By Tyler Worden, PE
Model Based Definition (sometimes Drafting or Dimensioning or Design) is an iteration to mechanical design that uses a CAD generated 3D model with dimensions, GD&T information, annotations, and other design/fabrication information. The 3D model with the imbedded information becomes the defining document for the design and is the controlling document for communicating the design internally, to customers, and to fabricators. Shops read and measure directly from the 3D model as opposed to 2D engineering drawings. The MBD method has the potential to accelerate and sharpen manufacturing procedures from design to fabrication and quality control. MBD is slowly replacing other methods of mechanical design in some industries, but it is still rather unfamiliar within many industries. The purpose of this blog is to shed some light on MBD, discuss how it compares to more common design methodologies, and conclude by analyzing whether it has a place in the fired heater and combustion equipment industries.
Figure 1: Example of Model Based Definition
What Does MBD Replace?
As most in the engineering, design, and manufacturing world know, the status quo for communicating mechanical designs relies on 2D engineering drawings which can be viewed on paper or digitally. Engineering drawings are a graphic universal language, playing a fundamental role in human progress from ancient pyramids and the Palace of Knossos to today’s geodesic domes and the space station. The fundamental purpose of an engineering drawing is to carry, control, and maintain a product’s definition in a precise and clear way with minimal risk of misinterpretation or faulting assumptions. 2D engineering drawings are usually produced by two means: 1. Direct interpretation on a 2D space (paper or CAD) and 2. As technology allowed, 2D views as projected (by a CAD program) from a CAD produced 3D model with little to no human interpretation of the views created.
Since the 1980s, the advent of solid modeling and specialized drafting software packages have contributed significantly to streamlining the production of engineering drawings. The solid model defines the product’s geometry, and the engineering drawing consists of only projections of that defined geometry for 2D viewing.
Despite its increased precision, there is a redundancy in effort to produce engineering drawings when they are created as a byproduct of a solid model, since these drawings only echo the information already defined in the solid model. Dimensions, material, and sometimes dimensional tolerances are built into the solid model. Assembly, subassembly, and weldment intent is often built into the solid model by means of nested assemblies and parts. If done correctly, the nesting (along with properties assigned to every part/assembly) results in a prebuilt BOM. Much of this information was assigned to the 3D model during creation, but it must be extracted later in a functional view within the 2D drawing. Most of the effort spent producing a 2D drawing from a 3D model is verifying that the imbedded design information of the 3D model is clearly represented on a 2D drawing and configuring the drawing to have the expected aesthetic look.
However, not all of the time spent on the 2D drawing is redundant. The true complexity exists in the fact that most 3D models do not hold all the necessary information. Surface conditions, like those addressed in ASME Y14.5 GD&T sections, are not as easy to imbed in the solid model. This also applies to heat treatment, weld notes, pressure testing properties, general notes, and many other properties that are considered during the design of equipment but are not stored directly in the model. These are often added to the 2D engineering drawing.
Room for improvement in the solid model to engineering drawing process is quite apparent. To be fair, there are solutions and workarounds that address some of the shortcomings. For example, it is typical to pull any dimension twice. First, during creation of the solid model and second, when a (ideally driven) dimension is created on the engineering drawing. To combat this, some software allows for “smart view” creation and auto-dimensioning for the 2D engineering drawing, thus reducing that specific redundancy. Unfortunately, these tools often need a lot of massaging to work smoothly across similar geometries. Engineering and drafting departments often spend a lot of time creating drawing macros, customizing properties within software tools, and creating complex rules in PDM systems to reduce time spent on each drawing.
For the sake of argument, let’s assume the 3D modelling process is usually more robust than total 2D design. There are still some great tools out there for 2D design --- people with years of experience, and specific applications and designs would argue that point --- but in a general sense the statement has proven itself true. Why shouldn’t we consider foregoing all 2D related design and drafting and find a way to better illustrate the defining properties that went into the creation of 3D models? It seems like the next logical step.
This is where several software companies have tried to enhance the solid modelling process to eliminate the need for 2D engineering drawings. This led to Model Based Definition. Going forward, these are the important questions:
Is it really better than the status quo?
What are the limitations or barriers for implementation?
Does it work in most industries? What about the fired heater industry?
What are the Benefits of MBD?
