VP of Engineering:
“Huh? Really? I’m not sure about that. I think we just use regular-old-CAD. We’re on Pro/Engineering.”

FEA Analyst:
“Ok, whatever. Our design guys do that stuff for me, and they’re damn good at it.”

Designer:
“Well, duh, all 3D CAD is parametric. That’s kinda the point.”

History-based CAD is so prevalent that most people in industry have no idea what the “history-based” part even means. “It’s just CAD, right?”

So, let’s chip away at the semantics. History-based refers to the set of steps most CAD tools use to define and create 3D objects on your screen. This works a bit like MS Excel macros. Every step the CAD Designer used to create that object gets “replayed” (in order) to redraw or “regenerate” the object each time you want to see it.

This term “design intent” gets tossed around a lot in CAD. In a history-based tool, the Designer needs to carefully think through how the object should be constructed before he begins constructing it. He needs to envision what changes people might want to make to this model in future revisions, and make sure the steps required to regenerate the model allow for those tweaks.

For example, a CAE analyst typically wants to de-feature a CAD model prior to simulation. He may want to simply remove an unnecessary hole… but if the next 2 steps in the “macro” of creating the part reference that hole, all hell breaks loose.

The model will “fail to regenerate.” The process of killing holes, fillets, and other noogies unnecessary for simulation can become a grueling game of JENGA. Remove the wrong piece in the wrong order, and the tower falls.

Direct modeling (AKA history-free modeling) offers a beautiful alternative for CAE analysts needing to prep manufacturing-ready models. Direct modeling doesn’t replay a macro to display a part on the screen. It just displays the part in its current state. You are free to slice, dice, move, push, pull, and manipulate the model with no fear of regeneration failures. All the unaffected parts stay where they are supposed to on the screen.

In this rough analogy, it’s a bit like playing magnetic JENGA in outer space.

After realizing just how much time we were devoting to our heroin-like dependence on Tivo, we pulled the plug and went to rabbit ears. It’s been great. Our 3 young kids adjusted quickly and HDTV broadcast over-the-air is stunning.

Our main home PC is in the living room corner opposite our 44″ Plasma TV home theater system. Being the technophile that I am, I demanded we have dual 19″ LCD screens. Careful: once you go dual, you’ll never be satisfied with a single monitor again!

We have a Netflix subscription and recently heard they were offering a huge selection of streaming HD videos at no extra cost. Heidi and I watched one movie on the PC and quickly realized that was no fun. Movies are best enjoyed on 44″ Plasma TV screens… not 19″ PC screens!

After a little investigation, we discovered that several devices are out which can connect your Netflix account directly to your HDTV. The $99 Roku Player got my attention first because most of the other options were Blu-Ray players running at over $250.

I was about to pull the trigger when this thought suddenly occurred to me: “Hey, I wonder if there’s a way to connect my PC directly to the Plasma TV?”

A little more digging and I found a way to do it for about $100. This solution is much better because it opens an even larger world of online digital content. There are tons of great HD shows available (with commercials) for free from Hulu, Fox, NBC, CBS, and more. Plus, I can now connect our iTunes music library to the home theater speakers.

Here’s what I bought:

$40 ASUS EAH4350 Video Card
$65 50 foot HDMI cable

The video card features 2 PC monitor outputs and an additional HDMI output. The HDMI output carries HDTV and 5.1 surround sound (on a single cable!) to our HDTV. This allows our HDTV to act as a 2nd monitor… unfortunately the card only allows you to have 2 screens at a time running from the PC. But still, it worked like a charm!

As a side note… Heidi and I were most satisfied with all the movies and TV shows we’d missed. My 6 year old daughter, Emma, however, was most impressed with the fact that she can drive SpaceClaim on “the big TV!” Charlotte (5) and Henry (2) haven’t quite figured SpaceClaim out yet, however, and prefer catching up on the Wiggles and Scooby-Doo.

Got a loved one with Techno-lust? Try this as a gift idea!

I now have about an hour drive from my home in Rhode Island to SpaceClaim HQ in Concord, MA. Turns out, I’m actually enjoying the ride thanks to my iPod! The selection of free podcasts on iTunes is astounding. At any given time, I’m subscribed to about 30 shows. I listen to them until I get bored, and then give ‘em the boot to go discover something new.

If Santa is bringing you into the Apple family this year, congrats!
Here are my current top 10 podcasts to get you started:

1. Wall Street Journal This Morning (easily my favorite)
2. Dave Ramsey
3. Stuff you missed in history class (found this by accident… wow!)
4. Joel Osteen
5. TedTalks
6. NPR Fresh Air
7. PRI’s The World: Technology
8. Saddleback Church: DriveTime Devotionals
9. 60 Minutes Podcast
   (takes far less than 60 minutes w/ no commercials!)
10. Buzz Report

While CAD experts can rightly argue that the new generation programs don’t do everything that the old generation programs do, that’s not really the point is it?

