Currentstatusof’ beam’commissioning’andoperaon’ of… · 2017. 1. 17. · 2 /V 1 (%) φ 2 (deg) Δp/p (%) 1 5 5 5 5 2 100 5 5 5 3 5 80 100 0.0 4 5 80 100 0.1 5 5 80 100 0.2

Current status of beam commissioning and opera3on

of the J-‐PARC 3GeV RCS

August 21, 2013 at RAL, UK

Hideaki Hotchi (J-‐PARC, JAEA)

JFY 2009 JFY 2008 JFY 2006 / 2007

Neutrino Beam Line to Kamioka (NU)

Materials & Life Science Facility (MLF)

3 GeV Rapid Cycling Synchrotron (RCS)

Hadron Experimental Hall (HD)

400 MeV H-‐ Linac [181 MeV at present]

50 GeV Main Ring Synchrotron (MR) [30 GeV at present]

J-‐PARC (JAEA & KEK)

1 MW

0.75 MW

Contents u Design parameters and layout of the J-‐PARC 3-‐GeV RCS

u Current status of the RCS beam opera?on & residual radia?on level

u High intensity beam trial of up to 540 kW -‐ Beam loss reduc?on by injec?on pain?ng -‐ Intensity dependence of beam loss and its beam loss mechanism -‐ Comparison between experiment and numerical simula?on

u Beam commissioning plan for the next two years

u Summary

Design parameters and layout of the J-‐PARC RCS Circumference 348.333 m

Super-‐periodicity 3

Injec?on Charge-‐exchange, Mul?-‐turn

Injec?on period 0.5 ms

Injec?on energy 181 MeV ⇒ 400 MeV

Extrac?on energy 3 GeV

Repe??on rate 25 Hz

Harmonic number 2

Number of bunches 2

Par?cles per pulse 2.5e13 -‐ 5e13 ⇒ 8.3e13

Output beam power 300-‐600 kW ⇒ 1 MW

Transi?on gamma 9.14 GeV

Number of dipoles 24

quadrupoles 60 (7 families)

sextupoles 18 (3 families)

steerings 52

RF cavi?es 12 (11 at present)

H-

Linac upgrade u  Output energy upgrade by adding an ACS linac sec?on in 2013 summer-‐autumn maintenance period 181 MeV ⇒ 400 MeV u  Intensity upgrade by replacing IS and RFQ in 2014 summer maintenance period Peak current 30 mA ⇒ 50 mA

We will aim at the design output beam power of 1 MW aber upgrading the linac.

Pulse dipole magnet to switch the beam des?na?on

History of the RCS output beam power u Beam commissioning of the linac ; November 2006~ u Beam commissioning of the RCS ; October 2007~ u Startup of the MLF user opera?on ; December 2008~

4 kW 20 kW

120 kW

300-‐kW demo.

210 kW 210 kW

300 kW

540-‐kW demo. Summer shutdown

Summer shutdown

Summer shutdown

Recovery from damages caused by East Japan earthquake

Outpu

t pow

er to

MLF (kW)

Einj=181 MeV

u RCS is now stably providing 300 kW beam for users.

u The beam power ramp-‐up of RCS has steadily proceeded following; -‐ Progression in beam tuning and hardware improvements, -‐ Careful monitoring of the trend of residual ac?va?on levels.

Current residual radia3on level

Coll. Arc1 Arc2 Arc3 Inj. Radia?on level (mSv/h) at 30 cm from the chamber surface measured 2-‐hour aber the beam shutdown of 300 kW rou?ne user opera?on

0.17 ~ 0.5

0.02 ~

0.18

<0.07 <0.05 0.09 ~ 0.6

BLM signal distribu?on along the ring

u Beam loss amount is <1% u Min cause of beam loss is foil scaiering during injec?on. u  Most of the beam loss are well localized at the collimator sec?on.

300 kW rou3ne user opera3on

< 1 mSv/h (@30 cm) over the ring

RCS is now stably providing 300 kW beam for users within the permissible beam loss level.

Einj=181 MeV

Accident in the slow extrac3on from MR to HD target

2 s

5 ms

-‐ Normal slow extrac?on from MR to HD target: the beam (3 x 1013) was slowly extracted during 2 s using a 3rd order resonance.

