阅读说明：本技术 纤维材料层中熔化的薄纱绞线比率的表征 (Characterization of the ratio of melted tissue strands in a layer of fibrous material ) 是由 M·A·沙尔斯基 于 2019-06-05 设计创作，主要内容包括：本申请公开了纤维材料层中熔化的薄纱绞线比率的表征。提供了用于表征具有热塑性薄纱的纤维层的系统和方法。一种方法包括：获取纤维材料的图像,该纤维材料包括纤维的绞线并且还包括热塑性塑料的细丝的薄纱；将该图像细分成片；确定在每个片内描绘的熔化细丝的量；以及确定在每个片内描绘的未熔化细丝的量。(The present application discloses characterization of the ratio of melted tissue strands in a layer of fibrous material. Systems and methods for characterizing a fibrous layer having a thermoplastic veil are provided. One method comprises the following steps: acquiring an image of a fibrous material comprising strands of fibers and further comprising a veil of filaments of a thermoplastic; subdividing the image into slices; determining an amount of melted filament traced within each sheet; and determining the amount of unmelted filament traced within each sheet.)

1. A method for characterizing a fibrous layer having a thermoplastic veil, the method comprising:

acquiring an image of a fibrous material comprising strands of fibres and further comprising a veil of filaments of a thermoplastic (202);

subdividing the image into slices (204);

determining an amount of melted filament (206) delineated within each of the sheets; and

the amount of unmelted filament depicted within each of the sheets is determined (208).

2. The method of claim 1, further comprising:

quantifying (210) an amount of melting of the tissue for each sheet based on the number of melted filaments and the number of unmelted filaments for that sheet; and

quantifying (212) an amount of fusing of the tissue for the image based on an amount of fusing of the tissue for each sheet.

3. The method of claim 1, wherein:

determining the amount of the molten filament includes identifying a portion of each sheet that includes the molten filament; and is

Determining the amount of unmelted filaments includes identifying a portion of each sheet that includes unmelted filaments.

4. The method of claim 1, wherein:

the determining step is performed by a neural network that has been trained.

5. The method of claim 1, further comprising:

comparing the amount of fusing of the tissue against the image to a design tolerance; and

sending a notification in response to determining that the amount of fusing of the tissue for the image is not within the design tolerance.

6. The method of claim 1, further comprising:

the melted filaments and the unmelted filaments are distinguished based on a difference in at least one of brightness or color.

7. The method of claim 1, wherein:

the strands of fibers are selected from the group consisting of carbon fibers, glass fibers, metal fibers, and ceramic fibers.

8. A portion of an aircraft assembled according to the method of claim 1.

9. A non-transitory computer readable medium containing programming instructions that are operable when executed by a processor to perform a method for characterizing a fibrous layer having thermoplastic tissue, the method comprising:

acquiring an image of a fibrous material comprising strands of fibres and further comprising a veil of filaments of a thermoplastic (202);

subdividing the image into slices (204);

determining an amount of melted filament (206) delineated within each of the sheets; and

the amount of unmelted filament depicted within each of the sheets is determined (208).

10. An apparatus for characterizing a fibrous layer having a thermoplastic veil, the apparatus comprising:

an interface (716) that receives an image (742) of a fiber material (750), the fiber material (750) including strands (752) of a fiber and further including a veil (754) of filaments (760) of a thermoplastic; and

a controller (712) that subdivides the image into patches (744), determines an amount of fused filament depicted within each of the patches, and determines an amount of unfused filament depicted within each of the patches.

11. The apparatus of claim 10, wherein:

the controller quantifies an amount of melting of the tissue for each sheet based on a number of melted filaments and a number of unmelted filaments for that sheet, and quantifies an amount of melting of the tissue for the image based on an amount of melting of the tissue for each sheet.

12. The apparatus of claim 10, wherein:

the controller operates a neural network (724), the neural network (724) determining an amount of the melted filament by identifying a portion of each sheet that includes the melted filament; and is

The controller determines the amount of unmelted filament by identifying the portion of each sheet that includes unmelted filament.

