Resea ch a icle
E alua ion o addi i e manu ac u ing o sand co es in e ms o he esul ing
su ace oughness
Ma ina Gaw ono 
a
a
,
*
, Pe Lichý
a
, I ana K oupo 
a
a
, Tom
a
s Obzina
a
, Ja osla Be
no
a
,
Isabel Nguyeno 
a
b
,V

acla Me a
a
, Jan Jezie ski
c
, Filip Radko ský
a
a
Depa men o Me allu gical Technologies, VSB-Technical Uni e si y o Os a a, Facul y o Ma e ials Science and Technology, 17. Lis opadu 2172/15, 708 00 Os a a,
Czech Republic
b
B embo Czech, S. R. o., Na Ro ince 875, 720 00 Os a a, Czech Republic
c
Depa men o Found y Enginee ing, Silesian Uni e si y o Technology, Facul y o Mechanical Enginee ing, Ul. Towa o a 7, 44-100 Gliwice, Poland
ARTICLE INFO
Keywo ds:
Moulding mix u es
Su ace p ope ies
3-D p in ing
Cas ing
ABSTRACT
Ob aining a good su ace finish on cas ing is challenging and depends on he dimension o he sand pa icles and
he p ocessing me hod o he mold. E ol ing mode n ends in mould and co e p oduc ion as a binde je ing
echnology is an op ion and i is mo e han desi able o e alua e and emo e any possible nega i e e ec s. The
aim o his s udy is o compa e he influence o u an no-bake echnology and 3D p in ing me hod on he su ace
quali y o co es and ca i ies o med in aluminium alloy cas ings. In addi ion o he sie e analysis and mechanical
p ope ies o he moulding mix u es, he oughness (Ra, Rz) o he co es and esul ing cas ing su aces o indi-
idual samples we e compa ed in his s udy.
1. In oduc ion
The su ace o he p oduc de e mines he final appea ance, he e o e
oughness can be conside ed as a ce ain eflec ion o su ace quali y.
One o he ac o s influencing he appea ance and quali y o he su ace is
undoub edly he influence o he co emix u e, bo h in e ms o hei
mechanical p ope ies and he quali y o he p ocessing o he mix u e
i sel . Binde sys ems o he p oduc ion o co es achie e di e en
cha ac e is ics bo h in p ima y p oduc ion (moulding o shoo ing he
mix u e) and in me al cas ing [1]. This may a ec he su ace p ope ies
o he cas ing om he co e side. The inne su ace o he cas ing exposed,
o example, o flowing liquids (wa e pumps, nozzles, e c.) is signifi-
can ly in ol ed in he beha iou o he liquid flowing in he cas ing
ca i ies, and hus hyd aulic losses can occu . The e o e, he aimis o
achie e he fines possible su ace. Co es o complex shape can be p o-
duced using an ino ganic binde sys em. These co es a e cha ac e ised by
su ficien s eng h and ab asion esis ance and he esul ing high quali y
and smoo h cas ing su ace. On he o he hand, he p oblems wi h high
empe a u e expansion o he co es and poo collapsibili y and so de e-
io a ed deco ing need o be conside ed [2,3,4]. O ganic binde s a e
o en used as a solu ion o such p oduc ion. In he s udy he au ho [5]
ecommends o use co es bound wi h esin o he RCS ( esin coa ed sand)
ype, which gua an ee a e y good su ace wi h a minimum o de ec s in
cas ings o g ey cas i on and nodula cas i on. Howe e , he p oblem
wi h his sys em is impai ed collapsibili y in case o alloys wi h lowe
mel ing empe a u es (aluminium o magnesium alloys), and hus low
co e hea ing o achie e imp o ed collapsibili y. In he au omo i e and
o he indus ies, Al–Si alloys, so-called silumines, a e used as uni e sal
ma e ials o he p oduc ion o complex cas ings (such as pump bodies,
ansmission cases, housings) because o hei good echnological and
ound y p ope ies as low densi y combined wi h su ficien s eng h [6,
7,8]. One o he mos commonly used alloys is hypoeu ec ic alloy A356
because o i s good cas abili y [9], excellen mechanical p ope ies [10,
11], co osion esis ance [12,13] and good dimensional s abili y and
s eng h a di e en empe a u es [14]. Wi h aim o imp o e collaps-
ibili y, o he au ho s epo he sui abili y o using binde sys ems based
on u an [15,16], polyu e hane esin [17,18] o alkaline phenolic esin
ha dened by CO
2
[19] in he aluminium ound y. Al hough he me-
chanical p ope ies a e lowe han ha ones o RCS bound mix u es,
moulds and co es bound wi h u an esin can be ha dened a oom
empe a u e wi h achie ing su ficien s eng hs. Also, hei e y good
collapsibili y a e he mal exposu e defines hei ad an ages o
shape-complex componen s wi h p ecise dimensions, especially in he
au omo i e indus y [20,21]. Fu he mo e, echnological, ecological and
* Co esponding au ho .
E-mail add ess: ma ina.gaw ono[email p o ec ed] (M. Gaw ono 
a).
Con en s lis s a ailable a ScienceDi ec
Heliyon
jou nal homepage: www.cell.com/heliyon
h ps://doi.o g/10.1016/j.heliyon.2022.e10751
Recei ed 28 June 2022; Recei ed in e ised o m 5 Sep embe 2022; Accep ed 22 Sep embe 2022
2405-8440/©2022 The Au ho s. Published by Else ie L d. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/).
