas like helium is a very attractive alternative, if the pumping power can be limited to a reasonable value. Different concepts to cool diverter by helium gas were compared. It was found that it is feasible to remove significant steady state heat flux (10 to 20 MW/m²) by using helium at a pressure of 4 MPa (580 psia) and with pumping power less than 0.5% of the power removed, by using optimized designs. From pumping power consideration, various concepts rank in the following order: offset fins (best), fins, jets, 3-D roughness, 2-D roughness, smooth tubes (worst). A module based on this study has been designed and fabricated for a steady state operation at 10 MW/m² and was tested at the High Heat Flux facility at Sandia National Laboratory. This paper also presents some preliminary studies of helium cooled ITER diverter.

Recently, the promise of major reductions in the price of synthetic diamond has allowed its consideration for use in electronic thermal management applications, such as MCM substrates, requiring many tens of grams of material (e.g., 100 mm square substrates, 1 mm thick). Because of the combination of extremely high thermal conductivity and electrically insulating nature, diamond is an ideal IC packaging or MCM substrate material; diamond can cool, through lateral thermal conduction to board edges alone, MCMs having many hundreds of watts power dissipation, and at the same time allow for a high density of vertical via interconnects through the diamond substrate. This diamond 3-D packaging approach is capable of handling, with a modest temperature rise, a power density of 200 Watts per cubic inch, which corresponds to a potential computational density of over 1300 MFlops per cubic inch or 83 GFlops in a 4" X 4" X 4" cube with current processor technology.

Thermal management has become a limiting problem for both military and commercial systems. High thermal conductivity graphite fibers provide a breakthrough in the efficiency of conductive cooling. The fibers also have exceedingly high specific modulus and a negative coefficient of thermal expansion; these properties can be used to tailor the thermal expansion of composite materials to virtually any required level. This paper discusses the nature of the fibers, the translation of fiber properties into composites including metal, organic, and carbon matrix. Emphasis is on applications for the resulting high thermal conductivity composites. These applications include electronic packaging, aircraft, and spacecraft.

Analytical expressions for the thermal shock resistance (TSR) of IR missile domes involve a phenomenological stress factor that is not yet available but can be either extracted from the results of aerothermal shock testing or estimated on the basis of model-related considerations. The primary purpose of this contribution is to derive a simple formula for the allowable heat flux at the stagnation point, under conditions such that the boundary-layer flow remains laminar over the entire dome surface.

Flat Beryllium tiles brazed to actively cooled CuCrZr withstand power densities up to 17 MW/m². Tiles with 2 mm thickness have been cycled for 1000 pulses at power densities between 12 and 14 MW/m². Flat CFC tiles brazed to OFHC copper failed at power densities of 13 MW/m² with long (10 s) pulses. The fault developed from an area which showed signs of overheating from the start of the test and led to a complete detachment of one tile. For shorter pulse duration (2.5 s) higher power densities (17 MW/m²) could be applied. CFC cubes brazed around a central cooling pipe (monoblocks) have been tested to 15 MW/m². The power density was limited by the peak surface temperature of the tiles, which exceeded 1500 - 2000 degree(s)C. None of the tiles failed during the test, but there were considerable temperature variations from tile to tile.

A bonding strategy was developed for joining carbon/carbon composites and beryllium to copper. The methodology used was first to control the interface properties by employing ion beam enhanced deposition (IBED) on the surface of the specimen. This process allowed the formation of a graded interface layer to alleviate the mismatch in coefficient of thermal expansion and an additional surface layer to be deposited for subsequent bonding. Joining was then performed using a traditional diffusion bonding technique or an infrared assisted bonding technique. The integrity of the joint was determined by nondestructive measurement of the thermal conductance through the joint. Extremely high values of thermal conductance were registered, indicating that very high quality joints were obtained.

