SHORT TERM COURSE ON

VACUUM TECHNOLOGY AND

PROCESS APPLICATIONS

(17th Nov – 27th Nov 2007)

Prof. V. Vasudeva Rao Coordinator

Vacuum Technology Laboratory Cryogenic Engineering Centre

IIT Kharagpur

VVAACCUUUUMM ((LLaattiinn -- EEmmppttyy))

According to American Vacuum Society (1958)

Any given space filled with gas at pressures below atmosphere (or) Molecular density < 2.5 x 1019

mol/cm3…Vacuum.

Atomic diameter of typical gas = 3A0 = 3 x 10-8 cm ∴ 1 cm length contains 3 x 107 atoms for

solid with tightly arranged atoms 1 cm3 contains 33 × 1021 = 3 × 1022 atoms solid evaporates to gas,

volume changes by 1000 ∴ gas of 1 cm3 contains 3 x 1022 ÷ 1000 = 3 x 1019

atoms even at 10-12 torr (best possible vacuum in laboratory) we have 30,000 molecules per cm3

and mean free path > diameter of earth.

Vacuum is measured by measuring the absolute pressure in an enclosure.

In coherent unit system [F] = [l] [m] [t]-2 and [P] = [l]-1 [m] [t]-2

C.G.S dynes/cm2 = 0.1 pascal dyne = 1gm.cm/s2

S.I (M.K.S) newton/m2 = 1 pascal.

In non-coherent system, the pressure units Torr & mbar are popularly used in "Vacuum Technology"

1 atm = 760 mm of Hg 760 Torr = 1013 mbar

≈ 0.1 Mega Pascal = 1.03 Kg/cm2

NATURAL VACUUM

Human beings 740 Torr – Respiration / 300 Torr - Suction Octopus 0.1 Torr Space: Pressure decreases with the altitude

- up to 100 km (troposphere & stratosphere) Pr. Decreases by a factor of 10 for each increase in altitude of 15 Km…. "10-3 Torr at 90 Km"

- 100 - 400 Km (Inosphere) Pr. Decreases by a factor of 10 / every 100 - 200 Km….."10-10 Torr at 1000 Km"

- Above 1000 Km Pr. Decreases slowly.

“10-13 Torr at 10000 Km”

Nature is powerful

Altitude Composition < 200 Km Atm. is air. 200 - 1000 Km Atomic N & O 700 - 1000 Km Appreciable He >1500 Km Neutral atomic H,

Protons, electrons etc.

Vacuum of 10-10 Torr Highly expensive technology on Earth - Naturally available in large volumes [Universe] Above 1000 Km in Space.

Brief History of Vacuum Technology

1564 – 1642 Galileo Vacuum with a piston in cylinder

1643 Torricelli Vacuum produced at the top of a column of mercury

1623 – 62 Pascal Barometer

1654 Guericke Mechanical effects famous

1879 Edison’s Invention of the incandescent lamp

1879 Crookes Cathode ray tube & Evacuated flask by Dewar

1902 Vacuum diode / 1906 triode / 1909 tungsten filament electron & X-ray tubes electronics

1874 Mc Leod Primary gauge

1906 Pirani’s Thermal conductivity gauge

1915 Gaede’s and Langmuir’s Diffusion pump

1916 Buckley’s Hot cathode ionization gauge

1937 Penning’s Cold cathode gauge

After 1940 Vacuum Technology for nuclear research

1950 Bayard – Alpert Ionization gauge

1953 H.J. Schwarz & R.G. Herd Ion-pumps produced

1958 W.Becker Molecular drag of a high speed rotor

1912 W. Gaede Turbomolecular pumps

As pressure goes down in a vacuum system, a marked change occurs in the following physical parameters.

1. Molecular density 'n': Average number of molecules per unit volume. For a given temperature and volume n ∝ p density ∝ p

2. Mean free path 'λ': Average distance that a molecule travels in a gas between two successive collisions with other molecules of that gas.

λ = 1 / (√2 π n d2)….kinetic theory (not interatomic distance) – Calcutta population example. For air at room temp. λ=5.1 x 10-3 / P … (λ in cm and P in torr).

3. Time to form a monolayer 'τ': Time required for a freshly cleaved surface to be covered by a layer of the gas of one molecule thickness. τ is very long for UHV (2.2 x 106 sec at 10-12 torr)

Gas molecules impinging per sq. cm = ¼ (nvavg)

P (Torr) N (mol/cm3) λ (cm) τ (sec) 760 1

10-3

10-6

10-9

10-12

10-15

2.46 x 1019

3.25 x 1016

3.25 x 1013

3.25 x 1010

3.25 x 1017

3.25 x 1014

3.25 x 10

6.7 x 10-6

5.1 x 10-3

5.1 5.1 x 103

5.1 x 106

5.1 x 109

5.1 x 1012

2.9 x 10-9

2.2 x 10-6

2.2 x 10-3

2.2 2.2 x 103

2.2 x 106

2.2 x 109

By analysing the valves of η, λ, τ and D (characteristic dimension of the chamber), we can classify vacuum into three regions.

Low (medium) Vacuum:

The number of molecules of the gas phase is large compared to that covering the surface.

760 - 0.5 Torr Low Force effects can be felt. 0.5 - 10-2 Torr Medium λ << D

High Vacuum:

The gas molecules in the chamber are located principally on surfaces. Viscosity effects disappear. 10-3 to 10-7 Torr. λ ≥ D

Ultra High Vacuum:

The time to form a monolayer 'τ' is longer than the usual time for laboratory measurements; thus clean surfaces can be prepared and their properties can be studied.

10-7 to 10-15 Torr λ >> D

Pumping/measurement very difficult

Gas Compositions

Component Atmosphere Partial Pr. (Torr)

Ultra High VacuumPartial Pr (Torr)

N2 O2

Ar CO2 Ne He Kr H2 Xe

H2O CH4 O3

N2O CO

595 159 7.05 0.25 0.014 0.004

8.4 x 10-4 3.8 x 10-4 6.6 x 10-5

11.9 1.5 x 10-3 5.3 x 10-5

1.8 x 10-4 ------------

---------- 3 x 10-13 ---------- 6 x 10-12 ---------- ---------- ---------- 2 x 10-11 ---------- 9 x 10-13 3 x 10-13 ---------- ---------- 9 x 10-12

In the ultra high vacuum range hydrogen is the dominant component coming mostly from the bulk of the materials (permeation).

VVaaccuuuumm TTeecchhnnoollooggyy -- PPrriinncciippllee // AApppplliiccaattiioonnss

1. Pressure difference : (Force : 1Kg/cm2)

(a) Holding, lifting, transporting solids, liquids.

(b) Vacuum sniffers; mouth is placed on the object to be lifted very precisely. (Ex: Vacuum cleaner - 600 Torr).

(c) Chemical industry to accelerate filtering speed.

(d) Railway breaks Low vacuum > 10 Torr.

2. Removing chemically active elements :

(a) Electric bulb To avoid heated filament oxidation. 10-5 Torr & sealing or Inert gas filling after evacuation

(b) Vacuum Metallurgy : To protect active metals from oxidation during melting, casting, sintering etc. (0.1 Torr).

(c) Vacuum Packaging : of food materials sensitive to atmospheric reactions (low vacuum - 0.1 Torr).

(d) Vacuum encapsulation : of sensitive transistors & capacitors.

3. Removing humidity from foods & chemicals :

(a) Vacuum Concentration (removing water) at low heating of Fruit juice, concentrated milk etc.

(b) Freeze drying for storage : Cooling and removing water by sublimation under vacuum, preserving volatile constituents.

- instant coffee, blood plasma.

(c) Vacuum impregnation : Removing occluded humidity/ gases, filling their place by other materials.

- insulation of motor windings, capacitors, cables etc. 4. Thermal & Electrical insulation :

(a) Dewar flasks (LN2 & LHe), Thermos flasks - double walled with evacuated space – convection reduced.

(b) Electrical insulation - vacuum switches/interuptors, high voltage tubes - Fusion reactors for energy production etc.

5. Avoiding atomic collisions :

(a) Oscilloscopes, photo cells, X-ray tubes, Mass spectrometers, Electron microscopes etc. (10-6 Torr).

(b) Vacuum coating units : where coating materials evaporated from a source travels straight to the substrate with out collisions to create high quality thin film devices (< 10-8 Torr).

(c) Fresh surface analysis equipment - SEM, EDAX, ESCA etc. 10-10 Torr where τ is very long. 6. Space simulation chambers

Ex: space shuttles etc. 10-10 Torr. Simulating the conditions of far space.

7. Molecular distillation of pure fractions by evaporating and condensing.

http://acept.la.asu.edu/PiN/rdg/vacuum/vacuum.html

Vacuum Pump

13 m

20 m 40 m

IIT Kharagpur

JSC Houston Space Simulation Chamber Cryopump Assembly

760 102 100 10-2 10-4 10-6 10-8 10-10 10-12 10-14

Piston Waterjet

Rotary Sorption Roots

Ejectors

Diffusion

Molecular Ion Cryogenic

P (Torr)

Production of Vacuum

Pressure ranges of vacuum pumps:

No single pump exits which can cover all ranges Vacuum pumping is based on one of the following

Compression - expansion of gas Ex:- Rotary Drag by viscosity effects Ex:- Vapour ejector Drag by diffusion effects Ex:- Vapour diffusion pump Molecular drag Ex:- Turbo molecular Ionisation effects Ex:- Ion pumps Physical & chemical sorption Ex:- Sorption & Cryopumps

Pump specifications :- Lowest pressure, Pressure range, Pumping speed, Exhaust pressure etc.

Pumping Speed: (SP) - Volume of the gas per unit time which the pumping device removes from the

system at the pressure existing at the inlet to the pump. (lit/sec, m3/hr).

Throughput (Q): - Product of pumping speed and the inlet pressure. Torr lit/sec or atm. Cm3 /sec

Q is proportional to mass flow rate

⎟⎟⎠

⎞⎜⎜⎝

⎛==

dtdVPpPSQ

( )

( )PVdtd

KTM

KTMPV

dtddtNMd

dtdm

=

=

=

⎟⎟⎠

⎞⎜⎜⎝

⎛

7

8 9

6

4 3

5

1

2

10 11

1. Inlet tube, 2. Inlet port, 3. Top seal, 4. Vanes, 5. Oil, 6. Rotor, 7. Stator, 8. Exhaust port, 9. Exhaust flap valve with backing plate, 10. Exhaust outlet, 11. Oil splash baffles

AB A

B B

A

BB

A

(a) (b) (c) (d)

Hyper Link for Rotary Animation

Rotary Vane Pump

Stator – with eccentric Rotor Inlet & Exhaust Two vanes in diametrical slot (few degrees either side of vertical line) with filter Neoprene constrained to hinge between stator & metal plate stator – Rotor assembly is submerged in oil since both vanes operate, in one rotation the volume of the gas swept is twice that shown in figure (b). If rotational speed is n/unit time (min) pump displacement (pumping speed) = 2Vn

The lowest pressure of Rotary pump depends on "Compression Ratio" and hence on dead volume.

As vacuum improves, pressure after compression is not above atmosphere. Then the gas cannot be discharged and subsequent pumping action re-expands and recompresses the same gas without reducing the pressure.

For getting 10-2 Torr, compression ratio of 1,00,000 is required. Even with such ratios the lowest pressure in single stage pump is only 5 x 10-3 Torr.

Parallel connections of two identical pumps provides twice the displacement but same ultimate pressure. Series connection provides same displacement but greater pumping speeds at low pressures. Hence, a two stage rotary pump may reach 10-4 Torr.

The ultimate pressure is also limited by the leak across top seal and vapour pressure of lubricating oil.

How to get better vacuum < 10-4 Torr ?

How to get more pumping speed at low pressure limit ?

Pumping speed Vs Pressure Single Stage Double Stage

P (Torr)

S (L/m)

760 Torr

60

50

40

30

20

10

0 10-4 10-2 1 102

Gas Ballast Operation When water is present in a vacuum system, it turns out that if the compression ratio of the pump exceeds approximately 8:1 water will condense. To avoid this a solution proposed by Gaede (Figure). Here atmospheric air be admitted to the pump during the compression cycle to reduce the effective compression ratio and thereby increase the proportion of non-condensable gases in the pump. By this means, the partial pressure of the vapour being pumped does not exceed its saturated vapour pressure at the time the exhaust valve lifts (the exhaust valve lifts earlier in the pump cycle than it otherwise would) and consequently vapour is discharged without condensing. The extra work done in compressing the gas introduced at gas ballast causes a temperature rise which also assists in preventing vapour condensing within the pump. Gas ballasting also has the effect of transporting oil from the pump chamber and this oil appears as an oil mist. Since gas ballasting will usually be conducted for 20-30 minutes at a time, it is necessary to monitor the pump oil level. The reduction in compression ratio accompanying gas ballasting causes a reduction in the ultimate pressure attainable.

Exhaust

Vanes

Oil Spring

Stator

Rotor

Inlet

Pump Oil The Rotary-pump oil is usually a hydrocarbon oil, chosen for its low vapour pressure. The oil must also possess the appropriate viscosity for the pump, since too low a viscosity will result in noisy pump operation and too high a viscosity may result in seal failure, loss of vacuum and possibly pump seizure. Oil serves as a sealent, coolent and lubricant.

http://acept.la.asu.edu/PiN/rdg/vacuum/vacuum2.html

Additional Precautions The pump should be vented back to atmosphere as soon as it is stopped as otherwise the oil in the pump will enter into the system due to suction. During unattended operation, this situation may occur due to power failure. Solenoid valves can also be used to close off the pump from the rest of the system and to vent it to atmospheric pressure if the power is turned off. Further, a reservoir to catch the oil is added between the pump and the system as a safety precaution Inlet filters are added to filter glass particles or abrasive materials that may be present in the system under evacuation.

System

12

5

4

6

37

Diffusion Pump

1. High Vacuum 2. Water Cooling 3. First Stage 4. Second Stage 5. Pump Oil 6. Heater 7. Fore Vacuum (Rotary Pump)

The molecules of the oil vapour travel up the chimney and are deflected by the umbrella and jet towards the walls. There the vapour condenses back due to water cooling, return back to the boiler, gets heated to form the vapour streaming up. During expansion in the nozzles the oil molecules given a downward momentum to the air molecules to push them (compress them) towards fore vacuum pump. The air molecules from the system slowly diffuse down to bottom of the pump. The oil column can sustain the pressure difference and act like an atomic compressor. The air at relatively high pressures at the bottom can be removed by a matching rotary pump.

Properties of some common diffusion pump fluids

Fluid Composition Mol.wt. Ultimate pressure* @ 20oC mbar

Apiezon A Apiezon B Apiezon C Edwards L9 Silicone DC 702 Silicone DC 703 Silicone DC 704 Silicone DC 705 Santovac 5 Fomblin 18/8 Mercury

Mixture of hydrocarbon Mixture of hydrocarbon Mixture of hydrocarbon Napthalene based Mixture of Polysiloxanes Mixture of Polysiloxanes Single molecule siloxane Single molecule siloxane Polyphenylether Perfluoropolyether ---

354 420 279 407 530 570 484 546 446 2650 201

6.5 x 10-5 1.3 x 10-6 1.3 x 10-7 5 x 10-9

6.5 x 10-6

6.5 x 10-6

6.5 x 10-8

1.3 x 10-9

1.3 x 10-9

2.7 x 10-8 1.2 x 10-3

11

10

9

8

1

Fig. Pumping mechanism of a vapour jet.

Pump fluid Gas molecules

2

7

3

4

5

6

12 1. Water-cooling coils 2. First compression stage 3. Second compression stage 4. Third compression stage 5. Vapour condenses and

returns to boiler 6. Boiler 7. Electric heater 8. Pump fluid 9. Fourth stage compression 10. Foreline baffle 11. Foreline (pump outlet) High

pressure 12. Pump inlet (low pressure)

Rotary pump to Diffusion pump The rotary pump is matched to a diffusion pump according to the relation given for vapour booster. Critical backing pressures for diffusion pumps are typically0.35 torr and maximum throughput occurs in the constant throughput-pumping region at pressures between 10-1 and 10-3 torr and can generally be considered to occur at 10-2 torr. Thus for a diffusion pump with a speed of 700 1s-1 at 10-2 torr and a critical backing pressure of 0.35 torr the minimum rotary pump sped required is

Equation: 1min3ft51or1ls24100120

0.35

1210700rotaryS −−=×

−×= Holding rotary pump to diffusion pump large rotary pumps is required to produce fast roughing cycles and to back large diffusion plumps operating at high throughput. Once these conditions have been fulfilled the large rotary pump can be replaced with a much smaller pump giving lower power consumption and lower noise level. Holding pumps are use full to back large diffusion pumps operating at small throughput, i.e. low process pressures, or operating in an idle condition against a closed isolation valve.

Fractionating Pump The various constituents of the pump fluid are so selected that the high vacuum nozzle is only supplied by the fraction of the pump fluid which has the lowest vapour pressure. This assures a very low ultimate pressure. Fractionating occurs because the degassed oil first enters the outer part of the boiler, which serves the nozzle on the backing vacuum side. Here a part of the more volatile constituents evaporates. In this way, the already purified pump fluid reaches then the intermediate part of the boiler, which serves the intermediate nozzle. Here lighter constituents evaporate in greater quantities than the heavier. When the oil enters the central region of the boiler which serves the high vacuum nozzle, it is freed of the light volatile constituents.

http://www.2spi.com/catalog/vac/santovac-5-diffusion-pump-fluid-technical-paper.html

~10-8

Inlet Pressure (Torr)

~10-3 ~10-1Torr

1 2 3 4

Pum

ping

Spe

ed

PPeerrffoorrmmaannccee CChhaarraacctteerriissttiiccss The pumping performance of a diffusion pump is displayed in the form of a plot of pumping speed versus inlet pressure in the figure. The graph consists of four distinct sections. To the left, the speed is seen to decrease near the limit of obtainable vacuum. The constant speed section results from constant gas arrival rate at molecular flow conditions and a constant capture efficiency of the vapor jets. At molecular flow, the gas molecules arrive into the pump due to their normal molecular velocities, which depend on temperature and the molecular weight. The rate of arrival also depends on the conductance of the inlet ducts and the geometry of the pump entrance. A certain percentage of molecules reaching the vapor jets will be captured. The capture rate is usually constant until the vapor jets become overloaded. The part marked "over-load" is a constant-throughput section which indicates that the maximum mass flow capacity of the pump has been reached. In the last section, at the right, the performance is highly influenced by the size of the mechanical backing pump (critical backing pressure).

1. Ultimate Vacuum Limitation.

2. Constant speed. 3. Constant through put

(Overload). 4. Mechanical Pump effect

H2

He

N2

Ar

Rel

ativ

e pu

mpi

ng s

peed

Inlet pressure (Torr) 10-13 10-11 10-9 10-7 10-5 10-3 10-1

Typical performance of diffusion pumps with various gases

Typical speeds 100 l/s to 45000 l/sec, commercially available. The pumping speed of a diffusion pump can be obtained from S =11.6 A H lit/sec (for air) A = Area of intake annulus H = Ho-factor (0.3—0.5) S ∝ 1/ M M = mol. wt of gas. Pump fluids: 1. Silicone DC 702 – 705 (siloxane) mol.wt ≈ 500;

ultimate pr: 10-6—10-9 mbar higher resistance to oxdidation.

2. Apiezon (A, B, C) Mixture of hydro carbon

mol. wt. 350 to 480, ultimate pressure 10-5 –10-7 mbar 3. Santovac Polyphenyl ether mol.wt 446 ultimate vacuum 1.3 Χ 10-9 mbar.