Heliyon 8 (2022) e10751
economic aspec s o p oduc ion such as good sand eclama ion, lowe
in es men equi emen s o echnology (con en ional mixe s, non-me al
co e boxes), highe economics o co e p oduc ion (lowe acquisi ion cos s
pe mix u e, lowe ene gy consump ion due o he elimina ion o he
need o p ehea ing he co e boxes) [20,22,23], educ ion o ca bon
emissions achie ed by educing liquid me al consump ion h ough e fi-
cien mould design and minimal pos -p ocessing a e cas ing [24,25],
can ep esen a ce ain ope a ional ad an age and wi h a sui ably chosen
sand g anulome y a quali y su ace can be achie ed [21].
In addi ion o he con en ional p oduc ion me hods as no-bake
echnology, he e is now an e o o p oduce co es and moulds using
new me hods. One o hem is he inc easingly widesp ead and popula
he 3D sand p in (3DSP) o moulds and co es using binde je ing
echnology. The echnological p inciples o he p oduc ion o co es and
moulds using 3D sand p in ha e al eady been s udied by many au ho s
[22,26,27,28,29], when hey ag ee ha he use o such sys ems is
possible in p ac ice. 3DSP echnology enables a a ie y o p oduc ion
wi h he possibili y o p oduce high-p ecision and shape-complex pa s
wi h accu acy o 0.5 mm o an e o o less han 0.3%, and hus find
applica ion in p o o ype p oduc ion, au omo i e and in he design o
inno a i e p oduc ion p ocesses in he field o me al oams [30], o
expe imen al p oduc ion o ib en o ced shell moulds o con olled and
accele a ed cooling o aluminium alloy cas ings [31,32]. In o he wo ds,
he p oduc ion o co e boxes, and he associa ed dimensioning, and he
p oduc ion o me al mould [22,26] all away.
Di e en ypes o 3D sand p in e s use di e en chemical sand cu ing
sys ems. No-Bake sys ems based on u an and phenolic esin, o ino ganic
alkaline silica es a e mos commonly used o he p oduc ion o 3D sand
p in ed co es. I is also possible o use a sys em o phenolic esin bonded
wi h hea . Ve y fine ac ions o silica sand o syn he ic ce amic sand a e
o en used as a sand [26,33]. Unlike o he ypes o 3D sand p in e s using
one-componen binde sys em ha equi e mo e binde , ExOne p in e s
use a wo-componen binde sys em ha is e y simila o con en ional
u an no-bake echnology [34].
3D p in ing o sand moulds is, in p inciple e y simila o he well-
known 2D p in ing on pape using an inkje desk op p in e . Figu e 1
schema ically shows he binde je ing echnology p ocess. A job box is
placed on he p in e pla o m in which a ecoa e e enly sp eads
0.28–0.5 mm hin laye o powde pa icles, such as silica sand o ce amic
pa icles, mixed wi h a ca alys . D ople s o liquid binde a e hen applied
o his laye using a p in head mo ing a he speed o 60–80 mm s
1
. The
binde is only dispensed a ce ain posi ions and on con ac wi h he
ca alys , he chemical cu ing p ocess s a s a ha poin . The emaining
places in he powde bed emain filled wi h loose sand only. The pla o m
is hem mo ed downwa ds a p edefined dis ance, he ecoa e d ies he
bed and applies a new laye o powde wi h ca alys o e he al eady
cu ed bed. As he ecoa e mo es, he new laye o sand is also gen ly
comp essed. The build speed achie ed is usually 60–85 l h
1
. This p o-
cess is epea ed laye by laye un il he objec is comple ed o he
designed 3D specifica ion. The finished pa is he emo ed om he
powde bed [23,24,35,36].
Se e al di e en au ho s [28,37] obse ed poo e su ace quali y
o cas ings a e using 3DSP moulds o co es, ei he due o highe
su ace oughnesso hep esenceo mino su acede ec s.Themain
eason appea s o be he excess binde wi h which 3D sand p in e s
gene ally wo k. As can be seen om Figu e 2, he powde con ains
pa icles o di e en sizes and shapes (g anulome ic composi ion
o he basic sand), which he binde mus be able o coa a he
desi ed laye dep h. The e o e, he applica ion o he liquid
binde mus be p ecise, he indi idual d ople s mus be e y fine, and
hey mus only flow in a e ical di ec ion o he su ace o
he powde , so ha he binde does no un o e he powde .
The e is also a need o su ficien ha dening o he binde be o e
applying an addi ional laye o sand ( ecoa ing p ocess) o p e en
he sp ead o new sand on s ill mois pa icles [36,38,39]. Also,
he newly applied laye mus be comp essed wi h a su ficien ly
la ge o ce o educe he amoun o po es in he mix u e [38]. The
po osi y o he mix u es is impo an in p edic ing he pe meabili y o
he mix u e, while he po e size a ec s he pene a ion o he me al
in o he ound y mould and also di ec ly a ec s he esul ing ough-
ness o he cas ing o he occu ence o ound y de ec s. Thus, he
fine and mo e angula he g ains, he highe he po osi y o he
mix u e [40,41].
The binde dosage is gene ally abo e he dosage le el o no-bake
co e p epa a ion me hods (binde con en <1%). The e ha e been
se e al s udies [35,37,42] in es iga ing he e ec o binde dosage and
sa u a ion on mechanical p ope ies in e ms o s eng h, pe meabili y
and dimensional compliance. This au ho s a es [43] ha he 3D sand
p in echnology used by hem wo ks a a binde con en o 1.6–1.8%
( o u an), ano he sou ce epo s a u an dosage o 1–3% [39]and
ano he au ho [37] lis s con en s app oaching up o 8% o a binde .