-‐ Unusual extrac?on on May 23, 2013 : major part of the total beam (2 x 1013) was promptly extracted during a very short period of 5 ms due to the malfunc?on of the slow extrac?on device EQ, which is a quadrupole to adjust the distance between the opera?ng tune and the 3rd order resonance. u  The gold target in the HD hall used to produce

secondary par?cles was heated up and par?ally damaged, causing dispersion of radioac?ve material

in the chamber. u  The radioac?ve material then leaked into the hadron hall, and also to the outside of the radia?on controlled area by opera?on of the ven?la?on fans. u  Consequently, then, 34 radia?on workers in the HD hall were exposed to radia?on, 0.1-‐1.7 mSv/h (total internal and external dose).

Normal slow extr.

Unusual extr. on May 2013

Since then, J-‐PARC opera?on has been suspended.

u  The external expert review commiiee is now inves?ga?ng the detailed cause of the accident and its preven?ve measure. u  The re-‐startup date of the J-‐PARC opera?on is s?ll unclear, but we are now advancing following the original schedule -‐ re-‐startup of the user program in the end of January 2014 including the linac energy upgrade to 400 MeV.

Detailed report (in Japanese and English) is available on hip://j-‐parc.jp/

High intensity beam trial of up to 540 kW

15.012 32 −≈−

f

pt

Brnεγπβ

Significance of high intensity beam trial of up to 540 kW

Einj=400 MeV 50 mA linac peak current x 0.53 chopper beam-‐on duty x 307 turns (0.5 ms) x 25 Hz →1 MW at 3 GeV (8.3E13/pulse)

Einj=181 MeV 24.5 mA linac peak current x 0.60 chopper beam-‐on duty x 235 turns (0.5 ms) x 25 Hz →540 kW at 3 GeV (4.5E13/pulse)

24.012 32 −≈−

f

pt

Brnεγπβ >

Laslet value at injec?on; Laslet value at injec?on;

Design beam opera?on:

The space-‐charge effect at injec?on is 1.6 ?mes higher than that in the 1-‐MW design beam opera?on with the higher injec?on energy of 400 MeV.

where ε=216π mm mrad for both cases

The experimental data will serves as a valuable benchmark test for -‐ Realizing the 1 MW design beam opera?on -‐ Discussing the further RCS power upgrade scenario (1-‐2 MW) in future

u  Date ; Nov., 2012 u  Injec?on beam; 181 MeV/24.5 mA/0.5 ms/0.60 chopper beam-‐on duty factor ⇒ 4.5E13/pulse, corresponding to 540 kW output at 25 Hz. u  Opera?ng point; (6.45, 6.42)

Experimental setup

In this experiment, we measured; -‐ Injec?on pain?ng parameter dependence of beam loss 　 -‐ Intensity dependence of beam loss -‐ Time structure of beam loss -‐ Transverse & longitudinal beam profiles, and bunching factor . . . . . . Experimental results vs. numerical simula?ons

Simpsons (PIC par?cle tracking code developed by Dr. Shinji Machida) Imperfec3ons included: u Time independent imperfec?ons 　 -‐ Mul?pole field components for all the main magnets: 　 BM (K1~6), QM (K5, 9), and SM (K8) obtained from field measurements 　 -‐ Measured field and alignment errors u Time dependent imperfec?ons 　 -‐ Sta?c leakage fields from the extrac?on beam line: 　 K0,1 and SK0,1 es?mated from measured COD and op?cal func?ons 　-‐ Edge focus of the injec?on bump magnets: 　 K1 es?mated from measured op?cal func?ons 　-‐ BM-‐QM field tracking errors 　 es?mated from measured tune varia?on over accelera?on 　-‐ 1-‐kHz BM ripple 　 es?mated from measured orbit varia?on 　-‐ 100-‐kHz ripple induced by injec?on bump magnets 　 es?mated from turn-‐by-‐turn BPM data u  Foil scaiering: 　 Coulomb & nuclear scaiering angle distribu?on calculated with GEANT

Numerical simula3on setup

We are improving calcula?on model step-‐by-‐step following the progression of beam experiment in collabora?on with Dr. S. Machida, discussing space-‐charge effect and its combined effects with machine imperfec?ons.

Time-‐dependent imperfec?ons can be included easily, because “Simpsons” takes “?me” as an independent variable.