While carbon fibers are electrically conductive, the conductivity is poor compared to metals, and is anisotropic to a degree depending on the degree of graphitic perfection and orientation in the fiber. A novel graphite fiber reinforcement known as vapor grown carbon fiber (VGCF) is being used to fabricate a high thermal conductivity carbon/carbon composite for use as a plasma facing material. Composite processing has been developed, resulting in recent fabrication of a 1D composite having a thermal conductivity of 910 W/m-K. Further advances in composite fabrication are under study which may provide for 1D composites having thermal conductivity exceeding 1000 W/m-K.

Results of experience with single crystal monochromators gathered at the ESRF during the past year using the first undulator and wiggler beams are presented. It was found that: Cryogenic cooling of Si is a reliable technique and fully adequate for present and very likely also for future ESRF undulator beams. Diamond crystals can be routinely used as well, both in reflection and in transmission geometry, whenever very high resolution is not required. Side- cooling is presently sufficient. Side-cooled thin Si crystals appear to be suitable for power densities of a few W/mm² and 16 W total power and must be supported for extracting the heat at higher loads. Cooling by small fins gave good preliminary results on a wiggler beam with reduced power but further confirmation is needed for higher power and mounting strain should be still decreased. Combining several techniques may be necessary in the future, e.g., cryogenic cooling of thin crystals.

A silicon block (typical size 100 X 100 X (20 - 50) mm³) cooled by liquid nitrogen has been studied with various incident power densities and spot sizes on the surface. Gaussian power distribution was assumed. Both bottom cooling and side cooling have been considered. The thermal slope error has been minimized by optimizing the cooling conditions (cooling coefficient and bulk temperature of liquid nitrogen) and the thickness of the silicon block. Finite element analysis has been used because the material properties ((alpha) , k) of silicon are highly non linear. The limits of absorbed total power and power density are estimated for both undulator and wiggler beams with various spot sizes and for the requirement in terms of thermal slope error. Correlations between thermal slope error and power, power density have been established.

This paper describes some of the aspects of the cryogenic optics program at the Advanced Photon Source (APS). A liquid-nitrogen-cooled, high-vacuum, double crystal monochromator is being fabricated at Argonne National Laboratory (ANL). A pumping system capable of delivering a variable flow rate of up to 10 gallons per minute of pressurized liquid nitrogen and removing 5 kilowatts of x-ray power is also being constructed. This specialized pumping system and monochromator will be used to test the viability of cryogenically cooled, high- heat-load synchrotron optics. It has been determined that heat transfer enhancement will be required for optics used with APS insertion devices. An analysis of a porous-matrix-enhanced monochromator crystal is presented. For the particular case investigated, a heat transfer enhancement factor of 5 to 6 was calculated.

With the help of several computer codes especially developed or adapted for this application, the energy losses in the individual components of a high resolution undulator beamline for BESSY II were quantitatively calculated and their affect on the performance of the beamline determined. The critical portion of the beamline, a Kirkpatrick-Baez system in front of the entrance slit, can be designed so as to withstand the deleterious effects of the heat loading with essentially no loss of intrinsic brightness of the source. The absorbed heat load profile has been determined in a dense grid on the mirror surfaces by taking the variation of the spectral power distribution from the undulator and the reflectivity of the surfaces into account. The influence of the angle of incidence on the first horizontally deflecting mirror as well as of a vertical aperture in front of the beamline has also been studied.

The high brilliance of third generation synchrotrons makes the cooling of mirrors highly critical, but a careful matching between the critical energy (epsilon) _crit of the insertion device -- controlled by gap tuning -- and the energy cut-off (epsilon) _cut of the mirror allows us to lower the heat load considerably. A universal relationship between the absorbed power and the ratio (epsilon) _cut/(epsilon) _crit is derived numerically. The optical configuration of the beamline can be optimized by taking into account the competition between optical aberrations, figure errors, and the demagnification. An analytical expression of the optimal demagnification for high flux applications is given.