Gene ally he mechanical s eng h inc eases wi h inc easing amoun o
binde . Howe e , he highe binde con en leads o a de e io a ion o
he mould collapsibili y a e pou ing, as he s eng h o he mould
emains e en a e hea exposu e [44]. I has been also ound ha a
highe binde dosing a e in 3DSP may cause impai ed mould o co e
sh inkage and pe meabili y, highe gas de elopmen , esul ing in
de e io a ion o su ace quali y [28,38] o o ma ion o gas de ec s
[35]. On he o he hand, he de e io a ed oughness o he su ace can
also be caused by insu ficien comp ession o he newly applied laye o
he sand mix u e and ca alys , which is usually caused by he high speed
o sp eading he new laye by he ecoa e [26,29,38]. Since in he
3DSP echnology he e is also a equi emen o high p oduc ion speed
o he piece, his can lead o une en compac ion o he sand in he bed
and consequen ly o he educ ion o he densi y o applied sand due o
highe po osi y [37].
Ob aining a quali y cas su ace wi hou machining is challenging in
he case o wa e pump cas ing o in au omo i e indus y. The pa s a e
mainly manu ac u ed om A356 alloy in combina ion wi h complex
co es wi h o ganic binde s, and addi i e manu ac u ing echnology o
co es is also used. This s udy ocuses on he esul ing inne su ace
oughness o A356 aluminium alloy cas ing whe e di e en co es we e
used o p oduc ion. Fu hemo e, he mechanical p ope ies o co es
made by u an no-bake and 3D sand p in echnology will be in es i-
ga ed, as well as he influence o p oduc ion pa ame e s on he su ace
quali y o he cas ings.
Figu e 1. Binde je ing echnology schema.
M. Gaw ono 
a e al. Heliyon 8 (2022) e10751
2
2. Ma e ials and me hods
A schema ic o he es sample p oduc ion me hodology and he o de
o he indi idual measu emen g oups can be seen in Figu e 3.
2.1. Sand cha ac e isa ion
Silica sand om a locali y Biala G
o a in Poland, wi h wo di e en
mean g ain sizes, ma ked by he supplie as BG 21 and BG 27, was used
o he p oduc ion o u an no-bake co es. Gua an eed mean g ain size o
sand wi h he designa ion BG 21 was d
50
0.21 mm 0.02 mm, he second
BG 27 was d
50
0.27mm 0.02 mm mainly used o he echnology wi h
o ganic esins. Mean g ain size 0.21 mm was chosen despi e he ac ha
i was he sand a he used o g een sand mix u es, as i ep esen ed he
size in e media e s ep be ween he g anulome y o u an echnology
and o 3Dsand p in echnology.
A de ailed o e iew o he g anulome ic composi ion based on
clay wash analysis acco ding o ASTM C117-13 and sie e analysis
acco ding o ASTM E11 we e pe o med. The mass o each sample was
50 g. Labo a o y sand sie e (LPzE-2e, Mul ise w-Mo ek, Poland)
equipped wi h complian sie es was used. Shaking ime was se on 10
min, ib a ion 50 kHz, ampli ude was adjus able om 0 o 2.5 mm.
G ain fineness numbe acco ding o Ame ican Found y Socie y (AFS)
was de e mined om a pe cen age o sand ac ion cap u ed on
each sie e using a mul iplie . Cumula i e cu es o g anula i y we e
e alua ed and c i e ia o he g anulome ic composi ion as alues o
d
25
,d
50
and d
75
and he Homogenei y deg ee S we e calcula ed
(equa ion 1).
S¼ðd75 =d25Þ⋅100 ð%Þ(1)
whe e d
25
and d
75
o mesh diame e s co esponding o 25 and 75 % o
he o al sand mass (a e washing o he pa icles less han 0.02 mm).
The C i e ion o he g ain-size dis ibu ion p obabili y log W (equa-
ion 2) was also de e mined om he sie e analysis. Compa ed o he
e alua ion c i e ion o he g ain fineness numbe acco ding o AFS, his
analysis p o ides an o e iew o so ing le el (mono ac ional o poly-
ac ional sand).
Ni¼ðmimpÞ⋅100 ð%Þ(2a)
logW ¼100 ⋅logð100ÞXNilogðNiÞ(2b)
whe e N
i
o ac ion cap u ed on he sie e (%), m
i
o he mass o sample
sand on sie e (g) and m
p
o he mass o sample sand a e washing o
clay pa icles below 0.02 mm (g).
Ini ial da a o pa icle size dis ibu ion o sand used o 3Dsand p in
esul ed om sie e analysis was p o ided by he manu ac u e and was
u he used o de e mine he log W, S and cumula i e cu e o
g anula i y.
Figu e 2. D opping o binde in o a powde .
Figu e 3. A schema ic o used me hodology.
M. Gaw ono 
a e al. Heliyon 8 (2022) e10751
3
2.2. Co e p oduc ion echnology pa ame e s
Fu an no-bake co e p oduc ion simula ed usual p oduc ion dosing o
1 % o binde pe sand mass and 35 % o he ca alys pe esin mass (0.35
% o he ca alys pe sand mass). The mix u e was mixed on a labo a o y
paddle mixe . Fi s ly, he sand-ca alys mix u e was mixed o 1 min o
cu ing homogenei y, hen a e adding esin binde o he mix u e, he
mix u e was mixed o ano he 1 min. Then he mix u e was compac ed
using a labo a o y sand amme (LUA-2e/Z, Mul ise w Mo ek, Poland) in
o de o main ain cons an condi ions o co e p oduc ion. The mixed
mix u e was pou ed in o 22.5 mm 22.5 mm x 170 mm s eel mul i-
ca i y co e box and subsequen ly compac ed by h ee imes hi ing o
he s anda d sand amme wi h cons an impac ene gy o 3.3 J Figu e 4
shows samples made by u an no-bake echnology and 3D sand p in .
