Phase-change processes, such as boiling and evaporation, generally represent very effective modes of heat transfer. However, the demands of modern thermal systems have required the development of methods to achieve still higher performance by use of enhanced heat transfer techniques. While heat fluxes above 10⁸ W/m² have been accommodated in carefully controlled situations, the fluid and convective conditions usually dictate maximum heat fluxes several orders of magnitude lower. Two major contemporary areas, enhanced surfaces for pool boiling and enhanced surfaces for forced convection vaporization, are discussed. They illustrate the opportunities and challenges of enhanced heat transfer technology.

The advent of LSI/VLSI systems has made possible the development of advanced electronic systems operating in the multi-GHz regime. such high speed systems will be of multichip construction to increase miniaturization, packing, and heat dissipation density. Similar advances in high laser-power optics have resulted in significant increases in heat flux density. The stringent temperature uniformity specifications on these systems demand innovative means of applying state-of-the-art technology in enhancing heat removal. Promising cooling techniques that will meet the future thermal control requirements for these electronic and optics packages are presented. These concepts involve the use of microchannel, droplet impingement, jet impingement, and flow boiling in straight or curved channels.

Studies have been performed in spray cooling with phase change using water as the coolant. A gas atomizing nozzle was used with both air and steam as the driving gases. The effect of gas atomizing pressure and liquid flow rate on the heat transfer, specifically, the critical heat flux is studied. Spray droplet size and velocity, liquid film thickness and flatness were measured using phase Doppler, Fresnel diffraction, and holographic techniques, respectively. The effect of spray characteristics on film thickness and heat transfer is discussed.

Several general categories of porous media heat exchangers have been successfully demonstrated to meet the challenges of high heat flux cooling. Demonstrated types of porous media heat exchangers include mechanically pumped single-phase and two-phase porous media heat exchangers, as well as capillary-pumped (heat pipe) two-phase designs. These designs have demonstrated the capability to meet and exceed typical cooling requirements for high heat flux applications. A brief description of each heat exchanger type is presented, as well as a discussion of the recognized applications for each approach. An overview is presented of the current state-of-the-art for each category, and an assessment is given of the capabilities of each approach to meet the needs of several high heat flux cooling applications.

Excellent thermal performance of a silicon microchannel heatsink is demonstrated using liquid nitrogen as the coolant over a wide range of heat loads up to 1 kW/cm². This performance is partly due to the order of magnitude increase in the thermal conductivity of silicon near 77 degree(s)K compared to room temperature. Subcooled boiling of the liquid nitrogen in the microchannel heatsink further enhances the thermal performance but makes the thermal resistance a nonlinear function of heat load. For a 500 W/cm² heat load, a thermal resistance of 0.052 cm^2*0C/W for a 9:1 aspect ratio heatsink was measured.

Cooling systems for very compact electronic components and computer chips are being miniaturized to meet the need for smaller overall packaging. One of the important present directions has been to use laminar flow in very small channels with hydraulic diameters in the sub-millimeter range to get high heat transfer coefficients with low pressure drops. It has been speculated that there might be some advantage to having convective subcooled boiling (SCB) occur in the micro-channels. As a first step in the evaluation of the utility of subcooled boiling in these micro-channels, a model has been developed for subcooled boiling in sub-millimeter diameter microtubes subject to uniform heat flux. This model builds on a previously well-validated computer code for convective subcooled boiling in tubes down to 1.57 mm inner diameter. The basic features of the new microtube model are described and some predictions using this model for 0.3 mm and 0.1 mm microtubes subject to a high heat flux of 10 MW/m² are given.

Heat transfer experiments were carried out to investigate the boiling characteristics of LN2 in a narrow channel made by silicon crystal, which simulated one of the cooling channels of the first crystal of DXM. Heat transfer coefficients in nucleate boiling and DNB heat fluxes in the narrow channel were obtained and compared with some existing correlations. These values are used in the numerical simulation to calculate the thermal distortion of the crystal more accurately.

In 1988 I presented a paper, `Fly's Eye Modular Optic,' in the Los Angeles Symposium that described an optic for high power laser systems that provided for a modular system of hexagonal components that were independently cooled using a high velocity jet pointed normal to the back surface of the optical faceplate. In this paper we look at the use of diamond optical materials in concert with high velocity jet impingement heat transfer of various liquid metal mediums. By using this combination of techniques and materials we can push the laser damage threshold of optical components to even higher levels of absorbed flux density. The thrust of this paper is to develop a theoretical model for use on optical elements subject to very high continuous flux density lasers and to evaluate the use of commercial diamond substrates with conventional optical thin films and conventional substrates with CVD diamond films. In order to assume the very high absorbed flux densities, it is necessary to have a heat transfer technique capable of maintaining the optical component at a stable temperature and below the damage threshold of the optical materials. For the more common materials, thermal shock and subsequent failure in bi-axial shear have proven to be one of the major constituents of the optical damage. In this paper we look at the thermal shock, vis-a-vis, the melting point of some of the materials.

