Vacuum at CEBAF

Seminar for Accelerator Operators17 January, 2006

Marcy Stutzman and Philip Adderley

What is vacuum

The woods were dark and foreboding, and Alice sensed that sinister eyes were watching her every step. Worst of all, she knew that Nature abhorred a vacuum

What is vacuum

Vacuums are nothings. We only mention them to let them know we know they're there.

Middle school student’s answer on a science test

Outline

Vacuum Definitions Vacuum conditions at CEBAF Pumps Gauges Operator interface with vacuum Other considerations

Vacuum Definition

Vacuum is when a system is sub-atmospheric in pressure.

There are 2.5x1019 molecules of air in 1 cm3 at sea level and 0°C. PV=nRT, NA=6.02x1023, n=(NA/R)(P/T)

Any reduction of this density of gas is referred to as vacuum.

Nature doesn’t abhorre a vacuum Intergalactic space vacuum: ~1e-16 Torr

Scales to measure vacuum

Atmospheric pressure at sea level and 0°C 760 Torr 1013 mBar 101,330 Pa 14.7 PSI 29.92 inches of mercury 33.79 feet of water

Torr (USA)

mBar (Europe)

Pa (SI - Asia)

Vacuum regimes Low, Medium Vacuum (>10-3 Torr)

Viscous flow interactions between particles are significant

Mean free path less than 1 mm High, Very High Vacuum (10-3 to 10-9 Torr)

Transition region Ultra High Vacuum (10-9 - 10-12 Torr)

Molecular flow interactions between particles are negligible interactions primarily with chamber walls

Mean free path 100-10,000 km Extreme High (<10-12 Torr)

Molecular flow Mean free path 100,000 km or greater

Vacuum Conditions at CEBAF

Application Pressure Range Location Vacuum Regime

Beamline to dumps 10-5 Torr Target to dump line Medium

Insulating vacuum for cryogens

10-4 Torr to 10-7 Torr Cryomodules, transfer lines Medium to high

Targets, Scattering Chambers

10-6 to 10-7 Torr Experimental Halls High to very high

RF waveguide warm to cold windows

10-7 to 10-9 Torr Between warm and cold RF windows

High to very high

Warm beamline vacuum 10-7 to 10-8 Torr or better Arcs, Hall beamline, BSY, some injector

High to very high

Warm region girders 10-9 Torr or better Girders adjacent to cryomodules

Very high to ultrahigh

Differential pumps Below 10-10 Torr Ends of linacs, injector cryomodules and guns

Ultrahigh vacuum

Baked beamline 10-10 to 10-11Torr Y chamber, Wien filter, Pcup

Ultra high vacuum

Polarized guns 10-11 to 10-12 Torr Inside Polarized guns Ultra high vacuum

SRF cavity vacuum Well below 10-12 Torr Inside SRF cavities with walls at 2K

Extreme high vacuum

Why we need vacuum

Keep liquid helium from boiling off Prevent high voltage arcs inside SRF cavities and electron

guns To avoid destroying photocathode by bombardment of

ionized residual gasses To keep the chemical composition of the activated

photocathode at the correct ratios To allow electrons to get to the halls without scattering on

air molecules To avoid beam optics effects caused by the focusing from

a column of ionized residual gasses in the beam path

How to achieve vacuum

Low, Medium Vacuum (>10-3 Torr)

Rough PumpsRoots Pump good for large gas load, large volumes

Mechanical (Oil Seal) PumpBacks Turbo, Roots in systems where oil isn’t to detrimental

Dry pump Used to rough down systems that will go to UHV – no oil contamination

Generation of High, Very High Vacuum

Turbo pumps High speed, precisely tuned fan

blades Backed with mechanical pump

Ion pumps High voltage to ionize gas Magnetic field to direct ionized

gas into plates to trap gasses Systems with ion or turbo pumps

must be roughed down to medium vacuum before starting

Turbo Pump

Ion Pump

Ultra High Vacuum Pumps Getter Pumps

Chemically active surface Titanium sublimed from hot filament Non-Evaporative Getters

Molecules stick when they hit Does not work well for inert gasses such

as Argon, Helium or for methane Ion Pumps

Electric field to ionize gasses Magnetic field to direct gasses into cathodes

where they are trapped Has some pumping capability for noble

gasses Baking used to get pressures below 10-10 Torr

250°C for 30 hours removes water vapor bonded to surface that otherwise limits pressure

Contamination by oil from roughing pumps, fingerprints, machining residue must be avoided

NEG pump array on support grid

Ion Pump

Extreme High Vacuum Generation Typically a combination of Getter pumps and cryo or ion pumps is

used to achieve Extreme High Vacuum (XHV) At room temperature, materials selection and processing, pumping,

and gauging are huge issues

In cryomodules, we get XHV just by the fact that the walls are so cold that everything that touches them freezes solid (except He, which sticks as a liquid)

Virtually impossible to get a gauge into the region where pressure is so low, and turning on gauge would disturb the pressure

Calculations tell us that pressure is very, very good (<10-14 Torr or better)

Where does the gas come from?