3DSP co e samples we e p oduced by local comme cial supplie using
he binde je ing echnology on he ExOne's 3D sand p in e , which was
used in s udies by o he au ho s [22,27,30]. The size o p in ed sample
co es was 22.5 mm 22.5 mm x 170 mm (Figu e 4). Silica sand wi h d
50
0.14 mm (he ea e e e ed o as 3DSP sand) was used. Fu an esin
supplied by ExOne was used, which was ha dened using a ca alys based
on pa a oluenesulphonic acid. Acco ding o he pa ame e s ecom-
mended by he manu ac u e o he 3D sand p in e he ca alys dosage
was 0.21 % pe sand mass. The dosage o u an esin was no p ecisely
defined by he supplie , as well as he exac speed o p in ing head and
ecoa e . The inpu amoun o u an esin in he used es co es was
he e o e e e sely de e mined in he labo a o y by loss o igni ion es
and i was a ound 1.25% 0.02 % pe sand mass. Basic pa ame e s o
binde s and ca alys s used a e gi en in he able below (Tables 1a and 1b)
2.3. Expe imen al p ocedu e and measu ing
T ans e se s eng h
The mechanical s eng h o he co es was cha ac e ised by a mea-
su emen known as 3-poin bending s eng h and was ca ied ou on he
de ice o es ing he s eng h o moulding sand (LRu-2e, Mul ise w
Mo ek, Poland) a oom empe a u e. Cons an load a e was 0.1 MPa
s
1
. Th ee measu emen s we e made o e e y co e ype. The size o each
specimen was 22.5 mm 22.5 mm x 170 mm a we e aken o mea-
su emen 24 h a e cu ing (Figu e 5).
Ab asion esis ance
Ab asion esis ance was e alua ed on samples aken a e mechanical
s eng h measu ing and was ca ied ou on a labo a o y de ice wi h a
o a ing wi e baske . The sample was placed in a pe o a ed baske wi h a
diame e o 180 mm, and i was subsequen ly ab aded o 60 s a 57
o a ions pe minu e. The esul ing dec ease in mass wi h a
compa a i eness o he o iginal mass o he measu ed sample gi es
ab asion esis ance pe cen ages.
Loss on igni ion
The c ushed co es we e d ied ou in o labo a o y d ie a 105 C o
emo e all mois u e. Then app oxima ely 3 g samples in annealed
ce amic c ucibles we e placed in mu fle u nace hea ed a 900 C o 2 h.
Loss on igni ion di ec ly p opo ional o mass loss caused by ola iliza-
ion o o ganics, in his case o ganic binde s, and was de e mined by
means o he di e ence in mass o he p e-d ied samples be o e and a e
he igni ion. Measu ing was pe o med on he labo a o y scale wi h
sensi i i y o 0.0001 g.
Rela i e po osi y o sand co e
The ela i e po osi y o co es was de e mined om he ob ained
sample densi ies and Eq. (3).
m¼ð
ρ
1
ρ
2
ρ
1
Þ⋅100 ð%Þ(3)
whe e
ρ
1
o he specific mass o sand (kg
.
m
3
) and
ρ
2
o densi y o
compac ed mix u es (kg
.
m
3
).
2.4. P epa a ion o es cas ings
The p oduc ion o es cas ings ook place unde labo a o y condi-
ions, see Figu e 6 3D model o cas ing. The a ea o he cas ing ma ked as
“Top”was c ea ed by he lowe pa o he co e and ep esen s he uppe
Figu e 4. Fu an no-bake es samples: (a) silica sand BG 21, (b) silica sand BG
27, (c) silica sand used o 3D sand p in .
Table 1a. Basic pa ame e s o binde s.
Componen No-bake 3D sand p in
Binde addi ion [w . %] 1.0 1.25
Fu u ylalcohol [%] Min. 75 Min. 70
Phenol [%] <1<1
pH [-] 6–86–8
Figu e 5. Schema ic o 3D sand p in ed es specimen du ing 3-poin bending
s eng h measu emen (Inspi ed by [38]).
Table 1b. Basic pa ame e s o ca alys s.
Componen No-bake 3D sand p in
Dosing o he sand con en [w . %] 0.35 0.21
Pa a oluenesuphonic acid [%] <65 30–70
Sulphu ic acid [%] <2 Max. 5
pH [-] <1<1
M. Gaw ono 
a e al. Heliyon 8 (2022) e10751
4
inne su ace o he cas ing. The a ea ma ked as “Bo om”was c ea ed by
he uppe pa o he co e and ep esen s he bo om inne su ace o he
cas ing. The mould was made using a no-bake geopolyme -based binde
echnology wi h he ade name GEOPOL. Tes ed co es wi h an o iginal
leng h o 170 mm ha e been sho ened o 70 mm acco ding o he cas ing
design. The A356 aluminium alloy in a pu e s a e was chosen o esea ch
pu poses, wi hou p io mel ing, wi h a gua an eed chemical composi-
ion. The cas ing empe a u e was 720 C wi h a cas ing ime o 4 s.
2.5. De e mina ion o su ace oughness o co es and cas ings
The e alua ion o oughness was ca ied ou on digi al mic oscope
(OM, VHX 6000, Keyence, Japan) wi h a zoom ange o 20x up o 2000x.
E alua ion o oughness was ca ied ou on pho og aphs ha we e made
up o pa ial figu es using he "s i ching" unc ion o ma ch he leng h o
measu ing o oughness and e alua ed acco ding o he ISO 4288 s an-
da ds. E alua ed leng h alues ha e always been aken om h ee
“poin s”, ed ec angles, see Figu e 7 o sand co e and cas ing su aces.
