**2. Testing**

### *2.1. Materials Used for Tests*

In this article, the following specimens has been prepared for testing the mechanical properties of ESD pseudoplastic resins recommended as concrete structure joint fillers:


1A—resin based on thixotropic acrylic for injections,

2A—resin based on acrylic mixed with water,


6A—bitumen-based resin.


**Table 1.** Resin specimens—tensile strength *ft* elongation at break ε, density, viscosity and pH.

As shown in Table 1, six types of resins have been selected for tests. Those resins are universally used and available on the construction market and, more importantly, are used as expansion joint fillers. Each resin is characterised by different properties and technical parameters. The resins were fed into the gaps of the expansion joint model by means of the so-called gravity pouring. Subsequently, the expansion joint models were put aside until the resins hardened, i.e., for a period of 24 h. The entire testing process was divided into several series of tests. Each series of tests was conducted for a different resin. In total, tests for six different filler resins were carried out.

The axial tensile test for resins has been conducted in accordance with PN-EN ISO 37 standard [16] using the Instron 33R strength testing machine.

The testing procedure included the proper preparation of specimens. Resins in Table 2 were placed on a flat surface 400 × 300 mm, circa 5 mm thick. Afterwards, with the use of a template, shaped elements were cut out—the so-called "oars"—which were then placed, one by one, in the strength testing machine, as shown in Figure 2.

The testing was carried out until the moment of breaking of the resin specimens for which the results of maximum elongation at break and maximum breaking force were obtained.

**Figure 2.** Resin testing according to [16]—axial tensile test: (**a**) view of the strength testing machine—test stand, (**b**) view of the shaped resin elements, the so-called "oars", placed in the clamps and (**c**) view of the completed test at the moment of breaking of the resin specimen.

Examples of diagrams of stress as a function of deformation in axial tensile test—5 samples of resin 1 (Table 2) are presented in Figure 3.

**Figure 3.** The tensile stress-deformation curves as a function of deformation in axial tensile test of resin 1, Table 2 (5 samples - based on thixotropic acrylic) according to PN-EN ISO 37 [16].

The obtained results of the axial tensile test for the analysed resins according to PN-EN ISO 37 [16] are presented in Table 2.


**Table 2.** Results from the axial tensile test for the analysed resins according to PN-EN ISO 37 [16].

### *2.2. Preparation of Specimens for Tests*

The expansion joint model prepared for the tests was a model in which class C37 concrete specimens with 100 × 100 × 100 mm dimensions were used. The tests were carried out in laboratory conditions, with a temperature of 20 ◦C and stable air humidity. The preparation of the expansion joint model consisted in arranging two concrete specimens in parallel with each other, Figure 4.

For all expansion joint model specimens, a 10-mm-wide gap was prepared, and each of the resins and the reference cement mortar was then poured into that gap. In the case of all the fillers, there was the same method of preparation - mechanical cleaning of the concrete specimens surface (1A,2A,3A,4A,5A,6A–Table 1) and pouring the resin into the expansion joint.

**Figure 4.** Preparation of a specimen for tests: (**a**) cleaning the surface and (**b**) pouring the resin into the expansion joint.

### *2.3. Description of the Test Stand*

The expansion joint model specimens were tested using a Hounsfield 10K-S strength testing machine (Tinius Olsen TMC-United States) and the Horizon numerical processing software (ver.1). The adopted crosshead speed was 5 mm/min. A view of the test stand is presented in Figure 5.

Owing to the use of clamps on a harness (jointed system)—the expansion joint model displacement in one direction, perpendicular to the side surfaces of the gap, was simulated. The expansion joint model tension reflects the actual behaviour of filler materials in repaired reinforced concrete structures. The displacements in perpendicular direction are smaller and are not the main cause of destruction of pseudoplastic fillers in expansion joints.

The aim of the tests of expansion joint models filled with sealing resins was to determine the force–deformation correlation serving as the basis for the assessment of the usefulness of resins as expansion joint fillers. The test was based on the measurement of the displacement of crosshead. The stress–strain tests took into account the weight of the lower concrete cube. After the test, the resin condition after the break and the percentage degree of the resin adhesion to the concrete substrate were assessed visually.

