Mechanical Characterization of HPC with Limestone Filler

Technical literature on new generation of HS/HPC is available since a decade or more, with sev­eral examples of structural applications (Aitcin, 1999; Rols et al., 1999; Person, 1998; Toutanji et al., 1995; Rosati, 1999; Guerrini et al., 1999; Namiki, 2005; Nawy, 2001). The most significant mechanical parameters of the today well known HPC, refer to strength giving less attention to elastic modulus, or other characteristics related more to workability and durability. Because many characteristics of high performance concretes are interrelated, a change in one usually results in changes in one or more of the other characteristics. Consequently, if several characteristics have to be taken into account in producing a concrete for the intended application, each of these characteristics must be clearly specified in the contract documents. That is why the design documents of the two mentioned interventions should refer to a Guideline documents based on experimental investigations

Table 4. some significant figures of Limestone Concrete samples.



Compressive strength (MPa)

Flexural tensile strength (MPa)

Dynamic elastic modulus (MPa)

















Fig. 7. Technical problems related to silica fume utilisation: plastic shrinkage (constant water/binder ratio) (Cangiano, 2005).

on different mix design best suited for different structural elements of a complex whole building, like are, for examples, underground retaining elements, underground elements in presence of water, floor slab elements, linear horizontal elements, vertical “core” elements. In particular the paper wants to introduce, as mentioned before, the campaign of tests aiming to produce large and well defined spectrum of mix design for particular HS/HPC based on fine aggregate of limestone. LSC’s are classified as HS-HP concretes, with the advantages of being characterized by the complete absence of pozzolanic addition, i. e. silica fume, generally present as filling material in practically all type of HPC (De Larrard, 1993; Toutanji, 1995). Table 4 shows some first results of the mean significant mechanical characteristics of Limestone Concrete, and allows us to include Limestone Concrete in the range of HS/HPC, with advantages of being Self-Levelling and Rapid Hardening Concrete (Plizzari et al., 2003).

It is however recognized that the following problems are related to silica fume utilization: (a) the high cost (about 4-5 times the cost of cement), (b) the total shrinkage (plastic and hydraulic) of HPC containing silica fume (10%) may be greater than other HSC based on other mineral admixtures,

(c) HPC with silica fume may exhibit a high tendency to desiccation and hence to early micro­cracking, as a consequence the long term durability may suffer, (d) several researchers found a loss of compressive strength between 90 days and 4 years of concrete with silica fume. Figure 7 underline the problems related to shrinkage behaviour when utilizing silica fume and concrete mix designs according to European Code (EN No. 197).

The main task of the experimental research on new generation of HSC using limestone (without any content of silica fume), is to identify a size distribution curve of the system composed by cement and aggregates in order to produce HPC’s characterized by: (a) very good rheological properties and

Fig. 8. Strength development of two different mix of LSC.

Fig. 9. The favourable performance of LSC respect to hydraulic shrinkage.

(b) rapid development of mechanical performances. In this starting-up phase of the experimental research, just few mix have been proposed and tested with the primar goal to prove the sensibility of the mechanical characteristic to small variations of size aggregates in the early stage.

The graphs and table of Figure 8 show the upper and lower bound of the strength development performance due to slightly different mix, where it is clearly show the rapid hardening characteristic of Limestone Concrete mix.

The favorable performance of LSC respect to shrinkage behavior is summarized in the graphs of Figure 9 where LSC is compaired with different mixes of SCC.

Further development of the research will be focused on obtaining LSC with specific characteristic for the specific used, identified as follows:

– for beams, pillars and floor-slabs ^ LSC with properties of HSC and RHC;

– for “core” structure ^ LSC with high resistance to temperature and with high toughness;

– for conteining walls ^ LSC with properties of durability against salts, very low permeability, low hygrometric shrinkage.

In any case Limestone Concretes, also fiber added, should develop creep characteristics as good as those of SCC.


The paper starts from the assumption that the use of high-strength, high-performance concrete (HS-HPC) is rising up, not only for pillars, in high-rise buildings. The research program undertaken by Politecnico di Milano-Universities of Bergamo and Brescia, which starting-up phase results are here reported, is primarily focused on Limestone Concrete samples tests, to state the mechanical characteristics requested not only by new codes, but also by new building design philosophy in case of high-rise buildings. As a matter of fact, the research on the scaled structural elements (wind test), or/and on numerical models simulating structural elements under impact or blast loading condition will be a necessary completion in the overall knowledge for the optimization of strength and per­formance capacity of the new HPCSs and their selected use. The complete range of knowledge may produce a material highly competitive for the bearing structure of the high-rise buildings.