The di e ence in measu ing he oughness o sand co es, and by a
g ainy ma e ials, compa ed o solids (cas ings) was he need o se he
maximum (up) and minimum dep h (down) o he sha pness, which was
se o 500
μ
m.
The co e su ace oughness and in e nal su aces ( op and
bo om) o he cas ing c ea ed by copying he co e su ace we e
measu ed. The cas ings we e le in ough s a e wi hou any machi-
ning,Figu e 6 shows hese a eas wi h he e ms "Uppe " and "Bo om"
which will be e alua ed sepa a ely because he lowe su aces o he
cas ing due o g a i y achie e be e su ace quali y han he uppe
su aces.
Figu e 6. (a) Cas ing d awing. (b) 3D model o cas ing wi h co es ob ained om MAGMA simula ion p og am.
Figu e 7. De e mina ion o measu ed a eas: (a) on he sand co e, (b) on he
cas ing su ace.
M. Gaw ono 
a e al. Heliyon 8 (2022) e10751
5

3. Resul s
3.1. Basic pa ame e s o he used sands
The g anulome ic composi ion o all he used sands was de e mined
using he sie e analysis. F om he ob ained alues he AFS numbe
oge he wi h C i e ion o he g ain-size dis ibu ion p obabili y log W
was de e mined, and cumula i e cu es o g anula i y o indi idual
sands we e compiled (Figu e 8). The alues o d
25
,d
50
and d
75
we e
subsequen ly ead om hem and he homogenei y deg ee S was de e -
mined. The pa icle size dis ibu ion is included in Table 2 oge he wi h
he o he c i e ia.
The measu ed esul s indica e ha he sand used o 3DSP showed he
log W alue by 35 % lowe han he BG 27 sand and by 32 % lowe han
he BG 21 sand, and hus he log W alue o he 3DSP sand was close o
0. I can he e o e be concluded ha he sand supplied o he needs o
3DSP sand was e y sha ply so ed compa ed o he sands used o u an
no-bake co e p oduc ion, mainly con aining g ains o he same size, and
app oached a he mono ac ion wi h i s p ope ies. This conclusion is
consis en wi h he se cumula i e cu es o g anula i y, whe e i is
e iden ha he sand cumula i e cu e o 3DSP has a s eepe inc ease.
On he con a y, he cu es o BG 21 and BG 27 sands ha e mainly
ac ions wi h di e en g ain sizes, especially in case o BG 27 sand,
which has he highes log W alue and is mo e poly ac ional.
The alues o mean g ain size d
50
ead om he cumula i e g ain
cu es o each sand co espond o he supplied sand pa ame e s.
Since na u al silica sands we e used, i was also possible o e alua e
he homogenei y deg ee S, o which i is alid ha he sand is he mo e
uni o m he close he S is o alue 1 o 100 % espec i ely, and ice
e sa. The mos homogeneous sand can again be conside ed he 3DSP
sand due o he high deg ee o so ing and, on he con a y, he leas
homogeneous he BG 27 sand.
3.2. Mechanical p ope ies o es ed mix u es
T ans e se s eng h, ab asion esis ance, loss on igni ion and calcu-
la ed ela i e po osi y can be seen in Table 3 below. I ans e se s eng h
is conside ed as he main ini ial pa ame e , ce ain conclusions can be
d awn. Wi h inc easing s eng h, ab asion esis ance dec eases di ec
p opo ionally, when he highes ab asion was measu ed o he 3DSP
co es wi h a alue o 16.31 % e sus BG 21 wi h ab asion o 6.58 % and
BG 27 wi h ab asion o 9.67 %. The fine he g anulome ic composi ion
o he sand and he lowe he d
50
o he mono ac ional cha ac e o sand,
he highe he s eng h was achie ed, bu his was alid only o he
con en ional me hod o compac ing (BG 21 and BG 27). This s a emen
was no alid in he case o co es p oduced by he 3Dsand p in me hod
Figu e 8. Compa ison o cumula i e cu es o g anula i y o all es ed sands.
Table 2. Pa icle size dis ibu ion and basic pa ame e s o es sand specimens.
Mesh size [mm] Silica sand BG 21
[ e ained %]
Silica sand BG 27
[ e ained %]
3D sand p in
[ e ained %]
1.4 0 0 0
1 000
0.71 0 0 0
0.5 0.06 0.48 0
0.355 3.12 11.29 0
0.25 19.65 32.01 0
0.18 32.35 29.56 3.50
0.125 36.64 21.27 60.70
0.09 6.37 4.17 29.10
0.063 1.31 0.73 6.00
Pan 0.34 0.17 0.70
AFS Numbe 69 60 97
d
25
[mm] 0.24 0.30 0.16
Mean g ain size d
50
[mm]
0.19 0.23 0.14
d
75
[mm] 0.15 0.18 0.11
Log W [-] 61.33 63.97 41.49
Homogenei y
deg ee S [%]
62.50 60.00 68.75
Table 3. Measu ed pa ame e s o es ed sand co es.
Measu emen s pe o med BG 21 sand
co e
BG 27 sand
co e
3D sand
p in
T ans e se s eng h [MPa] 3.45 2.67 0.92
Bulk densi y o compac ed co e
[g⋅cm
3
]
1.50 1.51 1.37
Ab asion [%] 6.58 9.09 16.31
Loss on Igni ion [%] 1.14 1.12 1.20
Rela i e po osi y [%] 43.18 42.80 48.11
M. Gaw ono 
a e al. Heliyon 8 (2022) e10751
6
due o highe ela i e po osi y o he co e, al hough he defined d
50
o he
sand was lowe and sand is mo e mono ac ional. The lowe co e s eng h
i sel and high ab asion esis ance could be he esul o a gi en p o-
duc ion echnology.
