Method and system for inspecting ply-by-ply machining of multilayer materials

Abstract

The invention relates to achieve a rapid, reproducible and reliable characterization of the quality of ply-by-ply machining of multilayer materials. This method for inspecting ply-by-ply machining of a part (10) made of multilayer composite material under repair by machining a ply-by-ply staggered or continuously sloped cut out in a stack of plies of various successive orientations includes taking images (IA to ID), under lighting of different orientations (12), of a surface area (10a) of the machined part (10) to be inspected; performing an analysis by comparing the images (IA to ID) pixel by pixel (P0) in order to define the orientation of each pixel (P0) as corresponding to that of the image in which this pixel has a higher brightness; if the pixel has a similar brightness in all the images (IA à ID), this pixel (Pr) is attributed to a resin; constructing a map (5) in units of ply of the surface area to be inspected (10a) by applying the preceding analysis to all of the pixels; estimating a machining quality level from the map (5) produced, and archiving (2m) each map (5) thus produced as a machining result.

Claims

1. A method for inspecting a ply-by-ply machining of a part under repair made of a fiber-resin multilayer composite material by ply-by-ply machining a stepped or continuously sloped cut-out in a stack of plies with different successive orientations, the method comprising the steps of: capturing images, by illuminating from different orientations a surface of the part; transmitting the image to a digital processing unit; analyzing each one of the images by comparing a pixel by pixel to determine the image in which the pixel has a greater brightness; if the pixel has similar brightness in all the images, then, the pixel is attributed to the resin; constructing a map of the surface to be inspected by applying the analyzing step to all the pixels; estimating a level of machining quality from the map as a function of the brightness homogeneity of pixels on the images; and archiving each map produced in a memory on the digital processing unit.

2. The ply-by-ply machining inspection method according to claim 1, further comprising before the archiving step, the step of determining a zone by zone, the surface distribution of the pixels and a resin orientation to determine a predetermined machining tolerance.

3. The ply-by-ply machining inspection method according to claim 2, wherein the determining step is followed by a step of determining the machined depth.

4. The ply-by-ply machining inspection method according to claim 3, wherein the machined depth is determined by a graphical report of its surface distribution as a function of different depths.

5. An automated system for ply-by-ply repair machining inspection of a surface of a multilayer material part comprising: a digital data processing unit connected to a controller of light sources and a controller of at least one imaging video camera; wherein the light sources are distributed on linear lighting strips mounted on adjacent light walls successively oriented to form a regular polyhedron; wherein the video camera being disposed on a central axis and recording image brightness signals corresponding to the lighting of pairs of strips of light sources of opposite orientation on the part to be inspected and successively activated by the controller, and in that a converter of the image signals is adapted to transmit brightness digital data of images corresponding to the different ply orientations to the digital processing unit; wherein the digital processing unit performs the analysis of brightness digital data according to the method of claim 1.

6. The automated machining inspection system according to claim 5, wherein the video camera and the light sources are fixed to an XY mobile table controlled by the processing unit to position the video camera and the lighting strips in order to produce an assembly of elementary images recorded by the video camera on lighting oppositely oriented pairs of lighting strips activated successively by the controller.

7. The automated machining inspection system as claimed in claim 5, wherein the digital processing unit includes a memory module for archiving the brightness data, pixel orientation data, map data, and an estimated level of machining quality obtained by processing image signal data.

8. The automated machining inspection system according to claim 7, wherein the memory module of the digital processing unit also includes machining tolerance data predetermined as a function of the material and the mechanical characteristics of the part.

9. The automated machining inspection system as claimed in claim 5, wherein the polyhedron of the light walls is an octagon and the light sources are light-emitting diodes aligned along each face of that octagon.

Description

DESCRIPTION OF THE FIGURES

(1) Other data, features and advantages of the present invention will become apparent on reading the following nonlimiting description with reference to the appended figures, in which:

(2) FIGS. 1a and 1b (already commented on) are sectional views of a multilayer panel machined for a local repair (FIG. 1a) and plugged by a patch (FIG. 1b) using the known glued repair technique;

(3) FIG. 2 shows a top view of an example of an inspection system according to the invention with light walls forming an octagon around the imaging video camera;

(4) FIG. 3 shows four views of images (I.sub.A to I.sub.D) captured by the video camera from FIG. 2 with four respective lighting orientations produced successively by the light walls from FIG. 2;

(5) FIG. 4 shows a map of the machined surface obtained by selection of the orientation of each pixel based on a comparison of the brightness values of that pixel in the images from FIG. 3, and

(6) FIG. 5 shows a graph of the evolution of the inspected surface percentage of different ply and resin orientation phases for different depths expressed in ply units.

DETAILED DESCRIPTION OF EMBODIMENTS

(7) Referring to the FIG. 2 top view, an automated repair machining inspection system 2 according to the invention includes eight light walls 11a to 11 h forming a regular octagon 11 with central axis Z′Z. Each pair of adjacent walls, for example walls 11a and 11b, has an angular offset of 45° and two walls symmetrically opposite with respect to the axis Z′Z, for example walls 11a and 11e, are parallel.

(8) Each wall 11a to 11h integrates a linear strip of light sources 12, six light-emitting diodes “LED” 12a in the example shown. The lighting by each light strip 12 has for its angle that of the wall 11a to 11 h on which the lighting strip is fixed and the strip pairs of two opposite walls, and thus of opposite orientation, are electrically interconnected.

