CAESES 5.1 Released

We are excited to announce the new release of CAESES version 5.1, which can be down­loaded here.

Para­met­ric modeling, CFD automa­tion and shape opti­miza­tion — every­thing has been improved in CAESES, to let you design and optimize the most com­pet­i­tive products in the market!

A special thanks to our users who gave feedback and many great sug­ges­tions through our helpdesk and forum!

Visu­al­ize the Influ­ence of Design Variables

CAESES 5.1 offers con­ve­niently acces­si­ble func­tion­al­ity to visu­al­ize the influ­ence of the design vari­ables on the geometry’s shape, using so-called design velocities.

Design veloc­i­ties are defined as the normal dis­place­ment of the model surface due to a change in value of a design variable. They display the regions of influ­ence for a design variable from a geo­met­ric point of view and are espe­cially helpful in under­stand­ing the shape mod­i­fi­ca­tion pos­si­bil­i­ties for a given setup.

GUI Facelift

The graph­i­cal user inter­face has received some fresh updates that stream­line your inter­ac­tion with CAESES and make it more intuitive.

Appli­ca­tion-Specific Workspaces

New work­space tabs for maritime and tur­bo­ma­chin­ery appli­ca­tions create a single point of access for ded­i­cated func­tion­al­ity, making it easier to find the tools you need.

Bidi­rec­tional Coupling of Points

Moving points that are defined with expres­sions and design vari­ables will dynam­i­cally modify the cor­re­spond­ing values without breaking these pre­vi­ously set dependencies.

Selec­tion Filter

The new selec­tion filter in the 3D view allows the user to restrict the selec­tion to spec­i­fied object types, sim­pli­fy­ing accurate picking, espe­cially in complex models.

Easier Modeling and Cleaner Projects

Dynamic Object Types

Dynamic Object Types are now used as the default setting when creating new objects.

New generic point, curve and surface object types allow you to choose the creation method after the gen­er­a­tion of the object and will give hints as to which method is the most suitable for the given task. You can also flexibly change the creation method later as you build your model, without breaking depen­den­cies. It is no longer required to know all object types with their indi­vid­ual prop­er­ties and decide upfront!

Watch the tutorial video here and learn how to enhance your work with the new dynamic type objects.

Enhanced Editing of Curve Control Polygons

New post-pro­cess­ing oper­a­tions for curves create com­pletely new degrees of freedom by allowing down­stream editing of the control polygon for any curve type including:

• Insert­ing, replac­ing, and removing points,
• Refine­ment at given positions,
• Re-inter­po­la­tion with a selected number of points,
• Adjust­ment of start and end tangent,
• Pre­scribed point displacements.

Organize Your Project with Custom Scope Sorting

Fur­ther­more, there is a new option to cus­tomize the order in which objects appear in the tree by intro­duc­ing weight values. This helps you arrange the scopes and objects in the order you prefer, apart from alpha­bet­i­cal sorting. So, there is no need for num­ber­ing your objects anymore.

New Video Tutorial Platform

We have launched a new video tutorial platform that helps you get up to speed with learning CAESES.

Creating helpful videos is cur­rently a work in progress, and we will reg­u­larly add new ones to estab­lish a com­pre­hen­sive library. Feel free to send us your comments and feedback on these videos.

New Design Space Explo­ration and Opti­miza­tion Algorithms

New Adaptive Sampling Method for Initial Variant Creation

In addition to Latin Hyper­cube sampling, we now offer an intel­li­gent adaptive sampling method, which is included in the Dakota design engine. The goal of per­form­ing adaptive sampling is to con­struct a sur­ro­gate model that can be used as an accurate pre­dic­tor for a com­pu­ta­tion­ally expen­sive sim­u­la­tion. Thus, it is advan­ta­geous to build a sur­ro­gate that min­i­mizes the error over the entire domain of interest using as little sim­u­la­tion runs as possible.

The adaptive part refers to the fact that the sur­ro­gate will be refined by focusing sim­u­la­tion samples in par­tic­u­lar areas of interest, rather than entirely relying on random selec­tion or standard space-filling tech­niques. The method uses a Gaussian process (Kriging) model as a sur­ro­gate, based on an initial set of sim­u­la­tion results from a Latin Hyper­cube sampling, which is iter­a­tively aug­mented by addi­tional data points in regions of the design space where the error of the sur­ro­gate is pre­dicted to be the highest.

One Step Steepest Descent Opti­miza­tion with Penalty Functions

The One Step Steepest Descent is a very simple opti­miza­tion algo­rithm for small scaled, single-objec­tive opti­miza­tion problems. The main advan­tage of this algo­rithm is the very low number of eval­u­a­tions required to find a first improvement.

In the first stage, the n‑dimensional gradient of the objec­tive function is deter­mined by per­form­ing n+1 eval­u­a­tions, with n being the number of free vari­ables con­sid­ered. Once the gradient is deter­mined, a one-dimen­sional search for the optimum in the direc­tion of the steepest gradient is per­formed. Due to the sim­plic­ity of the approach (one-dimen­sional search), the result­ing design cannot be con­sid­ered the optimal solution, but rather a first step towards finding a better design.

And So Much More!

Thanks to the steady inflow of feedback and sug­ges­tions from our active users, CAESES 5.1 is even better, easier to use, provides the most robust geometry engine avail­able, and offers an intu­itive and effi­cient inte­grated frame­work for your automa­tion and opti­miza­tion tasks. What else is included?

• BRep of Rev­o­lu­tion: Easily create a BRep by revolv­ing a set of curves
• Design space dimen­sion­al­ity reduction:
  • Combine prin­ci­pal para­me­ters with CAD design vari­ables to give you more options and detailed control
  • Choose the alter­na­tive back-trans­for­ma­tion (from prin­ci­pal para­me­ter space to original CAD para­me­ter space) method to handle more prin­ci­pal para­me­ters and more accuracy for certain geometries
  • Use the dimen­sion­al­ity reduc­tion method with Dakota design engines
• Visu­al­ize and inspect cur­va­ture on BReps, too

Check out the changes log for all details and get the latest version today!

Follow Us

If you are inter­ested in updates about CAESES®, then sign up for our newslet­ter. Don’t worry, we won’t bother you with too many emails. Of course, you can unsub­scribe at any time :-)