1 Introduction

A concrete facing is a framed surface, which is generally visible when the work is finished. It constitutes the skin of the work. The appearance of these facings is completely due to the conditions of the development of the skin area and thus to the concrete/formwork interface [1]. The final quality of these facings also depends on the release agents used for lubrication of the formworks. Release agents must be chosen according to the type of formwork and their compatibility with the formwork skin. Moreover, they must be homogeneously applied to the whole formwork on a clean surface in an uniform layer before reinforcements are placed [2]. They are applied like a painting, in a uniform layer, without any running, with the help of a sprayer or a nozzle depending on the situation. The excess must be removed with a rag or a scrapper, if necessary. Release agents are either fluid or pasty. There are several types of them: synthetic oils, vegetable oils, emulsions, wax and resins. Our study mainly deals with oils.

Literature gives a different thickness according to the film. Indeed a 2 μm film can gives a good quality facing, but 10 μm can also gives good aesthetic results [3]. It is obvious that the nature of the release agent and the way it is applied will play an important role. Moreover, synthetic or mineral-based oils do not have the same physico-chemical characteristics as vegetable-based oils. At the moment, mineral oils are the most popular. However, these products are not environmentally friendly and cause inconvenience to the users (skin and respiratory tract irritations due to the inhalation of volatile organic components) [4]. Prolonged exposures are recognized to generate irreversible damages which are considered as occupational diseases. To tackle these inconveniences, vegetable-based formulations were created.

The thickness of the oil film is an essential parameter for the protection of metallic surfaces against corrosion and when fine facings are required. A method to determine the thickness of the oils consists in using a weighing method, nevertheless this method cannot be applied on life-size elements (formworks, surfaces of moulds at their prefabrication stage). Thus, a measuring method was devised by the Centre d’Energie Atomique (CEA) of Saclay.

In civil engineering, two cases can be encountered for the method of fabrication of the concrete works: prefabrication and setting of the concrete on a building site.

In prefabrication, moulds with weak pouring high are often employed which entails a quasi-static action of the concrete. The pouring of the concrete in vertical formworks entails a dynamic action of the concrete on the formwork.

The thickness of the oil film will be measured before and after the demoulding according to the setting of the concrete on building sites and on the moulds so as to determine if enough oil remains after the concrete pouring to ensure an effective protection against corrosion. For this part of this experimental study a plane/plane tribometer will be used. Results will be correlated to a study dealing with the aesthetic aspect of the facings so as to determine an interval of thickness of the oil film in order to obtain facings without any defect.

2 Properties of materials and oils

2.1 Concrete

To avoid physico-chemical reactions which can be difficult to interpret between components of the release agents and additive in the concrete, a traditional concrete (B25) with no additive was formulated for this study. A slump value of 10 cm is obtained at the Abrams cone. The composition of the concrete is given in Table 1.

Table 1 Concrete composition
Full size table

The concrete mixtures were prepared in batches of 30 l in a mixer equipped with a helicoïdal mixing paddle. The mixing is realized according to the French standard NFP 18-404 called ‘Concretes Tests for studies—Conveniences and checks—Making and keeping of the test tubes’. The mixing protocol is:

All tests were made in a laboratory at an ambient temperature (about 20°C).

2.2 Characteristics of oils

For this study, two oils were chosen: a vegetable-based formulation (V) and a mineral-based formulation (M). The properties of the oils used in this study are given in Table 2.

Table 2 Physico-chemical characteristics of the studied oils
Full size table

Oil’s capacity to spread over a surface can be estimated by the contact angle [5]. The wetting power of the oils affects the evenness of the deposited film. The contact angle is measured with a goniometer. A small drop angle defines a high wettability power. The drop was deposited on a sample (5 × 5 cm2) cut from a new formwork wall. The value of the drop angle is 19° for the mineral oil and 32° for the vegetable oil. The mineral oil has better wettability than the vegetable oil.

Thus, the mineral oil has a better wettability than the vegetable oil. The evenness of the film is not only affected by its wettability. Another parameter influences the affinity of the film with its support: its viscosity. This parameter is more important for vegetable oil since its viscosity is more sensitive to the temperature variations than the mineral oil. The viscosity was measured with a cone/plane rheometer [6].

