Hyperelastic curve fitting is a tool for estimating your material constants by inputting your experimental data and comparing it to the hyperelastic material models. Your stress-strain curves can be converted to any of the supported hyperelastic models, including Mooney-Rivlin, Ogden, Neo-Hookean, Polynomial, Arruda-Boyce, Gent, and Yeoh. Compressible hyperelastic Ogden hyper-foam and Blatz-Ko models are also supported.
To perform curve fitting, you input experimental data, select a model from one of the provided hyperelastic models, perform a regression analysis, graphically view the curve-fitting results, compare the fits to the experimental data, and write the fitted coefficients to the database as nonlinear data table commands for the subsequent finite element analyses.
Hyperelastic models can define three types of behavior: purely incompressible, nearly incompressible, and compressible. Hyperelastic curve fitting is based on the HYPER option of the TB command.
The steps for hyperelastic curve fitting are defined as follows:
| 1 | Step 1. Prepare Experimental Data | The experimental data must be a plain text file delimited by a space or a comma. |
| 2 | Step 2. Input the Experimental Data | Read in your experimental data by specifying the filename and path on the command line. |
| 3 | Step 3. Select a Material Model Option | The material options for the applicable curve-fitting regimen are defined in the TB command. Nine hyperelastic models are supported. After you pick a model, you can still change to another model if an ideal fit is not realized. |
| 4 | Step 4. Initialize the Coefficients | Depending on the model, hyperelastic curve fitting can be a linear or nonlinear regression process. The hyperelastic material models, along with the associated process for each are listed in Table 7.3: Hyperelastic Curve-Fitting Model Types. |
| 5 | Step 5. Specify Control Parameters and Solve | You will specify the type of error norm to be used to generate the curve fit. |
| 6 | Step 6. Plot Your Experimental Data and Analyze | You review and verify the results by comparing the experimental data and the regression errors. If the results you obtain are not acceptable, repeat steps 3 to 5 to perform a new curve-fitting solution. |
| 7 | Step 7. Write Data to the TB Command | Write your curve-fitting results to the database in the TB command table format. |
Curve fitting requires experimental test data. Your hyperelastic curve-fitting data needs to be a comma or space delimited file, referencing your stress vs. strain values.
Hyperelastic curve fitting supports three main behaviors:
Case 1 - Totally Incompressible Models (see Table 7.1: Experimental Details for Case 1 and 2 Models and Blatz-Ko)
Case 2 - Nearly Incompressible Models (see Table 7.1: Experimental Details for Case 1 and 2 Models and Blatz-Ko)
Case 3 - Compressible Models (see Table 7.2: Experimental Details for Case 3 Models)
The types of data required for each of these cases is defined in the tables below.
Table 7.1: Experimental Details for Case 1 and 2 Models and Blatz-Ko
| Experimental Type | Column 1 | Column 2 | Column 3 |
|---|---|---|---|
| Uniaxial Test | Engineering Strain | Engineering Stress | |
| Biaxial Test | Engineering Strain | Engineering Stress | |
| Planar/Shear Test | Engineering Strain (in loading direction) | Engineering Stress | |
| Simple Shear Test | Engineering Shear Strain | Engineering Shear Stress | (Optional) Engineering Normal Stress (normal to the edge of shear stress) |
| Volumetric Test | Volume Ratio (J) | Hydrostatic Pressure |
Table 7.2: Experimental Details for Case 3 Models
| Experiment Type | Column 1 | Column 2 | Column 3 |
|---|---|---|---|
| Uniaxial Test | Engineering Strain | Lateral Direction Engineering Strain | Engineering Stress |
| Biaxial Test | Engineering Strain | Engineering Strain (in thickness direction) | Engineering Stress |
| Shear Test | Engineering Strain (in loading direction) | Engineering Strain (in thickness direction) | Engineering Stress |
| Simple Shear Test | Engineering Shear Strain | Engineering Shear Stress | (Optional) Engineering Normal Stress (normal to the edge of shear stress) |
| Volumetric Test | Volume Ratio (J) | Hydrostatic Pressure |
J is the ratio of current volume to the original volume.
All stresses output in POST1/POST26 are true stresses and logarithmic strains.
Use the EADD option on the TBFT command to input your data
files. First, designate whether they are uniaxial, biaxial, shear, simple shear, or
volumetric, and then designate the location in the
Option2, Option3, and
Option4 fields. All of the stress values will be
engineering stress, except for the volumetric option (true stress). The file should
be a simple, delimited set of stress and strain values similar to the
following:
0.9703 60.00 0.9412 118.2 0.9127 175.2 0.8847 231.1
For temperature-dependent curve fitting, specify your temperature values at the top of the experimental data using the
/temp,value
line. This header format specifies the attribute (temp) and its value (100). An example of a typical data input using these attributes follows:
/temp,100 0.9703 60.00 0.9412 118.2 0.9127 175.2 0.8847 231.1
Adding this header introduces a temperature attribute of 100 degrees. You can add additional data sets at other temperatures, in additional files. One file can have data at only one temperature.
