4.2. Unidirectional Pressure Mapping: CFD to Mechanical APDL

Use this load-transfer method to transfer pressure loads from a CFD analysis to Mechanical APDL (via SFE pressure commands). This method is well suited for transferring phase-shifted (complex number) harmonic values from a CFX Transient Blade Row analysis for use in a harmonic forced response analysis (ANTYPE,HARM).

The CFD pressure data (coordinates and pressure values) are contained in a text file that you read in. The data is mapped to the target nodes and written as SFE commands to a file for subsequent reading into a structural static, transient, or harmonic analysis. The mapping and interpolation is performed using the same point cloud algorithm as *MOPER,,,MAP:

  1. Resume the model (RESUME) containing the nodes and elements to which the pressures will be applied. The target nodes must also have been surface meshed with SURF154 elements (ESURF) in order to accept the pressure loads.

  2. Enter the mapping processor (/MAP).

  3. Identify the target nodes (TARGET). A nodal component named TARGETNODES and an element component named TARGETELEMS are automatically created.

  4. Specify the type of file that contains the source points and pressures (FTYPE). The following file types are supported:

    • CFXTBR - File from a CFX Transient Blade Row analysis. Request this output in the Export Results tab of CFX-Pre with [Export Surface Name]: Option set to Harmonic Forced Response.

    • CFDPOST - File from a CFD-Post export. Select BC Profile from the Profile Type under Export action.

    • FORMATTED - A generic fixed-formatted text file containing, at a minimum, x, y, z, pressure (and the imaginary pressure values if pressures are complex) in that order. You may have other columns of data between the required data values.

    • CSV - A comma-separated values file, with each field (column) of data separated by a comma. The fields need not be of the same width, and the data can be in any order. The file must contain at least x, y, z, pressure values (and the imaginary pressure values if pressures are complex).

  5. Read in the data (READ). Specify the number of rows to skip if header information is on the file. For FORMATTED files, also specify the data format using the FORTRAN FORMAT convention. For CSV files, also specify the fields containing the x, y, z, pressure data. Upon reading, nodes are created for the source points. For CFXTBR and CFDPOST files, if face data ([Faces]) is available, SURF154 elements are also created. A nodal component named SOURCENODES and an element component named SOURCEELEMS are automatically created.

  6. Verify the alignment of the source and target nodes (PLGEOM).

  7. If the source nodes need to be moved to align with the target geometry, you may use the subset of /PREP7 node commands listed below. Target nodes may not be modified.

    MOVE - Calculates and moves a node to an intersection.
    NDELE - Deletes nodes.
    NDIST - Calculates and lists the distance between two nodes.
    NGEN - Generates additional nodes from a pattern of nodes (enforces ITIME = 2 and INC = 0; that is, you can only move the nodes)
    NLIST - List nodes.
    NMODIF - Modifies an existing node.
    NSCALE - Generates a scaled set of nodes from a pattern of nodes.
    NSYM - Generates a reflected set of nodes.
    TRANSFER - Transfers a pattern of nodes to another coordinate system.

  8. Perform the mapping operation (MAP). The source pressures are mapped to the centroids of the target SURF154 elements (and not the target nodes). The mesh of the structural model should be of sufficient refinement to capture the CFD pressure field.

  9. Verify the mapped values (PLMAP).

  10. Once mapped, the pressures on the target nodes may be modified via the following commands:

    SF - Specifies surface loads on nodes.
    SFE - Specifies surface loads on elements.
    SFSCALE - Scales surface loads on elements.

  11. Write the pressure values to a file (WRITEMAP). The data is written in the form of SFE commands applied to the target SURF154 elements.

  12. Exit the /MAP processor (FINISH). The original database (BeforeMapping.db) is resumed. (See the Note below.)

  13. Apply the mapped pressure loads to the model (/INPUT) when the analysis requires them.

Note:  After the target nodes are defined (TARGET), or before the source points are read in (READ, if issued first), the current database is saved as BeforeMapping.db and the nodes and elements not on the target surface are deleted to make it easier for you to work with the mapping entities.

If using NGEN to rotate the source nodes (DY ≠ 0), and the active coordinate system is either the cyclic coordinate system (CYCLIC) or the coordinate system defined by the CFX rotation axis (CFXTBR file only), the source pressures are also rotated according to:

Pθ = Peiθ

where P are the complex pressures on the source nodes that are rotated and θ is the rotation angle (DY). Rotating the source nodes is useful when only one passage is modeled in the CFD analysis, and the suction (or pressure) side of the one blade side must be rotated in order to assemble a full blade; that is, suction and pressure side on one contiguous blade for the structural analysis.

During the mapping operation (MAP), any target element centroids (output points) found to be outside the region defined by the source nodes are not mapped, and a warning message is issued. This may be due to the target mesh being irregular (such as a tetrahedral mesh) or coarse in areas of high curvature; a higher LIMIT value may help to map these points. The output point numbers in the warning messages start at 1. To see which target elements a point corresponds to, add the first target element number (minus 1) to the output point number. This number is output in a note following mapping. Use the following command input to view the elements that are not mapped:

offset = firsttargetelement – 1
cmsel,u,targetelems
esel,a,elem,,offset+pt#1    ! pt number from the warning message
esel,a,elem,,offset+pt#2
.
.                           ! for many points, edit the .ERR file
.                           ! and create this input file
plgeom

Following are restrictions and limitations for the mapping procedure:

  • You must perform all operations in one session. (For example, you cannot perform some operations, save the database, leave the /MAP processor (or the program), resume that database, and continue the mapping operation).

  • PLGEOM or PLMAP displays 2D graphics (-d win32 or -d x11 on the command line or the launcher, and any output to a graphics file (/SHOW,<format> or /UI,COPY,SAVE,<format>)) with all source nodes.

    3D graphics show only those source nodes not hidden by the current view.

    Issue the NSEL family of commands to display only those nodes on the visible surface.