Identifying fault-inducing phenomena in an old masonry building

(1)

ScienceDirect

Available online at Available online at www.sciencedirect.comwww.sciencedirect.com

ScienceDirect

Structural Integrity Procedia 00 (2016) 000–000

www.elsevier.com/locate/procedia

Peer-review under responsibility of the Scientific Committee of PCF 2016.

XV Portuguese Conference on Fracture, PCF 2016, 10-12 February 2016, Paço de Arcos, Portugal

Thermo-mechanical modeling of a high pressure turbine blade of an

airplane gas turbine engine

P. Brandão

a

_{, V. Infante}

b

_{, A.M. Deus}

c

_*

a_{Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa,} Portugal

b_{IDMEC, Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa,} Portugal

c_{CeFEMA, Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa,} Portugal

Abstract

During their operation, modern aircraft engine components are subjected to increasingly demanding operating conditions, especially the high pressure turbine (HPT) blades. Such conditions cause these parts to undergo different types of time-dependent degradation, one of which is creep. A model using the finite element method (FEM) was developed, in order to be able to predict the creep behaviour of HPT blades. Flight data records (FDR) for a specific aircraft, provided by a commercial aviation company, were used to obtain thermal and mechanical data for three different flight cycles. In order to create the 3D model needed for the FEM analysis, a HPT blade scrap was scanned, and its chemical composition and material properties were obtained. The data that was gathered was fed into the FEM model and different simulations were run, first with a simplified 3D rectangular block shape, in order to better establish the model, and then with the real 3D mesh obtained from the blade scrap. The overall expected behaviour in terms of displacement was observed, in particular at the trailing edge of the blade. Therefore such a model can be useful in the goal of predicting turbine blade life, given a set of FDR data.

Peer-review under responsibility of the Scientific Committee of PCF 2016.

Keywords: High Pressure Turbine Blade; Creep; Finite Element Method; 3D Model; Simulation.

* Corresponding author. Tel.: +351 218419991.

E-mail address: amd@tecnico.ulisboa.pt

Procedia Structural Integrity 11 (2018) 290–297

XI

I

Abst The whic build interv insta crack depth back prope the m accou nume Copy Peer-Keyw 1. In S any) * C E -3216 Copyright review under resp

IV Internati

dentifyin

tract structural beha ch is probably ding underwent ventions, the b alled to track th k pattern allow h analysis on th k to the differen erties of the dif model. Structur unt. The tensile erical results ar yright © 2018 E -review under r words: masonry; c ntroduction tructural anal ) and the ensu

onal Confer

ng

fault-Dept. of Ci avior of a maso the outcome o several interve building is still he major cracks s the causes th he building also nt construction p fferent material ral analyses w e stresses in so re used to propo Elsevier B.V. A responsibility o cracking; ground lyses of mason uing safety le thor. Tel.: +39-02 lberto.taliercio@p Available onlin

Scie

Structural Integ B.V. All rights re e CINPAR 2018 o

rence on Bu

-inducin

P. C

ivil and Environm

onry farmhouse of different con entions of conso l heavily dama s for a duration hat originated th o allows some g phases, and loc ls into account, were carried out me parts of the ose simple retro All rights reserve

f the CINPAR settlements; finit nry buildings evel, and to pr 2-23994241; fax: polimi.it e at www.scie

enceDir

rity Procedia 00 ( eserved. organizers.

uilding Path

ng pheno

Condoleo,

mental Engineerin e exhibiting se nstruction phas olidation, includ aged, with verti

n of 18 months he main faults, geometric anom cally weaken th was developed t taking gravity e building are f ofitting strategie ed. 2018 organizer te elements. are particular redict the pos

+39-02-2399430 encedirect.com

rect

(2018) 000–000

hology and C

omena in

A. Talierc

ng, Politecnico di everal faults is ses; the oldest ding the additio ical and inclin s: some were f including chan malies to be hig he structural str

d. The main cra ty loads, groun found to be suf es rs rly effective t ssible damage 00.. m

Constructio

n an old m

cio*

i Milano, Milan 2 investigated. T part dates bac on of ties, a but ned deep cracks found to be stil nges in geomet ghlighted. Thes rength. A detail acks existing in nd settlements, fficiently high t to determine t e evolution. T www.elsevier.

ons Repair –

masonry

20133 (Italy) The building ha k to the seven ttress, and micro s. A static mon l propagating. try, to be better e anomalies can ed finite eleme n the building w and possible s o induce furthe

the actual dam The advanced .com/locate/proce

– CINPAR 2

y buildin

as an irregular nteenth century ropiles. Despite nitoring system Interpretation o r understood. A an probably be t

ent model, takin were also includ seismic actions er crack growth mage conditio survey techn edia

2018

ng

plan, y. The these m was of the An in-traced ng the ded in s into h. The ons (if niques 10.1016/j.prostr.2018.11.038 2452-3216

Peer-review under responsibility of the CINPAR 2018 organizers

2452-Peer-r

XI

I

IV Internati

dentifyin

onal Confer

ng

fault-Dept. of Ci avior of a maso the outcome o several interve building is still he major cracks s the causes th he building also nt construction p fferent material ral analyses w e stresses in so re used to propo Elsevier B.V. A responsibility o cracking; ground lyses of mason uing safety le thor. Tel.: +39-02 lberto.taliercio@p Available onlin

Scie

rence on Bu

-inducin

P. C

ivil and Environm

f the CINPAR settlements; finit nry buildings evel, and to pr 2-23994241; fax: polimi.it e at www.scie

enceDir

uilding Path

ng pheno

Condoleo,

+39-02-2399430 encedirect.com

rect

(2018) 000–000

hology and C

omena in

A. Talierc

d. The main cra ty loads, groun found to be suf es rs rly effective t ssible damage 00.. m

Constructio

n an old m

cio*

ons Repair –

masonry

– CINPAR 2

y buildin

2018

ng

plan, y. The these m was of the An in-traced ng the ded in s into h. The ons (if niques

2 P. Condoleo and A. Taliercio/ Structural Integrity Procedia 00 (2018) 000–000

available nowadays allow the geometry of buildings to be reconstructed in detail taking the typical complex and stratified morphology of historical buildings into account.