For compressible materials, the curve-fitting tool's default behavior is to solve
only for stress as a function of strain and lateral strain. To force the
curve-fitting tool to also fit experimental lateral strain data to generate the
coefficients for the Ogden
compressible foam model, add the line /USEL,1
near the top of the experimental data file. This option is valid for uniaxial,
biaxial and shear test data.
TBFT,EADD,ID,Option1,Option2,Option3,Option4 ! experimental data input
where:
ID = Index corresponding to the material number |
Option1 = UNIA, BIAX, SHEA, SSHE or VOLU |
Option2 = name of file containing experimental
data |
Option3 = file name extension |
Option4 = file directory |
Use the FADD option on the TBFT command to define your constitutive model. Table 7.3: Hyperelastic Curve-Fitting Model Types below lists the models that are available for hyperelastic curve fitting. When volumetric data is supplied, a compressible or nearly incompressible model is implied, and only the designated options will be available. When no volumetric data is supplied, the model is understood to be incompressible, and those model options will be available. Supplying zero coefficients for the volumetric data field will also denote an incompressible model.
Table 7.3: Hyperelastic Curve-Fitting Model Types
| Model Name | Abbreviation | Order/Options | No. of Coefficients [1] | Linear/Nonlinear Fitting |
|---|---|---|---|---|
| Mooney-Rivlin | moon | 2, 3, 5, 9 | 2/3/5/9+1 | Linear |
| Polynomial | poly | 1 to N | see below [2] | Linear |
| Yeoh | yeoh | 1 to N | N+N | Linear |
| Neo-Hookean | neoh | - | 1+1 | Linear |
| Ogden | ogde | 1 to N | 2*N+N | Nonlinear |
| Arruda-Boyce | boyc | - | 2+1 | Nonlinear |
| Gent | gent | - | 2+1 | Nonlinear |
| Blatz-Ko | blat | - | 1 | Nonlinear |
| Ogden Hyper-foam | foam | 1 to N | 2*N+N | Nonlinear |
Blatz-Ko and Ogden hyper-foam are compressible models. For Ogden hyper-foam, the experimental data you supply will require additional fields.
For more information about the hyperelastic models available for curve fitting, see Hyperelasticity in the Material Reference.
TBFT,FADD,ID,HYPER,Option2,Option3 ! material model options input
where:
ID = Index corresponding to the material number |
Option2 = model name, as specified in Table 7.3: Hyperelastic Curve-Fitting Model Types (above) |
Option3 = order or number of coefficients, where
applicable. Table 7.3: Hyperelastic Curve-Fitting Model Types (above) specifies the number
and type of coefficient(s) necessary for each hyperelastic model
type. |
Depending on the model you choose, hyperelastic curve fitting can be a linear or a nonlinear regression process. In both cases, the initial coefficients you supply will determine how accurate and efficient your curve fit will be. The initial values of the coefficients generally come from experience, and also from studying the function that defines the model you are attempting to compare/fit your data to. For most hyperelastic models, 1 or -1 is a good starting point. However, coefficient values can vary greatly depending on the model chosen. The Gent model, for instance, provides good fit with initial coefficient values as high as 1000.
You can also fix (hold constant) your coefficients
(Option4 = FIX on the TBFT
command). You specify a value for a coefficient and keep it unchanged, while
allowing the other coefficients to be operated on. You can then release the
coefficient you fixed and operate on it after the others are optimized. By default,
all of the coefficients are free to vary. The capability to fix coefficients applies
only to nonlinear curve fits (as listed in Table 7.3: Hyperelastic Curve-Fitting Model Types).
For temperature-dependent experimental data, enable temperature-dependency and specify a reference temperature before solving for the coefficients. You can set the reference temperature only to values specified using the /temp,value header line in the experimental data.
You can also specify tref = all and initiate multiple solves to evaluate coefficients at all available discrete temperature values. In this case, for data at three temperatures (t1, t2, and t3), a single TBFT,SOLVE entry will initiate three different solve operations at those three discrete temperatures.
With temperature dependency on and the reference temperature set to a particular value, a TBFT,SOLVE command solves for coefficients only at that temperature. To solve for coefficients at other temperatures, you set the reference temperature to each desired discrete temperature value and solve again.
You can initialize the coefficients before or after turning temperature dependency on. If the coefficients are initialized before turning temperature dependency on, the specified coefficients will become the initial coefficients for all future solves for that particular model. These coefficients are, however, overridden when temperature dependency is turned on and another set of initial values are specified at discrete temperature values. The curve-fitting tool looks for the initial coefficients at a particular temperature. If no coefficients are specified at a discrete temperature, the initial coefficients set before temperature dependency was turned on are used.