Figure 6 shows an exemplary load–deformation curve for an expansion joint model specimen marked as 1AR5 (resin 1, humidity or contamination of the substrate A: dry, substrate cleaned manually R, specimen 5) with photographs depicting the testing process. Manual cleaning concrete surface results in less resin adhesion (than mechanical cleaning).

**Figure 5.** A view of the test stand: (**a**) testing machine, (**b**) close-up of a specimen during testing and (**c**) the adopted tension method.

**Figure 6.** An exemplary diagram of the load–deformation curve for an expansion joint model specimen AR5 with photographs depicting the testing process—manual cleaning surface.

### *2.4. Diagram of Properties Assessment for ESD Pseudoplastic Materials*

Figure 7 shows a diagram for the assessment of mechanical properties based on the load–deformation curve of ESD pseudoplastic materials (*E*-elastic deformation, *S*-strengthening control, *D*-deflection control) in an axial tensile test according to [31–33].

The drawing outlines the adhesion loss area *A0X* in which the resin has been pull off from the concrete surface.

Characteristic areas have been determined: *AE*-elastic deformation (*fcr*, the proportionality range-Hooke's law), *AS*-strengthening control (the area between *fcr*, and the occurrence of maximum stress *fmax*), *AD*-deflection control (the area between *fmax* and *fd)* and *AP*-propagation area.

Point *fd* corresponds to the ability to carry stress *fcr* and is a determinant of the optimum ESD pseudoplastic filler.

Any point *fX*(*FX,*<sup>ε</sup>*X,WX*) on the load–deformation curve has been defined with the use of a corresponding force *FX*, deformation ε*X* and absorbed energy *WX* (area under the curve).

The characteristic points *fX* ending each of the areas *AX* marked respectively: *fcr, fmax, fd.*

**Figure 7.** A diagram for the assessment of mechanical properties based on the load–deformation curve of ESD pseudoplastic materials in an axial tensile test.

What has been identified is the pulling off or breaking off of the pseudoplastic filler at any point *f0X*(*F0X,*<sup>ε</sup>*0X,W0X*) and the correlating area *A0X* (point *f0E* in the proportionality area *A0E*, point *f0S* in the strengthening area *A0S* and point *f0D* in the deflection control area *A0D*—additionally point *f0P* in the propagation area *A0P*).

Point *f0X*, where the pull off/destruction of the material was recorded, was considered to be the end point of the deformation capacity assessment. If there was no pulling off the resin, point *fd* ending the deflection control range served as the determinant of the end of the test.

The assessment of the deformation capacity of the materials used in the Hooke's law range has been determined as *dx* = *tg*<sup>α</sup>.

### **3. Test Results**

The presentation of the results of 0C1, 0C2, 0C3—a model with M4 cement mortar filler-and the averaged result serving as the reference 0C for the tested resins, is shown in Figure 8.

**Figure 8.** Presentation of load–deformation curves for expansion joint model specimens 0C1, 0C2, 0C3 and the reference specimen 0C in the axial tensile test.

For the averaged reference specimen *0C*, the following have been determined: *fcr*(force-*Fmax*; deformation-<sup>ε</sup>*max*; absorbed energy-*Wmax*), and deformation capacity *d0*. Also *fcr* = *fmax* (5820 N; 0.81 mm; 2348 J) and *d0* = 7160 (tgα force-deformation correlation from the Hooke's law range) have been determined. The results are presented in Table 3.

Figure 9 shows a diagram for the reference specimen and the "strongest" resin 5A (based on elastic epoxide). Specimen 5A shows slightly higher deformation capacity *d0* = 6509 (lower tgα than the reference specimen *d0* = 7160), a larger proportionality area *fcr* and, additionally, a significant strengthening control area to point *f0S*, at which there was a catastrophic, rapid break of the resin.

**Figure 9.** The load–deformation curve for specimen *5A* and specimen *0C*.

Figure 10 presents collective curves for resins 1A, 2A, 3A, 4A, 5A, 6A and the reference specimen *0C*. Each resin was tested on at least three specimens, out of which the most representative one was selected for comparison purposes. It was not necessary to average the diagrams due to the deformation of the moment of resin destruction.