The transition from a very detailed geometric 3D model to a numerical model is often a cumbersome and tricky procedure. Apparently, it is necessary to simplify reality and, in particular, to carry out the subjective operation of discretizing the actual structure (Docci 1994). Whereas this procedure is almost spontaneous following a direct survey, it is somewhat more complex if advanced technologies are used. On the one hand, any geometric model needs to be simplified by eliminating details that are unnecessary from the structural point of view. On the other hand, construction techniques, materials, constructive faults and damages of the building should be known as accurately as possible, as they can significantly affect the structural behavior, although they may seem only minor issues.

As far as the analysis of the damage conditions is concerned, two distinct objectives can be pursued. The first one consists in taking the “as built” state into account, i.e. neglecting cracks and geometric variations, to check whether the detected type of damage is compatible with the presumed original conditions of the building. Here, the first major issue related to the interpretation of the building conditions arises. Is any detected anomaly in the geometry a symptom of failure, or was the building simply built in that way? The answer requires an analysis to link the surveyed crack pattern to the deformation of the building, to understand the mechanisms by which it originated. The local building techniques and the historic period in which they were used, as well as the available historical documentation, have to be taken into account to try and explain existing faults (Giuffrè 2010). This information allows a “virtual” model (Siviero et al. 1997) of the undamaged building to be developed, in which the possible fault-inducing effects can be incorporated. The model can be validated by checking its ability in reproducing the existing crack pattern, originated by stresses exceeding the material strength. The second objective is to analyze the building in its current, damaged configuration, both to check the possibility of damage evolution and to assess the effectiveness of retrofitting techniques. Another element to be taken into account is the specific vulnerability of the building, which is affected by the construction technique, the transformations that have occurred over time, and the possible repairs (Doglioni and Mazzotti 2007). Examples of these factors, which induce discontinuities similarly to cracks, are flues, walls built at a later time and weakly connected to the rest of the building, walls resting on vaults, etc. Whereas the specific vulnerability has only a minor influence on the behavior of masonry structures subjected to gravity loads, it can play a detrimental role under horizontal actions (e.g. earthquakes, or thrusts exerted by arches and vaults). These factors must be incorporated in the virtual model, to assess their effects and to make the model as representative as possible of the actual building.

In the present work, the above considerations will be applied to the assessment of a masonry farmhouse experiencing severe faults.

2. 2. Case study: a masonry farmhouse in Northern Italy

2.1. Description of the building

The analyzed building is a two-storey farmhouse located in a hilly region in Northern Italy. The building has a relatively simple L-shaped plan. The complex is rather irregular, probably as a result of different construction phases (Fig. 1a). Only a part of the building has a basement. The southern side of the farmhouse adjoins other buildings. Most of the building consists of load-bearing walls in solid bricks and earthen mortar. The inner and outer walls are plastered (Fig. 2a), except for the north façade where brickwork is exposed (Fig. 2b). All the rooms have timber floors. In the oldest part of the house, the staircase is surmounted by a barrel vault, with two cross-vaults at both ends. The vault at the first floor is segmental. Also the entrance hall to the courtyard is topped by a barrel vault.

A detailed analysis of the most ancient portion of the building allowed some geometric anomalies (specific vulnerabilities) to be highlighted: these “local weaknesses” might have fostered and affected the state of damage. The inner wall contains a flue that crosses the three floors, thereby reducing the wall section. Moreover, the openings are not aligned, thus leading to significant deviations of the stress path (Fig 1a). Another anomaly is in the entrance hall: by superposing the plans of the ground and of the first floor, it clearly stands out that a relatively thin wall of a bathroom (first floor) was built over the underlying vault (Fig. 1b); also the thickness of the wall

(2)

P. Condoleo et al. / Procedia Structural Integrity 11 (2018) 290–297 291 2452-Peer-r

XI

I

IV Internati

dentifyin

onal Confer

ng

fault-Dept. of Ci avior of a maso the outcome o several interve building is still he major cracks s the causes th he building also nt construction p fferent material ral analyses w e stresses in so re used to propo Elsevier B.V. A responsibility o cracking; ground lyses of mason uing safety le thor. Tel.: +39-02 lberto.taliercio@p

Scie

rence on Bu

-inducin

P. C

ivil and Environm

f the CINPAR settlements; finit nry buildings evel, and to pr 2-23994241; fax: polimi.it

enceDir

uilding Path

ng pheno

Condoleo,

+39-02-2399430

rect

(2018) 000–000

hology and C

omena in

A. Talierc

d. The main cra ty loads, groun found to be suf es rs rly effective t ssible damage 00..

Constructio

n an old m

cio*

ons Repair –

masonry

– CINPAR 2

y buildin

2018

ng

plan, y. The these m was of the An in-traced ng the ded in s into h. The ons (if niques 2452-Peer-r

XI

I

IV Internati

dentifyin

onal Confer

ng

fault-Dept. of Ci avior of a maso the outcome o several interve building is still he major cracks s the causes th he building also nt construction p fferent material ral analyses w e stresses in so re used to propo Elsevier B.V. A responsibility o cracking; ground lyses of mason uing safety le thor. Tel.: +39-02 lberto.taliercio@p

Scie

rence on Bu

-inducin

P. C

ivil and Environm

f the CINPAR settlements; finit nry buildings evel, and to pr 2-23994241; fax: polimi.it

enceDir

uilding Path

ng pheno

Condoleo,

+39-02-2399430

rect

(2018) 000–000

hology and C

omena in

A. Talierc

d. The main cra ty loads, groun found to be suf es rs rly effective t ssible damage 00..

Constructio

n an old m

cio*

ons Repair –

masonry

– CINPAR 2

y buildin

2018

ng

plan, y. The these m was of the An in-traced ng the ded in s into h. The ons (if niques

2 P. Condoleo and A. Taliercio/ Structural Integrity Procedia 00 (2018) 000–000

available nowadays allow the geometry of buildings to be reconstructed in detail taking the typical complex and stratified morphology of historical buildings into account.

