Confinement Effectiveness on The Strength and Ductility from Geopolymer Concrete

Geopolymer concrete made from fly ash as a substitute to Portland cement is a relatively new and environmentally friendly material. Various studies have revealed that geopolymer concrete is more brittle than Portland cement concrete. To increase its ductility in structures, geopolymer concrete needs to be confined. However, confinement models for geopolymer concrete are still limited. This paper compares some confinement models with stress-strain data obtained from laboratory tests. The confinement models investigated here were Ganesan et al. which was proposed for geopolymer concrete, and Mander et al. which were to Portland cement concrete. The comparison was conducted on the stress-strain action, including the characteristics from the ascending curve, the maximum stress and its corresponding strain, the descending curve. The outcomes show that the values of initial stiffness of geopolymer concrete simulated by models are higher than the actual values obtained from the tests and other confinement models. However, the maximum stress and the maximum strain values obtain from the models under estimate the test outcomes.


Background
Nowadays Portland concrete based concrete technology has widely advanced, not only in normal strength of normal concrete (f' c <50 MPa) that can easily be produced but also toward higher strength concrete or more popularly referred to as aloft strength concrete (50<f'c≤100 MPa) [1].High-strength concrete tends to be more brittle.Therefore, it has consequences for its use in seismic-resistant structural design, particularly in column structures where a higher volumetric ratio of confinement reinforcement is required compared to normal-strength concrete.Another alternative approach for enhance the ductility from aloft-strength concrete is by adding fibers such as steel fibers.The action from aloft-strength concrete with steel fibers has too been studied over the last three decades.This includes research on steel fiber-reinforced concrete at elevated temperatures [2-3].However, Portland cement-based concrete has one of disadvantages that leads to the greenhouse effect thus the earth gets warmer.

Problems
To overcome this environmental problem, another alternative is being developed that is geopolymer concrete, where the material uses environmentally friendly materials, Portland cement is returned by Fly Ash, and this material is efficient in terms of energy efficient use.The process from making geopolymer concrete with heating ± 60 o C. Therefore, making geopolymer concrete can reduce greenhouse gas emissions by only 20%.The mechanical action of geopolymer concrete is also continuously investigated with the essential aim of getting a comprehensive picture of its action hence it can be used as a planning reference [4][5].Variations in the use of geopolymer concrete forming materials are also continuing to get the optimum concrete mix [6].
One of the interesting geopolymer concrete action to discuss is its triaxial action.Understanding triaxial action is very important and useful to comprehend material comprehensively and be a determinant in the design of column structure components, the meeting of column structures [7].According to Haider [8], the application of lateral stress in geopolymer concrete with an active confinement system, will change the mode of collapse of geopolymer concrete from brittle to ductile properties.Muslikh et al. [9] conducted an experimental study in which they produced geopolymer concrete confined by hoop reinforcements.The outcomes of their research revealed a significant change in action, shifting from a brittle action in unconfined geopolymer concrete to a ductile action when geopolymer concrete was reinforced with hoop reinforcements at specific volumetric ratios.On the other hand, Herwani et al. [10] also conducted testing on confined geopolymer concrete, but with a square cross-section.Similar to the findings of Muslikh et al., ductile action was observed in confined geopolymer concrete with various reinforcement configurations.Herwani et al. proposed equations to the improvement from confined concrete strength for confined geopolymer concrete.However, because there is not much research on the problem of confinement on geopolymer concrete, the proposed confinement model is still very limited.Thus, the confinement model that is expected to be used as a planning reference is needed.

Objective
In this paper, the action from confined geopolymer concrete based on the experiments will be compared with existing confinement models.There are two confinement models discussed in this study.First, concrete confinement models made from normal concrete or Portland based cement.This model was proposed by Mander [11].Second, the geopolymer concrete confinement model.This model was reviewed and proposed by Ganesan [12].These two models will concentrate on the characteristics, effectivity in predicting the ductility, strength, and stress-strain action from confined concrete based on experimental outcomes.