TBFT,SET,ID,HYPER,Option2,Option3,Option4,Option5 ! initialize coefficients
where:
ID = Index corresponding to the material number |
Option2 = model name (see Table 7.3: Hyperelastic Curve-Fitting Model Types (above) for the available models) |
Option3 = order, if applicable |
Option4 = index of coefficient |
Option5 = value of the coefficient |
To set Option4 for a reference temperature, or for
temperature dependency:
TBFT,SET,ID,HYPER,Option2,Option3,Option4,Option5
where:
ID = Index corresponding to the material number |
Option2 = model name. See Table 7.3: Hyperelastic Curve-Fitting Model Types (above) for the available models. |
Option3 = order, if applicable |
Option4 = tdep or tref |
Option5 = If Option4 =
tdep, Option5 = 1 turns on temperature
dependency, and Option5 = 0 turns it off. If
Option4 = tref,
Option5 will be either “all,” or
the reference temperature. |
TBFT,FIX,ID,HYPER,Option2,Option3,Option4,Option5
where:
ID = Index corresponding to the material number |
Option2 = model name See Table 7.3: Hyperelastic Curve-Fitting Model Types (above) for the available models. |
Option3 = (Blank--not applicable) |
Option4 = index of coefficient |
Option5 = 1 for fixed and 0 to vary |
Depending on the model, your hyperelastic curve fitting will be either a linear or nonlinear regression process (see Table 7.3: Hyperelastic Curve-Fitting Model Types (above) for details). Your error norms can be either normalized or non-normalized. Normalized error norms (the default regression option) generally give better results than the non-normalized error norms, since normalized error gives equal weight to all of your data points.
The solution control parameters of a nonlinear regression include:
Number of iterations
Residual tolerance
Coefficient change tolerance
The solution stops when both the residual tolerance and the coefficient change tolerance of your error norm are met, or if the number of iterations criteria is met. When you use nonlinear regression, you must initialize your coefficients.
TBFT,SOLVE,ID,HYPER,Option2,Option3,Option4, ..., Option7 ! set control parameters and solve
where:
ID = Index corresponding to the material number |
Option2 = model name. See Table 7.3: Hyperelastic Curve-Fitting Model Types (above) for the models available. |
Option3 = order or number of your coefficients.
See Table 7.3: Hyperelastic Curve-Fitting Model Types (above) for possible values. |
Option4 = curve-fitting procedure: 0 =
non-normalized least squares, 1 = normalized least squares |
Option5 = maximum number of iterations |
Option6 = tolerance of residual changes |
Option7 = tolerance of coefficient
changes |
Other solution parameters are available. See the TBFT command for details.
After solving, the coefficient tables contain the fitted coefficients and also the residual errors. You can plot your data and visually interpret the results. The first column is always the X- axis, with each additional column plotted separately as a function of column one.
Try to reserve column one for the variable that you would like to see vary in the plot. For example, if your data contains time, temperature, stress, and creep strain, and you wish to see the creep strain vary as a function of time at different temperatures and stresses in the plot, you add your experimental data using multiple TBFT,EADD commands by splitting the file into multiple experimental files as prescribed earlier, one for each combination of temperature and stress.
Use the button to plot the data. Your plots will show columns 2 and above as separate curves, plotted as a function of column 1. The data in column 1 is always the X-axis. By default, all the experiments are plotted in the GUI window. To view specific data and its corresponding fitting result, you can click the right mouse button (RMB) on the specific data set, and pick a desired option to view the results. Other RMB plotting utilities can be found for different data fields in the curve-fitting GUI window.
Use RMB functions to Zoom, Fit, Save Plot to File, View/Hide Legend and View/Hide Grid. Two or more fitted functions can also be compared in the same plot. For example, you can view Mooney2 Uniaxial and Mooney9 Uniaxial plots directly on top of each other, provided you have already solved for both of these functions. RMB also allows you to set the number of points used to generate the plot, and also change the minimum X value and the maximum X value in a plot. You can also hide a particular curve within a graph, turn the legend and/or axis displays on and off, and switch the scales between log scale and regular scale. Use the middle-mouse button to eliminate a specific curve and clarify or refine the remaining curve.
The two factors you consider in determining results acceptability are visual fit and the error norm/residual values. When you plot the curve, the error norm/residual values are printed in the curve-fitting GUI window. Error norm values help you determine the quality of curve fitting and whether to accept the results, but are not always the best indicator of a valid curve fit. Plotting the curves and visually assessing the result is usually the best indication. If the results are unacceptable, you may want to go back to step 3 and solve again, either by picking a different model, increasing the order, or redefining your initial values of the coefficients or other control parameters. You can continue to use your original experimental data, repeating step 3 through step 7 until you get an acceptable solution.
When you are satisfied with the curve-fitting results, you can write the curve-fitting data to the database via TBFT,FSET. The command converts the coefficients to the appropriate form before writing to TB tables. The program stores the data as part of the material property set for use in subsequent analyses.
TBFT,FSET,ID,HYPER,Option2,Option3 ! write data to TB
where:
ID = Index corresponding to the material number |
Option2 = model name. See Table 7.3: Hyperelastic Curve-Fitting Model Types (above) for the models available. |
Option3 = order or number of your coefficients.
See Table 7.3: Hyperelastic Curve-Fitting Model Types (above) for possible values.
|