The transition from a very detailed geometric 3D model to a numerical model is often a cumbersome and tricky procedure. Apparently, it is necessary to simplify reality and, in particular, to carry out the subjective operation of discretizing the actual structure (Docci 1994). Whereas this procedure is almost spontaneous following a direct survey, it is somewhat more complex if advanced technologies are used. On the one hand, any geometric model needs to be simplified by eliminating details that are unnecessary from the structural point of view. On the other hand, construction techniques, materials, constructive faults and damages of the building should be known as accurately as possible, as they can significantly affect the structural behavior, although they may seem only minor issues.

As far as the analysis of the damage conditions is concerned, two distinct objectives can be pursued. The first one consists in taking the “as built” state into account, i.e. neglecting cracks and geometric variations, to check whether the detected type of damage is compatible with the presumed original conditions of the building. Here, the first major issue related to the interpretation of the building conditions arises. Is any detected anomaly in the geometry a symptom of failure, or was the building simply built in that way? The answer requires an analysis to link the surveyed crack pattern to the deformation of the building, to understand the mechanisms by which it originated. The local building techniques and the historic period in which they were used, as well as the available historical documentation, have to be taken into account to try and explain existing faults (Giuffrè 2010). This information allows a “virtual” model (Siviero et al. 1997) of the undamaged building to be developed, in which the possible fault-inducing effects can be incorporated. The model can be validated by checking its ability in reproducing the existing crack pattern, originated by stresses exceeding the material strength. The second objective is to analyze the building in its current, damaged configuration, both to check the possibility of damage evolution and to assess the effectiveness of retrofitting techniques. Another element to be taken into account is the specific vulnerability of the building, which is affected by the construction technique, the transformations that have occurred over time, and the possible repairs (Doglioni and Mazzotti 2007). Examples of these factors, which induce discontinuities similarly to cracks, are flues, walls built at a later time and weakly connected to the rest of the building, walls resting on vaults, etc. Whereas the specific vulnerability has only a minor influence on the behavior of masonry structures subjected to gravity loads, it can play a detrimental role under horizontal actions (e.g. earthquakes, or thrusts exerted by arches and vaults). These factors must be incorporated in the virtual model, to assess their effects and to make the model as representative as possible of the actual building.

In the present work, the above considerations will be applied to the assessment of a masonry farmhouse experiencing severe faults.

2. 2. Case study: a masonry farmhouse in Northern Italy

2.1. Description of the building

The analyzed building is a two-storey farmhouse located in a hilly region in Northern Italy. The building has a relatively simple L-shaped plan. The complex is rather irregular, probably as a result of different construction phases (Fig. 1a). Only a part of the building has a basement. The southern side of the farmhouse adjoins other buildings. Most of the building consists of load-bearing walls in solid bricks and earthen mortar. The inner and outer walls are plastered (Fig. 2a), except for the north façade where brickwork is exposed (Fig. 2b). All the rooms have timber floors. In the oldest part of the house, the staircase is surmounted by a barrel vault, with two cross-vaults at both ends. The vault at the first floor is segmental. Also the entrance hall to the courtyard is topped by a barrel vault.

A detailed analysis of the most ancient portion of the building allowed some geometric anomalies (specific vulnerabilities) to be highlighted: these “local weaknesses” might have fostered and affected the state of damage. The inner wall contains a flue that crosses the three floors, thereby reducing the wall section. Moreover, the openings are not aligned, thus leading to significant deviations of the stress path (Fig 1a). Another anomaly is in the entrance hall: by superposing the plans of the ground and of the first floor, it clearly stands out that a relatively thin wall of a bathroom (first floor) was built over the underlying vault (Fig. 1b); also the thickness of the wall

(3)

292 P. Condoleo et al. / Procedia Structural Integrity 11 (2018) 290–297 over layo Fig.1 to the a Fig. 3 the bu T inter and prob fit un rlooking the m ut can be con a

. (a) Plan of the g e plan of the first

a

3. (a) Ties at two uilding.

The building rventions. The

two anchor p bably to resist nder the exter

P. C

main street, th sidered a sour

ground floor and floor (green).

Fig. 2

openings in the

has suffered ese include tie plates beside

the movemen rnal walls faci

Condoleo and A.

at rests on the rce of structur

cross-section sho

2. (a) Facade over

b

south side of the

d structural p es located in o the entrance nt of this part o ing both the st

Taliercio / Struct

e underlying v ral weakness in

owing unevenly sp

rlooking the stree

attic. (b) Anchor

problems, as openings of th

hall (Fig. 3b) of the building treet and the e

tural Integrity Pr vault, should b n the event of b spaced openings a b

et. (b) Facade ove

r plates at both si

can be infer he attic in the

). A buttress g due to groun east side of the

rocedia 00 (2018)

be further inv f ground settle

and a flue. (b) Pla

erlooking the cour

c

ides of the entran

rred by num presumably o was also add nd settlements e courtyard. ) 000–000 vestigated. Thi ements or earth an of the ground f rtyard.

nce hall. (c) Buttr

merous, still v oldest part of t ed at the N-E s. In 2010 som is weird geom thquakes:

floor (grey) super

ress at the north s

visible retrof the house (Fig E corner (Fig me micropiles 3 metric rposed side of fitting g. 3a) . 3c), were 4 T con 2.2. T buil mar T side the also T the T the a 2.3. A focu Cra stay F outs wal chan its i faul

