(This article is one of my series of Frequently Asked Questions posts. See the rest of the FAQ pages.)
After a discussion with a colleague a few months ago, I went searching around the Internet to find a good definition of parametric history as it applies to product design CAD and jewellery CAD in particular. Unfortunately, after days of searching it appeared no one on the internet has ever bothered to explain this parametric modelling technology in layman’s terms, much less in terms relevant to a jewellery CAD designer. All my searching turned up was either very generic descriptions of parametric or direct modelling, highly technical articles from industrial engineering blogs, or salesmen selling their software without ever getting specific on technical details.
Indeed, during my research of for this article, it seems as if the most technical aspects of parametric and direct CAD modelling have been intentionally simplified to make the user interfaces more friendly for product designers. However, that did not change the fact nobody I had found had ever tried to explain what it is in a clear way before.
In an effort to solve this problem once and for all, I’ll now demystify what parametric solid modelling and its newer cousin direct modelling really are, and how they work with jewellery CAD.
Perhaps I’m getting a bit ahead of myself. First, let me describe what parametric modelling actually does.
We’ll start by considering what the Undo button has done for everyone who uses a computer.
As anyone who has ever had to write a book report in Microsoft Word can tell you, one of the most useful features in any design software package must surely be the Undo key. Being able to go back one or more steps and redo work is a tremedous asset of working with a computer. However, the Undo button can only help you when you only make small incidental mistakes. What if you make a mistake much further down the line, and you discover that you’d not only have to hit Undo 20 or more times to fix it, only to find you’ll now need to redo just as many complicated steps to get back to where you were, all in the name of fixing one mistake spotted too late?!? <Gasp>
This is why Parametric History was created. The basic principle of parametric history is that we can change an earlier stage of a model, and thanks to how the model has been constructed, all the subsequent stages will automatically adjust themselves, saving you a lot of undoing and reworking.
What’s more, when a model is designed, every designer inevitably has a Design Intent in mind for the model. Within Parametric history, you can specify which parts are connected to each other, or which properties of a given item are variable, thereby limiting what can and cannot be changed and ensuring your design will be functional.
At least, that’s the principle. In practice, the execution is a bit more complicated. As we discuss it in more detail, the reason for the existence of Direct Modelling will become more clear…
Where Parametric History Comes From
The origins of Parametric History go back to some of the second generation AutoCAD clones that appeared on the market sometime around the mid 90’s.
At that time, AutoCAD added a feature to their software allowing users to create 3D extruded projections based upon perfectly drawn 2D technical drawings, so you could see the shape your drawing would produce in 3 dimensions. In the years before 3D product design CAD became widespread, this saved a lot of testing time for AutoCAD draftsmen. This additional data stored inside the extruded shape which tied it to the original vector lines was called Parametric Data. Thanks to this parametric data, the lines could be changed, and the extrusion would be updated automatically to represent the new line shapes.
The first forms of parametric history modelling were an extension of this. By attaching additional information to CAD objects you created, you could make more dynamic changes to the model and see the results of those changes in 3D.
Once 3D CAD came about and became widespread, software developers decided to re-approach this old method of attaching key information to 3D CAD objects. By anticipating the most common combinations of solid or surface commands used to make objects, they could give CAD users an ability to change complex objects after they had already been made without having to undo any work. In order to do this, however, they had to keep track of a timeline of each step in the modelling process. This was how the concept of the Parametric History Tree was born.
What Parametric History Really Means in Practice
According to David Lehmann from 3Design CAD: “The idea of parametric modelling is to simplify the process of changing the model later. Each parametric object saves several parameters (hence the name) which are saved in that object, and can be modified later on…The way each object interacts with each other is drawn in a history tree, and you can go down into the tree to find and change the exact parameter you need.”
As you might expect, the parametric history tree brings with it an entirely different approach to modelling from other types of CAD. Specifically, the creation of every single object is now associated with a chronological “step” in the history tree. Thanks to this dedicated recording of the stages of your modelling process, you can potentially come back to an earlier stage and make changes after the fact. What’s more, if the parametric history tree is designed properly (and your requested change falls within the design intent for the model), then each change you made would then update all of the subsequent stages.
When you think about it, that’s quite powerful– you can design a full ring, go back to the original ring size or main gemstone, change it, and watch as all the subsequent steps will automatically adjust.
So what does this mean for your modelling process?
