Survey Of Heat Loads

Power dissipation is at least as important as power density on the crystal.

From Thomas Rabedeau of SLAC (19 March, 2008)

{{{Let's confine our discussion to BL11 since this is the hottest of the high field wigglers at SSRL. I can send you the engineering tech spec for the ID but frankly I'm not sure that will answer your questions very well. Instead here is what I think you need to understand the power radiated by BL11:

SPEAR - 3.0GeV/500mA

BL11 wiggler:

The first mirror (M0 in SSRL terminology) on BL11-2 is a Rh-coated, side-cooled, Si mirror which operates at variable incident angles from 2.7-6.8mrad as well as extracted from the beam. The M0 mirror is centered 14.0m from the source and accepts 1.5 horz_mrad. At 6.8mrad the absorbed power is a bit less than 2900W at 500mA. The maximum absorbed power density on the mirror surface in this configuration is about 0.32W/mm^2. We have yet to expose this mirror to 500mA so I don't have empirical experience yet to address questions about mirror stability under this loading. I have my concerns. We have experience another similarly cooled mirror to 500mA, but we haven't acquired detailed stability data on this mirror at 500mA. (Radiation physics has yet to authorize SSRL BL operations at 500mA, so experience is limited to short approved tests where our main focus was the LN-cooled monos.) We also don't have NSLS2 level emittance to preserve. My guess is your mirrors may want to employ internal cooling despite the extra manufacturing difficulties and associated costs.

At SSRL the hard x-ray ID beam lines all have LN-cooled monos. The mono first crystal is internally cooled with some interesting features for high heat exchange at reasonable flow rates and high liquid pressure as needed for wiggler operations. Specifically we can operate up to ~700lit/hr of LN flow and 10bar pressure. Above about 500lit/hr we start to see some increased mono noise owing to flow induced vibration. Typical operating parameters at 100mA are 2bar back pressure and 250lit/hr. An old paper describing the BL11 prototype LN mono is attached. Note the cooling channel inserts. We have tested an SSRL standard LN-cooled mono on BL6 (57ea 1T pole wiggler) at 500mA. In these tests we looked at the Si(111) and Si(333) rocking curves simultaneously under different power loading and chiller configurations. We found that we could handle about 900W and 13W/mm2 peak absorbed power with very little broadening of the Si(333). By the time we got to 1300W and 14.4W/mm2 the Si(333) rocking curve was significantly distorted but the naturally wider Si(111) rocking curve was relatively undistorted while showing about a 5% loss in normalized transimitted flux. In these later tests we were definitely close to the hairy edge based on the Si(333) curves. Examining the effect of changing the liquid pressure showed improved Si(333) rocking width with increased liquid back pressure from 5bar to 6bar. This suggests the limiter for these measurements was wet wall boiling. For an undulator BL wet wall is less likely to be the limiter whereas extracting heat away from the footprint in the hot wall is the problem. I've included an FEA of our mono crystal cooling scheme under both wiggler and undulator beam conditions which demonstrate some of the difference and limitations.

During the review at NSLS I gave Mike Sullivan a memory stick loaded with photos, top assm drawings, etc of the SSRL LN-cooled mono designs. You might see if you can down load from the stick.

Frankly, aside from the more aggressive emittance of NSLS2, SPEAR3 at 500mA and NSLS2 have a lot in common in terms of power loading issues on wiggler and IVUN BL components. I would be happy to exchange information as we both explore ideas for coping with the beam power. Thanks }}}

Tom also sent along these very useful documents describing the SSRL LN2-cooled monos:

NSLSII/XasBeamLine/ProjectBeamline/BAT/SurveyOfHeatLoads (last edited 2009-10-09 19:51:27 by localhost)