Wednesday, May 26, 2010

Design and Modeling Workflows

In my travels and conversations with architects and engineers, I encounter many different workflows for conceptual design and modeling. Some are clean and efficient, while others are clumsy and less efficient. For example, the old-fashioned workflow went something like this – you would create a set of 2D plan and elevation drawings, send blueprints to a model shop, wait for the first phone call, clarify, discuss, compromise, edit, and repeat. Then, a week or two later comes a 3D physical model that you would be proud to show your colleagues and client … if it wasn’t already out of date! Nevertheless, it was a model that could provide some design insight, and you would display it prominently because it probably cost quite a bit, and you wanted to point to something when your finance person needed to understand what the client bill-back item was all about. For the next design iteration, you likely thought twice about how you would communicate your design changes. Maybe an artist rendering would do the trick this time?

These days, in the 3D design world, a common workflow that I see includes a 3D rendering step in between the BIM/CAD model and the physical prototype. For example, an architect might design a building in Revit, and then export the file to 3dsMax. A visualization specialist would then edit this design to create a file that can be read by a 3D printer. The 3dsMax model becomes the master visualization file for renderings and physical prototypes. This workflow is more efficient than the first example, and results in more frequent, more accurate, and less expensive physical models for design visualization and collaboration.

Sometimes, I witness an STL file export directly from the BIM/CAD software, especially in early design stages where simple white massing models are required. By eliminating the rendering step, this is the most efficient workflow, as long as the 3D digital model is ‘watertight’ and does not require much editing for 3D printing.

On the other end of the efficiency spectrum, one customer described the following workflow:

Revit Architecture-->IFC-->Bentley Architecture-->U3d-->DeepExploration-->3ds-->ZPrint-->mono physical prototype

To be fair, this was not the typical workflow used for every physical model. In this case, the project was so close to CD release that the customer did not want to disrupt the Revit process, so they took the latest design into Bentley Architecture and proceeded from there. On a scale of 1 to 5, where 5 is the most efficient workflow, I would rank this example as a 1.

If you or your firm regularly creates 3D physical models which include 3D printed parts, please evaluate your workflows by answering the 4 questions in the survey link:
Comments welcome, as always!

Tuesday, May 18, 2010

Plaster or Plastic?

You may have heard Z Corp’s announcement today about ZBuilder Ultra, a new rapid prototyping system for producing high resolution prototype parts. This new system produces highly accurate thin walled plastic parts with smooth surface finishes. Some of you may be wondering if this system could be used for AEC models. Good question. In my travels around the world to various architectural firms, I occasionally see plastic models on display mixed among the wood, chipboard, and Plexiglas models in lobby exhibits. When I ask why that material was chosen, I get a variety of responses. Some say that their service bureau had an SLS or SLA machine, so this is what they delivered. Others say that clients perceive the plastic models to be more durable for long term display.

While architects are happy to pass on their costs to the client for a one-time presentation model, most agree that there is no substitute for inexpensive, fast turnaround composite material (gypsum-based “plaster”) models during the early conceptual design and design development process. These are the times during the project when designers want immediate feedback and designs change quickly. How does your firm incorporate physical models into their design process, and what materials are preferred and why?

Wednesday, May 12, 2010

3D Printing and 3dsMax

Today's blog posting was contributed by David Munson, principal of Munson3D, experts in virtual and physical 3D solutions for AEC.

For those who have invested the time and effort to learn Autodesk’s 3dsMax (or 3dsMax Design), there are special benefits. We can think of 3dsMax as a great receptacle for graphic information. It can act as a bridge, uniting the various software products we typically use. Be it from CAD, BIM, Google Earth, SketchUp, Illustrator, or Photoshop, 3dsMax is a place where all that information can find a home. Virtual models can then be expanded upon for not only on screen visualizations and renderings, but via 3D printing technologies, for physical model creation as well.

Awareness of how to prepare files for 3D printing needs to be understood when translating from one format to another, in order to be able to make the 'virtual' – 'physical' bridge happen. Crafting physical models was traditionally done with an isolated, independent process. Now physical models are made from the efforts put into making the virtual model. The physical model simply becomes a physical instance of the virtual model. Conversely, once you have prepared a model for 3D printing, you then have a virtual model which can be used for visualizations.

For efficient preparation for 3D printing, one needs to build the virtual model with solids rather than faces. So, new geometry creation should be with solids rather than faces and imported geometry should come in to the 3D printing software as solids. For that to happen cleanly, export settings need to be set properly. Also, one may think about how to thicken elements for the 3D printed model when creating your virtual model. Walls that may be best to have at true thickness for on screen visualizations may need to be thickened up when preparing files for 3D printing. This typically means extruding a face that is on the inside of an exterior wall.

Actually, the best example to consider may be a pane of glass. For small scale models, we usually print the glass such that the scale model meets the minimum thickness recommendations for the 3D printer. The answer to how one does that will lead you to the best procedure. For your process, that could mean editing a Family in Revit and re-exporting the model [e.g. directly to STL for 3D printing], or working with Instances in 3dsMax, or creating new glass on one facade as one entity spread out over the whole face. There typically are a few valid ways to produce similar results. The key is to understand the needs for both virtual and physical modeling, and to manage those needs along the way. Efficiency and a high degree of aesthetic control are then achieved, enabling you to communicate your design better, both virtually and physically.

Tuesday, May 4, 2010

The Recession and 3D Adoption

Recently, Charles Overy from LGM Models in Colorado offered the following assessment of the recession’s impact on 3D adoption in AEC …

We have been providing 3D printing services to Architects for over 10 years, primarily using Z Corporation printers. Before late 2008, 2D to 3D conversion was a significant part of our business. However in the last 12-16 month most AEC practitioners have "more time than money", a reversal of the previous situation. As a result of the recession, we are seeing far fewer 2D-3D conversions and many more 3D files. The new CAD is predominately from Revit and SketchUp.

Certainly there is the possibility that those who were doing 2D have just reverted to building their own models. I know some architects are. However, our customer research indicates that the recession has been a significant driver in an across-the-board move to 3D CAD in the AEC industry.

Is this true? Is the unintended side effect of the economic meltdown an acceleration of 2D to 3D migration? If so, then all the pain and suffering of the last year and a half may be worth it for the industry as a whole. With empathy to those individuals who lost their jobs, the net effect could be very positive for the long-term health and competitiveness of the industry. While AEC practitioners were not buying many new seats of 3D software last year (just ask the CAD vendors), perhaps they were finally taking the time to pull existing licenses off the shelf and learn 3D modeling, parametric design, BIM, etc. Whatever the motivation, if this is true, it is long overdue in AEC (see my first blog)!

Is there a silver lining in this recession cloud -- are more AEC professionals working in 3D?

I would love to hear your opinion about 3D adoption in AEC.