Influence of Polyurea Composite Coating on Selected Mechanical Properties of AISI 304 Steel

This paper contains experimental results of mechanical testing of the AISI 304 steel with composite coatings. The main goal was to investigate the impact of the applied polyurea composite coating on selected mechanical properties: Adhesion, impact resistance, static behavior, and, finally, fatigue lifetime of notched specimens. In the paper the following configurations of coatings were tested: EP (epoxy resin), EP_GF (epoxy resin + glass fabric), EP_GF_HF (epoxy resin + glass fabric hemp fiber), EP_PUA (epoxy resin + polyurea) resin, EP_GF_PUA (epoxy resin + glass fabric + polyurea) resin, and EP_GF_HF_PUA (epoxy resin + glass fabric + hemp fiber + polyurea) resin. The highest value of force required to break adhesive bonds was observed for the EP_PUA coating, the smallest for the single EP coating. A tendency of polyurea to increase the adhesion of the coating to the base was noticed. The largest area of delamination during the impact test was observed for the EP_GF_HF coating and the smallest for the EP-coated sample. In all tested samples, observed delamination damage during the pull-off test was located between the coating and the metallic base of the sample.

Polyurea coatings are increasingly popular in recent years [14][15][16]. They can be applied on metallic, wooden, and concrete surfaces, or even other plastics. These coatings allow for desired decorative properties to be obtained as well as specific mechanical properties. Polymers are used as cover layers for armed vehicles, ballistic shields [17], loading area of vehicles, and as a waterproof layer on concrete surfaces [18,19]. Additionally, they can absorb vibrations and sound waves. In order to strengthen the layers' properties, they were modified with glass fabric and hemp fiber.
The issue regarding the application of coatings is its poor adhesion to a metal surface. Due to this, in industrial conditions the surface is pre-processed by sandblasting and/or by application of an intermediate layer-primer-based on epoxy resin.
According to the practice of the polyurea coating application process, in the investigation presented in this paper the epoxy resin was used as an intermediate layer. The primer was modified in order to improve the impact resistance and vibration absorption by glass fabric and hemp fiber. In order to determine the influence of each constituent of the layer, there were also prepared samples with and without the polyuria layer. In order to eliminate the influence of mechanical treatment on the results of the experiment, the metallic base material was cleansed with acetone; the sand blasting process was not applied.
The coating was modified with natural fibers due to their low density and good ability to suppress acoustic waves. This application is often present in the automotive industry, where natural fibers are used as a filler in composite elements of vehicle interiors [20]. An important technological limitation in the natural fiber-reinforced composites industry is the temperature, which should not be bigger than 230 • C. Exceeding that temperature would cause the degradation of the fiber. Nevertheless, it is not an issue while using chemo-hardening resins.
The aim of the study was to determine the adhesion force of the polymeric coatings to the steel base, to compare the impact resistance of multilayer coatings based on the damage analysis caused by the impact of the energy of 17 J, to determine the coating resistance to cracking and peeling from the base as well as to investigate the influence of the coating on static mechanical properties and the fatigue lifetime of the sample, which is expected to improve. In general, the most widely used strategy of fatigue lifetime improvement is strengthening metallic structures using CFRP (carbon fiber reinforced polymer) patches [21][22][23]. The main reason for this is the redistribution of forces in metallic and composite structures. In this paper, the beneficial effect of polyurea composite on the fatigue performance of AISI 304 steel will be also demonstrated.

Materials and Methods
As a base material, austenitic steel AISI 304 in the form of 0.5 mm thick metal sheet was used. Chemical composition and static tensile results [24] of AISI 304 ((0.04%C, 1.1%Mn, 0.41%Si, 0.0437%P, 0.0044%S, 18.16%Cr, 8%Ni, 0.0335%Mo, 0.1%V, 0.32%Cu) steel are included in Table 1. The base was degreased with acetone. For the adhesion and impact tests 100 × 100 mm samples were prepared; for static tensile and fatigue tests oar-shaped samples were prepared ( Figure 1). with and without the polyuria layer. In order to eliminate the influence of mechanical treatment on the results of the experiment, the metallic base material was cleansed with acetone; the sand blasting process was not applied. The coating was modified with natural fibers due to their low density and good ability to suppress acoustic waves. This application is often present in the automotive industry, where natural fibers are used as a filler in composite elements of vehicle interiors [20]. An important technological limitation in the natural fiber-reinforced composites industry is the temperature, which should not be bigger than 230 °C. Exceeding that temperature would cause the degradation of the fiber. Nevertheless, it is not an issue while using chemo-hardening resins.
The aim of the study was to determine the adhesion force of the polymeric coatings to the steel base, to compare the impact resistance of multilayer coatings based on the damage analysis caused by the impact of the energy of 17 J, to determine the coating resistance to cracking and peeling from the base as well as to investigate the influence of the coating on static mechanical properties and the fatigue lifetime of the sample, which is expected to improve. In general, the most widely used strategy of fatigue lifetime improvement is strengthening metallic structures using CFRP (carbon fiber reinforced polymer) patches [21][22][23]. The main reason for this is the redistribution of forces in metallic and composite structures. In this paper, the beneficial effect of polyurea composite on the fatigue performance of AISI 304 steel will be also demonstrated.

