Model business

Acquisitive companies in the CAD/PLM industry are buying up technology to take their flagship products into new areas, in contrast to the slow consolidation that has been the hallmark of the CAD business for the past 25 years.

recently bought CoCreate Software, a competitive provider of product development solutions. CoCreate started life as the spin-off CAD organisation of HP, whose original products, ME-10, ME-30, Solid Designer and Work Manager, developed into the OneSpace brand of today.

PTC will maintain, enhance and develop all CoCreate products, including OneSpace Modeling, Drafting, Model Manager, Drawing Manager, Live! and OneSpace.net. It will continue to offer CoCreate solutions as stand-alone offerings.

The company also plans to integrate CoCreate solutions with the PTC Product Development System (PDS) to offer CoCreate customers complementary product development capabilities, including engineering calculations, dynamic publishing, visualisation, high-speed machining, and enterprise content and process management.

Although some of CoCreate's modelling technology is not an easy fit into PTC's technological framework, the company says it embraces all the accepted approaches to mechanical modelling - parametric, explicit, derived and 2D. It claims to be the only vendor to offer all these technologies to satisfy a customer's specific needs.

It is ironic that PTC, with its pioneering parametric approach, would acquire a company whose design philosophy in some respects runs counter to its own, although this confirms a MCAD industry trend towards non-history-based methodologies and direct geometry editing, something CoCreate calls Dynamic Modeling.

Since the acquisition announcement, PTC has used the term 'explicit' to describe CoCreate's Dynamic Modeling.

CoCreate's non-history-based modelling approach is unique, but not exclusive. Kubotek's KeyCreator, UGS' NX5, and IronCAD also employ similar approaches.

A history-based approach is what most parametric modellers (such as PTC's Pro/ENGINEER) employ, but it is not always the best approach, especially for changes made to a model late in the design process.

History-based systems use a history tree to track and replay modelling steps for generating and modifying 3D geometry. Each step is dependent upon the previous step. The further along you are in the design process, the more complicated the design and its associated history tree become, so any change late in the process is tricky.

On the other hand, a non-history based approach is well-suited for manufacturers that change designs late in the design process, so they are not bound by the constraints and the complications that a tree can impose.

Also, a non-history-based system can make data import easier because you do not have deal with a history tree from a different system.

Neutral file formats, such as IGES and STEP, are equivalent to native data to CoCreate OneSpace Modeling and other non-history based MCAD applications. This eliminates the burden of legacy data originating from another system.

CoCreate is the latest in a long series of acquisitions that PTC has made over the years, that have included Computervision, Mechanica, Arbortext, Mathcad and IsoDraw.

Another acquisitive vendor last year was

, which recently announced some interesting new modelling technology at its annual CATIA Forum event in Paris.

Dassault's research and development team has been developing a 'direct' modelling kernel for the past 10 years. The problem with the existing modelling technology is that much of it is built on maths and algorithms that have not changed fundamentally in nearly 20 years.

Most parametric and history-based systems struggle with geometric and topological changes. The history-based approach was developed to split the process of building both geometry and topology into discrete steps, which could be calculated and recalculated in a serial fashion.

The new Dassault technology allows you to interact with your 3D model and have the system solve both geometry and topology simultaneously.

The first incarnation of this technology will be a version of CATIA, which works in a similar way to other direct modelling tools. Dassault claims the new kernel can also overcome the drawbacks of existing direct modelling tools when major changes are made to the topology.

Dassault also plans to incorporate the Functional Design tools that it acquired from ImpactXoft. These allow you to design features within a part or assembly that have an underlying logic-base so they can be defined and interacted with without the need for history.

You can run CATIA as a standalone application but in the future, because of changes to the modelling philosophy, it will need to run closely coupled with intelligent data management.

Typically, CAD has an associated data management tool, which is essentially a separate system. If a part or sub-assembly changes, it is approved and everything 'up-revisioned'. If you change a part that is referenced by many assemblies, there is a cascade effect on all the other sub-assemblies and drawings and everything moves up a revision.

If you are working with a full vehicle description or an aircraft, you cannot have a system that manages design change at the file level only as minor design changes can cause serious 'housekeeping' problems.

CATIA's data manager (Enovia) manages design change at a much finer granularity at the feature, or entity level. This means the impact of a design change can be propagated and dealt with at a much lower level, so the system is more aware of the impact of change.

The new CATIA, closely coupled to an updated Enovia platform, will have the ability to create geometry easily, and will ensure this geometry is valid at all times. With the new system it is theoretically possible to have two or more people working on the same part at the same time. Features will be communicated between each member as they work on them, the model will be up to date, the geometry and topology are solved in real time, and design change is managed at the entity level — Utopia indeed.