How to get better at SolidWorks by doing less

“Speed Kills” is a phrase that is often heard in athletic circles. In sport, if you are faster, you win. The comparison of business to sport, and the inherent relevance of speed in both pursuits, is well documented. Today, I would like to shift the velocity conversation towards a different arena, Computer Aided Design.

At Forge we sell time. A lot of that time is spent in CAD software, specifically SolidWorks, creating fantastic new designs for our clients. The faster we are able to do that, the more value our clients get and the lower their costs.

In pursuit of speed, there are several tips that our engineers, new and old, have found valuable. Over the next few months we will be talking about those areas, in hopes that they can benefit you as well.

Some of the suggestions may seem obvious, but hopefully some are new to you, and all of them will be worth evaluating in your own workflow.

In the interest of speed, let’s begin…

Part 1 – Model Structure

SolidWorks gives engineers an enormous amount of control over how they create and modify designs. Its parametric modeling tools make it possible to build everything from simple components to highly complex assemblies while maintaining relationships between dimensions, features, and components. However, having access to powerful modeling tools does not automatically make the design process efficient.

A significant amount of CAD productivity comes from how the model is structured. Two engineers can create essentially the same part using SolidWorks and end up with very different results. One may create a model that is easy to modify and takes only a few minutes to build, while the other creates a collection of sketches and features that are difficult to understand and require significant work every time the design changes.

The difference often comes down to modeling strategy rather than familiarity with individual SolidWorks commands.

A few principles can make a substantial difference: avoid creating duplicate geometry, use relationships between components when appropriate, and build models with future changes in mind.

Use Patterns Instead of Modeling Repeated Geometry

One of the simplest ways to make a SolidWorks model more efficient is to avoid creating the same geometry more than once. If a feature is repeated or symmetric, there is a good chance that SolidWorks has a tool that can generate the additional geometry for you.

For example, consider a component with four identical mounting holes. One approach would be to create four circles in a sketch and dimension the location of each circle independently. Another approach is to create a single hole and use a pattern feature to generate the other three.

Both approaches can produce the same physical part, but the second approach communicates much more information about the design. It tells anyone reviewing the model that the four holes are intentionally related to one another. If the number of holes or their spacing needs to change, the pattern can be modified rather than requiring each hole to be edited independently.

SolidWorks provides several tools for creating this type of geometry, including linear patterns, circular patterns, mirror features, curve-driven patterns, sketch-driven patterns, table-driven patterns, and variable patterns. The appropriate tool depends on the geometry and the relationship you are trying to establish, but the underlying principle is the same: define the unique geometry once and allow SolidWorks to generate the repeated geometry.

This approach does more than reduce modeling time. It can also make the model easier to maintain because the relationship between the instances is explicitly represented in the feature tree.

Be Careful With Sketch Patterns

SolidWorks also allows geometry to be patterned within a sketch. This can be useful in certain situations, particularly when multiple downstream features need to reference the patterned sketch geometry. However, I generally prefer to use feature-level patterns whenever they can accomplish the same result cleanly.

For example, instead of creating four circles in a sketch and then extruding all four, create one circle, extrude it, and use a pattern feature on the resulting extrude.

This keeps the sketch simpler and places the pattern logic in the feature tree, where it is easier to understand and modify.

There are certainly exceptions, but as a general rule of thumb, trie to use a feature pattern first.

Use Top-Down Modeling When Component Relationships Matter

Another powerful way to improve both modeling efficiency and design robustness is to use top-down modeling for assemblies where the components have important relationships with one another.

Traditional bottom-up modeling starts with individual components. Each part is designed independently and then inserted into an assembly. This is a perfectly valid approach and works particularly well when the components are relatively independent.

However, many products contain components whose geometry is determined by the location, size, or movement of other components. In these situations, designing everything independently can create unnecessary work because the engineer has to manually maintain relationships that could instead be built into the model.

Top-down modeling takes a different approach. Rather than designing every component in isolation, the engineer establishes the important geometry and relationships that define the overall product and then uses those relationships to drive the individual components.

Two common techniques are the master model and skeleton sketch approaches. If you are unfamiliar with one or both, they are certainly worth future research., however their complexities and methods are outside the scope of this article.

Imagine a mechanism containing several moving parts. Rather than designing each component independently and manually maintaining the distances and interfaces between them, you can establish the critical locations, interfaces, and motion in a master model or skeleton sketch. The individual parts can then reference that information.

If the location of a mounting feature changes, the components that depend on that location can update accordingly. Instead of relying on the engineer to remember every downstream feature that needs to be changed, the model itself can carry much of that design intent.

This is one of the major advantages of parametric CAD. The model can represent relationships between geometry rather than simply representing the final geometry itself.

Manage Top-Down References Carefully

Top-down modeling does introduce additional complexity, and it needs to be implemented with some discipline.

If external references are not carefully managed, a model can develop excessive dependencies that increase rebuild times and make the design difficult to troubleshoot. Poorly planned dependencies can also create circular references.

A good top-down modeling strategy establishes a clear hierarchy. A master model or skeleton establishes the primary design intent. Individual parts reference that information, and the assembly brings those parts together. Information should generally flow in one direction rather than creating a network in which every component references several other components.

This requires additional planning at the beginning of a project. For complex products, however, that organization can be well worth the effort because it creates a model that is easier to update and less likely to develop inconsistencies.

A good parametric model does more than describe what the product looks like. It captures some of the reasoning behind why the geometry is arranged the way it is.

Think of the Model as a System

The most useful way to think about parametric CAD is as a system rather than a digital drawing.

That distinction becomes increasingly important as the complexity of the product increases. When a model contains dozens of components and hundreds or thousands of features, the ability to make a change without manually rebuilding large portions of the design becomes extremely valuable.

Patterns, symmetry, master models, skeleton sketches, and well-managed parametric relationships all serve the same larger purpose: they reduce redundant work and allow the CAD model to carry more of the design intent.

The best SolidWorks models are not necessarily the ones with the fewest features or the most sophisticated techniques. They are the models that make sense to another engineer, respond predictably to design changes, and allow the designer to spend time solving engineering problems instead of repeatedly rebuilding geometry.

Designing efficiently in SolidWorks starts with thinking about the structure of the model before thinking about the individual commands used to create it.

See you next month for Part 2 – APM