The Advantages and Disadvantages of Parametric Modelling
The most obvious advantage to parametric history this is that it means every single model you make can be changed easily on even the most fundamental level, no matter how long ago the model was made. The implications of this in practice can be quite impressive— stone setting sizes and shapes can be quickly replaced, drilled holes and filigree can be moved along a surface, and generating ring size ranges becomes simple and quick.
However, there are some drawbacks to parametric modelling over older modelling methods. Specifically, the concept of design intent itself forces you to plan ahead. Because everything you make is stored as a stage in the history tree, it can be possible to build things in such a way that the objects don’t relate to each other as subsequent steps in the way you’d want (a common sign of badly considered design intent). For example, if I am building a stone setting for a gemstone, I should build the gemstone first before building the setting. If I built the setting based on a different object or on its own, that setting would not respond if I changed the gemstone.
In effect, this planning ahead required by design intent and the history tree can become a separate item to manage in and of itself. Under most circumstances, this can actually help enforce good modelling practice onto your 3D modelling, but there are occasions (such as particularly unusual or unorthodox forms) where working within parametric history can end up making the process of modelling certain components more complicated than they would strictly need to be.
Also, while parametrics can make implementing changes to a model very quick indeed once that model has already been made, when it comes down to basic modelling from scratch, it’s not necessarily any faster.
To show you how it all works, here is a video showing how a 3 stone ring would be made with parametric history:
What Direct Modelling Does Differently
Direct 3D CAD Modelling (often shortened to just Direct Modelling or called Explicit Modelling) was developed as a response to the restrictions imposed by working with a history tree. Developers loved the idea of dynamic control of objects and surfaces, but didn’t want the hassle of worrying about the chronology of making the shape. So direct modelling was developed based on the principle of being able to change the key edges of surfaces of any object on the fly, providing a significant amount of ability to rework an object whether it was tied to a history tree or not.
Technically, direct modelling is parametric controls with less consideration about the history tree structure and more physical control within the actual workspace.
Advantages and Disadvantages of Direct Modelling Over Parametric Modelling
The advantage of direct modelling was to grant the best of both worlds— the reworkability of parametric history modelling combined with the flexibility of being able to work on any object you want anywhere you want without worrying as much about a history tree. In practice, this can lead to a very fast workflow once you understand the basics. According to Narine Babikian from Firestorm CAD: “People come to me after using Rhino and tell me ‘I wish someone had told me about Firestorm earlier. The interface is so much easier.’ And I say: ‘I know it’s easier.’”
However, while direct modelling does give more freeform control over surfaces, one big disadvantage is that most direct modelling doesn’t do quite as much to manage the history tree as parametric modelling would, so connections between objects may not be as certain to be maintained. In practice, this means making changes to one object is less likely to change other objects around it.
Here is an example of the Direct modelling program Firestorm in action:
The parametric history tree did make a proper splash on the CAD software industry when it first appeared. Since then, it has become one of the primary types of CAD software available for product design CAD modelling. Several software packages make good use of it, and some of the newest generations of the history tree are quite powerful indeed.
Direct CAD’s natural evolution from Parametric modelling gave users even more freeform flexibility with the same editing tools, opening the door to methods of reworking models never before imagined by CAD designers.
Would these tools be useful for you? That really depends. The big advantage to these tools is of course that they allow you to rework your old pre-existing models, and they are very helpful if you’re the type of designer who likes to repeatedly rework and experiment with a piece you’ve already made. Indeed, most CAD programs nowadays try to incorporate some limited parametric or direct modelling functionality (see the list below), but the main reason why modellers would opt to not use these tools is simply because they aren’t yet available for use with certain types of CAD geometry. For example, most mesh sculpting tools, subdivision modellers, and surface modellers don’t really have much in the way of parametric or direct modelling functionality.
In short: If you find you are the kind of designer who likes to rework your old models over and over once you’ve made them, these tools can prove to be a great asset for saving you time and making life easier for recycling components. As for the difference between parametric and direct modelling, it’s mainly one of control over your models once they’re made– With direct modelling, you have flexibility of modelling matched evenly with reworkability. With parametric modelling, you maximise your reworkability of your models in exchange for a more rigorous modelling process.
Further Reading – Parametric Jewellery CAD Software On the Market
Now that we have presented the technology, you can find out more about parametric jewellery CAD and direct modelling straight from the respective sources. At the present time there are two different jewellery CAD software packages which are based upon parametric or direct modelling, respectively:
There are also a few other software packages which have incorporated a few direct modelling features (such as Rhino) or parametric modelling features (such as PowerShape).
Note that this list may change over time, as new software appears or updates.