The installation of insertion devices at existing synchrotron facilities around the world has stimulated the development of new ways to cool the optical elements in the associated x-ray beamlines. Argonne has been a leader in the development of liquid metal cooling for high heat load x-ray optics for the next generation of synchrotron facilities. The high thermal conductivity, high volume specific heat, low kinematic viscosity, and large working temperature range make liquid metals a very efficient heat transfer fluid. A wide range of liquid metals were considered in the initial phase of this work. The most promising liquid metal cooling fluid identified to date is liquid gallium, which appears to have all the desired properties and the fewest number of undesired features of the liquid metals examined.

Proven bulk materials for mirrors with high surface qualities are classical optical glass, quartz, and low thermal expansion materials like ZERODUR^R. New requirements for mirrors with high heat load capabilities have established the need for different bulk materials. Metals, silicon and silicon carbide (SiC) are presently the favorite materials for the manufacturing of optical components for high brilliance synchrotron beam lines. This paper discusses properties of different types of material under manufacturing and applicational aspects.

In polycrystalline CVD diamond of useful macroscopic dimensions, which may be considered for high heat flux applications, thermal conductivity parameters are largely determined by grain size resulting from growth morphology, defects and impurities in the material. Thermal conductivity has been measured in a number of state-of-the-art diamond samples, by the steady state technique, over the temperature range 6 to 400 K. The results are presented, and discussed in terms of microstructural differences between samples. At approximately 30 K, a departure from normal Debye type behavior is observed as a lowering of the predicted conductivity. At higher temperatures, this departure becomes less significant so that above approximately 350 K, where only Umklapp processes contribute to phonon scattering, the measured thermal conductivity is close to that predicted by the model and in good agreement with reference data for natural type IIa single crystal diamond. To account for the observed temperature dependence of conductivity, an additional phonon scattering term is used which may be described as Rayleigh scattering at low temperature by defects of 0.7 to 1.3 nm in size.

A diverter unit, which consists of carbon armors brazed to a copper cooling channel, is under development for fusion devices. Isotropic graphite (IG-430U) and CFC (CX-2002U) are used for the armor, and a copper for the cooling tube. A technique named `dissolution and deposit of base metal' was employed for brazing. The reliability of the brazed components was evaluated both by a 4-point bending test and a thermal shock test. According to the results of the 4-point bending test under the temperature ranged from RT to 800 degree(s)C in a vacuum, it was found that the strength of the brazed surface at RT was maintained up to the higher temperature, 600 degree(s)C. A high heat load test has also been performed on the brazed sample in order to find out whether the samples meet the requirement of the diverter plates of LHD. Active Cooling Teststand (ACT:NIFS) with an electron beam power of 100 kW was used. In LHD, it is presumed that the maximum heat flux is 10 MW/m². In addition, the surface temperature of the diverter has to be kept below 1200 degree(s)C to avoid RES, by active cooling. The heat load test showed that the brazing components of CX-2002U (flat plate type CFC-Cu brazed) were stable at 1300 degree(s)C under a heat flux of 10 MW/m², when the flow velocity of cooling water was 6 m/s. No damage nor deterioration was found at the brazed zone after the heat load test.

The actively cooled version of the JET diverter plates use a finned structure as a heat exchange surface between the copper alloy base plate and the cooling water (hypervapotron). The turbulence created by the fin structure gives a remarkably good heat transfer even before the onset of boiling. The boiling heat transfer is stabilized by the colder fin structure. Finite element calculations confirm that the heat transfer can be explained by turbulent boiling heat transfer. Power densities of up to 25 MW/m² can be removed with a pressure drop of 4 bar per meter. Beryllium tiles brazed to the CuCrZr base plate can withstand a pulsed power loading of up to 16 MW/m². Limited in strength is the intermetallic layer between the braze and the beryllium tile. The test sections, mounted rigidly against a strong back, withstood the stress caused by thermal expansion.

JET is the largest magnetically confined fusion experiment in operation today. Plasma facing components of JET made of solid beryllium have sustained for periods of up to 1 second localized fluxes of up to 25 W/mm². In the new phase of operations foreseen for 1993 onwards peak heat fluxes of this magnitude will be swept across surfaces in contact with the plasma in order to reduce erosion and to increase the pulse length. Both low cycle (approximately 10 cycles) and high cycle (approximately 3 X 10³ cycles) fatigue response of prototypes have been studied in a test-bed for heat loads in the range 13 - 25 W/mm² and with peak strain rates of up to 1.5 mm/mm/sec.

