Ray Interviews SpaceClaim’s Blake Courter about their success with CAE customers

06 JUNE 2011: [The Bold questions are from Ray. The not bold replies are from Blake.]

Blake Courter

I wanted to explore the marketing thrust you have about improving product development time in CAE analysis by using SpaceClaim.

Sure!

It looks like it’s a pretty interesting idea, actually. So maybe you could tell me a little bit about what it is, how it works and what users need to make it work.

Well, CAE is one of our biggest markets. Not because SpaceClaim itself does simulation, the way people expect CAE tools to, but we serve CAE users’ geometry needs in a way that I don’t think any other product does. And we do that in two ways.

If you just start thinking about it for a second — the profile of someone that does a lot of simulation — you’re typically talking about someone who thinks about physics a lot. Part of that is really caring about the quality of geometry they use to create their meshes for FEA, for CFD; you name it. It’s going to be different for every study.

There are two workflows we see the most. One is when simulation users unfortunately have to do simulation after the CAD models have been made — which is usually a bit more work. And then there are the cases where companies use simulation before CAD to drive product development. That’s obviously a much more efficient strategy.

But I’ll start with the latter, because it happens all too often.

Ray: You know, of course, that post-CAD simulation is almost always required to verify the design. Right?

Blake: It can be a validation requirement.

Once the detail designs are done, they normally have so much junky geometry on them that’s really designed for the manufacturing specifications that it’s irrelevant to the engineering analysis. Rounds, small features and often the parts are more complicated.

Now we know that traditional CAD does a great job at producing those manufacturing-perfect models for machining, and represent how the part will actually be. But those are usually unsatisfactory for doing engineering. They’re just too darned complicated; the meshes blow up and there are other problems, as well.

So simulation users are faced with the challenge of reverting back to the engineering model from the manufacturing model that’s created in CAD. We’ve made SpaceClaim with help from our friends at ANSYS, as well as other vendors, to make SpaceClaim the best tool to turn the manufacturing model back into the engineering model, so it’s the right level of complexity for whatever simulation engineers want to do.

We have two whole toolbars of features for cleaning up dirty geometry and for extracting data. One good example of that is extracting mid-surfaces, which is often a much more efficient way to do simulation than a solid type of meshing.

Then we also have the ability to use our future-recognition technology to find things like… We can take a beam structure — a truss structure — that is all mitered and has the holes drilled in it and all that complicated stuff documented for manufacturing, and revert that back to actual engineering beam elements; trimming the corners together so they’re real beam nodes. And we extract from that solid geometry the cross-section.

We can automatically detect these and automatically extend them and move them into simulation tool, with all the right properties.

I mean they’re really killer features for people who do simulation.

In regard to these features, when have they been there since? Are they new for 2011? Or have they been there for many releases?

I believe that our first release to take a real focus with this was based on 2009. And we’ve made them better and we’ve added more, every step of the way.

We have a very close partnership with ANSYS, where we provide to them a private-label version of SpaceClaim that they sell to their customers as one of their geometry tools. For us directly and through our resellers, this is also a big market.

There’s just so much pent-up demand for simulation users. Because the CAD tools just aren’t very good at these things. And we’ve made it push-button simple.

So for any simulation user who wants to not only simplify geometry but edit it however they see fit, to do “What if?” studies and optimize a model — to put whatever parameters that want on the model — drive it through closed-loop iterations… These are the types of things that they haven’t been able to do before.

I would think that any CAD user would be able to do these kinds of operations directly on a typical CAD modeling system.

Well, what we hear the most from our simulation customers is that with SpaceClaim, they can actually reuse the CAD geometry, rather than having to remodel it from scratch. Here’s a good example… Let’s say you have something like a typical aerospace bulkhead. The CAD model probably has every little pocket and the bulkhead is probably developed with the same web thickness on the bottom.

On the other hand, a simulation user would typically want to optimize that to make the webs only as thick as necessary, given the stress on the model. So they’re going to want to change the depth of every pocket.

There’s no single parameter they can use in a CAD system to do it. They can hack it with offset features and something like that, but that’s ugly. In SpaceClaim, that’s a really simple thing to do.