2.3 Oil application methods

In prefabrication, release oil is applied by automated sprayers for a better control of the oil deposit in quantity and surface. On site, oil is most often sprayed on formworks or spread after spraying with a rubber scraper to remove excess oil from the single layer.

The selected nozzle is an important choice. It is recommended to use the thinnest type of nozzle and to use stainless steel nozzles [7]. For our study, the application is realized with a sprayer with a conic nozzle, such as an Ecospray type. The sprayed film is obtained after a short spraying (0.5 s). The distance of spraying is of about 20 cm. The spraying is done after spraying with a conic nozzle and a rubber scraper.

2.4 Measurement of oil film thickness by weighing

The thickness of the oil films was evaluated by weighing according to their application method. Then, the results were compared to a more precise evaluation done by a technique of analysis based on the alpha radiation (PIXE method) which was patented in 2006 by the CEA of Saclay.

On a 30 × 30 cm2 formwork sample, the oil is applied by spraying and either followed or not by a spreading with a scraper (Figs. 1, 2). Sample (5 × 5 cm2) cut from a new or worn formwork wall before and after the concrete pouring in the 30 × 30 cm2 formwork allow to determine the thickness of the deposited oil. Knowing the oil density and the surface of the wall, we only weigh the sample to determine the thickness.

$$ {\text{e}} = {\frac{\text{V}}{\text{S}}}\,{\text{avec}}\,{\text{V}} = {\frac{\text{m}}{\rho }} $$
(1)

where S is the sample surface (m2), ρ the oil density (kg/m3), m the oil film weight (kg) and e the film thickness (m).

Fig. 1
figure 1

Spraying of the mould surface

Full size image
Fig. 2
figure 2

Application by spraying followed by a scraping

Full size image

The measurements of the thickness of the oil films made by weighing and obtained on the average of four measurements are in Table 3.

Table 3 Values of thickness by the weighing method
Full size table

The thickness of the oil film is lightly linked with the viscosity of the oils (Table 2). When the oils were applied, the mineral oil presents a very volatile homogeneous film (it is necessary to check the application so as to avoid an over dosage) whereas the vegetable oil, which is more viscous, is essentially made of dispersed droplets. So, the vegetable oil is distributed unequally over the surface. Because of the viscosity of this oil coupled with the using of a conic nozzle, the deposit quantity after one short spray is inferior to the mineral oil. About 10 short sprayings are required to cover the whole metallic surface.

3 Method of measurement

3.1 The method principle

It is a variant of the Particle Induced X-ray Emission (PIXE) method known for its elementary analysis of the material by fluorescence X [8]. The excitation of the studied material by the particles (α for example) leads to a deenergization of the atoms, accompanied by α X-ray emissions which are characteristic of the atomic number. This enables to measure the atomic concentrations of the sample (Fig. 3).

Fig. 3
figure 3

Schematization of the PIXE principle

Full size image

The principle of measurement of the oil film thickness lays on the possibilities of the PIXE device and on the properties of the α rays which are particles of material (nuclei of helium made of two protons and two neutrons) launched at a high speed (energy equal to 5.3 MeV).

The thickness of the oil film is measured from the maximum fluorescence X of the steel contained in the metallic plate. The level of steel fluorescence is directly influenced by attenuation of α X-rays in the oil film. From the detection of the emitted X protons (little absorbed by the oil film), the number of α particles which have reached the wall (through the oil film) can be measured. Thus, it is a simple measure by metric absorption α.

The maximum distance covered is of 20 μm. After this value, the attenuation of α X-rays is too strong. A very good sensitivity of the method is observed for an oil film of 1 μm thick.