The crack pat nsolidation cou

Survey of the

To evaluate th lding was sur rked in red or The NE part o e of the extern same wall; th o found in the The out-of-plu courtyard of a The surveyed building tend Fig. 4. Fig. 5. Schema Static monito A static monit used on the ev ack movement y in place over Fig.s 6a,b sho side and insid ll facing the co nges. Howeve initial value, m lt induced by d P. Co tterns before uld not be asse

e crack pattern

he faults in th rveyed. In Fi

black, accord of the buildin nal wall facing his crack exte

entrance hall umb of the faç about 0.1 m at crack pattern ds to split. This Elevations of the atization of the ou oring toring system volution of so ts also allow th r a period of 1 ow the installa de the building ourtyard is plo er, after 12 mo meaning that f differential gr ondoleo and A. T

and after retr essed. n he building a ig. 4, cracks ding to their w g was found t g the street is ends to the op : they are like çades was also t a height of 6 , changes in g s is particularl e façade overlook ut-of-plumb (in c was installed ome through-w he type of for 18 months at l ation of a mil g during the m otted versus ti onths, the wid faults are still round settleme Taliercio/ Structur rofitting were after micropili passing acro width, whereas to experience also found at pposite wall fa ly to be the co o surveyed: it 6 m from the g geometry (out-ly evident in t

king the street (lef

b cm): (a) facade ov d in the farmho wall cracks, w rces acting on east, so that th limetre dial g monitoring pe ime. The obta dth of some of evolving. By ents at the NE

ral Integrity Proc

e not surveyed

ing, the crack ss the entire other cracks e the most sev the ground fl acing the cour ontinuation of t was found th ground (Fig. 5 -of-plumb) an the NE part, at

ft) and the opposi

b verlooking the str ouse, to asses which might si the building t he effects of s gauge on a cra eriod. In Fig. ained data sho f the cracks (l examining th E end of the bu cedia 00 (2018) 0 d and compar k pattern on t wall thickne are marked in vere faults. A oor and at the rtyard. Two w f those in the l hat the wall at

).

nd differential t the inner wa

ite façade overloo

reet; (b) opposite s whether fau ignificantly af to be identifie seasonal therm ack. Fig. 6c s 6d, the openi ow that crack o labelled G5 an e crack pattern uilding can be 000–000 red, so that th

the outer and ess (passing-th n green. vertical crack e first floor of wide, passing-living room at the entrance h ground settle all embodying

oking the courtya

facade overlooki

ults are still ac ffect the stabil d. Any monito mal changes ca shows the cha

ing of the crac opening is mo nd G6 in Fig.

n and the chan identified (M he effectivene inner walls o hrough crack k running alon f the internal s -through crack t the ground fl hall is tilted to ements indicat the flue. ard (right).

ing the courtyard.

ctive. Attentio lity of the bui toring system an be discarde ange in tempe acks on the ex ostly due to th 6d) did not re nges in geome Mastrodicasa 2 ess of of the s) are ng the side of ks are loor. oward te that . n was ilding. has to ed. erature xternal hermal ecover etry, a 2012).

(4)

over layo Fig.1 to the a Fig. 3 the bu T inter and prob fit un rlooking the m ut can be con a

. (a) Plan of the g e plan of the first

a

3. (a) Ties at two uilding.

The building rventions. The

two anchor p bably to resist nder the exter

main street, th sidered a sour

ground floor and floor (green).

Fig. 2

openings in the

has suffered ese include tie plates beside

the movemen rnal walls faci

at rests on the rce of structur

cross-section sho

2. (a) Facade over

b

south side of the

d structural p es located in o the entrance nt of this part o ing both the st

e underlying v ral weakness in

owing unevenly sp

rlooking the stree

attic. (b) Anchor

problems, as openings of th

hall (Fig. 3b) of the building treet and the e

vault, should b n the event of

b

spaced openings a

b

et. (b) Facade ove

r plates at both si

can be infer he attic in the

). A buttress g due to groun east side of the

be further inv f ground settle

and a flue. (b) Pla

erlooking the cour

c

ides of the entran

rred by num presumably o was also add nd settlements e courtyard. vestigated. Thi ements or earth an of the ground f rtyard.

nce hall. (c) Buttr

merous, still v oldest part of t ed at the N-E s. In 2010 som is weird geom thquakes:

floor (grey) super

ress at the north s

visible retrof the house (Fig E corner (Fig me micropiles metric rposed side of fitting g. 3a) . 3c), were T con 2.2. T buil mar T side the also T the T the a 2.3. A focu Cra stay F outs wal chan its i faul

The crack pat nsolidation cou

Survey of the

To evaluate th lding was sur rked in red or The NE part o e of the extern same wall; th o found in the The out-of-plu courtyard of a The surveyed building tend Fig. 4. Fig. 5. Schema Static monito A static monit used on the ev ack movement y in place over Fig.s 6a,b sho side and insid ll facing the co

nges. Howeve initial value, m

lt induced by d

tterns before uld not be asse

e crack pattern

he faults in th rveyed. In Fi

black, accord of the buildin nal wall facing his crack exte

entrance hall umb of the faç about 0.1 m at crack pattern ds to split. This Elevations of the atization of the ou oring toring system volution of so ts also allow th r a period of 1 ow the installa de the building ourtyard is plo er, after 12 mo meaning that f differential gr

and after retr essed. n he building a ig. 4, cracks ding to their w g was found t g the street is ends to the op : they are like çades was also

t a height of 6 , changes in g s is particularl e façade overlook ut-of-plumb (in c was installed ome through-w he type of for 18 months at l ation of a mil g during the m otted versus ti onths, the wid faults are still round settleme rofitting were after micropili passing acro width, whereas to experience also found at pposite wall fa ly to be the co o surveyed: it 6 m from the g geometry (out-ly evident in t

king the street (lef

b cm): (a) facade ov d in the farmho wall cracks, w rces acting on east, so that th limetre dial g monitoring pe ime. The obta dth of some of evolving. By ents at the NE

e not surveyed

ing, the crack ss the entire other cracks e the most sev

the ground fl acing the cour ontinuation of t was found th ground (Fig. 5 -of-plumb) an the NE part, at

ft) and the opposi

b verlooking the str ouse, to asses which might si the building t he effects of s gauge on a cra eriod. In Fig. ained data sho f the cracks (l examining th E end of the bu d and compar k pattern on t wall thickne are marked in vere faults. A oor and at the rtyard. Two w f those in the l hat the wall at