Confinement Models
Table 1 is the confinement models that researchers have developed and also explanation of discussion of each model.As can be seen in Table 1, each model was envolved based on the outcomes from confined concrete tests with parameter reviews that were not the same.This basically will affect the form from the stress-strain curve which of course is also different from each other.The confinement model based on Mander was derived from full-scale column testing, where the primary parameters under consideration included the cross-sectional form (circular and square), the type from confinement (spiral and hoop), and the rate from loading.The Mander model uses an equation to illustrate the stress-strain curve from confined concrete.The form from the curve is influenced by the factor from r, where the value from r depends on the magnitude from the elastic modulus (Ec) and the secant modulus (Esec).Mander further revealed that the top stress formulation from confined concrete (f'cc) was derived based on Willam-Warnke's yield criteria.Mander employs the concept of the effectiveness of confinement to compute the lateral stress from effective confined concrete (f'l).On the other hand, Ganesan et al. carry out few modifications to the Mander confinement model above after conducting a series of experiments on confined geopolymer concrete.As explained in Table 1, Ganesan only revised the equations for r and Ec derived from his experimental outcomes, specifically for geopolymer concrete, where the outcoming values would affect the post-top curves from the confined concrete.
Table 1.Summary of confinement models.

Model
Equations Comments Mander The confinement model by Mander is developed from the outcomes of the testing concrete made from Portland cement.The concept of confinement effectiveness is proposed in this model (ke ).This model is accurate in predicting the stress strain from confined concrete for normal strength concrete.ke is effectivity of confinement Ganesan Ganesan proposed the confinement model for geoplymer concrete.Mostly the confinement model is the same as model by Mander but with the modification with the parameter r and E c .The parameter of k is the confinement index.

Materials
For decide the sensitivity from the confinement models, a stress-strain action comparison was performed, where the test specimens were taken based on the outcomes from the experiment on the confined geopolymer concrete.The concrete mix design for geopolymer concrete is presented in Table 2. Coarse and fine aggregates were sourced locally, specifically using Fly Ash from the waste of Tanjung Jati B Jepara power plant.Other materials included Na2SiO3 and an Activator 8M NaOH.To achieve good workability in geopolymer concrete, a superplasticizer of the Sikament NN type was added at a dosage of 2% of the cement weight [13].

Specimens design
The test specimens were designed as cylinders with the diameter from 100 mm and the height from 200 mm.They were reinforced with confinement hoops with varying diameters from 5.5 mm and 6 mm.All specimens were tested without the use of concrete cover.Each of these reinforcement hoops was subjected to tensile testing, outcoming in yield stresses of 514 MPa and 466 MPa.Details of the specimens with the examined parameters are presented in Table 3.The hoop reinforcement was tied by four longitudinal bars with the diameter from 6 mm. the hoop and longitudinal reinforcements were equipped with FLA5-11 strain gauges.In the calculation of restrained concrete stress, the stress experienced by the specimen was adjusted by subtracting the stress experienced by the longitudinal bars.This longitudinal stress was calculated based on data of the strain gauges and multiplied by the elastic modulus from steel [14].The top unrestrained concrete stress was defined as 85% of the stress in the 150 mm diameter and 300 mm high cylinder concrete specimen at 28 days.The reinforcement of the test specimen was illustrated in Figure 1.

Testing and data acquisition
The testing of specimens were conducted using a 2000 kN capacity compression testing machine with a deformation control system.To measure axial deformation, 50 mm capacity LVDTs were installed on both sides of the specimen.Data readings during the test were recorded automatically, and a Data Logger was used as a backup.The testing setup is illustrated in Figure 2.

Specimens action
The parameters reviewed included the top stress from the confined concrete (f'cc), the strain from the confined concrete top (ε'cc), and the strain from the confined concrete after the top at the time the f'cc value had falled by 15% (εc85).The outcomes of the analysis of the f'c c, ε'cc.and ε'c85 values from each model are shown in Table 3, their comparison of experiments is presented in Table 4.The compressive strength from geopolymer concrete cylinders (fc') at 28 days is 32 MPa.The strength improvement from confined concrete (K) is higher in specimens with tighter hoop spacing.In general, the yield stress (yh) of the hoop reinforcement does not have a significant impact on the K value when comparing specimens with the same hoop spacing (G1 vs. G2 and G3 vs. G4).However, based on the experimental outcomes, the yield stress of the hoop reinforcement has a remarkable effect on determining the top strain value from restrained concrete, ε'cc.From Table 3, it can be observed that ε'cc values are higher and more effective when lower yield stress of the hoop reinforcement is used.Conversely, ε'c85 values are more efficient and yield optimum outcomes when a higher yield stress of the hoop reinforcement is employed.

