See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/265786432

Failure state of reinforced and prestressed concrete hollow-core slabs as a

result of a poor quality of precast elements

Chapter · January 2010

DOI: 10.13140/2.1.1836.6724

CITATIONS READS

0 7,459

2 authors, including:

Aldona Lowinska-Kluge
Independent researcher
28 PUBLICATIONS   114 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Buildings materials used in the nuclear industry View project

All content following this page was uploaded by Aldona Lowinska-Kluge on 18 September 2014.

The user has requested enhancement of the downloaded file.

Failure state of reinforced and prestressed concrete hollow-core
slabs as a result of a poor quality of precast elements

E. Przybyłowicza i A. Łowińska-Klugea
a
Poznan University of Technology, ul. Piotrowo 5, 60-965 Poznań

Precast reinforced or prestressed concrete hollow-core slabs are often used in civil engineering
structures of various purposes. The slabs suffer from frequent damage during their construction or service
period. In this paper causes leading to such a damage are discussed. A case study of a typical structure is
presented. The causes were determined during an experimental investigation of concrete specimens.
Compressive strength, apparent density, moisture content, adsorption and pH factor were measured and a
macroscopic assessment of precast elements, chemical and structural investigation of concrete including
SEM, X-ray and derivatography tests were carried out. Finally, methods to repair the damaged elements
designed after the analysis of damage causes and carried out during the construction stage are discussed.

Keywords: damaged precast elements, SEM, X-ray and derivatography tests, determination of damage causes,
repair of precast elements

1. Introduction
In contemporary school, public or industrial buildings precast reinforced or
prestressed concrete hollow-core slabs are frequently used. These elements have
several advantages as: relative large spans without temporary supports, high load
capacity, a short construction time, an immediate possibility to apply technological
loading during the construction time. On the other hand, the chief disadvantage, i.e.
slab faulting and opposite deflections in adjacent slabs must be taken into account.
However, this undesired feature is already eliminated in the new technology of
hollow-core slabs with HC ceiling slabs, where the precast element and reinforced
concrete cast on the slab carry the loads together. Thus, the load capacity is higher and
the slab faulting with negative deflections is not observed.
In this paper a case study of damage in typical reinforced concrete hollow-core
slabs and prestressed concrete SPIROLL slabs, used in a structure of one of university
buildings in Poznan is analyzed.
The damage occurred during a winter period in the open-state structure (without a
roof and window frames) in some floor slabs after 120-180 days of curing. The
concrete cover of channels at the bottom surface cracked and delaminated. In some
cases the entire channels burst due to freezing of precipitation water deposited inside.

177
 Damage was also observed in the case of prestressed concrete elements where no
ice in channels was formed. Hence, it was decided to carry out detailed macroscopic
and laboratory testing of drilled concrete cores taken from randomly selected floor
slabs. The results of this investigation are presented in the following sections.

2. General information about the building

The building is constructed for one of faculties at Adam Mickiewicz University in
Poznan. It consists of five segments connected by covered passages, which form
closed blocks of various height (Fig. 1).

Fig. 1. Plan view of structural segments

The prestressed slabs SP 26,5 with the reinforcement type 6 and the span of 6.53m
are used in the building „A” at the levels ±0.00, +4.08, +8.16. The composite slabs of
Filigran type with 2.40m span cover the corridor passages. The passage between the
segments „A” and „B” is covered with the slabs SP 40/16 spanning l = 15.65m and SP
32/10 spanning l = 12.38m. The building „A” is also connected with the segment „E”
by the passage covered with the slabs SP 40/16 spanning l = 15.85m.
Monolithic floors were designed over the staircases.
SP 26,5 slabs with 5/6 reinforcement type and the span l=6.50m and the slabs SP
32 spanning l = 14.19m are used in the building „B”. These floors are constructed at
the levels ±0.00, +4.08m, +8.16m. Besides, numerous cast concrete fragments are
designed at hollow-core slab channels, at the staircase and near the installation
channels.
In the building „C” the following prestressed floor slabs were used at the levels
±0.00m, 4.08m, 9.65m, 8.06m: SP 32/8 spanning l = 11.05m, SP 26,5/5/6 spanning
l = 9.88 – 7.16m, SP 32/8 spanning l = 13.11m, SP 26,5-5/6 spanning l = 9.88m and

178
SP 40-14 B spanning l = 15.51m. The composite slabs Filigran were used over the
corridor passages and lecture halls.
In the building „D” the following prestressed floor slabs were used at the levels
±0.00, 4.08m, 8.16m and 12.14m: SP 26,5/5/8 spanning l = 9.80m, SP 32/14 B
spanning l = 13.14m as well as the reinforced concrete hollow-core slabs of the school
type S-320×120 and S-320×90. The composite slabs Filigran and reinforced concrete
cast floors were used, too.
In the building „E” the prestressed slabs 32/14B spanning l = 13.14m were used at
the levels ±0.00, +4.08m, +8.16m and +12.24m. The composite floors Filigran were
used, too.
All the prestressed slabs had the modular width of 120cm with a possible
narrowing, e.g. 70 cm to match another slabs.
The floors are supported on masonry walls made of perforated bricks, thickness 25
cm, reinforced with RC columns and bolts. Some floors are placed on cast concrete
beams. Communication between the floors is provided by two-flight RC stairs, lift
shafts or, in the building „C”, by one-flight steel stairs with an intermediate floor
paved with stone.
The roof was designed as ventilated with RC slabs, prestressed hollow-core slabs,
composite Filigran slabs or cast concrete slabs. Attic walls made of hollow brick
masonry were constructed as a support for closed channel slabs of various width and
span.
Over window and door openings the typical L19 lintels were used. The walls have
continuous foundations and the columns – pad foundations.