Obviously switching to MBD requires a full commitment to 3D modeling. General benefits of 3D modeling include:
Ability to visualize clearances and interferences
Enables parametric modeling and interactive assemblies
Reduces a significant amount of room for misinterpretation of the design definition. Parts and assemblies within the 3D model are either accurate or not. Orthogonal, isometric, and section views are exact either way
Photorealistic renderings for marketing purposes can be created from models (Figure 2)
Design considering specific machining equipment, tooling, and operations using CAM imbedded features
Excellent tools for difficult design processes like pipe routing, structural steel framing, and sheet metal bending
Figure 2: Renderings created from 3D models
Obviously, the benefits of 3D modeling are limited regarding communicating the design to the client or fabricator. This is where the benefits of MBD begin to shine:
Going paperless! No need for shop drawings means no bulky filing cabinets or racks for storing drawings. Understandably this can be done in the 2D CAD world as well with digital pdfs, but that is a choice that is not always strictly followed. For MBD not-printing is a Hobson’s choice in that there is not anything to print.
Revision control is simplified in that only one document (or document set for assemblies) needs to be revised. There is no chance of making changes to a model/assembly and not producing the revised drawing.
Fewer documents means a simpler PDM system. One study showed a 25%-30% reduction in file sizes when switching from a drawing and non-annotated 3D model to a MBD 3D model (Quintana).
Less ambiguity. No need to interpret the design into 2D views. Information comes directly from the model itself with geometry and intent fully intact.
Greater potential for erection drawings and installation instructions. As opposed to notes and still images on a drawing, a 3D assembly can be dynamic with animations and the ability for the viewer to realistically manipulate the geometry using flexible constraints.
Automated tolerance analysis. All the tolerances are imbedded in the geometries and the overall effects are calculated by the software. This eliminates the need to flip between pages to sum up tolerance stacking and determining minimum material, etc.
What are the Barriers to MBD?
There is a lot that goes into dramatically changing the procedures that define a product lifecycle. Engineering groups must lay a stable foundation that defines adequate procedures and responsibilities. Management must understand the effects of fundamentally changing the way a product is defined, designed, and communicated. The implications not only affect the engineering group but also project management, applications, sales, nearly everyone in the organization.
Furthermore, the change directly affects the way business is done with groups outside of the organization. Are future customers equipped to handle 3D models as the sole document type to communicate design intent, etc.? What about existing customers that only use 2D drawings? Also consider vendors to and the fabrication chain. How will the shops operate with MBD models? Parts that require CNC fabrication will work perfectly, but weldments and assemblies likely require a hardware revamp for shops that haven not used MBD information before (e.g. tablets, etc.).
While these hurdles might seem high, many issues can be resolved with training and a plan for mitigating hardware inadequacies. The real challenge is distinguishing between the technical obstacles and the cultural barriers. It is often the resistance to a cultural change that creates doubt in an ability to overcome the technical obstacles.
Current Status Across Industries
What does American Society of Mechanical Engineers (ASME) say about MBD? ASME published Digital Tolerance in 2010 discussing MBD and mentioned that it “shows promise for improving and accelerating design, manufacturing, and inspection processes.” In the 2003 ASME Y14.41 Digital Product Definition Data Practices standard, requirements were set for CAD software developers regarding tolerances, dimensional data, and other digital design annotations on 3-D solid models. That standard has been the foundation for many MBD software tools to date. It seems ASME is aware of the advancements in 3D modeling and put energy into standardizing the movement.
The main industry to test and successfully incorporate MBD is aerospace. On the 787 Dreamliner Program, Boeing and Saab engaged in a Virtual Product Development approach where the entire product design, tooling and manufacturing processes (prior to fabricating parts or tools) were verified virtually and thus eliminated the need to generate 2D drawings. Through the application of this approach, they achieved a 62% reduction in product development time and a 42% reduction in development costs (Quintana).
From a study that noted the approach Boeing took came an interesting but important observation: There was a misconception that once the tools were in place the design engineer did not need to interact with the reviewers and approvers (Thilmany). Some design engineers thought they could “throw the design over the fence”, and the tool would see that it was approved on the other end. This, however, is not the case. Personal communication is needed between the design engineer and other team members with today’s technology just as much as in the past. The tools should make the design engineer’s job easier, but they do not relieve the design engineer of the responsibility to get the design approved and resolve any issues along the way.
Many smaller aerospace companies are still hesitant to trust an electronic file over a printed 2D drawing. ASME has written that those companies may have to justify making the changes sooner rather than later to supply or compete with the bigger companies who are early adopters (Thilmany).
Despite cases of success stories, some of the major players that use MBD still admit to not completely switching over. It takes a lot of coordination between parties to adopt a new program. Even the early adopters are not perfect, but it should be noted that the aerospace industry has a rather unique system of an abundance of regulations and inherent complications for making sweeping procedural changes.
Not to throw the fired heater industry under the bus, but it has not historically been the first to adopt cutting-edge methodologies with respect to engineering and design. Except for burners, it is still typical to see general arrangement drawings produced entirely in 2D CAD and detail drawings getting the same treatment. The system of contracting the production of the detail drawings can cause problems by further diluting design information and intent in the communication chain.
A few perks of being