[1] the grouted pocket foundation (PF) solution showed the smallest strength degradation during the cycles. The column strength at the 5% drift cycles is equal to 82% the maximum column strength. Collapse was reached during the first cycle at 5.5% (168mm), due to buckling of the longitudinal bars. Specimen PF showed the most stable behaviour up to collapse among all of the specimen tested;

• all of the grouted sleeves (GS) column-to-foundation connections showed a considerable strength degradation during the cycles. In all cases, the strength of the column at the 5% drift cycle was approximately equal to 2/3 the maximum strength. The maximum strength at the 2.5% drift was approximately equal to 90% the maximum strength. The observed strength degradation is due to the progressive damage of the 20 mm grout layer existing between the precast column base and the foundation. This grout layer eventually crushed and was expelled

[2] Since the conventional and shear modes are not identified jointly, the overall matrices [C;k] and [Bk are not diagonal – only their principal sub-matrices exhibit this property.

[3] Although most of the novel GBT formulations have not yet been applied to other than isotropic and linear elastic members, this is a straightforward (even if time-consuming) task that is planned for the near future.

[4] No validation is presented for the elastic-plastic GBT-based results, as the authors know no commercial FEM code capable of calculating plastic bifurcation moments – e. g., neither Adina (Bathe, 2003) nor Abaqus (HKS, 2002) offer such possibility.

[5] It was found that the shear modes associated with non-linear warping (see Figure 9) do not contribute to the beram fundamental vibration modes. Nevertheless, they were included in all analyses.

[6] crushing in the angle legs denoted by the ultimate tension stress being reached in the angle legs parallel to the beam web (conservative);

• crushing in the beam web denoted by the ultimate tension stress in the beam web being reached (again, conservative);

• 20% increase above the ultimate bolt shear stress magnitude.

The initial stiffness, KC1 , is defined using the smaller two stiffness magnitudes. The first is based upon web yielding and the second is based upon angle leg yielding. Post-yield stiffness is defined rather arbitrarily using a 0.5% multiplier to account for moderate strain hardening in the material on the way to crushing.

It should be noted that the behavior of the supporting element (e. g. a column flange, a column web, a girder flange) is omitted. This is likely very important, but the complexity incurred through consideration of this behavior would render the analysis proposed intractable. Expected yield and ultimate tensile stresses for the materials are used as recommended in the GSA guidelines (GSA 2003). Further details of the formulation and example computations can be found in Foley et al. (2006).

The tension and compression response for the bolt elements are shown in Figure 6 for the W18x35 and W21x68 wide flange shapes, respectively. These wide-flange shapes are consistent with the 3-story SAC-FEMA Boston building assumed as the analysis prototype (FEMA 2000b; Foley et al. 2006).

[7] if the current stress point is on the memory surface and (———– ) t ■ dah > 0, this signifies a

daij aij J

loading case;

if the current stress point is on the memory surface and loading to unloading occurs;

if the current stress point is inside the memory surface, i. e. – a (aj) – Rmem < 0, it is then an unloading case.

For the loading case the spring stiffness of the ‘Kelvin-Voigt type’ elements is defined as a function of the equivalent stress, Et = E1(aeq). For the unloading case, it is assumed that Ei remains the same

during the entire unloading process, and its value is that of the Et at which the switch from loading to unloading took place.

[8] In the sequel we write „displacements”, „strains”, „stresses” and „loads” having in mind generalized variables taken usually in Structural Analysis.

[9] For structures made from the rigid-perfectly plastic material the kinematic unknowns Я, w should be replaced by their rates Я, w.

[10]I. e. STL files that will procure a successful build on a Rapid Prototyping machine.

[11] Exposure (15 ± 1) h at a temperature (+40 ± 2) °C

• Change within (60 ± 20) min to a temperature (-20 ± 3) °C; exposure (2 ± 1) h

• Change within (80 ± 20) min to a temperature (+70 ± 2) °C; exposure (4 ± 1) h

• Change within (60 ± 20) min to a temperature (+40 ± 2) °C

The pictures in Table 1 show selected specimens after 25 cycles given above. All specimens showed large-area debondings; partly microcracks occurred in the Lexan PC which were availed by the surface pre-treatment. The wished compound effect could not be achieved. The experimental program was changed with regard to plastic, surface pre-treatment and adhesive.

In Table 2 shows specimens of another series of experiments with changed adhesives and surface pre­treatment methods. Using the Pyrosil method had a very positive influence on the compound effect between glass and plastic. In this process a pre-treatment of the glass surface as well as of the plastic surface was carried out. The Pyrosil method produces an increased surface energy and thus improves the bonding.

After 25 cycles no microcracks or debondings occurred in polycarbonate Makrolon GP 099 (see Table 2, specimen 3-3). In the case of polycarbonate Lexan PC (see Table 2, specimen 4-3) microcracks oc­curred and – at the edge – debondings.

In addition to an acrylic adhesive two other adhesives were deployed (Epoxy and PUR). Both speci­mens showed large-area debondings leading to the destruction of the glass in the case of specimen 6-3. Both adhesives are probably not adequate to produce a durable compound between glass and Lexan. The Pyrosil method had no influence on the compound effect.