Large CAD assembly management | The value of top-down design strategy

WRITTEN BY: ROBERTO PAPALIA

12 MARCH 2019

Since the good old days of the glorious drawing table and the pencil, the development of a new assembly starts always with the graphical representation of the project's layout. In particular in cases where the final object was an assembly of multiple parts or sub-assembly the purpose of the graphical model was to capture specifications ang organize components contained in a variety of other technical documents.

 

since that time to the era of 2D CAD (substantially a digital extension of the drawing table, from a methodological point of view) modern 3D parametric and explicit CAD systems make the ideal platform for developing, in practice, products that include mechanical hardware.

Roberto Papalia

Through my long experience as an engineering consultant in various industries and different domains, it is evident that it’s not so simple to transfer the classical concept of a layout from the old 2D approach to modern 3D complex assemblies. Often the result of a large 3D CAD assembly is becoming too complex to manage mainly for the following reasons:

  • The evolution of each project, due to revisions and modifications during its development determines disorder in the structure of the CAD assembly. Objects are not always nested in an easy and logical structure. It becomes too difficult to understand in which subassembly a certain object is contained
  • The regeneration of some parts or sub-assemblies often fails or slows down because multiple restructuring actions within the assembly brings to alterations among components constraints

Often (is there any mechanical designer who hasn’t already experienced it?!) the feeling of the designer at the end of the project is that, once the large assembly is finished, it’s better not to touch anything else. Just leave it as is and hope that no further modification is needed.

So, what are the strategies to avoid this extremely problematic scenario?

We can see a multitude of different hybrid strategies that are difficult to categorize.

Let’s mention at least a few of the common cases:

  • Positioning of sub-components constraining them to one or more given principal coordinate systems. This avoids the problem of regeneration failure once a components that is positioned in the middle of a sequence or chain of other components is suppressed, causing the majority of CAD systems to get stuck, face to a lack of constraints.
  • Creating logical sub-assemblies and trying to connect them with geometrical constraints that enables, at least at the top level of the assembly, an easier management.

Both of these systems have a number of drawbacks.

If components are fixed to a principal coordinate system this will determine a lack of “adaptivity” of the whole assembly face to further modifications. Reciprocal positions and clashes will be handled manually (typical of explicit 3D modelling systems).

 

If components are organized in “sub-groups” which is certainly very good, however this strategy doesn’t prevents the danger of regeneration failures of other problems mentioned above. We always have to see at which level this grouping is made and, further on, what happens inside each single subassembly. Moreover, things become more difficult when inside the assembly contains a mechanism. In this case subgroups shall be connected by kinematic constraints.

 

There are quite a few things that can be done to organize properly our CAD assembly.

Top down design strategy in 3 easy steps

The solution that I prefer and I propose in all cases when dealing with large assembly with kinematics constraints is the top-down design strategy.

This approach goes somehow in parallel with the organization of a typical design office of a modern company: there is someone that generates an output of a concept layout and there are people, at different levels who develop the assembly to the minimum detail. How to harmonize and conjugate all this with the requirement of minimizing the development time?

Well, top down design and concurrent engineering seem to give the most effective answer. Let’s make a very simple example (classical connection rod-crank-piston) of a typical work flow that respects the rules of this method. This might seem quite evident, but judging from the results that we can see from many design offices, it’s not!

  1. The specifications of the system that has to be designed are captured by very simple 3D sketches (datum points, axes, planes and reference surfaces can be used as well). For instance, in the image below representing the profile of a simplified connecting rod, the rod length is part of  them. Please note that the more these “skeleton” references are accurate, the easier will be to define:
    1. The reciprocal positioning of parts (sub-assemblies) within the subassembly in object
    2. The shape of the main components of the assembly in object (with no reciprocal movement; so, no kinematics, so far)



  2. All the “skeletons“ are connected one each other with kinematic constraints, in order to represent the mechanism we are studying. Normally in this phase, a kind of “dialogue” between the simulation and calculation people and the design people is open in order to optimize (under a dynamics point of view as well) and give sense to the specifications.



  3. Once the mechanism is defined, each subassembly receives its “solid dress”, giving the final shape to each single component, although this final shape could be driven by the layout skeleton within the subassembly. Thus, the optimization of the whole mechanism and the detail of each single part can be parallel, giving a sense to the “concurrent engineering” approach.

Obviously This is a simplified example. The number of components is small,  and therefore  very easy to understand what will happen if the assembly  that we are  working on is large. The good side of this approach is, for instance, the possibility of extending it to an assembly that contains a very large number of components and, more important, with very high degree of complexity.

About Nova

Nova is a leading engineering company providing advanced technological solutions, product design and development in a broad range of mechanical engineering disciplines as well as comprehensive engineering consulting services to leading companies worldwide.