).

nd differential t the inner wa

ite façade overloo

reet; (b) opposite s whether fau ignificantly af to be identifie seasonal therm ack. Fig. 6c s 6d, the openi ow that crack o labelled G5 an e crack pattern uilding can be red, so that th

the outer and ess (passing-th n green. vertical crack e first floor of wide, passing-living room at the entrance h ground settle all embodying

oking the courtya

facade overlooki

ults are still ac ffect the stabil d. Any monito mal changes ca

shows the cha ing of the crac opening is mo nd G6 in Fig.

n and the chan identified (M he effectivene inner walls o hrough crack k running alon f the internal s -through crack t the ground fl hall is tilted to ements indicat the flue. ard (right).

ing the courtyard.

ctive. Attentio lity of the bui toring system an be discarde ange in tempe acks on the ex ostly due to th 6d) did not re nges in geome Mastrodicasa 2 ess of of the s) are ng the side of ks are loor. oward te that . n was ilding. has to ed. erature xternal hermal ecover etry, a 2012).

(5)

294 P. Condoleo et al. / Procedia Structural Integrity 11 (2018) 290–297 D beha a b Fig. 6 Open 3. N 3.1. T (Fig the m 3.2. In from linea beam key ques T mate Damage is lik avior of the bu 6. (a) Preparation ning of the cracks Numerical mo

Geometrical m

The numerical . 7). Plans of t model, but rath

Materials and n the absence m the literature ar elastic and ms basically u elastic param stionable for m The software u erial densities P. C kely to be en uilding (as flo

n of the wall surf in the wall facin odel

model of the f

model was de the floors wer her taken into

Fig. 7. AutoCA d gravity load of experimen e (NTC 2008 isotropic. Thi undergo uniax meters are the masonry vault used for the FE s. In addition

Condoleo and A.

nhanced by th ors consist of

face to install the g the courtyard. farmhouse eveloped usin re available, to o account as de AD model of the b ds

ntal data, the e , UNI-EN 33 is assumption ial stress. Mas e axial Young s (Anzani et a E analyses (Ab to that, dead c Taliercio / Struct he geometrica f simply suppo e dial gauge. (b) M ng AutoCAD a ogether with e ead load building (grey: m elastic proper 8, etc.) and ar is reasonable sonry walls ar g’s modulus a al. 2010). baqus) autom loads taking tural Integrity Pr al singularitie orted beams). Measuring the cr according to a elevations and

masonry; red: woo

rties of the ma re listed in Ta not only for s re mainly subj and Poisson’s matically comp the weight of rocedia 00 (2018) es recalled ab rack width. (c) In an accurate su d cross-section

oden beams; brow

aterials formin able 1. All th steel and gran jected to in-pl s ratio. The a putes gravity lo f the ceilings d ) 000–000

bove, and the

nner and outer tem

urvey carried o ns. The roof w

wn: steel beams).

ng the farmho e materials w ite, but also fo lane vertical s assumption of oads accordin and the roof

e limited box

mperature vs. tim

out on the bui was not includ

ouse were ded were assumed for wood, as ti stresses, so tha f isotropy is ng to the presc into account 5 x-like me. (d) ilding ded in duced to be imber at the more cribed were 6 also acti kN/ resi and Fina 3.3. T (Fig the tow bou base 3.4. T (NT cho con A belo defi was give q = o applied on t ing on the upp /m2_{was cons}

idential buildi d attic, the liv ally, the weigh

Ground settl

The most seve g. 4). This crac

inner courtyar ward the courty

undary conditio e of its norther Fig. Seismic actio The intensity TC 2008). A h oice comes fro nstraint in case According to ongs to zone ine the seismi s considered ( es a soil facto 3.6 was assum Sd(T) = ag S P. Co the beams, by per side of th sidered, acco ings, in additi ve load is red ht per unit are

Table 1. Prope Material Steel Masonry Wood Granite lements

ere crack dete ck is likely to rd (Fig. 8a). T tyard because ons. A vertica rnmost part, as a

8. (a) Plan view o

ons

of the seismi horizontal gra om the presen e of seismic ex

the most rec 3, correspond ic action. For (deposits of lo or S = 1.35. Fi med (two-stor S 2.5Tc/(qT) = ondoleo and A. T y converting t e beams. Reg rding to the on to a conve duced to 0.5 k ea of the roof,

erties of the mater

ected in the bu cross the enti To explain thes of ground se al displacemen s shown in Fig of the surveyed o ic action was avity load wa nce of other si xcitation, prov cent territoria ding to a PGA lack of inform oose-to-mediu nally, to conv rey masonry b = 2.207 m/s2_, Taliercio/ Structur

the load distri garding the ce new Italian entional dead kN/m2_{, as in}

including tile

rials forming the Elements IPE beams Walls, vaults Ceilings, roof Stairs uilding is a su ire building, a se faults, the n ettlements. Th nt, linearly var g. 8b, in order out-of-plumb; (b) defined acco as added to th imilar building vided that ham l seismic cla A of 0.15g = mation on the um cohesionle vert the elastic buildings). Thu

ibuted over ea eiling between Technical St load of 1 kN/ principle the es, covering an farmhouse (E=Y E [M 2100 150 f 1100 5000 ub-vertical crac and is matched northernmost p hese were inc

rying across th to simulate a r b simulation of gro ording to the I he model para gs adjacent to mmering effec ssification of 1.47 m/s2_{. N} e stratigraphy ess soil, or pr c response spe us, the design

cedia 00 (2018) 0 ach tributary n ground floor tandards for /m2_{. Regardin} attic can be nd joists, was Young’s modulus, Pa] 000 00 00 00 ck visible on d by a signific part of the bui orporated in t he depth of th rotation about

ound settlements

Italian Techni allel to the lon o the longest w cts can be negl f Italy (2015) NTC standards of the soil be redominantly ectrum into de spectrum Sd(T 000–000 area into an e r and first flo