13. The apparatus of claim 10, wherein:

the controller determines the amount of the melted filaments and the amount of the unmelted filaments by a neural network that has been trained.

14. The apparatus of claim 10, further comprising:

comparing the amount of fusing of the tissue against the image to a design tolerance; and

sending a notification in response to determining that the amount of fusing of the tissue for the image is not within the design tolerance.

15. The apparatus of claim 10, wherein:

the strands of fibers are selected from the group consisting of carbon fibers, glass fibers, metal fibers, and ceramic fibers.

Technical Field

The present disclosure relates to the field of composite materials, and in particular to fiber reinforced composites including thermoplastic tissue.

Background

Dry carbon fibre materials typically comprise thermoplastic veil (veil) (comprising a plurality of filaments) fused to strands (strand) of carbon fibre. The strands of carbon fibre provide enhanced material stability and toughness, while the thin yarns of thermoplastic (thermoplastic) bind the strands together. Dry carbon fiber material is used as an input for the manufacture of composite parts. To ensure consistent manufacturing quality, it is desirable that the thermoplastic veil of dry carbon fiber material have a certain amount of melting into the strands of carbon fibers. However, inspection of dry carbon fiber materials remains a labor and labor intensive process, which in turn increases the cost of manufacturing composite parts. This is particularly noticeable because thousands of feet of dry carbon fiber material may be used in a single composite part (e.g., the skin of an aircraft wing).

Accordingly, it may be desirable to have a method and apparatus that takes into account at least some of the issues discussed above, as well as other possible issues.

Disclosure of Invention

Embodiments described herein utilize a feature detection process to characterize the amount of melting of thermoplastic veil within a fibrous material (e.g., a unidirectional dry carbon fiber material that is not impregnated with a thermoset or thermoplastic resin, a glass fiber material, a material having metal fibers or even ceramic fibers, etc.). The amount of melting can be characterized by comparing the number of filaments of the tissue that have melted (resulting in a color/brightness change) with the number of filaments of the tissue that have not melted. For example, the machine learning process discussed herein may divide an image of the fibrous material into sheets (slices), detect fused and unmelted filaments within the tissue using a trained convolutional neural network, and determine a ratio of fused and unmelted filaments for each sheet. Then, based on the ratios determined for each patch, an overall measure of the amount of fusing can be determined for the image.

One embodiment is a method for characterizing a fibrous layer having a thermoplastic veil. The method comprises the following steps: acquiring an image of a fibrous material comprising strands of fibers and further comprising a veil of filaments of a thermoplastic; subdividing the image into slices; determining an amount of melted filament traced within each sheet; and determining the amount of unmelted filament traced within each sheet.

Another embodiment is a non-transitory computer readable medium containing programming instructions that, when executed by a processor, are operable to perform a method for characterizing a fiber layer having thermoplastic tissue. The method comprises the following steps: acquiring an image of a fibrous material comprising strands of carbon fibers and further comprising a veil of filaments of a thermoplastic; subdividing the image into slices; determining an amount of melted filament traced within each sheet; and determining the amount of unmelted filament traced within each sheet.

Another embodiment is an apparatus for characterizing a fibrous layer having a thermoplastic veil. The device includes: an interface that receives an image of a fibrous material that includes strands of fibers and also includes a veil of filaments of a thermoplastic; and a controller that subdivides the image into slices, determines an amount of melted filament depicted within each slice, and determines an amount of unmelted filament depicted within each slice.

The apparatus and method of the present invention are also referred to in the following clauses that are not to be confused with the claims.

A1. A method for characterizing a fibrous layer having a thermoplastic veil, the method comprising:

acquiring an image of a fibrous material comprising strands of fibers and further comprising a veil of filaments of a thermoplastic 202;

subdividing the image into slices 204;

determining the amount of melted filament depicted within each sheet 206; and

the amount of unmelted filament depicted within each sheet is determined 208.