A cost-effective high heat load mirror based on the compensation for thermal deformation by the bimetallic strip effect has been studied. A theoretical description of the heat loaded bimetallic strip is given. Careful selection of two materials having complementary thermal properties allows efficient correction of thermal bending. Such a system could also be used as a very simple bender by tuning the temperature gradient across the two layers.

We investigated the performance of a cryogenically cooled silicon monochromator crystal exposed to high power x-ray undulator radiation. The heat transfer in this nonlinear material was studied analytically by approximating the thermal conductivity and then scaling relations for the temperature distributions with the cooling temperature and the power load were found. The strain distributions in this nonlinear material were studied analytically by approximating the thermal expansion coefficient. The broadening of the rocking curve was found to be determined, to the first order, by the maximum temperature and a load factor (gamma) which is determined by the properties of the source and the crystal, and independent of the optical geometry. Major conclusions were verified through numerical analysis with the program ANSYS. We concluded that cryogenically cooled silicon Laue monochromator should work with Undulator A at the Advanced Photon Source.

The x-ray monochromator SX700 High Flux design is based on the successful basic SX700 system which is installed in various beam lines. It consists of a rotatable plane tilting mirror, a rotatable plane grating, and a fixed ellipsoidal focusing mirror. By using two different gratings, the SX700 covers the energy range from 5 eV to 2800 eV. Together with the control software it is possible to move the tilting mirror together with the rotating grating continuously in such a way that a fixed source focus is achieved over a wide energy range (from 38 eV to 2800 eV for 1220 l/mm and from 11.4 eV to 860 eV for 366 l/mm). The most recent software update now includes any user defined angle relation between mirror and grating. To withstand the high heat loads of synchrotron beams without any compromise in optical performance, the SX700 monochromator was upgraded. This is achieved by a proper material selection for the optical components (plane mirror and grating) and by the installation of a respective cooling system for these components. In this paper the development of the chosen design solution and first beam line results are reported.

Subcooled flow boiling in water is thought to be advantageous in removing high heat load of more than 10 MW/m². Characteristics of the critical heat flux (CHF), which determines the upper limit of heat removal, are very important for the design of cooling systems. In this paper, studies on subcooled flow boiling CHF, which have been conducted by the authors, are reported. Experiments were conducted using direct current heating of stainless steel tube. For uniform heating conditions, CHF increment in small diameter tubes (1 - 3 mm inside diameter) and the CHF characteristics in tubes with internal twisted tapes were investigated, and also the existing CHF correlations for ordinary tubes (more than 3 mm inside diameter) were evaluated. For peripherally non-uniform heating conditions using the tube, whose wall thickness was partly reduced, the CHF for swirl flow was higher than the CHF under uniform heating conditions with an increase of the non-uniformity factor.

In a tokamak fusion reactor, the frequency of occurrence and the severity of a plasma disruption event will determine the lifetime of the plasma facing components. Disruptions are plasma instabilities which result in rapid loss of confinement and termination of plasma current. Intense energy fluxes to components like the first wall and the diverter plate are expected during the disruptions. This high energy deposition in short times may cause severe surface erosion of these components resulting from melting and vaporization. Coatings and tile materials are proposed to protect and maintain the integrity of the underneath structural materials from both erosion losses as well as from high thermal stresses encountered during a disruption. The coating thickness should be large enough to withstand both erosion losses and to reduce the temperature rise in the substrate structural material. The coating thickness should be minimized to enhance the structural integrity, to reduce potential problems from radioactivity, and to minimize materials cost.

This paper describes a UHV-compatible double crystal monochromator with independent drives for 2 linear and 2 angular crystal motions. Precise angular crystal positioning is achieved by using a spindle with a double gimbal mechanism, which converts linear motion of 0.1 micron to an angular motion of 0.042 arcsec. In order to decrease thermal distortions the crystal intercepting white beam is watercooled.