3. Macroscopic investigation of floor slabs

A macroscopic investigation of the constructed RC and prestressed floor slabs in

the analyzed structure was carried out.
In the case of some slabs (about 10%) the following damage types were observed:
- delamination of a concrete cover to an ordinary or prestressing reinforcement,
- longitudinal cracks along channels in the slabs (in zones of the minimal concrete
thickness, both at the bottom and the top side of the slabs),
- a significant porosity of the slabs at their top sides at some locations.
Additionally, drilled cores were taken for strength and laboratory testing.
The photographic documentation illustrating the observed damage as well as
construction errors, like an absence of irrigation openings for channels and porous top
surfaces of the slabs, was prepared. The detailed pictorial inventory of the damaged
slabs destined for repair with their positions marked in the plans of subsequent
building levels was done, too.
Below, some examples of damaged RC and prestressed hollow-core slabs are
presented in photographs.

179
Fig. 2. One of channels in a SP 26,5/6 slab in Fig. 3. Damaged concrete cover to prestressing
the segment „A” burst by ice cables in a SP 26,5/6 slab in the segment „A”

Fig.4. Damaged concrete at the bottom side of Fig. 5. Damaged concrete in channels of a SP
SP 26,5/6 slabs in the segment „A” 26,5/6 slab and exposed prestressing
reinforcement

Fig. 6. A damaged SP 32/8A-1105×70 slab in Fig. 7. Longitudinal cracking along one of channels
the segment „C”. Note the rough and damp slab in a SP 32/14B slab of L = 13.14m span in the
surface building „D”

180
Fig. 8. An example of channels damaged by ice Fig. 9. RC hollow-core slabs damaged by ice in
in RC ceiling slabs S-320×120 in the segment the corridor passage over the ground floor of the
„D” segment „D”

4. Laboratory testing of selected concrete properties

During the macroscopic investigation 2 RC hollow-core slabs and 12 prestressed

hollow-core slabs (damaged and undamaged) were selected and specimens were taken
to carry out the laboratory testing. The specimens were taken as drilled cores using a
HILTI boring rig with a crown drill φ51mm.
The testing included:
- determination of concrete compressive strength,
- macroscopic investigation,
- determination of apparent density,
- determination of adsorption,
- pH-factor measurement in central zones of specimens
- microscopic observations,
- determination of chemical and phase composition by X-ray, derivatography and
chemical testing.
Moreover, an assessment of set concrete composition including a sieve analysis of
aggregate separated during testing was carried out.

4.1. Destructive testing of concrete compressive strength

The current value of concrete compressive strength in RC and prestressed hollow-
core slabs was determined by pressing cylindrical concrete cores of the nominal
diameter 45.7mm drilled at randomly selected ceiling slabs.
The specimens taken from RC (series „B”) and prestressed slabs (series „A”) were
trimmed in the laboratory using a diamond saw to the height corresponding to their
diameter. Then they were dried and pressed in a testing machine connected to printer.
The obtained results for the specimen series „A” were analysed statistically taking into
account the specimen number exceeding 15. The series „B” specimens were assessed

181
using the compatibility criterion given in [9]. During the preparation of the results the
remarks included in [1]÷[10] were taken into account.
For the series „A” specimens the concrete class C35/45 was obtained for which
fck,cube = 38MPa, while for the series „B” specimens – the class C30/37 was determined
with fck,cube = 31MPa. This means that even after 180 days concrete in the prestressed
hollow-core slabs did not reach the class C40/50 (B50) designed and specified by the
slab producer. On the other hand, concrete in the RC hollow-core slabs exceeded the
class C25/30 (B25) declared by the producer.

4.2. Macroscopic investigation

This investigation was carried out for drilled concrete cores, fragments of cut slabs
and remaining fragments from the cut concrete specimens. Some of them were broken
and others – crushed in order to obtain information about the grain size of aggregate
used in concrete production.
A certain number of specimens was sprayed by the 1% solution of phenolphthalein
to check the level of concrete carbonation.
The macroscopic investigation for both concrete specimen series „A” and „B”
exhibited numerous air voids of various sizes, often interconnected. In some
specimens internal cracking (Fig. 10) and concrete mix segregation was observed. In
the series „A” specimens the coarser fraction consisted mainly of a granulated
aggregate. In both specimen series („A” and „B”) small amounts of lime (carbonate)
aggregate were found – Fig. 11.
Fractures of „A” and „B” concrete specimens had different colours indicating a
probability of different cement types used in the production.

Fig. 10. A concrete specimen „B” with internal Fig. 11. Evident carbonate aggregate reacting
cracking with 10% HCl solution

It was found that roughness and porosity on external surfaces of the specimens and
the slabs allow for water penetration, what leads to a freezing delamination and,
finally – to the destruction of the internal structure. Additionally, numerous internal
pores observed locally near surfaces facilitate water permeability.
Concrete fractures treated with the phenolphthalein solution indicated various
violet hues, what proved differing values of pH factor.

182
 Water solutions prepared to assess concrete pH factor for the specimens „B”
exhibited significant foaming while shaking. In the case of the specimens „A”
foaming was less intense.
An aggregate separated from the specimens „A” and „B” featured higroscopicity
and a tendency to form patches (Figs. 12, 13). These observations may indicate a too
large amount of chemical admixtures or their poor quality (e.g. exceeded expiry date).