A2. There is also provided a method according to paragraph a1, the method further comprising:

quantifying 210 the amount of melt of the tissue for each sheet based on the number of melted filaments and the number of unmelted filaments for that sheet; and

the amount of fusing of the tissue for the image is quantified 212 based on the amount of fusing of the tissue for each sheet.

A3. There is also provided the method of paragraph a1, wherein:

determining the amount of the molten filament includes identifying a portion of each sheet that includes the molten filament; and

determining the amount of unmelted filaments includes identifying a portion of each sheet that includes unmelted filaments.

A4. There is also provided the method of paragraph a1, wherein:

these determining steps are performed by a neural network that has been trained.

A5. There is also provided the method of paragraph a4, wherein:

the neural network comprises a convolutional neural network.

A6. There is also provided a method according to paragraph a1, the method further comprising:

comparing the amount of fusing of the tissue against the image to a design tolerance; and

a notification is sent in response to determining that the amount of fusing of the tissue for the image is not within the design tolerance.

A7. There is also provided the method of paragraph a1, further comprising:

the melted filaments and the unmelted filaments are distinguished based on a difference in at least one of brightness or color.

A8. There is also provided the method of paragraph a1, wherein:

the strands of fibers are selected from the group consisting of carbon fibers, glass fibers, metal fibers and ceramic fibers.

A9. There is also provided the method of paragraph A8, wherein:

the strands of fibres are strands of carbon fibres.

A10. A portion of an aircraft assembled according to the method of paragraph a1.

According to another aspect of the medium of the present invention, there is provided:

B1. a non-transitory computer readable medium containing programming instructions that when executed by a processor are operable to perform a method for characterizing a fibrous layer having thermoplastic tissue, the method comprising:

acquiring an image of a fibrous material comprising strands of fibers and further comprising a veil of filaments of a thermoplastic 202;

subdividing the image into slices 204;

determining the amount of melted filament depicted within each sheet 206; and

the amount of unmelted filament depicted within each sheet is determined 208.

B2. There is also provided the medium of paragraph B1, wherein the method further comprises:

quantifying 210 the amount of melting of the tissue for each sheet based on the number of melted filaments and the number of unmelted filaments for that sheet; and

the amount of fusing of the tissue for the image is quantified 212 based on the amount of fusing of the tissue for each sheet.

B3. There is also provided the medium of paragraph B1, wherein:

determining the amount of the molten filament includes identifying a portion of each sheet that includes the molten filament; and

determining the amount of unmelted filaments includes identifying a portion of each sheet that includes unmelted filaments.

B4. There is also provided the medium of paragraph B1, wherein:

these determining steps are performed by a neural network that has been trained.

B5. There is also provided the medium of paragraph B2, wherein:

the neural network comprises a convolutional neural network.

B6. There is also provided the medium of paragraph B1, further comprising:

comparing the amount of fusing of the tissue against the image to a design tolerance; and

a notification is sent in response to determining that the amount of fusing of the tissue for the image is not within the design tolerance.

B7. There is also provided the medium of paragraph B1, wherein:

the machine learning model distinguishes between melted and unmelted filaments based on a difference in at least one of brightness or color.

B8. There is also provided the medium of paragraph B1, wherein:

the strands of fibers are selected from the group consisting of carbon fibers, glass fibers, metal fibers and ceramic fibers.

B9. There is also provided the medium of paragraph B8, wherein:

the strands of fibres are strands of carbon fibres.

B10. A portion of an aircraft assembled according to a method defined by instructions stored on a computer readable medium according to paragraph B1.

According to another aspect of the apparatus of the present invention, there is provided:

C1. an apparatus for characterizing a fibrous layer having a thermoplastic veil, the apparatus comprising:

an interface 716 that receives an image 742 of a fiber material 750, the fiber material 750 including strands 752 of fiber and further including a veil 754 of filaments 760 of thermoplastic; and

a controller 712 that subdivides the image into slices 744, determines the amount of melted filament depicted within each slice, and determines the amount of unmelted filament depicted within each slice.

