IronCad constraint theory

[bmckelvie]

I have been using IronCad for almost 1 year now, and have never yet understood the constraint theory behind IronCad.

 

Coming from a Pro/E background I used to try to constrain every part to make it fully incapable of moving any way that I did not want, but I have found that after models get sufficiently complex, too many constraints just start causing problems. both with the computer slowing down and IronCad becoming impossible to continue to add new parts.

 

Then after downloading other IronCad users models from grabcad to see how they constrain their work and I found that the majority of users did not constrain anything, they just positioned everything with the TriBall and left it. After trying that method I find it to be way faster and have less problems putting everything together. The downside to that is, Mechanism mode completely destroys the model; making any form of simulation impossible.

 

So I am wondering, What is the theory on constraining models in Ironcad, how do you (other users) constrain your models and why do you use that method?

 

 

[ysinitsyn]

Hello.

Even, old school PTC understood that the "Triaball" way is faster.

New tool in Creo

So, if you want to quickly build an assembly - use Triaball.

About mechanism mode

[tlehnhaeuser]

The beauty of IronCAD is its "on-demand" technologies. The abilities to assign parameteres or constraints at will allows you to control your models more effectively becuase you are no burdened with unnecessary constraints that have the potential to impede your goals.

 

In addition, the ability to LOCK/UNLOCK your design constraints on-demand provides unprecented control of your designs before, during and after all phases of your design.

 

So really the theory ( for me ) is to design freeform without constraints, should I need a some sort of constraint in a particaular area of my design, I simply add it on-demand. Then if I need to rework something, I can CHOOSE whether or not to unlock the constraints temporarily , make changes and then Re-Lock.

 

I have used many CAD apps in my life and have never understood why anyone would want to work any other way then this IronCAD approach.

 

Hope this helps.

Tom

 

[bmckelvie]

The bi directional constraint solver in 2012 is infinity better than what 2011 used, I really like it.

 

Assembling using only Triball positioning, then only constraining parts that need to move, or are tricky to get into place without the bidirectional solver appears to be working well. Then when I need to use Mechanism mode, i use "fixed in parent" for everything I can, to try to keep the assembly in the correct shape.

 

Thanks for the suggestions

[tlehnhaeuser]

Just a note, that if you use the Mate/Align Positioning tool, this often bypasses the need for the TriBall and is much faster. Just learn the positioning the indicators and you can increase your speed significantly.

Edited March 6, 2012 by tlehnhaeuser

[Robert Andersson]

The basic principle is: Dont put in more info in to the 3D models than you need (and know).

For instance, in parametric systems, positioning objects is a two step process:

Position and constrain objects.

In Ironcad you only position objects. Constrain is an option if needed, not a basic way of how the program is designed to work.

 

Ironcad is designed from scratch to work without constraints.

(There is no "Rebuild assembly" button for instance.)

 

Booth methods have their ups and downs.

Ironcads way is better if you want to start modeling directly and get feedback from the screen how the model will look witch may make you want to change the models.

Parametric modeling is better if you mostly change parameter values on existing models.

Its more tricky to change how the models are built, cause you have to understand the parametrical relations between the objects.

Edited March 6, 2012 by Robert Andersson