We are all very familiar with 2D drawings. They’ve been used for hundreds of years and they still work. Why should we bother with model-based definition (MBD)? What are the concrete benefits of MBD? Let’s have a look at the top five reasons to use MBD.
1. MBD further automates manufacturing with software-readable product and manufacturing information (PMI)
Let’s start with computer aided manufacturing (CAM). CAM software programs read CAD models to automate numerical control (NC) code generation. The benefits of CAM have been proven and this automation has been widely adopted.
However, there is one problem: Some key requirements, such as tolerances and surface finishes, are typically defined and presented in 2D drawings. In most cases, CAM software cannot read drawings, so manufacturing engineers have to look back and forth between drawings and CAM programs to manually extract and re-enter these requirements. This step not only slows down the process, but also introduces data duplication, human interpretation and re-entry errors.
One solution to this problem is to define software-readable PMI directly in 3D models, rather than in 2D drawings. This is exactly the gist of MBD. This way, CAM software can automatically read and act upon the 3D PMI. This automation avoids human interpretations and data re-entries, which speeds up production and reduces errors. Figure 1 shows CAMWorks reading 3D surface finishes defined in SOLIDWORKS to automate NC programming.
Figure 1. CAMWorks reuses defined 3D surface finishes to automate NC programming.
After machining, another example is inspection. Acting upon 3D geometric dimensioning and tolerancing (GD&T), CheckMate by Origin International can automatically program coordinate measuring machine (CMM) paths and soft gauges. Furthermore, CMM sample points or 3D-scanned point clouds can be overlaid and compared with the nominal CAD model. Then, CheckMate automatically generates a quality heat map per the semantic 3D GD&T as shown in Figure 2.
Figure 2. CheckMate automates the CMM programming and generates a quality heat map per 3D GD&T.
Along with these two examples in machining and inspection, model-based software-readable PMI can automate many other procedures such as cost analysis, quoting, process planning, robot programming, tolerance stack-up analysis and so on.
It’s important to note that none of these automations would matter if they didn’t bring tangible benefits. To prove the quantitative value metrics of MBD, the National Institute of Standards and Technology (NIST) in the United States conducted a study, Testing the Digital Thread in Support of Model-Based Manufacturing and Inspection.
The research team compared drawing-based and model-based approaches side by side in three steps: annotation, machining and inspection. It found that the model-based approach saved over 60 percent of the net hours across various practical test models as shown in Table 1. The time savings primarily came from the automations powered by the software-readable PMI.
|
|
|
|
|
|||
|
|
|
|
|
|||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|||||
2. MBD increases technical communication efficiencies
We all live in a 3D world and 3D is intuitive to us. When it comes to technical communications, we have to project 3D objects down to a 2D plane to author a drawing. Then, to interpret it, somebody else has to mentally reconstruct this 2D abstraction up to 3D again. This is a detour and it becomes excessive when you consider that most designs are built as 3D CAD models anyway.
This detour not only requires heavy mental coding and decoding, but also invites ambiguities. For example, look at the simple drawing in Figure 3. Is it a cut or an extrusion?
Figure 3. Ambiguity in a simple 2D drawing.
We don’t know, so we have to wait for clarifications, or find another view and correlate multiple perspectives to make a judgment. A simple drawing may be quick to figure out, but if we have to interpret a normal drawing such as in Figure 4, waiting, correlations and judgments are compounded substantially. This can make communication even harder and less efficient.
Figure 4. A normal drawing.
These issues impact business bottom lines enormously. For example, in the NIST study in Table 1, three simple and practical test models by Rockwell Collins were sent to two suppliers for machining and inspection. One took the model-based approach as an experiment and the other took the drawing-based approach as a controlled comparison.
The model-based supplier delivered parts in approximately five weeks, but the drawing-based supplier spent approximately eight months, or 27 weeks longer. The root cause was that the drawing-based supplier had to raise 12 questions related to interpreting the product definition from drawings, which led to work stoppages because the job had to be removed from the queue until clarifications were provided. In contrast, the model-based supplier asked no questions during its machining and inspection work.