System

BBaacckk SSttrreeaammiinngg -- BBaafffflleess // CCoolldd TTrraappss Back-streaming occurs when pump fluid molecules move above the upper jet so that they can enter the chamber, causing possible contamination. This can be largely prevented by the use of chilled baffles or a cold trap. It can also be greatly reduced by proper design of the top jet and by the use of a large cold cap. An important element of any vacuum system is the baffle or cold trap. A trap (cold trap) is actually an entrapment pump for condensable vapours. A baffle is a device designed to condense pump fluid vapours and return them to the pump. It is therefore generally associated with diffusion pumps. Although modern diffusion pump fluids such as DC705 or Santovac have vapour pressure in the region of 10-10 mbar at room temperature, some decomposition of the pump fluid does occur in the pump boiler and lighter fractions are generated. Many of these may be trapped by means of a water-cooled (chilled) baffle situated above the pump first jet. An even more effective trap is provided by liquid nitrogen cooling of such a baffle. A more generally useful arrangement is the liquid nitrogen cold trap situated immediately above the mouth of the diffusion pump. In this position the cold trap not only holds the more volatile oil fractions arising from the pump, but also water vapour or other condensables arising from the vacuum chamber.

Chevron Baffle

Cold Trap

Diffusion Pump

LN2

CCrriittiiccaall bbaacckkiinngg pprreessssuurree During normal operation the supersonic high pressure region of oil vapour overtakes the slower moving gas molecules which have sonic speeds. There is a consequent pressure rise resulting in a steep and stable wave front, i.e. a shock wave is formed in which the gas is rapidly compressed. This shock wave acts as a "dam" or "seal" across the pump so that gas from the backing region cannot surmount the pressure step of the shock wave and return the high vacuum inlet. If the backing pressure is too high the shock wave front will be too near the nozzle outlet giving a less satisfactory sealing effect. In this way if the backing pressure rises to a critical point (typically 0.5 mbar) the vapour jets break down (due to the increased gas density in the pump) and gas molecules then back diffuse to the pump intake aperture increasing the ultimate pressure drastically. The Critical backing pressure.

P1

P2

103 1 10-3

1

10-7

P2 (mbar)

P1 (mbar)

Typical High Vacuum Pumping System

• Eliminates back streaming of pump fluid and increases pumping speed for condensable gases

Cooling Water

Baffle * Cryobaffle N2

Main isolation Valve

Pirani/thermocouple gauge

Isolation/air Admittance valve

Backing line

Roughing Line Backing valve

<10-7

Mbar

Ionization/penning Gauge

Chamber

Air inlet valve Roughing Valve

Gauge

Rotary pump

Diffusion Pump

Liquid Nitrogen Traps : The above Figure (a) shows a re-entrant trap of low conductance used in backing lines to suppress back migration of rotary pump oil vapours to high vacuum systems, and occasionally above low speed diffusion plumps. These traps are sometimes used in backing lines to protect a rotary pump from condensable vapour loads that are beyond its capacity on full gas ballast, but they are comparatively ineffective, since under viscous flow conditions, vapour is swept through without hitting a cold surface. At system pressures below 1 torr however, they protect the vacuum chamber against hydrocarbon and water vapour contamination originating from the rotary lump. The application of backing line traps for us e with clean system is more fully discussed in section 3.2 Figure (b) shows a high conductance high vacuum trap thermally insulated by the high vacuum achieved but not optically dense, i.e. a molecule can travel from pump connection to system connection making collisions with room temperature surfaces only and a certain proportion of molecules will find these paths Figure (c) is an optically dense version of that shown in Figure (b). This has been achieved by the use of the two overlapping cylinders at the lower end.

Fig. Typical liquid nitrogen traps

FFoorree--lliinnee TTrraapp The presence of oil vapour in the work-chamber is often thought to arise from back-streaming from the diffusion pump, but investigations have shown that this is in fact rotary pump oil which has entered the chamber via the roughing line. When the roughing pressure drops below 10-2 mbar rotary pump vapour molecules can move in any direction as they are in molecular flow. This can be greatly reduced by the use of a correctly maintained fore-line trap. It is important however that if a roughing line is used then once the chamber pressure reaches the correct value, the roughing valve should be closed to minimize the back-streaming of rotary pump vapour into the vacuum chamber. It is necessary to isolate the trap with suitable valves when system is exposed to atmosphere. Saturated alumina can be reactivated by heating to 2500C with the help of a heating rod.

Rotary Pump

Perforated Can Zeolite pellets /Alumina

Diffusion

A fore-line trap

Rotary Pump

Roots Blower

Pumps without a discharge valve, which move gases by the propelling action of rapidly rotating members, are called Rotary Blowers. A fairly common representative of this type is the Roots pump. This type of pump contains two counter-rotating lobes, each with a Figure-eight cross-section. The lobes do not touch each other or the casing. The clearance between lobes and between the lobes and the casing is of the order of 0.010 inch. A single shaft may drive the machine with the second rotor synchronized and driven through a set of timing gears.

Fig. Roots blower (cross-section)

Fig. Operating principle of Roots pump.

In the first position, air enters on the inlet side. The lower impeller, as indicated in the second position traps part of this air. This volume is discharged in the third and fourth positions. The latter also shows another quantity of gas trapped by the upper impeller, which will be discharged during the next quarter revolution.

The main advantage of Roots pumps is their ability to handle large gas loads in a pressure region where neither rotary nor diffusion pumps are fully efficient It is sometimes desirable to make use of a pump capable of reaching lower pressure than the single or double-stage rotary pumps and having very high throughput at these low pressures. In addition, it may be desirable to use such a pump to prevent the migration of oil molecules from the mechanical pump into the system. For these purposes, a Roots pump (booster) is frequently employed. Commercial models are available with high pumping speeds in the range 104 to 106 liter/min. They are rarely used in high vacuum systems, but offer economic advantages in industrial applications requiring 10 to 10-4 Torr pressure. By staging two such blowers in series with a mechanical pump, even lower pressures can be reached, although their general field of usefulness lines in the range of 1 x 10-2 Torr to 5 x 10-4 Torr. Due to the initial high viscosity of the air at atmospheric pressure, it is necessary to delay the start of the Roots pump until the backing pump has reduced the pressure to 100 mbar or to a use by-pass valve. This prevents the rotors becoming overheated, expanding and coming into contact with each other or the walls. Canned motors are sometimes used as drives for Roots pumps, where the rotor operates in vacuum but the stator windings are at atmospheric pressure, separated by a non-magnetic sleeve. This removes the need for a shaft seal and is often used when clean gas recovery is needed. (Hyper Link for Roots Blower Animation)

The RSV “B”models have a by-pass valve and manifold integrated into the Roots stator. The concept allows extended operation at a high inlet pressure or pump down from atmospheric pressure simultaneously with a mechanical roughing pump. The by-pass action is activated by high differential pressure inside the Roots vacuum chamber which opens the by-pass valve. The “excess” gas is metered to the mechanical pump with a portion recycled into the Roots to assist in cooling the lobes during high pressure operation. The by-pass reduces the demand of the drive motor, reducing energy consumption and the need for an external pressure switch.

If higher speeds are required near 1 torr, Roots Pump is used as a booster to backing rotary vane pump, whose individual speed is falling down. Both sides bearings/ one side synchronizing gears, located in separate chambers. Shaft seals both sides to isolate the auxiliary chambers from the pumping chambers Higher pressures heating problems sound Pressure ratio at atmosphere – 3 & at high vacuum 40 - 50 by pass valve for initial evacuation. Two roots blowers in series < 10-4 Torr

Valid for all roots pumps

0

20

40

60

80

100

0.0001 0.001 0.01 0.1 1 10 100 1000

Inlet pressure(Torr) ---->

Rel

ativ

e pu

mpi

ng s

peed

---->

In compression mode the staging ratio can range between 2 - 15 while the compression ratios achieved range between 5-40, depending upon combination selection. Initially, pumping is initiated at atmospheric pressures by Rotary pump and after achieving the recommended cut in pressure the booster is switched on. A bypass line around the booster may be provided for the initial pump down period. Boosters with hydrokinematic / electronic drive are also available which allow simultaneous start-up of the booster & the fore pump. This initial pumping by fore pump is necessary since pumping gas at high pressures with the booster generates considerable heat and the power input is also considerably higher. For this reason the booster is generally switched on at cut-in pressures of 20-60 Torr. A suitable vacuum switch can be installed between the booster & the fore pump, set for cut-in pressure, so that the booster is switched on only on achieving the designed cut-in pressures. However, for short duration the booster can withstand excessive differential pressure across it. The Booster-Rotary Pump combination are generally recommended when speed of 3000 LPM or higher are required since the combination is most economical and power saving than any rotary pump of similar capacity.

Combination 1: Everest Booster EVB30 Backed by 3000 LPM rotary pump Single stage Combination 2: Everest Booster EVB30 Backed by 3000 LPM Rotary pump Double stage

PUMP PUMPING

SPEED POWER ULTIMATE

VACUUM Rotary Pump 5000 lit 10 HP 2 × 10-3 Rotary Pump (5HP) + Roots Blower (2HP)

5700 lit 7 HP 2 × 10-4

RRoottaarryy eecccceennttrriicc ccyylliinnddeerr vvaaccuuuumm ppuummpp Instead of employing moving vanes a tube (F) of rectangular cross-section which is a sliding fit in an auxilIary small cylinder, connects the gas inlet port to the rotor or plunger. This plunger is mounted eccentrically about the motor driven revolving axle (E) and is in two parts: the inner drum (C) rotates with the axle (E) but the cylindrical shell (D) is a sliding fit on (C) and, since it is rigidly attached to the inlet sliding tube (F), will not rotate with (C) but undergoes a rocking motion, whereby the point G (where there is close contact between the plunger and stator) sweeps round the inner wall of the stator.

As the plunger moves in the direction of the arrow it rapidly creates extra space at (A) into which some of the gas is admitted through the inlet port. Simultaneously, compression of the gas previously trapped in volume (B) is taking place. When the plunger has almost reached its highest point it expels all air or gas and surplus sealing oil through the outlet valve and nozzle (H) into the oil separator tank, where the oil is retained and the air or gas is discharged into the atmosphere. A baffle trap is incorporated into air exhaust line to trap any oil mist which is carried over when pumping large quantities of air.

1. Housing 2. Cylindrical piston 3. Eccentric 4. Pump chamber open to intake port 5. Hinge bars 6. Flat slide valve 7. Oil-immersed pressure valve 8. Filter 9. Intake port 10. Exhaust port 11. Baffle 12. Shut-off pump chamber 13. Temperature regulator 14. Gas ballast duct 15. Oil-drain plug

These days most mechanical pumps are fitted with the gas ballast facility and in the example shown in figure the air admittance hole to the chamber is seen to be just to the right of the air inlet tube. Cross-section of a rotary piston pump

Operation and processes of a Rotary piston pump

Position 1. T.D.C Position 2. The slot in the intake-duct of the slide valve begins to open. Start of intake process. Position 3. B.D.C. The slot in the intake duct is completely open. The gas to be removed (arrow)

flows freely into the pump chamber (shaded). Position 4. The slot in the intake duct is closed again by the hinge bars. End of the intake

process. Position 5. T.D.C. Maximum volume of pump chamber Position 6. Just before the start of the compression processes, the front end of the pump

piston moves away from the gas ballast orifice so that it is exposed. Start of gas ballast cycle.

Position 7. Gas ballast orifice completely free. Position 8. End of gas ballast cycle. Position 9. End of pumping processes.

OOiill VVaappoouurr BBoooosstteerr PPuummpp Capacities available: Up to 23000 1 s-1 (50,000 ft3 min-1) and throughputs of up to 1500 Torr 1 s-1 at

0.1 torr. Operating pressure range: 1 torr to 10-4 torr using water-cooled baffles. A typical 3- stage vapour booster is illustrated in Fig.2.10 and consists of an annular jet followed by an ejector jet. If the pump consists only of an ejector stage then it is generally referred to as vapour diffusion pump, and requires initial evacuation to below 1 torr. The booster pump differs from the diffusion pump primarily with regard to boiler pressure, which is normally about 30 torr. This is achieved by using a high heater input and volatile oils such as pentachlor-diphenyl, which has a vapour pressure of somewhat less than 10-4 torr at 15ºC. The oil return to the boiler is arranged to flow through pipes, and during operation the head of oil established in these return lpipes balances the boiler presdsure. The hydrogen speed of this type of pump is normally about twice its air speed. The vapour booster pump exhibits considerable plumping speed for permanent gases below its ultimate pressure, which is limited to 10-4 to 10-5 torr by the vapour pressure of the oil. The high critical backing pressure, between 2 and 6 torr, allows the use of relatively small backing pumps. Vapour booster pumps are used for high-speed duties in the pressure range 10-1to 10-3 torr, where rotary pumps are at their limit and diffusion plumps unstable.

Fig. Vapour booster pump They are particularly suitable for dirty and mainly hydrogen loads such as are widely encountered in metallurgical applications in combination with a small rotary pump they provide a compact arrangement for backing large diffusion pumps. Slotted Cathode: A slotted cathode presents a surface to a portion of the impinging ions such that glancing incidence and high sputtering rates occur, see Fig. 2.13(a). The bottom of the groove is then subjected to this high sputtering rate and argon ions buried at the bottom of the groove are substantially covered and permanently trapped. This arrangement increases the argon speed from 1% to 6% of the nitrogen speed.

(a

Fig. Methods for increasing pumping speed for argon (a) Slotted (b) Triode pump (b)

Triode pump: The electrode arrangement of the triode pump is shown in Fig. 2.13(b) and consists of a stainless steel anode held at earth potential a titanium cathode in the for a of an open structure honeycomb held at minus 5kV and a collector which is normally the plump envelope and hence of stainless steel and at earth potential. The pumping mechanisms are exactly the same as in the diode plump. The significant difference arises from the fact that most of the positive ions striking the cathode do so at glancing incidence so that there is a substantial increase in the amount of titanium sputtered from the cathode. The bulk of the sputtered titanium showers onto the collector and will cover any inert gases that stick to the collector. With the honeycomb type of cathode argon speeds are raised to about 30% of the rated air speed. The principal advantage of the triode pump is its increased pumping speed for the inert gases, it should be used when inert gas pressures exceed 1x 10-7 torr, and a secondary advantage is a some what faster start up compared to diode plumps and the ability to start readily at high roughing pressures of up to 10-1. Titanium Sublimation Pump

Capacities available: Up to many thousands of liters per second. Operating pressure range: 10-3 torr (normal upper limit) to below 10-11 torr.

Titanium is evaporated from a tungsten filament over wound with titanium wire or from a filament of titanium molybdenum ally (McCracken, G.M & Pashley, N.A.,1966) onto a substrate or vacuum chamber wall. Active gases are pumped by chemical combination, but there is no pumping sped for inert gas or saturated hydrocarbons so that sublimation plumps are always used in conjunction with diffusion or sputter-ion plumps. The rate at which gas is taken up by the titanium, film, is determined by the rate at which titanium is sublimed, the chemical nature of the compound formed, the nature of the film and the gas access to the film.

Filaments designed for high-vacuum application are normally operated at constant voltage and the filament current is used to indicate the completion of useful filament life. Cartridges carrying multiple filaments are available and control units have the facility of switching between filaments and also provide automatic sublimation cycles.

Impeller Blades

Discharge Suction

Liquid Ring

Impeller

Shaft

Suction PortDischarge Port

Fig. Operating Principle of Liquid Ring Pump

Liquid Ring Vacuum Pumps Liquid ring vacuum pumps are used to evacuate the environments having condensable vapours or wet loads. They are used through out the process industries. These pumps are the only alternative to steam jet ejectors for handling large amounts of wet vapour or liquids and small amount of solids as the pump operates in a liquid environment. It is ideal for wet processes such as filtration, drying, condenser exhausting and distillation. Figure shows the schematic diagram of a typical liquid ring vacuum pump. It is basically a type of rotary positive displacement pump. In this pump, liquid is used as an element to compress the gas molecules. The compression is achieved by a ring of liquid and a rotating multi blade impeller located eccentrically in the pump casing. The eccentricity causes the filling and emptying of the each rotor (impeller) chamber. The concentric liquid ring is formed as a result of centrifugal force by the rotation of the impeller. The liquid ring also seals the space between the impeller blades and the casing. The pump is driven by an electric motor at standard speeds from 400 to 1750 rpm.

At the inlet port, the space available between the impeller blades and the liquid ring increases with rotation of impeller. Due to this increase in area, suction of the process load takes place and the gas enters the space between the impeller blades and the liquid ring. Further rotation isolates this trapped gas from the inlet. Now onwards the area between the impeller and the liquid ring starts reducing. This will cause the gas to be compressed. After the desired compression is attained, the gas will escape through the discharge port. The liquid ring acts as coolant to absorb the heat of compression and heat due to friction. Further, it also acts as condenser for condensable vapours.

Operating pressures of liquid ring vacuum pumps are in the range of 760-100 torr for single stage and 760-25 torr for two-stage. Liquid ring vacuum pumps with different capacities upto 6000 m3/hr are available. Any type of liquid can be used as sealant as long as it is not prone to vaporization (and thus cavitation) at the process conditions. Popular sealant liquids are water, glycol, mineral oils and organic sealants. Compression is accomplished without any metal-to-metal contact. This eliminates the need for lubrication and reduces pump wear to a minimum. The clearances between metal surfaces are large compared to other mechanical pumps. This allows the pumps to handle small solid particles as long as they are not abrasive. Liquid ring pumps are used as backing pumps for steam jet ejectors, mercury ejectors, mechanical boosters and condensers. Combinations of liquid ring pumps and steam jet ejectors reduce the noises and maintenance hazards. Finally, the advantages of liquid ring vacuum pumps are simple design, ease of fabrication, low cost, minimum noise/vibration, compatibility to liquids or vapours. The disadvantage of liquid ring vacuum pump is the limitation of the vacuum attainable by the sealant liquid due its vapour pressure.

SStteeaamm JJeett EEjjeeccttoorrss Steam jet ejectors are widely used in various process industries, especially for operations such as exhausting fumes; exhausting air from condensers; vacuum evaporation, distillation, crystallization; refrigeration; filtration; drying; air conditioning, and for pumping large volumes of vapours and gases at low pressures. They provide the best way to produce a vacuum in the process plants because they are rugged and simple in construction—therefore, easily maintained. Their capacities can be varied from small volumes to enormous quantities (30000 m3/min). They are simple in design, have no moving parts, operate with cheap, readily available fluids, and are reliable in service. Ejectors are basically momentum exchange pumps and the working of a typical single stage ejector is shown in figure.1. The ejector consists of three basic parts: steam nozzle, suction chamber and diffuser. High-pressure motive steam expands adiabatically in a converging-diverging nozzle. As a result, the velocity of the motive steam reaches to supersonic value having a mach number 3 - 4 (typically 1000 m/s). During the expansion process, the motive steam expands to a pressure below the suction fluid pressures. This will act as the driving force for the suction of fluid. Both steam and the process load mix in the mixing chamber. The resulting mixture is still supersonic. This supersonic mixture is passed through a diffuser where the velocity is reduced and the pressure is increased. In the converging portion of the diffuser, the velocity is reduced as the area is reduced. The throat section is designed to reduce the supersonic velocity into sonic velocity. Thereafter in the diverging portion of the diffuser, the velocity is further decreased as the area increases. This causes an increase in pressure (exhaust pressure). The capacity of steam jet ejector system can be increased by connecting them in parallel. The ultimate vacuum can be improved by staging them in series. The ultimate vacuum attainable with different stages of ejectors is given in table. Materials used in construction of steam jet ejectors are stainless steel, bronze, gun metal, cast-iron, Monel, carpenter 20, hastalloy, and titanium. Non-metals that are frequently specified for highly corrosive applications involves porcelain, impervious graphite etc.

Table: Vacuum Ranges for Different Ejectors

No. of stages Minimum practical absolute Pressure Torr

1 50 2 5 3 2 4 0.2 5 0.03 6 0.003

Ejector motive steam requirement increases as the compression ratio across the ejector increases. An ejector discharging to the atmosphere can be designed for a compression ratio of 20: 1, but economics normally limits the maximum compression ratio to approximately 10:1. Intercondesners are used in between the stages to reduce the load on the next stage. A precondenser is also used to reduce the load on the first stage ejector. Figure.2 shows an example for a four-stage ejector with intercondensers. Intercoolers are used between stages to condense steam from preceding stage or stages, thus reducing the load to be compressed in succeeding stages. The first stage discharges directly to the second stage because the interstage pressure (typically 3 to 6 torr) is too low to permit condensation of motive steam using a water-cooled condenser. The first stage and the second stage operate as a two-stage “noncondensing” unit. Two-, three-, four-, and five stage condensing jets are routinely specified for process applications. Six-stage systems have been built for steel degassing and other metal processing applications.