Beam loss reduc3on by injec3on pain3ng -‐ H. Hotchi et al, PRST-‐AB, 15, 040402 (2012). -‐ M. Yamamoto et al, NIM., Sect. A 621, 15 (2010). -‐ F. Tamura et al, PRST-‐AB 12, 041001 (2009).

Transverse pain3ng

Horizontal pain?ng by a horizontal closed orbit varia?on during injec?on

Ver?cal pain?ng by a ver?cal injec?on angle change during injec?on

Transverse pain?ng makes use of a controlled phase space offset between the centroid of the injec?on beam and the ring closed orbit to form a different par?cle distribu?on of the circula?ng beam from the mul?-‐turn injected beam.

εtp= 0~216π mm mrad

“Correlated pain?ng”

Transverse pain3ng

No pain?ng

100π transverse pain?ng

Horizontal Ver?cal

Numerical simula?ons Transverse beam distribu?on just aber beam injec?on (at 0.5 ms)

x‘ (m

rad)

y‘ (m

rad)

x (mm) y (mm) y (m

m)

x (mm) Posi?on (mm)

x‘ (m

rad)

y‘ (m

rad)

x (mm) y (mm)

y (m

m)

x (mm) Posi?on (mm)

Density

(Arb.)

Density

(Arb.)

Longitudinal pain3ng Longitudinal pain?ng makes use of a controlled momentum offset to the rf bucket in combina?on with superposing a second harmonic rf to get a uniform bunch distribu?on aber the mul?-‐turn injec?on.

Momentum offset injec?on

Δp/p=0, -‐0.1 and -‐0.2%

RF voltage paiern

Uniform bunch distribu?on is formed through emiiance dilu?on by the large synchrotron mo?on excited by momentum offset.

The second harmonic rf fills the role in shaping flaier and wider rf bucket poten?al, leading to beier longitudinal mo?on to make a flaier bunch distribu?on.

Fundamental rf

Second harmonic rf

V2/V1=80%

Time (ms) RF voltage (kV) V1

V2

Longitudinal pain3ng

V2/V1=0

Vrf=V1sinφ-‐V2sin{2(φ-‐φs)+φ2}

Phase sweep of the second harmonic rf

(A) φ2=-‐100 deg (B) φ2=-‐50 deg (C) φ2=0

The second harmonic phase sweep method enables further bunch distribu?on control through a dynamical change of the rf bucket poten?al during injec?on.

Addi?onal control of longitudinal pain?ng ; phase sweep of V2 during injec?on

φ2=-‐100⇒0 deg

V2/V1=80%

φ (Degrees)

RF poten

?al w

ell (Arb.)

Longitudinal pain3ng

No longitudinal pain?ng

V2/V1=80% φ2=-‐100 to 0 deg Δp/p= 0.0%

V2/V1=80% φ2=-‐100 to 0 deg Δp/p=-‐0.1%

V2/V1=80% φ2=-‐100 to 0 deg Δp/p=-‐0.2%

Measurements (WCM) Numerical simula?ons

Longitudinal beam distribu?on just aber beam injec?on (at 0.5 ms)

Bf=0.15 Bf=0.40

Δp/p (%

)

φ (degrees)

Density

(Arb.)

φ (degrees)

Δp/p (%

)

φ (degrees)

Density

(Arb.)

φ (degrees)

Δp/p (%

)

φ (degrees)

Density

(Arb.)

φ (degrees)

Δp/p (%

)

φ (degrees)

Density

(Arb.)

φ (degrees)

Beam loss reduc3on by pain3ng

ID εtp (π mm mrad)

V2/V1 (%)

φ2 (deg)

Δp/p (%)

1 -‐ -‐ -‐ -‐

2 100 -‐ -‐ -‐

3 -‐ 80 -‐100 -‐0.0

4 -‐ 80 -‐100 -‐0.1

5 -‐ 80 -‐100 -‐0.2

6 100 80 -‐100 -‐0.0

7 100 80 -‐100 -‐0.1

8 100 80 -‐100 -‐0.2

9 150 80 -‐100 -‐0.2

10 200 80 -‐100 -‐0.2

11 216 80 -‐100 -‐0.2

● Measurements ○ Calcula?ons

Pain?ng parameter ID

Beam

survival

Beam survival: output intensity (DCCT) /input intensity (SCT76)

By longitudinal pain?ng

By adding 100π transverse pain?ng

Min. loss:~2% Max. output: 539 kW

No pain?ng

ID1⇒ID8 Beam loss reduc?on by pain?ng ID8⇒ID11 Beam loss increase caused by larger transverse pain?ng (εtp>150π) due to the aperture limit.

