PAPER • OPEN ACCESS
Beam-ion losses eloci y-space dis ibu ion unde
neu al-beam injec ion on EAST
To ci e his a icle: S.S. Wang
e al
2025
Nucl. Fusion
65 016026
View he a icle online o upda es and enhancemen s.
You may also like
EMC3-EIRENE simula ions o edge
plasma and impu i y anspo by o oidally
localized a gon seeding on CFETR X-
di e o
T. Xie, H. Li, W. Zhang e al.
-
In luence o he densi y g adien on
u bulen hea anspo a ion-scales: an
in e -machine s udy wi h he gy okine ic
code s ella
H. Thienpond , J.M. Ga cía-Regaña, I.
Cal o e al.
-
Expe imen al cha ac e is ics o los as
nega i e ions on EAST okamak
Z.X. Zhang, J. Huang, J.F. Chang e al.
-
This con en was downloaded om IP add ess 193.147.173.203 on 24/04/2025 a 09:01
In e na ional A omic Ene gy Agency
Nuclea Fusion
Nucl. Fusion 65 (2025) 016026 (15pp) h ps://doi.o g/10.1088/1741-4326/ad933
Beam-ion losses eloci y-space
dis ibu ion unde neu al-beam
injec ion on EAST
S.S. Wang1,3,a, Z.X. Zhang1,2,a, J. Huang1,∗, J.F. Chang1,∗, J. Galdon-Qui oga4,
L. Sanchis4, W. Gao1, J. Fu1,2, Y.X. Sun1,2, X.H. Wang1,2, C. Shi1,2
and he EAST Team1,b
1He ei Ins i u es o Physical Science, Chinese Academy o Sciences, He ei 230031, China
2Uni e si y o Science and Technology o China, He ei 230026, China
3Qingdao Si ang SRI In ellec ual Technology Co., L d, Qingdao 266031, China
4Depa men o A omic, Molecula and Nuclea Physics, Uni e si y o Se ille, Se ille 41012, Spain
E-mail: [email p o ec ed] and [email p o ec ed]
Recei ed 21 Augus 2024, e ised 5 No embe 2024
Accep ed o publica ion 15 No embe 2024
Published 26 No embe 2024
Abs ac
The eloci y-space dis ibu ion o he as -ion loss in EAST neu al-beam injec ion (NBI)
hea ing discha ge is ob ained bo h om Scin illa o -based as -ion loss de ec o (FILD) signals
and by ASCOT5 and FILDSIM simula ions. The esul s o simula ions a e in good ag eemen
wi h he dis ibu ion o beam-ion losses measu ed wi h FILD o EAST and he co ec ness o
he as -ion loss dis ibu ion has been demons a ed. Simula ions indica e ha he beam-ion
losses obse ed by he FILD p obe a e a ibu ed o he as ions om bo h he high- ield side
(HFS) and he low- ield side (LFS). Howe e , he beam-ion losses om he HFS (associa ed
wi h NBI1L) ha e no been obse ed expe imen ally due o he limi ed de ec ing ange o he
FILD p obe. The e o e, an upg ade and modi ica ion o he FILD p obe was ca ied ou in 2022
o enable he de ec ion o as -ion loss wi h smalle pi ch angles. Compa a i e analysis is
conduc ed in neu al-beam injec ion (NBI2R) discha ges a e he upg ade, which indica es ha
he eloci y-space dis ibu ion o beam-ion losses om he HFS has s ong ag eemen be ween
expe imen al measu emen s and simula ion esul s. Howe e , he expe imen al and simula ed
esul s o he eloci y-space dis ibu ion o beam-ion losses om he LFS shows
inconsis encies, p ima ily because he BBNBI module in he simula ion does no conside he
con ibu ions o bounda y neu al pa icles o neu al-beam deposi ion (ioniza ion eac ions).
These conclusions no only p o ide aluable e e ences o imp o ing he neu al-beam
deposi ion model bu also es ablish a undamen al basis o u he explo ing he mechanisms o
as -ion loss unde a ious condi ions on he EAST okamak.
aThese au ho s con ibu ed equally o his wo k and should be conside ed co- i s au ho s.
bSee Wan e al 2017 (h ps://doi.o g/10.1088/1741-4326/aa7861) o he EAST Team.