3DSP co es showed he highes loss on igni ion o 1.20 %, which is
by 5–7 % highe han ha one o BG 21 and BG 27 co es o which he
same dosing o indi idual binde sys em componen s was used. This
esul co esponds o he expec ed and subsequen ly wi h he e e se
me hod de e mined highe dosage o u an esin o he 3Dsand p in ed
co es.
The ela i e po osi y alue o 48.11 % o he 3DSP co es compa ed
wi h he BG 21 and BG 27 co es was by 11.4 % espec i ely 12.4 %
highe . A de ailed iew o he in e -g ain po es and binde b idges on he
ac u e su aces o he co es a e ab asion measu emen is shown in
Figu e 9. The sand g ains we e pu posely illumina ed o g ea e isi-
bili y o he binde b idges. Pho os we e aken a 500x magnifica ion on a
digi al mic oscope Keyence VHX 6000. Visually, he g ains o all he
samples can be assessed as angula shape. The con ac s be ween he
g ains we e he e o e no poin con ac s, bu we e mo e like su ace
con ac o di e en shapes. The po osi y o he mix u e does no depend
on he diame e o he g ains bu on he s uc u e o he mix u e, i.e. he
maximum possible numbe o con ac s wi h neighbou ing g ains a a
gi en deg ee o compac ion, which was influenced no only by he shape
o he g ains bu also by he composi ion o he mix u e (mono ac ion
and poly ac ion). The dis ibu ion o po es be ween he g ains co e-
sponded wi h he calcula ed alues o he ela i e po osi y o he co es. In
poly ac ion mix u es, unlike mono ac ion, he addi ion o fines be-
ween he coa se ac ions significan ly educes he absolu e po e size o
he mix u e, which helps o p e en me al pene a ion in o he mould.
The fine ac ions ake he place o he coa se ac ions in he in e -
g anula spaces. The highe alue o 3DSP co es is he esul o fine
mono ac ional sand used. Mono ac ion sands, on he o he hand, ha e a
lowe o al specific su ace a ea and we e he e o e able o achie e
highe s eng hs due o he o ma ion o s onge bonding b idges when
he binde was homogeneously dis ibu ed and compac ed. Po osi y and
insu ficien s engh can hen be suppo ed by une enly applied indi id-
ual laye s o sand-ca alys mix u e du ing he ecoa ing p ocess o oo
high ecoa e speed, whe e insu ficien comp ession o he applied sand
laye may occu . In bo h cases he esul would be a low deg ee o
comp ession o he mix u e and hus a g ea e amoun o po es be ween
indi idual sand g ains, which co esponds o he measu ed low bulk
densi y o he compac ed 3DSP co e.
The esul s indica e he e was no significan di e ence in loss on
igni ion, bulk densi y o ela i e po osi y o BG 21 and BG 27 sand co es,
and wi h ega d o he same dosage o he binde and he ca alys and he
p oduc ion echnology i was no e en expec ed.
3.3. De e mina ion o oughness o he co e su ace
The oughness o he co es can be seen om Figu e 10. 3DSP co es
show he lowes su ace oughness as expec ed as hey we e composed o
sand g ains wi h he fines d
50
and a e y sha ply so ed na u e o
g anulome ic composi ion compa ed o u an no-bakes co es. The di -
e ence in su ace oughness be ween he 3DSP and BG 21 co es was o
Ra ¼13.67
μ
m and Rz ¼61.68
μ
m. The di e ence in oughness be ween
3DSP and BG 27 co es was o Ra ¼16.42
μ
m and Rz ¼73.14
μ
m. The
la ges measu ed di e ences in oughness co espond o di e ences be-
ween he highes and lowes d
50
o he sands used, i.e. o 3DSP and BG
27 samples. In con as , BG 21 and BG 27 co es do no show such sig-
nifican di e ences in oughness as migh be expec ed gi en he di e en
sizes o he mean g ain size d
50
. The sands used o hese co es did no
di e ha much om each o he in he dis ibu ion o g ain ac ions on
he sie es. The o al measu ed di e ences in oughness be ween BG 21
and BG 27 co es we e o Ra ¼2.75
μ
m and Rz ¼11.46
μ
m. Consis en
su ace oughnesses we e measu ed on all co e su aces.
3.4. De e mina ion o he cas ing oughness
Du ing he filling o he mould ca i y wi h liquid me al, he flow
causes all he impu i ies and inclusions o floa o he op su ace o he
Figu e 9. De ail o binde b idges ( ed a ows) and po es be ween g ains (g een a ows).
Figu e 10. Roughness Ra and Rz o e alua ed co es.
M. Gaw ono 
a e al. Heliyon 8 (2022) e10751
7
cas ing o o be apped on he bo om su ace o he co e. This leads o a
highe su ace quali y on he lowe su aces han on he uppe su aces
wi h deposi ed impu i ies. The same is also applicable in he case o he
inne ca i ies o cas ings o med by co es wi h he uppe inne su ace
(see Figu e 6) impu i ies and an inc eased su ace oughness by loose
g ains om he co e mix u e can be ound.
The oughness o he cas ings a e shown in Figu e 11, whe e he
designa ion 3DSP was o co es made by 3D sand p in ing and BG 21, BG
27 o u an no-bake co es. Pic u es show he oughness o a ea ma ked
as “bo om”, made by uppe pa o he co e. In he pho os wi h he
highes magnifica ion (500x), la ge ca i ies we e isible. These we e no
su ace po osi y o gas de ec s o he cas ing, bu he imp in o he in-
di idual g ains o he sand a high dep h image. A fi s glance, he di -
e ence be ween cas ing a eas copying 3DSP co es was no iceable.