Constructions ng the ceiling accessed only assumed to be =Poisson’s rati  0.29 0.25 0.46 0.23 the façade ov cant tilt of the ilding was assu

the model by he building, w the side along

in the numerical ical Standards ngest wing of wing, which a lected. , the municip s consider five eneath the farm soft-to-firm c esign spectrum

T) is given by

equivalent pre oor, a live load s (NTC 2008 between first y for mainten e 1.14 kN/m2 io, =density).  [kg/m3_] 7900 1800 420 2600 verlooking the façade overlo sumed to be ro y means of su was prescribed g the street. model. s for Constru f the building act as an addi pality of Gam ve classes of s m, a soil of cl cohesive soil) m, a behavior y essure d of 2 8) for t floor nance. . street ooking otating uitable at the ctions . This itional malero soil to lass D . This factor (1)

(6)

D beha a b Fig. 6 Open 3. N 3.1. T (Fig the m 3.2. In from linea beam key ques T mate Damage is lik avior of the bu 6. (a) Preparation ning of the cracks Numerical mo

Geometrical m

The numerical . 7). Plans of t model, but rath

Materials and n the absence m the literature ar elastic and ms basically u elastic param stionable for m The software u erial densities kely to be en uilding (as flo

n of the wall surf in the wall facin odel

model of the f

model was de the floors wer her taken into

Fig. 7. AutoCA d gravity load of experimen e (NTC 2008 isotropic. Thi undergo uniax meters are the masonry vault used for the FE s. In addition

nhanced by th ors consist of

face to install the g the courtyard. farmhouse eveloped usin re available, to o account as de AD model of the b ds

ntal data, the e , UNI-EN 33 is assumption

ial stress. Mas e axial Young s (Anzani et a E analyses (Ab to that, dead c he geometrica f simply suppo e dial gauge. (b) M ng AutoCAD a ogether with e ead load building (grey: m elastic proper 8, etc.) and ar is reasonable sonry walls ar g’s modulus a al. 2010). baqus) autom loads taking al singularitie orted beams). Measuring the cr according to a elevations and

masonry; red: woo

rties of the ma re listed in Ta not only for s re mainly subj and Poisson’s matically comp the weight of es recalled ab rack width. (c) In an accurate su d cross-section

oden beams; brow

aterials formin able 1. All th steel and gran jected to in-pl s ratio. The a putes gravity lo

f the ceilings d

bove, and the

nner and outer tem

urvey carried o ns. The roof w

wn: steel beams).

ng the farmho e materials w ite, but also fo lane vertical s assumption of oads accordin and the roof

e limited box

mperature vs. tim

out on the bui was not includ

ouse were ded were assumed for wood, as ti stresses, so tha f isotropy is ng to the presc into account x-like me. (d) ilding ded in duced to be imber at the more cribed were also acti kN/ resi and Fina 3.3. T (Fig the tow bou base 3.4. T (NT cho con A belo defi was give q = o applied on t ing on the upp /m2_{was cons}

idential buildi d attic, the liv ally, the weigh

Ground settl

The most seve g. 4). This crac

inner courtyar ward the courty

undary conditio e of its norther Fig. Seismic actio The intensity TC 2008). A h oice comes fro nstraint in case According to ongs to zone ine the seismi s considered ( es a soil facto 3.6 was assum Sd(T) = ag S the beams, by per side of th sidered, acco ings, in additi ve load is red ht per unit are

Table 1. Prope Material Steel Masonry Wood Granite lements

ere crack dete ck is likely to rd (Fig. 8a). T tyard because ons. A vertica rnmost part, as a

8. (a) Plan view o

ons

of the seismi horizontal gra om the presen e of seismic ex

the most rec 3, correspond ic action. For (deposits of lo or S = 1.35. Fi med (two-stor S 2.5Tc/(qT) = y converting t e beams. Reg rding to the on to a conve duced to 0.5 k ea of the roof,

erties of the mater

ected in the bu cross the enti To explain thes of ground se al displacemen s shown in Fig of the surveyed o ic action was avity load wa nce of other si xcitation, prov cent territoria ding to a PGA lack of inform oose-to-mediu nally, to conv rey masonry b = 2.207 m/s2_,

the load distri garding the ce new Italian entional dead kN/m2_{, as in}

including tile

rials forming the Elements IPE beams Walls, vaults Ceilings, roof Stairs uilding is a su ire building, a se faults, the n ettlements. Th nt, linearly var g. 8b, in order out-of-plumb; (b) defined acco as added to th imilar building vided that ham l seismic cla A of 0.15g = mation on the um cohesionle vert the elastic buildings). Thu ibuted over ea eiling between Technical St load of 1 kN/ principle the es, covering an farmhouse (E=Y E [M 2100 150 f 1100 5000 ub-vertical crac and is matched northernmost p hese were inc

rying across th to simulate a r b simulation of gro ording to the I he model para gs adjacent to mmering effec ssification of 1.47 m/s2_{. N} e stratigraphy ess soil, or pr c response spe us, the design

ach tributary n ground floor tandards for /m2_{. Regardin} attic can be nd joists, was Young’s modulus, Pa] 000 00 00 00 ck visible on d by a signific

part of the bui orporated in t he depth of th rotation about

ound settlements

Italian Techni allel to the lon o the longest w cts can be negl f Italy (2015) NTC standards of the soil be redominantly ectrum into de spectrum Sd(T area into an e r and first flo

Constructions ng the ceiling accessed only assumed to be =Poisson’s rati  0.29 0.25 0.46 0.23 the façade ov cant tilt of the ilding was assu

the model by he building, w the side along

in the numerical ical Standards ngest wing of wing, which a lected. , the municip s consider five eneath the farm soft-to-firm c esign spectrum

T) is given by

equivalent pre oor, a live load s (NTC 2008 between first y for mainten e 1.14 kN/m2 io, =density).  [kg/m3_] 7900 1800 420 2600 verlooking the façade overlo sumed to be ro y means of su was prescribed g the street. model. s for Constru f the building act as an addi pality of Gam ve classes of s m, a soil of cl cohesive soil) m, a behavior y essure d of 2 8) for t floor nance. . street ooking otating uitable at the ctions . This itional malero soil to lass D . This factor (1)

(7)

296 _{P. Condoleo and A. Taliercio / Structural Integrity Procedia 00 (2018) 000–000}P. Condoleo et al. / Procedia Structural Integrity 11 (2018) 290–297 ₇

where Tc is the (corner) period at the end of the plateau in which the response spectrum is constant (0.8s for class

D soils). Assuming a natural period of vibration T = 0.5s, the value in eq. (1) is obtained.