Velocity Profile

Pressure Profile

Mixing Zone

Diffuser Section

Nozzle

Suction Fluid Mixture

Motive fluid

Suction Port

Motive Fluid

Fig.1. Schematic diagram of Steam jet ejector with pressure and velocity profiles

Z stage Y stage W stage

X stage

Inte

rcon

dens

er 1

Inte

rcon

dens

er 2

Aft

erco

nden

ser

Barometric Legs

Fig.2. Multistage Steam Jets with Condensers

Measurement of Pressure in Vacuum Systems 1. Primary Gauges: Respond directly to the pressure of ambient. Manometers, Mcleod gauge

(10-7 Torr). 2. Secondary gauges: Respond to the pressure dependent property of the rarified gas. 10-3 T –

Thermal Conductivity - Pirani, Thermo Couple. < 10-3 T - Ionization Current - Bayert Albert, Penning.

Mc Leod Gauge : (1874)

To system

atm Vacuum System

A

Mercury

h2

h0C1

C1

Bulb

h3 By lowering mercury, bulb is brought to the system pressure.

Compress the bulb gas into capillary C1.

C1 & C2 same dia - capillary effect is same

U-tube manometer Mc.Leod gauge

The pressure of the compressed gas in the closed capillary is P + (h2 - h1) where 'P' is system pr. According to Boyle's law [P + (h2 - h1)] A (h0 - h1) = PV Where V Bulb vol., A Cross sectional area of capillary.

[P + (h2 - h1)] A (h0 - h1) = PV P = A(h2 - h1) (h0 - h1) /[V - A (h0 - h1)] Bring the mercury level upto h2 = h0 (end of the closes C1) or h1 = hs standard level. For h2 = h0 and h2 - h1 = (Δh1) P = A (Δh1)2 /[V - A(Δh1)] ≈ (A/V) (Δh1)2 -------------------------------------- (1) For h1 = hs and h0 - hs = const. P = A(ho - hs) (h2 - hs) /[V - A (ho - hs)] ≈ (A/V) (h0 - hs)(Δh2) -------------------------------------- (2) where, (Δh2) = h2 - hs 1st method - pr. is proportional to square of the reading. 2nd method - pr. is proportional to the first power of the reading - linear scale. By properly designing the volume of the bulb and the diameter of the capillary, we can use this gauge as primary standard down to 10-7 Torr. Not to be used for condensable vapours. Cold trap to be used to avoid Hg vapour in system. Discontinuous measurement.

To system

A

Mercury

h2

h0C1

C1

Bulb

h3

Mc.Leod gauge

SSeeccoonnddaarryy GGaauuggeess ((TThheerrmmaall CCoonndduuccttiivviittyy))

These two bulbs with their platinum filaments are as nearly alike as practical

Wheatstone bridge circuit

This bulb is highly evacuated and sealed

To system

R3

R2R1

Pirani vacuum gauge

Heater

Microammeter Milliammeter

115 V 60 ∼

To system

Thermocouple junction

Calibration curve for a commercial thermocouple gauge

Principle:

1. Both depend on the fact that the thermal conductivity of a low pr. Gas depends upon the pressure.

2. Cooling of the heated element changes with

pressure. Pirani - This is picked up by change in Resistance. Thermocouple - This is picked up by change in the equilibrium Temperature.

IIoonniizzaattiioonn GGaauuggee :: HHoott CCaatthhooddee Electrons from hot filament are accelerated towards anode, miss it and oscillate, thereby ionizing the gas molecules.

The negative gas ions are collected by ion collector set at a negative potential. This ion current is a measure of pressure.

System

Anode +150 V

Filament OV

Ion Collector -20V

This gauge works down to 10-8 Torr. Photo electric emission error.

_+

Cathode

Anode System

CathodeH.V

Penning Gauge: (1937) Cold cathode Electrons are emitted due to field emission discharge (small in number). These electrons are made to travel in helical path due to high electric and magnetic field ionise the ambient.

The positive ion current is measured in the anode circuit and is calibrated in terms of pressure units. This gauge is rugged, no heating required sensitivity is low at lower pressures.

B

+150V- 40V

Filament

Anode Grid

Ion Collector

System

Anode +150 V

Filament OV

Ion Collector -20V

BBaayyaarrdd -- AAllppeerrtt GGaauuggee:: ((11995500))

The residual current found in a conventional ionisation gauge is caused by photo electrons ejected from the ion collector by soft X-rays produced by 100-200 V electrons striking the cylindrical anode grid. External circuit current is both due to ions incident on collector (α pressure) and Photo electrons emitted from collector (error). At 10-8 Torr Photo effect is 100%. Bayard - Alpert Inverting (or) exchanging the positions of filament and ion collector, this problem is solved. The filament is now placed outside the cylindrical anode grid, and the ion collector which is now a fine wire rather than a large area cylinder, is suspended at the center of the anode grid. As usual electrons from hot cathode accelerated to grid Ionization by collisions A large fraction of ions collected by center wire. But solid angle presented by the ion collector to the X-rays emitted by the anode grid is heavily reduced (factor of several hundreds).

∴ X-rays error limit extended to 10-11 Torr.

TABLE : Calibration Factors (True pressure = indicated pressure x gauge calibration factor)

Gas Penning and Ionization Gauge

(approx.)

Alphatron ® Gauge

Air Hydrogen Helium Argon Neon Nitrogen Oxygen Carbon Monoxide Carbon Dioxide Methane Water Vapour Mercury Vapour

1.0 3.1 7.7 0.9 3.9 1.0 0.9 1.0 0.7 0.7

0.85 – 1.16 0.37

1.0 4.0 4.8 0.85 1.6 1.0 - -

0.6 -

1.16 -

Interpretation of Spectra: The below figure Illustrates a typical residual gas spectrum traced by a strip chart recorder. The spectrum was obtained with a nude ion source quadrupole analyzer fitted with an electron multiplier from an unbaked sputter-ion pumped system. The pertinent data were a nitrogen equivalent total pressure of 1.3 × 10-9 torr.

Absolute Pressure

Px

Getter

Differential Pressure

PxPr

CCaappaacciittaannccee MMaannoommeetteerr ((SSeeccoonnddaarryy ssttaannddaarrdd))

Typical arrangements for the pressure sensor of a capacitance manometer are shown in figure. The differential pressure design can be used for absolute measurements by connecting the reference pressure port Pr to a high vacuum pumping system. A more practical variation makes use of a permanently sealed reference cavity, avoiding the need for a pumping system on the reference side. With either variety, the latest sensor technology incorporates a single electrode on the reference side of a thin metal diaphragm under radial tension. The higher pressure on the process side will deflect the flexible diaphragm toward the reference side. A deflection in the metal diaphragm causes a change in capacitance between the diaphragm and the electrode. The change in capacitance is converted to a frequency change, which is passed through another converter to produce a change in the output voltage of the sensor. The signal from the sensor is modified in a unit called the signal conditioner and the output is then registered on an analog or digital meter and/or recorder as a pressure change. The output can also feed a pressure controller.

With the single-side sensor, all that the incoming process gases encounter is a deflection baffle to prevent the high-speed impingement of incoming particles on the diaphragm. The process side can even be cleaned with solvents if necessary. Diaphragm deflections as small as 10-9 in (10-11m) can be detected with capacitance manometers. With the thinnest diaphragm this corresponds to a pressure of about 10-7 torr.

The useful range for accurate measurements extends down to approximately 10-5 torr. The use of stronger diaphragms allows for pressure measurements at higher pressure. Capacitance manometers are sensitive to changes in ambient temperature since sensor materials like the ceramic electrode and metal diaphragm have different thermal expansion coefficients. Temperature compensation can be used successfully if the sensor temperature does not fluctuate rapidly.

BBoouurrddoonn GGaauuggee Because these gauges are rugged and inexpensive, they are the usual choice as a permanently mounted local gauge in rough vacuum systems. The bourdon gauge named for its main pressure-sensing element a bourdon tube. This is a thin walled metal tube with an elliptical cross section, which is usually bent into an arc resembling the letter C. The tube is closed at one end and open to the system pressure at the other. A sketch of the gauge internals is shown in figure.

Pointer

Pointer leakage

To vacuum system

Bourdon tube

Bourdon-tube vacuum gauge

The outside of the tube is subjected to atmospheric pressure. If the gauge is connected to a system under vacuum, the inside of the tube will begin to be evacuated. The resulting difference between the internal and external pressures will cause the Bourdon tube to curl inward and indicate a lower pressure through the pointer linkage. An increase in pressure will produce the opposite effect, with the curved element opening slightly to indicate a rise in pressure. With proper design and adjustment of the pointer linkage, the relationship between movement and pressure changes can be made very nearly linear. Since the Bourdon element moves with respect to the difference between atmospheric pressure and the measured pressure, readings are affected by change in the local barometric pressure. The measured pressure should be corrected by using the barometric pressure and calibrated at the time of the reading in calculating the absolute pressure.

MMeecchhaanniiccaall DDiiaapphhrraaggmm GGaauuggee

The pressure –sensing element is a beryllium copper capsule which has been evacuated to a pressure well below the lower limit of gauge pressure measurements and has been hermetically sealed. The out side of the capsule, as well as the linkage mechanism and the pointer, are all exposed to the unknown system pressure with in a sealed instrument housing. As the gauge is evacuated, the sealed capsule begins to expand. The flexible capsule face acts like a diaphragm, pushing a connecting rod as it expands. The capsule expansion is eventually translated into pointer movement by the linkage mechanism. Since the capsule flexure is with reference to essentially zero pressure (the pressure inside the hermetically sealed capsule), the indicated readings are not affected by the barometric pressure. Measured pressure is total system pressure independent of gas composition. Typical gauge accuracies range from 0.05 to 1.0 percent of the full-scale reading. Pressures approaching 0.5 torr can be accurately measured on gauges designed with thinner sensing capsules. Since their entire instrument housing of this gauge is exposed to the process gases and vapors, the gauge must be protected from liquids and dust particles, which could enter the gauge and bind the linkage mechanism. The gauge can hold its calibration for years if reasonable care is exercised in its use.

Pointer

Capsule

Capsule Stop

Calibration Adjustment Pinion

Geared Sector

Backlash Eliminator Revolution Indicator Flexure

(a) front view (b) gauge internals.

Mechanical diaphragm gauge

CCaalliibbrraattiioonn ooff VVaaccuuuumm GGaauuggeess The calibration of any vacuum gauge is important taking into account the pressure range of interest and accuracy necessary to ensure operation within this pressure range. Perhaps the most important feature of any gauge is the repeatability. One calibrate the vacuum gauge by measuring it relative to a direct reading standard gauge or by measuring the gauge at a fixed known pressure. In the region from atmospheric to about 0.5 torr the use of U-tube manometers of one sort or another, corrected for local atmospheric pressure and temperature. In the region from 0.5 torr to 10-5 torr, McLeod gauges with calibrated volumes can be used as standards, but care must be taken to eliminate condensable vapours and the possibility of mercury contamination of sensor and system the spinning rotor gauge can also be used as a calibration standard. Static Expansion Method: In this method a known small volume of a known gas (usually nitrogen or argon) of known pressure measured by direct reading gauge such as a U-tube is allowed to expand into a large volume. Then all, or in most cases a small volume, of the expanded gas is allowed to expand into a larger volume, and so on, until the desired pressure is obtained. With care, one can obtain known final pressures as low as 10-7 torr. Gauges can be calibrated at the final pressure or at several points in the expansion process. Dynamic Method : A gas introduced into a vacuum chamber at a constant known throughput Q. At the same time the gas is pumped at a constant speed S. The equilibrium pressure P can be calculated from P = Q/S and used as a reference pressure for calibration.

P1 V1 = P2 (V1 + V2)

Pressure gauge

V1 V2 V3

PPaarrttiiaall PPrreessssuurree GGaauuggeess//RReessiidduuaall ggaass aannaallyyzzeerrss It has already been noted that the sensitivity of pressure gauges varies for different gas species so that if it is important to have an accurate knowledge of total pressure, it is necessary to know the composition of residual gases and appropriate gas sensitivity. Further more for many application of vacuum technology, it is more important to know the identity of residual gases rather than the actual value of total pressure. For example for surface studies, it may necessary to know that the partial pressure of particular chemically active gases in the residual system gas are at or below a certain level, or that others are not present at all. Thus there is need to incorporate a partial pressure analyzer (PPA) into a system. The PPA instrument is more frequently referred to as a residual gas analyzer (RGA). An RGA is a essentially a mass spectrometer designated specifically for investigating residual gases in vacuum system so that it has normally has a higher sensitivity but lower resolution than a conventional analytical mass spectrometer, and more limited mass range-typically 1-100 or 1-200 amu. The typical mass spectrum of unbaked vacuum system is given in Figure 6.10 It can be seen that the largest peak is at mass 18, which is that of water vapor. Devices in which mass separation is based on either magnetic sector or electrostatic quadruple filtering had come to be the most widely used. These instruments also serve as detectors in leak detection of vacuum systems, explained later. The working principle of these spectrometers are now discussed below.

Typical mass spectrum of an unbaked vacuum system

2 28 36

44

18 Water vapour

WWoorrkkiinngg PPrriinncciippllee ooff mmaassss ssppeeccttrroommeetteerr A positive ion of mass 'M' and charge "ne", moving with velocity 'V' undergoes a deflection in the magnetic field B; which results in a path along the arc of a circle of radius 'R' given by Mv2/R = B nev -------------------- (1) The velocity 'v' is acquired by acceleration through the potential difference 'V' volts ne.V= 1/2 M v2 -------------------- (2) From (1) and (2) M = [B2 R2 ne]/2V Now for given values of R & B, M = K ne/V Where K = 1/2 B2 R2 Usually n = 1 . Therefore Scan of various values of V , is also a scan over the mass spectrum

d.c amplifier

m

Intermediate Slit

M2 > m

m1 < m

Slit S

Ion Source V

R

Ion Source Entrance aperture

Filament

Ion Path Quadrupole rods

Collector

Electron Collector

Residual Gas Analysis Quadrupole mass spectrometer

Working range 10-4 to 10-12 Torr. The analyser in this spectrometer does not require magnet. The ion source and collector are connected in a straight line path. The focused ions travel along the symmetry axis of the four-rod structure in case of a quadrupole analyzer, receiving r.f and d.c. voltages. Ions passing through the analyser are filtered according to mass by a quadrupole electric field. The collector current is proportional to partial pressure of allowed constituent. The R.F. field causes a charged particle traveling in the z-direction to oscillate. The amplitude of oscillation is a function of both the mass number of the particle and the voltage at the electrode.

Vdc + Vac Cos ωt

Ions, where amplitude of oscillation remains smaller than 'ro' (distance from central z-axis to a rod electrode) are free to pass through the quadrupole field. The field parameters can be adjusted so that the filter will pass only the mass 'M' related to applied voltage, so long as Vdc/Vac ≤ 0.167. The filtered mass number is given by M = Vac/ω2 r2

o = const. The spectrum is obtained by scanning r.f. voltage while keeping the ratio or r.f. to d.c voltage remains constant. Species of different e/m ratios are thereby brought one by one into the region of stable paths. Minimum detectable partial pressure is as low as 5 x 10-14 mbar. Identifies gases at levels of 1 - 10 ppm. While the sensitivity and resolution of a magnetic sector analyser varies with slit width, the same parameters are electronically controlled in the in the r.f. quadrupole. The sensitivity of the instrument varies with the efficiency with the ion source and its filtering capacity. All these factors are closely inter-related. In RGA work it is necessary resolve adjacent peaks separated by one mass unit so that the minimum absolute resolution needed piece unity. At present such RGA systems with full computer control are commercially available to qualitatively analyse the residual contents in any vacuum application.

Fig. Typical residual gas spectrum

Resolution and Sensitivity: Resolution is a measure of the ability to separate adjacent mass peaks on the recorded output of the RGA and sensitivity is the comparison of the current measured at the detector with partial pressure of the component. Sensitivity at a specified emission current is defined as, Sensitivity =

ssurepartialprecurrentpeaksignal amps torr-1

Sensitivities arte measured by introducing a purer gas into a UHV system containing RGA and measuring the total pressure (equal to the partial pressure under these conditions) with a calibrated ionization gauge. Interpretation of spectral: Figure illustrates a typical residual gas spectrum traced by a strip chart recorder. The spectrum was obtained with a nude ion source quadrupole analyzer fitted with an electron multiplier from an unbaked sputter-ion pumped system. The pertinent data were a nitrogen equivalent total pressure of 1.3 x 10-9. Qualitative interpretation: The spectrum is qualitatively described as containing approximately equal parts of hydrogen (mass 2) and nitrogen or carbon monoxide (mass 28) together with approximately half as much water vapour (Mass 18) and carbon dioxide (mass 44) and traces of methane (mass 16) and argon (mass 40) .

Cracking pattern: A single gas species gives rise to multiple peaks in a mass spectrum. A primary peak occurs at the mass corresponding to a singly ionized molecule, for instance at amass 32 for oxygen (O2

+) and mass 18 for water vapour (HOH+). Secondary peaks occur due to multiple ionization giving an oxygen peak ate 16(O2

++ or O+) or due to dissociation of the molecule within the ion source giving water vapour peaks at mass 17(OH+), mass 16 (O+), mass 2 (H2

+) and mass1 (H+). Magnitudes of the secondary peaks as percentages of the primary peak are given as a cracking pattern. Nitrogen equivalent partial pressure: The nitrogen equivalent partial pressure exerted by any one peak can be quickly but approximately calculated by dividing the peak current.(A) by the nitrogen sensitivity corresponding to the resolution used. In the present example the nitrogen sensitivity corresponding to the resolution used. In the present example the nitrogen sensitivity is 1 x 10-4 A torr-1 and the nitrogen equivalent partial pressures (x 10 –11 torr) are listed in column 4 of Table 4.6 A total nitrogen equivalent pressure obtained by summing these partial pressures gives 5.2 x 10-10 torr compared with the measured pressure of 1.2 x 10-9 torr. A better way of calculating nitrogen equivalent partial pressures is tko consider the contribution that any one peak makes to the total and equating this total to the measured nitrogen equivalent pressure. Thus the hydrogen partial pressure is (97/316) x 1.3 x 10-9 torr, and all such partial pressures ( x 10-11 torr) are given in column 5 of below the table. True partial pressures: The most significant factor affecting the peak height of any one mass is the ionization probability within the ion source, and since ion source conditions are similar to those used in ionization gauges the ion gauge calibration factors given in Table 4.2 are used to correct the nitrogen equivalent partial pressures. The data obtained in columns 4 and 5 are presented iln columns 6and 7 respectively of Table 4.6 with ionization gauge calibration factors incorporated. The true total pressures obtained by summing these true partial pressures are 8.6 x 10–10 torr and 2.2 x 10-9 torr respectively.