∗Au ho s o whom any co espondence should be add essed.
O iginal con en om his wo k may be used unde he e ms
o he C ea i e Commons A ibu ion 4.0 licence. Any u -
he dis ibu ion o his wo k mus main ain a ibu ion o he au ho (s) and he
i le o he wo k, jou nal ci a ion and DOI.
1741-4326/25/016026+15$33.00 P in ed in he UK 1 © 2024 The Au ho (s). Published by IOP Publishing L d on behal o he IAEA
Nucl. Fusion 65 (2025) 016026 S.S. Wang e al
Keywo ds: beam-ion losses, eloci y-space dis ibu ion, ASCOT, FILDSIM
(Some igu es may appea in colou only in he online jou nal)
1. In oduc ion
I is well known ha e ec i e con ol o as -ion loss is one o
he key physical challenges in achie ing s eady-s a e and high-
pe o mance plasma o u u e usion de ice such as ITER
and CFETR. Fas -ion losses in usion de ices can educe hea -
ing and cu en d i e e iciency and may e en cause localized
damage o he i s wall i concen a ed [1]. Fas -ion popula-
ions c ea ed by usion eac ions, by neu al-beam injec ion
(NBI) and by adio equency (RF) hea ing a e usually concen-
a ed in he cen e o he plasma [2]. The eloci y-space dis i-
bu ion o los as ions is a key physical quan i y o s udying
he con ol mechanisms o he as -ion loss, as i enables he
esolu ion o los as -ions o bi s h ough ace e e sal.
Scin illa o -based as -ion loss de ec o (FILD), which can
measu e pi ch angle and gy o adius simul aneously, is an
impo an diagnos ic ool o s udy as -ion loss. The FILD dia-
gnos ic has been applied in se e al okamaks, such as ASDEX
Upg ade [3], MAST-U [4], DIII-D [5], JET [6], KSTAR [7]
and NSTX [8]. FILD has been ins alled on Expe imen al
Ad anced Supe conduc ing Tokamak (EAST) o in es iga e
he beha io o as -ion losses. I p o ides eloci y-space dis-
ibu ion du ing NBI and ion cyclo on esonance hea ing
(ICRH) [9].
In his wo k, he eloci y-space dis ibu ion is in e ed om
FILD measu emen s and ASCOT simula ion in he EAST. The
da a analysis ools consis o FILDSIM [10] and ASCOT [11].
The model is based on he FILD s ike-map and FILDSIM
model. FILDSIM is a Mon e Ca lo o bi - acing code ha can
es ablish connec ions be ween losses a he i s wall and a
FILD scin illa o and i has been applied in ASDEX Upg ade
and MAST-U [12] amongs o he s. The eloci y-space dis-
ibu ion o beam ion p omp loss a FILD loca ion can be
ob ained by ASCOT because i can also simula e in sc ape-
o laye (SOL) and FILD p obe loca e a SOL. ASCOT has
been applied in ASDEX Upg ade [13], W7-X [14], ITER [15]
and JET [16] unde NBI [17] and ICRF [18] hea ing o as -
ion loss [19], wall load [20] and neu on [21] s udies in usion
de ices.
The wo k is s uc u ed as ollows. A e he in oduc ion in
sec ion 1, he FILD sys em on EAST is in oduced in sec ion 2.
Sec ion 3desc ibes he e i ica ion and analysis o beam-
ion loss eloci y-space dis ibu ion by FILD on EAST. The
impac o neu al pa icles on beam-ion losses is discussed in
sec ion 4. Sec ion 5summa izes he ea u es o beam-ion loss
eloci y-space dis ibu ion in he EAST.
