Analysis of the mechanical characteristics of facade render samples retrieved in-situ António Armando Ortiz Soares Extended abstract Júri Presidente: Prof. Augusto Martins Gomes Orientadora: Profª. Inês dos Santos Flores Barbosa Colen Orientador: Prof. Jorge Manuel Caliço Lopes de Brito Vogal: Doutora Maria do Rosário da Silva Veiga Março de 2011
14
Embed
Analysis of the mechanical characteristics of facade …...Analysis of the characteristics of facade samples retrieved in-situ 1 1. Introduction and objectives Coating mortars (renders)
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Analysis of the mechanical characteristics of facade render samples
retrieved in-situ
António Armando Ortiz Soares
Extended abstract
Júri
Presidente: Prof. Augusto Martins Gomes
Orientadora: Profª. Inês dos Santos Flores Barbosa Colen
Orientador: Prof. Jorge Manuel Caliço Lopes de Brito
Vogal: Doutora Maria do Rosário da Silva Veiga
Março de 2011
Analysis of the characteristics of facade samples retrieved in-situ
1
1. Introduction and objectives
Coating mortars (renders) perform the function of protecting the walls from the degradation agents to which
they are subjected. It becomes necessary, then, to evaluate their service performance, so as to understand if
they can carry out the function attributed to them. For this purpose, the chemical, physical and mechanical
properties of the render under analysis must be studied.
In what concerns the mechanical characteristics of the renders, there are still some difficulties in their
assessment, for which reason the present work focuses on this matter, with particular detail on the evaluation
of the compressive strength of samples retrieved in-situ. In this manner, this work allows to deepen the
knowledge on the parameters to be considered in the compressive strength tests of render cores, with the
main objective of relating it with the compressive strength of normalised test specimens
2. Mechanical characteristics of coating renders
Several authors have tried to evaluate the compressive strength of coating mortars from samples retrieved in-
situ. In order to study the compressive strength of irregular samples of historic mortars retrieved in-situ, a
project was developed by LNEC (Magalhães & Veiga, 2006) in order to create a simple adaptation of the
method of determining the compressive strength defined by EN 1015-11 (CEN, 1999), to allow compressive
strength test of irregular samples in old buildings. The adaptation in question consists of producing a
confinement mortar, with a strength superior to that of the sample being tested, applied over the two parallel
faces of the irregular sample (Figure 1).
Figure 1 - Detail of samples prepared with the confinement mortar over the two more parallel faces of the samples
(Magalhães & Veiga, 2006)
As a means to evaluating the method, reference specimens were used, with dimensions 20 x 40 x 80 [mm3],
resulting from the cutting of specimens with 40 x 40 x 160 [mm3], to which the confinement mortar was
applied for the compressive strength test. The values of the compressive strength obtained with these
specimens were compared to those resulting from the testing of normalised specimens, and a high correlation
coefficient (R2 ≈ 0.96) was achieved for a linear relation.
Flores-Colen (2009) also studied the compressive strength of samples of render by comparing the
compressive strength of square cores with 50 mm length and 15 mm height retrieved in-situ, resulting from
the pull-off adhesion test with the compressive strength obtained through the testing of normalised
specimens of the same product, relating the obtained strengths through a power trend line with a correlation
coefficient R2 ≈ 0.78.
However, the results of the compressive strength of cores after the adhesion tests displayed variation
coefficients between 25-51% in some products, which the author attributes to the state of the core after the
process of heating to remove the metallic disk, to the control of the compressive strength testing machine, to
eventual variations of the thickness of the cores or to the sample’s state of degradation.
Extended abstract
2
The same author studied the possibility of directly relating the value of the compressive and tensile strength
with the adhesion strength. This study resulted in the graphic presented in Figure 2, where satisfactory
correlation coefficients are found for the relationship between adhesion strength and strength of normalised
specimens of the same products.