Plasma facing components in tokamak fusion reactors are faced with a number of difficult high heat flux issues. These components include: first wall armor tiles, pumped limiters, diverter plates, rf antennae structure, and diagnostic probes. Peak heat fluxes are 15 - 30 MW/m² for diverter plates, which will operate for 100 - 1000 seconds in future tokamaks. Disruption heat fluxes can approach 100,000 MW/m² for 0.1 ms. Diverter plates are water-cooled heat sinks with armor tiles brazed on to the plasma facing side. Heat sink materials include OFHC, Glidcop^TM, TZM, Mo-41Re, and niobium alloys. Armor tile materials include: carbon fiber composites, beryllium, silicon carbide, tungsten, and molybdenum. Tile thickness range from 2 - 10 mm, and heat sinks are 1 - 3 mm. A twisted tape insert is used to enhance heat transfer and increase the burnout safety margin from critical heat flux limits to 50 - 60 MW/m² with water at 10 m/s and 4 MPa. Tests using rastered electron beams have shown thermal fatigue failures from cracks at the brazed interface between tiles and the heat sink after only 1000 cycles at 10 - 15 MW/m². These fatigue lifetimes need to be increased an order of magnitude to meet future requirements. Other critical issues for plasma facing components include: surface erosion from sputtering and disruption erosion, eddy current forces and runaway electron impact from disruptions, neutron damage, tritium retention and release, remote maintenance of radioactive components, corrosion-erosion, and loss-of-coolant accidents.

Current trends in the microelectronic industry suggest that by the mid-1990s successful thermal management will require removal of as much as 500 W and 100 W/cm² from a single chip and in excess of 10 kW and 10 W/cm³ from a multichip module. These cooling requirements pose a serious challenge to today's cooling technology and have spurred extensive research and development of advanced thermal control techniques for microelectronics. The choice of the thermal management strategy for an electronic product has a large impact on its cost, reliability, operating environment, and performance. Thus, while thermal control is just one of several enabling packaging technologies, it deserves and often receives special attention in the development of leading-edge electronic systems. Many observers believe that direct cooling with inert, dielectric liquids may become the method of choice for electronic systems of the late 1990s. This presentation begins with a review of the trends in IC technology and the state-of-the-art in single chip packages and multichip modules. It then examines the packaging development process and defines the future requirements, options, and limits of thermal management techniques. Attention is then turned to direct liquid cooling, including implementation schemes, governing phenomena, research issues, and prospects for widespread implementation.

A graphite foil heat absorber protecting a beryllium window from thermomechanical damage was overheated, sublimated, and became partially transparent by multipole wiggler radiation. A simple theoretical model on heating of the foil-type filter by multipole wiggler radiation was presented, confirmed experimentally, and applied to analyze the failure.

Recent advances in chemical vapor deposition (CVD) technology have made it possible to produce thin free-standing diamond foils that can be used as the window material in high heat load synchrotron beamlines. Numerical simulations suggest that these windows can offer an attractive and at times the only alternative to beryllium windows for use in third generation x-ray synchrotron radiation beamlines. Utilization, design, and fabrication aspects of diamond windows for high heat load x-ray beamlines are discussed, as are the microstructure characteristics bearing on diamond's performance in this role. Analytic and numerical results are also presented to provide a basis for the design and testing of such windows.

The ISOFLOW cooled mirror technology was developed at Itek with the goal of producing a high performance heat exchanger with emphasis on coolant efficiency and low jitter. Design and analysis capabilities, as well as manufacturing processes were developed and demonstrated in 8.0 inch diameter mirrors made from single crystal silicon, silicon carbide, and Ultra Low Expansion glass. The ISOFLOW mirror design has been tested and has demonstrated excellent thermal distortion for very low coolant flowrates and pressure drops. The ISOFLOW mirrors have been polished to better than 3 angstroms rms. The mirrors are bonded with a composite lead borosilicate glass frit blended to match the heat exchanger materials. The heat exchanger was designed to operate over a wide flow range with water or ammonia as the coolant. The high performance turbulent flow regime requires 20 gpm with a 40 psi pressure drop through the mirror. The mirror can also provide low flow low jitter operation with 6.6 gpm and 4.5 psi pressure drop. The demonstrated thermal performance obtained with low flowrates and pressure drops is what makes the ISOFLOW design unique.

Laser facilities for inertial confinement fusion (ICF) experiments require laser and x ray optics able to withstand short pulse conditions. After a brief recall of high power laser system arrangements and of the characteristics of their optics, we present some x ray optical developments.

