Why Aren’t There Basic Dimensions in MBD?

Jennifer Herron shares her thoughts on one of the questions we get asked a lot in the MBD transition: why aren’t there basic dimensions? 

We get asked this all the time by folks who haven’t been through our training. Typically, they think basic dimensions have to be measured, which is inaccurate. There are theoretically perfect dimensions that are used to describe the geometry so that you can recreate a 3D shape from a 2D drawing.  

The question then becomes, if you have a 3D model ready, why do you need those? In all the cases where the model is used in the manufacturing, like CNC path programming, you don’t need dimensions. When you’re automatically generating a bill of characteristics from the product definition for a quality plan, you don’t need the basic dimensions because there’s not somebody manually creating geometry and doing math in quality to figure out how to set it up. 

People ask for basic dimensions because they’re manually doing geometry from a 2D drawing to create manufacturing work instructions, CNC programs, or measurement plans. If you don’t have that, you don’t need them. If you have that, there’s definitely a case for them. And for the majority of the parts that we see in MBD, the ones with complex geometry are already being consumed directly into a machine tool. For those parts, you definitely don’t need them. If you have manual operations happening where people need to read and do some math and geometry, there’s definitely an argument for them. 

What does ‘minimally dimensioned’ mean? 

We also used to call that a ‘reduced dimension drawing’ or a ‘limited dimension drawing.’ That occurred in the 2000s and still persists: the drawing just has the tolerance information on it, but leaves the basic dimensions in the model because they knew they were consumed already into a machine tool.  

We have a hodgepodge of different scenarios going on. People really want to be drawing-only or model-only, and they really struggle with the hybrid approach. Right now, realistically, you’re going to have stuff that’s drawings, probably with basic dimensioning on it, but it is better to have good GD&T versus plus/minus tolerancing. The big difference is that people haven’t transitioned from 2009. 

When the standard in 2009 was released for geometric tolerancing, they were still using 1994 ways and didn’t bother to get retooled. So in the 2018 standard, they’ve refined the 2009 standard nine years later. Took them nine years. They had something like 5,000 comments from the public to adjudicate as feedback. And we’re still struggling to get people to even go to 2009. The bottom line is that if you’re going to go to the 2009 standard for ASME-Y14.5, you might as well go to 2018. It’s just a better 2009. And there’s no reason that you should not be going there. There’s literally zero reason. It’s been analyzed heavily at different companies and standards. 

Why does the hybrid approach seem to make people more reluctant to move to model–based? 

It’s related to the classification codes: 

• Classification code 5: Model-Only 
• Classification code 4: Model-Only, drawing displays the same information
• Classification code 3: Model and drawing, with the model holding all the tolerance information

Once you try to talk yourself into three and four, you go, “Why don’t we just do model only?” Everybody always ends up there, but it does generally take them two years. So basic dimensions are already in the model. They are the model.  

The basic dimensions are the model. 

But people really struggle with that because they’ve been applying basic dimensions for so many years they just can’t get through it. Once they take our MBD Using Modern GD&T course, then it’s done. They get it. 

Which is the bigger culture shift: going from old-fashioned tolerancing to modern GD&T, or going from old-fashioned dimensioning to modern GD&T? 

Dimensioning is absolutely is the harder part, because they’ve built all these crazy processes and qualification programs and stuff that they do. They’ve made a bunch of Rube Goldberg mechanisms because they had to go from 3D in the brain to 2D on the paper, from 2D on the paper back to 3D in the brain, from 3D in the brain to 2D on the paper. There are all these crazy gyrations to do all that transfer from design to engineering to quality. And they’re so invested in all those processes and people do very specific things in those processes that they won’t give it up. 

Where’s the benefit of taking this approach? 

I typically say, “As a rule, don’t display the basic dimensions and then add them back in when you need them.” Don’t take away engineering efficiency for 80% of the stuff that you don’t need. And that’s probably what we’re talking about, is 80% of the parts do not need the basic dimensions. They may sometimes need manufacturing dimensions, which is different, and that’s held in the manufacturing definition. 

What are the 20%, the ones where they really do need something more? 

As an example, sheet metal. Where the layout is for the bend line, that’s an excellent spot for a basic dimension, or some kind of dimension that can be coded in. Sometimes they’re automatic, but sometimes they’re just manual layup and setups. Basic dimensions are for anything that’s done with a manual operation that requires a manual layout. But, for instance, on the sheet metal flat pattern, where they cut before they bend it, they don’t need the basic dimensions because it’s always cut with a program. Nobody goes in there and hand snips the thing. 

You’re going to have to look at it. You’re going to have to look at each of the parts you build and ask the question, can I get away without displaying the basic dimensions? And that normally comes from the manufacturing and inspection feedback, which is why it’s always so critical to get that manufacturing inspection feedback early on. 

---

---