It is well-known fact that to draw out maximum penetrant entrapment from a defect the presence of direct contact between a penetrant in defect's mouth and a developer is required. According to EN 571-1 the removal of excess penetrant and following application of developer have to be accomplished in such a way when penetrant should not be wiped out or evaporated. The evaporation of penetrant results in the displacement of meniscus position in defect's channel. An air gap between a penetrant and a developer, subsequently applied to tested surface, arises and hinders their direct contact. Consequently, the drawing of a penetrant from defect's cavity due to capillary penetration does not take place.
We carried out some special experiments and established that even in the absence of direct penetrant-developer contact the liquid's drawing out from a defect and subsequent formation of an indication take place. A number of video-images of such a process for various liquids drawn from model defects were obtained.
The experiments were conducted with model defects. As model defects the glass cylindrical dead-end capillaries of length l = 3...6 mm and radius R = 100...240 m
m were used.
The solutions of ethanol in water with the additive of a dye were used as the penetrants. The concentration of ethanol in solution was 25, 50, 75 %. The presence of a dye allowed an indication of a model defect on a developer surface to be visualized.
Metallic and kaolin powders and various kinds of wet developers containing an alcohol as basic component were used as the developers. The powder developers were applied on a glass plate and with a given effort were pressed by other glass plate. The transparency of a glass allowed to observe the processes of forming and growth of an indication. Suspension developers were put also on a glass plate with subsequent drying. As a result of drying process the wet developer forms a uniform, dry, porous structure with good adhesion to a glass plate, that allows to conduct researches at different positions of a model defect and developer layer.
The glass plate with a developer layer was anchored on the holder. Another holder fixed the capillary filled with a liquid. To increase the filling depth the capillary was heated in the oven up to temperature T = 80 C, then it was immersed in penetrant. Before fixation the capillary was wiped with a napkin for removal penetrant excess. To realize different contact conditions between a capillary and developer layer the wiping was implemented in two ways: in one case the capillary was wiped dry on a lateral and end area, in the other case the small drop of a liquid was left on its end. Through the moving gear of a microscope the bracket with a fixed capillary moved down to touch the capillary end with a developer layer (in this case there is a small deformation of a developer layer).
The observations of mass transfer processes were conducted in two directions: from above, with the purpose of visualization the indication, and from sideways for visualization the process of meniscus movement. The registration of indication growth was made through a microscope. For obtaining a necessary contrast the appropriate concentration of special dye in a liquid was selected. Mass transfer processes in model defects were observed with a help of video camera. The analog signal from the video camera was digitized by video blaster and transferred to IBM PC.
The experiments on development process show that depending on contact conditions between a liquid in a capillary and developer layer the liquid's drawing from a capillary is determined by one of two different physical mechanisms. The first one corresponds to the case of direct contact of a liquid with a porous layer, when there is no an air gap between a liquid column and powder's particles at the contact moment. In this case liquid's extraction from a capillary into the developer layer is determined by classical capillary penetration process. The liquid is soaking into the pores and the principal moving force is capillary pressure in porous layer.
Some stages of such extraction process are shown in Fig. 1. Here the capillary dimensions and the characteristics of developer layer are as follows: capillary diameter d = 120m
m, the length l0 = 5,8 mm, the width of developer layer h = 136 m
m, the porosity of the developer P
= 0,36, effective radius of pores Ref = 3 m
m, maximum diameter of the indication D = 600 m
m. Such development process is described by classical penetrant imbibition due to capillary pressure in porous developer . One can see from Fig. 1 that the development process duration takes just a few seconds. It is in good agreement with the calculated data.
Fig 1: Penetrant extraction under classical development mechanism.
a - initial moment of the contact; b - t = 1 s; c - t = 2 s; d - t = 3 s.
In the second case, when the capillary end is completely dry, there is a concave meniscus in the channel and, moreover, - an air gap. The pictures appropriate to this case of a development process are shown in Fig. 2. The pictures show that the development is accompanied by appearance and growth of a vapour-gas mixture column in the mouth of capillary channel. Capillary dimensions and developer layer characteristics in this case are following: d = 103 m
m, l = 5,4 mm, h = 48 m
= 0,34, Ref = 3 m
m , D = 112 m
m . To the point to note that the lower meniscus is motionless, and duration of extraction process is large enough (more than 5 min).
Fig 2: Penetrant extraction in the case of formation and growth
of vapour-gas column near the capillary outlet.
a - initial moment of the contact; b - t = 1 min; c - t = 3 min; d - t = 6 min.