Partial pressure ( × 10-11 torr) analysis of a residual gas spectrum

1 2 3 4 5 6 7 8 Species Mass

No. Peak

height Recorder divisions

N2 equiv.

pp from peak

height

N2 equiv.

pp from peak

height

True pp from

Column 4

True pp from

Column 5

True pp

Hydrogen Helium Methane Water vapour Neon Nitrogen Carbon monoxide Argon Carbon dioxide

2 4

16

18

20

28

40

44

97 2

20

49 2

(appox. 10)(appox. 88)

98 4

44

16

.33

3.3

8.2

.33

16.2

.66

7.3

40

.82

8.2

20

.82

42

1.7

1.8

50

2.5

2.4

8.2

1.3

16

.58

5.2

125

6.4

5.9

20

3.2

42

1.6

13

179

3.6

7.1

12.5

5.2

16.2

.58

6.5

The true partial pressures indicated above have not taken into account the variation of sensitivity of the analyzer for different gases. Corrections are applied from calibration data for different gases supplied by the manufacturer or plotted by the user, in the form of sensitivity vs. true resolution or as sensitivity vs. nominal resolution. In the first case peak resolution has to be experimentally measured since resolution will vary to some extent with mass over a single scan (particularly important for low masses); in the second case this variation is automatically taken t into account. True partial pressure obtained with such data are given in column 8 of Table 4.l6 and the corresponding true total pressure is 2.3 x 10-9. The largest variation that occurs with the additional refinements in interpretation arises with hydrogen and helium since they are light ions with large ionization calibration factors, and in particular with hydrogen which often forms a major constituent of a residual atmosphere. The total gas load of a vacuum system originates from the following sources: Release of gases and vapours from system surfaces.

(a) Release of gases and vapours from within system materials (b) Permeation of ambient gases through permeable materials, notably elastomer gaskets and

glassware. (c) Leakage. (d) Release of gases and vapour from virtual leaks, notably caused by poor fabrication methods

or liquid nitrogen trap warm-up. (e) Back migration of gases and vapours through or from a diffusion pump, or release of

trapped gases from a sputter-ion pump.

Gas Flow at Low Pressures

Atmosphere Low Vacuum High Vacuum Viscous flow (Intermediate) Molecular flow Viscous λ <

100d (flow governed by viscosity)

Turbulent Re = ρνd / π > 2200(flowing gas layers are not parallel)

Laminar (Re < 1200)(flowing gas layers are parallel)

MMMooollleeecccuuulllaaarrr λ >> d (no viscous effects, molecules independent)

φ = d

φ = d

φ = d

Conductance in Vacuum Systems Rate of flow "Q" of a gas through a pipe which offers resistance to flow is proportional to the pressure drop (P2 - P1). Q α (P2 - P1) Q = C (P2 - P1) C = Q / (P2 - P1); C conductance C units Lit/min same as pumping speed, S C = f (pipe dimensions, properties of gas etc). If conductances are connected in series If pump of speed 'S' is connected by a pipe of conductance 'C' to a chamber, the effective pumping speed S' is given by 1/S' = 1/S + 1/C.

Backing Line

Rotary Pump

Valve

Diff. Pump

System

Roughing Line

Q

System to be evacuated

“Seff”

Pump “S”

1/C = 1/C1 + 1/C2 + 1/C3 + ………

C1

C1

C2

C3C1C2

If parallel

C = C1 + C2 + C3 + ………

Conductance in high Vacuum lines: Mean free path >> lateral dimensions of the channel Flow at these pressures is molecular flow if PD < 15 P Average gas Pr in microns D Lateral dimensions in cm In such case, the conductance of a circular pipe is given by C = 12.1 (28.7/M)1/2 ( T/293)1/2 D3 / L M Molecular wt, T Abs. Temp. D, L diameter & length of the pipe. This formulas neglects the end correction. The conductance of an aperture is CA = 11.6 (28.7/M)1/2 ( T/293)1/2

A. A Area of the aperture in cm2. For circular aperture CA = 9.1 (28.7/M)1/2 ( T/293)1/2 D2 The net conductance (both tube & orific contributions) is usually given by C = (28.7/M)1/2 ( T/293)1/2 f D2 f Dimensionless factor which depends on 'L/D' ratio.

L D

If Seff should be 75 l/s 'D' ? 1/S' = 1/100 + 1/C C = 300 l/s C = (28.7/M) .( T/293)1/2 f D2 T = -2000C, M =2

300 = 29373.

27.28 f D2

fD2 = 159. From figure, D is found to be 11.5 cm. Practical vacuum systems may contain one or more bends in connecting lines. For right angle bend it is common practice to increase the length by one or two diameters.

EX:

System H2 -2000C

Diffusion. Pump

100 L/S

System Roughing Line

Rotary

Diffusion pump

CCoonndduuccttaannccee ooff rroouugghh--vvaaccuuuumm lliinneess The mode of gas flow in fore vacuum region may range form viscous flow (λ<D) to molecular flow (λ>D) Viscous flow PD>500 P- pressure in microns D in cm 15<PD<500 transition region Here flow is mixed, having some attributes of viscous flow and some of free molecular flow. From poiseuille’s law of viscous flow, conductance of a long cylindrical tube is C = ( π /128η) (D4/L)P C cm3/sec, η viscosity in poises (C.G.S) ,P pressure in dynes/cm2 C = 3.3 x10-5 (D4P/ηL) c lits/sec, P microns C α P, P must be known to get ‘C’ The conductance in mixed flow 15<PD<500 is C = 3.3 x10-5 (D4P/ηL) + (10D3/L) for air at room temperature Ex: Find the conductance of a pipe (I.D-2.5cm, L-200cm, Air, T= 20OC, Pmean = 25 microns) Since PD = 25x2.5 = 62.5, we are in mixed flow. Therefore, C = 3.3x 10-5 (D4P/ηL) + (10D3/L) = 3.3x 10-5 ___2.54 x 25___ + 10 x 2.53 1.78 x 10-4 x 200 200 = 0.905+0.78 = 1.68 lit/sec use this ‘C’ in calculating Seff by 1/Seff = 1/S + 1/C http://www.e-insite.net/semiconductor/index.asp?layout=article&articleid=CA170179

SORPTION The group of interactions in which the gas in retained by the solid received the name 'sorption'

SORPTION ADSORPTION ABSORPTION (Gas enters into the solid) PHYSISORPTION CHEMISORPTION

Adsorption process whereby gas (adsorbate) molecules are attracted to and become attached to the surface of a solid (adsorbent). Physisorption Attracting forces are physical. Chemisorption Attracting forces are chemical. Gas can be removed effectively from a volume by adsorption, a process utilized in "sorption pumps". "Pumping Efficiency" depends on starting pressure of the volume, adsorption coefficeient of the adsorbent for a perticular gas and the temperature of the adsorbent. Usually adsorbent is kept at -1960C (LN2) to reduce desorption and increase adsorption net effect is evacuation Cryosorption pump.

Energy (E)

Distance from surface (h)

Ha = Ed

E

h

EA

EP ED

HC

Chemically adsorbed molecule

TThheerrmmooddyynnaammiiccss ooff AAddssoorrppttiioonn A) Physisorption:

It involves vanderwalls inter molecular forces, like those occuring in liquefaction of gases. The attracting forces are comparatively weak and the "heat of adsorption" is small (≈8K cal/mole). Since the forces are attractive work is done in adsorbing molecules and heat is generated, thus the adsorption is exothermic phenomena. The gas molecule assumes an equilibrium position at minimum potential energy, called heat of adsorption HA and is equal to energy of desorption, ED.

HA = ED

B) Chemisorption:

Attractive forces are much larger than in physisorption. Heat of chemisorption ≈ 250 Kcal/mol.

Two step process Molecules are initially adsorbed physically and the, with the provision of "Activation Energy EA" they are chemisorbed. Energy of desorption = Heat of chemisorption + Activation Energy ED = HC + EA The overall chemisorption process is "Exothermic".

SORPTION PUMP

The Sorption Pump Consists of a st. steel body with internal copper fins to facilitate heat transfer to the zeolite. A liquid nitrogen container is attached to three support brackets. As the temperature of zeolite (molecular sieve) falls, it sorbs more gas from the system to cause reduction in the pressure. After pump down to the equilibrium pressure, the valve to the system is closed. At this stage the molecular sieve is saturated. The reactivation can be carried out by allowing the pump to warm up to room temperature, care being taken to vent the pump.

Two sorption pumps connected for multistage pumping

Vent

Valves

Rubber stopper

Pump being reactivated

Liquid nitrogen pump

Operation

Dewar

Molecular sieve

Schematic diagram of a Sorption pump

Rubber stopper

Liquid nitrogen container support

14”

9.5”

4.5”

These pumps are used in ultra high vacuum system for preevacution to 10-2torr, when other UHV pumps can start. Rotary pumps cannot be used (oil vapour contamination). Some times it is necessary to take the molecular sieves to 3000c for several hours to drive of water vapour. The rubber stopper acts like safety pressure release valve. Sorption pumps are generally used to pump from atmospheric pressure, in 10-2 torr provided the sorpive capacities correctly matched to the volume of the system. The final pressure achieved by sorption pumping can be improved by pre pumping the sorption pump with another sorption pump or a mechanical pump. P – t Curves : Sequential pumping reduces the pressure of N2 but cannot achieve lower ultimate pressures due to Neon & Helium, which cannot be adsorbed at 77K. If mechanical pump is used as first stage to pump the system, it reduces, the partial pressure of Ne &He so that second stage sorption pump results lower pressure. But the vacuum is not oil-free which is necessary in U.H.V systems.

0 10 20 30 40 50

10-4

10-3

10-2

10-1

1

10

102

P(to

rr)

20 60 1010-3

10-2

10-1

1 ---- Volume = 90 l ---- Volume = 20 l ---- Volume = 2 lit

t (min)

---- 1st stage: sorption +

2nd stage: sorption

---- 1st stage: rotary

2nd stage: sorption

Cold cathode ion - pumping

Ion Pumps If a gas is ionized and the resulting +ve ions are accelerated to a negatively charged plate, atoms of the gas are effectively removed. In order to increase the pumping efficiency, sorption and gettering phenomena are combined with ionization. Getter: A material able to chemisorb gases. Ex : - Titanium.

Hot cathode – pumping Cold cathode ion – pumping Collisions of gas molecules with electrons emitted from a hot filament

Collisions of gas molecules with electrons emitted from a cold cathode discharge.

Getter is intermittently vaporized by Thermal

evaporation.

Evapor – Ion pump

Getter is evaporated by sputtering (cathode sputtering).

Getter – ion

Sputter - ion pumps

Hot cathode ion - pumping

Ionization

Ion – sorption pump Getter – ion pump

+ Gettering + Sorption

Ion pump

IIoonn PPuummppiinngg Electrical gas discharge Ion produced Chemisorbed by a Getter Pumping

Efficiency ∝ PI+

I+ Ion current, P Pressure

I+ ∝ number of molecules entering the pump per unit time (throughput ‘Q’)

I+ = KQ

KSPQK

PI ==+

⎟⎠⎞

⎜⎝⎛

∴ S, Pumping speed = P

Iβ+

=

K

β 1= → constant of the pump.

Unlike the diffusion pump, Ion pumps do not require a fore pump to pump the collected gas to atmosphere, since the pumped gases are in effect trapped inside. How ever a pre vacuum (10-3 torr) is necessary to start the pump. Once Ion pump is started, the roughing pump can be isolated.

EEvvaappoorr –– IIoonn ppuummpp

Combine ion pumping with gettering process. Gettering effect is used both during evaporation (dispersal gettering) and in the form of fresh

film on a surface (contact gettering). The gas is ionized to ensure transport by electrical pumping of the inert gases (which are not

gettered) to the getter (usually Titanium) coated wall at which they are made to arrive with energies of few hundred electron volts. At these energies 20% of the ions are retained and embedded in the film as the fresh getter is vaporized.

Small, single pumping operation:

Hot cathode ionization gauge + getter mechanism. First evacuate to 10-3T. Fire getter, active gauge get sorped. Operate ionization gauge, drive +ve ions to ‘Ti’

coated walls, where they are retained by subsequently deposited titanium. Application: - Improvement of vacuum in special tubes after sealing. Ex: X- ray tube.

Evapor ion pumps

Orbitron pump Large high speed with continuous ‘Ti’ feed

Small, single Pumping operation

Titanium coil on tungsten

Electrode (Formed by evaporated Ti

Anode ring

Filament

Glass envelope

LLaarrggee EEvvaappoorr -- IIoonn PPuummppss A spool carrying titanium wire, is externally controlled so that the wire is fed downward onto a post of refractory conducting material at 1000V positive with respect to pump wall. Electrons produced at the circular filament (100V +ve w.r. to wall) bombard the post and heat it to 20000C. This causes rapid evaporation of 'Ti' wire which then condenses on the cooled pump walls. This continuous evaporation of the wire ensures a continuous pumping action by both dispersal and contact gettering. A wire mesh grid, a +1000V also attracts electrons from the filament and these cause ionization of the gas, resulting +ve ions travelling to and being retained by the pump walls. Applications: Accelerators, large X-rays devices etc.

1. - wire feed spool. 2. - Titanium wire. 3. - Heated Post. 4. - Grid. 5. - Spool shaft driven from outside. 6. - Wire guide. 7. - Cooling water coil. 8. - Filament.

15

4

6

3

27 +100V (filament) +1000V (grid) +1000V

SSppuutttteerr--iioonn ppuummpp A potential of 3000V is maintained between the electrodes and a magnetic field of 1500 G is applied along the axis by external permanent magnet. Positive ions of system gas which are formed in the region of the electrodes are accelerated to the cathode and acquire sufficient energy to sputter titanium. The sputtered titanium condenses mainly on the open structure anode and chemisorb active gases by both dispersal and contact gettering. Insert gas ions (+ve) are propelled into the cathode and gets burried. Some times they may again get desorbed when the cathode is getting sputtered due to several complex mechanisms. Nevertheless good pumping is achieved down to 10-11to10-12T.

1. St. Steel anode (Honey Comb). 2. Titanium cathodes.

V+

V-

B 1

2

2

10

8

6

4

2

10-8 10-6 10-4 P(Torr)

H2

N2

O2

A

S (l/s)

DDiiaapphhrraaggmm ppuummppss Ultimate vacuum form 100 to 0.1 torr with pumping speeds up to 200 L/min. As compared to mechanical vacuum pumps, diaphragm pumps are made of the materials with high resistance to the chemical attack. Because of this reason these pumps are extensively used in the chemical laboratories. For attaining the oil free high and ultra high vacuum, the diaphragm pumps are used as the backing or auxiliary pumps. These diaphragm pumps are used in the coating and semiconductor industry, vacuum metallurgy, and analytical instruments where ultra high vacuum is required without any oil traces. The diaphragm is attached to the connecting rod with the help of diaphragm clamping disc and the connecting rod is attached to crank shaft is rotated with the help of the motor. The construction of the diaphragm pumps is similar to piston and cylinder arrangement. The diaphragm is attached to the connecting rod with the help of diaphragm clamping disc and the connecting rod is attached to crank shaft. The crankshaft is rotated with the help of the motor. Due to the rotation of the motor, the crankshaft is moved in linear direction and this will cause the diaphragm to move in forward and reverse direction. Due to movement of the diaphragm the volume between the cylinder head and the diaphragm is varying. As a result, the volume of the gas in between the cylinder head and the diaphragm undergoes compression and expansion. The inlet and outlet valves are located between the head cover and housing cover. The valves are opened and closed according to the pressure difference between the atmosphere and the volume of the chamber.

The working of the diaphragm pumps is explained below. The inlet port is connected to the chamber which is to be evacuated. The outlet port is connected to the atmosphere. The diaphragm is pulled down when connecting rod goes down causing the volume of the space to increase and the pressure decreases so that the inlet valve (port) is opened. The gas from the chamber enters the space and after reaching a certain pressure the valve is automatically closed. During this time the outlet valve is closed. In the next step the captured gas is compressed due to upward movement of the diaphragm. When this gas reaches a certain pressure, the outlet valve opens automatically and the high-pressure gas is expelled to the atmosphere. At the end of the stroke the diaphragm reaches the top position. During this process the inlet valve is closed. After the entire gas is vented to atmosphere, the exhaust valve closes and the diaphragm begins to come down and due to the pressure difference the inlet valve opens and the above steps repeat again.

Schematic section of a diaphragm Pump stage

1. Housing 2. Valve 3. Head cover 4. Diaphragm clamping disc 5. Diaphragm 6. Diaphragm supporting disc 7. Connecting rod 8. Eccentric bushing

The operating principle of a two stage diaphragm pump:

Opening and closing of the valves, mechanism of gas flow from inlet to outlet during one pumping cycle

The typical compression ratio of a single stage diaphragm pump is around 10. Single stage diaphragm pumps therefore have an ultimate vacuum around 70 torr. For lower ultimate vacuum diaphragm pumps are connected in series. In the series connections, the outlet of the first stage is connected to the inlet of the second stage. During the expansion of one stage the compression process is going in the second stage. The crankshaft is designed for adjusting the compression and expansion strokes. In the multi stage diaphragm pumps the crankshaft is designed such that the operations happen at the desired intervals. Due to the problems occurring in series connections, only four pumps are connected in series. Being totally oil free and made of materials with high resistance against chemical attack, diaphragm pumps can be used for pumping aggressive and condensable solvents. Large amounts of condensates may cause mechanical damage and thus reduce the life time of diaphragms and valves. It is therefore recommended that gas ballast be used on chemical versions of diaphragm pumps. The major restriction was mainly the wear of the diaphragms caused by the rocking of the connecting road. In addition, the stroke was comparatively small, and the pumping speeds and ultimate vacuum per “cylinder” were thus limited. Modern intrinsically corrosion resistant diaphragm pumps use PTFE compound materials for parts coming into contact with vapors and gases and are therefore ideal vacuum pumps for the chemical laboratory.

MMoolleeccuullaarr PPuummppss The principle of molecular drag based on the directional velocity imparted to gas molecules which strike a fast moving surface is applied in modern “TURBO MOLECULAR PUMP”. These pumps contain alternate axial stages of rotating and stationary discs & plates. The discs and plates are cut with slots set at an angle so that gas molecules caught in the slots of the moving disc are projected preferentially is the directions of the slots in the stationary plates. The running clearances between the rotating and stationary plates are of the order of 1 mm. the rotational speed for a pump having a rotor diameter of 17 cm ‘42000’ r.p.m.

D

B

C

A

(a)

Rotor & Stator

Inlet Outlet

Net gas flow s

b

φ φ

Vb

(b) V

Moving wall ‘V’ speed

With a comparison ratio of 5/stage, a 9. stage pump can have 59 = 2 × 106 ratio. Pumping speeds ------------- 3500 l/sec Ultimate pressure ------------- 10-10 Torr Backing pump effectively removes the gas from the exhaust at 10-2 Torr. The great advantage of molecular pump compared to diffusion pumps is that they are free from oil vapours. Under molecular flow conditions the rotor velocity approaching the average velocity of gas molecules, a significant pressure difference can be maintained with out developing back flow. λ >> D. Collections with the moving surfaces dominate the collisions with one another. ∴ The particles attain their wall velocity as an additional component.

b

a S (m3/h)

100 500 200

10-8 10-6 10-4 10-2

P (Torr)

a ----- 45 m3/h backing pump b ----- 10 m3/h backing pump

TTuurrbboo--MMoolleeccuullaarr PPuummpp

1. High Vacuum Connection 2. Rotor disk 3. Stator disk 4. Venting Connection 5. Molecular-drag-rotors 6. Molecular-drag-stators 7. Fore vacuum connection 8. Motor 9. High precision ball bearing with ceramic

balls 10. Operating medium reservoir

View of an axial flow pump. Two rotors and one stator are shown

Typical assembly of rotors and stators Transmission angles for a moving blade row

A molecule colliding with the moving blade at the entrance to the blade now can pass to the right side (angle 3) or return to the entrance (angle 1). A molecule at the discharge side has a much greater chance of remaining at the point of discharge (angle1). Thus the inclination of the blades produces a net flow toward the right for a row of blades moving downward. Role of Stator

Deceleration of flow (stator function) between two rotors

An illustration of stator function is shown in figure. Conside paddle wheels rotating within a rising trough. The wheel accelerates the fluid at a certain distance downstream from the wheel, the velocity is decreased and the pressure increased in accordance with Bernoulli’s principle. In this case the inclined plane between the two wheels serves the function of the stator (i.e to decelerate the flow and increase the pressure). With the following rotor were placed too close to the former (dashed circle), it would essentially be idle, because its paddle velocity will be nearly equal to the velocity of the fluid. In turbo molecular pumps the stator should be thought of as a baffle that redirects, straightens, or readjusts the gas flow to make the next rotor more effective. The stators could be entirely omitted if the distance between the rotors were perhaps 10 times greater than the diameter. For practical reasons the stator geometry should be such as to accomplish its function in a shortest possible distance while providing adequate conductance to flow.

