Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

33
Beam-based vacuum observations and their Beam-based vacuum observations and their consequences consequences Sergei Nagaitsev August 18, 2003

Transcript of Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Page 1: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Beam-based vacuum observations and their Beam-based vacuum observations and their consequencesconsequences

Sergei NagaitsevAugust 18, 2003

Page 2: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 2

Design goalsDesign goals

Stochastic cooling only Emittance growth rate (n, 95%): 2 μm/hr Lifetime of 10-μm emittance pbar beam with

cooling: ≥ 200 hrsStochastic+electron cooling Emittance growth rate (n, 95%): 4 μm/hr Lifetime of 10-μm emittance pbar beam with

cooling: ≥ 200 hrs

The lifetime cannot be greater than:

%95,%95,

30)(

6nn

A

AA

AAΛ

yx

yx

Page 3: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 3

Design goalsDesign goals

Stochastic cooling only Emittance growth rate (n, 95%): 2 μm/hr Lifetime of 10-μm emittance pbar beam with cooling: ≥ 200 hrsStochastic+electron cooling Emittance growth rate (n, 95%): 4 μm/hr Lifetime of 10-μm emittance pbar beam with cooling: ≥ 200 hrs

Relative Coulomb scattering loss rate for the Recycler beam, assuming a 40-μm acceptance in both planes.

The design emittance is 10 μm (n, 95%)

0 10 20 30 400.01

0.1

1

Normilized 95% emittance [mm mrad]

Rel

ativ

e be

am li

fetim

e

Page 4: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 4

Today’s menuToday’s menu

I will present the beam-based measurements and demonstrate that the results of these measurements are consistent with Coulomb scattering off residual gas.

Page 5: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 5

Today’s menuToday’s menu

I will present the beam-based measurements and demonstrate that the results of these measurements are consistent with Coulomb scattering off residual gas.

Assuming that all scattering is elastic, I will derive the partials pressures. These are an order of magnitude higher than what we obtain from our vacuum model.

Page 6: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 6

Today’s menuToday’s menu

I will present the beam-based measurements and demonstrate that the results of these measurements are consistent with Coulomb scattering off residual gas.

Assuming that all scattering is elastic, I will derive the partials pressures. These are an order of magnitude higher than what we obtain from our vacuum model.

I will critique my own model and consider uncertainties.

Page 7: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 7

Today’s menuToday’s menu

I will present the beam-based measurements and demonstrate that the results of these measurements are consistent with Coulomb scattering off residual gas.

Assuming that all scattering is elastic, I will derive the partials pressures. These are an order of magnitude higher than what we obtain from our vacuum model.

I will critique my own model and consider uncertainties.

Finally, I will demonstrate that the goal of 4 μm/hr could be achieved with a successful bakeout. The goal of 2 μm/hr is likely to be unattainable with the present vacuum system.

Page 8: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 8

Observations (coasting beam) Aug 2, 2003Observations (coasting beam) Aug 2, 2003

One hour of observations with 9E10 protons (no MI ramps)

Initial emittance: 10 μm (95%, Schottky) 7 μm (100%, scraper) Final emttance: 20 μm (95%, Schottky) 17 μm (95%, scraper) Growth rate: 10 ± 1 μm/hr Zero-current, pencil beam lifetime: 90 hours Acceptances: Ax = 60 μm, Ax = 40 μm The emittance growth rate and the lifetime are

selfconsistent!

Page 9: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 9

Emittance growth rateEmittance growth rate

Our measurements demonstrate that: The transverse emittance growth is linear with time The rate is insensitive to beam intensity and

bunching structure The rate does not depend on beam emittance The rate does not depend on tunes (within limits)

Page 10: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 10

Emittance growth rateEmittance growth rate

Our measurements demonstrate that: The transverse emittance growth is linear with time The rate is insensitive to beam intensity and

bunching structure The rate does not depend on beam emittance The rate does not depend on tunes (within limits)

