The Field data models menu option generates new amplitude samples from selected field quantities. For each active design, a new set of amplitudes is computed based on the random field model that is associated with the selected field quantity.
The respective random field models of the selected quantities must exist in the database.
You must select the complete set of field quantities. For example, if you have a cross-correlated random field model, then you must select all field quantities that are part of the model.
oSP3D determines the random field models that are associated with the selected quantities. For each random field model, the generation of amplitude samples is carried out.
The new scalar quantities are named as follows:
amp[quantity.ident][object type]_shape[number]
Usage
Parametrization of signals: Represent an input signal by few scalar parameters
Import signals from a DOE.
Create a random field model for the signal.
Delete all samples, save the database, and reload the database.
Import new signal data.
Project new signal samples into subspace of random field amplitudes.
Recompute smoothened field data from amplitudes to compare the original signal with the approximative representation.
Identify meta models between signals as input and scalar responses
Import signals from a DOE.
Create a random field model for the signal.
Export amplitude samples to optiSLang.
Create a MOP between the amplitudes and the scalar response.
Delete all samples, save the database, and reload the database.
Import new signal data.
Project new signal samples into subspace of random field amplitudes.
Export the new amplitude samples to optiSLang.
Approximate response values for the new amplitude samples using optiSLang's MOP
Also apply this for 3D field data on the input side.