ACIS R23 is out, and here’s why you should care

The new R23 version of that ACIS geometric modeling kernel, from the Spatial division of Dassault Systemes, is now available. This will probably be of the most interest to you if you happen to be a CAD developer, and are using ACIS. But it’ll also (eventually) be of interest to you if you’re a user of CAD software that incorporates ACIS. It may take 6 months or a year for the next version of your software to come out, incorporating ACIS R23, but, while you’re waiting, here are some of the improvements (from Spatial’s Developer 2 Developer blog):

  • Support for multi-threaded faceting. To simplify implementation for the user, thread management is handled completely within the modeler. Restoring of SAB (Standard ACIS Binary) models, a function used by most ACIS applications, brings dramatic speed improvements of up to 40%.
  • Extended Boolean and Stitching capabilities. 3D ACIS Modeler includes enhancements to several of the Fuzzy Boolean operations. These types of operations are especially useful in applications which use an approximation of a piece of geometry in a Boolean operation. ACIS is now better able to handle near-coincidence between surfaces that previously would have resulted in sliver faces, performance problems, or even failure.
Booleansm
  • Cellular topology allows developers to attach attributes, such as material properties, to volumetric cells. New improvements now allow these attributes to migrate through Boolean and Glue operations.
  • Seeded feature detection is also available in 3D ACIS Modeler R23. This new capability offers significant benefits to direct-modeling applications and other workflows that operate on logical groups of faces. Feature detection functionality helps in the identification of protrusion and depression features as well as blends. Developers have the flexibility to specify whether or not the blends at the boundary of a depression or protrusion feature are to be included as part of the feature.

 

Here are some images showing examples of feature detection:

260px ACIS Feature Detection001

The Body and Seed Face

260px ACIS Feature Detection002

The Detected Depression

260px ACIS Feature Detection003

The Body and Seed Face

260px ACIS Feature Detection004

The Detected Protrusion

260px ACIS Feature Detection005

The Body and Seed Face

260px ACIS Feature Detection006

Detected Blend Sequence

260px ACIS Feature Detection007

Detected Blend Network

A little back story

Back when Keith Mountain was CEO of Spatial (the company was, already at the time, a division of Dassault Systems), he took a chance, and agreed to help a start-up CAD company named SpaceClaim solve a big problem. They had a vision for a new kind of CAD program, that would allow users to push and pull on their models to edit them. It’s something we now call “direct editing.” This capability required significant updates and improvements to ACIS. It probably didn’t hurt that the CEO of SpaceClaim was Mike Payne, Keith Mountain’s predecessor at Spatial, but, still, it was a big investment for Spatial to make. I understand that it took on the order of 4 years to accomplish the bulk of the work.

The seeded feature detection in ACIS R23 is likely built on top of that work originally undertaken for SpaceClaim. So, Keith Mountain’s gamble paid off. For all users of ACIS.

I’m pleased to see Spatial continuing to develop ACIS. The company is heavily focused on Dassault Systemes’ CGM modeler these days, and I’m always a little concerned, in the back of my mind, that ACIS might not get the attention it deserves.

For a complete list of improvements in ACIS R23, see “R23 Release Notes for 3D ACIS Modeler”.

 

3D CAD Tips