Fig. 12. Aggregate fractions in concrete „A”. Fig. 13. Aggregate fractions in concrete „B”. Visible
Visible aggregate patches → large aggregate patches → a smaller higroscopicity than
higroscopicity in concrete „A”

4.3. Testing of apparent density

Within this testing carried out for the concrete specimens „A” and „B” mean
volumetric density was determined. For concrete taken from the prestressed hollow-
core slabs the mean apparent density of γ=2413 kg/m3 was found, while in the case of
the RC hollow-core slabs – γ=2317 kg/m3. Hence, both concrete types fulfilled the
standard requirements [9] in this respect.

4.4. Adsorption testing

Weight adsorption was determined for both series of concrete specimens. The
resulting values of 3.01% for the specimens „A” and 4.17% for the specimens „B”
indicate that in both cases the standard requirements [9] concerning concrete exposed
to a direct action of atmospheric factors were fulfilled (weight adsorption < 5%).

4.5. Testing of pH factor

The pH factor for a water extraction from the drilled cores (20% of the solid phase
in the suspension) was determined.
The measurements of pH factor were carried out using the pH-meter N517OE with
a combined glass electrode. The pH values range from 10.10 to 11.6 for concrete „A”
and from 10.30 to 11,27 for concrete „B”. They differ significantly from typical values
for fresh concrete [12-13].

183
4.6. Testing of concrete morphology and microstructure
Microscopic investigation was carried out in order to determine morphology,
structure and types of chemical compounds formed in concrete as well as types of
additives used. Both types of concrete „A” and „B” were analyzed according to [11].
The microscopic investigation was also aimed at an assessment of possible future
chemical reactions in the analyzed concrete types.
The testing was carried out in accordance with generally accepted principles for
microscopy. The scanning microscope VEGA TS 5135 MM with a X-ray microprobe
was used for the following parameters:
- acceleration voltage – 25 kV,
- low vacuum – 10 Pa,
- reflected electrons BSE.
The specimens were not covered with films. Photographs of characteristic fragments
of the specimens were taken during the testing. Some results are presented in Figs. 14-
17.
X-ray microanalyses EDX were also carried out. Some examples of reaction
products in the tested concrete types are presented in Figs. 18 and 19.

Fig. 14. Microscopic view of concrete „A”. Fig. 15. Microscopic view of concrete „B”.
Tight structure. Visible portlandite, ettringite, Visible portlandite, gel and crystallised C-S-H
gel and crystallised C-S-H phases and small phases, products of pozzolanic reaction, small
amounts of loose fly ash with products of amounts of ettringite.
pozzolanic reaction on surfaces.

184
Fig. 16. Microscopic view of concrete „B”. In Fig. 17. Microscopic view of concrete „A”. Tight
the central fragment an aggregate with surface structure. A void with gas pore (chemical
delamination, visible products of reactions in the admixture used). Visible ettringite, gel and
aggregate-paste interface. Cracks in the central crystallised C-S-H phases, portlandite.
fragment – at the grain border.

Fig. 18. EDX patternless microanalysis of reaction products in concrete „A” (in the vicinity of a gas pore
void)

185
Fig. 19. EDX patternless microanalysis of aggregate from concrete „A”

It was found during the analysis of the investigated fragments and the photographs
that the following compounds were present in the concrete specimens „A” and „B”:
- portlandite,
- calcite,
- gel and small crystalline tobermorite phases,
- ettringite,
- insignificant amount of fine fly ash,
- relic clinker grains – alite and belite.
In the case of concrete „B” more aggregate grains with delamination and cracks in
the microstructure were observed. In both concrete types a certain amount of lime
aggregate was found, slightly larger in concrete „B”. The internal structure of concrete
„A” was tighter. Some porosity due to chemical admixtures was observed.
The EDX patternless microanalyses carried out using the X-ray microprobe for
concrete „A” exhibited presence of oxygen, silicon, aluminium, calcium, sulphur, iron,
magnesium, sodium and potassium. In the case of concrete „B” calcium, oxygen,
silicon, aluminium, sulphur, iron, magnesium, sodium and potassium were found.

4.7. Testing of chemical and phase composition

The chemical composition of concrete was determined using the standard
procedure – [15] and [16].
The phase composition of concrete was determined using the X-ray diffractometer
BRUKER D8 ADVANCE. Diffraction patterns for the concrete specimens „A” and
„B” were obtained and the phase identification was carried out. The powder
diffraction patterns were prepared using the CuKα1 radiation with the Bragg angle
ranges 2Θ: 5° to 70°, 14o to 54o and 14o to 70o. The measurements were carried out at
goniometer increments of 0.01° and an impulse sampling time of 4.5s at every step.
Radiation was registered using a silicon strip detector.

186
 The concrete specimens („A” and „B”) were subjected to a thermal analysis to
check for possible lime addition and to assess contents of compounds attached during
cement setting and hardening. The testing was carried out using the derivatograph
Setsys TG-DSC15 made by SETARAM France with the following conditions.
Testing parameters:
Temperature range: 20-1000oC
Temperature rate : 10oC/min
Reference compound: A12O3
Oven atmosphere: N2
As a result of these X-ray testing the diffraction patterns for the averaged concrete
specimens were obtained. They formed a base for the phase identification. In the
presented patterns (Fig. 20) only those phases were marked for which the pattern fit
was reliable. The given values of the phases contents cannot be treated as absolute. In
the analysis no effects of texture and possible partial substitution in the phase
crystalline mesh were considered; they might influence the reflection intensity.