C2. There is also provided apparatus according to paragraph C1, wherein:

the controller quantifies an amount of melting for the tissue of each sheet based on the number of melted filaments and the number of unmelted filaments for the sheet, and quantifies an amount of melting for the image based on the amount of melting for the tissue of each sheet.

C3. There is also provided apparatus according to paragraph C1, wherein:

the controller operates a neural network 724, the neural network 724 determining an amount of melted filament by identifying a portion of each sheet that includes melted filament; and is

The controller determines the amount of unmelted filament by identifying the portion of each sheet that includes unmelted filament.

C4. There is also provided apparatus according to paragraph C1, wherein:

the controller determines the amount of melted filaments and the amount of unmelted filaments by a neural network that has been trained.

C5. There is also provided apparatus according to paragraph C4, wherein:

the neural network comprises a convolutional neural network.

C6. There is also provided the apparatus of paragraph C1, further comprising:

comparing the amount of fusing of the tissue against the image to a design tolerance; and

a notification is sent in response to determining that the amount of fusing of the tissue for the image is not within the design tolerance.

C7. There is also provided apparatus according to paragraph C1, wherein:

the strands of fibers are selected from the group consisting of carbon fibers, glass fibers, metal fibers and ceramic fibers.

C8. There is also provided apparatus according to paragraph C7, wherein:

the strands of fibres are strands of carbon fibres.

C9. A part of an aircraft is manufactured using the apparatus according to paragraph C1.

Other illustrative embodiments (e.g., methods and computer-readable media related to the above-described embodiments) may be described below. The features, functions, and advantages that have been discussed can be achieved independently in various embodiments or may be combined in yet other embodiments further details of which can be seen with reference to the following description and drawings.

Drawings

Some embodiments of the present disclosure will now be described, by way of example only, with reference to the accompanying drawings. Throughout the drawings, the same reference numerals refer to the same elements or to the same type of elements.

FIG. 1 is a block diagram of a fiber evaluation system in an illustrative embodiment.

FIG. 2 is a flow chart showing a method for evaluating a fibrous material in an illustrative embodiment.

Fig. 3 is a diagram showing a picture of a fiber material in an illustrative embodiment.

Fig. 4-6 are diagrams showing slices (slices) from a picture in an illustrative embodiment.

FIG. 7 is a block diagram of a fibrous material evaluation system in an illustrative embodiment.

FIG. 8 is a flow chart of an aircraft production and service method in an illustrative embodiment.

FIG. 9 is a block diagram of an aircraft in an illustrative embodiment.

Detailed Description

The figures and the following description illustrate specific illustrative embodiments of the disclosure. It will thus be appreciated that those skilled in the art will be able to devise various arrangements that, although not explicitly described or shown herein, embody the principles of the disclosure and are included within its scope. Moreover, any examples described herein are intended to aid in understanding the principles of the disclosure and are to be understood as not being limited to the specifically described examples and conditions. Accordingly, the disclosure is not limited to the specific embodiments or examples described below, but by the claims and their equivalents.

Composite parts, such as Carbon Fiber Reinforced Polymer (CFRP) parts, are initially stacked in layers that together form a laminate. The individual fibers within each layer of the laminate are aligned parallel to each other, but different layers may exhibit different fiber orientations in order to increase the strength of the resulting composite along different dimensions. The laminate may include an adhesive resin that cures to harden the laminate into a composite part (e.g., for an aircraft). Carbon fibers that have been impregnated with an uncured thermosetting resin or thermoplastic resin are referred to as "prepregs". Other types of carbon fibers include "dry fibers" that are not impregnated with a thermosetting resin but may include a tackifier or binder. The dry fibers may be infused with a resin prior to curing. For thermosetting resins, hardening is a unidirectional process known as curing, whereas for thermoplastic resins, the resin may reach a viscous form if reheated. The systems and methods discussed herein describe the evaluation of dry fiber materials that include a binder in the form of a thermoplastic veil.