3.2 Experimental device (C.E.A. patent)

The experimental device (Figs. 4, 5, 6) includes:

  • a ring shaped radioactive source of polonium 210 (210Po) which emits α particles of a 5.3 MeV energy in a very pure way (with no emission of β rays or γ rays). The activity of this source is of 15 MBq. The half-life period is of 138 days. This geometry enables an optimization of the activity of the source,

  • a metallic ring allowing to keep a constant distance between the source-detector device and the sample which contains the oil film,

  • a scintillation detector including a iodide of sodium scintillator (NaI) with a diameter of 32 mm and a thickness of 1 mm, protected by a Beryllium (Be) window whose thickness is 0.2 mm (to obtain a transparency to the X-rays of the iron of more than 90%) then a photomultiplier and a spectrometric chain (pre-amplifier, amplifier and multi-channel analyzer).

This kind of cheap detector, working at an ambient temperature enables a large surface of detection.

Fig. 4
figure 4

System excitation–detection using 210Po

Full size image
Fig. 5
figure 5

Experimental device

Full size image
Fig. 6
figure 6

PIXE instrumentation

Full size image

3.3 Gauging

For practical reasons, Mylar sheets are used to simulate the oil [9]. Indeed, the chemical composition of Mylar ((C10H8O4)n) is similar to the oil and the low thickness of the sheets available on the market enables a precise gauging.

The reduction of the α particles essentially depends on the surface mass (exponential law) and little on the chemical composition (for products chemically similar such as the oil and the Mylar made up from carbon, hydrogen and oxygen).

Two spectrums were created (Fig. 7): one with a clean metallic surface and the other with a layer of Mylar which is several millimeters thick (infinite thickness for the α rays).

Fig. 7
figure 7

Total spectrum of a steel surface (threshold 10–100%)

Full size image

When moving a threshold at about 4 KeV and counting all the impulses which are superior to this threshold, the signal noise ratio is excellent. Indeed, the left hand side of the curves (<4 KeV), which is essentially due to an electronic noise, can be eliminated. The counting rate (number of detected photons divided by the counting time) is obtained with integration times of 100 s. The gauging is done from Mylar sheets with 3.5 μm thick. But as this thickness is not a precise value and is not easy to control, the mass surface of each disc is measures, which have a 6 cm diameter by weighing. The Mylar has a density of 1.39, so the thickness of the discs can be calculated (Table 4). As Mylar sheets are used to simulate the oil, it is possible to define an equivalent thickness of oil from the following equation:

Table 4 Thicknesses of Mylar disc measurement
Full size table
$$ {\text{Equivalent}}\,{\text{thickness}} = {\text{Thickness}}\,{\text{of}}\,{\text{the}}\,{\text{disc}} \times {\frac{{{\text{density}}\,{\text{of}}\,{\text{the}}\,{\text{mylar}}}}{{{\text{density}}\,{\text{of}}\,{\text{the}}\,{\text{studied}}\,{\text{oil}}}}} $$
(2)

Results are presented in Table 5. The gauging curves are given in Figs. 8 and 9. With the Tcwin 4.06 program, it is possible to obtain an approximate equation of the equivalent thickness (x) of the oil film according to the attenuation α (α = a + bx3 + c.lnx).

Table 5 Recapitulative values of gauging measurement
Full size table
Fig. 8
figure 8

Gauging of mineral-based oils

Full size image
Fig. 9
figure 9

Gauging of vegetable-based oils

Full size image

After the determination of the coefficients, the following equations are obtained:

$$ {\text{For}}\;{\text{the}}\;{\text{mineral}}\;{\text{oil:}}\;\alpha = 40.59 + 1.33 \times 10^{ - 6} {\text{x}}^{3} - 9.12\;\ln \;{\text{x}} $$
(3)
$$ {\text{For}}\;{\text{the}}\;{\text{vegetable}}\;{\text{oil:}}\;\alpha = 36.93 + 1.17 \times 10^{ - 6} {\text{x}}^{3} - 8.28\;\ln \;{\text{x}} $$
(4)

3.4 Precision obtained on a measure

The disintegration of the radio-elements is a random phenomenon. The counting of the emitted photons are random variables which follow the Poisson law. The absorption law of the α particles is similar to an exponential law (Figs. 8, 9):

$$ {\text{R}} = {\text{R}}_{0} {\text{e}}^{{{{ - {\text{x}}} \mathord{\left/ {\vphantom {{ - {\text{x}}} {{\text{x}}_{0} }}} \right. \kern-\nulldelimiterspace} {{\text{x}}_{0} }}}} $$
(5)

R is the measured counting rate, R0 is the counting rate with no oil, x is the thickness of the measured oil and x0 is a characteristic constant of the absorption of the oil.