Do you have any other examples?

Well, I should make another point, based on your comment. The CAD vendors have been selling and acquiring analysis companies — often lower-end analysis companies — for the past decade and a half, as a way of moving more product through their channels, as an ala carte offering. They all will make some version of a claim that says, “Hey. To do simulation, we’ve made it so easy that you can just use this little wizard and get results.”

It’s absolutely true that users can get results. We find that a lot of the time, without really a strong engineering background, it’s hard to know how accurate the results are. I find it fascinating talking to people whose career is dedicated to simulation when I ask them questions like, “How do you know how accurate your results are?” They’ll tell me about the work they go through to make sure that they’re confident in the system!

There really seem to be two use concept out there. There’s the use concept put forth by the CAD vendors, which is for someone who doesn’t know what the units of stress are — they can do simulation. Or they can do the fluid dynamics without knowing what a Reynolds Number is. There’s an argument that they might say, “You can get insight,” but without knowing the accuracy, I’m not sure what the point is of doing the simulation.

Then you have the other use concept, which is, “I want to have accurate answers.” As the vice-president of engineering, I’m going to hire someone who is a PhD at doing this type of work, to make sure that I know when I turn to this person for results, that person put error bars on it and lets me know how confident we can be that this is going to work. Obviously I’d like to have that person working on this problem as soon as possible, rather than as late as necessary. That’s a very different philosophy.

So organizations that really want to innovate and want to maximize the efficiency of their engineering process are going to invest in tools and training to get the results that they want. Those companies tend to buy the more-dedicated simulation packages, like ANSYS. The ones where someone is going to have dedicated CAE users who need to work with that geometry. That’s predominantly our market.

I just want to explore that for a minute. I’m just trying to get really to the heart of this matter, and I think you’re saying it quite succinctly.

But all the CAD systems can export their models to analysis software. Right? So if you have a user running on, say, CATIA, he could buy external analysis software and run it using his CAD model. Right?

Then I guess the difference here is whether you’re going post-CAD model or pre-CAD models. If you’re running it through for validation, then don’t you want to stick with the design data that’s been already built within CAD? Put some additional features in to simplify the model, and then drive it through CAE analysis? I mean all of these vendors support that kind of meshing and submission to external analysis solvers. Right?

But the added benefit they seem to have is that they’re in the mainline stream of design changes. So if a design change came along, it’d be easy to just rerun the analysis.

Well, that’s theoretically the situation that almost all of our customers are in before they buy SpaceClaim. I can tell you why they buy SpaceClaim. One, because that situation is a nightmare.

You mean it really doesn’t work. Right? Or it doesn’t work well enough. And doesn’t it require the CAE user to be a CAD expert then? Is that what you’re saying, as well?

Well, that’s one of the problems. And typically, the CAE guy is not going to be a CAD expert. Typically, if they’re using the in-house CAD system, CAE users are remodeling things from scratch. Because it’s almost impossible to simplify the model.

More typically, they’re asking the CAD team to do it. Which means every time they want to model a new feature or every time they want one little rib moved, they’re looking at a week-long iteration cycle, just because that’s the nature of it.

You go to the CAD guys and say, “Hey, can you make this change for me?” They’re like, “Yes. We’re busy. We’re in the middle of this other ECO. We’ll get to in a couple of days.”

Then you have to remind them again in a couple of days. Then a day later, they get back to you.

So for situations where the CAE users aren’t able to use the in-house system… And keep in mind that often some companies like a Tier 1 automotive supplier will have three different in-house CAD systems. Those three different CAD teams, where the people are trained on the different systems. Right?

Yes.

There’s no way the simulation user is going to be able to use all three.

Right.

So we can go up and answer a simple question in a half-hour, where otherwise they would’ve had to wait for a week. That’s a huge improvement.

They can go through many more changes and they don’t have to be as selective about what parts they can perform simulation on.

So what happens if they read a CAD model and perform a simulation and run it through optimization and they say, “Aha! This is what I want!” Then how do they get that back into the CAD system?