Materials and Methods
As a base material, austenitic steel AISI 304 in the form of 0.5 mm thick metal sheet was used. Chemical composition and static tensile results [24] of AISI 304 ((0.04%C, 1.1%Mn, 0.41%Si, 0.0437%P, 0.0044%S, 18.16%Cr, 8%Ni, 0.0335%Mo, 0.1%V, 0.32%Cu) steel are included in Table 1. The base was degreased with acetone. For the adhesion and impact tests 100 × 100 mm samples were prepared; for static tensile and fatigue tests oar-shaped samples were prepared ( Figure 1). The type of applied coatings and layers configurations are presented in Table 2. The type of applied coatings and layers configurations are presented in Table 2. The coatings were applied manually. All samples were coated with epoxy resin LH 289 and characterized by low viscosity (Havel composites); more information about the resin are presented in Table 3. The first layer of the coating was modified by reinforcing it with glass fiber Areoglas 163 g/m 2 ( Table 4) and/or with cut hemp fibers of 40 mm length. Part of the samples was covered with two-component polyurea coating Almacoat Floor Sl (Table 5). Obtained surfaces of the samples are presented in Figure 2.

Adhesion of Coatings
Measurements of the adhesion of the coatings were carried out using a pull-off method, according to the standard PN-EN ISO 4624:2016-05 [25] using the PosiTest AT-A device (DeFelsko Corporation, Ogdensburg, NY, USA). During the test, the pull-off force of the stamp from the polymer coating based on the steel base was estimated. Prior to the test, measurement stamps were applied to the coating. After curing the glue, the circular notch around the stamp was cut and, subsequently, the pull-off test was carried out. There were five measurements done per each type of coating. In Figure 3 images of samples with glued measurement stamps are presented. The pull-off test classification according to the standard PN-EN ISO 4624:2016 [25] is presented in Table 6.

Adhesion of Coatings
Measurements of the adhesion of the coatings were carried out using a pull-off method, according to the standard PN-EN ISO 4624:2016-05 [25] using the PosiTest AT-A device (DeFelsko Corporation, Ogdensburg, NY, USA). During the test, the pull-off force of the stamp from the polymer coating based on the steel base was estimated. Prior to the test, measurement stamps were applied to the coating. After curing the glue, the circular notch around the stamp was cut and, subsequently, the pull-off test was carried out. There were five measurements done per each type of coating. In Figure 3 images of samples with glued measurement stamps are presented. The pull-off test classification according to the standard PN-EN ISO 4624:2016 [25] is presented in Table 6.

A Cohesive separation in the base A/B
Adhesive separation between the base and the first layer B Cohesive separation in the first layer B/C Adhesive separation between the first and the second layer N Cohesive separation in the n-th layer of the system n/m Adhesive separation between the n-th and the m-th layer of the system -/Y Adhesive separation between the last layer and the adhesive Y Cohesive separation in the adhesive Y/z Adhesive separation between the stamp and the adhesive The results of adhesion measurements are presented in Table 7. In all cases adhesive separation between the base and the first layer of the coating was obtained. In Figure 4, the results of the pulloff force measurement obtained during the PosiTest test are shown. Due to the different materials (including fibers) used for the coating, the total thicknesses of the layers were different. However, this did not change the reinforcement and redistribution of stresses, as shown by the results of the static tests. During the application of the layers, it was ensured that the thickness of the layers was equally distributed. Quality control with optical scanners revealed differences in thickness not greater than 8% of the applied layer.  The results of adhesion measurements are presented in Table 7. In all cases adhesive separation between the base and the first layer of the coating was obtained. In Figure 4, the results of the pull-off force measurement obtained during the PosiTest test are shown. Due to the different materials (including fibers) used for the coating, the total thicknesses of the layers were different. However, this did not change the reinforcement and redistribution of stresses, as shown by the results of the static tests. During the application of the layers, it was ensured that the thickness of the layers was equally distributed. Quality control with optical scanners revealed differences in thickness not greater than 8% of the applied layer.  [25].