During a measuring period equal to t, N photons can be count:

$$ {\text{N}}_{0} = {\text{R}}_{0} {\text{t}} $$
(6)

From Eqs. 5 and 6,

$$ {\text{N}} = {\text{N}}_{0} {\text{e}}^{{{{ - {\text{x}}} \mathord{\left/ {\vphantom {{ - {\text{x}}} {{\text{x}}_{0} }}} \right. \kern-\nulldelimiterspace} {{\text{x}}_{0} }}}} $$
(7)

So,

$$ {\text{x}} = {\text{x}}_{0} \ln \left( {{{{\text{N}}_{0} } \mathord{\left/ {\vphantom {{{\text{N}}_{0} } {\text{N}}}} \right. \kern-\nulldelimiterspace} {\text{N}}}} \right) $$
(8)

The thickness of the measured oil x is equal to a constant −x0ln(N).

When deriving and respectively replacing the increments dx and dN by their fluctuations σx and σN, the following equation can be written:

$$ \sigma_{\text{x}} = - {\text{x}}_{0} \times {\frac{1}{\text{N}}} \times \sigma_{\text{N}} $$
(9)

In the Poisson low, the coefficient of variation is equal to the average value, so \( \sigma_{\text{N}} = \sqrt {\text{N}} , \)

$$ \sigma_{\text{x}} = - {\text{x}}_{0} \times {\frac{{\sqrt {\text{N}} }}{\text{N}}} = {\frac{{ - {\text{x}}_{0} }}{{\sqrt {\text{N}} }}} = {\frac{{ - {\text{x}}_{0} }}{{\sqrt {{\text{R}} \times {\text{t}}} }}} $$
(10)

When replacing R by its value function to x,

$$ \sigma_{\text{x}} = - {\frac{{{\text{x}}_{0} }}{{\sqrt {{\text{R}}_{0} \times {\text{t}} \times {\text{e}}^{{{{ - {\text{x}}} \mathord{\left/ {\vphantom {{ - {\text{x}}} {{\text{x}}_{0} }}} \right. \kern-\nulldelimiterspace} {{\text{x}}_{0} }}}} } }}} $$
(11)

As an example, with the values found experimentally for the mineral oil (x0 = 8 μm and R0 = 30 c/s), for a counting of 1 min, the standard deviation on the measure of the oil thickness is given to the Table 6.

Table 6 Standard deviation on the thickness measure
Full size table

Of course, it is possible to count a period superior to a minute to increase its precision. The standard deviation on the measure is reduced with the square root of the measuring time.

3.5 Comparison between PIXE and weighing measurements

The thicknesses found for the sprayed films that are scraped with the PIXE method, are obtained with and average of four measurements (Table 7). The thickness of the oil film is given as an interval of value rather than an average because of the precision and the high sensitivity of the PIXE technique. Indeed, the weighing technique gives a global value of the studied disc (5 × 5 cm2) whereas the PIXE technique gives a value much more local on the studied surface. Oil does not have an homogeneous thickness. This can explain the difference between results with these two techniques (Tables 3, 7). The variation of the thickness of the oil film on the same surface may vary of a micron according to the used technique.

Table 7 Measures of the thicknesses of the oil films with the PIXE method
Full size table

In the case of the scraped films (homogeneous thickness on the whole surface), measures obtained from the PIXE technique show a better reproducibility.

The PIXE technique enables to realize measures directly on the surface of the formworks or of the prefabrication moulds.

4 Measurement of the thicknesses of the oil films when concrete is poured into a formwork

To create the conditions of a concrete poured into a formwork, a plane/plane tribometer is used. In building trade and civil engineering, each pouring of concrete requires particular pressure scales and velocities. Filling velocities are about 6 m/h for pressures inferior to 70 kPa for standard 2.5 m high formworks. The thickness of the oil films are measured before and after the concrete pouring.