Well, let’s talk about how they did this before SpaceClaim. So we can understand how bad the problem is.

Okay.

What they’ll typically do… To simplify, let’s try to start with suppress or delete features. That will inevitably, for any real part, cause regeneration failure. Because it’ll delete a little hole. But something’s aligned to that hole. So then a child feature fails. And then the model starts to blow up.

So you can’t really do that.

Typically, the approach to use is to create features that sort of fill in features by creating extrusions and to add material. That isn’t a very elegant way to do it.

Then there’s the other problem.

By the way — let me interrupt you, Blake. Is that true for all the CAD vendors out there, including Solid Works?

Any history-based modeler is going to have the problem that when you suppress a feature, in large part, there’s a pretty good chance that other features are going to depend on that. And you’re not going to be able to suppress the features. This is extremely typical.

Yes. Okay.

The other thing that can happen is when edges get referenced… You see things where the geometry can actually shift because you suppress a feature.

A good example of that is if you suppress a round. Sometimes a dimension that referenced the rounded edge now references the original edge, and a wall moved.

So when you go around suppressing features willy-nilly, it’s hard to be sure that the geometry you end up with afterwards is actually a fair de-featured reflection of the original part. It’s a very risky business. That’s why most CAE users find it’s easier to remodel from scratch.

With SpaceClaim, our de-featuring tools don’t have that problem, and you can read in that complex geometry and get a simplified model without a lot of sweat.

The point I’m trying to make is that if they manage to get that old feature model turned into the new one, the model that they make is going to have so many — “hack-and-stack,” I think is the industry term — features at the end that the CAD guys are going to have to remodel it off the original part, anyhow.

We didn’t even talk about re-parametrizing the model so that it has the right design intent to do their optimization. Which can involve re-constraining. It can be as simple as re-constraining sketches, if you’re lucky. But often, the change you want to make isn’t conveniently found in one sketch. You want to look at it in some other cross-section or you want to create mid-surfaces. There’s so much that goes into it.

We would put forth the idea that I think a lot of systems engineers have come up with — which is that the CAD guys are the ones that are the right professionals to know how to make those changes to correctly reflect the model-base definition correctly. To get managed into the PDM systems and have the right parent-child relationships and external dependencies and design-intent and constraints. All those good things that you can use to build really rich CAD models; if that’s done correctly by the right people. They’re going to use whatever comes out of CAE, regardless, of simply a specification for that.

SpaceClaim makes it much easier, because we give them two things, so they can get back to the CAD team. One, you have a precise, solid model at the end of the day, and you can overlay that in your CAD system with the original, and make the changes that you need to make.

Now it’s not going to have all the features. The rounds are going to be removed and the other things are going to be optimized. But they pull that up right next to it and make sure that they get the changes right.

The other thing is, we have this “track changes mode,” like in Microsoft Word, where it red-lines what’s changed. We can actually save that as a set of PowerPoint slides that show the different views of what changed; including dimensions that say, “This used to be this size; now it’s that size.”

I don’t think I know about this capability. Is this something new?

Oh, it’s awesome! We make these 3D mark-up things that show every face that’s been added, removed or deleted.

And what do you do? Do you do that by comparing the original model and the one after you’ve modified it a certain number of times? How does that work?

Exactly.

I can show it to you. It’s kind of what you’d expect. The things that move get colored, and then we can throw dimensions on it to show “Is this value; was that value.”

So this is something that might be done afterwards to document the CAE changes. Right?

Exactly. We see it used in a lot of other contexts like design reviews.

Well, since you’ve stripped all the nuts out of the model to run it through CAE… a lot of the details… right? Is that still valid?

Well, it was a valid engineering model to do the CAE calculation.

Yes. But compared to the original CAD model, it may be a LOT different. Right? Half the features could be missing. Maybe more.

Well, exactly. Certainly I’ve seen parts that have more round features on them than everything else put together. And all of those are gone.

Obviously there’s going to be some remodeling, but the important thing is that the simulation user can concentrate on doing simulation and getting answers and results, which is how they add value to the organization.

