Data processing: structural design – modeling – simulation – and em – Simulating nonelectrical device or system
Reexamination Certificate
1999-11-15
2002-01-22
Teska, Kevin J. (Department: 2123)
Data processing: structural design, modeling, simulation, and em
Simulating nonelectrical device or system
C703S007000, C345S420000, C700S182000
Reexamination Certificate
active
06341261
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates generally to a composite materials design process. More specifically, the present invention relates to a knowledge driven composite design optimization process and corresponding system. A computer program adapted to facilitate implementation of the knowledge driven composite design optimization process is also disclosed.
The use of composites in airframe construction is becoming an increasingly complex process. Composite structures are generally designed at the local laminate level using wireframes or solids in Computer Aided Design (CAD) systems such as the UNIGRAPHICS (™) CAD system. Analysis is performed using a separate analytical model(s), e.g., a finite element based model, which initially assumes laminate families which define a percentage composition (number) of each ply orientation and which is generally coordinated with design geometry through imported master datums and a design surface. A typical design process is illustrated in the functional flowchart of FIG.
1
. From the flowchart, it will be appreciated that the current design process has not been optimized. For example, finite element analysis (FEA) is conducted before the step of editing the ply lay-up for manufacturability, which virtually guarantees that the FEA step will have to be repeated and, in many cases, repeated several times. Moreover, it will be appreciated that changes in, for example, the wireframe model will require updates to, or recreation of, the finite element model and vice versa.
Furthermore, many design considerations are routinely addressed during the composite design process. For example, often in the analytical model, the order in which each oriented ply is found within the total number of plies is not considered. Typically, the analyst does not restrict the thickness map to “buildable” thicknesses of the material selected or specify families which produce “fabrication friendly” designs. These modifications are typically integrated by the designer in their geometry model. However, since the abovementioned modifications occur after the bulk of the analysis has been completed, there are often surprises, e.g., higher than expected component weight, at a point well into the design process. In addition, even when the analyst does define local stacking sequences, the analyst accomplishes this by performing optimizations which result in local laminate definitions which do not integrate well with one another.
It will also be noted that documentation for manufacturing is currently provided in two key ways:
(1) as a design surface with cured ply boundaries projected to a plane; and
(2) through manually created section cuts and laminate tables that contain text entities for each ply identifying orientation, material and number of each ply.
Since the textual data must be created manually, text data having missing information is frequently released to other manufacturing departments. When design changes occur, these errors and omissions are compounded because the text data is often only partially updated after each design change.
The inner moldline, which is often a tooled surface and always a structural mating surface, is defined by conceptually joining the resultant section cuts of the engineering, i.e., CAD, definition. Manufacturing personnel then flatten the ply geometry to create uncured ply boundaries that are cut for fabrication. Translation errors due to selection of the incorrect normal orientation during this step are a common occurrence when working with complex geometry, primarily because standards do not address this level of detail. Moreover, while the thickness of the composite material used to create the laminate is 18-25% thicker than cured material, this fact is normally not reflected in any of the traditional models. It should be mentioned that the one exception to this general statement is found in the unique files which are created by manufacturing for laser projection that use a “debulked” ply thickness to develop a three dimensional (3-D) plies representation of the part in the lay-up step of fabrication. The variability of the translation process is high, since design intent is not always clear because all ply boundaries are represented on one 3-D plane rather than in true 3-D space. An example of the type of defect created with no geometrical definition at the ply level in 3-D space is described immediately below.
The lack of geometrical accountability for ply overlaps leads to locally undersized areas in tools. This in turn results in increased local pressure on the component during the cure at ply overlap or splice areas. This may contribute to internal laminate defects in the form of porosity or resin poor areas in a structure if enough of these details occur through the thickness in an area. In addition, lack of geometrical accountability can lead to local distortion of the fiber architecture and can result in increased interlaminar shear stress, each of which adversely affects structural performance. These cause and effect relationships are viewed as too complex for the current conventional composite process to track and control during design and fabrication. It will be appreciated from the references, discussed briefly below, that a great deal of attention is paid to these effects at the micromechanics level in literature and in the typical fabrication shop for specific structures, yet no standard process allows easy incorporation of these considerations into composite design practices. The current approach to understanding these local effects is to build and then cutup composite parts, i.e., to perform destructive testing of the articles. It will be appreciated that this is an expensive and time-consuming approach to understanding a geometrical problem.
In addition, manufacturing constraints such as material width restrictions are not incorporated or reflected in the design data. Such design constraints are often considered only as a refinement (iteration) within the manufacturing definition cycle, i.e., when editing ply definition to ensure manufacturability. It will be noted that this results in additional ply splices which may not be accounted for in the design. While this often leads to structural degradation, these manufacturing constraints may not be reviewed during the design steps in current practice. Thus, the analytical community is forced to adopt conservative analysis approaches to avoid the risk created by uncertainties arising from manufacturing constraints. The most common impacts of this approach are greater structural weight, more stringent fabrication requirements, and, of course, higher costs. The lack of understanding of the structural design impacts on fabrication also leads to inconsistent disposition of discrepant fabrication events, since a “preferred” or “best” practice has never been identified. The impacts are considered part of the variability that leads to the reduced material allowables used for composite analysis.
It should also be mentioned that design changes often require updates to manufacturing data. Manufacturing recreates the textual data at least twice to produce data forms that meet the needs of the manufacturing database. The composite database becomes the source for all in-process inspections and fabrication. Final parts are inspected to design data as prepared by manufacturing personnel in their templates and database. Database coordination for design changes is a challenging, not to mention a continuous, process.
Moreover, the only software tool that attempts to integrate the definition and analysis of composites is the Northrop Grumman's, formerly Vought Aircraft Company's, Computerized Composite Development Project (CCDP) program. The CCDP program is limited in that it does not create a 3-D product definition, is hard coded in FORTRAN to run on a VAX computer or computer cluster, and includes no adaptive knowledge or objects that can assist the designer in tracking sensitivities of design changes. Moreover, the CCDP program is not focused on visualizat
Hale Richard D.
Schmitz Jeffrey J.
Vasey-Glandon Virginia M.
Frejd Russell W.
McDonnell Douglas Corp.
Powell, Jr. Raymond H. S.
Teska Kevin J.
Westerlund Robert A.
LandOfFree
Knowledge driven composite design optimization process and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Knowledge driven composite design optimization process and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Knowledge driven composite design optimization process and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2866982