From: Katalev [valery.katalev@desy.de] Sent: Wednesday, August 01, 2007 9:48 AM To: nezhev; Nantista, Christopher D. Cc: Adolphsen, Chris; Sergei Nagaitsev; Brian Chase; N.Solyak; cancelo@fnal.gov; Larsen, Ray S.; Bane, Karl L.; Shigeki Fukuda; Akira Yamamoto; Stefan Choroba; vladimir.vogel@desy.de; Jerry Leibfritz; John S. Reid Subject: Re: WR650_temperature Attachments: Thermal Phase Shifts_1.ppt Dear All, I added a little bit to the Chris Nantista's calculation. I think the thermal phase shift will be less then the first estimation. May be even it's not necessary to compensate it. With kind regards Valery nezhev wrote: > Chris, > A few comments to your letter: > 1. It is quite possible that discrepancy in the measured waveguide > temperature is just because at SLAC you didn't reach an equilibrium. > Thermal capacitance doesn't allow > for more than about 3 C/hour at this power for waveguide with 6 mm > wall. Valery should know better what time do one need to reach > equilibrium. > 2. Temperature dependence of the phase shift vs. > temperature is actually less than you calculated: > WR 650 ~ 1.12 deg/mm ~ 0.026 deg/m/C > WR 770 ~ 1.26 deg/mm ~ 0.029 deg/m/C. > 3. Accordingly, the overall phase shift will be just > 12.7 deg instead of 23.2 deg. 8 deg out of 12.7 deg is a common shift > for all cavities and consequently LLRF system can take care of it. The > rest of of the phase shift is only 4.7 deg maximum and practically > will not affect acceleration. > 4. So, I sitll don't see any necessity in the water cooling of the > waveguide. > > Regards, Oleg > > Nantista, Christopher D. wrote: >> Gentlemen, >> Since I won't be here next week for the meeting, here are some >> rough calculations I did with the data sent previously. I assume >> Valery's heating values, though the large discrepancy with Faya's >> should be understood. Perhaps it's a question of different air >> circulation. Also in the spirit of "worst case" evaluation, I use >> maximum klystron power without reduction due to attenuation along the >> waveguide circuit. On the other hand, I don't really account for >> enhanced heating in flex waveguides, bends, etc. The results should >> thus be taken only as ballpark figures to aid in the discussion. My >> feeling is that it may be worthwhile to insulate and water cool the >> long waveguide runs but not all the components. Of course, the heat >> load and phase shift can also be reduced ~22% by copper plating. >> Some cost analysis should be done. >> >> Chris N. >> >> >> -----Original Message----- >> From: nezhev [mailto:nezhev@fnal.gov] Sent: Wednesday, July 25, 2007 >> 9:50 AM >> To: valery.katalev@desy.de >> Cc: Adolphsen, Chris; Sergei Nagaitsev; Brian Chase; N.Solyak; >> cancelo@fnal.gov; Larsen, Ray S.; Bane, Karl L.; Nantista, >> Christopher D.; Shigeki Fukuda; Akira Yamamoto; Stefan Choroba; >> vladimir.vogel@desy.de; Jerry Leibfritz; John S. Reid >> Subject: Re: WR650_temperature >> >> Dear Valery, >> Very helpful. For the case of ILC (10 MW per >> 3 cryomodules and 5 Hz) the maximal temperature raise suppose to be >> 12 C according yours measurements. >> Of course it will go down along the cryomodule RF distribution system. >> Regards, Oleg >> >> Katalev wrote: >> >>> Dear all! >>> >>> May be it will be useful for you. It's my experimental data for WR650. >>> >>> Regards >>> Valery >>> >>> nezhev wrote: >>> >>>> Hello Chris, >>>> Th waveguide warming depends on orientation, air flow, etc., but at >>>> 100 W/m I would expect it to be 10-12 C in average (see attached file). >>>> Thermal expansion coefficient for Al is 23e-6/C and thermal >>>> expansion of 12 m long waveguide (along one cryomodule) will be ~ 3 mm. >>>> Phase shift in WR650 waveguide at 1.3 GHz is ~ 1.12 deg/mm and >>>> maximal temperature induced phase shift along one cryomodule is < 4 >>>> deg, which practically will not reduce acceleration. >>>> So, I just don't see any sense to employ water cooling. >>>> Oleg >>>> >>>> Adolphsen, Chris wrote: >>>> >>>>> For Friday at 1:00 PM at FNAL (11 AM at SLAC), Karl is willing to >>>>> answer questions about his gradient optimization calculations as >>>>> summarized in the attached PAC paper. >>>>> >>>>> For Oleg and the HLRF people, I note the following waveguide >>>>> heating data we took at SLAC to answer the question of how hot the >>>>> WG gets if it is not water cooled - DESY apparently has much >>>>> higher values but this may be because they are measuring the >>>>> heating near a load (the ferrites probably radiate heat down the >>>>> pipe) >>>>> >>>>> When the VTO is running at 3.2 MW (average power during a 1.2 ms >>>>> pulse) at 5 Hz, the rf on and off temperatures are listed below. >>>>> >>>>> Section RF_OFF RF_ON (more than 1 hour) >>>>> >>>>> Water-Cooled Ferrite Load 88 degF 135 degF >>>>> >>>>> Black Flexible Waveguide 83 degF 94 degF >>>>> >>>>> Straight waveguide 82 degF 87 degF >>>>> >>>>> For reference, the phase length along a 12 m section of WR650 will >>>>> change by roughly a degree L-band per degree C temperature change. >>>>> >>>>> -----Original Message----- >>>>> From: nezhev [mailto:nezhev@fnal.gov] Sent: Friday, July 20, 2007 >>>>> 6:49 AM >>>>> To: Sergei Nagaitsev >>>>> Cc: Adolphsen, Chris; Brian Chase; N.Solyak; cancelo@fnal.gov >>>>> Subject: Re: rf modeling meeting >>>>> >>>>> 1 pm is OK with me. >>>>> It will be interesting to hear news from Karl (driving cavities by >>>>> pairs) and results from Julien on the cavity excitation by the >>>>> beam in the case of klystron failure. >>>>> Oleg >>>>> >>>>> Sergei Nagaitsev wrote: >>>>> >>>>> >>>>>> I would like to have a meeting next Friday July 27 at 1pm (CST) >>>>>> on rf modeling. There are things that we need to discuss, >>>>>> however we have this school: >>>>>> http://www-bd.fnal.gov/2theILC/ >>>>>> going at that time. I think we may be OK at 1pm? >>>>>> >>>>>> If it is OK with you, please send agenda items to me. >>>>>> >>>>>> Thanks, >>>>>> Sergei >>>>>> >>>>>> >>>>>> >>>>> >> >> > >