*3.2. Apparent Density*

As it has been already determined in [40], the apparent density of polyurethane foam is among the key parameters, significantly affecting the product's physical and mechanical properties. The test results summarised in Figure 6 reveal that cured polyurethane adhesive in 8 mm and 15 mm thick bonds were characterised by apparent density from 19.3 kg/m<sup>3</sup> to 25.3 kg/m3. In general, the polyurethane foam's apparent density depends on the cellular structure [35,37]. Structures with larger cells are characterised by lower apparent density [37,40], which is confirmed by the study results. The highest densities were obtained for the samples taken from 8 mm thick bonds developed in laboratory conditions, at high temperature and low relative humidity, as well at low temperature, amounting to 24.8 kg/m3, 24.6 kg/m3 and 25.3 kg/m3, respectively. According to the description, the adhesive structure in the bonds developed in the abovementioned conditions was homogenous, the cells were uniform and well-defined, and their diameter was up to 300 μm (Figure 5a,c,d). The adhesive density in 15 mm thick bonds and in 8 mm thick bonds formed at high temperature and high relative humidity, for which a non-uniform structure and the presence of cells with ca. 350 μm diameter was observed (Figure 5b,e,f), was lower and amounted to 19.3 kg/m<sup>3</sup> and 21.2 kg/m3, respectively. As expected, the adhesive density in the bonds was higher than the density determined for a free-foamed product and amounted to 18 ± 2 kg/m<sup>3</sup> (Table 2). The cells in the freely applied products reach higher diameters than under limited product expansion conditions [13].

**Figure 5.** Microstructure of the cross-sectional surface of polyurethane adhesives in the bonds (magnification 100×) (**a**) OSB/23/50/8, (**b**) OSB/23/50/15, (**c**) OSB/25/30/8, (**d**) OSB/5/-/8, (**e**) OSB/25/90/8 and (**f**) OSB/25/90/8.

**Figure 6.** Apparent density of PU adhesive in bonds made under different thermal and moisture conditions. Error bars show standard deviation values.

A literature review revealed that the apparent density of polyurethane foam varies depending on the concentration of water as a blowing agent. The apparent density decreases with the increasing blowing agent content [20,39,40]. It was determined [40] that the polyurethane foam density decreased from 116 kg/m<sup>3</sup> to 42 kg/m<sup>3</sup> as the water content increased from 0.1 to 3.0 phr. The same observation was made during studies of closed-cell rigid polyurethane foams based on low functionality polyols [39]. A similar trend was observed in this study. The apparent density of the adhesive in the bonds made at low humidity (25 ± 2 ◦C, 30 ± 5%) was 15% higher than that for the adhesive in the bonds made at the same temperature but high humidity (25 ± 2 ◦C, 90 ± 5%).

#### *3.3. Mechanical Properties*

As mentioned, before their launch, construction products are verified for the building structure's meeting seven basic requirements, according to CPR [21]. In reference to ETICS, the bond strength, shear strength and shear modulus of the bond adhesive are among the essential requirements that determine the fulfilment of the fourth basic requirement, 'safety in use' [5,6,25].

Analysing the bond strength test results presented in Figure 7, it can be concluded that the bonds of the 8-mm-thick polyurethane adhesive for a system with MW and OSB, and FGB and CPB, had a bond strength similar to that of the reference concrete substrate used as a standard in tests of polyurethane adhesives for EPS-based ETICS. For bonds made under laboratory conditions, bond strength was from 85 to 100 kPa, at high temperature and low relative humidity from 83 to 93 kPa, at high temperature and high relative humidity from 85 to 93 kPa, and at low temperature from 81 to 89 kPa, while for systems with concrete substrate it was 89 kPa, 100 kPa, 87 kPa and 84 kPa, respectively. By analysing the minimum values of the bond strength (Figure 7, values in brackets), one might conclude that for 8 mm thick bonds made in laboratory conditions it ranges from 64 to 81 kPa, at high temperature and low relative humidity from 60 to 76 kPa, at high temperature and high relative humidity from 69 to 77 kPa, and at low temperature from 62 to 78 kPa, while for systems with a reference concrete substrate it is 72 kPa, 89 kPa, 77 kPa and 61 kPa, respectively.

