NMR INSPECTION ON CONCRETE-COATINGS

Transcription

1 NMR INSPECTION ON CONCRETE-COATINGS Jeanette Orlowsky IBAC, RWTH Aachen University, Germany Abstract The application of the NMR-MOUSE (Nuclear Magnetic Resonance Mobile Universal Surface Explorer, registered trademark of RWTH Aachen University) for the inspection of concrete-coatings is studied. This measurement method is non-destructive. In combination with a special lift, measurement depths from to 5 mm with a step size of better than 1 µm are possible. Geometrical properties such as the thickness of different coating layers can be determined as well as functional properties like the ingress of water into the coating and ageing processes of the coating material. The current results demonstrate that the NMR-MOUSE is a promising tool for non destructive testing of coatings on concrete. The thickness of the coating and its different materials layers can be measured. Furthermore, water ingress through the surface protection system inside the concrete can be detected as well as changes of the material due to weathering. Concerning the material changes a linear relationship between modulus of elasticity and relaxation time consist for the investigated acrylic dispersion. 1. INTRODUCTION Scores of infrastructural concrete buildings are protected with coating systems against aggressive exposure. The main aims which can be realized with these coating systems are: Protection of the concrete against the penetration of substances like water, chloride, carbon dioxide. Drying up of the concrete due to the coating barrier for water ingress and the moisture transport out of the concrete through the coating. Increase of the resistance against chemical and/or mechanical attack. These concrete-coatings usually consist of different layers of polymers or cement bonded materials. Concrete-coatings are also called surface protection systems in Germany. Up to now the use and application of surface protection systems in Germany was regulated in the guideline protection and repair of concrete structures published by the German Committee for Reinforced Concrete DAfStb [1]. As of recently, the European standard EN has been adopted in Germany by a DIN standard DIN V In both cases surface protection systems are divided into ten different classes depending on the application. 119

2 To ensure the target functions of a surface protection system, the following aspects are important: Thickness of the different coating layers, application method of the coating materials, consideration of environmental parameters like temperature, humidity and wind and preparation of the substrate. The last three points affect the reactive hardening of the coating materials, which is directly linked to the material properties, and the adhesion between concrete and coating as well as between the different layers. The thickness of the coating layers mainly affects the durability and properties like the crack bridging of the surface protection system. Up to now no non-destructive test-methods exist to determine: the whole thickness of the surface protection system as well as the thickness of the different layers, the water ingress through the surface protection system, the drying process of the concrete after application of a surface protection system, changes of the material due to weathering, quality and conformity of the applied material. It is shown below, that the NMR-MOUSE provides the potential to investigate these issues in a non-destructive way. 2. THE NMR-MOUSE WHAT S THAT? NMR is a method to interrogate molecular properties of matter by irradiating atomic nuclei in a magnetic field with radio-frequency waves. There is a big variety of the conventional NMR techniques that are used for characterization of polymers. This normally requires the sample to be positioned inside the NMR probe, which is placed inside a magnet with a strong homogeneous magnetic field. In unilateral NMR the magnetic field is applied to the sample from one side, so that the sample can be arbitrarily large. The use of this type of NMR is most profitable in quality control, but offers interesting perspectives also in other applications. A compact unilateral sensor developed for non-destructive tests of materials is the so called NMR-MOUSE (Nuclear Magnetic Resonance Mobile Universal Surface Explorer, registered trademark of RWTH Aachen University) [2], [3]. The main advantage of the NMR- MOUSE is its small size and weight. The NMR-MOUSE (Fig. 1, left) consists of four permanent magnet blocks positioned on an iron yoke. Two magnets are polarized along y, and two along y, producing a magnetic field B above the magnet parallel to the surface of the sensor. For analysis, a radio-frequency (rf) magnetic field B 1 is applied by means of a rectangular rf coil mounted in between the magnet blocks, which is perpendicular to B. Different types of the NMR-MOUSE are available. The types differ according to frequency, gradient, measuring depth and resolution [4]. For the investigations described in this paper the NMR-MOUSE PM 5 with a resonance frequency of 18 MHz was used. The PM 5 has a measuring field (sensitive volume) of 2 by 2 mm in the cross-section, while the thickness can be chosen between 1 and 1 µm. The NMR signal can be collected in the described sensitive volume which can be moved until a depth of 5 mm into the specimen by using a lift. At this lift the NMR sensor is mounted on a mobile plate while the specimen is placed on a fixed top plate which is parallel to the mobile plate (Fig. 2, left) [5]. The most important contrast parameters employed by the NMR-MOUSE are the signal amplitude corresponding to the proton density and the transverse relaxation time T 2 which relates to the molecular mobility and the effective self-diffusion coefficients of materials 12