XUV (10 nm <EQ (lambda) <EQ 100 nm) free-electron lasers (FELs) are potential light sources for materials research applications (to complement future synchrotron light sources) and soft x-ray projection lithography applications for the manufacture of electronic microcircuits. Requirements for compact physical dimensions for these devices, together with relatively low reflectivity mirrors and small optical beam sizes at these wavelengths, make thermal loading of the mirrors of XUV FEL oscillators a potentially serious problem. In this paper, we consider the effects of thermal distortion of the mirrors on the performance of an XUV FEL designed specifically as a 60 nm light source for lithographic applications. We formulate a mathematical model of the relevant physical processes and solve that model with the aid of a 3-D FEL simulation code. The numerical solutions demonstrate the existence of an unusual optical mode of the FEL oscillator in which the steady-state resonator power decreases linearly as the thermal distortion coefficient of the mirrors increases. The stabilizing effect of a scraper mirror as a mode control aperture is discussed. Limits on the magnitude of the mirror thermal distortion coefficient, for which near ideal FEL performance is retained, are discussed.

The purpose of this paper is to present REOSC experience in the field of high energy or high power optics. We review some challenging recent achievements which are: a 1.8 meter mirror assembly for the U.S. Navy, the 1 meter LATEX high energy beam expander telescope, large high energy optics for the CEA laser fusion laboratory, optics for the E.S.R.F. synchrotron prototype adaptive mirror and other beamlines, and the LASCO chronograph off-axis parabola.

Thermally induced optical distortions severely degrade the quality of a laser beam thus reducing the irradiance on target. The purpose of this paper is to formulate procedures for evaluating the impact of spherical aberrations generated by cooled cylindrical optics that transmit cw laser radiation; specifically, the paper concerns edge- and face-cooled laser windows and examines how beam shape and cooling strength affect the performance.

A new high power electron cyclotron heating (ECH) system designed to operate at 110 GHz with a power output of 2 MW has been introduced on DIII-D. All components of the system are capable of handling a 10 second pulse at an interval of 10 minutes. Transmission of ECH power from the source (a millimeter-wave gyrotron) to the plasma through waveguide miter bends may change the polarization and rotate the polarization major axis. Polarizing elements are therefore required to correct for the effect of transmission lines and also to generate proper polarization for coupling into the plasma. Rotating mirrors with different rectangular grooved gratings in two successive miter bends can generate the required wide range of elliptical polarizations. Peak heat fluxes due to ohmic losses in these mirrors are several MW/m² for a 0.5 MW gyrotron power. The complex distribution of losses in the grooves requires a detailed thermal stress analysis to ensure that temperature and stress limits are not exceeded. The desired pulse length is 10 sec, with a cooling time of 10 min between pulses. The temperature rise in the polarizing mirrors must be limited to less than 300 degree(s)C to prevent thermal fatigue and outgassing in the vacuum lines. This paper presents an analysis for the polarizing mirrors for the DIII-D ECH system.

Experiments were performed to investigate the optical and thermal responses of a cryogenically cooled beryllium mirror to pulsed laser irradiation. The mirror has flow channels on its backside for rapid cooling using liquid nitrogen and solidifying argon. The tests were carried out at solid argon temperatures. Laser fluence levels up to 18.5 kW/cm² were incident on the mirror. The optical distortion of the front surface of the mirror was monitored with an interferometer. The thermal response of the mirror was measured using a platinum resistance thermometer and a differential thermocouple on the back of the mirror. The surface distortion due to laser irradiation and the recovery of the mirror to its undistorted figure are described.

The Thermal Distortion Test Facility (TDTF) provides precise measurements of the distortion of laser mirrors which occurs when their surfaces are heated. The TDTF has been used for several years to evaluate mirrors being developed for high power lasers. The facility has recently undergone some significant upgrades to improve the accuracy with which mirrors can be heated and the resulting distortion measured. The facility and its associated instrumentation are discussed.

Third generation synchrotron radiation sources currently being constructed worldwide will produce x-ray beams of unparalleled power and power density. These high heat fluxes coupled with the stringent dimensional requirements of the x-ray optical components pose a prodigious challenge to designers of x-ray optical elements, specifically x-ray mirrors and crystal monochromators. Although certain established techniques for the cooling of high heat flux components can be directly applied to this problem, the thermal management of high heat load x-ray optical components has several unusual aspects that may ultimately lead to unique solutions. This manuscript attempts to summarize the various approaches currently being applied to this undertaking and to point out the areas of research that require further development.

Perfect crystal monochromators cannot diffract x rays efficiently, nor transmit the high source brightness available at synchrotron radiation facilities, unless surface strains within the beam footprint are maintained within a few arcseconds. Insertion devices at existing synchrotron sources already produce x-ray power density levels that can induce surface slope errors of several arcseconds on silicon monochromator crystals at room temperature, no matter how well the crystal is cooled. The power density levels that will be produced by insertion devices at the third-generation sources will be as much as a factor of 100 higher still. One method of restoring ideal x-ray diffraction behavior, while coping with high power levels, involves adaptive compensation of the induced thermal strain field. The design and performance, using the X25 hybrid wiggler beam line at the National Synchrotron Light Source (NSLS), of a silicon crystal bender constructed for this purpose are described.

