What's new in FEFLOW 7.5 (Update 1)?

In this FEFLOW release we have new exciting software features to support your current modelling applications and more:

  • Support of cloud computing directly from the FEFLOW and FePEST GUIs.

  • Update models faster using a new conceptual modelling approach.

  • Better management of selections and their workflows.

  • Improvements in FEFLOW’s Numerics.

  • Get closer to integrated groundwater resources management.

  • New input unit for BHE: Energy demand

  • Several improvements in IFM, Python and FEFLOW console.

  • Last but not least, FePEST comes reloaded with a new set of interesting features. You can look the online FePEST documentation for all the details.

A detailed description of the major software features is presented below:

The new Cloud Simulation panel

With the release of FEFLOW 7.5.1,  model run (s) can be executed in the cloud using DHI MIKE Cloud infrastructure. The deployments of the model run (s) are done following a one-click solution approach directly from FEFLOW and/or FePEST graphical interfaces. You can forget about preparation of remote desktops, administration cost/time of managing multiple machines, among other. The new Cloud Simulations panel is capable to run several jobs at the time with a flexible set of virtual machines, which could support you in all the different groundwater projects.

During the runtime, you can easily monitor the FEFLOW results through the new interactive Cloud Charts on real time. There is no need to wait until the run is completed! You can find more information how to get started with FEFLOW Cloud Simulations here.

simulation_problemsummary.png 

Example of model list with several model runs

Update models faster using a new conceptual modelling approach

From the FEFLOW 7.5 version, we introduce a new conceptual modelling approach. This major usability improvement will allow you to update and build numerical models in a faster manner.

You can define Material regions now in the FEFLOW Supermesh. You can create your own materials (e.g., sand, clay, etc.) and these can be associated to any FEFLOW property (constant and/or transient) even before the mesh is generated. The Database with the material information can be easily reused in your next groundwater projects.

After the meshing process is finalized, FEFLOW automatically creates the association between elements and FEFLOW parameters in the resulting mesh. You can go back and forward as many times as you need, without losing any elemental information. With this update, material regions can be defined directly during the import of map files by creating the 2D or 3D Supermesh.

For more details, please refer to the modelling workflows using Material Regions and Parameter Lookup Table (here).

Material regions in the 3D unstructured finite-element mesh.

Better management of selections and their workflows

FEFLOW selections are the core of any workflow, either for model parametrization or for recording any specific resulting quantity. Now FEFLOW provides you a convenient way for operating with selections. You can group (and ungroup) individual selections, for example if you would like to organize specific information into one location. You can sort your list of selections by simply drag-and-dropping. Operations of Budget-History Charting (for nodal selections) and Content-History Charting (for elemental selections) can be applied to the group level. Enhanced visualization feedback in FEFLOW gives you an overview of the selections being used for history charting.

For more details, please refer to Group Selections and Selection Management

Selections panel with groups and new icons for budget/content history recording.

Improvements in FEFLOW’s Numerics

FEFLOW 7.5 version introduces new features associated to the numerics and calculations. The new SAMG version 2020 is available and comes with enhanced memory management and improve parallelization for more efficient computation. Also, FEFLOW brings a new solver package PETSc. The Portable, Extensible Toolkit for Scientific Computation. PETSc (v. 3.17.0) comes with three game changers: Krylov-based methods solvers, AMG solver and GPU-accelerated preconditioning (for CUDA devices). Moreover, this FEFLOW version offers also additional flexibility for computing different problem classes (flow, mass and heat) in FEFLOW. The numerical parameters (Error norm and Error Tolerance) are separated for each problem class. The change gives a new way of tuning numerical problems to the user.

Now with the release of FEFLOW 7.5 Update 1, PETSc is further equipped with a GPU-accelerated AMG using Hypre v.2.24.0. FEFLOW 7.5.1 Linux version supports the GPU-accelerated AMG solver to speed-up even more your groundwater projects.

New solver package PETSc in the FEFLOW Problem Settings dialog.

Multiple error norm support for FEFLOW models with more than one Problem Classes (flow, mass transport and heat transport).

Get closer to integrated groundwater resources management

This release means also that FEFLOW strengthens its links to the MIKE family with two new features.

