The new GRASP – solve bigger problems in less time!
During the last year we have taken every opportunity to present some of the exciting features to come in the next GRASP release. It has been very well received and we are now happy to tell our users that your patience is finally rewarded:
GRASP 10.4 is here
The new GRASP introduces a wide a range of improvements, primarily to allow easier, faster and more accurate simulations of electrically large structures, together with some long awaited features from our user community:
Major updates to the MoM add-on:
- The Multi-Level Fast Multipole Method (MLFMM) has been implemented and tailored for use with Higher-Order (HO) basis functions, allowing very high speed while using a very modest amount of memory. Compared to MLFMM based on RWG basis functions, the typical choice in the industry, our HO-MLFMM requires significantly less memory, is roughly a factor of 3 faster, and allows high accuracy to be obtained at a very low computational cost. The new MLFMM solver represents a dramatic improvement relative to previous generations of GRASP and allows full-wave solution of electrically large structures using a fraction of the memory needed in the past.
Currents resulting from a full MLFMM analysis of a telecommunications antenna, including scattering in the full-size satellite. The source is the dual-reflector Ku-band antenna on the West wall of the spacecraft. Credits to M. Sabbadini of ESA for providing the geometry of the spacecraft.
- To ease the task of discretizing the geometry, both MoM and MLFMM can now be applied to non-connected meshes. This greatly simplifies the meshing process for more complex structures by allowing users to mesh different parts of the structure independently. The individual parts can then be added without worrying about mesh connectivity.
Advanced application of non-connected meshing to an offset-reflector antenna support structure
- MoM, BoR-MoM and MLFMM have been extended to allow the treatment of a wider range of dielectric materials. Specifically, it is now possible to define relative permittivities and permabilities less than one (e.g. plasma), just as materials with magnetic losses are now also supported.
- MoM and MLFMM can now treat multi-mode waveguide ports, allowing the user to compute S-parameters of multiple horn antennas while taking into account the presence of reflectors, mounting structures, e.t.c.
MoM calculation of coupling from the transmit feed to the receive feed in a Ku-band reflector system. Feeds are modelled as piece-wise linear bodies of rotation and fed by waveguide ports excited by the TE11 mode.
Improved Physical Optics performance:
- An accelerated algorithm for computing the field from a current distribution, the major bottleneck of a Physical Optics simulation, has been implemented.
- For the most time consuming problems with dense current and output grids, the reduction in time can be as much as by a factor of 1000, keeping the error level better than -100 dB relative to peak.
The field from this 2.5m Ka-band antenna was calculated in the shown grid with the new fast PO algorithm, reducing the current integration time from 37 to less than 3 minutes.
- To facilitate PO/PTD simulations on more complicated structures, the PO, Multi-Face Scatterer object can now automatically detect the edges where there is a PTD contribution without user intervention. This makes it much easier to perform PTD on structures that are defined as a combination of several individual scatterer objects.
Improvements to the GUI, particularly for larger projects:
- A range of improvements to the GUI have been made. The 3D viewer in particular now features a simple show/hide function for individual objects, that is much easier to use, and GRASP 10.4 is much more responsive when the user has a large number of hidden objects.