At the very last stage, the stator can be and often is omitted. It would serve no useful function, because any surface in the discharge area is adequate to decelerate the flow.

TTuurrbboo--MMoolleeccuullaarr PPuummpp

Advantages

Roughing valve may not be necessary

High vacuum valve may be omitted

No contaminating motive fluid

Less chance of back-streaming accidents

Quick startup & shutdown

Less need for cold traps

Low operating expenses

Pumps all gases effectively

More predictable performance

Disadvantages

High initial cost • Advanced technology • Sophisticated machining • Advanced bearings

Oil lubricated

Grease lubricated

Magnetic Passive • Passive • Active

Precision Balancing

Moving parts subject to wear

Proper operation is important

Not available in large size

Pumping limitation with LIGHT gases

CCRRYYOOPPUUMMPPSS

Require Liquid Helium (4.2 K)

Refrigerator Cooled Cryopumps Liquid Pool Cryopumps

Continuous Flow Cryopumps.

Cryo Refrigerator

Cryo surface (20K)

Vacuum Chamber

Compressor

Vacuum

Cryo Surface (10K)

Coil

LHe

Vacuum Cryo surface (20K)

LHe

HPLP

Gaseous He

Water vapour (Chevron

Baffle)

Second stage can

(10 – 20 K)

First stage can

(50 – 80 K)

Remote temperature

monitor

Hydrogen Helium Neon

pumping

Nitrogen Oxygen Argon

pumping

First stage

Regeneration purge tube

Hydrogen vapour therometer (2nd stage temperature)

Schematic drawing of a typical cryopump

Gifford-McMahon cycle Cryorefrigerator Charcoal -- 1000m2/g, Typical cool down time 1 hr Clean vacuum 10-10 torr

Pumping speed α Area

ApS

= ⎟⎟⎠

⎞⎜⎜⎝

⎛

− f22f

4avpγ

f sticking factor Vavp average velocity at probe

CCyyrrooppuummppiinngg MMeecchhaanniissmm Cryopumping is the process by which gases are condensed at low temperatures in order to reduce the pressure. More analyzed this phenomenon by considering molecular flow of a gas between two infinite parallel planes. 1. Source 2. Condenser ASSUMPTIONS:-

1. L << λ 2. Condenser is covered with solid at T2 3. Fraction f of input stream sticks on the condenser. 4. (l – f)ω1 reflected molecules have a velocity distribution corresponding to T2. 5. Solid deposit emits molecules by evaporation at T2, and at the same rate as if it were in

equilibrium with a gas at T2. 6. The mass flow from source ω1 = W1 + ω2

ω1 = Mass flow emitted from source ω2 = Reflected from (2) and then (1) 7. The velocity distributions of all molecular streams are Maxwellian.

(ω1-w2)

L W1

T1 T2

ω1 ω2

f

2cW

(1) (2)

f

L W1

T1 T2

ω1 ω2

2cW

(1) (2)

ω1 = flowing from source to condenser ω2 = flowing from condenser to source. Net mass flow input W1 = ω1 − ω2

= ω1 − [(l − f) ω1 + W c2] ∴ Wc2 = Evaporation from condenser at T2 = ω1f − Wc2

ω1 = f2Wc1W +

ω2 = ω1 − W1 1Wf2cW1W

−+

ω2 = W1 (I − f)/f + W c2/f ω1 & ω2 expressed in terms of W1 & Wc2

If ω1 & ω2 are considered as being gas streams with molecular densities n1 & n2

Mass flow rate 41avγ1mn

A1ω

= , m = molecular wt., γ1av = Average mol.vel. from source

γ1av = ( 21

)πm18KT

) Maxwell – Boltzman statistics

∴ 21

)2π2mKT

(1nA1ω

= , similarly 21

)2π2mKT

(2nA2ω

=

n1 = Af2cW1W

21

)1mKT

2π(f2cW1W

1avmAγ4

+=

+

n2 = Af2cWf)(l1W

21

)2mKT

2π(f2cWf)(l1W

2avmAγ4

+−=

+−

The total molecular density is the average of these two expressions

Pp = n1 kT1 21

)1T

Tp( , Pp = ]21

)2TpT

(2T2n21

)1TpT

(1T1[n2k + , Pp = n2 k T2 2

1)

2TpT

(

Pp = Pressure inside an open ended probe Substituting for n1 & n2

Case – A :- Pp = Af2CW1W

21

)mpkT2

(+π

-------------- (1)

Case – B :- Pp = ]Af2Wc

2Aff)(21W

[21

)mpkT2

( +−π

---------------- (2)

Case – C :- Pp = Af2Wcf)(11W2

1)m

pkT2 +−π ---------------- (3)

n1 n2 T1 Case C T2

Case B ω1 ω2

Case A Tp Pp

Contribution to pressure from the molecules evaporation from the deposit is the same

Af2cW

21

)mpkT2

(2cP

π= ------------------------(4)

We have assumed that evaporation rate (

2cW ) from deposit- as if it were in equilibrium with a gas at T2

21

)2kT2

m(2vfPA

2cW

π= ----------------------(5)

From equation (4) & (5)

21

)2TpT

(2vP

2cP = ----------------------------(6)

21

)2TpT

(2vP

2cP =

i.e Contribution due to reevaporation from the condensing surface is equal to the vapour pressure of the deposit corrected for the probe temperature. Pressure due to net mass flow W1 to the system is dependent on f and probe orientation. If W1 >>

2cW (i.e 2cW ≈o)

Pp Pp f<<1 f = 1

Case A ‘’ A1W2

1]m

pkT2π ---------Max

Case B Af1W2

1]m

pkT2π 2A

1W21

]mpkT2π

---------21 Max

Case C Af1W2

1]m

pkT2π o --------- o

Pressure is independent of Very sensitive to Probe orientation Probe orientation.

Pumping speed is defined Sp = Q/p

∴ Sp is different for different orientations. As P changes with Probe orientation. Std. Cryopump:

Q Pp = npkTp , Q np = Pp/kTp , pAmP

pkT1W)Ap(mn

1WApS

==

Source Condenser

Probe

W1 = Mass flow rate gm/s, Density 8/cm3 Substituting for Pp and on simplification

21

)m2pKT

](21

)2TpT

(pP2vP

[1f22f

ApS

π−

−=

Where 2vP is given by

21

)2kT2

m(2vfPA

2Wcπ

=

We know γav p = 21

)8(πm

kT

γav p = Average velocity at probe

]21

)2

(21[422

TpT

pPvPavp

ff

ApS

−−

=γ

Q

To get maximum pumping speed we require 2vP << Pp. Then

4]22[/A

Sp avpf

fMaxγ

−=

Pumping speed is independent of Pr. For f << 1 Sp/A = f γavp/4 And for f = 1, Sp/A = γavp/2 With chernove baffle and radiation shields f G overall capture probability.

)4)(2

2( avpG

GApS γ

−=

ω4

gs -------- Probability that a molecule impinging on chevron shields will pass through. gc --------- Probability that a molecule impinging on the condenser will pass through. f --------- Sticking coefficient G --------- over all capture probability F (gs,gc,f, etc) W1 = ω1 − ω2 ω2 = (1 − gs) ω1 + gs ω4 ω3 = gsω1 + (1− gs) ω4 ω5 = gc ω3 + (1− gc)(1− f) ω5 ω4 = gc ω5 + (1− gc)(1− f) ω3 Simultaneous solution of all these equations gives

G = ]22)1(2)1)(1()2)(1)(1(1

22)1(2)1()1)(1(21[11

cgfcgsgsgfcgcgfcgfcg

sgW

−−−−+−−−−

−−−+−−−=ω

The pumping speed of such an array per unit area of the cryosurface is given by

4].22[ avp

GG

ApS γ

−=

ω3

gs f gc

ω5

Source Condenser Back Shield 2 3 4 Cold Baffle

ω5 ω2

ω1

20K 100K

100K

CCyyrrooppuummppiinngg AArrrraayyss Cryo pumping surfaces cannot generally be exposed directly to a source of gas at room temperature because the heat load due to radiation would exceed that due to be condensation of gas molecules. Therefore the cryogenic surface is protected on the side facing the gas surface by an optically opaque baffle at an intermediate temperature to act as a radiation load.

Fig. Shows some common an ansements used in space simulation.

20K100K

100

3 4 5 6 1.Radiation 2. Gas 3. Chevron baffles (100K) 4. Condenser 2K 5. Back shield 6.Chamber wall (300K)

1 2

100K20K

100K

20K 100K

CYROPUMPING ARRAYS The pumping effectiveness of the array depends on the overall capture probability ‘G’ rather than sticking coefficient. G – Capture probability is the fraction of the total number of molecules incident on the inlet of cryopump, which is finally captured.

The figure illustrates the Conflat Flange, which is used up to seal diameters of9 in. This design fulfils all of the design –feature requirements given above. In particular, fixed com presion is achieved by bolting the flanges together and very high interface pressures are developed which are not relieved by baking since the gasket is initially a very good fit within the flange recess. The seal is made on the sloping surface of the flange using the captive part of the gasket, and the extent of the bite (0.15 in on each side) that is used to establish the high interface pressure produces sufficient gasket cold flow to seal imperfections and scratches up to 0.01 in deep.

Fig. Conflat and Wheeler seals

Couplings, Fittings and Feedthroughs

Vacuum System Design / Pumpdown Times/ Materials If η = coefficient of viscosity then η ∝ mv Where m = mass of molecule (g) v = arithmetic average molecular velocity = 0.918 C T Hence η ∝ T. It is surprising that η is independent of pressure. This is verified by experiment at normal pressures where molecular collisions are frequent, that is where the assumption of the kinetic theory holds well. At low pressures, less than 10 –3 torr, when the mean free path is 5 cm or more the molecules can travel directly from a moving surface to the walls of the vacuum chamber. Thus the momentum exchange between a moving surface and the walls or the drag on the moving surface depends on the number of molecules present, i.e. the pressure, and viscosity gauges are constructed to measure pressures below 10-3 torr. Thermal conductivity is the transfer of thermal energy whereas viscosity is the transfer of mechanical energy through a gas and kinetic theory gives an expression for the coefficient of thermal conductivity k similar to that for viscosity. In fact K = η Cv

Where C v = the specific heat of the gas at constant volume and is constant over wide ranges of pressures and temperature. With the same reasoning as applied to viscosity it is possible to construct a thermal conductivity gauge to measure below 10-1 torr.

Typically a clean unbaked stainless steel surface has an out gassing rate, after one to two hours pumping, of 10-8 to10-9 torr 1 s-1 cm –2. Outgassing rates are particularly influenced by surface history such as exposure to moist atmosphere, baking etc. At a pressure of 760 torr and 20ºC the time to form a monolayer of nitrogen molecules is 3 x10-9 seconds. At pressure of 10-9 torr and 20ºC the time to for a monolayer of nitrogen molecules is about 2000 seconds. These values have been calculated assuming an accommodation coefficient of 1, i.e. that an incident molecule sticks on first impact. The pressure in a vacuum system falls as gas is removed, and the rate of gas removal is the throughput of the vacuum pumps. Assuming the pumps. Assuming the pumping speed to remain constant, the throughput falls as the pressure falls, and is in fact exponential, given by:

P2 = p1 exp - St V Where p2 = pressure (torr) at time t (second), p1= pressure (torr) at time t = 0 , V= system volume (litres) and S = pumping speed (1 s-1). Thus the pressure falls to half its original value in 0.69 V/S seconds to quarter that value in twice as long, to 1/8th that value in twice as long again and so on. At low pressures the evacuation rate falls off as a result of leakage and outgassing from system walls. A general relation for vacuum systems sis that throughput= pumping speed times pressure, and this is generally used in the following forms: Required pumping speeds S = Q/p 1 s-1 = estimated total throughput. Required process pressure Required system pressure p = Q/S torr = estimated total throughput. Required process pressure Actual system throughput Q = S x p torr 1 s-1= available pumping speed x achieved pressure.

Q total = (qe + 0.1 qg) air + (0.9 qg) water vapor + (q0) oxygen.

The ultimate pressure attainable under these conditions is given by

)(q

)(9.0

)(1.0q

(total)P 0e

OxygenSvapourWaterSq

airSq gg ++

+=

Where S defines the pumping speed for the various constituents. A diffusion pump will have essentially the same speed for all of these constituents, but the pumping speed available at the chamber will vary, due to changes in the conductance of the accessories above the pump with the molecular weight of the constituent. A major change occurs if a liquid nitrogen trap is fitted, since it has a very high (cryo) pumping speed for water vapour, and water vapour often constitutes the largest part of the total gas load. A sputter-ion plump exhibits different pumping speeds for different gases. Fig: Pumping speed available for various constituents from a nominal 500 1 s-1 pump

Pumping Speed Pumping speed ls-1 for Diffusion

pump alone Diffusion pump + baffle

Diffusion pump + Liquid nitrogen trap

Sputter-ion pump

Air Oxygen Water vapour

500 500 500

270 260 300

180 175 1900

500 285 525

We can construct a composite pump down cure that will illustrate the relative roles of these phenomena.

Fig: Rate limiting steps during the pumping of a vacuum chamber

Figure shows the high-vacuum pumping portion of an unbaked, metal-gasketed system. In the initial stages the pressure is reduced exponentially with time as the volume gas is removed. This portion of the pumping curve takes only a short time because a typical system time constant is less than 1 s. It is expanded here for clarity. In the next phase, surface desorption controls the rate of pressure decrease. In a typical unbaked system most of this gas load is water vapor, however, nitrogen, oxygen, carbon oxides, and hydrocarbons the material and its history. Glass or steel that has been exposed to room ambient for extended periods may contain up to 100 monolayers of water vapor, whereas a carefully vented chamber may contain little water vapor. If the system has a number of large elastomer O-rings or a large interior surface area, a slow decrease in pressure is to be expected. Unbaked, routinely cycled vacuum systems are never pumped below this pressure range. If the system is allowed to continue pumping without baking, the surface gas load will ultimately be removed and the out diffusion of gases in solution with the solid walls will be observed. The sl-op of the curve will change from t-1 to t-1/2, for example, hydrogen that diffused into the steel in a short time at high fabrication temperatures diffuses out very slowly at room temperatures.

If we were continue pumping until the dissolved hydrogen was removed, the pressure would become constant even though the system was leak free. The system would now be at its ultimate pressure given by P = Q k /S. Experimentally, hydrogen permeates the walls of metal systems and helium permeates glass walls. Notice that the time required to reach the ultimate pressure in this hypothetical example of an unbaked metal-gasketed system is 10 8 h. This clearly demonstrates the absolute necessity of baking in ultrahigh vacuum technology. The order of importance of the processes shown in Fig.6.5 is not always as given here for this example. Elastomer gaskets have a high permeability for atmospheric gases, and if the system contained a significant amount of these materials the limiting permeation rate would be several orders of magnitude higher than illustrated here. In fact, it could be large enough to mask the diffusion process. Structural Metals / Ceramics /Glasses for vacuum systems

Metals : Metals are used both in the vacuum chamber and to form its walls. Each metal must have a vapor pressure low enough to prevent vaporization from occurring at the highest temperature encounte-red. Each should also have a low outgassing rate. The particular application in the chamber, such as the filament, radiation shield, and thermal sink, will more narrowly define the selection. Metals used for vacuum chamber walls should be easily joinable and sealable to one another and to ceramics and have strength. They should not be porous or permeable to atmospheric gases. This section presents the vaporization, permeation, and outgassing properties of several metals of interest along with the structural properties of austenitic stainless steels. Stainless steel and aluminum are the two metals most commonly used in the fabrication of vacuum chambers. Aluminum is inexpensive and easy to machine but hard to join to other metals. It is oft-en used in the fabrication of vacuum collars for glass bell jar systems which are sealed with elastomer O-rings, and in some internal fixturing. Because of these difficulties and its inability to make a seal via a metal gasket, aluminum has been largely bypassed in modern vacuum system construction.

Fig. Stainless steels used in vacuum equipments (AISI designation) CR = corrosion resistance. W = weldability, Y = yield strength, and M = machinability.

W

Y

W

316 CR

Y

W

W 302

M

304

304 L

304 LN

303

317 CR

316 L

316 LN

317 L

For ordinary laboratory high vacuum systems stainless steel is the preferred material of construction. It has a high yield strength, is easy to fabricate, and is stable. The AISI 300 series is basically an “18/8” steel that contains 18% chrome and 8% nickel. To this basic composition additions and changes are made to improve its properties. Carbide precipitation in stainless steel is a concern in certain applications. Carbon that has precipitated at grain boundaries as a result of welding or improper cooling after annealing removes with it a substantial fraction of the chrome from nearby regions. See below Figure. The nearby regions then contain less than13% chrome and are no longer stainless steel. They are subject to corrosion if exposed to a corrosive atmosphere; subsequent baking in the presence of water vapor can increase hydrogen permeation through the affected region. The formation of micro cracks is also a concern for stainless steel subjected to low temperatures. A crack in a carbide-rich zone can cause a leak in cold trap or cold finger. Carbide precipitation may be prevented by the use of an alloy with a low carbon content, a stabilized alloy, or a minimum-heat welding technique. The best solution is the use of low-carbon steel alloys such as 304L and316L, but they are not so strong as their higher carbon counterparts and may not suffice for some applications, because they require more nickel and are more expensive to manu-facture. It is assumed that all welding is done by the tungsten inert-gas (TIG) process to avoid oxidation. TIG welding, welding, also known as heliarc or argon arc welding, is a technique for forming clean oxide-free, leak-tight joints by flooding the arc with an inert gas, usually argon.

For some applications such as cryogenic vacuum vessels it is desirable to reduce the thickness of the structural steel wall to reduce heat loses. This can be accomplished without loss of any other properties by the use of a nitrogen-bearing alloy such as 304LN or 316LN or by cold stretching and annealing. Alloys that contain zinc, lead cadmium, selenium, and sulfur, for example, have unsuitably high vapor pressures for vacuum applications. Zinc is a component of brass; cadmium is commonly used to plate steel screws, and sulfur and selenium are used to make the free machining grades of 303 stainless steel. Under no circumstances should these materials be used in vacuum system construction. Glasses and Ceramics : A glass is an inorganic material that solidifies without crystallizing. The common glasses used in vacuum technology are formulated from a silicon oxide base to which other oxides have been add-ed to produce a product with specific characteristics. Soft glasses are formed by the addition of so-dium and calcium oxides (soda-lime glass) or lead oxide (lead glass) and hard glasses are formed by the addition of boric oxide (borosilicate glass). The physical properties of a glass are best described by the temperature dependence of the viscosity and expansion coefficient. Glass is brittle, and because of its high thermal expansion and low tensile strength it can shatter if unequally heated. Its expansion coefficient is therefore important when selecting the components of a glass-to-metal or glass-to-glass seal; for example, Corning 7720 glass is formulated to seal to tungsten, whereas the expansion coefficient of 7052 glass matches Kover. The expansion coefficients of soda lime and borosilicate glasses are so different that they cannot be sealed directly to each other but only through a graded seal consisting of 5 to 7 intermediate glasses whose expansion coefficients differ successively from one another by about 10 x 10-7 / °C. The viscosity-temperature and thermal expansion characteristics determine the suitability of a glass for a specific application. Borosilicate glasses are used whenever the baking temperature exceeds 350 °C, whereas fused silica is required for temperatures higher than 500°C. The thermal expansion coefficient and strength determine the maximum temperature gradient that a glass can with-stand and to what it can be sealed.