This all points to Coulomb scattering I will be using the following growth rate

expression for my analysis:

i i

ieiniiave

p

Z

ZZZn

crn

)()(12 2

2

%95,

Page 11: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 11

Observations (coasting beam) Aug 2, 2003Observations (coasting beam) Aug 2, 2003

1 hour, 9E10 p’s, long. Schottky at 1.75 GHz (no MI ramps) The average revolution frequency has decreased by 32 mHz

This corresponds to a 0.37-MeV energy shift

The rms momentum spread has increased (1.0 to 1.3 MeV/c) The low energy tail has developed

-12 MeV/c 12 MeV/c -12 MeV/c 12 MeV/c

Page 12: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 12

Observations (coasting beam) Aug 13, 2003Observations (coasting beam) Aug 13, 2003

30 min, 1E10 p’s, long. Schottky at 79 GHz (no MI ramps) Energy loss: 0.42 MeV/hr The rms momentum spread has increased. The low energy tail has developed.

dB

Frequency

Page 13: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 13

The low-energy tailThe low-energy tail

We’ve proved that the low-energy tail comes from ionization losses.

1.9E11 protons before scrape 0.9E11 protons after scrape

-12 MeV/c 12 MeV/c

Page 14: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 14

The low-energy tailThe low-energy tail

The scrape was done with a horizontal scraper at a location with a horizontal beta-function of 52 m, zero dispersion (~20 cm) and with equal tunes.

The scraper was stopped 6.2 mm away from the beam center, which corresponds to a 7-μm acceptance, and then withdrawn.

How can one scrape the low energy tail at zero-dispersion location?

Page 15: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 15

The low-energy tailThe low-energy tail

The scrape was done with a horizontal scraper at a location with a horizontal beta-function of 52 m, zero dispersion (~20 cm) and with equal tunes.

The scraper was stopped 6.2 mm away from the beam center, which corresponds to a 7-μm acceptance, and then withdrawn.

How can one scrape the low energy tail at zero-dispersion location?

The answer is: proton-electron collisions.

Page 16: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 16

Proton collision with a stationary electronProton collision with a stationary electron

Max. energy transfer Tmax= 91 MeV

0 20 40 60 80 1000

1

2

3

4

5

Tra

nsve

rse

mom

entu

m, M

eV/c

pperp ( )

t ( )

Energy transfer per collision, MeV

Outside of ±0.3% momentum acceptance

Page 17: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 17

Proton collision with a stationary electronProton collision with a stationary electron

avep

n cM

pJ

22

2

2

1

0 5 10 15 200

10

20

30

40

Bet

atro

n ac

tion,

mm

-mra

d

J ( )

t ( )

Energy transfer per collision, MeV

Scraped to 7 mm-mrad

Page 18: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 18

Observations (coasting beam) Aug 2, 2003Observations (coasting beam) Aug 2, 2003

1 hour, 9E10 p’s, long. Schottky at 1.75 GHz (no MI ramps) The average revolution frequency has decreased by 32 mHz

This corresponds to a 0.37-MeV energy shift

The rms momentum spread has increased (1.0 to 1.3 MeV/c) The low energy tail has developed

-12 MeV/c 12 MeV/c -12 MeV/c 12 MeV/c

Page 19: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 19

The rms momentum spread increaseThe rms momentum spread increase

IBS

Lc - is the Coulomb logarithm, C – is the ring circumference, σ – is the rms beam size and θ – is the rms angular spread.

21

2||

233

2

2

2||2

||

C

cNLr

p

p

dt

d

dt

d Cp

Page 20: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 20

IBS (coasting beam)IBS (coasting beam)

The rms momentum growth rate (MeV/c per hour) as a function of the rms momentum spread (MeV/c) for various transverse (n,95%) emittances.

N = 9x1010 protons Measured:

Initial 1.0 MeV/c Final 1.3 MeV/c

Need to add rms spreads in quadrature!

The ibs can explain a portion (0.1-0.2 MeV/c) of the measured momentum spread.

