Load Identification

You can use Load Identification to compute the operational loads applied to translational or rotational nodal degrees of freedom in a FEM model from:

• Measured or computed vibrations.

• Frequency Response Functions (FRFs).

You use Load Identification to determine operational loads from measurements or simulation data, such as for automotive NVH, ride handling, and durability needs.

When you perform a Load Identification analysis, you can use an existing FEM and Simulation files, or you can create an empty FEM and Simulation files, to access the Vibrations and Acoustics tab on the Ribbon bar. You can then import data from .op2 or .lms files.

You can identify loads using one of two methods:

• Direct Stiffness MethodIn the Direct Stiffness Method, the software uses the relative displacement (or velocity or acceleration) from the input vibration data and the input FRFs to compute the forces on the nodes. You can then convert the forces to a load recipe, for further analysis, or for use in downstream analyses.The FRF data is represented as force per displacement, velocity, or acceleration (vibrations). You can input the structure stiffness manually or use imported measurement data.You import measurement data or FRF data from .op2 or .lms files.You can automatically map connector elements to user-defined labels or names.You can manually map connector elements to user-defined labels.You can select different subcases from the input file. You can select different stiffness units.You can convert measurement and FRF data to the absolute coordinate system.

• Inverse Stiffness MethodWith the Inverse Stiffness Method the software computes an estimate of the operating loads, based on operational measurements, such as accelerations, and measured or computed FRFs.The input measurement data is acceleration of the structure or system.The FRF data is displacement, velocity, or acceleration (vibrations) per force.You import FRF data from an.op2 or .lms fileYou can apply a Single Value Decomposition (SVD) relative tolerance percentage to reduce the noise induced by overdetermining data from the FRFs. This will produce a more accurate substructure characterization.Inversion of the FRF matrix makes small singular values significant and they can dominate the data.Forces are the product of the inverse of the FRF matrix and the operational acceleration vector. Since the FRF matrix is mostly rectangular, an SVD of the FRF matrix is performed and then forces are computed.The Relative Threshold and Number of Singular Values give you control over what data to retain.Since the SVD is computed once per frequency, singular values vary with frequency. The first singular value is the dominant one, followed by the second most dominant and so on. By plotting the SVDs, you can determine the relative importance and decide which ones to retain in the calculation. The plots show the contribution of each singular value. As shown, only the top eight singular values are important, and the bottom 10 less so.You choose the Number of Singular Values to retain, which will determine the overall accuracy of the computed forces. The Relative Percentage is the relative percentage at each frequency, for the condition number criteria to be met. A 1% relative percentage is equivalent to a condition number of 100 and 2% is equal to a condition number of 50.You can use the Model and Load Pre-Processor to compute and export:Singular Values as a function of frequency. Singular Values above the actual rank for a given frequency are exported as zeros. You can specify the number of singular values to be included in the solutionCondition Numbers of the matrices corresponding to each output frequency. The Condition Numbers are the ratio of the first diagonal value of the SVD matrix to the last non-zero SVD.A condition number greater than 150 indicates ill-conditioning and therefore inaccurate results.Matrix Rank of each SVD matrix. The Matrix Rank is the inverse of the Condition Number.You enter a Relative Tolerance , which is the inverse of the conditionnumber. Singular values less than the condition number are set to zero, which makes the rank frequency dependent.

The plots shows the contribution of each singular value. In this example, only the top 8 singular values are important, and the bottom 10 less so.

Top SVDs versus frequency

Bottom SVDs versus frequency

Condition numbers versus frequency. Inaccurate data below 100Hz, when using 10 SVDs.

Matrix Rank plotted as a Relative Percentage. A Rank of 6 is required to include low frequency results.

Load Identification workflow

1. On the Noise and Vibration tab on the ribbon bar, click Load Identification to create the solution.

2. Define and input vibration data.

3. Define the computation method as Direct Stiffness or Inverse Stiffness.

4. Input the FRF data.

5. Map the vibration and FRF data from active (excited) to passive (response) nodes or IDs.

6. Post-process the results.

Post-processing

The Load Identification post-processing scenario is available for you to display and interpret the results.

XY plot of Force in dB versus frequency (Hz)

How do I

Create a Load Identification solution

Learn more

Remote solving workflow (Linux)

Quick links

Command reference

Pre/Post video examples

Bulk Entry Descriptions

Simcenter 3D tutorials

Browse Simcenter 3D help by product area

Load Identification, Simcenter 3D 2021.1 Series

© 2020 Siemens

window.mainLanguage="en_US"

window.delivId=""

window.projectId=""

MathJax.Hub.Config({ TeX: { extensions: ["autoload-all.js"] }, tex2jax: { displayMath: [ ] }, "SVG": { scale: 125 } });