2. Expe imen al se ups
FILD has been ins alled on EAST o in es iga e as -ion loss
beha io in high pe o mance plasma [9]. The NBI hea ing
sys em is shown in igu e 1(a) and loca ion and di ec ion o
NBI2 has been changed in NBI upg aded a 2020. The ecip-
oca ing FILD is ins alled abo e he ou e midplane on EAST
Po J as shown in igu e 1(b). FILD diagnos ic sys em is com-
posed o he de ec o head, he exchange box, he op ical spli -
e , he de ec ion/da a acquisi ion sys em, and he long sha
sys em [22]. A high-speed cha ge-coupled de ice (CCD) cam-
e a sys em (Phan om V2010) cap u es images o he scin il-
la o sc een, p o iding in o ma ion abou he pi ch angle and
gy o adius o los as -ions. Addi ionally, a 25-channel pho-
omul iplie ube (PMT) da a acquisi ion sys em acks he
e olu ion o he as -ion loss signal o e ime [23]. The main
componen s o he de ec o head a e he ion collima o , which
consis s o a pinhole, sli and he scin illa o sc een, as shown
in igu e 2. The p obe geome y was upg aded du ing expe -
imen s conduc ed in he i s hal o 2022. The pinhole was
eloca ed om he cen e o he collima o o i s pe iphe y
o ex end he de ec ion ange o he FILD, as illus a ed in
igu e 3(a). The key pa ame e s o p obe such as leng h and
wid h o he collima o pinhole, he dis ance om collima o
pinhole o he scin illa o , he leng h and wid h o he scin il-
la o de e mine he posi ion o as ions s iking he scin illa o
wi h di e en ene gy and pi ch angles. The ajec o y o as -
ion in he p obe is shown in igu e 3(b). The los as ions pass
sequen ially h ough pinhole and sli and inally s ike o he
scin illa o sc een.
Be o e his wo k, he gy o adius o los ions was calcula ed
based on he assump ion o a ci cula o bi passing h ough
he cen e o he pinhole, sli , and he s ike poin on he scin-
illa o . Howe e , his me hod is inaccu a e due o he limi -
a ions imposed by he ini e sizes o he pinhole and sli . In
ac , as ions wi h ixed ene gy and gy o adius a on ape -
u e will p esen a dis ibu ion when hey s ike o scin illa o
ins ead o a poin . Hence, i is necessa y o es ablish a connec-
ion be ween eloci y-space dis ibu ion a on ape u e and
eloci y-space dis ibu ion a scin illa o , o e en he CCD. In
o de o cha ac e ize he esponse o he FILD de ec o , we
in oduced FILDSIM o ajec o y calcula ion [10]. To e i y
he dis ibu ion o beam-ion loss, ASCOT has been applied
o simula e he eloci y-space dis ibu ion o beam-ion loss a
FILD p obe loca ion.
3. Beam-ion losses eloci y-space dis ibu ion
measu ed by FILD
3.1. Analysis and modelling ools
The main analysis and modelling ools in his wo k a e
FILDSIM and ASCOT. FILDSIM is a code o ajec o y cal-
cula ions o he ions in he FILD p obe. Based on he ini ial
eloci y-space dis ibu ion o he incoming ions ha each he
2
Nucl. Fusion 65 (2025) 016026 S.S. Wang e al
Figu e 1. (a) Dis ibu ion o he neu al-beam injec ion (NBI) hea ing sys ems and loca ion o FILD ( op iew). The con igu a ion o NBI2
has been changed du ing he NBI upg ade in 2020. The black beam line ep esen s NBI2 a e 2020, while he blue beam line ep esen s
NBI2 be o e 2020. The NBI1L and NBI2R a e mo e angen ial han NBI1R and NBI2L be o e 2020. The NBI1L and NBI2L a e mo e
angen ial han NBI1R and NBI2R a e 2020. (b) Loca ion o FILD a R-Z space a J po ( o oidal angle =22.5◦).
Figu e 2. The 3D model o de ec o head. The pinhole o he de ec o head has been mo ed om he cen e o he edge o he collima o .
(a) Be o e 2022. (b) A e 2022.
pinhole, FILDSIM is employed o gene a e he co espond-
ing dis ibu ion on he scin illa o pla e [10,24]. FILDSIM
model uses a weigh unc ion o malism o ela e he eloci y-
space dis ibu ion o as -ion losses eaching he de ec o pin-
hole o he scin illa o pa e n ob ained expe imen ally, which
can be unde s ood as a dis o ion o he eloci y-space dis i-
bu ion due o he ini e esolu ion o he sys em. The weigh
unc ion consis s o e iciency unc ion and p obabili y ma ix.