3.5. Finite element model

The finite element model used in the numerical analyses consists of approximately 715000 4-node constant stress tetrahedra, with 500000 degrees of freedom. A sufficiently refined mesh is required to match the irregular geometry of the model (including cracks) with good accuracy. The base of the model and the external sides of the foundation walls are fixed, except for a part of the base where vertical displacements can be prescribed (Fig. 8(b)).

4. Numerical results

The FE model of the farmhouse was analyzed under the following load conditions: (i) weight; (ii) self-weight + ground settlements; (iii) self-self-weight + ground settlements + seismic actions. For each load condition, both the presumed original situation, in which the building is uncracked, and the current situation, in which cracks are open, were considered. The results will be presented in terms of contours of the maximum principal (tensile) stress, as compressions were never found to exceed the presumed compressive strength of the materials.

4.1. Original conditions (closed cracks)

Fig. 9 shows the stress distribution in the building assumed to be uncracked. Unsurprisingly, Fig. 9(a) shows that the only service loads acting on the structure cannot explain the crack pattern in the building. On the contrary, the prescribed ground settlements induce significant tensile stresses in wide regions of the building: the grey areas in Fig. 9(b) correspond to stresses exceeding 0.2 MPa, which can be conventionally assumed to be a critical value for masonry in tension according to the literature (Lourenço 2002). Note that these areas roughly match the regions where most of the cracks are found in the walls overlooking the inner courtyard (see Fig. 4). Fig. 9(c) shows a detail of the region experiencing ground settlements: stress concentrations can be noticed on the front of the house facing the street, which match the deep vertical crack shown in Fig. 4. The situation worsens if the seismic action is taken into account (Fig. 9(d)). In the latter case, also the walls parallel to the assumed seismic load experience tensile stresses at approximately 45° degrees, which might induce the typical crack pattern of in-plane loaded shear walls.

4.2. Current conditions (open cracks)

Fig. 10 shows the stress distribution in the building if cracks are assumed to be open, as in the current conditions. If no ground settlements existed, the stress in the building would not be of concern and would basically be the same computed in the uncracked building (compare Figures 9(a) and 10(a)). The stress in the building experiencing ground settlements is not significantly affected by the presence of cracks as far as the walls overlooking the inner courtyard are concerned (compare Figures 9(b) and 10(b)). On the contrary, in the wall facing the street the extension of the regions experiencing high tensile stresses is greater than in the uncracked building (compare Figures 9(c) and 10(c)), which means that the existing cracks are likely to grow further. The stress in the building experiencing seismic loads in addition to the previously considered actions does not significantly differ from that shown in Fig. 9(d) and is not reported here.

5. Concluding remarks

The structural analysis of the farmhouse allowed the assumption regarding the origin of the main cracks in the building to be confirmed. A ground settlements, basically consisting in the rotation of part of the base of the building about the road axis, might explain most of the cracks (Sec. 4.1). If the existing cracks are incorporated in the numerical model, tensile stresses are found to increase further in some regions of the building, which indicates that cracks are likely to develop further (Sec. 4.2). Seismic loads might aggravate the crack pattern and additional cracks might develop.

8 I evo mak mig Fig. the r Fig. regio Ref Anza o Docc Dogl E Giuf Lour Mast NTC Sivie a In the continu olution will be ke the buildin ght also be imp a c 9. Maximum prin region experiencin 10. Maximum p on experiencing g ferences

ani, A., Condoleo on Structural Ana ci, M., Maestri, D lioni, F., Mazzot Edition, Ed. Regio ffrè, A.,2010. Leg renço, P.B., 2002 trodicasa, S., 201 C, 2008. Italian Te

ero, E., Forabosch

P. Co uation of the r assessed. As ng act as a box proved to tie t ncipal stress in th ng ground settlem principal stress in ground settlement o, P., Gobbo, A., alysis of Historic D., 1994. Manuale tti, P., 2007. Cod one Marche, Anc ggendo il libro de 2. Computations o 12. Dissesti static echnical Standard hi, P., Barbieri, A ondoleo and A. T research, the the micropilin x-type structur the building w he model of the u ments; (d) self-we

n the model with ts. Taliercio, A., 20 Constructions. Sh e di rilevamento a dice di pratica p cona.

elle antiche archit on historic mason i delle strutture e ds for Constructio A., 1997. Lettura Taliercio/ Structur effectiveness ng system inst re and resist o walls more effe

uncracked buildin eight + ground se

open cracks: (a)

10. Modeling the hanghai, China (6 architettonico e u er gli interventi tetture, 1st Edition nry structures. Pro

dilizie, 9th Editio ons (in Italian), D strutturale delle c

b

of possible i talled so far h overturning of fectively. b d ng: (a) self-weigh ettlements + seism ) self-weight only e static behavior o 6-8/10), Vol. 1, 3 urbano, 1st Editio di miglioramento n, Gangemi, Rom ogress in Structur on, Hoepli, Milan D.M. (Ministerial costruzioni, 1st E

cedia 00 (2018) 0

nterventions o has not proven

f the external ht only; (b) self-w mic action. y; (b) self-weigh of a double curva 367-372. on, Laterza, Bari.

o sismico nel res ma.

ral Engineering an n (I).