Besides these focused studies, drawing communication issues become even more alarming in today’s manufacturing industry, which has grown exponentially more complicated. For example, a Boeing 787 Dreamliner contains about 2.3 million parts according to Jeff Plant with Boeing commercial airplanes. These are just final parts. Now let’s consider the engineering changes generated in the decades of product development and sustainment.
Regarding a similar aircraft, Bob Deragisch with Parker Aerospace pointed out that one change to a simple manifold created 1,700 changes to other related models and systems. The engineering change order (ECO) drawings would be 100 pages for this single change alone. If all the drawings of an airplane were printed, the package would be even bigger than the airplane, to which Deragisch declared “I can’t do that anymore with drawings!”
If a picture is worth 1,000 words, then a model is worth a million words because it’s in 3D and we can rotate and query it. The level of complexity in today’s manufacturing demands model-based communication to improve efficiency.
MBD provides a 3D presentation rather than a 2D abstraction. It minimizes the necessary mental coding and decoding and accordingly reduces miscommunication. In addition, dedicated MBD capabilities such as the cross-highlighting from a callout to its corresponding features provides an instant visual confirmation as shown in Figure 5.
Figure 5. Cross-highlighting from a 3D callout to corresponding features.
Many people believe that the majority of time saved with MBD comes from the avoidance of authoring a 2D drawing. It may indeed save time by not needing to create 2D drawing, but we need to create certain 3D callouts in models too. While 3D callouts may be faster than 2D callouts thanks to the feature-based 3D PMI automations as illustrated in a previous article, the real saving comes from the data consumption side, rather than the authoring side.
The reason is simple: The data, either in drawings or MBD, is created only once, but is consumed many times by many stakeholders. There are many consumption points across an organization and its supply chain, customer base and partner network throughout the entire product lifecycle, making consumption-side savings much larger than authoring-side savings.
3. MBD improves product quality
Much like model-based manufacturing automation, MBD can lead to significant quality improvements. Although the NIST report quoted the net hours and the total delivery time in a side-by-side comparison between drawing-based and model-based approaches, it turned out that time was not the full story. There were also major quality differences between drawing-based and model-based approaches. Figure 6 shows an unintended through-hole and a misshaped groove.
Figure 6. An unintended through-hole and a misshaped groove in the drawing-based part.
The unintended hole scrapped the entire part because there was no cost-effective way to fill it up and make it blind again. The root cause was that the drawing sent to the supplier missed a hole depth callout as shown in Figure 7.
Figure 7. The hole depth callout was missing in the drawing.
Without the depth, a hole defaults to a through-cut in drawings. How did this error slip through the cracks? By simply looking at the drawing in Figure 7, the machinist and even the inspector instinctively interpreted it as a through-cut. It didn’t even occur to them that this could be a blind one because there was no way to tell visually. As a comparison, the model-based supplier caught this issue because it used the model as the authority in numerical code (NC) programming.
In Figure 6, notice the surrounding seal groove on the drawing-based part on the right-hand side didn’t match the original design. This may not be a major issue, but does demonstrate another quality discrepancy due to the drawing-based approach. This type of issue prolongs the cycle time and erodes a manufacturer’s margin and can also compromise customer satisfaction.
Some may argue that these quality issues were the result of mismatching between 3D models and 2D drawings. If the drawings had matched the models perfectly, these issues would have been prevented.
Ideally, that would be true—but in reality, we all know that these discrepancies happen all the time. According to some manufacturers, up to 60 percent of 2D drawings don’t match 3D designs. The problem has more to do with years of drawings maintenance than initial creation. Shop floors could redline a paper drawing on the fly without notifying the design team, or a designer could update a 3D model but forget to update its drawings—especially in 2D PDF or paper formats. The link between models and drawings are broken, intentionally or unintentionally.