Thermal distortions of an inclined silicon crystal subjected to the high heat loads expected for a 2.5 m long undulator at the Advanced Photon Source were simulated, and the distortions were then used to calculate (111), (111) (+, -) double crystal rocking curves. The inclination angle for all the simulations was either 80 degree(s) or 85 degree(s). The first crystal was assumed to be a slab that was uniformly cooled on its underside with liquid gallium.

At the ESRF, the power absorbed by the first optical element will vary between a hundred watts on undulator beamlines, and several kilowatts on wiggler beamlines. One consequence is a distortion of the optical element, leading to a degradation of the performances of the beamline optics. Cooling certain materials (Si, Ge, InSb) to low temperature (typically around 125 K for Si) was proposed as a possibility to reduce the thermal deformation. In this paper, we report on the development of liquid nitrogen cooling of silicon crystal monochromators for ESRF beamlines. First, a prototype of a closed liquid nitrogen loop is presented., Then, hydraulic and thermal measurements are given, showing the influence of the flow rate, the pressure, and of the absorbed power.

Adaptive optics systems have proven their efficiency for obtaining high resolution images in the field of large astronomical telescopes. The same techniques can be applied to correct x-ray mirror shapes. The paper describes the principle of an adaptive x-ray mirror system (mirror architecture, measurement subassembly, control unit). In the second part, first results obtained during the design study of the ESRF adaptive x-ray mirror are given. The possibility of achieving cylindrical or elliptical mirror surfaces using adaptive optics techniques are suggested.

The focusing of highly brilliant synchrotron insertion device beams necessitates very precise and stable beamline optics in order to conserve the high brilliance of the source. Moreover, the heat load on mirrors and monochromators may strongly degrade the brilliance in the focus. It is shown here that insertion devices on the borderline between undulators and wigglers 2.5 <EQ K <EQ 5.0 give the highest possible brilliance of the focus over the energy range 0 <EQ (epsilon) (keV) <EQ 40 keV and at a modest overall heat load. Installed on a low-(beta) site the high intrinsic collimation produces a small effective source size which can be imaged using spherical optics with nearly vanishing aberrations.

SiC laminar gratings have been fabricated by reactive ion-beam etching and holographic exposure. Diffraction efficiencies were measured in the region between 17 angstroms and 300 angstroms. The efficiency of the +1 order was 5 to 20% in this region with a small amount of scattered lights. An irradiation test for the SiC gratings was performed by using intense radiation with a power density of 2.7 W/mm² emitted from a multipole wiggler installed into the 2.5 GeV Photon Factory ring. No visible damages and no reduction of the diffraction efficiencies were observed for the uncoated SiC grating after the irradiation, while a remarkable deformation of the deposited Au layer and increase of the scattered light component were observed for the grating coated with Au. In addition to these results, some experimental results of polishability of CVD-SiC are reported.

Preserving the high source brightness of the third generation of synchrotron radiation facilities will require that thermal and pressure deformations of the monochromator crystals be maintained within a few arc-seconds. Recent experiments at the National Synchrotron Light Source (NSLS) have demonstrated the potential of adaptive crystal optics to cope with high power densities. In this technique the crystals deformations are minimized both by an efficient water-jet cooling and by externally applied pressure loads. Thermal deformation can be reduced further with diamond crystals because of their high thermal conductivity and low coefficient of thermal expansion. In this paper we describe the results achieved by optimization of adaptive crystal optics for diamond crystals.

The effect of white wiggler radiation exposure on Mo/BN, W/BN, Mo/B₄C, and W/B₄C multilayers was evaluated by comparing soft x-ray reflectance at an incidence angle of 45 degree(s). All samples were prepared by magnetron sputtering onto SiC substrates. The choice of combinations was based upon annealing tests for Mo/X and W/X (X equals C, Si, BN, and B₄C) multilayers, prior to any exposure tests. The Mo/BN multilayer was found to be the most stable for an exposure time of 10 min under radiation power density of approximately 2.3 W/mm². The observed reduced reflectance for the Mo/BN and W/BN samples was found to be reducible to a radiation-induced effect on the BN layers.

An innovative diverter concept for a thermonuclear fusion reactor is being studied at the Institute for Advanced Materials of the Joint Research Centre of the European Communities. It is made of a single material, an ultrahigh thermal conductivity carbon fiber reinforced graphite (CFC) composite; the coolant is helium gas. When looking at the new diverter concept, it is necessary to make several numerical calculations in order to investigate the influence of the various parameters to be optimized. This paper shows a new numerical procedure which has been developed to avoid a complete 3D-calculation.

Contact thermal conductance is much lower in a vacuum than normal pressure conditions. When we cool SR beamline optics, such as mirror components, a water cooled copper plate which is attached to the optics is used. In order to increase the effect on cooling an indium sheet or liquid metal, such as Ga, In-Ga, is inserted between the optics and the plate. We experimentally obtained contact thermal conductance in a vacuum, which was essential to computing the thermal distribution of optics. And we simulated our cooling system using a beamline mirror component.