Legend: Rt - flexural tensile strength of normalised specimens; Rc - compressive strength of normalised specimens; Rc
ad -
compressive strength of samples after the adhesion test; fu - adhesion strength
Figure 2 - Relationship between adhesion strength and strength of the products tested (translated from Flores-Colen, 2009)
In the same study, mechanical tests were performed in campaigns conducted in-situ. It was then possible to
compare the results obtained in laboratory with real cases. In the case of the relationship between pull-off
adhesion strength and compressive strength extrapolated from the values of cohesive fracture (related to the
material’s straight tensile strength), a correlation coefficient of R2
= 0.95 was obtained between the
relationships established in laboratory and those verified in-situ.
Thus, the direct relationship between adhesion strength and compressive strength seems to be a good method
to evaluate the mechanical characteristics of the coating mortars. However, since fracture in the pull-off
adhesion test may be adhesive or cohesive on the substrate, part of the information pertaining to the internal
structure of the coating mortar is lost, since only the lower strength limit being known. As such, this method
is only possible when there are straight cohesive ruptures in the mortar, which occur mostly in “weak”
mortars.
Due to the few existing studies on the evaluation of the strength of mortars from samples of render, studies
related to the behaviour of concrete are referred to as a means to understanding some of the parameters
which may affect the value of compressive strength in test cores.
When a cylindrical concrete specimen is subjected to compression, it tends to expand sideways. However,
there is a friction force at the surface of contact between the plates of the testing machine and the specimen,
leading to lateral compression forces, which are responsible for the formation of a confinement area which
originates a cone at the moment of rupture, as can be seen in Figure 3. When that friction force is eliminated,
the lateral compression forces disappear, and a cracking rupture is reached. However, it is difficult to
eliminate the friction force, for which reason it is regarded as viable to consider the lateral restriction along a
certain length and consequent confinement area (Kim & Yi, 2002; Kim et al., 1998).
Analysis of the characteristics of facade samples retrieved in-situ
3
Figure 3 - Cylindrical concrete specimen subjected to a compression load (Kim et al., 1998; Kim & Yi, 2002)
Chung (1979), cited by Mohiuddin (1995), studied the effect of the h/d1 ratio on the compressive strength of
concrete cores and concluded that confinement, exerted by the plates of the compression machine on the
sides of the core, creates a triaxial state of tension in the concrete that originates an increase in the
compressive strength as the core’s h/d ratio decreases.
In the beginning of the 20th century, Connerman (1925) showed that the compressive strength decreases with
the increase of the h/d ratio of the specimen being studied, tending to infinite, for values of h/d lower than
0,5. The same author also defined an interval for the h/d ratio in cylinders, between 1,5 and 2,5, where the
difference in strengths is not significant, having as reference a specimen with h/d ratio = 2.
More recently, other researchers, while studying the effect of the relationship between height and diameter or
lateral dimension, also reached curves with power trend lines, as is the case of Leonhardt & Mönning (1977),
cited by Costa & Appleton (2008), as represented in Figure 4 where the reference specimen is the cube, for
which reason the reference value (of one) occurs for the geometric relation h/d = 1.
Figure 4 - Relation between h/d and the compressive strength of concrete specimens (translated from Leonhardt & Mönning,
1997, cited by Costa & Appleton, 2008)
When references are made between specimens produced with concrete identical to that used on site, their
strength will be different from the one determined in-situ, even if the possibility of achieving perfect cores,
identical in size to the test specimens, is considered, due to divergences in compacting and curing
(Indelicato, 1997).
Therefore, the need to estimate the strength of normalised specimens from cores retrieved in-situ emerged.
Thus, the concept of estimating the actual or in-situ strength referred to cubic specimens was used when
values of strength resulting from cores are used for structural calculations to represent the strength of the
existing structure, and the concept of potential strength estimate is used when there are doubts about the
quality of the concrete used (Concrete Society, 1976).
1 The h/d ratio defines the slenderness of the test specimen through the relation between height (h) and diameter or lateral dimension
(d)
Extended abstract
4
Thus, Equation 1 can be used (Concrete Society, 1976; BSI, 1983) to estimate concrete’s strength from cores
retrieved in-situ.