1 105

1 106

1 107

1 105

1 106

1 p( )

2 p( )

3 p( )

4 p( )

p Mp

dWithout any cooling the momentum spread diffusion rate is (for a 10 pi emittance):

εn =5 μm

10 μm

15 μm

20 μm

Page 21: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 21

0 2 105

4 105

6 105

8 105

1 106

0

2 106

4 106

6 106

8 106

1 105

1.2 105

g p( )

p

Energy loss, MeV

The rms momentum spread increaseThe rms momentum spread increase

IBS

Energy-loss straggling

Landau probabilitydensity

Mean energy loss: 0.42 MeVMost probable loss: 0.3 MeV

Page 22: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 22

Momentum distributionMomentum distribution

A Gaussian(σ=1 MeV/c) distr. is folded with the Landau probability density.

The result is that by fitting the energy loss alone, I am able to reproduce both the low-energy tail and the rms spread increase

Energy loss due to resistive impedance is negligible

04.8 106

9.6 106

1.44 107

1.92 107

2.4 107

60

49.6

39.2

28.8

18.4

8

FD jj

GD jj

jj

Momentum spread: +/- 12 MeV/c

Page 23: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 23

Summary of beam-based measurementsSummary of beam-based measurements

Emittance growth rate:

Beam average energy loss: 0.40±0.04 MeV/hrThis corresponds to a mean energy loss of 0.42±0.04

MeV/hr

m/hr 1 10)(

)(12 22

%95,

i i

ieiniiave

p

Z

ZZZn

crn

i ii

i

i

I

Tx

A

ZMeV 0.040.42ln][g/cm

307.0MeV 2max2

2

Page 24: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 24

Partial pressures modelPartial pressures model

Assume that the residual gas consists only of Z=1 and Z=8 atoms (H2 and H2O)

Solving equations for 10±1 μm/hr and 0.42±0.04 MeV results in: pH = 3.3±1.7x10-9 Torr

pW = 1.0±0.2x10-9 Torr

Page 25: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 25

Partial pressures modelPartial pressures model

Assume that the residual gas consists only of Z=1 and Z=8 atoms (H2 and H2O)

Solving equations for 10±1 μm/hr and 0.42±0.04 MeV results in: pH = 3.3±1.7x10-9 Torr

pW = 1.0±0.2x10-9 Torr

If I assume that one of the gases is hydrogen with a know concentration nH, and then try looking for another gas with a new Z (A = 2Z), nZ ≤ nH, I am unable to find any solution unless pH > 1.5x10-9 Torr and Z ~ 5.

Page 26: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 26

Partial pressures modelPartial pressures model

Assume that the residual gas consists only of Z=1 and Z=8 atoms (H2 and H2O)

Solving equations for 10±1 μm/hr and 0.42±0.04 MeV results in: pH = 3.3±1.7x10-9 Torr

pW = 1.0±0.2x10-9 Torr

“Water” alone contributes 8.3 μm/hr to the emittance growth

Page 27: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 27

Partial pressures modelPartial pressures model

Assume that the residual gas consists only of Z=1 and Z=8 atoms (H2 and H2O)

Solving equations for 10±1 μm/hr and 0.42±0.04 MeV results in: pH = 3.3±1.7x10-9 Torr

pW = 1.0±0.2x10-9 Torr

“Water” alone contributes 8.3 μm/hr to the emittance growth

Before Jan 2003 shutdown the measured emittance growth rate was 5 μm/hr. There were no beam energy loss measurements.

Present measurements are consistent with the water content doubled after the shutdown.

Page 28: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 28

What could have increased the water content?What could have increased the water content?

Work done during Jan 2003 shutdown: Installed new ion pumps, capable of pumping

Ar at a higher rate. Installed some diagnostics (RGAs, ion gauges).

Out of 27 vacuum sectors, 20 were vented with “dry” nitrogen and then only 5 sectors baked to 120C. Of these five, three have been vented since then without a bake.

Many new ion pumps were not baked in situ.

Unrelated to the Jan 2003 shutdown: BPMs and bellows (total equiv. length of about 500m)

were never baked at all.

Page 29: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 29

So, what’s the problem with this model??So, what’s the problem with this model??