Glasses are used in the production of vacuum apparatus, ion gauge tubes, view ports, metal-to-glass seals, and internal electrical and thermal insulation. Solder glasses, a third class of glasses, have low melting points and are used to make glass-to-glass, glass-to-metal, or ceramic-to-metal seals. A review of modern seal glasses has been pub-shed by Takamori(35). A ceramic is a polycrystalline nonmetallic inorganic material formed under heat treatment with or without pressure. Ceramics are mechanically strong, with high dielectric breakdown strength and low vapor pressures. The general class of ceramics includes glass bonded crystalline aggregates, and single-phase compounds such as oxides, sulfides, nitrides, borides, and carbides. Because processes such as sintering wetted powers form ceramics, they contain entrapped gas pores and are not as dense as their crystalline counterparts. Their physical properties improve as their density approaches that of the bulk. Alumina is made with densities that range from 85 to almost 100% of its bulk density; most ceramics have a density of about 90% of the bulk. The important physical properties of ceramics are their compression and tensile strength and thermal expansion coefficient. High-density alumina, for example, has a tensile strength between four and five times greater than glass and compression strength 10 times greater than its tensile strength. Because of its high compression strength, it is not necessary to have a match as close as that of glass to a metal to which it is being sealed. This results in a more rugged seal than is possible between glass and metal. The properties of some ceramics are listed in Table 6.5. Alumina (Al2 O3) is the most commonly used ceramic in applications such as high-vacuum feed-throughs and internal electrical and thermal standoffs. Machinable glass ceramic also finds wide application in the vacuum industry for fabricating precise and complicated shapes. It is a recrystallized mica ceramic whose machinability is derived from the easy cleavage of the mica crystallites. Borides and nitrides have found application in vacuum technology. Evaporation hearths are made from titanium diboride and titanium nitride, alone or in combination. They are available in machinable and pyrolytically deposited form. Forsterite ceramic (2MgO:SiO2) are used in applications in which low dielectric loss is needed, and beryllia must be machined while carefully exhausting the dust, because it is extremely hazardous to breathe.

SSoouurrcceess ooff GGaasseess aanndd VVaappoouurrss iinn aa VVaaccuuuumm SSyysstteemm Figure describes the potential sources of gases and vapours in a vacuum system. The slow evolution of additional gases and vapours from the interior surfaces of vacuum system lengthens the pump-down time. This surface gas release is actually a result of several processes and is collectively referred to as outgassing. Gases and vapours released from the surface are a result of several physical processes like vaporization, thermal desorption, diffusion, permeation etc. Outgassing rates of steel, cooper, and aluminium after cleaning are given in Figure. With out prior heat treatment, these metals are desorb mostly water vapor.

Pump

4 53 2

1 7

System 6

1. Backstreaming. 2. Permeation. 3. Diffusion. 4. Desorption. 5. Vaporization. 6 & 7. Internal and

real leaks.

Outgassing rates for typical metals

Vacuum Materials

Properties Required The material must be able to be machined and fabricated to produce the component, adequately strong at the maximum temperature of operation and must not become brittle at the lowest operation temperature. The vapor pressure must remain adequately low at the highest operating temperature and the thermal expansion of adjacent materials must match and not cause any distortion or change of tolerance (e.g. flange bolts in UHV systems which undergo bakeout). The material must be impermeable to gases, must not be porous or have any cracks or crevices as these can create virtual leaks, trap dirt and retain cleaning fluids. The materials selected must not react with any other materials in the system and must not react with the vacuum process. The materials must not evolve excessive gas when subject to irradiation from neutrons, x-rays or high-energy particle beams, or degrade so as to be inadequate

MMaatteerriiaallss uusseedd iinn VVaaccuuuumm ssyysstteemm

1. High vacuum sealing at minimum material thickness 2. Low saturated vapor pressure at room temperature 3. Low out-gassing in vacuum and ease of degassing 4. Weldability

Stainless steel is ideal for UHV chambers Brass is ok for only poor vacuum as it contains Zn, which outgases. Weldability is good for stainless steel. Glass is good but not strong. Materials for scaling

Low / medium vacuum – Neoprine O – rings or Teflon or rubber etc. UHV – Copper/Gold gaskets

CCeerraammiicc,, GGllaasssseess aanndd GGrreeaasseess Ceramics : Fully vitrified electrical porcelain and vitrified alumina are most suitable for vacuum components and insulators and can be used up to a temperature of 1500oC. Care must be taken in handling these materials as they are brittle and the surfaces mark easily, so they should be handled wearing lint free gloves. Selected ceramics can be bronzed to suitable metals by the use of molybdenum - manganese fired coating onto the ceramic. Merchantable ceramics are available and are satisfactory for specific components. They may be machined, drilled etc following manufacturers' instructions.

Glass : Can be used for constructing vacuum systems or parts of systems. The most generally used is borosilicate glass (e.g. Pyrex), which can be obtained in matching components from stock. Glass is also used for viewing windows and selected optical glass can be used for the transmission of ultra-violet light. The permeability of glass varies with the temperature and glass has a high corrosion resistance.

Greases : These are generally mineral oil, silicone or Fomblin based and are used to lubricate moving shafts where they enter the vacuum system. Greases should only be used sparingly and are unnecessary on static flange seals if the ring and flange finishes are unimpaired. Some high viscosity pump fluids can be used as seal lubricants. Table: Working temperature of seals.

Seal Working temperature (oC) Neoprene 90 Indium alloy 90 Viton 150 Aluminum 200

Invar Iron (64%), Ni (36%) Very low thermal expansion coefficient Used in glass metal/ceramic to metal joints

Molybdenum Non magnetic refractory metal

Can be spark welded to the most of the metals Used in electron tubes Ideal for fillers, grids, anodes and supports etc.,

INCONEL (Ni 79.5, Cr 13, Fe 6.5)

Machinable and has good corrosion resistance at high temperatures Used in heating elements

CCoommmmoonnllyy uusseedd MMaatteerriiaallss Metals Austenitic stainless steels. Advantages: Types generally preferred are EN58A, EN58B(US321), EN58E (304) and EN58(347) with EN58B and EN58F chosen most frequently for satisfactory welding by an argon-arc. If low magnetic permeability required, EN58B is used. All have a low out gassing rate as shown in Figure. Disadvantages: EN58F will not accept a high polish and welding can cause considerable distortion in all grades. Aluminum and aluminum alloys: Advantages: aluminum and magnesium alloys are most used, but high zinc content should not be used. Good corrosion resistance, easily machined and jointed. Disadvantages: strength at high temperatures is poor. Alloys with copper content present welding problems. High distortion when welding, which may lead to further machining.

Outgassing rates for typical metals

Nickel alloys (Inconel, Kovar). Advantages: high strength at high temperatures, excellent corrosion resistance. Disadvantages: not readily available, high cost and present machining problems. Copper. Advantages: oxygen free, high conductivity grade (OFFC) excellent for vacuum systems, easily machined, good corrosion resistance. Brass. Advantages: suitable for some exceptional applications, care must be taken in grade selection, good corrosion resistance. The use of castings should be given careful consideration to avoid virtual leaks. Disadvantages: brass contains zinc (15-20 %). Zinc evaporates out at temperatures over 100oC. Mild Steel. Advantages: may be used generally down to 10-3 mbar, can be used at lower pressures if plated after welding. Disadvantages: liable to rust. Titanium. A good clean metal in vacuum, light in weight and ductile. Used mostly for its gettering properties e.g. ion pump cathodes, getter pump filaments.

SSeeaallss Elastomer seals Nitrile rubber: The most commonly used sealing ring is nitrile rubber, which can be easily jointed (in situ for large installations). Viton: Is most suitable for seals at lower pressures, has low outguessing rates and is heat resistant. It has a tendency to remain deformed for some time after compression. New viton should be vacuum baked at 100 oC for 1 hour to remove moulding release agents. PTFE: Has a very poor compression set and is not generally recommended. Metal seals Copper ring: Used with knife-edges machined into opposing flanges (e.g. Conflat). Can be baked up to 450 oC . Care must be taken that the knife-edges do not become damaged in store or during installation.

Out gassing rates for plastics often used for demountable seals

Aluminum diamond edged disc seal: Fits between flat-machined flanges. Self-aligning, bake out to 200oC. Flange finish required - 0.8μm. Indium wire: Very soft, continues to flow after initial tightening of flanges. Can be easily re-extruded. Aluminum wire: Easy to manufacture from annealed wire by electric butt-welding. Fits between flat-machined flanges. Requires aluminum foil centering tabs. Flange finish required - 0.8μm. Gold wire: Inert to all gases, bake out to 450oC. Needs high flange bolt loading and well-finished flanges faces (0.4μm). Expensive initially but good cost recovery for scrap.

GGLLAASSSS//CCEERRAAMMIICC -- MMEETTAALL SSEEAALLSS Glass-to-metal seals have been used for many years as the basis of electron tubes and vacuum devices. In order to form a glass-to-metal seal, the glass must wet the metal. As a rule, glass does not wet clean metal surfaces but it does wet an oxide-covered surface, hence metals must be carefully oxidized before sealing to glass. The second problem involved in a glass-to-meal seal is matching the coefficients of expansion (otherwise during baking or local heating the joints will leak) of the materials involved. Glasses are now available which have thermal expansion coefficients approximately the same as the more common metals used in vacuum technology.

One of the most common materials used in hard glass systems is Fernico or Kovar, an alloy of 54% iron, 29% nickel and 17% cobalt. It may be sealed directly to hard glass (Pyrex) and is available in many forms including electrical feedthroughs and optical windows.

Ceramic-to-metal seals are now more common than glass-to-metal seals; ceramics are stronger than glass, can withstand higher bakeout temperatures, have better dielectric properties and are less permeable to gases. There are four main categories of ceramic-metal seal: 1. diffusion seals; 2. sintered metal powder seals; 3. active alloy seals; 4. electroformed seals. The diffusion seal is achieved by pressing a metallized ceramic ring to a metal partner at several hundred degrees centigrade. In the sintered metal powder process, finely divided powders of tungsten, tantalum, molybdenum, rhenium, iron or mixtures of these, are mixed with an organic binder and applied to the surface of the ceramic. The piece is sintered by firing in wet hydrogen at 1250-15000C (depending on the ceramic). The resulting metallized ceramic is copper or nickel-plated and then brazed to the metal part by conventional techniques using silver bearing braze alloy in a hydrogen atmosphere.

Outgassing rates fro typical plastics and glasses

PPllaassttiiccss

Plastics generally desorb large quantities of gas and have a high permeability rate compared with metals, so that they must be carefully considered and generally kept to a minimum. Out gassing rates for some plastics are given in Figure.

PTFE: Low out gassing rate, good electrical insulator, can be used at a higher temperature than most plastics, self-lubricating. Glass-filled PTFE - a form of PTFE, which has fewer tendencies to cold flow.

Polycarbonate: Moderate out gassing rate and water absorption, good electrical insulator. Nylon and acrylic: High outgassing and water absorption rates, self-lubricating.

PVC: High out gassing and water absorption rates, flexible PVC tubing useful for backing lines and temporary connections (e.g. leak detectors).

Polyethylene: Only suitable if well outgassed. PEEK is a high temperature polymer, useful as an insulator where low temperature bake out is used. Electrical wires: Electrical wires coated with Kevlar have good electrical insulation. Ceramic beads are also useful and prevent trapped volumes. Synthetic resins Not recommended.

Fig.2. Rotary seal for simple thrust motion into

vacuum Fig.1. Two types of knife-edge vacuum seal with

metal gasket

Fig.3(a). Rotary motion into vacuum without sliding seals

employing bellows

Fig.3(c). Typical design of a ferrofluid seal

Flanges / Coupling / Seals

Fig.3(b). K-F Coupling

Fig.5. High vacuum Gate valve with packed shift seal

Fig.4. High vacuum Gate valve with packed shift seal

Fig.7. Disc valve

Fig.6. Diaphragm valve

Fig.8. Gas-admittance valve (a). Packless needle valve (b). Packless variable-leak valve

Vacuum Valves

Fig. Double O-ring gasket arrangement

If leakage is kept very low and good practice eliminates sources(e) and (f) the total gas load can be considered as originating from sources(a), (b) and (c) only, although in most cases sources(c) is so small that it can be disregarded. Under such conditions an out gassing rate (generally expressed as torr 1s-1 cm-2 ) is a measure of the rate of release of gases and vapours bound on and in constructional and process materials. The out gassing rate from an O-ring clamped between flanges approximates to the condition of freely exposing half of the surface area to the vacuum. Hot elastomers are highly permeable and give problems when exposed to differential pressure of an atmosphere when used as gaskets on ultra-high vacuum plant, in split of an adequately low final outgassing rate. These problems can be overcome by means of double gaskets on ultra-high vacuum plant, in spite on an adequately low final outgassing rate. These problems can be overcome by means of double gasket arrangement as indicated in Fig. The inters pace is pumped to as low a pressure as is convenient, using a rotary pump independent diffusion pump, or the backing diffusion pump where the main pumping group comprises a series connection of diffusion pumps. Rapid cooling of the gasket at the start of the cool-down cycle is also desirable so as to reduce permeation rates as quickly as possible. Typical permeation rates at room temperature for and atmospheric pressure differential for both nitrile rubber and Viton area 2 x 10-7 torr 1 s-1 cm-2 for dry air and 1 x 10-5 torr 1 s-1 cm –2 for water vapour (40% relative humidity).

At 100ºC these values would increase by a factor of 10. With the double gasket arrangement an improvement of some 103 would be obtained depending on inters pace pressure. Pressures below 10-10 torr have been attained using this arrangement with the gaskets cooled to –25º C . An advantage of the double gasket seal is that it can be quickly and reliably made without bolting, particularly in poorly accessible positions, to give minimum back to air times. Vacuum Brazing: Vacuum brazing is a process that can produce multiple complicated fabrications in a single operation. Small machining tolerances are required and the fabrication is securely jigged or spot-welded. Brazing alloy powder is supplied with a carrier that evaporates under vacuum, and a barrier past is applied to confine the spreading of the molten brazing alloy. Resistance heating is normally used to attain brazing temperatures between 1000ºC and 1200ºC and the furnace is pumped to about 10-4 torr. Pump-down and heating times are short but cool-down times are lengthy in vacuum. Cool-down times are reduced by forced internal circulation of an inert gas in conjunction with a water-cooled heat exchanger Advantages of vacuum brazing are that the fabrication is heated uniformly and distortion is eliminated, the fabrication is cleaned by the high temperature vacuum bake, the braze penetrates completely through the joint and is resistant to mercury attack or corrosion in the presence of water. Leaking welds can be sealed by vacuum brazing, and leaking brazes can be reprocessed since the braze alloys with stainless steel producing a joint that does not re-melt when reheated to the melting point of the braze power.

LLEEAAKK -- DDEETTEECCTTIIOONN An ideal vacuum chamber should maintain for ever the vacuum reached at the moment of its separation from the pumps. But a real chamber presents a rise in pressure after being isolated from the pump. Pressure rise can be either due to Leak or due to outgassing or due to both Leak rate Q = V(dp/dt) in T - L/S units. Real and virtual leaks Leaks are normally referred to as real or virtual, where in the former the gas passes from the external atmosphere into the vacuum chamber and the latter arises either due to the evolution of gases or vapors trapped inside the vacuum envelope in holes or channels or due to desorption of adsorbed molecules or vapors on the inside walls and components in the vacuum system. In the latter case the most important example of this is adsorbed water vapor.

dp/dt = x/2 dp/dt = x

Same leak

V

2V

Both Leak

Outgassing

Pumping

T Time

LN2 Isolation

Pres

sure

P

The leak rate QL can be defined as the quantity of gas, which enters the vacuum space per unit time from real or virtual leaks or both.

)/( dtdpvLQ =

Where dp/dt is the rate of rise of pressure in a closed volume, v, isolated from the pumps. If the system is being continuously pumped the rate of removal of the gas is equal to the rate of entry of the gas through the leak. Then we can write,

uPSLQ ×=

Where S is the pumping speed and Pu is the ultimate pressure in the system. Hence it is meaningless to speak about a 'large leak' unless we are able to specify the volume of the system. These considerations determine whether a leak is significant or not and in a continuously pumped system a leak may be present, which is small enough to be ignored. For example, in chamber with volume 125 litres and surface area 1.5×104 cm2, the total surface outgassing rate (for stainless steel) would be 1.5×10-8 mbar l s. With a pump of speed 1000 l s-1, the pressure with out leaks would be 1.5×10-11 mbar. If we now have leaks of various sizes, the resultant pressure is shown in Table below.

QL (mbar l s-1) QT (mbar l s-1) P (mbar) 1.0×10-5 1.00015×10-5 1.015×10-8 1.0×10-9 1.6×10-8 1.6×10-11 1.0×10-10 1.51×10-8 1.51×10-11

From these figures it can be seen that a small leak, well with in the detection limits of all teh helium leak detectors, has little effect on the final pressure. The decision has to be made whether a leak can be ignored.

SSiimmppllee mmeetthhooddss ffoorr lleeaakk ddeetteeccttiioonn Soap bubbles In this case the probe gas will normally be compressed air but greater sensitivities can be obtained using helium. The soap solution can be brushed on the suspected areas and bubbles will be seen if a leak is present, However it is more satisfactory if the chamber can be totally immersed in the liquid and the bubbles can be seen rising to the surface. Thermal conductivity detector As the thermal conductivity varies for different gases then a Pirani or a thermocouple gauge can be used as a leak detector. One of these devices is included in most vacuum systems and a leak can be indicated by an apparent change in the pressure when another gas e.g. carbon dioxide is directed over the leak. Leak detection over welds, joints, seals etc can be carried out using helium or carbon dioxide as the trace gas. Ionization gauge and ion pump detectors

Most high vacuum systems include some type of ionization gauge. The ion gauge current and the ion pump current are both dependent on the gas species in the system or pump and changes can be seen when a leak is present and covered with a search gas as used for the thermal conductivity gauge. Useful safe gases for both methods are helium, carbon dioxide and argon. Acetone and isopropyl alcohol can be used with care.

Detector probe In detector probe method, the test gas is filled into the system under test and a detector probe (sniffer) connected to the leak detector is passed over the suspected area to receive the test gas.

TTyyppiiccaall DDeetteeccttoorr

Magnetic deflection mass spectrometer

LLeeaakk llooccaattiioonn

1. Tracer probe technique. 2. Detector probe technique.

Test gas

Container 5

System 3

Detector

2

Pump 1

Probe

Fig. Tracer probe method

Electron Collector

Focussing electrodes

Iron Beam

Magnetic Field

Ion Collector

Angle of deflection is adjusted for Ze/m of the test gas ions.

Filament

Tracer Probe

A stream of test gas is spread on the suspected area and the gas penetrating into the system is pumped via the detector.

Flows through fine leaks. Partial Pressure is negligible in atmosphere.

Probe

Test gas

(He) Container

5

System 3

Pump 1

Detector

2

Fig. Detector probe method

The appearance and colour of a glow discharge can be used as a rough indication of system pressure and the composition of the residual gases. Data are given in Tables 4.1 and 7.1 TABLE 4.1: Characteristics of a glow discharge With a variety of test gases, notably hydrogen and helium, the Pirani gauge can detect leaks of the order of 5 x 10-3 torr 1 s-1. Backing line Pirani gauges are useful for high-vacuum system leak checking when the compression produced by a diffusion pump permits smaller leaks to be located. Volatile contaminants can be removed by flashing the filament to high temperature for about 2 minutes in a vacuum below 10-4 torr. Gauge heads should not be mounted horizontally to avoid filament sag, nor vertically downwards to avoid collection of dirt particles.

Pressure torr Appearance of discharge 7 – 10 1 – 0.5 0.5 0.3 0.1 0.05 0.03 0.005

General glow discharge Closely spaced striations in positive column 10 mm spacing of striations Crookes dark space 5 mm long Crookes dark space 10 mm long Crookes dark space 20 mm long Crookes dark space 30 mm long Discharge black out

Technologically Important :

1. Thermal evaporation (PVD). 2. Cathodic sputtering (DC/RF). 3. Chemical Vapour deposition (CVD) 4. Molecular Beam Epitaxy (MBE). 5. Electron beam Deposition. 6. Laser Evaporation.