Decree) of Janua dition, Città Stud

000–000

on the contain n effective, tie

walls. The sti

weight + ground s

ht + ground settle

ture brickwork va stauro del patrim nd Materials 4 (3 ary 14, 2008. di, Torino. c nment of the es could be add iffness of the settlements; (c) d ements; (c) detail

ault, Proc. 7th Int monio architetton ), 301-309. crack ded to floors detail of l of the t. Conf. ico, 1st

(8)

where Tc is the (corner) period at the end of the plateau in which the response spectrum is constant (0.8s for class

D soils). Assuming a natural period of vibration T = 0.5s, the value in eq. (1) is obtained.

3.5. Finite element model

The finite element model used in the numerical analyses consists of approximately 715000 4-node constant stress tetrahedra, with 500000 degrees of freedom. A sufficiently refined mesh is required to match the irregular geometry of the model (including cracks) with good accuracy. The base of the model and the external sides of the foundation walls are fixed, except for a part of the base where vertical displacements can be prescribed (Fig. 8(b)).

4. Numerical results

The FE model of the farmhouse was analyzed under the following load conditions: (i) weight; (ii) self-weight + ground settlements; (iii) self-self-weight + ground settlements + seismic actions. For each load condition, both the presumed original situation, in which the building is uncracked, and the current situation, in which cracks are open, were considered. The results will be presented in terms of contours of the maximum principal (tensile) stress, as compressions were never found to exceed the presumed compressive strength of the materials.

4.1. Original conditions (closed cracks)

Fig. 9 shows the stress distribution in the building assumed to be uncracked. Unsurprisingly, Fig. 9(a) shows that the only service loads acting on the structure cannot explain the crack pattern in the building. On the contrary, the prescribed ground settlements induce significant tensile stresses in wide regions of the building: the grey areas in Fig. 9(b) correspond to stresses exceeding 0.2 MPa, which can be conventionally assumed to be a critical value for masonry in tension according to the literature (Lourenço 2002). Note that these areas roughly match the regions where most of the cracks are found in the walls overlooking the inner courtyard (see Fig. 4). Fig. 9(c) shows a detail of the region experiencing ground settlements: stress concentrations can be noticed on the front of the house facing the street, which match the deep vertical crack shown in Fig. 4. The situation worsens if the seismic action is taken into account (Fig. 9(d)). In the latter case, also the walls parallel to the assumed seismic load experience tensile stresses at approximately 45° degrees, which might induce the typical crack pattern of in-plane loaded shear walls.

4.2. Current conditions (open cracks)

Fig. 10 shows the stress distribution in the building if cracks are assumed to be open, as in the current conditions. If no ground settlements existed, the stress in the building would not be of concern and would basically be the same computed in the uncracked building (compare Figures 9(a) and 10(a)). The stress in the building experiencing ground settlements is not significantly affected by the presence of cracks as far as the walls overlooking the inner courtyard are concerned (compare Figures 9(b) and 10(b)). On the contrary, in the wall facing the street the extension of the regions experiencing high tensile stresses is greater than in the uncracked building (compare Figures 9(c) and 10(c)), which means that the existing cracks are likely to grow further. The stress in the building experiencing seismic loads in addition to the previously considered actions does not significantly differ from that shown in Fig. 9(d) and is not reported here.

5. Concluding remarks

The structural analysis of the farmhouse allowed the assumption regarding the origin of the main cracks in the building to be confirmed. A ground settlements, basically consisting in the rotation of part of the base of the building about the road axis, might explain most of the cracks (Sec. 4.1). If the existing cracks are incorporated in the numerical model, tensile stresses are found to increase further in some regions of the building, which indicates that cracks are likely to develop further (Sec. 4.2). Seismic loads might aggravate the crack pattern and additional cracks might develop.

I evo mak mig Fig. the r Fig. regio Ref Anza o Docc Dogl E Giuf Lour Mast NTC Sivie a In the continu olution will be ke the buildin ght also be imp a c 9. Maximum prin region experiencin 10. Maximum p on experiencing g ferences

ani, A., Condoleo on Structural Ana ci, M., Maestri, D lioni, F., Mazzot Edition, Ed. Regio ffrè, A.,2010. Leg renço, P.B., 2002 trodicasa, S., 201 C, 2008. Italian Te ero, E., Forabosch

uation of the r assessed. As ng act as a box proved to tie t ncipal stress in th ng ground settlem principal stress in ground settlement o, P., Gobbo, A., alysis of Historic D., 1994. Manuale tti, P., 2007. Cod one Marche, Anc ggendo il libro de 2. Computations o 12. Dissesti static echnical Standard hi, P., Barbieri, A research, the the micropilin x-type structur the building w he model of the u ments; (d) self-we

n the model with ts. Taliercio, A., 20 Constructions. Sh e di rilevamento a dice di pratica p cona.

elle antiche archit on historic mason i delle strutture e ds for Constructio A., 1997. Lettura effectiveness ng system inst re and resist o walls more effe

uncracked buildin eight + ground se

open cracks: (a)

10. Modeling the hanghai, China (6 architettonico e u er gli interventi tetture, 1st Edition nry structures. Pro

dilizie, 9th Editio ons (in Italian), D strutturale delle c b of possible i talled so far h overturning of fectively. b d ng: (a) self-weigh ettlements + seism ) self-weight only e static behavior o 6-8/10), Vol. 1, 3 urbano, 1st Editio di miglioramento n, Gangemi, Rom ogress in Structur on, Hoepli, Milan D.M. (Ministerial costruzioni, 1st E

nterventions o has not proven

f the external ht only; (b) self-w mic action. y; (b) self-weigh of a double curva 367-372. on, Laterza, Bari.

o sismico nel res ma.

ral Engineering an n (I).

Decree) of Janua dition, Città Stud

on the contain n effective, tie

walls. The sti

weight + ground s

ht + ground settle

ture brickwork va stauro del patrim nd Materials 4 (3 ary 14, 2008. di, Torino. c nment of the es could be add iffness of the settlements; (c) d ements; (c) detail

ault, Proc. 7th Int monio architetton ), 301-309. crack ded to floors detail of l of the t. Conf. ico, 1st