FIG. 1 is a block diagram of a fiber evaluation system 100 in an illustrative embodiment. The fiber evaluation system 100 includes any system, device, or component operable to automatically evaluate an image of a fiber material (e.g., a unidirectional CFRP, a glass fiber material, a material with metal fibers or even ceramic fibers, etc.) in order to determine a ratio of melted thermoplastic filaments (filamentts) to unmelted thermoplastic filaments in the material. In this embodiment, the fiber evaluation system 100 includes a characterization unit 110 and an imaging system 130.

The imaging system 130 acquires an image of a sheet 140 of fibrous material 142 (e.g., unidirectional CFRP layers, fiberglass material, material with metal fibers or even ceramic fibers, etc.). These images may be acquired at various locations along the sheeting 140 and/or at various orientations. These images depict not only strands 150 of fibers 152 (e.g., carbon fibers, glass fibers, metal fibers, ceramic fibers, etc.) within the sheet 140, but also thin yarns (veils) 160 of thermoplastic filaments 162 that serve as binders or tackifiers for the strands 150. Each thermoplastic filament 162 may be, for example, seven thousandths of an inch thick, or even thinner. Each strand 150 may be smaller such that 1.2 to 4 thousand strands are found within a single linear inch. The imaging system 130 may include a camera (e.g., color camera, stereo camera, etc.) or other non-destructive imaging components, such as an ultrasonic imaging device or a laser imaging device.

Images acquired by the imaging system 130 are received at an interface (I/F) 116. These images may be stored by the controller 112 in a memory 114 (e.g., hard disk, flash memory, etc.) for later analysis. The controller 112 manages the operation of the characterization unit 110 to facilitate image reception, analysis, and reporting. For example, the controller 112 may access the neural network 124 at the memory 114 when evaluating the image. The neural network 124 may comprise, for example, a convolutional neural network that has been trained based on the training data 122 to detect unmelted filaments and melted filaments within the image.

The training data 122 may include a set of images or patches (e.g., thousands of such images) and accompanying labels indicating features found in these elements. For example, the training data 122 may include images with regions that have been marked as melted or unmelted. The training data 122 may also include a picture of a dry carbon material taken in which the in-plane fiber angle is varied (e.g., zero degrees, +45 degrees, -45 degrees, 90 degrees, etc.). This may be relevant in order to train the neural network 124 to account for differences in brightness or contrast found at these different fiber angles. Thus, the training data 122 may be used to test and refine the process by which the neural network 124 detects both melted and unmelted filaments. The controller 112 may be implemented, for example, as custom circuitry, a hardware processor executing programmed instructions, or some combination thereof.

Illustrative details of the operation of the fiber evaluation system 100 will be discussed with reference to FIG. 2. For this embodiment, assume that the skilled artisan wishes to characterize the sheet 140 of fibrous material 142 in order to determine whether the ratio of fused filaments to unmelted filaments within the sheet 140 is within a desired tolerance.

FIG. 2 is a flow diagram showing a method 200 for evaluating fibrous material in an illustrative embodiment. The steps of method 200 are described with reference to fiber evaluation system 100 of FIG. 1, but those skilled in the art will appreciate that method 200 may be performed in other systems. The steps of the flow diagrams described herein are not fully inclusive and may include other steps not shown. The steps described herein may also be performed in an alternate order.

In step 202, the imaging system 130 acquires an image of the sheet 140 of fibrous material 142. The fibrous material 142 includes strands 150 of fibers 152 and also includes a veil 160 of thermoplastic filaments 162 of thermoplastic. The image may be generated in any suitable format, and a digital version of the image may be retrieved by the I/F116 for storage in the memory 114. In one embodiment, an image is taken every few hundred meters of the length of the fibrous material 142, and the image represents a small portion (e.g., a2 inch by 2 inch portion) of the fibrous material 142. However, the image may depict any suitable region having any suitable desired size.

