Cryogenic Pipe Calculations

Cryogenic Pipe Calculations
VB
Jan 2008
idea
• Use superconducting pipe for atomic
beam experiments
– advantages
• cryopumping for better vacuum
• exclusion of magnetic fields inside pipe
– disadvantages
• cost & complications
• cryopump vibrations
basic concept
cryocoolers
insulation pipe
• use simple pipe with superinsulation (MLI) & two cryocoolers

– no liquids in the system (except in the cryocoolers themselves)
– use two cryocoolers (~ 1W cooling @ 4 K)
• use lead pipe (used for calculations)
– device is NOT a magnet
– Type I superconductor
– high critical temp ( ~ 7 K)
• Niobium has even higher critical temp (~9 K)
• Can probably use Type-II superconductors below lower critical temp
performance calculations
• use MATLAB to simulate pipe performance
– heat capacity as a function of temperature
– cryocooling as a function of temperature
• load map
– heat conductivity constant
– insulation on pipe + heat leak at pipe ends
– pipe divided into longitudinal segments
– program calculates new temp profile every fraction of a second
• for each segment
– conduction from adjacent segments
– heat gain through insulation
– arbitrary heat gain in a any segment (used for ends)
– cryocooling heat loss (if present for the segment)
– cryopumps turn on at an upper temp and off at a lower temp (for any
segment)
– temp cannot go below 4 K (cryopump limit)
Superinsulation
• http://www.cryogenicsociety.org/cryo_central/cryogenic_insulation/
– An insulation material's performance under a large temperature difference is given in terms
of milliwatt per meter-kelvin (mW/m-K) and is referred to as the apparent thermal
conductivity or k-value. To compare k-values for different materials one must understand
the warm and cold boundary temperatures, the vacuum level, the residual gas composition,
and the installed thickness. The designer has a very wide range of k-values with which to
work: as low as 0.03 mW/m-K for perforated MLI blankets up to approximately 40 mW/m-K
for cellular glass. As in all good designs, the performance must justify the cost. The
performance of the total thermal insulation system as it is actually put to use is defined as
the overall k-value for actual field installation or koafi.
– Several test methods are usually needed to adequately test and evaluate the overall
performance of an insulation system. Standardized material test methods can be employed
for basic thermal, mechanical, and compatibility properties. Cryostat test methods provide
the apparent thermal conductivity values for the insulation systems. Prototype testing is
then needed to determine the actual performance for a specific mechanical system. The
use of MLI systems illustrates the need for this three step testing process. The k-value for
an MLI system under ideal laboratory conditions may be around 0.05 mW/m-K while the
koafi can easily be 10 times worse.
Cryocooler Performance
Curve used in calculations

(1.0 watts @ 4 K)
SHI Cryogenics Group, a global manufacturer that includes the Cryogenics Division of
Sumitomo Heavy Industries, Ltd. and the former APD Cryogenics, delivers innovative
solutions to the semiconductor, research, optical coating, and medical industries.
http://www.shicryogenics.com/index.php?option=com_content&task=blogcategory&id=22&Itemid=169
Heat Capacity
• Handbook of Chemistry & Physics
– page 2357 for low temp behavior for lead
Heat Conductivity
• Handbook of Chemistry & Physics
– page 2528 for lead, relatively flat, ~ 0.1 cal per sec per cm**2 for 1 cm thickness
• http://prola.aps.org/pdf/PR/v80/i5/p859_1
– evidence of superconducting behavior of heat capacity (factor of 2.5
enhancement of the heat conduction in the 4-15 K temprange)
Used a constant
value in calculations
(0.5 or 1.0)
MATLAB simulations parameters (data1)
• rateloss=0.00003 % insulation heat loss W/meter/K
• ncool=1.0 % de-rating factor for the 1.0 watt cryocooler
• initT = 10 % starting temperature for the pipe
• roomT = 300 % temperature of the laboratory
• lowT = 5 % turn-off temp of cryocooler
• highT = 5.75 % turn-back-on temp of cryocooler
• Tmin=4.0 % min cryocooler temp
• maxIn = 30000 % number of seconds to run simulation
• pradius = 5.0 % pipe radius in cm
• pthick = 1.0 % pipe thickness in cm
• plength = 1000 % pipe length in meters
• pdensity = 13 % density of pipe g/cc
• nsegs = 100 % number of pipe segments in length
• cooling(8)=0.5 ; cooling(72)=0.5; % cryocooler power in
segments
• heatleak(1)=0.1; heatleak(100)=0.0; % heat leak in segments
• secsegs = 2 ; % number of time segments in a second
• hcond=1.0 ; % heat conductivity
Results – data1 – temp contours
Results – data1 – temp vs. time
Uptime=85%
Times on are 439 310 303 301 300 300 300 300 300 300
Times off are 2065 1769 1772 1771 1770 1770 1770 1770 1770
Cooldown from 300 K
•Would be nice to get
faster cooling
•Pre-cooling
•Better distribution
100 hour timescale