A Cohesive separation in the base A/B
Adhesive separation between the base and the first layer B Cohesive separation in the first layer B/C Adhesive separation between the first and the second layer N Cohesive separation in the n-th layer of the system n/m Adhesive separation between the n-th and the m-th layer of the system -/Y Adhesive separation between the last layer and the adhesive Y Cohesive separation in the adhesive Y/z Adhesive separation between the stamp and the adhesive

Coatings Impact Resistance
Coatings resistance to cracking or peeling from the base was evaluated according to the standard PN-EN ISO 6272-1:2011 [26] using the impact resistance testing device-TQC (TQC Sheen B.V., Capelle aan den IJssel, Netherlands), presented in Figure 5. The test consists of determining the minimum height of fall for 20 mm diameter mass, under normalized conditions, in order to damage investigated coatings. The research according to this procedure was conducted also in papers [27,28]. Initial tests were conducted on the metallic base without any coating with a load of 1 kg. The metal sheet was hit from different heights and there was no observed rapture of the material ( Figure  6). Due to the lack of visible material damage after dropping the weight of 1 kg from maximal height of 1 m, the weight was changed to 2 kg. Material damage was observed for the drop of 2 kg weight

Coatings Impact Resistance
Coatings resistance to cracking or peeling from the base was evaluated according to the standard PN-EN ISO 6272-1:2011 [26] using the impact resistance testing device-TQC (TQC Sheen B.V., Capelle aan den IJssel, Netherlands), presented in Figure 5. The test consists of determining the minimum height of fall for 20 mm diameter mass, under normalized conditions, in order to damage investigated coatings. The research according to this procedure was conducted also in papers [27,28].

Coatings Impact Resistance
Coatings resistance to cracking or peeling from the base was evaluated according to the standard PN-EN ISO 6272-1:2011 [26] using the impact resistance testing device-TQC (TQC Sheen B.V., Capelle aan den IJssel, Netherlands), presented in Figure 5. The test consists of determining the minimum height of fall for 20 mm diameter mass, under normalized conditions, in order to damage investigated coatings. The research according to this procedure was conducted also in papers [27,28]. Initial tests were conducted on the metallic base without any coating with a load of 1 kg. The metal sheet was hit from different heights and there was no observed rapture of the material ( Figure  6). Due to the lack of visible material damage after dropping the weight of 1 kg from maximal height of 1 m, the weight was changed to 2 kg. Material damage was observed for the drop of 2 kg weight Initial tests were conducted on the metallic base without any coating with a load of 1 kg. The metal sheet was hit from different heights and there was no observed rapture of the material ( Figure 6). Due to the lack of visible material damage after dropping the weight of 1 kg from maximal height of 1 m, the weight was changed to 2 kg. Material damage was observed for the drop of 2 kg weight from 0.9 m height. The obtained sample was used as a reference sample for further tests on coated samples (Figure 7).  The impact resistance test was conducted on previously-prepared coated samples, using the weight of 2 kg gravity dropped from 0.9 m height. Observed results of the test are presented in Tables  8 and 9.   The impact resistance test was conducted on previously-prepared coated samples, using the weight of 2 kg gravity dropped from 0.9 m height. Observed results of the test are presented in Tables  8 and 9.  The impact resistance test was conducted on previously-prepared coated samples, using the weight of 2 kg gravity dropped from 0.9 m height. Observed results of the test are presented in Tables 8 and 9.   The impact resistance test was conducted on previously-prepared coated samples, using the weight of 2 kg gravity dropped from 0.9 m height. Observed results of the test are presented in Tables  8 and 9.   The impact resistance test was conducted on previously-prepared coated samples, using the weight of 2 kg gravity dropped from 0.9 m height. Observed results of the test are presented in Tables  8 and 9.

Static Tensile Test of Notched Specimens
The static tensile test was conducted on MTS 810 Material Testing Machine (MTS Systems Corporation, Eden Prairie, MN, USA). The test was conducted for samples with EP_GF, EP_PUA, EP_GF_PUA, and EP_GF_HF coatings. The results are the mean of tests on five samples per coating and presented in Figure 8. All results correspond well with the previously obtained [29] experimental data for the same specimen configuration (without coating) made from AISI 304 steel. The critical gross-section tensile stress for the notched (k t = 5.88) AISI 304 steel specimen was estimated at the level 505 MPa. The results for the sample with EP coating were no different from the value for non-coated steel.

Fatigue Testing
The fatigue test was conducted on the uniaxial MTS 810 Material Testing Machine equipped with a 5 kN load cell under a stress-controlled mode for one selected load level. During the test stress, ratio (R = 0.05) and frequency (f = 20Hz) were kept constant. All specimens were loaded using sinusoidal waveform with the maximum load level FMAX = 1400 N and minimum load level FMIN = 70 N. In order to achieve proper surface and geometry of the notch, as well as to avoid delamination of the coating, the notch was cut out using the diamond string method. Obtained results are the mean of five samples and presented in Figure 9.