(9) Taking as reference a plane II-II perpendicular to two opposite walls, the walls 11a and 11e in the example, the lighting by the strips 12 fixed to two opposite walls 11a and 11e, 11 b and 11f, 11c and 11 g as well as 11d and 11 h is oriented with angle differences respectively equal to 0°, 45°, 90° and 135° on the composite material panel 10 to be inspected, e.g. disposed parallel to the plane of the lighting strips 12. The lighting strips 12 are advantageously fixed to an XY mobile table 13.

(10) The automated system 2 also includes a digital imaging video camera 21, also fixed to the XY mobile table 13, including a lens 2a matched to the spectral band of the LEDs 12a. The lens 2a, aligned on the central axis Z′Z, is advantageously fitted with a polarizing filter 2b in order to generate reflection-free images of the surface 10a of the panel 10 machined for glued repair. More generally, it is advantageous to adapt the lens, the addition of filters, the type of photosensitive cells of the video cameras as a function of the materials inspected.

(11) Moreover, account is advantageously taken of the main orientations of the layers of the multilayer material to be inspected in order to define the optical characteristics of the video camera and the number of lighting strips so as to use pertinent angle differences between the strips in order to cover all of the repair zone.

(12) Moreover, the automated system 2 includes a digital data processing unit 23 with an integral memory module 2m connected to a controller 12c of the lighting strips 12 and a controller 21c of the imaging video camera 21. A signal converter 25 also integrated into the processing unit 23 converts the image signals into digital data that can be exploited by the processing unit 23.

(13) In operation, lighting by the opposite strips 12 is successively activated by the controller 12c and the video camera 21 records an image for each orientation of the lighting strips 12, the digital processing unit 23 managing all of the controllers.

(14) Referring to the four image views I.sub.A to I.sub.D in FIG. 3 obtained in this way, the zone covered by the video camera 21 and the lighting strips 12 is advantageously positioned in the plane XY by the table 13 (cf. FIG. 2) to enable a complete or at least representative inspection of all of the machining. The four images I.sub.A to I.sub.D obtained for the four lighting orientations 0°, 45°, 90° and 135°, coinciding with the orientations of the plies, extend over a wide area 300×130 mm.sup.2. To be more precise, the XY mobile table 13 is controlled by the digital processing unit 23 to enable assembly of elementary images that individually extend over approximately 15×15 mm.sup.2. The converter 25 of the image signals recorded by the video camera 21 transmits digital data to the digital processing unit 23 to supply the image information exploited below (cf. FIG. 2).

(15) The images I.sub.A to I.sub.D show different semicircular steps Mi around the bottom ply Px, the steps Mi being obtained after ply-by-ply machining for subsequent repair of the machined composite material panel 10 by means of a patch with a complementary configuration (cf. FIGS. 1a and 1b). The brightness of the steps Mi differs according to the orientation of the lighting strips. These brightness differences are exploited by the generation of a surface map for estimating the level of machining quality (see below).

(16) This kind of surface map 5 of the machined panel 10 to be inspected is shown in FIG. 4. This map 5 is produced by analyzing the four images I.sub.A to I.sub.D from FIG. 3. The analysis consists in pixel by pixel comparison of the four images I.sub.A to I.sub.D of the machined surface 10a calibrated in terms of ply units in the following manner. To be more precise, if the brightness of a pixel P.sub.0 of one image, for example the image I.sub.A, is greater than that of the same pixel in the other images I.sub.B to I.sub.D, the pixel P.sub.0 is considered to have the orientation of the image I.sub.A. If a pixel Pr has similar levels of brightness in the four images I.sub.A to I.sub.D, that pixel Pr is considered to be from the resin.

(17) This surface map 5 then enables direct, automated, rapid and reproducible estimation by the processing unit 23 of the quality of machining as a function of the homogeneity of the assignment of pixels to the various images I.sub.A to I.sub.D corresponding to the various ply orientations. This direct estimate is digitally archived in the memory module 2m of the processing unit 23.

(18) The surface percentages of the five phases the four ply and resin orientations of the surface map 5 may advantageously also be exploited by the processing unit 23 per zone, in the example per step Mi, and define a percentage of each ply orientation for a reference orientation, a 45° orientation in the example: 85% of plies at 45°, 8% of resin and 7% of plies at 90°. These percentages are then used to validate the machining tolerance expressed in ply units for this panel 10, as referenced in the memory module 2m of the processing unit 23 (cf. FIG. 2). The map is also archived in this module 2m.

(19) The surface percentages of the five phases of the map 5 also make it possible to define machined depths expressed in ply units, as the graph “G” in FIG. 5 shows.

(20) This graph “G” shows the evolution of the surface percentage curve C.sub.P of controlled phase Ph % on two consecutive plies P+1 and P−1 for different depths P % expressed in ply units situated around the interply interface taken as a 100% reference. The “resin” phase percentage curve Cr is also shown.

(21) Each depth P % expressed in ply units, for example 80% on the FIG. 5 graph, corresponds to specific distributions substantially complementary to the inspected surface Ph % in the plies P+1 and P−1, respectively approximately 10% and 35% in the example, with approximately 55% resin. The graph “G” is also archived in the memory module 2m (cf. FIG. 2).

(22) The invention is not limited to the examples described and shown. The method according to the invention may be totally, semi or partially automated. Also, the number of light walls may be 6, 10, 12 or more, and the number of light sources per strip may also vary. Moreover, two video cameras positioned face to face may be used. Additionally, the light walls need not be coupled in pairs.