Imagine that a company has 100 engineers and designers, of which only a small percentage (choose any number you want) are truly proficient using existing CAD tools. How might it affect the company’s overall product development productivity to give the rest of the engineers and designers new generation CAD tools?

Read Evan’s full article here.

This will not be a perfect analogy.
It’s just a collection of random, navel-gazing thoughts.

In the late 90s, I participated in the skunk works of GM’s electric vehicle program. I’ve never been one to hug trees, but that experience really opened my eyes to global environmental issues as well as domestic dependence on foreign oil.

But the coolest part was driving the world’s first production electric vehicle, the EV1! Some trivia: the original name of the EV1 was the “Impact.” The marketing folks later realized that might not be the best name for a vehicle.

We wanted to log real world miles on these things, so employees were encouraged to regularly sign out one of our handful of prototypes for a night or weekend. So, I got to take this rocket for a ride several times.

Let me tell you, this thing could accelerate! At the time, I drove an ill-advised Porsche 944 Turbo… but that’s a story for another day. Point is, an electric vehicle gives you instant torque. My Porsche didn’t stand a chance off the line. The EV1 got lots of attention on the open road, too. People would literally follow me for miles back to my apartment to ask about it!

I learned an interesting fact during my time in the EV trenches: EVs weren’t a new idea. Cars powered by electricity actually predated those driven by internal combustion engines. In the early 1900s, there was a mix of gas, steam, and electric vehicles. EVs only started to die out as our cities became more connected by roads. The IC engine could better handle the required long distances that came with this infrastructure.

By the way, I’m not nearly as smart as that last paragraph might suggest. I lifted some of it from this Wikipedia entry.

Now for the “potential” parallel that popped into my head:

When 98% of the people involved in R&D today think about CAD, they are thinking about history-based, parametric modeling ala Pro/Engineer, SolidWorks, CATIA, Inventor, and NX. Few people in industry are even aware of direct modeling. However, both have been around for ages… software-years are roughly equivalent to dog-years, by the way.

Just as the idea of electric vehicles has recently re-entered the public consciousness, so will we see a rapid growth in direct modeling awareness over the next 12 months. It’s already happening, actually.

One of the few major exceptions to the history-based, parametric rule in modern CAD times was SolidDesigner. SolidDesigner later became CoCreate and was purchased by PTC in 2007. PTC’s Pro/Engineer pretty well fathered or influenced all the other CAD tools to follow. So, I’d say that’s a significant nod to the benefits of direct modeling!

Today, Siemens is pouring mountains of marketing cash into the category.  It’s latest version of SolidEdge added some direct modeling capabilities under the brand-name “Synchronous Technology.” Even more startling, the world’s largest CAD company, Autodesk, recently unveiled a prototype product based on direct modeling techniques.

Direct Modeling is here. And very shortly, everyone will be aware of it.

What about those hybrids?

I predict we’ll see every major CAD company race towards direct modeling. Most will take the Siemens approach of grafting it on top of their existing history-based product. Siemens even seems to embrace this term “hybrid modeling environment.”

Opinions: everyone has them.
Watch out, I’m about to show some of mine.

Hybrid vehicles are stupid. It’s easy to produce an EV with a 150 mile, single-charge range today. That probably covers the needs of 80% of the driving public. For me, the beauty of an EV lies largely in the minimal number of moving parts and simplicity. There is very little to service when compared to all the friction, explosions, and linkages in an internal combustion engine. A hybrid vehicle has to somehow harmoniously marry both technologies within a single chassis… talk about two cats in a bag!

Because I’m a competitive guy, I really wanted to find a way to make this analogy blow a hole in Siemens’ new boat. I gotta be honest, though: that would be dishonest.

I don’t anticipate an immediate war between direct and history-based parametric modeling. There are real advantages to both. History-based tools will continue to be popular with detail design and drafting specialists.

These are skilled professionals who draft for a living. They have mastered their CAD tools and are ultimately responsible for compliance with detailed design standards. I applaud Siemens and Autodesk for adding some direct modeling capabilities to make parts of the job easier for these CAD gurus.

I like our positioning here at SpaceClaim, though. SpaceClaim isn’t really built for detail CAD design specialists. (Here’s where I can make the parallel work again.) It’s like the simplicity of a pure electric vehicle. It fits the needs of the 80% of folks on the product development team who can’t or won’t climb the learning curve of a traditional, history-based CAD tool.

Direct modeling is here to stay.

It will become a vital feature of hybrid CAD modeling environments for full-time designers. They will see it as a welcome addition to their toolkit. But, they’ll put the headphones on and rock out while doing what they’ve always done for a living.

I’m personally more excited about the impact direct modeling will have on people outside the design department. FEA and CFD specialists who currently struggle with traditional CAD will easily adopt a pure direct modeling tool like SpaceClaim. Same goes for guys on the CAM side. Same goes for those artsy thinkers who just need to work out some conceptual thoughts in 3D before elevating an idea to the detail design department. Imagine the productivity gains this opens in all new areas!