Tune footprint at the end of injec3on

Par?cles here suffer from emiiance dilu?ons, leading to the big beam loss.

ν x

νy

ν x

νy

-‐ 100π transverse pain?ng -‐ Full longitudinal pain?ng (V2/V1=80%, φ2=-‐100 to 0 deg, Δp/p=-‐0.2%)

-‐ No transverse pain?ng -‐ No longitudinal pain?ng

Numerical Simula?on

The beam loss reduc?on achieved by injec?on pain?ng can be interpreted as the outcome of the space charge mi?ga?on led through the charge density control by injec?on pain?ng and its resultant mi?ga?on of the influence from the betatron resonances.

ID1 ID8

x=6

y=6

x=6

y=6

The tune shib is mi?gated by injec?on pain?ng

Intensity dependence of beam loss, bunching factor, extrac3on beam profile・・・

Beam

survival

Li pulse length (µs)

539 kW 433 kW

326 kW

217 kW

104 kW

539 kW (Li pulse 500 µs) 433 kW (Li pulse 400 µs) 325 kW (Li pulse 300 µs) 217 kW (Li pulse 200 µs) 104 kW (Li pulse 100 µs)

Beam survival : ra?o of output intensity (DCCT) to input intensity (SCT76)

Pain?ng parameter ID8 : -‐ 100π transverse pain?ng -‐ Full longitudinal pain?ng

Measurement vs calcula3on: intensity dependence of beam loss


539 kW (500 µs) 433 kW (400 µs) 326 kW (300 µs) 217 kW (200 µs) 104 kW (100 µs)

Time (ms) 0.7 ms : End of foil scaiering

SBLM

signal (A

rb.)

Beam

loss (%

)

Time (ms)

# of lost par?cles/turn

Beam

loss (%

)

Calcula?ons

0.7 ms : End of foil scaiering

Measurement vs Calcula3on: intensity dependence of beam loss (3me structure)

Scin?lla?on type BLM @ Collimator

Normalized to be 2%

(A) (B)

The beam loss appears only for the first 4 ms in the low energy region, and has a characteris?c two peak structure (A) and (B) for higher intensity beams.

Measurement vs Calcula3on: bunching factor from injec3on to extrac3on

Time (ms)

Bunching factor

Time (ms)

539 kW (500 µs) 433 kW (400 µs)

326 kW (300 µs) 217 kW (200 µs)

104 kW (100 µs)

― Calcula?ons

Measurement vs Calcula3on: extrac3on beam profile at 3 GeV

― Calcula?ons

MWPM @ extrac?on beam line

539 kW (500 µs)

433 kW (400 µs)

326 kW (300 µs)

217 kW (200 µs)

104 kW (100 µs)

Charge den

sity (Arb.)

Posi?on (mm)

Li pulse length (µs) RM

S width (m

m)

Ver?cal

Horizontal


Horizontal Ver?cal Intensity dependence of RMS beam width

Our numerical simula?on well globally reproduced the experimental results up to 540 kW intensity beam.

Discussions for remaining beam loss


Time (ms) 0.7 ms : End of foil scaiering

SBLM

signal (A

rb.)

Beam

loss (%

)

Time (ms)


Beam

loss (%

)

Calcula?ons


Scin?lla?on type BLM @ Collimator

The beam loss appears only for the first 4 ms in the low energy region: u Cause of (A) Foil scaiering during charge-‐exchange (H- to proton) injec?on u Cause of (B) 100-‐kHz dipole field ripple induced by injec?on bump field

(A) (B)

(A) (B)

2%

Possible causes of remaining beam loss

Time (ms)



Time structure of beam loss calculated for 540 kW (500 µs) intensity beam

Without 100-‐kHz ripple induced by injec?on bump field

With 100-‐kHz ripple induced by injec?on bump field (all machine imperfec?ons)

Loss by 100 kHz ripple ~1.6%

Loss by foil scaiering ~0.5%

Possible causes of remaining beam loss

Conclusion from the numerical simula?on: the main part of remaining beam loss arises from foil scaiering and 100-‐kHz ripple induced by injec?on bump field.