Time dependence of the extracted liquid volume, the volume of impregnated layer and the area of the indication are presented in Fig. 3. As one can see, the liquid has appeared on the outer surface of developer layer after 1 min, and the area of impregnated zone was growing during 4-5 min. Fig.3 illustrates the fact that the volume of the liquid absorbed in the developer is substantially smaller than the liquid volume loss in the capillary.
Fig 3: Time dependencies of the volume of extracted liquid (1),
volume of impregnated layer (2) and area of the indication (3).
The analysis of these dependencies enables to clarify the mechanism of development process. It follows from our results that new mechanism of penetrant drawing from defect's channel is based on liquid film flow. The liquid flows due to the action of so called disjoining pressure pd in thin adsorptive film, which is defined as the difference between the normal component of the pressure in a film pN and the pressure p0 in a bulk phase of a liquid :
pd = pN - p0.
In an equilibrium state the value pd is constant along the film width and is equal to pressure of a gas phase P, whereas the tangential components of pressure are variable quantities along film cross section, but do not depend on surface coordinates. Thus, the film in a capillary is not in the equilibrium state. The width of a film in a capillary is decreasing with the distance from a meniscus, and the flow of a liquid takes place in a direction of film width decrease. It means that the gradient of the value Pi = P - pd(h) plays a role of hydrodynamic pressure in a film.
The dependence of disjoining pressure on a width of liquid film h - the isotherm of disjoining pressure pd(h) - is main characteristic of a liquid thin film. Disjoining pressure is determined by the surface forces. The analytical expression for main component of disjoining pressure, derived with a use of molecular - kinetic theory, is as follows :
where A is the constant. We use this well-known approximation for the isotherm of disjoining pressure and solve classical Navier-Stokes equation for film flow in z direction with ¶
z as the moving force.
The adsorptive film of a liquid on an internal lateral area of a capillary tube contacts in the outlet of a channel (on capillary's butt end) with a grains of developer. The particles of a powder are also covered with adsorptive films of a liquid, and the curvature of these films is larger than the curvature of a film in a capillary channel, with which they contact. A liquid flows towards the particles of a developer from a capillary wall covering the grains. Thus, as a result of film flow the liquid is extracted from a capillary channel. The similar process takes place at double-sided filling of conical capillary with a liquid, and this process has good conformity between theoretical and experimental results . As far as a liquid only envelops the grains of developer, in contrast to complete filling of porous region at capillary imbibition, there is a vapour-gas exchange between a cavity of a capillary channel and the atmosphere through the inter-grain cavities. .
It is possible to assume that penetrant evaporation from a surface of a meniscus will cause the transfer of a dye to a developer layer. Special experiments were conducted to clarify this hypothesis. The cooled object was inserted into the test-tube above a penetrant at temperature of 40-45 C, then the condensed drops of liquid were observed through the microscope to detect the presence of dye. The results of experiment have shown that there is no dye transfer during penetrant evaporation - condensation. It allows explaining, why the volume of a liquid absorbed in a developer is substantially smaller than the volume extracted liquid.
The processes of film flow and evaporation of a liquid from a meniscus surface take place simultaneously, but the film flow transfers dye particles and results in colouring a developer. Thus, the processes of film flow and the evaporation have the competing character. The closer a meniscus of a liquid to a developer layer, the less hydraulic resistance of a film and the greater the influence of mass transfer of a liquid due to film flow. It explains the termination of indication growth after 4 min (Fig. 3) at presence of a liquid loss from the capillary channel. It means, that the process of evaporation already plays a determinative role. It is necessary to note that for such action of a development, the process of evaporation results in decrease of penetrant testing sensitivity.
The determined new regularities of mass transfer processes during the extraction of a liquid from a capillary by a porous layer is important for understanding the physical processes underlying the stages of a development at penetrant testing. Besides they allow optimizing the procedure of penetrant testing.
This paper was funded by Belarussian Foundation of Fundamental Research (Project T97-051).
- Prokhorenko P. P., Migoun N. P. Introduction to Theory of Penetrant Testing (in Russian), Minsk, Science & Engineering, 1988, 207 p.
- Deryagin B.V., Churaev N.V. Wetting films (in Russian), Moscow, Science, 1984, 160 p.
- Prokhorenko P.P., Migoun N.P., Stadthaus M. Theoretical Principles of Liquid Penetrant Testing, Berlin, DVS-Verlag, 1999, 252 p.