After acquiring the image, the controller 112 continues to subdivide the image into slices (step 204). As used herein, a "slice" may include any suitable portion of an image. For example, a tile may include a portion that occupies the entire width of the image but only a small portion of the image height, may include a portion that occupies the entire height of the image but only a small portion of the image width, may include a rectangular section, and so forth. Ideally, the size of the patch (e.g., narrow scale) would be small enough so as not to delineate multiple filaments, but large enough so that a convolutional neural network trained for region detection can operate effectively when attempting to classify portions of the patch as representing either melted filaments or unmelted filaments. For example, the tile size may be between sixty and one hundred sixty pixels (e.g., one hundred pixels). It may be desirable for each sheet to depict a number of different filaments, for example fifty to one hundred filaments.

In further embodiments, the tile size may be selected such that the image height may be evenly divided by the tile size, or the image may be pre-processed (e.g., cropped, scaled, filtered, etc.) to enhance image quality and/or slicing (slicing).

In step 206, the controller 112 determines the amount of melted filament depicted for each sheet. As used herein, "fused filaments" (e.g., as depicted by fused portion 324 in fig. 3) include filaments that have been fused into strand 150. This may be performed by the controller 112 operating the neural network 124 to detect a characteristic in the sheet indicative of the presence of a melt filament. For example, a melt filament may be associated with a particular curvature, brightness, color, etc., and the neural network 124 may be trained by the training data 122 to identify these features. In one embodiment, neural network 124 performs a signature analysis on a patch to detect the presence of one or more signatures associated with a melt filament. The size of each region considered by the neural network 124 may be equal in scale to the size of the slice being considered. If enough features are detected with sufficient confidence, the neural network 124 may indicate the presence of a melting filament within the region of the sheet being analyzed.

Further, for each sheet, the controller 112 determines the amount of unmelted filament described (i.e., step 208). As used herein, "unmelted filaments" (e.g., as depicted by unmelted portions 326 in fig. 3) include filaments that remain on strand 150 without being melted into strand 150. This may be performed by the controller 112 operating the neural network 124 to detect a characteristic in the sheet indicative of the presence of an unmelted filament. For example, unmelted filaments may be associated with particular curvatures, intensities, colors, etc., and neural network 124 may be trained by training data 122 to identify these features. When performing the determination on the unmelted filament, the neural network 124 may perform a feature recognition task similar to the feature recognition task described above in step 206.

Certain portions of the filament may appear to partially melt, in which case the controller 112 may classify the filament as a molten state or an unmelted state depending on how the neural network used by the controller 112 is trained. In further embodiments, certain regions may depict both melted and unmelted filaments, and the controller 112 may classify such regions as melted or unmelted based on how the neural network used by the controller 112 is trained.

In steps 206 and 208, not all regions within the sheet necessarily include filaments (e.g., melted filaments or unmelted filaments). These regions may be reported as empty regions. The size of the void region may be relevant when determining whether the sheet 140 is within a tolerance that is independent of the melting rate.

The amount determined in steps 206 and 208 may indicate the size of the area (e.g., linear area or planar area) in which the filaments of a given type are determined to be present, may indicate the number of filaments of a given type, or may include other suitable metrics (e.g., the number of pixels depicting the filaments of a given type). These amounts are used to determine the ratio of melted to unmelted filaments.

In step 210, the controller 112 quantifies the amount of melting of the thermoplastic filaments 162 for each sheet of the tissue 160. This may be accomplished, for example, by summing the amounts of melted and unmelted filaments for a sheet, and then dividing the amount of melted filaments by the sum. In further embodiments, this may include determining a ratio of melted filaments to unmelted filaments. The percentage of melting desired by the design specifications may vary depending on the application, but one example of a percentage of melting is between 30% and 50%.

Step 212 includes the controller 112 quantifying the amount of fusing of the tissue 160 for the image based on the amount of fusing of the tissue 160 for each sheet. For example, the controller 112 may average the amount of melting determined for each sheet in step 210 to determine the amount of melting found in the entire image.

In further embodiments, the method 200 may be repeated for the same image by slicing the image in a different manner (e.g., slicing into "wide" slices followed by "high" slices, slices of different sizes, mirrored or rotated slices, slices with adjusted color or brightness or contrast, etc.). In yet another embodiment, the method 200 may be repeated for a plurality of images in order to quantify the entirety of the sheet 140. For example, in embodiments where an entire roll comprising hundreds of feet of dry carbon fiber material 142 is characterized, it may be desirable to acquire and analyze a large number of images of the material.