**Figure 10.** The load–deformation curves for specimens 1A, 2A, 3A 4A 5A, 6A and the reference specimen 0C.

For each of the presented resins, characteristic points *fX* and *f0X* (moment of destruction/pulling off the resin) were determined, which is shown in Table 3.


**Table 3.** Presentation of the results (characteristic points on the force–deformation curve) for the reference specimen 0C and tested resins 1A, 2A, 3A, 4A, 5A, 6A.

The obtained data show more precisely the behaviour of expansion joint filler materials in a tensile test, enabling the assessment and comparison of various resins. The presented data enable the assessment of the behaviour of resins in each of the proportionality, strengthening control and deflection control areas. A tabular presentation of data enables the characteristic points *fx* from a number of tests to be averaged for the purpose of the assessment of ESD pseudoplastic materials.

Figure 11 is a presentation of load–deformation curves for the reference specimen 0C and resins 1A, 2A, comparing the obtained values of mechanical properties with the defined reference specimen.

**Figure 11.** The load–deformation curves for specimens 1A, 2A and specimen 0C.

Table 4 contains a comparison of the obtained values of the mechanical properties of the tested resins 1A, 2A, 3A, 4A, 5A and 6A in relation to the reference specimen 0C.

Table 5 presents the results of resins 1A, 2A, 3A, 4A, 5A, and 6A compared to the linear correlation of specimen *5A*, with the largest Hooke's law area.

The proposed method of describing points on the force–deformation curve enables the comparison of any selected points *fx1* (or areas *AX1*) and their comparison with any selected points *fx2* (areas *AX2*) chosen for the analysis of the obtained effects: *fx1*/*fx2(FX1*/*FX2,*<sup>ε</sup>*X 1*/<sup>ε</sup>*X2,WX1*/*WX2)*.

The results presented in Tables 4 and 5 indicate the possibility of juxtaposing freely the data selected for analysis, enabling the comparison of multiples of the achieved effects on various test stands. However, it is recommended to quickly introduce a standardised testing procedure that will enable an accurate comparison of the results obtained at various research centres.

Examples of deformations of resins 1A, 2A and 3A filling the expansion joints before and after the loss of adhesion to the substrate (point *f0X*) in an axial tensile test is presented in Figure 12.

A linear decrease in the load–deformation correlation in an axial tensile test of the presented model indicates the pulling off the fillers from the substrate. That process may be more or less dynamic. In the case of resin 5A, there was a catastrophic break.


**Table 4.** Resins 1A, 2A, 3A, 4A, 5A and 6A compared with the reference specimen *0C*.

**Table 5.** Resins 1A, 2A, 3A, 4A, 5A and 6A compared to the linear correlation of specimen 5A.


Resins 2A and 1A were characterised by a less rapid loosening process—as seen in the diagram. Resins 4A, 3A and 6A, after a partial loosening, were characterised by the greatest capacity for deformation and energy absorption in the destruction propagation area. In that area, the loss of tightness of the joints occurs, which is why that range is not taken into consideration when interpreting the obtained results.

 **Figure 12.** Examples of specimens of resins filling the expansion joints before and after the loss of adhesion to the surface (point *f* 0*X*) in an axial tensile test: (**a**) resin 1A, (**b**) resin 2A and (**c**) resin 3A.

### **4. Discussion of the Results**

The assessment of tensile tests of a model of an expansion joint filled with ESD resin mass is made possible by the recording of data: force/stress, corresponding deformation and work-as the quantity of absorbed energy (surface area under the stress–strain curve), Figure 7. The characteristics of each point *fx* on the load–deformation curve are presented in the form of *fx* (force/load; deformation; absorbed energy), Figures 8–11.

The proportionality area *AE* determines the quality of the filler, whereas the areas of strengthening control *AS* and deflection control *AD* (characterised by a greater ability to carry stress than the Hooke's law proportionality area) contribute to the improvement of durability of sealants in the expansion joints of concrete structures, and constitute an additional safety range.

The characteristics of ESD pseudoelastic resins serving as sealants in expansion joint gaps indicate an increased deformation capacity in areas *AS* and *AD*—resulting in a considerable quantity of absorbed energy compared to the proportionality area *AE*. The strengthening control area *AS* is more significant in that type of fillers (it does not generate a destruction of the filler structure).