The e iciency unc ion is associa ed wi h scin illa o yield and
p obabili y ma ix can be calcula ed by FILDSIM. A de ailed
desc ip ion o he FILDSIM model can be ound in [10]. In his
wo k, we only conside ed he p obabili y ma ix and se he
e iciency unc ion as cons an o quali a i e analysis. Since
we only in es iga e NBI ions ha p ima ily lose ene gy close
o he NBI injec ion ene gy (which is nea ly monoene ge ic),
he use o a cons an yield is jus i ied. We will also assess
he luminescence e iciency (yield) o he scin illa o ma e -
ial (ZnS:Ag) in he u u e. The esolu ions o he FILD on
EAST we e ob ained using FILDSIM a e inco po a ing he
geome ical pa ame e s o he FILD in o he simula ion. The
esolu ion in gy o adius is illus a ed in igu e 4(a), which
shows gy o adius dis ibu ion p o iles ob ained in he scin il-
la o eloci y-space along a line o cons an pi ch angle. The
di e en colo s co espond o di e en alues o he pa icle
gy o adius s a ed a he de ec o pinhole. The s ike poin dis-
ibu ions can be ai ly well modeled as skew Gaussian, simila
3
Nucl. Fusion 65 (2025) 016026 S.S. Wang e al
Figu e 3. (a) The posi ion o pinhole ela i e o scin illa o . The
g een line is posi ion o pinhole be o e 2022 and ed lines a e
posi ions o pinhole a e 2022. (b) Fas -ion o bi which s ikes o
scin illa o in FILD p obe.
o he FILD on ASDEX Upg ade. The esolu ion o gy o adius
is poo e o la ge gy o adius o he FILD. The esolu ion in
pi ch-angle illus a ed in igu e 4(b) indica es he s ike-poin
dis ibu ions o pi ch-angle i s o Gaussian unc ion. The es-
ul s sugges ha he FILDSIM model is applicable o he FILD
on EAST.
ASCOT is a code used o simula ing he mo ion o as -
ions and impu i y pa icles h ough Mon e Ca lo pa icle
acking me hods [11]. The in o ma ion (e.g. loca ion, o bi ,
momen um, ene gy, and dis ibu ion) o ma ke s in usion
de ice can be ob ained by ASCOT simula ion. The eloci y
dis ibu ion o beam ion p omp loss a FILD loca ion can be
ob ained since FILD is si ua ed in he SOL, and ASCOT also
allowed simula ions in he SOL egion. The inpu pa ame -
e s o ASCOT mainly include ma ke s, magne ic ield, elec-
ic ield, plasma p o iles, and wall model. The ma ke s we e
gene a ed by an ASCOT model named BBNBI5 [25], which
akes in o accoun he ine s uc u e o he injec o , ollows
he injec ed neu als un il ioniza ion, and gene a es a sou ce
ensemble o ionized NBI ma ke s o slowing down calcula-
ions. Figu e 5shows he di ec ion o he magne ic ield and
plasma cu en o discha ge #85626 in ASCOT, along wi h
he 2 million ma ke s o NBI1L gene a ed by BBNBI5. The
magne ic ield and plasma cu en a e bo h di ec ed coun e -
clockwise in he op iew o discha ge #85626.
3.2. Resul s o expe imen s and simula ions
To esea ch he eloci y-space dis ibu ion o beam-ion loss
in EAST, we chose h ee ime poin s ( 1=3.1 s, 2=4.5 s,
3=6.5 s) in discha ge #85626 o analysis. Discha ge #85626
is a H-mode discha ge be o e FILD p obe geome y was
upg aded in 2022 and beam geome y was upg aded in 2020.
As shown in igu e 6, ed dashed line indica es he h ee ime
poin s and each ime poin co esponds o only one NBI beam-
line. The ene gies o h ee NBI beamlines a e all abou 50 keV.
The plasma cu en is 400 kA and B =2.40 T. The plasma
p o iles o discha ge #85626 a e shown in igu e 7, he elec on
densi y p o iles a e de i ed om e lec ome y sys em [26] and
Pola ime e -In e e ome e (POINT) sys em [27], while he
elec on empe a u e and ion empe a u e p o iles a e espec -
i ely acqui ed h ough TS (Thomson Sca e ing) Diagnos ic
Sys em [28] and CXRS (Cha ge Exchange Recombina ion
Spec oscopy) [29].