Fig. 20. Comparison of diffraction patterns for: A – concrete specimen „A”, B – concrete specimen „B”.

187
The following letters set in alphabetic order were used in the diffraction patterns to
distinguish between the reflections corresponding to the different phases:
CA - 2CaO·Al2O3·8H2O; CA’ - CaO·Al2O3·10H2O;
CA’’ - 4CaO·Al2O3·xH2O; CS - 2CaO·SiO2·5H2O;
CF - 3CaO·Fe2O3·6H2O; E –ettringite-3CaO·Al2O3·3CaSO4·32H2O;
E’ - ettringite- 3CaO·Al2O3·3CaSO4·31H2O;
G- gypsum- CaSO4·2H2O; G’ - gypsum- CaSO4·0,67H2O;
Gs - gismondite- CaO·Al2O3·2SiO2·4H2O;
K– calcite-CaCO3 ;M –monosulphate-CaO·Al2O3·CaSO4·13H2O;
P– portlandite- Ca(OH)2; D - dolomite- CaMg(CO3)2;
Q- α quartz- SiO2; Bm –brownmillerite-2CaO⋅Al⋅FeO3;
T– tobermorite-5CaO·6SiO2·2,5H2O
X- reflections due to several possible different hydrated aluminates, silicates or
calcium aluminium silicates present in small amounts.
The phase composition of both analyzed concrete types is given in Table 1.
Most probably the phase depicted as „X” in the diffraction patterns represents anortite
(Ca, Na) (Si, Al)4O8. Its content was not determined due to the difficulties in fitting
its proper structural model.
Table 1. Assessment of concrete phase composition
Content [%]
Phase „A” „B”
Q – quartz 65.8 69.5
K - calcite 14.3 13.4
P - portlandite 7.4 5.7
A - alite 4.4 4.4
B - belite 2.8 1.2
Bm - brownmillerite 2.7 2.4
E+T ~1.8 ~2.0
Dolomite 0.8 1.3

The combined content of ettringite and tobermorite was only roughly assessed
basing on the intensity of just one line corresponding to a low angle value.
Besides that, the diffraction patterns indicate presence of some crystalline phases
featuring a too small number of reflections to attribute them to any specific phase.
Their identification was not possible because their location did not correspond to any
identified cement phase.
The values of the percentage content given in Table 1 are referred to the analysis
carried out by the program PowderCell, where unidentified crystalline phases and the
ones present in trace amounts were not taken into account.

188
 The results in Table 1 indicate that the content of cement phases (portlandite,
calcite, alite+belite) was slightly higher in concrete „A”. On the other hand, the
content of quartz and dolomite was lower.
The results of derivatography for the averaged concrete specimens „A” and „B”
are presented in Figs. 21 and 22 as well as in Table 2.

Table 2. Description of thermal effects from DTA curve and the corresponding mass decrease
Temperature range of endothermic effects –
Accumulated mass
Specimen Effect DTA curve [°C]
decrease [%]
Tp Tmax Tk Mass decrease [%]
„A” I 50 – 98 – 185 1.8
II 420 – 450 – 480 1.2 10.8
III 680 – 755 – 785 4.4
„B” I 48 – 92 – 140 0.9
II 410 – 445 – 480 1.3 10.5
III 680 – 755 - 780 4.6

Fig. 21. Derivatogram for the concrete specimen „A”

189
Fig. 22. Derivatogram for the concrete specimen „B”

In the DTA curve for both specimens three endothermic effects in a close
temperature range were observed (Table 2).
Effect 1 in the temperature range (50-98-185)°C for the specimen „A” and (48-92-
140)°C for the specimen „B” is related to the loss of water bound in hydrated calcium
aluminates, sulphur aluminates and silicates. This loss was twice as big in the case of
the specimen „A”.
Effect 2 in the temperature range (420-450-480)°C for the specimen „A” and
(410-445-480)°C for the specimen „B” is related to the decomposition of Ca(OH)2.
The mass losses for both specimens were similar.
Effect 3 in the temperature range (680-755-785)°C for the specimen „A” and
(680-755-780)°C for the specimen „B” is related to the decomposition of calcium
carbonates. In the case of the specimen „B” the mass loss (%) was slightly higher than
for the specimen „A”.
Both specimens kept the powder form during the testing. No melting was
observed.
The mass loss in the temperature 580°C for the specimen „A” was 5.30%, and for
the specimen „B” – 4.94%.
In order to determine exactly the amounts of compounds which might influence
properties of tested concrete, chemical analyses of the concrete samples taken from the

190
hollow-core slabs denoted as „A” and „B” were carried out. The samples were
prepared by quartering method from the ground concrete and the testing was done in
accordance with the standards PN-EN 196-2, PN-EN 196-21, PN-EN 196-21/AK
„Methods of cement testing” as well as the methods of the classical chemical analysis.
The results obtained [%] are presented in Tables 3 and 4.