Rather than creating perfectly matched drawings separate from the models, why can’t we put them together? Why can’t we bypass drawings and put 3D PMI into models directly in one document?
4. MBD establishes manufacturing competitive advantages
For this reason, more and more organizations and manufacturers are moving toward the MBD approach. In the public sector, the Department of Defense (DoD) in the United States released the Military Standard 31000 revision A in 2013 to specifically define the requirements and best practices for its supply chain. In the private sector, General Electric (GE) named model-based manufacturing as one of the four pillars in its factory initiative, along with automation powered by sensors and the Industrial Internet of Things, process prototyping and informatics.
It isn’t just North America, either. The Japan Electronics and Information Technology Association, or JEITA, is the governing body of the Japan Industrial Standards (JIS), or equivalent of ASME standards in the U.S. In 2014, JEITA members paid special visits to manufacturers and software vendors across Europe and the U.S. to learn about MBD developments. A JIS standard for MBD is currently in the works.
These driving forces from the top of the global supply chain are generating strong ripple effects in the manufacturing industry. In order to be eligible in bids, stay competitive, win contracts and move up in the supplier tiers, manufacturers have to catch up and plan ahead. For example, Figure 8 shows growing percentages of SOLIDWORKS customers using or planning to use MBD. In many cases, small- to medium-sized machine shops have moved to MBD at the request of clients.
Figure 8. Growing percentages of SOLIDWORKS customers using or planning to use MBD (Survey sample sizes: 700 in 2009 and 524 in 2015).
5. MBD unleashes the power of emerging technologies
We are living in an exciting age for manufacturing. Emerging technologies, such as 3D printing, big data analysis, sensors, artificial intelligence and connected machines, are pushing manufacturing forward every day. There have been many initiatives around the globe, such as Industrial Internet of Things in the United States, Industry 4.0 in Germany and Made in China 2025.
MBD holds the potential to unleash the power of these emerging technologies and facilitate these initiatives. For example, 3D printing a part is very easy today with a 3D CAD model, but is unfeasible with 2D drawings. In addition, after printing, the part needs to be inspected according to its dimensioning and tolerancing requirements. Typically, these callouts are conveyed in 2D drawings. Since the design, printing and finished products are all in 3D already, it’s much more useful to avoid generating and maintaining a separate 2D drawing solely for inspection purposes by instead putting PMI directly into the 3D models.
Tolerance analysis is another example. Traditionally, all the tolerances are defined and locked in 2D drawings. Engineers have to visually read and manually re-enter the tolerances from drawings into a spreadsheet to calculate. But with the MBD approach, the digital semantic tolerances are liberated and analyzed by software applications automatically. Even better, the actual downstream as-built quality and cost data can be mined and correlated back with the upstream as-designed tolerances to optimize designs.
The GE “Brilliant” Factory initiative in Figure 9 illustrated them as the production feedback loop and the design feedback loop. The closed-loop analysis can reveal meaningful and actionable insights to cut costs while improving quality. The cost and quality goals may sound conflicting, but the reality is most tolerances are overly conservative. We all would love to loosen them to increase pass rates, but don’t necessarily have the clarity to pinpoint where to loosen without compromising the quality, so end up with tolerance overkills just to be safe. The closed-loop tolerance analysis powered by MBD can provide that much needed clarity.
Figure 9. The GE ”Brilliant” Factory initiative. (Image courtesy of Stephan Biller/Mfg4.)
Although these are the five biggest reasons to use MBD, there are also other benefits to consider, such as reduced paperwork, streamlined processes, workforce hiring/training and job satisfaction. To learn more about how SOLIDWORKS MBD can help you with your MBD implementations, please visit its product page.
About the Author
Oboe Wu is a SOLIDWORKS MBD product manager with 20 years of experience in engineering and software. He is an advocate of model-based enterprise (MBE) and smart manufacturing.