Fatigue Testing
The fatigue test was conducted on the uniaxial MTS 810 Material Testing Machine equipped with a 5 kN load cell under a stress-controlled mode for one selected load level. During the test stress, ratio (R = 0.05) and frequency (f = 20 Hz) were kept constant. All specimens were loaded using sinusoidal waveform with the maximum load level F MAX = 1400 N and minimum load level F MIN = 70 N. In order to achieve proper surface and geometry of the notch, as well as to avoid delamination of the coating, the notch was cut out using the diamond string method. Obtained results are the mean of five samples and presented in Figure 9.
The fatigue test was conducted on the uniaxial MTS 810 Material Testing Machine equipped with a 5 kN load cell under a stress-controlled mode for one selected load level. During the test stress, ratio (R = 0.05) and frequency (f = 20Hz) were kept constant. All specimens were loaded using sinusoidal waveform with the maximum load level FMAX = 1400 N and minimum load level FMIN = 70 N. In order to achieve proper surface and geometry of the notch, as well as to avoid delamination of the coating, the notch was cut out using the diamond string method. Obtained results are the mean of five samples and presented in Figure 9. The macroscopic images of broken specimens are presented in Figure 10. The macroscopic images of broken specimens are presented in Figure 10. Coatings with glass fiber as a constituent had a visible asymmetric fracture surface. In case of all samples, there was visible delamination of the coating from the metallic base. The area of delamination was different for each sample, including samples with the same coating. Nevertheless, the delamination process was observed to start behind the notch area, or was even not observed in the notch area at all. This indicates that the notching method was selected and conducted properly. There was no visible delamination between the layers of the coating. The observed break of adhesion forces was, in the case of all samples, between the metallic base and the coating.

Conclusions
Based on the performed experimental campaign, the following conclusions can be drawn: (1) It was observed that the highest value of force required to break adhesive bonds was achieved for the EP_PUA coating, the smallest for the single EP coating.
(2) The largest area of the delamination during the impact test was observed for the EP_GF_HFcoated sample and the smallest for the EP-coated sample.
(3) The static tensile test did not show a significant difference in the influence of the coating on the tensile strength of the material.
(4) Fatigue tests results showed that the difference in the number of cycles to failure depends on the type of coating used. For coatings with polyurea and glass fiber as constituents, the increase of fatigue lifetime was significant.
(5) The macroscopic analysis of the fracture area of damaged samples confirms that the method of notch preparation was correct and had no influence on the behavior of individual samples during the fatigue test.
Due to the possibility of manual application of the coating, if further research on the fatigue lifetime and fatigue crack growth confirm the preliminary results presented in this paper, the coating might be used as an "on-site" fatigue lifetime enhancer and fatigue crack growth retardation tool on the existing structures.

Funding:
The publication has been prepared as a part of the Support Programme of the Partnership between Higher Education and Science and Business Activity Sector financed by City of Wroclaw.

Conflicts of Interest:
The authors declare no conflicts of interest. Coatings with glass fiber as a constituent had a visible asymmetric fracture surface. In case of all samples, there was visible delamination of the coating from the metallic base. The area of delamination was different for each sample, including samples with the same coating. Nevertheless, the delamination process was observed to start behind the notch area, or was even not observed in the notch area at all. This indicates that the notching method was selected and conducted properly. There was no visible delamination between the layers of the coating. The observed break of adhesion forces was, in the case of all samples, between the metallic base and the coating.

Conclusions
Based on the performed experimental campaign, the following conclusions can be drawn: (1) It was observed that the highest value of force required to break adhesive bonds was achieved for the EP_PUA coating, the smallest for the single EP coating.
(2) The largest area of the delamination during the impact test was observed for the EP_GF_HF-coated sample and the smallest for the EP-coated sample.
(3) The static tensile test did not show a significant difference in the influence of the coating on the tensile strength of the material.
(4) Fatigue tests results showed that the difference in the number of cycles to failure depends on the type of coating used. For coatings with polyurea and glass fiber as constituents, the increase of fatigue lifetime was significant.
(5) The macroscopic analysis of the fracture area of damaged samples confirms that the method of notch preparation was correct and had no influence on the behavior of individual samples during the fatigue test.
Due to the possibility of manual application of the coating, if further research on the fatigue lifetime and fatigue crack growth confirm the preliminary results presented in this paper, the coating might be used as an "on-site" fatigue lifetime enhancer and fatigue crack growth retardation tool on the existing structures.

Funding:
The publication has been prepared as a part of the Support Programme of the Partnership between Higher Education and Science and Business Activity Sector financed by City of Wroclaw.

Conflicts of Interest:
The authors declare no conflicts of interest.