Co e ed he s ike map g id o he FILD signal o discha ge
85626 a 3.1 s, 4.5 s and 6.5 s, he eloci y-space dis ibu-
ion o los as -ion has been shown in igu e 8. FILD can
de ec as -ion wi h pi ch angles anging om 50◦ o 140◦
based on he s ike map g id, and we selec ed 1 cm o 12 cm
in he g id because he esolu ion o gy o adius is poo e o
la ge gy o adius. Mo e in ui i e eloci y-space dis ibu ion
has been acqui ed h ough FILD signal by FILDSIM emap-
ping as shown in igu e 9. Figu e 9(a) showed a b igh spo
wi h gy o adius abou 2.1 cm and pi ch angle abou 57◦,
which co esponding o he as -ion loss om NBI1L injec-
ion. In he same way he gy o adius and pi ch-angle o beam
ion p omp loss o NBI2L a e espec i ely abou 2.1 cm and
66◦ om igu e 9(b). A e emapping FILD signal a 6.5 s,
he gy o adius and pi ch-angle o beam ion p omp loss o
NBI2R a e espec i ely abou 2.1 cm and 57◦ om igu e 9(c).
Conside ing he ion species (deu e ium) and he magne ic ield
a he p obe loca ion (∼1.9 T), he ene gy o los beam ions can
be calcula ed based on he gy o adius. Figu es 9(d)–( ) indic-
a e ha he ene gy o he los as -ions p oduced by he h ee
beamlines is app oxima ely 45 keV. As shown in igu e 1,
4

Nucl. Fusion 65 (2025) 016026 S.S. Wang e al
Figu e 4. (a) Gy o adius dis ibu ion wi h pi ch angle held cons an in he scin illa o eloci y space; (b) pi ch angle dis ibu ion wi h
gy o adius held cons an in he scin illa o eloci y space.
Figu e 5. Di ec ion o he magne ic ield and plasma cu en and
ma ke s om NBI1L. The blue a ow indica es he di ec ion o he
magne ic ield, while he ed a ow indica es he di ec ion o he
cu en . The black do s ep esen he es pa icles.
he NBI2L and NBI2R beam lines o igina e om he same
sou ce bu ha e di e en di ec ions, esul ing in di e en pi ch
angles in hei eloci y dis ibu ions. Meanwhile, he NBI1L
and NBI2R beam lines ha e di e en sou ces bu simila beam
line di ec ions, which leads o simila pi ch angles in hei
eloci y dis ibu ions, as he pi ch angle o he eloci y-space
dis ibu ion is ela ed o he beam line di ec ion.
The wall load da a ha e been ob ained h ough ASCOT5
simula ion o discha ge 85626 a 3.1 s. Each ma ke was
acked along i s ull o bi o up o 10 ms, un il i impac ed
he i s wall and he acking was e mina ed. A dis inc b igh
spo is isible on he wall a he FILD loca ion, wi h a o oidal
angle o app oxima ely 22.5◦and a poloidal angle o a ound
25◦, as shown in igu e 10. The ou egions o beam-ion loss
a e he FILD, he Lowe Hyb id Wa e limi e , he main lim-
i e , and he di e o . The majo i y o los beam ions a e s uck
on he main limi e . Two million ma ke s we e simula ed o e
10 ms (using 160 co es on he ShenMa High-Pe o mance
Compu ing Clus e wi h a un ime o app oxima ely 70 h) o
inc ease he numbe o ma ke s collec ed a he FILD p obe
Figu e 6. Time aces o main pa ame e s o EAST #85626. (a) Ip.
(b) Densi y. (c) Dαsignal. (d) NBI ene gy o NBI1L, NBI2L and
NBI2R. The ed dashed line indica es he h ee selec ed momen s.
loca ion. The eloci y-space dis ibu ion o beam-ion loss wi h
di e en beam injec o s ha e been ob ained by ASCOT5, and
he esul s a e shown in igu e 11. The dis ibu ion consis s o
wo a eas wi h only NBI1L injec ion as shown in igu e 11(a)
and ini ial posi ion o ma ke s o wo a eas a e di e en as
shown in igu e 12. The pi ch angle o ma ke s nea he high-
ield side (HFS) anges om 20◦ o 40◦. Fo ma ke s nea he
low- ield side (LFS), he pi ch angle is app oxima ely 57◦, and
his obse a ion is consis en wi h he dis ibu ion measu ed
by FILD. Ano he dis ibu ion a ea canno be measu ed by
FILD because he FILD pi ch-angle measu ing ange is 50◦–
140◦. Figu es 11(b) and (c) show eloci y-space dis ibu ions
o beam-ion loss wi h NBI2L and NBI2R.