VVaaccuuuumm bbaasseedd ccooaattiinngg uunniittss ffoorr tthhiinn ffiillmm ddeeppoossiittiioonn

x x x x B x x x x x x x x x x

Electron Bombardment Heating: Simple resistive heating Comtamination from support materials. Disadvantages. Input power limitations for high melting point Excess consumption of cost materials.

∴ Electron Beam Deposition by electron beam heating. beam focussed on a spot to cause spot melting without evaporating bulk of the material. Tungstem filament supply of electrons +ve potential accelarates then Magnetic

field to bend & focus then on one spot Electrons loose energy Rapidly Melting locally deposition

PVD By Resistive Heating Resistively heated filament/boat Mo, Ta, W, Nb. Alloy Films Two sources. Sequentially or simultaneous. Shutters/high current feed throughs/substrate heating and cooling attachments.

Filament H.V.-ve

Substrate

Electron beam

Spot Melting

Material Water Cooled Crucible

Substrate

HV (-v)

Vacuum

Ar

DD..CC//RR..FF.. SSppuutttteerriinngg Ejection of atoms from the surface of a material (the target, water cooled) by bombarding with energetic particles is called 'Sputtering'. Positive ions-bombard Cathodic sputtering Ejected particles on (Glow discharge) substrate for thin film Control V, I , Pressure to get uniform films of composition same as cathode. Ar gas used for sputtering ion production 20-100 m Torr. Ionization efficiency increased by transverse magnetic field Magnetron sputtering.

R.F (13.56 MHz) insulating targets For < 10-3 Torr, because of scarcity of ions, one uses Ion-beam-sputtering separately created differientially pumped.

Molecular Beam Epitary Recently availability of large capacity UHV Pumps - High quality thin films are bein made by Molecularbeams.

Effusion cells (Ga,As,Si,Cu,HighTc) for microelectronic devices

Gas in

UHV Gauge

Gate Valve

RGA

RHEED

UHV Pumps (Sputter Ion, Cryo, Turbo) < 10-10 Torr.

X, Y, Z Tilt/Rotation

Analysis Chamber

A typical oil-diffusion pump evaporation station. The notations are: A, quartz iodine lamp heater; B, substitute; C, quartz-crystal rate controller and deposition monitor; D, substrate mask; E, shutter(mechanical or electromagnetic); F, vapours from evaporation source; G, adopter/coller between the bell jar and the pump base plate flange; H, air-inlate valve; I, baseplate flange; J, Pirani or thermocouple gauge; K, roughing valve; L, liquid air trap; M, cooled chevron baffles; N, diffusion pump; O, cooling coils; P and Q, backing valves; R, Pirani gauge; S, fore-pump with air-inlet valve; T and U, diffusion-pump heater; V, filament holders; W, multiple feedthrough; X, ionization gauge; Y, Meissner trap; Z, baffles valve.(Reprinted from reference 3.)

The attainment of a superior vacuum is more easily accomplished in vessels for liquid helium storage because liquid helium provides a cryopumping surface which can pump all gases excepting helium. However, care must taken to ensure that there is no leak of helium from the inner vessel to the vacuum space.

Copper

High Vacuum

The Stainless steel

Liquid nitrogen

Charcoal adsorbent Liquid Helium

or Hydrogen

Fig. Simplified cross-section of a commercial dewar for the low-cross storage and transportation

of liquid hydrogen and helium.

Cryogenic Storage Dewars

The inner line is supported within the outer line by suitable low conductivity spacers. Further, some adsorbing materials are kept in contact with the inner pipe to help maintain the vacuum after closing the valve. Flexibility is introduced into the outer piping to withstand the thermal contraction effects of the inner pipe during transfer of cryogenic liquids. Stainless steel is usually used in making transfer lines to avoid outgassing at high vacuum levels.

Fig. Typical vacuum - insulated transfer line

Stainless steel flexible line

Inner pipe Outer pipe

Safety pressure relief

Vacuum Evacuation valve

Cryogenic Transfer Line

Vacuum Furnaces Vacuum furnaces are finding increasing application in the metallurgical processing of high temperature materials where pick up of Oxygen and Nitrogen both during melting and casting as well as during subsequent heat treatment is minimum. The vacuum environment is a pure protective atmosphere than any inert gas. The products are processed in vacuum furnaces with minimum contamination/distortion. High purity gas approximately contains a total impurity of 10 ppm = 10 in 106 1 in 105 at 750 torr (atm) 10ppm means 750 / 105 torr which is equivalent to a 7.5 x 10-3 Torr. Today vacuum furnaces having a conventional diffusion pump, mechanical pump can have a vacuum level of better than 1 x 10-3 torr. Role of Leak Rates When a typical vacuum furnace pumped down to 10-4 Torr using conventional diffusion pump/mechanical pump, the water vapour exists predominantly even after pumping for 4 hrs. This situation continues till the furnace is heated approximately to 6500C. Above this temperature the water vapour molecules start dissociating and react with the elements like Carbon (coming from residual Hydro carbons, Graphite heating elements and Graphite insulation). Thus the vacuum atmosphere can be treated as slight oxidizing till 6500C. This point has to be kept in mind while designing furnaces for treatment below 6500C. This will not be a problem during cooling cycles as the moisture inside the furnace is completely removed. Thus if a component is heat treated above 6500C and still get oxidized, it means the leak rates are not within acceptable limits. Even though vacuum atmosphere is better than inert/reducing atmosphere, the total atmosphere purity is drastically altered by leaks and adulterations from the product.

In case of furnaces which involve heavy degassing a roots pump (increasing the effective pumping speed of roots-rotary combination) is essential for maintaining the required backing pressure of the diffusion pump for its efficiency. In a vacuum furnace for leaks at 10-3 Torr (λ > d) where molecular flow exists, leaking gas (O2) molecules will have a tendency to touch the job before being removed by the pump. This results in oxidised product after the vacuum brazing process. Thus leak rate should be very small in high vacuum furnaces for brazing/sealing applications. Similarly during off period the vacuum furnaces should not be exposed to air as it results in trapped humidity, which will not be removed easily in the next run and contaminates the hot zone. Leak checking of vacuum furnace should be done under high vacuum conditions initially when job is not hot and then with job and finally with job heated so that trouble shooting is easier with out confusion. MSLD leak detectors are usually employed. Design of Vacuum Furnaces Vacuum furnaces are broadly classified into two types: A. Cold wall type : The vacuum chamber is water cooled and kept at room temperature. The job is

kept in a hot zone (resistance heating/induction heating) supported by thermally insulating structures. Here size of the hot zone is small and energy input is less. But degassing from the walls of the vacuum shell remains to be a problem to be solved.

B. Hot wall type : The vacuum chambers/muffle is externally heated (by resistive heating) and job

located inside receives heat mainly by radiation. Degassing/water vapour problems are absent. It can handle large quantities of jobs to be processed.

The design of vacuum furnace depends on : Materials being processed, size of the useful hot zone, melt capacity in case of melting & casting, Heating rate and temperature control accuracy and cooling schedule. Further Vacuum considerations such as pumpdown time, operating vacuum and tolerable leak rate also should be considered. Design of vacuum chamber considerations for cold wall type : Mechanical strength : The chamber should not collapse under vacuum. The thickness of the chamber should be optimised to stand the pressure difference (1 Kg/Cm2 + cooling water pressure in the outer jacket) and not to increase the mass and out gassing rate. Further it should be able to prevent in-leakages from outside atmosphere when it is pumped down to a low pressure and to permit the material to be processed without any damage to the chamber as well as to the product. However thinner sheets (to create lower weight) can also be used by re-inforcing the strength of the chamber by welding stiffeners. For vacuum applications, argon arc welding is preferred as the contamination from flux is totally avoided. If complete argon arc welding is not possible, we can weld few tax and then remaining by brazing with compatible metals/flux. It is essential to have good surface finish in all the grooves meant for 'O' rings,Wilson seals for rotating devices and door seals. Regarding shape cylindrical and rectangular chambers are very often used. The best method of cylindrical chamber fabrication is from a sheet cylindrically rolled up and dished heads on both ends welded throughout. Although this type of construction offers good mechanical strength, some designers select a rectangular cross section for efficient space utilization and use of low capacity vacuum pumps. However, rectangular chambers have to be designed for higher stresses and obviously require higher shell thickness.

With respect to orientation/operation the furnace chambers are classified as :

1. Vertical type, bottom loaded for heavy jobs. 2. Vertical type, top loaded where top dished end can be lifted by an over head hoist and the job can

be loaded by a crane. If necessary System can be designed in such a way that both the chamber and the top dome can be lifted for loading the job(horijantally by a forklift) to avoid striking of the jobs with the heating elements/shields .

3. Horizontal type, front loaded (by opening the door) for long light weight components/flat pieces etc.

For the purpose of external water cooling most of the manufacturers prefer double walled configuration as the cooling efficiency is maximum and so the furnace can be designed for very high temperatures (14000C). The basic disadvantage in this design is the repairing of any water leak developed in the inner chamber.

1. One can also have single walled configuration (advisable upto 11000C) with channels welded/copper tube soldered on the outer surface of the chamber.

2. This has lower cooling efficiency but repairing is quick.

Materials for chamber fabrication:

For the fabrication of chamber, either hot rolled mild steel or stainless steels of grade 304, 304L and 316 are used. Vacuum theorists do not prefer bare hot rolled mild steel for the chamber, as the material will be rusted during usage. This not only will affect the compressive strength of the material but may also increase the surface available for condensation of water vapour, whenever the chamber is exposed to atmosphere. However, economic considerations, mild steel may be used but with a certain special surface treatment. By coating the inner shell walls with epoxy (gas loads from such paints are minimal) mild steel vessels are widely used for many applications. Whenever there is any suspected water contamination, a hot argon scrub is to be given.

The arguments in favour of a polished stainless steel chamber are the availability of less surface area for condensation of water vapour and a low degassing rate of the polished surface. Some designers prefer a stainless steel inner chamber and mild steel outer chamber due to economic consideration. However, it is essential to take care of the quality of the weldments, as galvanic corrosion of welded joints can occur due to water circulation. Stainless steel 304 grade weldments are susceptible to pitting corrosion when the sediments present in the water accumulate near the weldments, especially when such welded joints are present at the bottom of the chamber.

Even though stainless steel grade 316 can overcome this problem of pitting corrosion, the associated welding problems and high cost prevent its widespread use.Stainless steel grade AISI 304L, with low carbon content, is the most suitable material for a vacuum furnace chamber for Indian condition. If there is any possibility of heat dissipation to the 'O' rings, the flanges have to be water cooled. The various types of materials used for 'O' rings for operations upto 1 x 10-6 Torr are neoprene (upto 500C) silicone (upto 1000C), and viton (upto 2000C).

Chamber design for hot walled type: Since hot walled furnaces are heated externally, the furnace shall be designed keeping in view the yield strength, modulus of elasticity and creep properties of the shell material at the rated temperature. This restricts the usage of hot walled furnaces above 9000C. However, with the development of better class of nickel base super alloys, presently hot walled furnaces are being designed for certain metallurgical applications upto 11000C for short duration operations. Basic design calculation procedure applicable to cold wall type are also applicable here. For hot walled furnaces, design of the cooling water circulation near the flanges needs careful consideration as those areas are more prone for thermal stresses. Generally, these designs have internal insulation near the flange end, which enables in decreasing the temperature near the flange much below the actual furnace shell temperature. For hot walled furnaces, cylindrical shapes are only used.

For most of the metallurgical processes, such as sintering, brazing annealing, and melting & casting vacuum of the order 1 x 10-6 Torr is adequate. Hence, the vacuum system consists of a combination of the following pumps: i) Mechanical vane pump/piston pump. ii) Roots pump. iii) Oil diffusion pump. The combination includes other sub-assemblies such as chevron baffles, valves and pipe line. Regardless of the efficiency of the pumps, the cross sectional area of the pumping line, valves, manifolds etc., must be equal to or more than the inlet diameter of mechanical pump itself used as the backing pump for the diffusion pump, the diameter of the pipe line upto the inlet of diffusion pump must correspond to the diameter of the diffusion pump while the dia. of pipe from the outlet of diffusion pump must correspond to the inlet diameter of the mechanical pump.

Vacuum valves : The various types of valves used in the pumping line are : i) Angle valves ii) Ball valves iii) Butterfly valves iv) Gate valves. Gate valves are the best suited for a pumping line, as they offer minimum impedance to gas flow. For high vacuum line, angle valves fabricated out of stainless steel are also used. The other type of valves i.e., ball valves and butterfly valves, are used in the low vacuum mechanical pumping lines. Electromagnetic valves help in automatic closing of the valves in case of power failures.

HHoott zzoonnee ffoorr ccoolldd wwaalllleedd ffuurrnnaacceess Resistance type: The primary criterion for the hot zone of any furnace is that the entire volume of the work piece should experience the same temperature. The important parameters on which the thermal uniformity of the hot zone depends are size, shape and location of the heating elements, rate of heating, the type of insulation/shields used etc. In general, the hot zone of the furnace consists of: i) Suitable heating elements. ii) Thermal insulation/radiation shields to conserve the heat. iii) Ceramics for electrical and thermal insulation. Heating Elements: The various heating elements and their physical characteristics are presented in Table 3. Kanthal is not recommended for temperatures above 9000C in vacuum as its electrical characteristics change in course of time due to chromium losses. Molybdenum provides the high power density required for vacuum furnaces. However above 14000C, the discharge voltage is 40V, and so the cross sectional area of the rod/strip has to be increased which inturn imposes constraints on the formability. Hence, for operations above 14000C molybdenum is not preferred. Besides, after a few heats, grain growth occurs which will be sufficient to cause the material to become very brittle, thereby seriously limiting its ability to withstand any thermal shock due to which frequent replacement is required. For temperatures between 1400-18000C tungsten/tantalum is preferred due to its capability to withstand thermal shock. However, fabrication difficulties and higher initial costs are significant points to be considered. With the availability of high quality graphite, graphite elements have been widely used upto 20000C.

The advantages of graphite heating elements are : 1) Low initial cost. 2) No change in resistance at high temperatures. 3) High hot strength. 4) Ability to withstand reasonable mechanical shocks and accidental high temperature air breaks. 5) Ease of fabrication of the hot zone due to its availability in the forms of thin sheets and strips

and cloth. However, the user has to ensure that the heating elements do not come in contact with the job, as carbon can diffuse into base metal interstitially. If carbon contamination is not tolerable even in ppm level, these elements are not recommended. Heating elements design: Three types of heating element arrangements are available for heating element location in the hot zone. 1. Locating the heating elements circularly along the length as well as on top & bottom sides - This

design is generally followed for circular hot zones. The uniformity of temperature on the job depends on the location of it with reference to hot zone size.

2. Locating the heating elements along the sides. Depending on the uniformity of temperature

required the heating elements are located on either on two sides, four sides or six sides. 3. Zone trimming - In this the heating elements are located on all six sides. However power coupling

is adjusted depending on the size, shape and uniformly of temperature required on the Jon. Generally this configuration is followed in brazing furnaces. The heating elements are generally used either in rod form or strip from.

IInnssuullaattiioonn//SShhiieellddss The efficiency of the furnace depends on the way in which the heat is conserved. This can be maximised by properly designing the furnace to reduce heat losses. The main mode of heat loss in any furnace is through radiation and hence designing radiation shields/insulation plays an important role in the overall design of the furnace. The essential requirement is the use of highly polished metallic sheets as shields. Depending on operating temperature and the shield materials, the number of shields needed may vary. The common materials widely used for shields are stainless steel, molybdenum and tantalum. Of these molybdenum and tantalum have very low emissivities (Table 3) in comparison to stainless steel and hence minimize heat losses. Thus, for high temperature applications above 8000C, four polished stainless steel sheets are used. For temperatures upto 16000C, six shields, consisting of a combination of SS + Mo, are used. Graphite felt can also be used as insulating material in place of shields where carbon contamination is not critical. The features in favour of graphite felt insulation are: a) The thermal efficiency of the furnace increases by 33% and thereby, power and water

requirements are less. b) The initial cost is low (66% of the cost of metallic shields) and c) Frequent replacement is avoided. The main disadvantage with graphite insulation is that it quickly absorbs moisture, from the atmosphere. Whenever the furnace is opened prolonged initial pumping is required. In the chamber, carbon monoxide may form and react with the job. By opening the furnace chamber at temperatures 120-1500C for loading and unloading the job, and always keeping the chamber under vacuum when furnace is not in operation can cut down the initial pumping time considerably.

CCeerraammiiccss High quality ceramics which can withstand thermal shocks are used for element sports/spacers below shields, ( to ensure the elements do not touch during heating). These ceramics are fabricated from 99% Al2O3 and the degassing rate of alumina is 7.5 x 10-15 Torr 1/s. Depending on the design of heating elements, the shape of the ceramic components is decided.

Electrical feed-throughs for resistance heating: These require both electrical insulation and vacuum tight seal. The type of feed through generally used for vacuum resistance furnace application upto 1 x 10-5 mm Torr are dismountable type with conventional 'O' ring seals. Since high current passes through the feed through, water cooling is essential to avoid excess heating. The copper feed-through assemblies are fabricated either as a single piece or in two parts brazed together.

Induction heating: Induction heating is characterised by excellent control over alloy composition because of the stirring action generated by eddy currents. Hence it is preferred for alloy melting and refining processes. Induction stirring homogenises the melt as well as brings the reactants to the melt vacuum interface so that the reactions can proceed rapidly. Induction is also used for applications like sintering, brazing, annealing etc, where temperature accuracy on the component need not be so accurate. For these applications a graphite scepter is used which in turn radiates the heat to the job.

With the development of medium frequency generators using thyristor techniques, it has become possible to design vacuum furnaces having induction coils located inside the cold walled furnace. The frequencies of these generators is in the range of 0.5 to 10 KC/s. The induction coils are fabricated out of conventional copper tubes which are water cooled using softened water. If needed, magnetic shunts shall be used to prevent induction leakage.

CChhaarraacctteerriissttiiccss ooff HHeeaattiinngg EElleemmeennttss

Heating Element Normal Operating Temp. C

Electrical Resistivity (Microohm m) at operating temp.

Kanthal A - 1 900 1.52 Molybdenum 1600 0.5

Tungsten 1800 0.59 Tantalum 1800 0.81 Graphite 2200 19

Comparison of hot walled and cold walled furnaces

SL. No Specification Hot-walled type Cold-walled type 1. Furnace shell material Thick nickel base superalloys Mild steel & stainless steel of

relatively low thickness 2. Volume of the furnace

chamber for a given hot zone size

About half of the volume required for a cold walled furnace

---------------

3. Pumping system capacity Low High 4. Power requirements to heating

elements Conventional power supply Low voltage, high amperge

supply 5. Heat losses High Low 6. Heating elements Conventional heating elements

such as SiC, Kanthal etc. Specialised heating elements such asMo, W, Ta etc.

7. Operating temperatures 10000C max As high as 24000C 8. Maintainance/Running cost High Low 9. Cost of furnace for a given hot

zone size and vacuum Cheaper Costlier

Today, microprocessor based temperature programmers/controllers, wherein it is possible to preset heating rates, holding time, cooling rates etc., are available.

Temperature Recorder/Over temperature protection:

Where multiple thermocouple are used for recording the temperature at different locations of the job a multipoint recorder or scanner is incorporated. In order to protect the hot zone materials/job from damage due to distortion or melting in a vacuum furnace, it is mandatory to incorporate an over temperature protector.

Safety interlocks:

Safety interlocks are essential to ensure that proper sequential operations are followed. This helps in increasing the life and preventing frequent breakdown of the furnace and its subsystems.

Important interlock combinations are:

1. Water flow and switch ON of the systems. 2. Diffusion pump ON and foreline vacuum after attaining proper. 3. Diffusion pump and thermal switch for water flow. 4. Heating element ON only under high and vacuum valve open. 5. Roughing pump OFF and DP should cut OFF.

Indicating lights corresponding to different on conditions display the status of the furnace operation. In addition, alarms and fault indication lights are also to be provided in.