Table 3 Total SO3całk, MgO, Na2O, K2O and Na2Oeq contents in bulk concrete
Concrete Content in the averaged concrete sample [%] *
sample SO3całk MgO Na2O K2O Na2Oeq: = CaO SiO2 Fe2O3
Na2O+0.658K2O
„A” 0.55 0.96 0.099 0.44 0.39 27.9 55.08 1.94
„B” 0.41 0.97 0.097 0.43 0.38 28.2 54.04 1.91
* the results are mean values from 6 tests

Table 4 Chlorine contents in concrete

Concrete Cl– content in the averaged concrete sample [%]

„A” 0.20
„B” 0.28

The percentage contents of the given compounds in the averaged concrete samples
are different for both concrete types.
The Cl– content (0.20% and 0.28%) in concrete bulk „A” and „B” fulfils the
requirements of the standard [9] (Table 10 therein). Concrete „A” meets the
requirements for prestressed concrete (Cl 0.20) and concrete „B” meets the
requirements for steel reinforced concrete or concrete with another metal elements
(Cl 0.40).

5. Assessment of set concrete composition

The true mix composition of concrete „A” and „B” used in the production of the
hollow-core slabs was not available. That is why an assessment of the compositions
was carried out according to ITB Instruction No. 277/86 [11].
This assessment allows to relate the determined compound contents of SO3całk, MgO
and Na2Oeq to the approximate cement amount used in both concrete types and to find
if the respective quantities fall in the admissible ranges or if they pose internal
corrosion hazards should appropriate conditions occur in future.
The lime aggregate amounts in both concrete types were similar: 4.3 to 4.8% in
the entire aggregate amount (104-111 kg).
The aggregate size in concrete „A” ranged from 0-16mm with 47.5% of the 0-
2mm fraction, 25.83% of the 2-8 mm fraction and 26.67% of the 8-16mm fraction.
The aggregate size in concrete „B” ranged from 0-16mm with 34.59% of the 0-
2mm fraction, 34.16% of the 2-8 mm fraction and 31.25% of the 8-16mm fraction.

191
 Concrete „A” featured a larger amount of granite aggregate (fraction 2-16mm) and
doubled amounts of very fine fractions 0-0.125mm and 0.125-0.25mm when
compared to concrete „B”.
The results of an approximate recalculation of the determined contents of chemical
compounds onto concrete in [kg] and relating the values to the ones calculated from
the probable cement content are presented in Table 5.
The SO3cal content in analyzed concrete ranges from 2.48% („B”) to 3.20% („A”)
when related to the designed cement mass. These values meet the requirement given in
Table 3 in the standard [12] for all the cement types; the standard limit is 3.5% - 4.0%
depending on the strength class.

Table 5. Assessment of SO3całk., MgO, Na2Oeq contents in concrete

Content in concrete
Concrete type Related to:
SO3całk. MgO Na2Oeq
„A” Concrete mass 0.55 ⋅ 2413 0.96 ⋅ 2413 0.39 ⋅ 2413
- apparent density [kg] 100 100 100
γ=2413 kg/m3 ≅ 13.27 kg ≅ 23.16 kg ≅ 9.41 kg
- cement content Cement mass 13.27 ⋅ 100 23.16 ⋅ 100 9.41⋅ 100
according to ITB [kg]
277/86 415.04 415.04 415.04
415.04 kg/m3 ≅ 3.20 % ≅ 5.58 % ≅ 2.27 %
„B” Concrete mass 0.41 ⋅ 2317 0.97 ⋅ 2317 0.38 ⋅ 2317
- apparent density [kg] 100 100 100
γ=2317 kg/m3 ≅ 9.50 kg ≅ 22.47 kg ≅ 8.80 kg
- cement content Cement mass 9.50 ⋅ 100 22.47 ⋅ 100 8.80 ⋅ 100
according to ITB
277/86 [kg] 382.54 382.54 382.54
3 = 2.48 % = 5.87 % = 2.30 %
382.54 kg/m

The determined MgO contents of 5.58 („A”) and 5.87% („B”) related to the
cement mass exceed the admissible value of 4% according to [13]. For many years in
the standards of many countries the admissible value of MgO content in cement was
as high as 5%.
The alkali contents (Na2Oeq) of 2.27% („A”) and 2.30% („B”) related to the
cement mass exceed the admissible value of 0.6% according to [13].
The sulphur compounds recalculated to the total amount of sulphur, MgO and Na2Oeq
contents in the analyzed concrete types can come from aggregate and their
contamination or from cement and filler.

6. Analysis of testing results and assessment of ceiling slabs state

It was found during the macroscopic investigation of the RC and prestressed

hollow-core slabs used to construct the buildings „A”-„E” that the state of concrete

