“Class A” surfacing is an occasional topic on this forum, with some participants being very passionate. It might be worthwhile to consider where the concept orgintated and how it evolved.
The term “Class A” originated in the auto industry when auto design did not involve any computers and no math was used in designing the free-form surface shapes or creating the dies used to stamp sheet metal and the molds used to make cast metal and plastic parts. Class A surfaces are the primary exterior and interior surfaces which are visible with doors, truck lids, hoods, etc. closed. There surfaces need to have the highest visible quality with no visible defects, hence “Class A”. The quality of reflections and the feel of the surfaces are two ways the quality of the surfaces was and still is evaluated. Degree of mathematical continuity was not used.
Much of the focus on achieving Class A surfaces was on the die and mold making and the manufacturing processes. Achieving Class A quality was more the responsibility of the manufacturing staff rather than the design staff. The dies and molds were made and then refined to eliminate surface defects and ensure sufficient visible and tactile matching between parts. Potential defects in stamped metal include warping, bunching, and wrinkling. Potential defects in cast parts include prominent parting lines, sink marks and warping.
It was only after math modeling of surfaces was introduced in the process that discussion of geometric continuity and similar became relevant.
I’m surprised I didn’t know that the term class A is older than the heart of the debates that often take place around the famous g1 and g2 continuity and their brothers and sisters. one more info.
The first time I heard the term Class A , was in the 80’s and it was explained in a way of surface quality and that was in composites. How to achieve a surface that looked like a piece of glass out of resins and fiberglass. After listening to a lot of explanations of the term, I like all of them because it means that you are trying to achieve something better than just good enough.—-Mark
Design and final surfacing have been and are two distinct functions in the auto industry. The design is created in the studio, and once the design is completed it is passed to a group who refine the surfaces to Class A standards and add details such as flanges and similar. There is no need for the initial designs in the studio to meet Class A standards and to do so would be a waste of time and slow the design process.
Also even within a studio (when I left the auto industry in 2009) the creative designers were distinct from the digital and clay sculptors. Creative designers would draw on the computer but they did not directly create digital 3D surface models.
I left an Automotive exterior Class-A surfacing department in 2019.
The ‘Designers’ were responsible for defining the shapes and would use anything at their disposal (Photoshop, hand sketching, Alias…) to get the message across.
The designs were then given to the CAS department who were expected to create 3d models quickly and usually used Alias (not to Class-A standard but many of the guys were pretty damn good). They had to hit hard points and consider packaging constraints.
These models were then milled in clay.
The clay modellers would then refine the clay after which several rounds of modifications were made (mill - refine - change clay - scan clay - create new 3d model - repeat…)
At a certain point, after they were way over budget and missed all their gate-way dates (which resulted in compressed timescales for the Class-A bit…) the design was then handed to the Class-A department.
The Class-A department was tasked to -
Refine the surfaces.
Work with the Designer to implement any style changes.
Work with the Engineering department/Manufacturing/Tooling to ensure 100% feasibility for manufacture.
The resulting surfaces were then handed over to Engineering who then carried on with all the b-surfaces and all the necessary stuff that isn’t ‘visible’ to the customer.
Traditionally ICEM Surf was the software used for Class-A. It was developed by VW (in the late 80’s I believe) and then made available to buy throughout the industry. These were the days before NURBS. As a side note - I remember working as a design engineer in the early 90’s using CADDS4X where you couldn’t trim holes in surfaces. You just had to create some curves on a particular layer which were defined as ‘trim curves’. Then CADDS5 came along with it’s NURBS and blew my mind!
Alias is a relative new-comer to the Class-A scene but seems to be adopted more and more.
In my opinion there are many good things to be learnt from the ‘traditional’ Class-A way of 3d modelling with single spans. It’s not suitable for every job and some aspects are totally unnecessary for the level of many products. It’s also very time consuming. But at least if you have the knowledge of the techniques used and why, then you can make an informed decision about whether or not to use these techniques in a particular situation. That’s of course if your software lets you…
Rob’s description above of the automotive design process including the iterative mill in clay - sculpt the clay - scan the clay - revise the math model - mill - etc is very similar to what I observed when I left the automotive industry in 2009. I was on the product engineering side and we along with manufacturing worked with the studio to get as much criteria into the models as possible before they we released to the Class A department. That was a change from the previous “throw it over the wall” approach.
What I begin teaching trainees from day one. The simpler the curve or surface, the easier to control its shape, and the easier to achieve a proper monolithic look (perfect visual continuity) across surfaces. As you say, once you know, you can then make an informed decision if it’s worth it, depending on client and product.
I always like to refer Rhino beginners to Autodesk Alias tutorials help section, as these are mostly principles that apply, to a certain degree, even to surface modelling in, for example, parametric CAD packages like SolidWorks and Creo. I often see good Rhino beginners having much more success with matching or the sometimes dreaded filleting, when they understand to keep things lightweight and simple.
This guy has quite a few good beginner videos about these issues beginners often get wrong, never worrying about spans. He also sends people to the Autodesk Alias golden rules site, lol.
I agree 100% these rules are not exclusively to (alias).
they are applicable on all nurbs software.
I went to read them out of curiosity.
they describe the right way to approach modelling. on the other hand, I think that the benefits that can be drawn from these rules are not enough to approach modeling effectively in rhino.
but it remains a solid starting point to start
Here I would disagree. Although Rhino lacks many easy to use manipulation and analysis tools, the basics are all present. I opened, via IGES, really good Rhino models in Alias and ICEM, from a beginner or mildly advanced beginner point of view. The thing is, Rhino, as per the underlying tools/math has the toplologic information present. That much is certain. It is just that it is not revealed to the user in an easy way or easy to manipulate, like in Alias, for example. Maybe there are software/algorithm patents that get in the way. But what do I know, lol.