Method 200 provides an advantage over prior art techniques and systems in that it enables manual inspection techniques to be replaced by a cheaper and more accurate automated process. This means that the technician has more time devoted to other aspects of the manufacturing, which enhances the process of manufacturing composite parts from dry carbon fiber materials.

The techniques of method 200 may be used to inspect the surfaces of a large number of layers to confirm a desired melt rate at each of the layers before stacking occurs. Where the melt rate for each layer is known, any final preform produced from these layers will also have a known melt rate. Even for a layer placed on the inside of the preform after inspection.

Examples of the invention

In the following examples, additional processes, systems, and methods are described in the context of characterizing dry carbon fiber materials. That is, by discussing the systems and methods provided above for analyzing dry carbon fiber material, the following figures provide examples indicating how images may be sliced and characterized.

Fig. 3 is a diagram showing a fiber material in an illustrative embodiment. In particular, fig. 3 is a top-down view of a single layer unidirectional dry fiber material in the form of a CFRP. Fig. 3 depicts one or more strands 310 of carbon fibers held together by a veil 320 of filaments 322. Fig. 3 depicts slight non-uniformity in the distance between the strands 310 of carbon fiber, as the strands 310 in the bundle 300 are expected to be generally uniformly distributed, but not completely uniformly distributed. Some filaments have melted portions 324 while other filaments have unmelted portions 326. The single filament may have a melted region and an unmelted portion.

In this example, picture 300 would be subdivided into slices at boundary 330, resulting in slice 342, slice 344, and slice 346. Fig. 4 depicts the sheet 342 in detail. As shown in fig. 4, at indicator bar 400, portions of sheet 342 are labeled with different identifiers based on analysis performed by the neural network. "E" indicates the area without any kind of filament, "UM" indicates the area occupied by the unmelted filament, and "M" indicates the area occupied by the melted filament. The size of the portion M (in this case the linear distance) may be compared to the size of the area UM to determine the amount of molten filament for the sheet 342.

Fig. 5 depicts a sheet 344 and, by means of an indicator bar 500, a region M, a region UM and an empty region. In slice 344, the size of region M is substantially smaller than the size of region UM. Fig. 6 depicts a patch 346 having the highest ratio of area M to area UM indicated by indicator bar 600.

FIG. 7 is a block diagram of a fibrous material evaluation system 700 in an illustrative embodiment. According to fig. 7, the system comprises a carbon fiber material 750 made of a strand 752 of carbon fiber and a veil 754 of filaments 760. Filament 760 includes a melted portion 764 and an unmelted portion 766. The imaging system 730 generates an image of the carbon fiber material 750, which is acquired (in digital form) through an interface (I/F) 716. The controller 712 may direct these images to the memory 714 for storage. The memory 714 stores a neural network 724, as well as training data 722 and objective functions 726 for scoring the output of the neural network 724 during training. Memory 714 also stores images 742, slices 744, and slice data 746. The patch data 746 may indicate, for example, the location and amount of melted and unmelted portions of the filament detected in each patch. Although the system described above focuses on dry fibre materials, in further embodiments it may be used to perform a similar function on prepreg materials.

Referring more particularly to the drawings, embodiments of the disclosure may be described in the context of aircraft manufacturing and service in method 800 as shown in FIG. 8 and aircraft 802 as shown in FIG. 9. During pre-production, method 800 may include specification and design 804 of aircraft 802 and material procurement 806. During production, component and subassembly manufacturing 808 and system integration 810 of aircraft 802 occurs. Thereafter, the aircraft 802 may pass certification and delivery 812 for commissioning 814. When placed in service by a customer, aircraft 802 is scheduled for routine work (which may also include modification, reconfiguration, refurbishment, and so on) in maintenance and service 816 on a regular basis. The apparatus and methods embodied herein may be used during any one or more suitable stages of production and service described in method 800 (e.g., specification and design 804, material procurement 806, component and subassembly manufacturing 808, system integration 810, certification and delivery 812, service 814, maintenance and repair 816) and/or any suitable component of aircraft 802 (e.g., fuselage 818, system 820, interior 822, propulsion system 824, electrical system 826, hydraulic system 828, environmental system 830).