As mentioned, the assessment of the suitability of use of ETICS is carried according to EAD 040083-00-0404 [5] and EAD 040089-00-0404 [6]. Comparison of bond strength values, obtained in our experiment, with the criterion specified at [5,6] for polyurethane adhesives in EPS-based ETICS, which is at least 80 kPa for the average value and at least 60 kPa for the minimum value, allows for a conclusion that the analysed solution is characterised by adhesion at the level higher than the mentioned threshold values. The above could be considered as an important indication for a more favourable assessment of the applicability of polyurethane adhesive as a component of a mineral wool-based ETICS. The obtained results are also in line with essential characteristic of polyurethane adhesives for EPS-based ETICS existing on the market [9,10]. To date, no more information in the

literature on the performance of polyurethane adhesives in ETICS has been presented. The researchers' attention has been directed towards cement-based adhesive systems. The results obtained show that polyurethane adhesives bond strength is significantly lower than bond strength between cement-based adhesive and the concrete [5,31,32]. As it has been already determined in [31], bond strength between cement-based adhesive and the concrete, after 28 days under laboratory conditions, can achieve values above 250 kPa. In other works, bond strength at the level to 1000 kPa was noted [33]. The difference may be explained by differences in structure and material nature of the polymer foams and cement-based products [13]. However, as regards bond strength between cement-based adhesive and the concrete after 28 days under laboratory conditions and 2 days in water, bond strength similar to bond strength of polyurethane adhesives [9,10] can be noted. The test of bond strength between cement-based adhesive and the thermal insulation material is performed separately [5,6]. As it has been already determined in [31,32], it depends strongly on the type of insulation material and the model of damage. For EPS systems, values in the range from 120 kPa to 270 kPa and cohesive rupture in the insulation material were achieved [31,33,34]. However, for MW system values in the range from 30 kPa to 80 kPa, cohesive damage in the insulation material was noted [5,9,10].

**Figure 7.** Bond strength results for PU adhesive bonds made under different thermal and moisture conditions. Error bars show standard deviation values. The minimum value for the series is given in brackets.

The effect of bond thickness was prominent in the tests conducted. For the 15 mm thick bonds, noticeably lower bond strength values were obtained than for the 8 mm thick bonds, as expected. The results were 71 kPa for OSB/23/50/15, 73 kPa for FBG/23/50/15, 76 kPa for CPB/23/50/15 and 76 kPa for the reference substrate C/23/50/15 (Figure 7). Therefore, when compared to the bond strength of bonds made under the same conditions but with a thickness of 8 mm, they were lower by 16%, 19%, 24% and 15%, respectively. These differences are due to differences in the adhesive cellular structure [35]. According to the experience of other researchers, in wider bonds carbon dioxide has the ability to form larger bubbles, resulting in a more porous structure [40]. The performed SEM analysis indicates cells less than 300 μm in diameter predominated in the 8 mm bond (Figure 5a). Cells of the adhesive in the 15 mm bond were noticeably larger. The predominant cells were about 450 μm and larger in diameter (Figure 5b) as a previous study showed more porous polyurethane foam may have a lower tensile strength [37]. By comparing the test results for 15 mm thick bonds with the criterion specified for PU adhesives in EPS-based ETICS of at least 80 kPa [5,6], it can be seen that significantly lower values were obtained. In this case, consideration should be given to limiting the use of the adhesive on substrates where no irregularities are necessitating the use of 15 mm thick bonds. Taking into account

that the deviation from the flatness of OSB, FBG and CPR is usually less than 5 mm [41,42], this condition does not pose a serious problem.

A correlation between the bond strength and the apparent adhesive density was noted. As has been already determined [40], higher apparent density of polyurethane foam resulted in higher mechanical properties. A similar effect was observed in this study. The highest bond strength was obtained for bonds developed in laboratory conditions at high temperature and low relative humidity, as well as at low temperature, whose densities were 24.8 kg/m3, 24.6 kg/m3 and 25.3 kg/m3, respectively. No such regularity was observed for bonds developed at high temperature and high relative humidity.