Outgassing from the system Metal and non-metal (viton o-rings, ceramics) all outgas Primarily water in unbaked systems Primarily hydrogen from steel in baked systems

Leaks Real

Gaskets not sealed Cracks in welds, bellows, ceramics, window joints Superleaks that only open at very low temperatures

Virtual Small volumes of gas trapped inside system (screw threads, etc.) that

pump out slowly over time Gas load caused by the beam

Desorption of gases by elevated temperatures, electrons or photons striking surfaces, etc.

Loads (targets, etc.) where gas is added Permeation of gasses through materials

Viton gaskets worse than metal seals Hydrogen can permeate through stainless steel!

Vacuum Measurement - Gauges Convectron for low and medium vacuum

Heat transfer from heated strip inversely proportional to pressure

Ionization gauges for high-ultra high vacuum Hot filament ionizes gasses, voltage accelerates them and

sensitive ammeter reads current, proportional to density of gas

Residual Gas Analyzer Hot filament gauge with Quadropole Mass Analyzer to

determine gas species Ion pump current

High voltage ionizes gas, current hitting plates is measured and proportional to vacuum

Always available for monitoring when ion pumps are used Frequently used in alarm handler

Ops interface with vacuum

Alarms Spike commander Halls? UHV supply readouts

Ops interface with vacuum

Ion pump current monitored throughout the machine

Ion pump current corresponds to pressure Different curve for different

pumps Chart gives typical

pressure/current curve Vacuum level determines

If beam can get through to halls Optics effects when a column

of polarized gas is formed Useful indicator of problems

with steering, beam profile

UHV ion pump vs. extractor gauge

1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-12 1.E-11 1.E-10 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05

Pressure (Torr)

Cu

rre

nt

(A)

pump 1pump 2pump 1 calibrationPhysical Electronics

Vacuum Alarms Alarm handler designed to

let ops know when something is wrong

FSDs will trip around scattering chambers etc. at 10-5 Torr

This means beam is hitting something bad or something is leaking

FSDs in linacs, arcs will trip at ___

Burn throughs, lots of beam scraping

ESD (electron stimulated desorption), synch light (photons hitting), thermal heating

Not all pumps are alike – some are aging badly and will read higher currents even at good pressures

Trip levels, actions taken when they trip (FSD, alarm handler), fast valves in some SRF

UHV ion pump power supplies

Ion pump current in baked beamline and electron guns typically read 0 uA

J. Hansknecht developed UHV ion pump power supplies

UHV ion pump vs. extractor gauge

1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-12 1.E-11 1.E-10 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05

Pressure (Torr)

Cu

rre

nt

(A)

pump 1pump 2pump 1 calibrationPhysical Electronics

Sensitive circuit to measure ion pump currents as low as 10-10 Amps

Pressures as low as the 10-12 Torr range

UHV ion pump power supplies

Real vacuum event vs. communication issue Real vacuum events are

typically seen on several pumps at once

Real vacuum events have a sharp rise time, then a slow decay time

Communication issues show up as a spike, typically in only one pump, and do not have a slow decay time

UHV supplies

Most of the supplies are very steady

Some have odd atmospheric dependence

When they go into alarm, let gun on call know It might be a disaster It is often just

something weather related

The gun on call needs to make the determination

Steady readout with comm.errors

Dewpointrelatedpressure readings

Cryocycling

Helium can leak from where it should be into the beamline

Helium accumulates in the beamline A cryocycle is needed to periodically to remove the

Helium. This consists of warming up the cavities to a

temperature above the condensation point of the helium and actively pumping on the beamline with a turbo cart.

Additional Considerations Placement of pumps vs. gas source

Conductance can limit the effective pump speed of a pump. Narrow tubes, elbows, valves limit the effective pump speed. Placement of pumps along a beamline is a significant design issue Distributed pumping is used in storage rings where vacuum must be

even better than at CEBAF RF arcing is partly related to vacuum condition, vacuum pump activity Pump maintenance

DP can activation pump carts on cryo modules Ion pump bakeouts

Failure modes of pumps – want fail safe: don’t destroy equipment when there are power glitches

(turbos and HV don’t go well together) Leak checking

Use RGA and spray helium outside, look for when helium signal shows up on inside of chamber

Summary

Vacuum is essential in CEBAF for many reasons

Different techniques for generating and measuring vacuum depending on need

Operations interface through alarm handler, vacuum spike chart, UHV ion pump monitors

Vacuum readbacks can be a useful diagnostic for problems with beam

Questions?