FEFLOW - MIKE SHE Integration

FEFLOW supports the direct import of DFS2 files, for example typically created during MIKE SHE runs. The information available behind the DFS2 file can be linked to any FEFLOW parameter. A common application would be linking MIKE SHE's estimated groundwater recharge into FEFLOW's In/outflow on top/bottom or source/sink parameters. The functionality offers a great opportunity for modelling integrated groundwater resources management and is not limited to only water quantity. Water quality information such as concentration or mass rates behind the DFS2 file format could be utilized as well.

Area-weighted projection is a map-data projection method developed to support recharge data import from MIKE SHE, DFS2 files.

New way of coupling groundwater and surface water

The former IfmMIKE11 is now officially replaced by FEFLOW piMIKE1D. This solutions offers a native coupling between FEFLOW engine and MIKE 1D engine for surface water modelling. FEFLOW piMIKE1D offers a coupling between an existing FEFLOW Fluid-Transfer BC and a river network file from MIKE 1D. If the user imports an old MIKE 11 file, the file will be migrated automatically to the new MIKE 1D format. For those users already working previously with MIKE11.

FEFLOW piMIKE 1D was initially introduced with the release of FEFLOW 7.5. Now with the MIKE 2022 Update 1 (FEFLOW 7.5.1 and MIKE Zero 2022.1), several extensions are offered:

  • Major flexibility to control the coupling settings through FEFLOW nodal selections, controls for coupling distance, definition of transfer rates, definition of exchange terms and more.

  • General overview of river geometrical information used for the coupling (branches, nodes, cross-sections, etc.) displayed in FEFLOW GUI.

  • Control of global grid spacing directly from the FEFLOW GUI.

  • Possibility to execute multiple model runs in parallel.

You can find more details how to get started with the coupling in the documentation (here).

New customer features "Hydrodynamics" in the FEFLOW Problem settings dialog.

Usability improvements for borehole heat exchangers (BHE)

FEFLOW 7.5 introduces a new unit type Energy in BHE Editor and BHE Interconnection dialogs. This provides a much intuitive manner for defining energy demand (cooling or heating) in a closed-loop system. Further details about this change in the workflow can be found in Energy demand.

There is a new implementation for assignment of BHEs through external map files, i.e., Parameter Association dialog. This usability improvement is added now to support the assignment of large numbers of BHEs as well as with different BHE datasets.  

BHE Dataset name for the Parameter Association.

IFM, Python and FEFLOW console

The Open Programming Interface (Interface Manager, IFM) comes with additional functions to support the new conceptual modelling approach. The APIs offer additional degree of flexibility for assigning and editing material regions.

FEFLOW 7.5 introduces a new switch for multi-threading control in the FEFLOW Python runs. In the FEFLOW console, you have now two switches to define the equation solver and to monitor exactly the CPU consumption (as well the number of iterations).

 

FEFLOW IFM Functions

Support the new conceptual modelling approach:

  • IfmLookupFindMaterialByName

  • IfmLookupFindMaterialById

  • IfmLookupGetNumberOfMaterials

  • IfmLookupGetMaterialName

  • IfmLookupGetMaterialId

  • IfmLookupGetPropertyCount

  • IfmLookupGetPropertyId

  • IfmLookupHasProperty

  • IfmLookupGetPropertyDoubleValue

  • IfmLookupGetPropertyIntValue

  • IfmLookupGetPropertyStringValue

  • IfmLookupAddMaterial

  • IfmLookupRemoveMaterial

  • IfmLookupMoveMaterial

  • IfmLookupSetMaterialName

  • IfmLookupSetMaterialId

  • IfmLookupSetPropertyDoubleValue

  • IfmLookupSetPropertyIntValue

  • IfmLookupSetPropertyStringValue

  • IfmLookupRemoveProperty

FePEST Improvements

FEFLOW 7.5 comes with a superior FePEST version for supporting a new framework for history-matching (automatic calibration) and uncertainty quantification with the PEST++ package.

The MIKE Cloud implementation in FePEST GUIs offers you a level of parallelization without levels. You just require few clicks to define the number of virtual machines (pools) and worker in each pool. FePEST takes care of all the rest: deployment and management of the cloud runs. You do not need to wait until the end of the PEST/PEST++ optimization. All results are monitored in real-time for your convenience.

You can get the overview of all these new features and more in the separate FePEST Online Help.

 

Table of Contents

Index

Glossary

-Search-

Back