Four kinds of large mirrors, which are 0.8 m SiC flat, 1 m SiC flat, and two 1 m Si cylindrical mirrors (coated with Pt or Ni), are developed for high brilliance synchrotron radiation with the cooperation of the suppliers. The reflectivity, surface finish, and surface figure are characterized by not only laser interferometers but also x-ray at ISAS 36 m long beamline.

The photon shutters (PS) on the insertion device front end of the beamlines at the Advanced Photon Source (APS) are designed to fully intercept powerful 7-GeV undulator radiation. Traditional materials (oxygen-free copper and Glidcop) are used in their construction. The heat flux from the undulators is enormous. For example, in the later phase of the project, the first photon shutter (PS1) placed at a distance of 17 m from the Undulator A source is subjected to 1400 W/mm² at normal incidence with a total power of 11.4 kW. Our analysis of PS1 indicates that the face plate made of either graphite or beryllium retains its integrity in most of the cases. The maximum effective stress of the absorber plate (made of annealed OFHC) exceeds the yield strength (50 MPa) except in the case of an absorber with a 10-mm graphite face plate.

High power wiggler or undulator beamlines at synchrotron radiation sources need devices that stop or reduce the size of the beam. This paper describes a new photon shutter/slit assembly that takes advantage of the high collimation of the x-ray beam in the vertical direction. In the center of a water-cooled copper block, an opening has been configured that has an aperture of 10 X 120 mm² with a length of 100 mm. This copper block is rotated around the center of the length axis. There are two modes of operation for the rotation. One mode switches the beam on and off. The other controls the rotation, which results in a variable vertical slit. The angle of this rotation reduces the power density on the cooled surface of the slit by a factor larger than ten.

In this paper we address the transient temperature distribution and subsequent optical distortion resulting from an annular high power continuous wave laser beam on to coated, non-cooled optical components. The transient temperature distribution is described using The Reverse Thermal Wave Transform. The optical components are uncooled and subject to thermal shock. The standard equations are used for evaluating thermal shock as a function of time, coating expansion coefficient, and material for both substrate and coatings.

To be useful in high power CO₂ laser systems, both the substrate and the antireflective (AR) coatings should exhibit a high laser induced damage threshold which depends principally on both mechanical and optical properties, especially on optical absorption of materials of the substrates and AR coatings.

Several temperature field solutions due to bending magnet and undulator x-ray heating are developed and presented in this paper. The Gaussian power distribution is simulated as the bending magnet whereas a Gaussian-parabolic type of power distribution is used for the undulator/wiggler heating. The heating on a 2-D plane, 3-D block, thin disk, infinite wedge plane, infinite wedge block, and beryllium-copper composite are analyzed. Parametric studies are also included to determine the optimized temperature.

Solar thermal energy systems can produce heat fluxes in excess of 10,000 kW/m². This paper provides an introduction to the solar concentrators that produce high heat flux, the receivers that convert the flux into usable thermal energy, and the instrumentation systems used to measure flux in the solar environment. References are incorporated to direct the reader to detailed technical information.

The major concern in advancing the state-of-the-art technologies for hypersonic vehicles is the development of an aeropropulsion system capable of withstanding the sustained high thermal loads expected during hypersonic flight. Even though progress has been made in the computational understanding of fluid dynamics and the physics/chemistry of high speed flight, there is also a need for experimental facilities capable of providing a high heat flux environment for testing component concepts and verifying/calibrating these analyses. A hydrogen/oxygen rocket engine heat source has been developed at the NASA Lewis Research Center that is capable of providing heat fluxes up to 450 W/cm² on flat surfaces and up to 5,000 W/cm² at the leading edge stagnation point of a strut in a supersonic flow stream. Gas temperatures up to 3050 K can also be attained. Two recent experimental programs conducted in this facility are discussed.