The deflection control area *AD*, with loads greater than those occurring in the proportionality area and comparable to those in the strengthening control area, additionally enables an effective implementation of the applicability range of the ESD resin material in expansion joints. That effect is also controlled by the possible appearance of the first forms of destruction of the structure on the side surface of the pulled material (it usually cannot be seen with a naked eye).

The area of propagation, weakening *AP*, with the decreasing stress and increasing deformation, is characterised by the moment of the resin specimen being pulled off from the concrete surface together with the absorption of another portion of energy in the resin structure destruction process. The area of propagation (of carrying stress that is lower than the critical one) is not recommended for ESD pseudoplastic sealing resins (it may also be assessed).

The area of the loss of adhesion *A0* (Figure 7), with the decreasing stress and increasing deformation, is characterised by the moment of the resin specimen being pulled off from the concrete surface together with the destruction of the resin structure. The deformations breaking and destroying the structure of

the material result in a significant linear decrease in stress (linear decrease in the force–deformation correlation). What determines the maintenance of reliability of the whole system, i.e., the expansion joint gap filled with sealing resin, is the adhesion control area *A0*. The loosening of the resin from the concrete surface is considered an emergency situation. The loss of adhesion of the sealing resin in the expansion joint is, from the perspective of the maintenance of the sealant's reliability, the moment when water tightness is lost. At that moment, the water influencing the whole system is able to permeate through the filled expansion joint.

From the perspective of the sealant's reliability, it is assumed that the loss of adhesion should occur as late as possible. The breaking o ff, loosening of ESD pseudoplastic resins should take place in the propagation area *AP*, which is not significant as regards the values of carried stress.

What is an important component of the analysis of the obtained results is the correct determination of the Hooke's law range (*fcr*). The linear correlation should be determined on large-scale diagrams. It should be noted that the correlation should be determined from a section between 50% *fcr* and *fcr*. The force-deformation correlation in the initial range is not linear due to the specimen's arrangemen<sup>t</sup> on the slings and, therefore, should not be taken into account (when interpreting the test results).

A need to standardise the testing procedure has been emphasised. Because the tests are conducted on di ffering test stands (e.g., with di fferent crosshead speed, ambient temperature, etc.), the obtained data cannot be compared with one another. Until relevant standard provisions are adopted, it is proposed to compare the results as a multiple of the established reference item. It may be a specimen with defined parameters or e.g., the Hooke's law range of the best of the tested materials.

The obtained results of the tests of resins presented in Table 3 enable a preliminary assessment of the possibility of using them as expansion joint fillers. Depending on the obtained deformations, an appropriate filler can be selected for a specific case of an expansion joint on a structural element. Thus:


A comparison of the tested resins with the reference specimen 0C (Table 4) confirms the conclusions presented above. An analysis of the results in relation to a predefined reference item, based on a multiple of changes, enables a comparison of the results obtained at various research centres—until an applicable standard (norm) is established. In summary, it should be noted that only specimens 1A and 2A could be classified as ESD pseudoplastic resins. Specimen 1A shows lower deformation capacity (dx = 715) in comparison with specimen 2A (dx = 135), with a much greater ability to carry stress and higher quantity of absorbed energy in the areas: elasticity, strengthening control and deflection control.

What is worth noting is the comparison of the results of the proposed testing method with the traditional method of assessing the mechanical properties of resins presented in Figure 3 and Table 2. The presented data show that resin 1 (Figure 3) has di fferent ESD mechanical parameters compered to resin 1A (Figure 11), which leads to various conclusions regarding the assessment of the usefulness of pseudoplastic materials proposed as fillers in working elements of concrete structures—taking into account the assessment of their adhesion to the concrete surface.

The comparison of Figures 6 and 11 shows that mechanical cleaning of the concrete surface 1A results in better resin adhesion than manual preparation of the concrete surface 1AR5.

As mentioned above, the existing hyperelastic models used for a description of the correlation of non-elastic materials (including hyperelastic resins) do not take into account the impact of a large number of variables determining the usefulness of ESD materials as expansion joint fillers. In particular, they do not take into consideration the materials' adhesion to concrete surfaces and the strengthening control and deflection control areas.