The syn he ic eloci y-space dis ibu ion simula ed by
ASCOT is no a FILD scin illa o bu a he a he pin-
hole. The e o e, we should use FILDSIM model o ob ain
syn he ic scin illa o eloci y-space dis ibu ion. Figu e 13(a)
shows he syn he ic pinhole eloci y-space dis ibu ion by
5
Nucl. Fusion 65 (2025) 016026 S.S. Wang e al
Figu e 7. Plasma p o iles o EAST 85626 a =3.1 s (NBI1L),
=4.5 s (NBI2R) and =6.5 s (NBI2L) a e shown in (a) elec on
densi y. (b) Elec on empe a u e, and (c) ion empe a u e.
FILDSIM wi h ASCOT dis ibu ion (LFS: pi ch-angle ∼57◦)
as inpu . The syn he ic scin illa o eloci y-space dis ibu ion
o beam-ion loss can be gained by FILDSIM o wa d model
which conside ed p obabili y ma ix. F om igu es 9and 13,
he gy o adius o beam ion p omp loss o NBI1L and NBI2R
injec ion is abou 2.1 cm and he pi ch-angle is abou 57◦. A
he same ime, he gy o adius o he beam ion p omp loss om
NBI2L injec ion is app oxima ely 2.1 cm, wi h a pi ch angle
o abou 66◦. The beam lines NBI2L and NBI2R a e coun e -
IPbeams, demons a ing signi ican ly la ge losses compa ed
o he co-IPbeam (NBI1L). The esul s also indica e ha
he mo e angen ial beamlines (NBI1L, NBI2R) exhibi lowe
p omp losses compa ed o less angen ial beams (NBI2L),
which is consis en wi h p io s udies [30,31]. By compa ing
he gy o adius p o iles o beam-ion loss dis ibu ions ob ained
om ASCOT simula ions and FILD measu emen s, while
keeping he pi ch angle cons an (a he cen e o he b igh
spo ) ac oss di e en NBI injec ions, we ind good ag eemen
Figu e 8. Veloci y-space dis ibu ion o beam-ion loss o #85626
by FILD CCD mapping wi h di e en NBI injec ion: (a) NBI1L, (b)
NBI2L, and (c) NBI2R.
be ween he ASCOT simula ions and FILD measu emen s o
all h ee NBI injec ions, as shown in igu e 14.
3.3. O bi analysis o beam-ion losses
In o de o analyze he loss mechanism o di e en pi ch-angle
a eas in igu e 10(a), he o bi o los as -ions in wo a ea ha e
been calcula ed by ASCOT5. Figu es 15(a) and (b) espec -
i ely show o bi o HFS and LFS los beam ions wi h NBI1L.
The o bi o los as -ions whose ini ial posi ion nea HFS is
passing los o bi while he o bi o los as -ions whose ini ial
posi ion nea LFS is apped o bi . Figu e 16 shows he o bi
classi ica ion and opological bounda ies o hese wo a eas.
The o bi s o los beam ions whose ini ial posi ion nea HFS
a e all co-passing los o bi s and o bi s o los beam ions whose
ini ial posi ion nea LFS a e nea ly all apped-los o bi s.
The esul s need sho loss ime because he bounda y he e
based on conse a ion o ene gy, magne ic momen and o -
oidal canonical angula momen um. The e o e, we can dis in-
guish be ween apped pa icles and passing pa icle losses by
pi ch angle o beam-ion loss eloci y-space dis ibu ion du ing
NBI1L injec ion.