Intense heat sources are needed to address new manufacturing techniques, such as, the Rapid Thermal Process for silicon wafer manufacturing. The current technology of high heat flux sources is the laser for its ability to do welding and cutting is well-known. The laser with its coherent radiation allows an image to be focused down to very small sizes to reach extremely high heat flux. But the laser also has problems: it is inefficient in its use because of its singular wave length and brings up OSHA safety related problems. Also heavy industrial manufacturing requires much higher total energy in addition to the high heat flux which makes the current laser system too slow to be economical. The system I am proposing starts with a parabolic curve. If the curve is rotated about the axis of the parabola, it generates the classical parabolic reflector as we know it. On the other hand, when the curve is rotated about the chord, a line passing through the focal point and perpendicular to the axis, generates a new surface called the Orthogonal Parabolic Surface. A new optical reflector geometry is presented which integrates a linear white light (continuum spectra) source through a coherent path to be focused to a very small area.

In this paper we address the transient temperature distribution and subsequent optical distortion resulting from an annular high power continuous wave laser beam on to uncoated, non-cooled optical components. The transient temperature distribution is described using The Reverse Thermal Wave Transform. The optical components are uncooled and subject to thermal shock. The standard equations are used for evaluating thermal shock as a function of time, thickness, and material.

Synchrotron x-ray windows are vacuum separators and are usually made of thin beryllium metal. Filters are provided upstream of the window to filter out the soft x-rays to protect the window from overheating and failing. The filters are made of thin carbon products or sometimes beryllium, the same material as the window. Because the window is a vacuum separator, understanding its potential structural failure under thermal load is very important. Current structural failure models for the brazed windows and filters under thermal stresses are not very accurate. Existing models have been carefully examined and found to be inconsistent with the actual failure modes of windows tested. Due to the thinness of the filter/window, the most likely failure mode is thermal buckling. In fact, recent synchrotron tests conducted in Japan on window failures bear out this position. In this paper, failure criteria for filters/windows are proposed, and analyses are performed and compared with the experimental results from various sources. A consistent result is found between the analysis and reported experiments.

There are obvious differences but some surprising similarities between high-power laser mirrors designed for use in the visible and infrared wavelengths and synchrotron mirrors designed for use at x-ray wavelengths. The use of synchrotron mirrors at grazing incidence results in a relaxation of figure, surface microroughness, and thermal heating tolerances relative to the wavelength of nearly two orders of magnitude. As a result, the tolerances become roughly comparable to those desired for high-power laser mirrors for the visible region of the spectrum. Experience gained over the years by laser mirror designers in substrate design, the influence of metallic coatings on thermal and optical performance and limiting values for surface microroughness, may be helpful to designers of synchrotron systems. Exploitation of the promise of the new synchrotron systems still represents a challenging problem for researchers. The specifications and operational approaches to meeting them for the new Advanced Photon Source at Argonne National Laboratory are given.

A-high-heat-load, horizontally deflecting/focusing mirror is designed for installation on an APS undulator beamline. The main design objective has been to keep the total tangential RMS slope error, including the thermally induced component, to less than 2 (mu) rad with an absorbed beam power on the mirror of 2 kW and a peak flux of 3.2 W/mm². Extensive examination of various design parameters and detailed thermal/structural analyses has resulted in a mirror design that meets the tight slope-error requirement. Design features include a silicon substrate, a tailored pin-post cooling scheme, a moderate coolant flow rate, primary and secondary cooling areas, a multi-strip coating on the reflecting surface, and inlet/outlet cooling manifolds through an attached Ni-Fe mounting structure.

A new ultra-high-vacuum (UHV) compatible coaxial water cooling structure has been designed for the Advanced Photon Source high-power bending-magnet front-end photon shutters. Laser- beam thermal-simulation test results show that this new cooling structure can provide more than 1.56 kW total power cooling capacity with 12.3 W/mm² maximum surface heat flux. The maximum surface temperature is lower than 116 degree(s)C.