Operating at 8 GeV with 1 amp of beam current, the CESR B-factory produces highly intense synchrotron radiation (SR). In response to other vacuum chamber design concerns, a discrete bar absorber has been designed to receive the x rays in the transition region. The absorber is an actively cooled copper extrusion. Finite element analysis has been used to predict the thermal and structural response of the absorber. Elastic/plastic stress analyses predicted localized plastic deformation and since operation is cyclic, fatigue failure has been considered. Preliminary testing using a rastered electron beam for both steady state and fatigue testing has begun.

In SPring-8 beamline for insertion devices heat load problems of the optics become serious. Efficient cooling methods must be improved to make use of the original photon characteristics. Here we indicate some simple and analytical calculations for monochromator crystal. Angle distortions due to the thermal deformation for Si crystal are indicated by using simple formulas. Critical heat flux for liquid nitrogen cooling is estimated to be a few W/mm². The thermal conductance and the loss of head are calculated as a function of fin number in the cooling channel considering the fin efficiency. The possibility of using a diamond crystal for the undulator beamline is also discussed.

The re-emitted spectra and re-emitted power of hard x rays traversing solid material have been formulated analytically. The re-emitted x rays are due to the interaction of the incident x rays with the material through Compton scattering, Rayleigh scattering, and inner-shell vacancy radiative decay. Based on the formula, the re-emitted power of Compton and Rayleigh scatterings are calculated for the x rays from a SPring-8 Multi-Pole-Wiggler (MPW) traversing a graphite filter, a Be window, and an Al or a Cu filter. The percentage of the re-emitted power from an Si crystal due to Compton and Rayleigh scatterings as a function of the incident angle is also discussed.

Conceptual design of an International Thermonuclear Experimental Reactor (ITER) was underway from 1988 to 1990 and an engineering design of ITER started in 1992. Diverter plates of ITER are exposed to severe heat loads and particle fluxes from fusion plasma. A peak heat flux of the diverter plates is estimated at 15 to 30 MW/m². In the present study, monoblock diverter mock-ups have been manufactured and tested in an electron beam test facility in JAERI, which consist of carbon reinforced carbon composite (CFC) materials brazed directly on an OFHC copper tube. Thermal cycling experiments have been carried out with a peak heat flux of 15 MW/m². It has been successfully demonstrated that the present design of the ITER diverter plate can endure a stationary heat load of 15 MW/m² for more than 1000 cycles. Brazing interfaces between the CFC armors and cooling tubes were investigated after the tests.

One of the most critical components on the Advanced Photon Source (APS) insertion device (ID) beamline front ends is the first photon shutter. It operates in two modes to fully intercept the high total power and high-heat-flux ID photon beam in seconds (normal mode) or in less than 100 ms (emergency fast mode). It is designed to operate in ultra high vacuum (UHV). The design incorporates a multi-channel rectangular bar, bent in a `hockey stick' configuration, with two-point suspension. The flanged end is an articulated bellows with rolling hinges. The actuation end is a spring-assisted, pneumatic fail-safe flexural pivot type. The coolant (water) channels incorporate brazed copper foam to enhance the heat transfer, a tube technology particular to the APS. The design development, and material aspects, as well as the extensive thermal and vibrational analyses in support of the design, are presented in this paper.

Flat transmissive material windows are often used in high power cw laser systems to separate a low pressure resonator cavity from a one atmosphere beam train. Recently a requirement was identified for a pressure cell window that would transmit intensity peaks of 31 KW/cm² for 5 seconds, react a differential pressure of 52 atm without active cooling, and focus the incident light as a lens. Budgeted optical distortion levels were 400 nm of rms wave front error and no more than 0.01% conversion of s polarization state into p state through window birefringence. The design of this window is described, including material selection, thermal and structural finite element analysis, Wiebull strength evaluation, measurement and selection of single crystal blanks with low internal stress, evaluation of phase and polarization distortion and mounting considerations.

We describe a computer code that aids in the thermal, structural, and optical analysis of laser optical systems. Inputs to the computer model include energy absorption by coatings and substrates, thermal diffusion, expansion, and heat removal parameters. The program performs steady-state thermal and structural analyses based on geometrical ray traces through the optical system. We show examples to illustrate how this capability can assist in the optical system design process.

The proposed Cornell B-factory offers great possibilities for the next generation of synchrotron sources and at the same time presents many challenges to today's high heat-load cooling technology. One of the challenges, the exit port crotch, is discussed in this article. Its preliminary design consists of a beryllium thin tube with water cooling. Finite elements analysis shows that such a design should be able to tolerate a normal incident power density of 4.7 kW/mm² at 2 meters from bending magnet radiation with a 55 meter radius at 8 GeV and 1 ampere of positron current. Our design consists of a tube that is almost parallel to the synchrotron beam with an incident angle of 2 degree(s). This dilutes the power density by a factor of approximately 30 reducing the heat load to a more manageable level. The tilt angle can be 15 degree(s) if the crotch is located 4 meters instead of 2 meters from the source.