4. The impac o neu al pa icles on beam-ion
losses
4.1. Beam-ion losses wi h modi ied FILD p obe head
In o de o u he e i y he simula ed eloci y-space dis i-
bu ion, as shown in igu e 17, he FILD p obe was upg aded
6
Nucl. Fusion 65 (2025) 016026 S.S. Wang e al
Figu e 9. Veloci y-space Gy o adius-Pi ch angle dis ibu ion o beam-ion loss o #85626 by FILD CCD emap wi h di e en NBI
injec ions: (a) NBI1L, (b) NBI2L, and (c) NBI2R. Veloci y-space Ene gy-Pi ch angle dis ibu ion o beam-ion loss o #85626 by FILD
CCD emap wi h di e en NBI injec ions: (d) NBI1L, (e) NBI2L, and ( ) NBI2R.
Figu e 10. EAST #85626 a 3.1 s wall load wi h NBI1L injec ion ob ained by ASCOT5.
in he i s hal o 2022, allowing i o ecei e beam ions wi h
pi ch angles anging om app oxima ely 30◦–100◦. Discha ge
#115236 was conduc ed wi h only NBI2R injec ion a e he
FILD p obe and NBI we e upg aded on EAST. The ime aces
o main pa ame e s o EAST #115236 a e shown in igu e 18.
The ene gy o NBI2R beamline is all abou 60 keV. The
7
Nucl. Fusion 65 (2025) 016026 S.S. Wang e al
Figu e 11. The eloci y-space dis ibu ions o beam-ion loss a FILD p obe loca ion by ASCOT wi h di e en NBI. (a) NBI1L. (b) NBI2L.
(c) NBI2R. The o ange do ed lines indica e he p obe de ec ion ange. The eloci y-space dis ibu ions o ma ke s om di e en beam lines
a e used as inpu condi ions o FILDSIM, wi h he esul ing ou pu shown in igu e 13.
Figu e 12. The ini ial posi ions o he es pa icles collec ed a he FILD p obe loca ion shown in igu e 11(a). The g een do and he blue
do espec i ely ep esen es pa icles wi h a smalle pi ch angle and a la ge pi ch angle in he as -ion loss eloci y-space dis ibu ions.
plasma cu en is 500 kA and B =1.64 T. NBI2R injec ion
is a e icali y beam line a e NBI upg ade.
The plasma p o iles o #115236 a 4 s a e shown in
igu e 19(a), whe e he elec on densi y and empe a u e, as
well as he ion empe a u e, a e ob ained h ough diagnos ic
measu emen s and i ing. These da a se e as inpu s o he
plasma pa ame e s in ASCOT. The eloci y-space dis ibu-
ion o beam-ion losses a he FILD p obe loca ion, o i-
gina ing om NBI2R and simula ed by ASCOT, is shown
in igu e 19(b), wi h he o ange do ed line indica ing he
p obe de ec ion ange. In o de o alida e he consis ency
be ween expe imen al measu emen s and simula ion esul s,
he eloci y-space dis ibu ion o as -ion losses a he FILD
p obe ape u e, ob ained om ASCOT, was used as an inpu
o FILDSIM. Figu e 20 p esen s he compa ison o he
eloci y-space dis ibu ions be ween he discha ge #115236
simula ion and he expe imen al measu emen s. Bo h he sim-
ula ed and expe imen al beam-ion loss signals exhibi wo
pi ch angle egions. The los beam ions exhibi a small pi ch
angle o app oxima ely 45◦–50◦, wi h a gy o adius ange o
3–4 cm. Bo h simula ed and expe imen al da a show good
ag eemen in his ange. Howe e , disc epancies a e obse ed
be ween he simula ed and expe imen al esul s o beam ions
wi h la ge pi ch angles. The simula ed pi ch-angle ange is
app oxima ely 60◦–70◦, co esponding o a gy o adius ange
o 1.8 cm o 3.8 cm. In con as , he expe imen al meas-
u emen s e eal a pi ch-angle ange o abou 70◦–75◦and a
gy o adius ange o app oxima ely 3 cm–5 cm.