Time (ms)

Kick angle of ripple (m

rad)

Flaiop Fall ?me of injec?on bump

Time structure of 100-‐kHz dipole field ripple

Horizontal Ver?cal

~100 kHz

Frequency (MHz)

Power (d

B)

FFT spectrum of BPM signal frev x 5 β side-‐band

peak Side-‐band peak excited by 100-‐kHz ripple induced by injec?on bump field

This fast ripple excites addi?onal betatron resonances at 0.2 during injec?on (first ~1 ms) and affects the beam.

100-‐kHz dipole field ripple induced by injec3on bump field

Horizontal

ν x

νy

539 kW (500 µs) 433 kW (400 µs)

326 kW (300 µs) 217 kW (200 µs) 104 kW (100 µs)

0.2 (100 kHz/frev at injec?on)

0.2

Effect of 100-‐kHz dipole field ripple Tune footprint calculated at the end of injec?on

u  100-‐kHz ripple makes addi?onal betatron resonances at 0.2 for the first 1 ms period. u  A part of beam par?cles reaches to 0.2 lines due to space-‐charge tune depression, where the effect of ripple builds up, leading to emiiance growth.

0.2

0.2

0.2

0.2

0.2

0.2

0.2

0.2

Core on 0.2 lines

Tail on 0.2 lines

Present opera?ng point

The situa?on for higher intensity beams is more cri?cal in terms of halo/tail forma?on and its resultant beam loss, because the 100-‐kHz ripple directly affects a tail part of the beam.

Effect of 100-‐kHz dipole field ripple

x (m) y (m)

539 kW

433 kW

326 kW

217 kW

104 kW ― Without 100-‐kHz ripple ― With 100-‐kHz ripple

Beam profile calculated at the end of injec?on (ploied in log scale)

Horizontal Ver?cal

Charge den

sity (Arb.)

Larger beam halo/tail forma?on takes place for higher intensity beam, leading to beam loss.

This is the main cause of the second beam loss structure (B) and also this is the reason why the second beam loss structure (B) is observed only for higher intensity beams.

RF shield strip (Cu) Capacitor

Flange

Schema?c drawing of the rf-‐shielded ceramics chamber installed in the injec?on bump magnets

Equivalent circuit of the RF shield

Source of 100-‐kHz dipole field ripple

Resonant current in the rf shield loop induced by dB/dt of the injec?on bump field

u  If the rf shield keep a symmetric configura?on, the ripple field can be canceled out through the four injec?on bump magnets (- + + -). u  But now the symmetric condi?on is destroyed because a part of capacitors is now removed; some capacitors burned out in turning on the injec?on bump field because of the lack of withstand voltage. u Such asymmetric configura?ons of the rf shield spoil the cancela?on and the remaining field ripple component affects the beam at present.

New ceramics chamber with modified rf-‐shield structure without the chip of capacitors will be installed. ⇒　The excess 2% beam loss observed for 540 kW intensity beam will be decreased to less than 1%.

Y. Shobuda et al, PRST-‐AB 12, 032401 (2009)

Further beam loss reduc3on (first-‐aid treatment)

November 2012 April, 2013


Time (ms)

SBLM

signal (A

rb.)

Beam

loss (%

)

BLM @ collimator

~0.7 ms : End of foil scaiering


Time (ms)

SBLM

signal (A

rb.)

Beam

loss (%

)

BLM @ collimator

~0.7 ms : End of foil scaiering

~2% ~2.7%

2nd High intensity trial in April 2013

Injec?on beam : 24.5 mA, 100-‐500 µs, 640 ns, 2 bunches

Injec?on beam : 25.5 mA, 100-‐500 µs, 640 ns, 2bunches

Effort for further beam loss reduc3on

Time (ms)

Horizon

tal tun

e Ve

r?cal tun

e

(A) Ac?ve tune change to mi?gate the resonant condi?on of ν=0.2

Tune 1 (original)

Tune 2

Measurement

Tune 1 (original)

Tune 2

Foil

Δx

(B) Re-‐adjustment of foil posi?on to reduce the foil hi|ng rate

Δx : 13 mm ⇒ 8 mm

Injec?on beam (H : 7 mm, V : 7mm)

Main cause of remaining beam loss observed for ~560 kW beam : -‐ 0.2 resonance excited by 100 kHz dipole ripple -‐ Foil scaiering

This kinds of ac?ve tune change should lead to the mi?ga?on of the build-‐up of the ripple kick.