192
therein was not uniform. The upper sections of the slabs were porous, rough as if
smoothed with wooden boards. The lower sections were also rough, uneven, what
might indicate unsatisfactory cleaning of moulds used in the production.
It was also found that some slabs were lacking openings for channels draining. It was
the case in all the school type RC slabs and some of the prestressed concrete slabs.
This failure combined with porous upper surfaces led to water penetration to the
channels. Water deposited there froze and burst the channels what was indicated by
longitudinal cracking of channel walls in their thinnest part. In the extreme cases the
concrete cover to rebars (in the RC slabs) or to prestressing cables (in the prestressed
slabs) was delaminated.
The macroscopic investigation proved that some slabs underwent damage even
without ice presence in channels, e.g. the slab SP 32/8A-1105×70 over the 1st floor
(level +8.06) of the building „C” in the field limited by the axes B27/B26-B22-A10-
A9. It cannot be explained otherwise than by the porous structure of concrete,
presence of carbonate aggregate and overdosing or using the outdated aerating
substances and plastifiers, which weakened the concrete structure. This is confirmed
by the results of the compressive strength testing for concrete taken from this slab.
Concrete from the upper zone had a much smaller strength reaching 51.5MPa than the
one from the deeper zones – 92.5 MPa. The degree of damage in the prestressed
concrete hollow-core slabs is unprecedented. The lack of tightness in class C35/45
prestressed concrete used in the slab production is also hard to explain. The tightness
of such concrete should be absolute and the problem of precipitation water migration
through the channel walls should not exist at all.
Additionally, the analysis of compressive strength values for the concrete specimens
taken from the RC and prestressed concrete hollow-core slabs proved, that concrete
lacked uniform quality and strength. A very high value of standard deviation in this
testing disqualifies these products and poses serious doubts concerning the quality
control system in the production plant.
Concrete in the RC slabs was visually better and its strength and class deviated only
slightly from the class of concrete used in the prestressed slabs despite the use of
broken granite aggregate, microsilica and cement of a different quality and strength.
However, due to the production process, concrete in the upper zones of these slabs
was significantly less tight than in the case of the prestressed slabs. In some slabs
longitudinal cracks were observed on the top side along channels. Hence, the presence
of water or ice in the majority of channels what led to damaging of vast areas of floors
in the buildings „C” and „E”.
High content of magnesium compounds and alkali in the averaged concrete samples
and presence of carbonate aggregate indicate that in an appropriate humidity and
temperature alkali expansion might occur. This is confirmed by the damage type to
some of the prestressed slabs.
It was found that all the floor spans where RC hollow-core slabs (65) were used had to
be reinforced. The appropriate method is described in Section 7.1.

193
Some of the damaged prestressed hollow-core slabs (46) required reinforcing, too.
This was the case for the slabs where the concrete cover to prestressing cables burst or
in the case of large delaminated sections of concrete between channels. The repair
method is described in Section 7.2.
The analysis of static load bearing capacity for the prestressed slabs with longitudinal
cracking along channels pointed out, that after an appropriate protection of these
elements against moisture influence the elements may be left without reinforcing.
One more surprising feature is a relatively low and varying pH factor (about 10-11) in
the RC and prestressed hollow-core slabs. In fresh concrete this value should be about
13. Such a situation may be related to the type of chemical substances and additives
used in the concrete production. This low value of pH factor may cause undesired
phenomena during the service period of the prestressed slab structure. It is so because
with concrete aging a possible carbonation reduces pH factor further. When it sinks
below the value of 9 concrete loses its protective properties with respect to steel and
becomes a corrosion factor. It may lead to a permanent damage or even a failure state
of the structure. This was the reason for the recommendation to protect concrete using
corrosion inhibitors at bottom surfaces of all the slabs and top surfaces of the slabs SP
32 and SP40 with prestressing cables at their upper section (see Section 7.3).

7. Repair recommendations

7.1. Repair of RC hollow-core slabs

7.1.1. Repair of RC hollow-core slabs spanning not more than 320 cm
The following actions in order to repair and reinforce (see [14]) the RC hollow-
core slabs of the school type S - 320×120 used in the building „D” over the basement
(level ±0.00) in the fields limited by the axes A29-A30-B11-B16 and A29-A30-B20-
B31 were recommended
− shore all the spans at 1/3 and 2/3 of the length using truss girders and regulated
spacers PERI,
− remove loose concrete pieces from bottom surfaces,
− cut longitudinal grooves of 15cm width and entire clear length at every second
channel using a grinder with a diamond wheel,
− crush the floor clamping at the walls where they meet grooves to 10 cm depth,
− clean and remove dust from the grooves,
− attach reinforcement made of A-I or A-IIIN steel φ6 stirrups spaced at 15cm and 4
longitudinal A-IIIN steel φ12 rebars placed at the stirrup corners (see Fig. 23),
− lay Ecophon plates of 4.9 mm thickness covered with strips of PE foil of 0.2mm
thickness on the top side of the slabs,
− attach an additional upper reinforcement for the RC slab instead of the designed
layer of jointless floor of 5 cm thickness; the reinforcement should have the form
of a steel grid of A-IIIN steel φ6 bars spaced at 20×20cm (see Fig. 23),

194
− cast the concrete slab and the reinforced ribs using vibrated C20/25 class concrete;
carefully level the concrete surface to form the direct substrate for the designed
terracotta tile floor;
− protect the exposed reinforcement a the bottom surface of the slabs with the
concrete repair mortar type PCC/SPCC of SCHOMBURG or SIKA make
consisting of: binding layer, anti-corrosion protection of reinforcement, reprofiling
layer, levelling layer and acrylic resin protective cover against concrete
carbonation,

Fig. 23. Repair method for the RC hollow-core slabs of the school type and span up to 320cm in the
building „D” in the fields limited by the axes A29-A30-B11-B16 and A29-A30-B20-B31

− remove regulated spacers and truss girders PERI,

− attach the remaining installations,
− glue Ecophon plates of 4.9 mm thickness at the bottom surfaces of the RC floor
slabs,
− attach the false ceiling.

7.1.2. Repair of RC hollow-core slabs spanning not more than 560 cm

The similar actions as given in Section 7.1.1 were recommended in order to repair
and reinforce the RC hollow-core slabs of the school type S340×90, S-440×90, S-
440×120, S-550×90, S-550×120, S-560×90, S-560×120 used in the building „C” over
the basement (level ±0.00) in the fields limited by the axes A15-A13/A12-B29-B19
and over the first floor (level +9.65) in the fields limited by the axes A15-A13/A12-
B29-B19. The scheme of the reinforcing is given in Fig. 24.