The e a e wo ac o s con ibu ing o he disc epancies
be ween he simula ed and expe imen al eloci y-space dis-
ibu ions o beam-ion losses a la ge pi ch angles. Fi s ly,
disc epancies a ise om e o s in aligning he expe imen al
beam-ion loss signals wi h he s ike-map de i ed om
FILDSIM, and inaccu acies in he magne ic ield a he FILD
p obe loca ion can also a ec he simula ed eloci y-space
dis ibu ion o beam-ion losses a he p obe. Secondly, he
simula ion esul s indica e ha he a e age ime o beam-
ion loss om he HFS o he p obe (τHFS_loss ∼6.6 ×10−4s)
is signi ican ly sho e han ha om he LFS o he p obe
(τLFS_loss ∼6.4×10−3s). The la e may cons i u e he p e-
dominan ac o . The beam ions o igina ing om he LFS
unde go a longe pe iod o slowing down be o e being los
8
Nucl. Fusion 65 (2025) 016026 S.S. Wang e al
Ma in R., McClemen s K.G., Sanchis L. and Zole nik S.
2019 J. Ins um. 14 C09015
[13] Sanchis L. e al 2019 Plasma Phys. Con ol. Fusion 61 014038
[14] Kon ula J., Koschinsky J.P., Äkäslompolo S. and
Ku ki-Suonio T. (W7-X Team) 2021 Plasma Phys. Con ol.
Fusion 63 035022
[15] Sanchis L. e al 2021 Nucl. Fusion 61 046006
[16] Si én P., Va je J., Weisen H. and Koskela T. (JET Con ibu o s)
2017 J. Ins um. 12 C09010
[17] Äkäslompolo S. e al 2019 J. Ins um. 14 C10012
[18] Sipilä S. e al 2021 Nucl. Fusion 61 086026
[19] Sä kimäki K. 2020 Nucl. Fusion 60 036002
[20] Ku ki-Suonio T. e al 2017 Plasma Phys. Con ol. Fusion
59 014013
[21] Si én P., Va je J., Äkäslompolo S., Asun a O., Gi oud C.,
Ku ki-Suonio T. and Weisen H. (JET Con ibu o s) 2018
Nucl. Fusion 58 016023
[22] Wu C.R. e al 2018 Re . Sci. Ins um. 89 10I144
[23] Jin Z., Chang J.F., Huang J., Wu C.R., Xu Z., Chu N., Lin S.Y.,
Hu L.Q., Isobe M. and Ogawa K. 2017 Fusion Eng. Des.
125 160
[24] Rod iguez-Ramos M. e al 2017 Plasma Phys. Con ol. Fusion
59 105009
[25] Asun a O., Go enius J., Budny R., Go elenko a M.,
Ta dini G., Ku ki-Suonio T., Salmi A. and Sipilä S. (The
ASDEX Upg ade Team and The JET EFDA Con ibu o s)
2015 Compu . Phys. Commun. 188 33–46
[26] Qu H. e al 2015 Plasma Sci. Technol. 17 985
[27] Liu H.Q. e al 2014 Re . Sci. Ins um. 85 11D405
[28] Yang L., Baonian W., Junyu Z., Qingsheng H., Yanqing J.,
Xiaoqi X., Xiao eng H. and Qing Z. 2010 Plasma Sci.
Technol. 12 284
[29] Li Y.Y. e al 2016 Re . Sci. Ins um. 87 11E501
[30] Huang J. e al 2020 Nucl. Fusion 60 016002
[31] Fu J. e al 2022 Plasma Phys. Con ol. Fusion 64 095006
[32] Gao W. e al 2011 J. Nucl. Ma e . 415 S391–4
[33] Liang L., Wei J. and Hu C. 2015 J. Nucl. Sci. Technol.
52 481–9
[34] Chen Y. 2011 Phys. Plasmas 18 062506
[35] Zhang L. e al 2011 Plasma Sci. Technol. 13 431
[36] Ollus P., Ake s R., Colling B., El-Ha oun H., Keeling D.,
Ku ki-Suonio T., Sha ma R., Snicke A. and Va je J. 2022
Plasma Phys. Con ol. Fusion 64 035014
[37] Suzuki S., Shi ai T., Nemo o M., Tobi a K., Kubo H., Sugie T.,
Sakasai A. and Kusama Y. 1998 Plasma Phys. Con ol.
Fusion 40 2097
[38] Kim J.Y., Rhee T., Kim J., Yoon S.W., Pa k B.H., Isobe M.,
Ogawa K. and Ko W.H. 2016 AIP Ad . 6105013
[39] Vianello N. e al 2020 Nucl. Fusion 60 016001
[40] Redl A. e al 2024 Nucl. Fusion 64 086064
15