Δy 40 mm Pulled out

Li pulse : 500 µs

Time (ms)

SBLM

signal (A

rb.) BLM @ collimator

Beam

loss (%

)

Tune 1 Tune 2

Tune 1 Tune 2

2.7% ⇒ 2.0%⇒1.4%

Foil pos. Δx 13⇒8 mm

Foil pos. Δx 13⇒8 mm

Effort for further beam loss reduc3on

u Beam loss for 560 kW beam was reduced almost half by (A) and (B). u Most of the remaining beam loss of 1.4% was well localized at the collimator regions. u This 1.4%-‐loss corresponds to 470 W in power, which is s?ll less than 1/8 of the current collimator limit 4 kW.

We hope to try 500 kW rou?ne beam opera?on, but the RCS beam power is now limited to 300 kW to minimize the damage of the neutron produc?on target as possible.

Beam commissioning plan for the next two years (original plan before the accident on May 23, 2013) u We con?nued 300 kW rou?ne user opera?on un?l the end of May, 2013 u Summer-‐autumn shutdown in 2013 -‐ Linac upgrade: ＊Installa?on of the ACS linac sec?on (Output energy ; 181 MeV⇒400 MeV) -‐ RCS upgrade & maintenance: ＊Upgrade of the power supply of injec?on bump magnets ＊Installa?on of 12th RF cavity ＊Re-‐alignment, etc. u  December 2013~ -‐ Beam commissioning of 400 MeV linac u  January 2014~ -‐ Beam commissioning of RCS with 400 MeV injec?on energy u End of January 2014 -‐ Re-‐startup of the user program u  Summer shutdown in 2014 -‐ Linac upgrade ＊Upgrade of the front-‐end system IS & RFQ 　 (Maximum peak current ; 30 mA⇒50 mA) u October 2014~ -‐ Linac and RCS start beam tuning toward the 1MW output beam power

The re-‐startup date of the beam opera?on is s?ll unclear. Possibly the schedule will be shibed following the response of external review commiiee for the accident in the HD hall.

1-‐MW beam simula3on with Einj=400 MeV

Time (ms)


Beam

loss (%

) Time (ms)

Beam loss for the first 4 ms region

Loss ~0.5% at injec?on, which is mainly from foil scaiering ⇒ 670 W << 4 kW (collimator limit)

Injec?on beam ; 181 MeV/50 mA/0.5 ms/0.533 chopper beam-‐on duty factor ⇒ 8.3E13/pulse, corresponding to 1 MW output at 25 Hz

The numerical simula?on gives a posi?ve sign for realizing the 1 MW design opera?on. But we s?ll have several issues for 1 MW beam opera?on.

u Beam instability arising from the extrac?on pulse kicker impedance: -‐ Introduc?on of matching resistor and diode to the power supplies for damping beam-‐induced currents in the kicker magnets -‐ Modifica?on of the coaxial cable lengths for reducing a pile-‐up of beam-‐induced currents among the kickers . . . u Quality of the 1-‐MW extrac?on beam, namely the beam halo/tail reduc?on, which is a key issue especially for the beam injec?on to the MR -‐ Re-‐op?miza?on of the opera?ng tune -‐ Introduc?on of longer-‐dura?on 2nd harmonic rf to further improve the bunching factor in the low energy region . . . . . . etc.

Issues for the 1-‐MW beam opera3on

Summary u  The RCS beam power ramp-‐up has steadily proceeded since the startup of

user program following the progression in beam tuning and hardware improvements. u  The RCS is now stably providing 300 kW beam for users. u We have successfully achieved a high intensity trial of up to 540 kW. u  The numerical simula?on well reproduced the experimental results up to 540 kW intensity beam. u  Accelerator modeling and quan?ta?ve benchmarking between experiment and simula?on becomes feasible. u  Beam commissioning of the RCS proceeded efficiently through the itera?on between experiment and numerical simula?on. u We are progressing toward star?ng the 1 MW beam tuning in October 2014 aber upgrading the linac.

Documents

Currentstatusof’ beam’commissioning’andoperaon’ of… · 2017. 1. 17. · 2 /V 1 (%) φ 2 (deg) Δp/p (%) 1 5 5 5 5 2 100 5 5 5 3 5 80 100 0.0 4 5 80 100 0.1 5 5 80 100 0.2