195
Fig. 24. Repair method for the RC hollow-core slabs of the school type and span up to 560cm in the
building „C” in the fields limited by the axes A15-A13/A12-B20-B29 – the floor over the basement
(±0.00) and the 1st floor (level +8.06)

7.2. Repair of prestressed hollow-core slabs

7.2.1. Repair and reinforcing of the damaged slab SP 32/8A 1105× ×70
It was recommended to repair the slab SP 32/8A-1105×70 in the building „C” over
the first floor (level +8,06) in the field limited by the axes B27/B26-B22-A10-A9 in
the following way:
− protect the floor below in the field B27/B6-B27-A10-A9 against possible damage
due to falling concrete debris and increased loading by shoring and laying the
foamed polystyrene sheets,
− shore the damaged slab using boarding plates, truss girders and regulated spacers
PERI,
− cut off the extreme slab fragment including the rib and one half of the first channel
using a grinder with a diamond wheel; before the longitudinal cut make transverse
incisions at every 200 cm to cut the first prestressing cable,
− remove debris from the floor boarding,
− crush both floor clamping zones at the axes B27/B26 and B22 forming the socket
of the width ~30cm, depth of 15cm and height of 32cm,
− remove dust and clean the cut channel and the sockets formed at the supports,
− attach the profile NP [300 of the length equal to the slab clear span +2×150mm
with the A-I or A-IIIN steel φ6 stirrups spaced at 12cm through which 10
longitudinal A-IIIN steel φ12 rebars of the length equal to the NP[300 length
should pass; NP [300 should be placed on the low-shrinkage mortar Addiment VB
55-3N of SIKA make (see Fig. 25),

196
Fig. 25. Repair method for the damaged prestressed slab SO 32/8A-1105×70 in the building „C” over the
1st floor (level +8.06) in the field limited by the axes B27/B26-B22-A10-A9

− lay the Ecophon plates of 4.9mm thickness covered with PE foil of 0.2mm
thickness on the existing slabs,
− moisturize the rim of the first half-channel and the supporting socket for the steel
beam after placing the grid reinforcement of the concrete slab; the grid to be made
from A-IIIN steel φ6 rebars spaced at 20×20cm,
− cast concrete over the prestressed slabs and in the first rib of the floor slab forming
a composite steel-RC element using the vibrated concrete of C35/45 class; it was
recommended to carefully level the concrete surface to form the direct substrate
for the designed terracotta tiles,
− remove the boarding plates, supporting girders and regulated spacers when
concrete reaches the strength of 0.7 f cG, cube ,
− glue the Ecophon plates of 4.9mm thickness at the bottom surface of the
prestressed hollow-core slabs,
− attach the steel railing using the embedded [ 300 profile,
− lay the terracotta tiles,
− attach the false ceiling.

7.2.2. Repair and reinforcing of the damaged slabs SP 26,5/6-560× ×120

It was recommended to repair the damaged slabs SP 26,5/6-653×120 in the
building „A” over the first floor (level +8,16) in the fields limited by the axes B4-B7-
A14-A15, B5-B7-A24/25-A27/28, B4-B7-A31-A32, B4-B7-A33/34-A34/35 and over
the basement in the field limited by the axes B4-B7-A46-A47 in the following way:

197
− protect the floors below in the above mentioned fields against possible damage
due to falling concrete debris and increased loading by shoring and laying the
foamed polystyrene sheets,
− shore the damaged slab using boarding plates, truss girders and regulated spacers
PERI,
− cut off the extreme slab fragment including the first and the second rib and one
half of the adjacent channel using a grinder with a diamond wheel; before the
longitudinal cut make transverse incisions at every 200 cm to cut the prestressing
cables,
− remove debris from the floor boarding,
− crush both floor clamping zones at the axes B4 and B3 at the places designed as
supports for steel beams made from the profile I240 PE at the depth of 15cm and
height of 27cm and at the axes B7 and B1 through the entire thickness (to the
exterior side of the external walls),
− remove dust and clean the cut channel and the sockets formed at the supports,
− attach the profiles I240 PE of the length equal to the slab clear span +2×150mm
wrapped with the steel mesh; the profiles I240 PE should be placed on the low-
shrinkage mortar Addiment VB 55-3N of SIKA make,
− attach the A-I or A-IIIN steel φ6 stirrups spaced at 12cm between the profiles I240
PE welding them to the profile webs; then pass through the stirrups 8 longitudinal
A-IIIN steel φ12 rebars – 4 on the top side, and 4 on the bottom side, see Fig. 26,
− lay the Ecophon plates of 4.9mm thickness covered with PE foil of 0.2mm
thickness on the existing slabs,
− moisturize the rims of the adjacent channels and slabs after placing the grid
reinforcement of the concrete slab forming the direct substrate for terracotta tiles
or a roll floor finish; the grid to be made from A-IIIN steel φ6 rebars spaced at
20×20cm,
− cast concrete over the prestressed slabs and ribs of the prestressed floor slab
reinforced with steel beams and rebars using vibrated concrete of C35/45 class; it
was recommended to carefully level the concrete surface to form the direct
substrate for the designed terracotta tiles or the roll floor finish,
− remove the boarding plates, supporting girders and regulated spacers when
concrete reaches the strength of 0.7 f cG, cube ,
− glue the Ecophon plates of 4.9mm thickness at the bottom surface of prestressed
hollow-core slabs,
− lay all the installations under the ceiling,
− make the plastering layers,
− attach the false ceiling,
− lay the terracotta tiles or the roll floor finish.