3 containing hydrogen isotopes within the sensitive volume. This information is obtained by using the Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence (Fig. 2, left). To determine the transverse relaxation time, the decay curve measured with the CPMG sequence is fitted with an exponential function (Equation 1). The proton density is estimated from the integration of the signal corresponding to the first part of the CPMG decay (Fig 1, right). b 1 t s( t) = A exp b T2 with amplitude A [-], scaling factor b [-], time t [ms] and relaxation time T 2 [ms] normalized signal s(t) (1) proton density Figure 1: Left: Magnet geometry used to generate a highly flat sensitive volume [3]. Right: Decay curve fitted with equation 1 and proton density. time in ms Figure 2: Left: NMR-MOUSE with lift. Right: CPMG pulse sequence and the NMR response in inhomogeneous fields. 121

4 3. APPLICATION OF THE NMR-MOUSE 3.1 Thickness of coatings Measuring the proton density in steps of 3, 5 or 1 µm allows conclusions about the composition and layer thickness of coating systems on concrete. In a first step a reference specimen consisting out of polyurethane between two glass plates was investigated (Fig. 3, left). As shown in Fig. 3, right, the proton density inside the glass is nearly zero while the polyurethane has a significant higher proton density. Due to the chosen thickness of the sensitive volume of 1 µm the resolution is about 1 µm. To determine the polyurethane thickness the measuring depth at half of the averaged maximum proton density was calculated (Fig. 3, right) Proton density Step size in 1 µm Mean 1/2 Mean Thickness polyurethane Thickness glass plate Measuring depth in µm Figure 3: Left: Draft of the specimen. Right: Proton density as function of the specimen depth. Table 1 gives an overview about the mean glass plate and polyurethane thickness calculated from ten profiles in each case. Sensitive volumes (similar to step sizes) of 1, 5 and 3 µm were chosen for the measurements. The standard deviation shows an increasing scatter with decreasing step sizes. An accuracy of 5 µm can be reached with a step size of 5 µm for this example. Table 1: Thickness of the glass plate and polyurethane layer depending from the step size. (Mean and standard deviation calculated from 1 profiles in each case.) Step size Glass plate Polyurethane layer Mean Standard deviation Mean Standard deviation µm Figure 4, left, shows a microscopic image of a cross section of a coating system. This coating system has a high crack-bridging ability and robustness. Figure 4, right, shows the 122

5 proton density as function of the measuring depth inside the coating. Due to the arrangement of the specimen on the lift (Fig. 2, left) the measuring depth is top down. The top coating begins at a measuring depth of 5 µm. Up to 5 µm the proton density is low due to the roughness of the coating surface. The thickness of the sensitive volume was 1 µm. The increasing polymer content leads to an increase of the proton density. This enables us to differ between the top coating, the PUR-layer with sand and the elastic coating. Figure 4 clarifies the good correlation between the measured thicknesses of the coating layers via microscope and NMR-MOUSE. 25 Proton density 2 15 PUR Top-Coating PUR Elastic- Coating 1 5 PUR with Sand 3 µm 12 µm Measuring depth in µm Figure 4: Left: Cross-section of a coating system. Right: Proton density as function of the measuring depth in the coating system. 3.2 Water content An increasing content of water leads to a higher amount of hydrogen isotopes in the specimen. The outcome of this is a higher proton density and a slow down of the signal decay (increase of T 2 ). Figure 5 left shows the proton density inside a concrete specimen. The concrete core was water saturated, conditioned at a water content of 74 M.-%, 52 M.-%, 36 M.-% and dried at 5 C until constant mass before the NMR measurements started. As a result the correlation between measured proton density and water content is printed in figure 5, right. The calculation of the mean amplitude (proton density) was done with the mean value of three profiles on four different measuring points. 3.3 Material changes due to weathering For the investigation of material changes due to weathering using the NMR-MOUSE, coatings consisting out of acrylat-dispersions were used. Table 2 shows the composition of the investigated coatings. The coating DII-2 was produced as a film, thickness approx. 11 µm, and applied on concrete. Producing a free film the acrylat-dispersion was applied with a lambskin-roll on a coated glass plate. After 24 hours the film was removed from the glass plate. The specimens with coating DII-2 were weathered for 3 (D3), 5 (D5) and 12 (D12) years in the industrial city Duisburg in Germany [6]. 123