Eq. 1
Where:
Estimated strength [N/mm2];
Measured strength of the concrete core [N/mm2];
h/d ratio between the height (h) and the lateral dimension or diameter of the concrete core;
Experimental constant which depends on the type of estimate (D = 2.5 for the estimate of the actual
strength of a cubic specimen from a core retrieved horizontally in relation to concreting).
Another way of relating the strength of a normalized concrete specimen with a non normalized one is
through Equation 2, obtained through the theory basis of the non linear mechanical fracture of concrete,
presented by Kim et al. (1998) for cylindrical specimens and not taking into account the concrete’s strength
class, nor any restrictions on the maximum dimension of the aggregate used.
Eq. 2
Where:
Compressive strength of normalised cylinders [N/mm
2];
Compressive strength of a non normalised cylinder [N/mm2];
Cylinder height[mm];
Cylinder diameter [mm];
Ratio between height (h) and lateral dimension or diameter (d).
3. Characterisation of the experimental work
In the experimental phase, the following were applied over bricks: traditional multi-layer renders
(spatterdash, under coat and final coat), traditional control renders (a layer with the composition of the under
coat of a multi-layer render) and industrial renders with the thicknesses presented in Table 1.
Table 1 - Thicknesses of render
Type of
render
Traditional render (mm) Industrial
render
(mm) Spatterdash
Under
coat
Final
coat
Type 1 3 15 6 -
Type 2 3 25 6 -
Type 3 - - - 15
Type 4 - - - 30
Type 5 - 15 - -
Type 6 - 25 - -
For the production of traditional render, river sand was chosen as aggregate, and CEM II/B-L 32.5 N cement
was used as binder, presenting the compositions indicated in Table.
Analysis of the characteristics of facade samples retrieved in-situ
5
Table 2 - Characteristics of each layer of the traditional render used
Layers Ratio (volume) Water/cement ratio
Spatterdash 1:2 0.70
Under coat 1:4 1.00
Final coat 1:4.5 1.06
In the case of industrial render, a mineral regularization render was used, with the characteristics presented in
Table 3, to display a compressive strength value (5.06 N/mm2) close to the average indicated by producers of
general use industrial mortars (Flores-Colen, 2009).
Table 3 - Characteristics of the industrial mortar used, adapted from Flores-Colen (2009)
Components Cement (type I)
[%]
Water repellent
[%]
Water retainer and plasticiser
(cellulose ether) [%] Water/product ratio
Type Binder Admixture Admixture Characteristic
Quantity 15 - 25 0.10 – 0.50 0.05 – 0.10 0.17
Reference specimens (Figures 5 and 6) with the same characteristics of the renders applied (as a way of
simulating cores) and normalised cores, were used.
Figure 5 - Moulds with the under coat in fresh state and
spatterdash in hardened state
Figure 6 - Final aspect of a reference specimen
Upon application of the plaster and production of the specimens, the samples were stored in a curing
chamber of controlled environment up to the test date, with a temperature of 20 ± 2 ºC and relative humidity
of 65 ± 5 %, according to EN 1015 - 11 (CEN, 1999). For the initial curing, polyethylene bags were resorted
to for the durations indicated in Table 4.
Table 4 - Initial curing time in polyethylene bags
Type of
product or
render layer
Types 1 and 2 in brick or
reference specimens Types 3 and 4
in brick or
reference
specimens
Types 5 and 6
in brick or
reference
specimens
Specimens with
normalised size
Spatterdash Under
coat
Final
coat
Traditional
mortar
Industrial
mortar
Initial
curing
(days)
14 11 3 11 11 11 11
Compressive strength tests were performed at 7, 14, 28 and 90 days. The flexural strength test was
performed to obtain 2 half normalised specimens (NS) for compressive strength tests for each NS produced,
and the pull-off adhesion test to obtain cores for the compressive strength test, using a heating process to
Extended abstract
6
remove the pull-off metallic disk. Table 5 presents the general distributions of tests by age, according to the
type of section, specimen and products used.
Table 5 - General distribution of the compressive strength tests by test age