Using Datasets to Define Transient Parameters

Time-varying parameters are defined using piecewise linear (pwl or pwl_periodic) or piecewise constant (pwc) functions. These functions reference datasets with Time as the X-coordinate and the appropriate parameter and units as the Y-value. You must define a Design Dataset; Project Datasets are not currently supported for this purpose.

Design datasets are supported for the following time-varying properties in Transient Thermal solutions:

You can predefine the design dataset before referencing it in a boundary, excitation, or solution setup dialog box. Alternatively, you can enter the function, a new dataset name, and the X-dimension of the dataset, such as pwl(ds1,Time) or pwc(ds1,Time), into a supported input field (text box) of a boundary, excitation, or solution setup dialog box. If you choose the latter method, the Add Dataset dialog box appears automatically when you press Enter or click OK to complete the assignment or setup. In either case, the process of defining the time-varying value is the same.

Note:

In additions to using datasets for time-varying parameters in transient thermal solutions, you can also use datasets to define certain temperature-dependent excitation parameters in Steady-State Thermal and Transient Thermal solutions. You can also use them to define temperature-dependent material properties in Structural solutions. See the following topic for details:

Datasets in IcepakFEA Designs

For additional, general information on the definition and usage of datasets, see Working with Datasets and its subtopics.

Example – Dataset for Time-Varying Heat Generation:

The preceding figure shows a pwl curve of power in milliwatts (mW) versus time is seconds (s). There is a heat generation of 85mW beginning at time=0, a quasi-step increase to 150mW at 3.01s, and a quasi-step increase to 375mW at 5.01s. The heat generation decreases linearly from 375mW to 85mW over a 1s period beginning at 7s and then remains constant for a short time. At 9.5s, the power decreases linearly to 0mW over a 0.7s period and remains zero until time = 12s.

Note:
  • For solution timesteps lying between points defined in the design dataset, the Y values are interpolated.
  • For solution timesteps that are at a time beyond the last time value in the dataset, the behavior depends on the function used:
    • pwl: The last dataset value is used for all remaining timesteps. Downstream Y values are not linearly extrapolated from the curve trend.
    • pwl_periodic: Periodic extrapolation occurs. That is, the pattern of Y values is repeated in the same sequence for solution times beyond the last time value in the dataset. Use this function when you wish to define a repeating pattern, such as a set of intermittent power levels.
    • For example, assume there is a set of values, Y1 through Y10 defined for X = 0 through 10 seconds, respectively. For Time = 10 through 20 seconds, the corresponding Y values will be a repetition of Y1 through Y10. The same is true for Time = 20 through 30 seconds and so forth.

      Important:

      For a dataset to be supported within a pwl_periodic function, and the extrapolation performed accordingly, the first and last Y values must be identical. The equality of these two values ensures that there are no discontinuities in the resulting data.

  • If the first value in the dataset is specified for a time later than the first solution timestep: The first Y value in the dataset is used for all timesteps preceding the first X (time) value in the dataset. Upstream extrapolation is not performed based on the curve trend.
  • Best Practice: To avoid confusion and unexpected thermal source or load magnitudes, define the datasets to fully cover the range of time used for the solution setup. This practice is not a requirement, but exercise caution when forcing the solver to use assumed constant values beyond the range of a dataset. The exception is when you are using the pwl_periodic function, and the expected periodic extrapolated data is clearly known.
  • See Transient Thermal Solutions for a list of boundary and excitation parameters that support time-varying values.

How to Create and Use a Design Dataset

This procedure shows the creation of a dataset for use within a piecewise linear (pwl or pwl_periodic) function, suitable for time-varying parameters such as temperatures, convection film coefficients, or power (as in heat flux or heat generation excitations).

Additionally, see Using a Piecewise Constant (pwc) Function, which you can also use to time-vary parameters.

  1. Access the Add Dataset dialog box using one of the following two methods ("a" or "b"):
    1. Before beginning to assign a boundary or excitation (predefining the dataset):
      1. From the menu bar, click IcepakFEA > Design Datasets.
      2. The Datasets dialog box appears.

      3. Click Add.
      4. The Add Dataset dialog box appears

      5. Continue to step 2.
    2. While assigning a boundary or excitation (dataset does not yet exist):
      1. Type a function and dataset reference into a supported parameter text box using the following syntax:
      2. pwl(<dataset_name>,Time)   or   pwl_periodic(<dataset_name>,Time)

        where pwl indicates the piecewise linear function, pwl_periodic indicates piecewise linear with periodic extrapolation, and Time indicates the X-dimension type.

        The following table show examples of acceptable dataset names and how they would be referenced within a boundary or excitation dialog box:

        Design Dataset Name Examples:

        Dataset Name

        How to Reference it from a Boundary or Excitation Dialog Box

        ds1

        pwl(ds1,Time)

        ConvectionFilmCoef

        pwl(ConvectionFilmCoef, Time)

        HeatGen2

        pwl_periodic(HeatGen2, Time)

      3. Press Enter or click OK.
      4. The Add Dataset dialog box appears with the Name parameter and column data types already specified along with the default units.

      5. Skip step 2 and continue at step 3.
  2. When predefining a dataset, set up the parameters as follows:
    1. Specify the desired dataset Name.
    2. Refer to the preceding table for typical name examples.

    3. Click the X-column header and choose Time from the available choices that appear.
    4. Optionally, click the X-column header again, to change the units of time, and select the desired units.
    5. Click the Y-column header and choose the desired data type from the available choices that appear.
    6. Suitable data types for transient thermal analyses are as follows:

      • Power – for Heat Generation or Heat Flux (with Total Power option selected)
      • Irradiance – for Heat Flux (with Surface Flux option selected)
      • Temperature – for Temperature boundaries or for time-varying Ambient Temperature in a Convection boundary
      • Thermal Convection - for the Film Coefficient of a Convection boundary
      Note:

      If you accidentally select the wrong data type for either the X or Y column, click the column header again and choose No unit. This step will clear the units list and restore the list of data types. Then, select the desired data type.

  3. Optionally, click the X-column header to change the units of time, and select the desired units.
  4. Optionally, click the Y-column header to change the units of the Y-value, and select the desired units.
  5. The available units depend on the specified data type.

  6. Fill in the individual X and Y values to define how the Y-quantity varies with time.
  7. Verify that the appearance of the piecewise linear plot is as expected.

  8. Click OK to close the Add Dataset dialog box.
    • If you chose to define a new dataset while assigning a boundary, excitation, or solution parameter (Step 1.b), you are done creating both the dataset and the assignment or solution parameter.
    • If you chose to predefine the dataset before starting to assign a boundary or excitation (step 1.a), you will be returned to the Datasets dialog box at this point. Next, perform the following additional steps:
      1. Click Done to close the Datasets dialog box.
      2. Optionally, for a boundary or excitation, select the entities to which you wish to assign a time-varying parameter.
      3. Use the desired command to open the boundary or excitation dialog box (such as, Assign Boundary > Convection or Assign Excitation > Heat Generation). Alternatively, add a solution setup or edit an existing solution setup's properties if you wish to vary the time step size or field-saving interval.
      4. Type the function, dataset name, and X-quantity in the text box for the parameter to be time-varied, as previously described (in step 1.b.i).
      5. Note:

        As you begin to type a function name or dataset name, you will be presented with a list of functions and existing dataset names matching the characters typed thus far. You can select from the list and do not have to type the remainder manually. (See the note and figures in the topic, In-Place Editing of Datasets for more information.)

      6. Click OK to complete the assignment.