There are two problems with my pressure model.

1. Terry’s vacuum model predicts an order of magnitude low pressures.Recycler Pressure Profile Using Average RGA Data from Pump Locations

1.0E-13

1.0E-12

1.0E-11

1.0E-10

1.0E-09

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

Distance (cm)

Pres

sure

(To

rr)

Total

IG Readings

Hydrogen

Water

Carbon Dioxide

Carbon Monoxide

Methane/Ethane

Argon

Unknown

Page 30: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 30

So, what’s the problem with this model??So, what’s the problem with this model??

1. Terry’s vacuum model predicts an order of magnitude low pressures.None of the ion gauges show pressures above 5x10-10

TorrRecycler IG Readings

1.E-11

1.E-10

1.E-09

1.E-08

Location

Pres

sure

(T

orr)

1/10/2003 7/11/2003

Page 31: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 31

So, what’s the problem with this model??So, what’s the problem with this model??

1. Terry’s vacuum model predicts an order of magnitude low pressures.

There is only one ion pump that show pressures above 1x10-

9 Torr.Recycler 30 - 40 House Ion Pump Pressure Profile

1.E-10

1.E-09

1.E-08

1.E-07

R:I

P220

R:I

P221

R:I

P222

R:I

P224

R:I

P225

R:I

P226

R:I

P227

R:I

P228

R:I

P229

R:I

P230

R:I

P231

R:I

P232

R:I

P301

A

R:I

P301

B

R:I

P302

A

R:I

P302

B

R:I

P303

A

R:I

P303

B

R:I

P304

A

R:I

P304

B

R:I

P305

A

R:I

P305

B

R:I

P306

A

R:I

P306

B

R:I

P306

C

R:I

P307

R:I

P308

R:I

P309

R:I

P310

R:I

P311

R:I

P312

R:I

P313

R:I

P314

R:I

P315

R:I

P316

R:I

P317

R:I

P318

R:I

P320

R:I

P321

R:I

P323

R:I

P324

R:I

P325

R:I

P326

R:I

P327

A

R:I

P327

B

R:I

P328

R:I

P329

R:I

P330

R:I

P331

R:I

P332

R:I

P333

R:I

P334

R:I

P335

R:I

P336

R:I

P337

R:I

P338

R:I

P339

R:I

P340

R:I

P341

R:I

P400

R:I

P401

R:I

P402

A

R:I

P402

B

R:I

P402

C

R:I

P403

R:I

P404

R:I

P405

1

R:I

P405

2

R:I

P406

R:I

P407

R:I

P408

R:I

P409

R:I

P410

R:I

P411

R:I

P412

R:I

P413

R:I

P414

R:I

P415

R:I

P416

R:I

P417

R:I

P418

R:I

P419

R:I

P420

R:I

P421

R:I

P422

R:I

P423

R:I

P424

R:I

P425

Location

Pre

ssur

e (T

orr)

1/10/2003 7/11/2003

Page 32: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 32

So, what’s the problem with this model??So, what’s the problem with this model??

2. We know that the total gas capacity of our TSPs is 0.2-0.4 Torr-L. Assuming we understand the TSP’s pumping speed at the beam pipe, pressures of 3x10-9 Torr would saturate the TSPs in 10-20 days, yet the TSPs last on the average 100 days.

Page 33: Beam-based vacuum observations and their consequences Sergei Nagaitsev August 18, 2003.

Recycler vacuum shutdown work review - Nagaitsev 33

ConclusionsConclusions

The evidence for the residual gas scattering from beam-based measurements is overwhelming.

I can not reconcile the beam-based measurements with the instrument measurements.

All beam-based measurements point to the fact that amount of heavy molecules (most likely water) in the system has doubled after the Jan 2003 shutdown.

The model can not answer how much CH4, CO or CO2 is in the system. If I assume that all heavy molecules are water, eliminating them completely reduces the emittance growth rate to 2 μm/hr. It is likely that we will never reach this value with a present system. I estimate that reaching a 4-μm/hr rate is possible with a successful bakeout.