Figu e 12 shows, ha in he case o he “bo om”inne side o he cas ing,
he oughness di e ence be ween he 3DSP co e and he BG 21 co e was
Ra ¼1.85
μ
m, Rz ¼6.85
μ
m. The e was a di e ence o Ra ¼1.48
μ
m and
Rz ¼5.36
μ
m be ween he 3DSP and he BG 27 one. Fo he u an no-
bake ( he BG 21 and he BG 27), he di e ence was Ra ¼0.37
μ
m and
Rz ¼1,49
μ
m.
The “ op”inne su ace o he cas ing a ea o med om he co e side
shows he same end, namely inc eased su ace oughness in he case o
3DSP co es. The di e ences we e as ollows: 3DSP and BG 21 he di -
e ence was Ra ¼1.8
μ
m and Rz ¼8.12
μ
m, 3DSP and BG 27 Ra ¼2.32
μ
m and Rz ¼9.94
μ
m. The di e en ia ion be ween BG 21 and BG 27 was
Ra ¼0.52
μ
m and Rz ¼1.82
μ
m only and i can be concluded ha he
g ain size in his case does no ha e a significan influence on su ace
quali y.
By a e aging he alues o oughness o he op and bo om inne pa
o he cas ing, an o e all iew on he achie ed su ace quali y o he
p oduc was ob ained (see Figu e 13). Ob ained esul s o he cas ing
su aces (a e age Top and Bo om alues) we e lowe when using co es
p oduced by he 3D p in ing me hod. In he case o a eas ma ked 3DSP s
BG 21 he alue was dec eased by Ra ¼1.83
μ
m and Rz ¼7.49
μ
m. The
di e ences be ween he a ea o 3DSP and BG 27 we e almos iden ical
( he dec ease Ra ¼1.9
μ
m and he dec ease Rz ¼7.65
μ
m). The esul ing
alues we e app oxima ely he same o a eas whe e BG 21 and BG 27
co es we e used, espec i ely negligible di e ences Ra ¼0.007
μ
m and
Rz ¼0.16
μ
m. Thus, he achie ed esul s indica e a sligh de e io a ion o
he su ace quali y o he a eas whe e he co es p oduced by he 3D
p in ing me hod we e used. I e alua ed he a e age su ace quali y be-
ween BG 21 and BG 27 a eas, i can be conside ed ha he size o he
g ains o he used sands does no cause a significan dec ease o su ace
quali y in o al.
4. Discussion
Bo h, he oughness o he co e su aces and he esul ing oughness
o he su aces o cas ing ca i ies o med om co es we e compa ed in
his s udy. The esul s o his measu emen co espond o he measu ed
mechanical p ope ies o u an no-bake co es and he 3DSP echnology.
Co es p oduced by he 3DSP echnology compa ed o he o he es ed
sands, BG 21 and BG 27, we e made using e y homogenous sand o high
fineness, wi h sha ply so ed ac ions and an o e all g anulome ic
composi ion app oaching a he o mono ac ion, as confi med by he
lowes measu ed log W alue o 41.49 and he highes homogenei y
deg ee S o 68.75 %. As expec ed, he 3D sand p in ed co es i sel show
significan ly lowe su ace oughness. In con as , he esul ing su ace o
he cas ing ca i ies showed he highes oughness and o e all poo e
quali y o he cas su ace. The o al a e age measu ed di e ence o
cas ing ca i y su aces o med om co es was o 3D sand p in Ra by
31.6 % highe compa ed o he BG 21 and by 33.2 % espec i ely
compa ed o BG 27. Rz oughness was by 28.99 % highe in case o
3Dsand p in compa ed o he BG 21 and by 29.79 % compa ed o he BG
27. The di e ences in oughness be ween he BG 21 and he BG 27 co es
we e Ra 1.22 % and Rz 0.62 % o cas ing ca i y su aces and he e o e
hose we e almos negligible. None o he su aces o he made cas ings
showed su ace de ec s o ca i ies o a gas na u e.
Con a y o expec a ions, he esul s ob ained o he oughness o he
in e nal su aces o samples BG 21 and BG 27 we e di e en . Howe e ,
he measu ed esul s show a oughe in e nal su ace o he cas ing in he
case o BG 21 co es, al hough he o e all di e ence in oughness Ra and
Rz was only minimal. F om he esul s, i can be concluded ha he
o e all homogenei y and he dis ibu ion o he indi idual sand ac ions
on he sie es o bo h samples ha e a g ea influence on he esul ing
su ace quali y o he cas ing. Despi e he ac ha he BG 27 co es we e
composed o sand wi h a coa se d
50
(0.23 mm) han BG 21 (0.19 mm),
he sand in he case o BG 27 was mo e he e ogeneous in composi ion
wi h a cumula i e cu e o g anula i y showing a mo e poly ac ional
cha ac e (Figu e 8) han was he case o BG 21 sand, which was mo e
homogeneous wi h a g ea e olume o g ains in se e al indi idual
ac ions. Also, he calcula ed ela i e po osi y o he co e is highe in he
case o BG 21 by 0.89 % compa ed o BG 27. Thus, i can be seen ha he
sands ha ing a a he he e ogeneous and poly ac ional cha ac e , and
hus con aining bo h coa se and e y fine pa icles in simila amoun s,
con ibu e o he esul ing smoo hness o he cas su ace. This was
caused by he abili y o he fine pa icles o fill he in e g anula spaces
be ween he coa se sand g ains, making i di ficul o liquid me al o
pene a e and copy he g ain shape and po es. This could be seen om
Figu e 11 ( he highes magnifica ion), whe e he smoo hes su ace wi h
ewe deep sand g ain imp essions was obse ed in BG 27 samples.