4.1 Principle of the tribometer

A plate slides in a translatory movement between two samples of concrete (Fig. 10). The plate has been cut into formworks walls by a formwork building firm.

Fig. 10
figure 10

Schematic representation of the tribometer principle

Full size image

The movement is done with a motor connected to an endless screw. The samples are pressed thanks to jacks. The sample-holders are cylinders with a 120 mm diameter (Fig. 11). These one are equipped with a moving bottom to introduce the concrete. A gasket system is set on the sample-holders so as to avoid any leak of concrete and not to damage the oil film which is applied on the plate. This device has already been described in a lot of publications [10].

Fig. 11
figure 11

Details of sample holder

Full size image

4.2 Measurement of the thicknesses of the oil films before and after the pouring of the concrete

The aim of this test is to know if an oil film remains on the formwork wall after the concrete demoulding.

4.2.1 Influence of the watertightness system on the oil film

The influence of the watertightness system on the oil film is determined while measuring after the friction of the joint of the sample-holders. The impact of the concrete friction will be observed by measuring the thickness of the oil film. So as to avoid the influence of the calcium element (Ca2+) from the concrete on the measurement of the film thickness, the spectrum will be yielded from 37 to 60%. The spectrum is thus focused on the iron element.

The measurements are done on the oil film 30 min after its application according to the advice of the contractor. Several measurements are done on the length of the plate so as to obtain an average thickness.

The acquisition time is of 120 s. To measure the film, the PIXE drill is directly placed on the plate (Fig. 12).

Fig. 12
figure 12

Measurement of oil film on the tribometer plate

Full size image

4.2.2 Measurement of the oil film after the concrete friction

Results of the thicknesses of the film after the concrete friction under pressure for an application of the sprayed oil are given in Table 8. For an application of the film by spraying followed by a scraping, results are given in Table 9.

Table 8 Thickness of the oil films after the concrete sliding for a sprayed application
Full size table
Table 9 Thickness of the oil films after the concrete sliding for a sprayed application followed by a scraping
Full size table

For sprayed films, a reduction of the thicknesses of the film can be observed. This reduction is more important for the vegetable oil. For scraped films, the thickness of the film is reduced to about 1 μm for a pressure of 150 kPa. The reduction is more important for vegetable-based oils. A film of 1 or 2 μm is enough for vegetable-based oils to obtain a good removing of the concrete [3]. This condition is respected until 150 kPa (a concrete wall of 6 m high).

For more important heights, the reliability of an oil film for the removing of the concrete is not guaranteed. Some defects could thus appear on the facing.

Thicknesses variations on the initial film are really reduced after the concrete friction, mainly at 150 kPa. This means that the final thickness little depends on the oil (viscosity) but is mainly due to the method of application.

5 Measurement of the thicknesses of oils films before the pouring and after the removing of a concrete in a metallic mould

Some 30 × 30 × 30 cm3 moulds have been created by the same formwork builder as the one which made the plate for the tribometer (Fig. 13). The metallic plates of the formwork are independent. The volume of the moulds is of 27 l.

Fig. 13
figure 13

Photograph of a mould

Full size image

The roughness of the moulds surfaces with new plates is of Ra = 1.30 μm. With worn plates the roughness is Ra = 1.70. Ra is the international parameter of roughness which is the most used. It represents the mean arithmetic deviation in comparison with a medium line. This parameter is measured with the help of a mobile roughness device called Surtronic 3+.

The oils have been applied on the moulds surfaces with a conic-base sprayer. The value of the oil film is the average of two points taken at the top and at the bottom of the mould (Fig. 14). The films measurement is the same as the one described previously.

Fig. 14
figure 14

Presentation of the zones on the surface of the mould where the thickness is measured

Full size image

The concrete is poured at the same time in all the moulds. Each mould is filled with two layers followed by a vibration at five different places during 10 s for each layer. The concrete has been removed after 16 h. The results are presented in Table 10. For low viscous oils, droplets are very fines and beyond our control to avoid a surplus of film, volatility drops can entail a superposition of the film. The thickness is increased. This is not observed for a viscous film.

