PMUTs are often set up in an array formation for imaging. OnScale is capable of simulating large 3D PMUT arrays, which helps designers get a full picture of the performance of the array.
The typical workflow for the design and development of MEMS devices is based on a mask layout design. It combines the 2D mask layout with the process flow to create 3D solid models of the MEMS devices. The layout-based discretized 3D models need to be converted into system models. Now, these high-fidelity MEMS 3D models can contain between 100,000 to 1,000,000 degrees of freedom to fully capture the effect of process flow on 3D geometry and its effect on device performance. FEA simulations of these models with legacy software can take a few days to run and sometimes it is not even possible to run these models. Given this a MEMS design engineer must rely on reduced-order or lumped model approximations, which do not truly capture the effect of etching, residual stresses and packaging effects. This makes the system design optimization of MEMS devices a big challenge to achieve the required performance.
At OnScale we have developed fast time-domain multiphysics 3D finite-element (FE) solvers and seamlessly integrated them with cloud high performance computer (HPC) capabilities that address these constraints. MEMS design engineers can now import full 3D high-fidelity models of MEMS devices created in Mentor Graphics / SoftMEMS and run FEA simulations in OnScale to accurately understand the performance of these devices in less than a few hours.
In this example we will show you how to run the model on the cloud and post-process the results using the OnScale Post Processor.
Model Description
Model of a 3D Piezoelectric Micromachined Ultrasonic Transducer (PMUT) array section with an Aluminum Nitride active layer operating into a water load.
Download and Open the Model
Download: PMUT Array 3D SoftMEMS Import Model
OnScale input files have the file extension *.jfp. To open the downloaded file:
- Open OnScale.
- Select Designer mode.
- Click the menu button in the top-left of the application and select Open Project.
Once the model is open you won't need to do any set-up as we have set up the model in advance. All you will need to do is submit the job to the cloud to start a simulation.
Preview Model
Preview Model is used to view both the geometry dimensions and the material assignment of a model to ensure that everything is correct. To preview a model, click Preview Model in the Home tab of the application ribbon. To close the preview, click the red x in the top right-hand corner of the widget.
Run Model on Cloud
To run the model on the cloud, click Run on Cloud in the Home tab of the application and follow these steps:
- Click Estimate.
- Set CPUs to 2.
- Click Run.
Download Results
- From the Home tab, click Storage.
- Click the refresh icon to refresh storage.
- Select the job name from the dropdown menu.
- Click Download and select Download All.
Choose a directory to download the results to. These results will be stored in the folder: pmut_3D_Load-YYYYMMDD-HHMMSS\1 where YYYYMMDD-HHMMSS is the current date and time.
Switch to Post Processor
Post Processor is a suite of tools used to visualize and analyze output data generated from OnScale simulations. To switch to Post Processor, click the button in the top-right:
What kind of analysis can I do in Post Processor? Find out here.
Analyze Results
To plot the membrane displacement of the PMUT array, follow these steps:
- In File Explorer, navigate to the download directory and double-click the file pmut_active.flxhst to load it.
- In Results Manager, double-click zdsp to plot the Z displacement of the membrane.
Have a play around with the model in Designer. Add a few more outputs and open them in Post Processor once you've run the simulation. You can also calculate impedance/admittance on the electrodes by selecting the voltage or charge records from the Results Manager and clicking the impedance and admittance buttons found in the XY Tools section of the Home tab. Below is an FFT of the Z displacement output shown above.