Wo se cas ing su ace quali y in case o 3DSP co es was also indica ed
by he measu ed mechanical p ope ies o he co es, whe e he 3DSP
co es showed ex emely low s eng h ( ans e se s eng h) compa ed o
he s eng h decla ed by he manu ac u e (~4 MPa) and namely only
0.92 MPa on a e age compa ed o 3.45 MPa o he BG 21 and o 2.67
MPa o he BG 27. Then he esul s o o he es ed p ope ies co espond
wi h hese s eng hs, when a e y low bulk densi y was measu ed o he
3DSP co es, as well as high ab asion, ela i e po osi y and loss on igni ion
esul ing in dis u bing o su ace laye o he co e du ing pou ing o
liquid me al a de e io a ion o quali y o he cas su ace.
The app oxima e used u an con en in he 3DSP co es was de e -
mined e ospec i ely om he loss on igni ion. The obse ed u an
dosage was highe han he con en ional u an no-bake me hod o co e
p oduc ion, bu no as high as commonly epo ed. The e ec o he
dosage o u an binde and he se ing o op imum p in ing pa ame e s
was also shown in a s udy by ano he au ho [36], whe e when he binde
con en was educed du ing 3DSP, he pieces eco ded he bes su ace
quali y in e ms o pi ing and gas po osi y, bu he co es p oduced in his
way achie ed low s eng hs. Possible explana ions o he insu ficien
mechanical p ope ies o he co es and he high oughness and low
quali y o he cas su ace can be se e al, o a combina ion he eo as
ollows:
Compa ed o he dosing o u an gene ally gi en in he li e a u e, he
de e mined dosage o u an in he case o hese specific 3DSP co es
could be in eal p ocess lowe han equi ed. This may be he eason
o he lowe binde capabili ies o he mix u e and hus o he lowe
co e s eng hs and ela ed p ope ies. In p inciple, he binde mus be
able o ha den be ween he indi idual laye s o he applied mix u e in
such a way as o p e en he new laye om sp eading ou on s ill
mois pa icles. This usually leads o he use o excess binde o 3D
sand p in ing.
The use o lowe quali y u an esin wi h lowe u u yl alcohol
con en
The use o e y fine and sha ply so ed sand o e y low he e oge-
nei y. This can lead o a highe numbe o po es in he mix u e and
highe equi emen s o su ficien dosage o binde , he lack o which
can lead o educed s eng h and his a de e io a ed cas ing su ace.
Too high p in speed and he e o e ecoa e speed oo, causing un-
e en applica ion o a new laye o sand wi h a ca alys and insu ficien
comp ession o he applied laye . The esul was a low compac ion o
M. Gaw ono 
a e al. Heliyon 8 (2022) e10751
8
he mix u e and he o ma ion o ee in e g ain spaces and po es in
he applied laye o sand, which can be u he suppo ed by he
choice o sha ply so ed sand wi h li le o no ep esen a ion o fine
ac ions o he sand.
Too high p in speed and he e o e une en dosing o u an in he o m
o small d ople s, which should be e enly applied by he p in head.
Conside ing he possible la ge a ea o po es be ween he indi idual
g ains o he moulding mix u e, as well as be ween he indi idual
laye s, he e was also he isk o en e ing o a pa o he binde in o
he lowe laye and he o ma ion o an impe ec ly ha dened uppe
laye .
5. Conclusions
The co es es ed we e p oduced using u an no-bake echnology and
3D sand p in ing. The g anulome ic composi ion o he basic sand was
e alua ed as well as he mechanical p ope ies o he co es. The ob ained
co es we e used o he p oduc ion o cas ings o de e mine he esul ing
su ace oughness in he cas s a e. The main conclusions a e as ollows:
Co es p oduced by3D sand p in ha e low su ace oughness as a
esul o a fine mean g ain size d
50
compa ed o con en ional u an
no-bake echnology.
3D sand p in ed co es achie e significan ly lowe s eng h (0.92 MPa)
han u an no-bake co es (3.45 MPa), 2.5 imes highe ab asion, 5–7
% highe loss on igni ion and 12.4 % highe ela i e po osi y han
con en ional u an no-bake co es.
The esul ing Ra oughness o he cas su ace in he case o 3D
p in ed co es was 33.2 % highe han in he case o u an no-bake
echnology as a esul o low mechanical p ope ies o p in ed co es.
No significan di e ences in su ace oughness we e obse ed when
using co es o silica sands BG 21 and BG 27 wi h di e en mean g ain
size d
50
.
No gas de ec s o po osi y we e obse ed on he su ace o he
cas ings.
Decla a ions
Au ho con ibu ion s a emen
Ma ina Gaw ono 
a: Pe o med he expe imen s; Analyzed and
in e p e ed he da a; W o e he pape .
Pe Lichý, Jan Jezie sk: Concei ed and designed he expe imen s.
I ana K oupo 
a: Pe o med he expe imen s; W o e he pape .
Tom
a
s Obzina, V
acla Me a: Pe o med he expe imen s.
Ja osla Be
no, Filip Radko ský: Analyzed and in e p e ed he da a.
Figu e 11. Su ace oughness a cas ed condi ion made by co es om sand BG 21, BG 27 and 3D sand p in ed.
Figu e 12. Roughness Ra and Rz o he su aces o he op and bo om inne
pa o he cas ing.
Figu e 13. O e all su ace oughness Ra and Rz o cas ings.
M. Gaw ono 
a e al. Heliyon 8 (2022) e10751
9