Thermocouple feed - throughs: These (electrical feed through in the forms of flange) are also positioned on the chamber wall through an 'O' ring seal. Depending on the requirement, one or more thermocouple can be positioned in the feed through by using a suitable glass-to-metal seal. It is essential that all the joints on the thermocouple feed through are tested by helium mass spectrometer leak detector, at a standard leak rate of 1 x 10-8 std cm3 /s.

TThheerrmmooccoouupplleess If two wires of unlike metals or alloys are joined together firmly at one end, and this junction is heated or cooled, a small voltage (emf) will appear at the open ends. This emf is a function of the temperature difference and of the kinds of metals used. Thermocouples for use in vacuum devices are to be chosen on the basis of (a) useful temperature range, (b) sensitivity, and (c) speed of response. Metals and alloys with high vapor pressures should be avoided where elevated temperatures and bake-out are scheduled. The range of emfs produced by most conventional couples in use is up to about 50 μv/oC. The speed of response is largely determined by the size of wire used for the two metals, and is greater with finer wires than with larger sizes. But for use in oxidizing or corrosive atmospheres at high temperatures, as in furnaces and ovens where speed of response is not an important factor, the thermocouple should be made of heavy-gauge, corrosion-resistant material. Platinum and platinum-rhodium alloy couples are expensive but can be used at temperature upto 15000 C in air. Exposure to high temperatures (up to 16000 C) may change the calibration at lower temperature and should be checked after each such exposure. Chromel-Alumel thermocouples, are widely used in industrial and laboratory equipment because of their uniformity, corrosion resistance, sensitivity, and relative low cost (compared to platinum). Chromel-Alumel thermocouples can be used between -2000 and +12600C in air or in non-reducing environments. The alloys cannot be used at high temperatures in hydrogen, sulfur, or carbon monoxide atmospheres.

Copper-Constantan junctions can be used in the range -2600 to +3500C, the upper value being set by the rapid oxidation of copper at higher temperatures. The copper preferred for making these couples is OFHC which is very homogeneous. Constatan is a somewhat variable material, having the composition Cu 50-60%, Ni 50-53%.

Iron-Constantan thermocouples are very widely used in oxidizing atmosphere up to 7600C and in reducing atmospheres up to about 9800C.

Tungsten-tungsten 74%, rhenium 26%. This couple can be used up to 28000C in vacuum, hydrogen, and inert gases such as nitrogen, argon, and helium, but it cannot be used in oxidizing or hydrocarbon atmospheres at high temperatures.

Thermocouples can be made in several ways. For use in electron (vacuum) devices, an arc-, spot-, or butt-welding technique can be used, as illustrated in Fig. 145, (a), (b) and (c). Platinum thermocouples can be arc-welded without using any flux, or plain borax can be used in torch-welding to prevent the effects of a reducing flame which may embrittle the wires. In spot-welding, if copper electrodes are used, any copper

contamination should be removed by appropriate chemical treatment after the weld is made. For furnace and oven thermocouples, the wires may be twisted together for a short distance at the end, and the junction made by welding or brazing, as shown in Fig. 145(d) and (e). In the case of the brazed (or soldered) junction (e), the presence of braze material between the elements does not introduces an error so long as the material is homogeneous and the temperature along it is uniform. Water flow switches: These are incorporated in every outlet water line of the furnace assembly which requires water cooling and interlocked with respective subsystem.

Vacuum in Chemical Process Industries In chemical industries, there is a huge demand for separation and enrichment processes at low temperatures. This is possible only by vacuum distillation/drying/degassing techniques. Vacuum distillation is routinely used in petroleum refineries to separate crude oil into fuels and petrochemicals. Vacuum processing is used in chemical industries to refine a wide range of products and recover raw materials for reuse. The purification of food products such as vegetable oil also requires the use of vacuum distillation. Molecular distillation is used to purify and separate high-boiling materials such as vitamins, oils, waxes, fatty acids, glycerides and plasticizers. Vacuum concentration is extremely useful in pharmaceutical and biomedical industries involving heat sensitive chemicals. Vacuum Distillation: In distillation, a mixture of two or more components is separated by using successive evaporations and condensations. Products with desired purity can be obtained by this process. In vacuum distillation, the column condenser is not vented to the atmosphere but to a vacuum pump for maintaining sub-atmospheric pressure. Further, in some processes, in order to prevent undesirable reactions/thermal degradation/ polymerization of process materials, it is necessary to avoid high distillation temperatures. This is achieved by processing under vacuum. Similarly most of the food products and pharmaceutical compounds are processed under vacuum. In other applications, vacuum is used so that less expensive, low-pressure steam can be employed as the heating medium. Low-temperature vacuum operation increases the mixture separability because the ratio of vapor pressures for two compounds will increase with decreasing temperature. If the two compounds form an ideal solution, reducing the temperature will increase the relative volatility. The equilibrium diagram shown in Fig. 1(a) illustrates the effect of vacuum on the phase equilibria of a typical binary mixture. As the volatility, K =

1

1

xy (ratio of mole fraction in vapour phase to liquid phase) more the

(b). Elimination of an azeotrope by vacuum

operation

X1

Y1

VacuumAtmospheric

Pressure

Fig.1 (a). Typical Equilibrium Diagram

bending of the curve away from the diagonal, more the possibility of the separation. Processing under vacuum reduces the operating and/or capital costs by lowering the reflux ratios or the number of equilibrium stages.

Molecular Distillation : Some heat sensitive commercial products have such a high molecular weight that conventional rough-vacuum distillation techniques cannot be used as a means of separation. Because of their low volatility, these products begin to decompose before they can be heated to the temperatures required for rough vacuum distillation. Some compounds may exert a vapor pressure of only a fraction of a torr at temperatures as high as 3000C. These products must be distilled at comparatively high vacuum to maintain moderate temperatures, and exposure to heating must be minimized. The device that meets this need is the molecular still. Molecular stills are used on an industrial scale to purify and separate high-boiling materials such as vitamins, oils, waxes,

On the other hand, in some highly nonideal binary mixtures the equilibrium curve at atmospheric pressure crosses the diagonal (fig. 1(b)) indicating an azeotrope. At this point of intersection, the equilibrium compositions of both liquid and vapor are identical. As a result this further enrichment of the vapor phase beyond this azeotrope composition is impossible by conventional distillation techniques. If the distillation operation is carried out under vacuum, further enrichment is possible as no azeotrope is formed. Vacuum is also used to distill toxic materials which will allow the removal of toxic gases formed during the process operations.

fatty acids, glycerides, and plasticizers. Industrial molecular distillation is generally regarded as a major vacuum processing operation. Molecular distillation requires operating pressures between 0.1 and 10 microns (10-4 and 10-2 torr) and a short, unobstructed path between the evaporator and condenser surfaces. The total separation between these two surfaces is of the order of the mean free path of the distilling molecules at these pressures. DDeeggaassssiinngg : In degassing operations, the volatile components to be removed are dissolved gases in the liquid solvents. Small quantities of volatiles are removed overhead, and the desired product is removed underneath. The product need not be heated to its bubble point. Dissolved gases are removed by changing their solubility in the product. This is primarily accomplished in vacuum degassers by reducing the pressure. Vacuum degassing is widely used to prevent corrosion in equipment sensitive to the presence of gases like oxygen and carbon dioxide. The feed water to many industrial boilers undergoes vacuum deaeration to reduce the levels of oxygen and carbon dioxide and prevent corrosion with in the boiler. Dissolved oxygen is usually the controlling factor in the corrosion of iron in water. Vacuum degassers are also used in conjunction with molecular stills and other operations in order to remove dissolved gasses under a rough vacuum before subjecting the feed to a higher vacuum. Because the pumping of dissolved gasses is less expensive at higher pressures, the feed to many high vacuum operations is degassed in several stages, each stage at a lower pressure than the one before.

Filtration : Vacuum filters are used extensively in the process industries in liquid-solid separation operations involving chemicals, pharmaceuticals, food, beverages building materials, and other products. A solid can be separated from a liquid by pulling a vacuum across a filter medium that passes the liquid but retains the solid (maximum pressure differential of 1 atm). Capital cost for a vacuum filter is considerably less than that of a centrifuge.

VVaaccuuuumm DDrryyiinngg ooff TTrraannssffoorrmmeerrss

The manufacturing of good transformer involves drying of the solid insulation (paper, wood, epoxy etc) of transformer by heating (oil heat exchanger) the assembly in a vacuum vessel (auto clave) and then applying the vacuum to remove the moisture. After the drying is completed, the transformer is allowed to cool down, and oil is filled in the transformer under vacuum. Before filling the oil it is filtered, and dehydrated by a separate vacuum filteration and degassing plant.

For drying of power transformer it is essential to have a combination of mechanical booster and Rotary oil sealed Vacuum Pump having enough water vapour tolerance. As such the vacuum drying process is slow. 1 cc of water = 1150 litres of vapours @ 1 torr. = 11,500 litres of vapours @ 0.1 torr. = 1,15,000 litres of vapours @ 0.01 torr. As the vacuum improves, the volume of vapour also increases and to remove small amount of water, the vacuum pump has to pump out lot of vapours and hence booster stage is employed.

Hot Thermic Oil

Sensor

Oil

Radiator CoilGlass wool insulation

Condensate Collecting Tanks

Condenser

Insulation Valve

Collecting Tank

Exhaust

Condenser Rotary Pump

Roots Pump

Chilled water

Vacuum Gauge

Air Release Valve

Autoclavee

Transformer

Vacuum Drying System

HHeeaatt TTrraannssffeerr iinn vvaaccuuuumm ssyysstteemmss Above approximately 10 Torr, the heat transfer through a gas inside a relatively small chamber is dominated by convection (bulk motion of gas). Between 1 to 10-3 torr the heat transfer is through gas conduction which is a linear function of pressure (hence applied in thermal conductivity gauges).

In high vacuum conditions (Below 10-4 Torr) the heat transfer is mainly due to radiation and to a some extent by solid conduction through support structure. It is to be mentioned here, if two solid pieces are in just apparent contact in a vacuum chamber the heat transfer is very poor, as the real area of contact between two rigid bodies is usually only 0.1% of the apparent area. Heat is transferred only through a few touching high points between the plates. To increase the heat flow, one (or both) of the plates should be soft and high clamping forces must be used. Even then, the heat transfer by conduction will be limited to the small areas of contact created near the clamping bolts.

In view of the above facts, we can calculate the temperature rise of the job in a vacuum furnace due to resistive heating from surrounding heating elements, based on radiation formula.

Q12 = δ12 A1 (T1

4 - T24)

where δ12 = Cb/1/E1 + A1/A2 (1/E2 - 1)

where Cb = Radiation coefficient of the black body i.e. 5.7 x 10-11 KW/(m2 T4) E1 ,E2 = Emissivities of 1st and 2nd shields A1,A2 = Areas of the 1st and 2nd shields T1, T2 = Temperatures of the 1st and 2nd shields.

As the radiation between two surfaces is proportional to difference in fourth power of temperature, the heat flux in the initial period to the job from a heating element (reached fast to the equilibrium, say 8000C, due to resistive heating) is very high. Hence if we want more or less linear slow increase of job temperature, it is very important to raise the temperature of the heating elements in steps and allow the job to reach that set limit. For heating in several steps using suitable temperature controllers, we can apply radiation formula piece wise taking the necessary temperature limits. Once heat flux is known, the temperature raise can be calculated using the physical properties of the job namely specific heat (more or less a constant value in high temperature range) and mass. For calculating heat flux due to support solid structures, one can use Fourier equation by taking the thermal conductivity value at the average temperature. However, in high temperature region for most of the metals thermal conductivity does not change much and can be taken as constant. Q = - A / L K(T) dT

VVaaccuuuumm iinnssuullaattiioonn Vacuum insulation alone is used extensively for small laboratory-size dewars. Heat is transferred across the annular space of a vacuum-insulated vessel by radiation from the hot outer jacket to the cold inner vessel and by gaseous conduction through the residual gas within the annular space.

The radiant heat-transfer rate between two surfaces is given by the modified Stefan-Boltzmann equation,

⎟⎠⎞

⎜⎝⎛ −−= 4

1T42T12σA21FeFQ ----------------------------- (1)

Where Fe = emissivity factor, F1-2 = configuration factor, σ = Stefan-Boltzmann constant ,

σ = 56.69 nW/m2 – K4 = 0.1714 × 10-8 Btu/hr-ft2 –0R4 , A1 = area of surface 1, T = absolute temperature.

For cryogenic-fluid storage vessels, in which the inner vessel is completely enclosed by the outer vessel, F1-2 = 1, in addition, the emissivity factor for diffuse radiation for concentric spheres or cylinders is given by

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛−+= 1

2e1

2A1A

1e1

F1 ----------------------------- (2)

where e is the emissivity of the surface and A is the surface area.

Radiant heat transfer can be reduced by interposing floating (thermally isolated) radiation shields of highly reflective material between the hot and cold surface. For N shields between the hot and cold surfaces, the emissivity factor for a shield emissivity es is (parallel flat plates)

( )⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛−++−−+−+= 1

se1

2e11

se21N1

se1

1e1

eF1 -------------------------------(3)

As an example of the effectiveness of radiation shields, suppose we consider a pair of surfaces having an emissivity e1 = e2 = 0.80 for parallel flat plates, using eqn. (4), we obtain 1/Fe (no shields) = 1/0.8 –1 or Fe (noshields) = 0.6667 If 10 shields having emissivities of es = 0.05 are placed between the surfaces, the emissivity factor from eqn. (15) is 1/Fe (10 shields) = (2)(1/0.8 + 1/0.05 – 1) + (10 – 1)(2/0.05 –1) or Fe (10 shields) = 0.00255 By using 10 radiation shields, the radiation heat-transfer rate is reduced by a factor of (0.00255/0.6667) = 0.00383 ⇒ 0.4% of original valve (250 times improved) In addition to the heat transferred by radiation, energy is transmitted by gaseous conduction through the residual gas in the vacuum space. If the pressure of the gas is low enough that the mean free path of the gas molecules is greater than the distance between the two surfaces, the type of conduction differs from the usual continuum-type conduction a ambient pressure. For ordinary conduction with constant thermal conductivity, there is a linear temperature gradient within the medium transmitting heat. On the other hand, for free molecular conduction, the gas molecules rarely strike each other, thus an individual gas molecule travels across the gas space without transferring energy to other gas molecules.

Fig. Energy transport and molecule “temperature” for free-

molecular conduction.

Consider two parallel plates maintained at T1 and T2, respectively, as shown in Figure. A gas molecule collides with the cold surface at T1 and transfers some energy to the surface. Because the molecule may not remain on the surface long enough to establish thermal equilibrium, in general, it will leave the surface with a kinetic energy corresponding to a somewhat higher temperature, say T1’ This molecule at T1’ travels across the gas space and collides with the warm surface at T2. Again, the molecule generally does not remain on the surface long enough to establish thermal equilibrium; therefore, it leaves the warm surface with a kinetic energy corresponding to a temperature some what less than T2, say T2’. The degree of approach of the molecules to the thermal equilibrium upon collision is expressed by the accommodation coefficient, defined by transferenergypossiblemaxmum

trnsferenergyactuala =

The accommodation coefficient depends upon the specific gas surface combination to the surface temperature. Some typical values of accommodation coefficient are given in Table. The accommodation coefficients for the two surfaces shown in above figure are

Cold surface, ⎟⎠⎞

⎜⎝⎛

⎟⎠⎞

⎜⎝⎛ −−= 1T1

2T/'1T1

2T1a

Warm surface, ⎟

⎠⎞

⎜⎝⎛

⎟⎠⎞

⎜⎝⎛ −−= '

1T2T/'1T1

2T2a

Solving for the temperature difference between the warm and cold surfaces we obtain

⎟⎟⎟⎟

⎠

⎞

⎜⎜⎜⎜

⎝

⎛

⎟⎠⎞

⎜⎝⎛

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛ −=−−+=−

aF

'1T1

2T'1T1

2T12a1

1a1

1T2T ---------------------(4)

Where Fa , is the accommodation factor, which has exactly the same form as the emissivity factor. For concentric spheres and cylinders, the accommodation coefficient factor is given by Table : 4 Approximate accommodation coefficients

Temperature Gas

K 0R Helium Hydrogen Neon Air 300 78 20

540 140 36

0.29 0.42 0.59

0.29 0.53 0.97

0.66 0.83 1.00

0.8-0.9 1.00 1.00

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛−+= 1

2a1

2A1A

1a1

aF1 ------------------------(5)

Where subscript 1 denoted the enclosed surface and subscript 2 denotes the enclosure.

A comparison of the advantages and disadvantages of the insulations used in cryogenic systems is given in the following summary.

SL.No. Advantages Disadvantages 1. 2. 3. 4. 5. 6.

Expanded foams low cost. No need for rigid vacuum jacket. Good mechanical strength. Gas-filled powder and fiberous materials low cost. Easily applied to irregular shapes. Not flammable. Vacuum alone Complicated shapes may be easily insulated. Small cool-down loss. Low heat flux for small thickness between inner and outer vessel. Evacuated powders and fibrous materials. Vacuum level less stringent than for multilayer insulations. Complicated shapes may be easily insulated. Relatively easy to evacuate. Opacified powders Better performance than straight evacuated powders. Complicated shapes may be easily insulated. Vacuum requirement is not as stringent as for multilayer insulations and vacuum alone. Multilayer Insulations Best performance of all insulations. Low weight. Lower cool-down loss compared with powders. Better stability than powders.

High thermal contraction. Conductivity may change with time. Vapor barrier is required. Powder can pack and conductivity is increased. A permanent high vacuum is required. Low-emissivity boundary surfaces needed. May pack under vibratory loads and thermal cycling. Vacuum filters are required. Must be protected when exposed to moist air (retains moisture). Higher cost than evacuated powders. Explosion hazards with aluminum in an oxygen atmosphere. Problems of setting of metallic flakes. High cost per unit volume. Difficult to apply to complicated shapes. Problems with lateral conduction. More stringent vacuum requirements than powders.

MMuullttiillaayyeerr IInnssuullaattiioonnss Multilayer insulations consist of alternating layers of a highly reflecting material, such as aluminum foil, copper foil, or aluminized Mylar, and a low conductivity spacer, such as fiberglass mat or paper, glass fabric, or nylon net. The reflecting layers may also be separated by crinkling or embossing the sheets so that they touch only at a few discrete points, and a spacer is not required. Multilayer insulation must be evacuated to pressures below 10 mPa. (7.5 X 10-5 torr) to be effective. The dependence to the apparent thermal conductivity on residual gas pressure for a typical multilayer insulation shown in figure.

Fig. Variation of mean apparent thermal conductivity with residual gas pressure for a typical multilayer insulation. The insulation layer-density is 24 layers/cm (60 layers/in.), and the boundary temperature are 300K (5400R) and 90.5K (1630R).

VVaaccuuuumm--iinnssuullaatteedd lliinneess

Inner line

Spacer

Outer line

Roller

Vacuum-insulated lines consist of an inner line, in which the liquid flows, concentric to an outer vacuum jacket. The annular space may contain a multilayer insulation or vacuum alone. The vacuum-insulated line may be used with any cryogenic fluid from liquid oxygen to liquid helium to attain low-loss transfer. The ever-present thermal-contraction problem can be solved in cryogenic transfer-line design through the use of expansion bellows and U- bends. It is good practice to locate the expansion bellows only in the outer line and to achieve flexibility of the inner line through the use of U-bends, especially for large-diameter lines with high working pressures.

Long section of vacuum-jacketed transfer lines require spacers to support the inner line with in the outer line. In addition to having a low thermal conductivity, the spacer should have a high mechanical strength, low specific heat (to make cool-down losses small), and a low outstanding rate (since the spacer is located in the vacuum space). The spacer configuration should be such that it does not seriously block the annular space, so that the annular space can be evacuated from a single connection.

Fig. Typical spacers used to separate the inner and outer lines in a vacuum-jacketed cryogenic-fluid transfer line.