Analysis of the cross-sections of the samples after testing clearly indicates the cohesive model of the damage. For the 8 mm thick adhesive bonds made under laboratory conditions, high temperature and low relative humidity, as well as low temperature, damage within the MW was predominant. In these series, the average proportion of damage within the MW was up to 80 to 95% (Figures 8a and 9a–c), 50 to 95% (Figure 8b) and 70 to 90% (Figure 8d), respectively. The above indicates that the bond strength exceeded the perpendicular tensile strength of the thermal insulation material itself. A similar effect was observed for MW-based ETICS with cement-based adhesive [32]. Cohesive damage was observed also for bonds made at high temperature and high relative humidity but with predominant damage within the polyurethane adhesive. The proportion of damage in MW ranged from 35 to 48% (Figure 8c). Cohesive damage within the polyurethane adhesive was also recorded for 15 mm thick adhesive bonds (Figure 10a). The proportion of MW damage ranged from 22% to 28%, which is noticeably lower than for the 8 mm thick bonds where it ranged from 80 to 95% (Figure 10b). Again, these differences can be explained by the differences in the cellular structure of adhesive. More porous polyurethane adhesive may obtain lower tensile strength [35,37,39].

**Figure 8.** Model of damage—average values for the series: (**a**) 23/50/8, (**b**) 25/30/8, (**c**) 25/90/8 and (**d**) 5/-/8 (C/MW—cohesive damage within the MW, C/PU—cohesive damage within the PU adhesive).

**Figure 9.** Illustration of the model of damage of 8 mm thick bonds (**a**) CPB/23/50/8 series sample— C/MW damage, (**b**) OSB/23/50/8 series sample—C/MW damage combined with C/PU damage and (**c**) CPB/23/50/8 series sample—C/MW damage (C/MW—cohesive within the MW, C/PU cohesive within the PU adhesive).

**Figure 10.** Illustration of the damage of 15 mm thick bonds (**a**) OSB/23/50/15 series sample, (**b**) average values for individual series (C/MW—cohesive damage of the MW, C/PU—cohesive damage of the PU adhesive).

There was no significant effect of sheathing type (OSB, GFB and CPB) on bond strength. The same observation was made during studies of cement-based adhesive [34]. In the series prepared under laboratory conditions, the highest value was for CPB/23/50/8— 100 kPa and the lowest was for OSB/23/50/8—85 kPa; in the series prepared at high temperature and low relative humidity, the highest value was for CPB/25/30/8—93 kPa and the lowest for FGB/25/30/8—83 kPa; for series prepared at high temperature and high relative humidity, the highest value was for CPB/25/90/8—93 kPa and the lowest for FGB/25/90/8—85 kPa; and for series prepared at low temperature, the highest value was for CPB/5/-/8—89 kPa and the lowest for OSB/5/-/8—81 kPa. The above indicates that the performance evaluation process may consider limiting the number of test runs to one type of sheathing.

No effect of substrate type on the model of the damage was observed. The GFB/25/30 /8 series highlighted samples slightly in this respect, for which, as for the OSB/23/50/8 series, the proportion of damage within the polyurethane adhesive was recorded at 50%, while for samples on other substrates, it ranged from 5 to 25%. No such regularity was observed in the other test series.

Summarising the experimental data on bond strength obtained in this study, it can be stated that the tested polyurethane adhesive showed satisfactory adhesion to both mineral wool (MW) and boards typical for sheathing of walls of frame structure—oriented strand boards (OSB), fibre-reinforced gypsum boards (FGB) or cement-bonded particleboards (CPB). The cohesive property of the damage, predominantly within the thermal insulation material, indicates that the polyurethane adhesive bonds' bond strength may exceed the perpendicular tensile strength of the thermal insulation material itself. It should also be noted that mineral wool lamella, without coatings or facing, with a perpendicular tensile strength ≥80 kPa (TR80) was used in the tests. The factor determining the bond strength was, as expected, the thickness of the adhesive bond. Increasing the thickness from 8 mm to 15 mm resulted in a decrease of approximately 20% in bond strength. The effect of the thermal and moisture conditions under which the bonds were made and cured was also outlined. The lowest values of bond strength were recorded for the series prepared at low temperature, next at high temperature and high relative humidity, high temperature and low relative humidity, and the highest at laboratory conditions. In contrast, it should be noted that only in the series prepared at high temperature and high relative humidity the damage of the polyurethane adhesive predominated. In contrast, the damage of MW predominated in the other cases, so the decisive influence on the values obtained was the properties of the thermal insulation material. No effect of substrate type (OSB, FGB, CPB or concrete) on bond strength was observed.

