Home / MCAD / Is Our CAD Use All Wrong?

Is Our CAD Use All Wrong?

Are CAD jockeys just that—folks who are adeptly skilled at working the CAD tools to do surface modeling and other complex tasks—or are they engineering wizards simply employing a tool to get the job done?

The question of whether the “D” in CAD is more about enabling good drafting or good design is the subject of a heated debate in the Desktop Engineering LinkedIn Group. Cydesign’s Derek Cheng posited the idea that drafting is not the same as design. He says companies can run into trouble if they leverage CAD too early in the cycle to create beautifully meshed and rendered 3D models prior to do the mathematical work that is foundational to good design.

“Design means understanding what the requirements are, the goals of the project, and ensuring that what you do will work,” Cheng maintained.” Comparatively, he said drafting is spatial, answering the questions of will something fit and how will it look when it fits. “Design is much, much bigger,” he contended. “It requires an understanding of physics, motion dynamics, cyber-electromechanical engineering, and the user experience. Sadly, we often skip the math and go right to drafting it, prototyping it, and testing it.”

Matthew Ian Loew, an independent consultant and a former engineering professor, was equally passionate that CAD is oftentimes misused, particularly when the workflow involves doing detailed CAD designs (specifically, solid models of production intent geometry) too early in the cycle. Loew argues that the industry needs to get back to the process of having disciplined engineers and machine/system architects drive the conceptual development right from the start and only add detailed CAD when appropriate. “My teams use 3D CAD early, but we are not creating ‘product geometry’ and certainly not committing to a product structure before it is time,” he explained.

Loew is also concerned about the CAD industry’s push for greater of ease-of-use, making it even simpler for engineering teams to get to elaborate geometry faster and earlier in the process. “My gripe is that in many organizations, the product development process STARTS with the creation of solid models that can very quickly appear to be production quality,” he said.  If the basic engineering at the product and system levels are not performed up-front, what business does anyone have generating solid models? Engineering belongs up-front — detailed CAD generation belongs at the end of the product development process.”

Premature solid model generation is replacing sound engineering practices. — Matthew Ian Loew

Others in the discussion took exception to this premise, sounding the argument that engineers, not just CAD jockeys, are now taking advantage of the easier-to-use, more accessible, and functionally-rich CAD tools to do  design and drafting as part of an integrated process. “Increasingly CAD users are engineers first, who happen to know how to draft as a skill,” said Steven Weinberg, a mechanical engineer. “The design is embedded in the drafting in a way far more direct than it used to be.

Weinberg and several other commentators made the case for starting designs with a solid modeling approach, using that functionality of CAD to prove out the concept along with FEA, CFD, and a variety of other simulation tools. “CAD allows drafting and design to happen simultaneously, which is how things should be,” Weinberg said. “There might be growing pains along the way, with both design and drafting suffering until we get it right, but I see the end result as both desirable.”

What do you think about where CAD fits into the design process? Do you worry that detailed solid models are created too early in the process, causing teams to skip over critical engineering and mathematical work? Feel free to wade in on the discussion here and read Derek Cheng’s post, which started the conversation. 

If you enjoyed this post, make sure you subscribe to my RSS feed!

About Beth Stackpole

Beth Stackpole is a contributing editor to Desktop Engineering. Send e-mail about this article to DE-Editors@deskeng.com.


  1. I can absolutely stand by the notion that creating detailed solid models too early on is a very frequent and in the end painful, slow and resource hungry process.

    It’s too easy to put in details and make solid parts before the concepts are thought out.

    Management often will push for detailed solid parts as early as they can swindle engineers into doing so. because then they think “wow look at us we are going so fast!” until they hit a wall down the road, where 60% of everything is already highly detailed, yet the concept – at its core – fails.

    Then try redesigning that in a fast and easy way, without wasting or throwing away much of those detailed models. It’s emotionally hard for the engineers to throw away the work already done, even if it turns out to be weong. And the managers have already invested too much money into it that they rather just flesh out the “current” design until it is “acceptable enough”

    Result is painful, time intensive and designs that are only (eventually) satisfying the most critical and basic of requirements

    But I understand the temptation of this! it’s super easy to fall into, I have, everyone has

  2. I do not agree that creating 3D Solid model should be done last. There is nothing wrong with solid geometry early on to establish a concept that will fit geometrically and identifies space and proximity constrains and/or design envelopes. In fact, this is a wonderful advantage of CAD modeling.

    Such preliminary modeling does not have to be expensive. From my experience preliminary 3D CAD geometry can be generated quickly so long as excessive detail is not created too early.

    The danger comes from the penchant for schedule driven projects to treat such preliminary models as complete and ready or nearly ready for presentation as a solution. Sometimes the preliminary work, because it looks so detailed and mature gets mistakenly deemed ready to be turned into production data.
    This is more of a problem in recent years because the CAD software is so good it can render great looking geometry rather quickly and easily.

    It is often the non-technical team members who make the erroneous conclusion that this design is done or nearly done. “That looks good. Let’s go with that. What’s the hold up?” Engineering staff can also make that same error, especially when under pressure to meet schedule.

    There is nothing wrong with preliminary 3D Cad models so long as all involved understand that they are just that, preliminary, the beginning of the design process and not the end.

    Time is the key. Time must be provided for thorough engineering study, investigation, synthesis (design) , analysis and review. Otherwise problems have a way of surfacing at very inopportune points in the project cycle, often with expensive and potentially disastrous consequences.

    The trick is in deciding how much time is reasonably needed. How much analysis is enough? How much engineering design study & consideration is appropriate?

    These questions have always been present in the engineering project and well before CAD software ever existed. It’s a matter of judgement and computers and machines of any kind can’t make judgements. Only people can. It’s best if those people are properly trained and educated to do so.

    The modern problem today is the predilection to assume the software is creating quality designs or analysis because of its speed and power. There is also the tendency for management to think all that is required is the ability run the software.

    There is also the natural tendency to want to believe the computer isn’t wrong. We rely on them so much how unacceptable it seems to find out the computer has given us a wrong answer. This forces us to consider what other wrong answers it might have given us.

    This is why all computer generated data, be it CAD models or Finite Element analysis needs to checked, reviewed and questioned by at least one human being.

    Synthesis and design also still seem to be only properly done with the human brain, at least at present. There is no reason that this cannot be done concurrently with preliminary CAD layouts. In fact such work complements the design process with early information that the designer, engineer or architect needs. There is a big advantage to seeing a 3D render of an idea early.

    However there is a need for a return toward more reliance on the trained and educated professional and not as much assumption that the software can replace that skill set.

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>