MATLAB simulations parameters (data2)
segments
Results – data2
Uptime=64%
Times on are 714 608 606 602 602 601 602 601 602 601
Times off are 1203 1084 1085 1084 1084 1084 1084 1084 1084
Results – data2 – secsegs=10
uptime=64%
Times on are 713 608 606 603 603 603 603 603 603 603
Times off are 1203 1084 1085 1085 1085 1085 1085 1085 1085
MATLAB simulation parameters (data4)
segments
Results – data4
Uptime=65%
Times on are 275 268 265 265 264 265 264 264 264
Times off are 511 502 501 501 501 501 501 501 500
Cryocooler - Vibrations
Information from …
Vibration Reduction Methods:
Active Cancellation
[13]
Vibration reduction – PT cryocoolers
CC
sizes
Sumitomo
Heavy
Industries
~ 50 cm scale
Mounting & Other Issues
• Need a system to support pipe vertically
• Need to connect cryocoolers to pipe
– Copper collars (Cu conductivity ~ 5-10 higher)
– Can we use flexible metal hose between collars and
cryocoolers (reduce vibrations)
– How many points on the pipe do we connect
– What happens at the “warm” end that is connected to
rest of the apparatus
• ES&H issues with lead?
– We have other options for metals (e.g. Nb)
CC Spec Sheet
Sumitomo
Heavy
Industry
Summary
• Simple model of pipe shows promise
– Timescales look reasonable
– “brute force” vibration control (i.e. CC off)
works
– Still have options to improve cooling and
vibrations
• Next step – get professional help

Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, De
tectors and Associated Equipment
Volume 538, Issues 1-3, 11 February 2005, Pages 33-44

Reduction of field emission dark current for high-field gradient electron gun by using a molybdenum
cathode and titanium anode
Enhancement effect of dark current by electron and ion

impact on electrodes. (1) Primary field emission, (2) Desorption
Cathode flattop = 18 mm
of ions and molecules by electron bombardment, (3) Ionization
Anode flattop = 2 mm by electron impact, (4) Back bombardment, (5) Emission of
secondary ions and electrons.


1 nA plots
The free parameter α was

adjusted in each case, but had
E(I,10mm) = 124/(1+4) = 25 MV/m for Ti an average value of 0.4±0.02
E(I,10mm) = 170/(1+10) = 15.5 MV/m for Mo for Ti and 1.0±0.04 for Mo. This
constancy of α over the entire
range of dark current indicates
Can we make a flat beam?? that the gap separation
dependence is well
approximated by Eq. (2).
Nuclear Instruments and Methods in Physics Research Section A: Accelerators,
Spectrometers, Detectors and Associated Equipment
Sacrificing some
gradient can greatly
reduce the dark current
Surface preparation
and cleaning is
critically important

Cryogenic Pipe Calculations

Uploaded by

Copyright:

Available Formats

Cryogenic Pipe Calculations

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cryogenic Pipe Calculations

Uploaded by

Copyright:

Available Formats

Cryogenic Pipe Calculations

• use simple pipe with superinsulation (MLI) & two cryocoolers

Curve used in calculations

100 hour timescale

• Next step – get professional help

Volume 538, Issues 1-3, 11 February 2005, Pages 33-44

Enhancement effect of dark current by electron and ion

Volume 538, Issues 1-3, 11 February 2005, Pages 33-44

Volume 538, Issues 1-3, 11 February 2005, Pages 33-44

The free parameter α was

You might also like