Shear strength and shear modulus were analysed in terms of the influence of the type of substrate, taking into account the boards standard for the sheathing of timber frame walls and the adhesive thermal and humidity conditions bonds. The shear strength values are shown in Figure 11, and the shear modulus values are shown in Figure 12. The highest values of the considered properties were recorded for the samples prepared at high temperature and low relative humidity, obtaining shear strength of 55 kPa for OSB/25/30 series, 75 kPa for FGB/25/30 and 69 kPa for CPB/25/30 and shear modulus of 605 kPa, 920 kPa and 940 kPa, respectively. The bonds made at high temperature with high relative humidity showed significantly lower values concerning their properties, which may be dictated by the difference in the adhesive cell structure (Figure 5). Shear strength of 56 kPa for OSB/25/90 series, 52 kPa for FGB/25/90 series and 52 kPa for CPB/25/90 series was obtained, while for shear modulus, it was 455 kPa, 510 kPa and 590 kPa, respectively. The properties of bonds made at low temperature were of average values, except for the shear strength of the OSB/5/- bonds where a value of 71 kPa was recorded, while it was 57 kPa for FGB/5/and 47 kPa for CPB/5/-. Shear modulus was 610 kPa, 720 kPa and 660 kPa, respectively. All tested samples proved to be vulnerable to cohesive damage, in 100% within the adhesive bond, which confirms the high adhesion of polyurethane adhesive to all considered substrates—OSB, FGB and CPB—recorded bond strength tests. No significant effect of substrate type on the properties considered was observed.

**Figure 11.** Shear strength tests results of bonds made under different thermal and moisture conditions. Error bars show standard deviation values. Data supplemented with a description of the damage: C/PU—damage of cohesion in PU adhesive.

**Figure 12.** Shear modulus tests results of bonds made under different thermal and moisture conditions (error bars show standard deviation values).

The shear strength values obtained in this study were slightly lower than those approved for typical adhesives intended for use in EPS-based ETICS [9,10]. The shear modulus values were close to those indicated in [9] and significantly higher than those specified in [10]. It should also be noted that within the framework of the above-mentioned ETA procedures, the tests of bonds made under laboratory conditions on standard particleboards were carried out. Shear strength and shear modulus, according to both EAD 040083-00-0404 [5] and EAD 040089-00-0404 [6] guidelines, should be considered as a property of adhesive bond that can be used in the insulation design process.

#### **4. Conclusions**

The analysis of the experimental data obtained in this study proves that there are indications for an upbeat assessment of the applicability of polyurethane adhesive as a component of a mineral wool-based ETICS, intended for fixing thermal insulation material to the sheathing of walls with the timber frame structure.

It has been shown that polyurethane adhesive can achieve satisfactory adhesion to mineral wool lamella (TR80) without coatings and facing in the form of fabric, veil, film, etc. Bond strength of bonds made in thermal and moisture conditions limited for the considered application, with a bond thickness of 8 mm, achieved a satisfactory for ETICS value above 80 kPa.

It was also found that polyurethane adhesive has good adhesion to boards typical for timber frame wall sheathing—oriented strand boards (OSB), fibre-reinforced gypsum boards (FGB) or cement-bonded particleboards (CPB). No significant effect of board type on bond strength, shear strength and shear modulus was determined.

The methodology of testing the performance of polyurethane adhesives intended for fixing mineral wool boards requires the analysis of the specifics of the polyurethane applied on site and the thermal insulation material and the sheathing boards. The test shows that the introduction of appropriate modifications to standard procedures established for EPSbased ETICS makes it possible to obtain data indispensable for assessing the performance of adhesives intended for MW-based ETICS.

Taking into account the diversity of both polyurethane adhesives and mineral wool face finishes, the authors intend to continue work focused on the aspect of adhesion. Furthermore, verification of the performance of MW-based ETICS made with the use of polyurethane adhesive is planned on facade models, including all components of the system.

**Author Contributions:** Conceptualization, E.S.; Methodology, E.S.; Formal analysis, E.S.; Investigation, E.S. and E.K.; Writing—original draft preparation, E.S. and E.K.; Writing—review and editing, E.S.; Visualisation, E.S. and E.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Ministerstwo Nauki i Szkolnictwa Wyzszego (NZM-48/2020). ˙

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** Special thanks for Iwona Gał ˛aska, Jarosław Sówka and Anna Baranowska for technical support.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the study's design, in the collection, analyses, or interpretation of data, in the writing of the manuscript; or in the decision to publish the results.

#### **References**