6 The coating OS5a-2 is an actual available product from the market. It was produced as a free film, as described above with an average thickness of 46 µm, and applied on concrete. The specimens were subjected to accelerated ageing in the laboratory (three different combinations of UV-light, temperature cycles, wetting and drying). Before and after the accelerated ageing NMR-measurements were done. On the coating film tensile tests were done to calculate the E-modulus Proton density water saturated 12 1 Proton density mean ampiltude including standard deviation (12 measued profiles) 8 water content 74 % dried at 5 C until constant mass Measuring depth in µm Water content related to constant mass at 5 C Figure 5: Left: Proton density in a concrete core as a function of measuring depth and water content. Right: Correlation between Proton density and water content in the concrete core. Table 2: Composition of the investigated coatings Code DII-2 OS5a-2 Base coat only for concrete specimens Acrylate-dispersion (solid content: ~ 18 Vol.-%) Copolymer-dispersion (solid content: ~ 23 Vol.-%) Inner layer Top layer Acrylate-dispersion Acrylate-dispersion (solid filled with sand content: ~ 53 Vol.-%) - Acrylate-dispersion (solid content: ~ 55 Vol.-%) Using the NMR-MOUSE the decay curve, measured with the CPMG sequence with an echo-time of.3 ms, was determined inside the top layer of the coatings after the different weathering conditions. From these curves the amplitude, b-factor and transverse relaxation time T 2 was calculated with equation 1. Figure 6 compares the results after different weathering times in Duisburg. The weathering time has an influence on the b-factor and on the T 2. During the first three years no significant changes appears but after five years a decrease of both factors was determined. The decrease of T 2 seems to go on slightly with increasing weathering time. The coating film shows the same tendency as the coating applied on concrete. Due to the higher flexibility of the film the T 2 is higher than the T 2 of the coating applied on concrete. The decline of T 2 indicates a decrease of the mobility of the polymer chain molecules. Therefore the weathering leads probably to an increase of the coating stiffness. Due to the inner layer of the coating film DII- 2 the tensile tests didn t allow a conclusion about the stiffness of the top layer after weathering. For this reason further investigations were done with the coating film OS5a

7 Amplitude [-] b-factor [-] T [ms] DII-2 - film DII-2 on concrete DII-2 - film DII-2 on concrete DII-2 - film DII-2 on concrete Figure 6: Amplitude, b-factor and T 2 of the coating-film and applied coating on concrete after 3, 5 and 12 years in Duisburg (: erence, D: Duisburg, mean value and standard deviation). In Figure 7, left, a tensile test on a coating film is shown exemplarily. Figure 7, right, shows the relaxation time T 2 calculated with equation 1 of the coating film OS5a-2 as function of the E-modulus. Each black symbol symbolizes one kind of accelerated weathering at one point of time and is calculated as mean value of three specimens (UV-chamber with 2, 125, 25 cycles; 2, 4, 8 freeze-thawing-cycles and complex weathering for 56, 9, 18 days as well as the reference). A nearly linear relationship between the E-modulus of the film and the relaxation time was determined. While the E-modulus increases due to the longer weathering time, the relaxation time decreases. This result underlines the decrease of T 2 due to an increase of the polymer stiffness. 2, 1,8 1,6 1,4 1,2 Relaxation time T 2 in ms Coating film OS5a-2 1, erence Freeze-thawing-cycles,8 UV-chamber Comblex weathering, E-modulus in N/mm² Figure 7: Left: Tensile test on a coating film. Right: Correlation between E-modulus and relaxation time T 2 of the coating film OS5a-2 125

8 4. CONCLUSION AND OUTLOOK The current results demonstrate that the NMR-MOUSE is a promising tool for non destructive testing of coatings on concrete. The thickness of the coating and its constituting layers can be measured. Furthermore, water ingress through the coating system can be detected as well as changes of the material due to weathering. The accelerated ageing leads to an increase of the E-modulus of the acrylat-dispersion OS5a-2. This material change was also detected by the NMR-MOUSE. A linear correlation between the E-modulus and the relaxation time T 2 was determined. While the E-modulus increases with ageing time the relaxation time decreases. This underlines the assumption, built up on the investigation of outdoor weathered specimens of DII-2, that the mobility of the polymer chain molecules decrease due to weathering. This leads to a decrease of the relaxation time. In the future the NMR-MOUSE shall be used as non-destructive method to investigate building surfaces. To realise the application on building surfaces further questions like the influence of steel reinforcement, environmental conditions, connection and movement of the NMR-MOUSE on the building have to be answered. ACKNOWLEDGEMENTS The investigations were done partly in the research project Determination of the ageing behavior of polymer coatings on concrete by NMR-techniques financed by the German Federation of Industrial Research Associations (AiF). The other part was done in a three-year research scholarship of RWTH Aachen University within the framework of the DFG (Deutsche Forschungsgemeinschaft) Excellence Initiative. The support is gratefully acknowledged. REFERENCES [1] Deutscher Ausschuss für Stahlbeton (DAfStb), 'DAfStb-Instandsetzungs-Richtlinie: Schutz und Instandsetzung von Betonbauteilen'. Berlin, 21 (Guideline 'protection and repair of concrete structures') [2] Blümich, B., 'Essential NMR' (Berlin, Springer, 24) [3] Perlo, J., Casanova, F., Blümich, B., 'Profiles with microscopic resolution by single-sided NMR', J. Magn. Reson. 176 (25) [4] [5] Orlowsky, J., Raupach, M., Baias, M., Blümich, B., 'Application of the NMR-Technique to Concrete Coatings', Proceedings of the 2nd International Conference on Concrete Repair, Rehabilitation and Retrofitting (ICCRRR), Cape Town, South Africa, , (28) [6] Antons, U., Orlowsky, J., Raupach, M., 'Results of a 12-year outdoor weathering of surface protection systems in different locations', Proceedings of the 13 th International Congress on Polymers in Concrete (ICPIC), Madeira, Portugal, , (21) 126