Each of the processes of method 800 may be performed or carried out by a system integrator, a third party, and/or an operator (e.g., a customer). For purposes of this description, a system integrator may include, but is not limited to, any number of aircraft manufacturers and major system subcontractors; the third party may include, but is not limited to, any number of suppliers, subcontractors, and suppliers; the operator may be an airline, leasing company, military entity, service organization, and so on.

As shown in fig. 9, an aircraft 802 produced by method 800 may include a fuselage 818 with a plurality of systems 820 and an interior 822. Examples of system 820 include one or more of propulsion system 824, electrical system 826, hydraulic system 828, and environmental system 830. Any number of other systems may be included. Although an aerospace example is shown, the principles of the invention may be applied to other industries, such as the automotive industry.

As noted above, the apparatus and methods embodied herein may be used during any one or more of the stages of production and service described in method 800. For example, components or subassemblies corresponding with component and subassembly manufacturing 808 may be manufactured or processed in a manner similar to components or subassemblies produced while aircraft 802 is in service. Further, one or more apparatus embodiments, method embodiments, or a combination thereof may be used during subassembly manufacturing 808 and system integration 810, for example, by greatly accelerating assembly of aircraft 802 or reducing the cost of aircraft 802. Similarly, one or more apparatus embodiments, method embodiments, or a combination thereof may be used while aircraft 802 is in service, for example, during maintenance and service 816, but not limited to during maintenance and service 816. For example, the techniques and systems described herein may be used for material procurement 806, component and subassembly manufacturing 808, system integration 810, service 814, and/or maintenance and repair 816, and/or may be used for airframe 818 and/or interior 822. These techniques and systems may even be used for system 820, including, for example, propulsion system 824, electrical system 826, hydraulic system 828, and/or environmental system 830.

In one embodiment, the part comprises a portion of the fuselage 818 and is manufactured during component and subassembly manufacturing 808. The part may then be assembled into an aircraft in system integration 810 and then used in service 814 until wear renders the part unusable. The part may then be discarded and replaced with a newly manufactured part in maintenance and service 816. The components and methods of the present invention may be used in component and subassembly fabrication 808 to fabricate new parts.

Any of the various control elements (e.g., electrical or electronic components) shown in the figures or described herein may be implemented as hardware, a processor implementing software, a processor implementing firmware, or some combination thereof. For example, the elements may be implemented as dedicated hardware. A dedicated hardware element may be referred to as a "processor," "controller," or some similar terminology. When provided by a processor, the functions may be provided by a single dedicated processor, by a single shared processor, or by a plurality of individual processors, some of which may be shared. Moreover, explicit use of the term "processor" or "controller" should not be construed to refer exclusively to hardware capable of executing software, and may implicitly include, without limitation, Digital Signal Processor (DSP) hardware, network processor, Application Specific Integrated Circuit (ASIC) or other circuitry, Field Programmable Gate Array (FPGA), Read Only Memory (ROM) for storing software, Random Access Memory (RAM), non volatile storage, logic, or some other physical hardware component or module.

Further, the control element may be implemented as instructions executable by a processor or a computer to perform the functions of the element. Some examples of instructions are software, program code, and firmware. The instructions, when executed by the processor, are operable to instruct the processor to perform the functions of the element. The instructions may be stored on a storage device readable by the processor. Some examples of storage devices are digital or solid state memory, magnetic storage media such as magnetic disks and magnetic tape, hard disk drives, or optically readable digital data storage media.

Although specific embodiments have been described herein, the scope of the disclosure is not limited to these specific embodiments. The scope of the disclosure is defined by the following claims and any equivalents thereof.