Experimental outcomes vs confinement models
All confinement models reviewed (Mander and Ganesan models) values f'cc turned out to be under estimation of the experimental outcomes, but the difference was not excessively large because it was still below 10%.Profoundly significant differences occur in the predicted value of ε'cc and εc85 values.As shown in Table 4, most of the differences are in the range above 40%.These outcomes indicate that the estimated ductility from the confined concrete like the Ganesan model for geopolymer concrete is still far from the experimental outcomes from the confined geopolymer concrete.
Hereafter, the differences from the stress-strain action from the confined concrete shown in Figures 1 and 2 will be explored in greater depth.From the four stress-strain curved concrete stress curves reviewed can be identified that the increasing branch curve in the geopolymer concrete confinement model by Ganesan is slightly lower than the same action in another confinement models.The design of a confined concrete top (f'cc) between the model by Mander and the Ganesan model can be declared to be relatively coincide.This is because the f'cc equation used is also the same (can be seen in Table 1).It appeared that the Mander model was only slightly corrected, namely factor r and modulus of elasticity (Ec) by the Ganesan model.It can be seen from the two models that the ductility from confined concrete according to the Ganesan model is lower.
Compared with the two confinement models discussed above, the stiffness from the confined geopolymer concrete based on the experiment looks more rigid, and the outcoming f'cc value is also higher (Figures 1 and 2).Except for the comparison of the G1 specimen in which the descending branch between the Ganesan model and the imitated experiment, the other specimens show that the descending branch in the Ganesan model is under estimation or ductility outcoming from experimental testing is more ductile.

Conclusion
The strength improvement from confined concrete in geopolymer concrete is more effective and optimal when hoop reinforcement with tighter spacing is employed.The yield stress of the hoop reinforcement plays a significant role in deciding the values from ε'cc and εc85.The confinement models discussed in this paper tend to under estimate the experimental outcomes from confined geopolymer concrete, chiefly the stresses and strains from the top from the confined concrete and the strain from the confined concrete after the top while the stress drops by 15 %.The geopolymer-confinement experiments has a slighly lower initial stiffness than the initial normal stiffness models (i.e.Mander and Ganesan models).Post top action or the ductility of confined-geopolymer concrete tends to be different than the confinement models.Some from the differences that occur are indications that material properties like elastic modulus and secant modulus developed in confinement models and with experimental outcomes are not the same, so they also have an influence on determining the form from the confined stress-strain curves.The outcomes of this discussion have recommended the need for further confined-geopolymer concrete research with a broader-parameter review to produce a more general confinement model.

Acknowledgement
The experimental program presented in this paper granted by Universitas Semarang, Indonesia.The support accepted to this research successfully is gratefully acknowledged.

7 Figure 4 .
Figure 4.The behaviour of confinement models and experimental outcomes of G3 and G4 specimens.

[ 1 ]
Antonius, Imran, I, P Setiyawan 2017 On the confined high-strength concrete and need of future research.Procedia Engineering 171 pp 121-130 [2] Antonius, Purwanto and P Harprastanti 2019 Experimental study of the flexural strength and ductility of post burned steel fiber rc beams International Journal of Technology 10(2) 428-437 [3] Purwanto, Antonius and P Setiyawan 2022 Stress-strain behavior of normal and high-strength steel fiber concrete post burning.International Journal of GEOMATE 21(85) 61-70 [4] E I Diaz-Loya, E N Allouche, S Vaidya 2011 Mechanical properties of fly-ash-based geopolymer concret.ACI Materials Journal 108(3) 300-306 [5] K T Nguyen, N Ahn, T A Le and K Lee 2016 Theoretical and experimental study on mechanical properties and flexural strength of fly ash-geopolymer concrete.Construction and Building Materials 106 65-77 hoops f'c = compressive strength of standard cylinder test at 28 days fcc = stress of confined concrete f'cc = top stress of confined concrete fyh = yield stress of hoop fyl = yield stress of longitudinal reinforcement ε'cc = top strain of confined concrete εc85 = strain corresponding to the 0.85 top stress of confined concrete at descending curve K = strength improvement of confined concrete = f'cc/0.85f'cs = spacing of hoop measured centre-to-centre of the steel

Table 3 .
Specimen details of confined geopolymer concrete cylinders.

Table 5 .
Comparison from analytical outcomes with experiments.