A crotch absorber design for use in the Advanced Photon Source (APS) has been proposed and analyzed. The absorber is placed downstream of sectors S2 and S4 in the curved storage ring chamber and is subjected to a peak power of 120 W/mm² per 100 mA synchrotron radiation. A beryllium ring is brazed on the GlidCop cooling cylinder in order to diffuse the concentrated bending magnet heating. One concentric water channel and two annular return water channels are arranged in the GlidCop cylinder to enhance the cooling. A Bodner-Partom thermoviscoplastic constitutive equation and a modified Manson-Coffin fatigue relation are proposed to simulate the cyclic thermal loading, as well as to predict the thermal fatigue life of the crotch absorber. Results of temperature and stress using finite element computations are displayed and a series of e-beam welder tests and microstructure measurements are reported.

SPring-8 is a third generation synchrotron facility that provides high-brilliance synchrotron radiations in the x ray region. The electron or positron beam with an energy of 8 GeV is stored up to 100 mA in the storage ring with a circumference of 1,435.95 m. Of 44 straight sections in the storage ring, 38 straight sections are available for insertion device beamlines. The insertion devices installed in the 8 GeV storage ring produce higher total power and power density compared with those in the 2 - 3 GeV storage rings since the total power is proportional to the square of the stored electron or positron energy and the aperture of the radiation cone is inversely proportional to the electron or positron energy. Such substantial amounts of heat power radiated from the undulator and wiggler are deposited on the components of the front-end. The heat absorber intercepts the synchrotron radiation beams to protect front-end components placed downstream. In this work, we have studied the heat absorber in the insertion device front-end by the finite element analysis.

Single crystal silicon has been the material of choice for x-ray monochromators for the past several decades. However, the need for suitable monochromators to handle the high heat load of the next generation synchrotron x-ray beams on the one hand and the rapid and on-going advances in synthetic diamond technology on the other make a compelling case for the consideration of a diamond monochromator system. In this paper, we consider various aspects, advantages and disadvantages, and promises and pitfalls of such a system and evaluate the comparative performance of a diamond monochromator subjected to the high heat load of the most powerful x-ray beam that will become available in the next few years. The results of experiments performed to evaluate the diffraction properties of a currently available synthetic single crystal diamond are also presented. Fabrication of a diamond-based monochromator is within present technical means.

The directly water-cooled silicon crystal used on the multipole wiggler beam line BL16 at the Photon Factory has extended cooling fins below the crystal block. Such extended fins are useful to reduce bowing of the crystal due to thermal distortion induced by synchrotron radiation. Experimental and calculated Si(111) and Si(333) double crystal rocking curves are compared in the present paper. Thermal deformation is calculated using a finite element analysis program ANSYS. Fairly good agreement was obtained between experiment and calculation.

The next generation of synchrotron radiation sources will produce very high power and power density x-ray beams. For example, the Advanced Photon Source (APS) under construction at Argonne National Laboratory will produce beams containing up to 5 kW of power and peak normal power densities in excess of 150 W/mm². Normally, the first optical component to intercept the x-ray beam is a crystal monochromator. This device typically uses a single crystal of silicon or germanium as a band-pass filter according to Braggs' law of diffraction. Under the severe heat loading of modern synchrotron beams, the performance of the monochromator is degraded by reducing the photon throughput and increasing the beam divergence. This paper describes the methods used to calculate the thermally induced deformations in standardly configured monochromator crystals using finite element analysis. The results of these analyses are compared to recent experiments conducted at the Cornell High Energy Synchrotron Source (CHESS) using a high-performance, gallium-cooled crystal. Computer simulations can be used to evaluate the performance of high-heat-load x-ray optics for future synchrotron sources.

Thermal and structural analysis was carried out to evaluate the thermal distortion of a silicon crystal, which was irradiated by x ray with high intensity. Liquid nitrogen, gallium, and water were considered as coolants in this numerical simulation. These results were applied to the evaluation of a rocking curve. According to the results obtained by using the present model, cooling by liquid nitrogen was the most advantageous in view of the rocking curve when photon energy was under 20 keV. This advantage, however, disappeared gradually as the photon energy was increased to more than 20 keV.

Our recent experimental and analytical results obtained so far for the surface roughness of the Pt-coated SiC flat mirrors are reviewed. Total reflectivity and angle resolved scattering (ARS) curves were measured using CuK(alpha) x ray for an 800 mm long mirror and three kinds of small mirrors having different surface roughness without heat load. The convolution analysis of ARS curves derived the power spectra of the surface waving of the mirror. The root mean square surface roughness calculated from the integral of the power spectrum is consistent with that estimated from the total reflectivity data. Also, the range of the surface wave number contributing the x-ray reflection was estimated and compared with that measured with the other types of experimental methods, heterodyne interferometer and scanning tunnel microscopy.