198
Fig. 26. Repair method for the damaged prestressed slabs in the building „A” over the 1st floor between
the transverse axes A34-A35, A31-A32, A24-A27, A14-A15 and between the axes B1-B3-A36-A37 and
the floor over the basement between the axes B4-B7-A47-A46

7.3. Impregnation of concrete in the prestressed slabs

Because concrete in the prestressed slabs featured a relatively low value of the pH
factor (between 10 and 11) it was recommended to protect the prestressing cables by
an additional in-depth impregnation using corrosion inhibitors, e.g. SIKA Ferrogard
903.

199
 The prestressed slabs SP 26,5 where the prestressing cables are placed in the lower
section were to be impregnated from the bottom surface, while the slabs SP 32 and SP
40 with the cables at both, upper and lower section were to be impregnated on both
surfaces. The consumption of the inhibitor should range from 0.3 to 0.5 kg/m3
depending on the substrate.

7.4. Repair of concrete losses in prestressed hollow-core slabs

It was recommended to fill the existing losses of concrete in the prestressed slabs
using the manual repair system for concrete PCC III of Sika make which includes the
following materials:
− the one-component mortar to form the adhesive layer protecting the exposed
rebars based on the cement modified with polymers and microsilica Sika
MonoTop 610 additive (consumption of 2kg/m2 of dry mortar for two layers of
1mm combined thickness, in the adhesive layer – 1.5-2 kg/m2 of dry mortar),
− repair mortar to reprofile the loss based on the cement modified with polymers
with microsilica additive, reinforced with fibres, with 0-2mm aggregate – SIKA
MonoTop 612 or with 0-4mm aggregate – SIKA MonoTop 612 (consumption of
18.5-19.5kg/m2 of dry component with 1cm thickness),
− filler mortar based on the cement modified with polymers with microsilica
additive – SIKA MonoTop 620 to smooth the repaired surfaces (consumption of
1.3-1.6kg/m2 of dry component depending on the surface structure).

8. Summary

The damage types presented in this paper, which were found in the RC and
prestressed hollow-core slabs result from failure of the slab producer to assure proper
control procedures during the concrete production and concrete placing in moulds.
This led to significant discrepancies in concrete strength in prestressed elements,
porosity, lowered values of pH factor and large roughness of surfaces. The use of
carbonate aggregate (4.3÷4.8%) in concrete and increased alkali content combined
with critical permanent access of water evoked swelling processes of some concrete
contents what led to bursts in some elements. The destruction of some of the slabs was
sped up by the lack of factory-made openings for channel draining what evoked bursts
due to freezing water which penetrated through the porous slabs to the channels.
The repair works described in Section 7 concerning the damaged RC and
prestressed slabs were carried out in their entire range. The contractor is currently
terminating the finishing and furnishing works. The current state of some slabs after
repair is presented in Figs. 27 and 28.

200
Fig. 27. The repaired slab SP 32/8A spanning l = Fig. 28. The repaired hollow-core slabs S
11.05m in the building „C” – see Figs. 6 and 25 spanning l = 3.2 m

Bibliography
1. L. Czarnecki Ed.: Concrete in the standard PN-EN 206-1 – a commentary to the
standard, Wydawnictwo Polski Cement, Kraków 2004 (in Polish).
2. Ł. Drobiec, R. Jasiński, A. Piekarczyk: Diagnostics of RC structures,
methodology, in-situ tests and laboratory tests for concrete and steel, PWN,
Warszawa 2010 (in Polish).
3. S. Kaufman Ed.: Concrete structures, vol. 1, part 2, Concrete technology, Arkady
1972 (in Polish).
4. A.M. Neville: Properties of concrete, Wydawnictwo Polski Cement, Kraków 2000
4th Edition (in Polish).
5. PN-EN 12390-3:2002 Concrete testing, part 3 Compressive strength of specimens
used in strength testing (Polish standard).
6. PN-EN 12390-4:2001 Concrete testing, part 4 Requirements for testing machines
(Polish standard),
7. PN-EN 12524-1:2001 Concrete testing in structures. Part 1. Drilled cores, cutting
and testing of compressive strength (Polish standard),
8. PN-EN 13791:2008 Assessment of concrete strength in structures and concrete
products (Polish standard),

201
 9. PN-EN 206-1:2003, PN-EN 206-1:2003/A1:2005, PN-EN 206-1:2003/A2:2006,
PN-EN 206-1:1003/Ap1:2004 Concrete part 1 Producton requirements and
compatibility (Polish standard),
10. PN-ISO 3534-1:2002 Definitions and notation, part 1: General definitions in
probability and statistics (Polish standard),
11. ITB Instruction No. 277/86 Determination of composition of set concrete (in
Polish).
12. PN-EN 197-1:2002 Cement. Part I. Composition, requirements and compatibility
criteria for general purpose cements (Polish standard),
13. Peukert S.: General purpose and special cements. Polski Cement Sp. z o.o.,
Kraków 2000 (in Polish).
14. W. Buczkowski Ed.: Civil engineering vol. IV. Section V, Arkady Warszawa 2009
(in Polish).
15. Chaipanich A., Nochaiya T. 2010. Thermal analysis and microstructure of
Portland cement-fly ash-silica fume pastes. Journal of Thermal Analysis and
Calorimetry, vol. 99, p. 487-493.
16. Talero R., Rahhal V. 2009. Calorimetric comparison of Portland cements
containing silica fume and metakaolin. Journal of Thermal Analysis and
Calorimetry, vol. 96, p. 383-393.

202

View publication stats