A final, unrelated rant:

Let me conclude by showing more of my, ummm, opinion:

Vehicles

• Hybrids are too complex and will prolong the use of fossil fuels.
• Don’t talk to me about fuel cells. Fuel cells require hydrogen. Hydrogen is costly to produce.
  (Please consider the full energy balance!)
• Pure EVs will win for 80% of local transportation.

Power

• Fossil fuels gotta go.
• Wind power ain’t gonna cut it.
• Wave power ain’t gonna cut it.
• Solar power probably ain’t gonna cut it.
• Bring on the Nukes, baby! Seriously.

“This year, our enhanced solver is 20% faster! Buy now!”

That stuff doesn’t hold my attention unless we’re talking about a 10x or better speedup. If you want faster performance today, you can simply go out and buy a faster PC. A $6k investment in new hardware will probably get you 20% improvement in solve times.

My real problem with all this solver speed talk, however, is that it redirects customer attention from the important costs buried in the CAE life cycle.

Who cares if a big, non-linear FEA run takes 6 hours instead of 4?
The engineer behind that simulation is off doing other productive work during that time. It’s “hands off” during the number crunching phase!

The most costly part of CAE happens during the manual model prep and simulation setup!

If Joe Engineer spends 5 hours translating and repairing the CAD geometry, 3 hours creating a mesh, and an hour applying the proper boundary conditions… that’s where I want to see CAE vendors focus!

It’s tough for CAE companies to offer improvements on the geometry side.
I give ‘em lots of credit when they do, but there haven’t been many earth shattering advances on that front in awhile.

Luckily, you can largely attack the wasted geometry man-hours by giving your CAE engineers access to direct modeling tools such as SpaceClaim. Most of these hours in traditional CAD environments (Pro/E, SolidWorks, CATIA, etc) are eaten up fighting complex, history-based feature trees.

When I read a new marketing blurb on FEA tool X or CFD tool Y in Desktop Engineering, I’m looking for word of enhancements to the manual aspects of meshing and model setup. Save me 20% on that phase, and we are saving the most precious kind of time: man-hours.

First things first… what the hell is direct-modeling?

Your car, dishwasher, mouse, stereo, chair, lawnmower, TV, and toilet all started life as 3D virtual models on some designer’s workstation. That designer most likely created the model with a feature-based, parametric modeling tool such as Pro/Engineer, CATIA, SolidWorks, Solid Edge, UG (NX), or Inventor.

The models created by these tools are difficult to modify by anyone except the designer who created them. Features are tied to sketches and other features in a rigid hierarchy that needs to be well thought out prior to building the model. The designers creating these models are CAD experts and drive these design tools for a living. So, what happens when they hand this model off to a CAE engineer?

Pain
Well, the CAE engineer rarely has the expertise of a full-time designer when it comes to CAD. The first thing they’ll want to do is de-feature the CAD model, removing small features such as holes and fillets. At best, those elements are typically unnecessary for CAE simulation. At worst, they wreak havoc on the meshing process and solve times. It is at this point when CAE engineers begin the pulling of hair and gnashing of teeth.

Turns out, the de-featuring process can be brutal in this scenario. The engineer has no idea how the designer laid out this model. Nor does she have a designer’s deep CAD expertise, experience, or training. The process of removing parts and suppressing unwanted features often “breaks” the model resulting in the infamous “regeneration failure” message.

Relief
Direct-modeling avoids this horror by stripping out all of the painful, relational feature history. The engineer is able to directly and precisely edit any piece of the geometry without fear of destroying the rest of the model. Most engineers exposed to this style of modeling for CAE say it gives them the freedom to “slice & dice” at will.

How the heck is that possible? Sounds too good to be true. Well, that’s what I’ll be exploring in the next few blog posts. Meantime, I recommend you install the trial version and experience it for yourself.

Oh… and did I mention, you can even edit and modify “dumb” models imported as IGES and STEP? Stay tuned.

Blue Ridge Numerics is an incredible company with awesome development prowess and a deep concern for the customer. More important than any of that, however, are the people. Back home, we used to say an honorable guy was “good people.” Blue Ridge Numerics is good people.

I thought it fitting to leave by letting one of the best men I know give an overview of CFdesign and upfront CFD on my blog:

I drew the line at Twitter, though. Firstly, because the name sounds distinctly unmanly. Secondly, because I thought it was just a bunch of kids IMing the fact that they just ate a sandwich or lost a favorite XBOX game.

After enough high praise from Derrek, I gave it a try. The jury is still out for me, but I like what I see initially. Twitter is a great way to get plugged into like minded industry folks to whom you’d otherwise not have much access.

If you like CAD & CAE, Josh Mings does a much better job of explaining the benefits here: 90+ SolidWorks and 3D CAD Users to Follow on Twitter