High energy insertion devices, used as x-ray monochromators on synchrotron storage rings, require high performance cooling of the primary optic. Monocrystalline materials, such as silicon, inherently provide crystal lattice properties suitable for x-ray diffraction. Silicon, provides excellent thermal and structural properties as well. Free electron lasers also require high performance heat exchanger technology for mirrors. A highly efficient approach to cooling, called `pin post cell,' was developed and fully validated in silicon. However, an additional criteria is imposed on the optic when used as a diffractive crystal. The crystalline structure of the material must not be altered during any step of fabrication. A test program has been completed which evaluated the existing fabrication technology for crystal lattice distortion. X-ray diffraction test results are presented. Currently, we are fabricating an actively cooled crystal that will undergo dynamic testing on the CHESS F2 beamline later this summer.

Electronic and atomic processes under high brilliance x-ray irradiation are simulated for the estimation of heat stored in the optical devices. X-rays interact with materials through photo-electric affects and Compton scattering. As for the photo-ionization cross sections, multi-pole transition matrix elements using the Dirac-Hartree-Slater (D-H-S) wavefunctions are calculated. The Compton scattering cross sections are calculated by the use of the Klein-Nishina formula. Besides these inelastic scattering processes, an elastic process, i.e., the Rayleigh scattering process is also considered by the use of the same wavefunctions. High energy electrons as a result of photo-electric affect or the Compton process can be slowed down by the process of atomic excitation or ionization. These processes are calculated by using the Born-Bethe formula. The electrons with 1 keV energy in the present case are assumed to contribute to the heating of local area, because the mean free path of 1 keV electron is less than 100 nm. Spatial distribution of deposited heat by x-ray irradiation for various substances of optical interests such as Be, C, Al, Si, and so on are obtained. The results are shown with discussion from the physical point of view. Empirical formulae for the spatial distribution are discussed and given in some cases.

Detailed thermal stress analyses of beamline and optical components subject to high heat loads require an accurate determination of the absorbed power profile for accurate prediction of the temperature profile and structural parameters. This is particularly important for high power beams from wigglers and undulators at the third generation synchrotron sources because components must, in general, be designed and maintained with strict mechanical tolerances. The spatial distribution of the power density of an undulator is a rapidly varying function of the energy of the photons suggesting that approximative methods based on a smooth spectral variation may not be valid. In this paper, a fast code for calculating undulator spectra is developed and compared with a wiggler code for approximation of the same spectra. Results from numerical simulations, including the emittance of the stored particle beam, are presented for the absorbed power density in a beryllium window. We find markedly different results for the two models for far off-axis radiation indicating the inadequacy of the wiggler model applied to an undulator spectrum in this case. The wiggler model overestimates the total absorbed power by as much as 82% for the beryllium window.

The high-power and high-flux x-ray beams produced by third generation synchrotron radiation sources such as the Advanced Photon Source (APS) can cause significantly high gas desorption rates on beamline front-end components if beam missteering occurs. The effect of this gas desorption needs to be understood for dynamic vacuum analysis. To simulate beam missteering conditions, optical ray-tracing methods have been employed. The results of the ray-tracing analysis have been entered into a system-oriented vacuum program to provide dynamic vacuum calculations for determination of pumping requirements for the beamline front-ends. The APS will provide several types of synchrotron radiation sources, for example, undulators, wigglers, and bending magnets. For the purpose of this study, the wiggler source was chosen as a `worst case' scenario due to its high photon flux, high beam power, and relatively large beam cross section.

Power density distribution in multipole-wiggler radiation was measured on the beamline BL-28 of the Photon Factory using a photocalorimetric device. The beamline had no beryllium nor graphite windows, which allowed exact comparison between measured power and calculated value using Kim's formula for undulator radiation. Although the measured density was lower by an average 20% at the peak density, it was ascertained that Kim's formula gives roughly the power density of multipole-wiggler radiation. In addition, the photocalorimetry was found to be a useful means for measuring high-power x-radiation.

We have modeled dynamic stresses in the envelopes of pulsed xenon flashlamps, treating stresses produced by three different sources: the heating of the envelope by the plasma; the pressure rise of the xenon gas; and magnetic forces, due to currents flowing in nearby lamps. The heat-induced stresses were calculated by the finite element method, using uniform heating rates for the inside surface of the envelope that were inferred from flashlamp radiant efficiency measurements. Pressure-induced stresses were calculated analytically, using empirical relationships for temperature and pressure in terms of current density. Magnetically-induced stresses were also calculated analytically, for flashlamps packed parallel to each other in linear arrays.