Table 10 Values of the film thickness before the pouring and after the removing of the concrete
Full size table

For the mineral-based oil, an oil film very thin remains on the mould surface after dismantling this one and the concrete. The thickness of this film is about 0.6 μm.

For the vegetable-based oil, the value is very small (thickness = 0.15 μm). Oil traces remain on the formwork. The initial thickness is smaller than the thickness of the mineral-based oils and this can explain the slight presence of oil on the mould after the removing of the concrete.

The presence of release agent after the concrete removing enables to protect the surface of the formwork against corrosion and an easy removing of the concrete. It thus plays its role at the interface concrete/formwork but does it allow to obtain a high quality facing?

6 Study of the aesthetical aspect of the facings and the surfaces of the moulds

For this part of the study, aesthetical defects [11, 12] observed on the concrete wall after its removing is investigated. The two ways of applications described previously are analysed. The parameters retained are the following ones (Table 11):

Table 11 Visual criteria of grading
Full size table

  • for concrete facing: the variation of colour which is evaluated by a grey scale (Pa), The micro-bubbling (B), the dusting of the concrete facing (Pop),

  • for formwork surface: the dirtying (E), the dusting (Pom) and the attachments points (Pt Ac).

For each aesthetical defect, an appreciation of the facing quality is made with a visual approach (Table 11). For example, when no dusting on the concrete facing is observed, facing quality is considered excellent (++). In the same manner, for formwork surface with a lot of attachment point, the facing quality is considered bad (− −).

6.1 Application of the oil by spraying

The grading of the facing for an application by spraying is given in Table 12. This grading is obtained from visual appreciations presented in Table 11. They correspond to an average of three independent mixing. The films have a thickness of about 2 μm.

Table 12 Average of the grading for the facings and the aspects of the surface of the mould for a sprayed film
Full size table

For the mineral-based oils, the quality of the facings is not high but quite acceptable. An important dirtying of the surface of the moulds can be observed (Fig. 15).

Fig. 15
figure 15

Dirtying of a mould

Full size image

For the vegetable-based oils, the facings are globally of the better quality compared to those with mineral-based oils. The facings have little micro-bubbling and dusting on the surfaces of the moulds.

6.2 Application by spraying followed by scraping

The average of the grading for facings and aspects of the mould surface with a spraying followed by a scraping is done in Table 13.

Table 13 Average of the grading for the facings and the aspects of the surface of the mould with a spraying followed by a scraping
Full size table

For mineral-based oils, results are quite similar to those obtained with a spraying. However, a weak diminution of bubbling on concrete facing for oil M is denoted.

For vegetable-based oils sprayed then scraped, results are of a better quality than with the sprayed oils. However, the facings have more micro-bubbling and dusting. The facings have still a very good aesthetical quality. The scraping thus improves aesthetical aspect of the facings.

7 Conclusions

To conclude, the PIXE method seems to be really adapted to this kind of measurement. The range of this study (thickness between 0 and 10 μm) is particularly well covered. The thicknesses of film vary according to the viscosity of the employed oils. When scraping a sprayed film, thicknesses are between 1 and 2 μm.

Therefore, for the sprayed oil films, the thicknesses vary from 2 to 5 μm. Moreover, a very thin film remains on the walls of the formwork after the removing of the concrete. This film is slightly more important for the mineral-based oils due to the difference of the sprayed quantity. The surfaces of the formwork will than be better protected against corrosion.

Vegetable oil films have a good resistance after the concrete friction. The static study of the concrete in the moulds shows that a film from 1 to 2 μm is enough to obtain a high quality facing, mainly with the vegetable oil. Indeed, they better reduce the concrete friction on the formworks walls and this entails a low reduction of the applied film.

Moreover, the studied vegetable oil allows obtaining a facing with a better aesthetical quality than the mineral oil. The low thickness of the vegetable oils is to be used on building sites and at a prefabrication stage so as to obtain high quality facings without any defects.