    1. The shape of the main components of the assembly in object (with no reciprocal movement; so, no kinematics, so far)



Since the good old days of the glorious drawing table and the pencil, the development of a new assembly starts always with the graphical representation of the project's layout. In particular in cases where the final object was an assembly of multiple parts or sub-assembly the purpose of the graphical model was to capture specifications ang organize components contained in a variety of other technical documents.

 

since that time to the era of 2D CAD (substantially a digital extension of the drawing table, from a methodological point of view) modern 3D parametric and explicit CAD systems make the ideal platform for developing, in practice, products that include mechanical hardware.

Large CAD assembly management | The value of top-down design strategy

WRITTEN BY:

ROBERTO PAPALIA

12 MARCH 2019

Roberto Papalia

Through my long experience as an engineering consultant in various industries and different domains, it is evident that it’s not so simple to transfer the classical concept of a layout from the old 2D approach to modern 3D complex assemblies. Often the result of a large 3D CAD assembly is becoming too complex to manage mainly for the following reasons:

  • The evolution of each project, due to revisions and modifications during its development determines disorder in the structure of the CAD assembly. Objects are not always nested in an easy and logical structure. It becomes too difficult to understand in which subassembly a certain object is contained
  • The regeneration of some parts or sub-assemblies often fails or slows down because multiple restructuring actions within the assembly brings to alterations among components constraints

Often (is there any mechanical designer who hasn’t already experienced it?!) the feeling of the designer at the end of the project is that, once the large assembly is finished, it’s better not to touch anything else. Just leave it as is and hope that no further modification is needed.

So, what are the strategies to avoid this extremely problematic scenario?

We can see a multitude of different hybrid strategies that are difficult to categorize.

Let’s mention at least a few of the common cases:

  • Positioning of sub-components constraining them to one or more given principal coordinate systems. This avoids the problem of regeneration failure once a components that is positioned in the middle of a sequence or chain of other components is suppressed, causing the majority of CAD systems to get stuck, face to a lack of constraints.
  • Creating logical sub-assemblies and trying to connect them with geometrical constraints that enables, at least at the top level of the assembly, an easier management.

Both of these systems have a number of drawbacks.

If components are fixed to a principal coordinate system this will determine a lack of “adaptivity” of the whole assembly face to further modifications. Reciprocal positions and clashes will be handled manually (typical of explicit 3D modeling systems).

 

If components are organized in “sub-groups” which is certainly very good, however this strategy doesn’t prevents the danger of regeneration failures of other problems mentioned above. We always have to see at which level this grouping is made and, further on, what happens inside each single subassembly. Moreover, things become more difficult when inside the assembly contains a mechanism. In this case subgroups shall be connected by kinematic constraints.

 

There are quite a few things that can be done to organize properly our CAD assembly.

Top down design strategy in 3 easy steps

The solution that I prefer and I propose in all cases when dealing with large assembly with kinematics constraints is the top-down design strategy.

This approach goes somehow in parallel with the organization of a typical design office of a modern company: there is someone that generates an output of a concept layout and there are people, at different levels who develop the assembly to the minimum detail. How to harmonize and conjugate all this with the requirement of minimizing the development time?

Well, top down design and concurrent engineering seem to give the most effective answer. Let’s make a very simple example (classical connection rod-crank-piston) of a typical work flow that respects the rules of this method. This might seem quite evident, but judging from the results that we can see from many design offices, it’s not!

  1. The specifications of the system that has to be designed are captured by very simple 3D sketches (datum points, axes, planes and reference surfaces can be used as well). For instance, in the image below representing the profile of a simplified connecting rod, the rod length is part of  them. Please note that the more these “skeleton” references are accurate, the easier will be to define:
    1. The reciprocal positioning of parts (sub-assemblies) within the subassembly in object
    2. The shape of the main components of the assembly in object (with no reciprocal movement; so, no kinematics, so far)



  2. All the “skeletons“ are connected one each other with kinematic constraints, in order to represent the mechanism we are studying. Normally in this phase, a kind of “dialogue” between the simulation and calculation people and the design people is open in order to optimize (under a dynamics point of view as well) and give sense to the specifications.



  3. Once the mechanism is defined, each subassembly receives its “solid dress”, giving the final shape to each single component, although this final shape could be driven by the layout skeleton within the subassembly. Thus, the optimization of the whole mechanism and the detail of each single part can be parallel, giving a sense to the “concurrent engineering” approach.

Obviously This is a simplified example. The number of components is small,  and therefore  very easy to understand what will happen if the assembly  that we are  working on is large. The good side of this approach is, for instance, the possibility of extending it to an assembly that contains a very large number of components and, more important, with very high degree of complexity.

About Nova

Nova is a leading engineering company providing advanced technological solutions, product design and development in a broad range of mechanical engineering disciplines as well as comprehensive engineering consulting services to leading companies worldwide.

Roberto Papalia