And the CAD team can make the changes correctly once, rather than having to make 4 or 5 different changes once a week of what the CAE user dreamed up. They can make those changes only once, once the CAE user is done, and save a lot of time by doing that remodeling only once, with a really well specified design for it.

Have you got any good productivity stats or info from users, other than generic, “This is x-percent faster?” Anything interesting that’s usable?

Sure. I could just come up with quotes. I can tell you how it averages out. But we have — I don’t know — many, many case studies from simulation users, at this point, to talk about how much faster they are.

Sometimes customers are hesitant to specify the precise detail. You know?

Well, I hold their feet to the fire when we have the opportunity to write up a case-study and get a real metric out of them. I can tell you that there are two pretty common themes, and I’ve already given you some of them.

When the CAE users aren’t using CAD themselves and must rely on the CAD team to make changes, their cycle times go up by a factor of 10. I hear that when they are doing their own 3D, the geometry is 3/4 of the work of getting a part ready for simulation. And that’s the major bottleneck. When we remove that bottleneck, they’re typically able to get twice as many simulations set up in the same time. And assuming they have the compute power, twice as many simulations running in the background at the same time.

Those are consistent. I’ve heard those numbers pretty much from most of the people that do simulation. And you’ll see quotes like that in the case studies.  (http://www.spaceclaim.com/en/Resources/CustomerSuccesses.aspx)

If I want to take this route, what are my options? Obviously, SpaceClaim. By the way, you have only one product, now. SpaceClaim Engineer?

That’s right. We also have variants that we make for folks like ANSYS, and so forth.

Yes. I know they have something called ANSYS SpaceClaim Direct Modeler.

Right.

But here’s my question, then.

If somebody decides to go this route, what are their options in terms of CAE vendors? Can they use their existing ones? I mean let’s suppose that I’m again a CATIA shop. So I could go out to SpaceClaim, but if I already paid for a license of, say, MSC… or NASTRAN… Is it just as easy to use that as ANSYS? What are some of the alternatives for a user?

There are three different paths. The bottom line is that the amount of productivity benefit you get using SpaceClaim to create new concepts for simulation or to prepare models is so much faster. The integration itself is probably the least-important part of the value that we add.

In other words, if we can make it so that you can get a new simulation running in an hour instead of a day, whether it takes a minute to transfer or five minutes to transfer isn’t that big a deal.

But certainly we’ve gone to great lengths to have integrations. There are several vendors with which we have direct integration with SpaceClaim. In addition to ANSYS, Thermal Desktop, COMSOL for multi-physics, EnMesh, and Autodesk CFdesign, to name a few.

So these vendors have made, or I guess you both have made, modifications in your software to make passing down the data easier. Right?

Yes.

Do you support assemblies?

Yes.

Or assemblies within a model of an assembly? How does that work? Does it work like SolidWorks does, where you have an assembly attached with attached models to it? Or what?

Sort of. Our data model’s a little more flexible, so we don’t have a different assembly part or part parts. The other thing is, we don’t have any requirements that every part lives in its own document, the way most feature modelers do — including SolidWorks. Which is a major pain in the butt, when you want to restructure an assembly.

So we just sort of did that right. We have one type of document, and you can insert documents into documents.

When you import assemblies for most CAD systems and from STEP, we do the right thing when it comes to instancing. I downloaded [a bearing] the other day from this company. Anyway, what they ended up sending me a Pro-E assembly. So it was a Pro-E assembly with a couple of parts. And it had instances, but of course with SpaceClaim, each roller bearing in the bearing was an instance-of the same part. We only had to read that in once, of course. So we get all that stuff right.

I was just wondering about… You don’t combine any pricing or anything. Right? There’s no combination. I guess ANSYS might have one, but I don’t know… I’m just trying to get a handle on what this stuff might cost.

It’s reasonable! I mean it’s less expensive than a CAD system, or than most CAD systems, anyhow.

What — the solvers and the meshers? You don’t do any meshing. You have external meshers, right?

Meshing is not a simple thing. It’s really better that meshing be coupled with the simulation code.

Right. So that’s for the CAE guys. SpaceClaim does not provide that – right?

Our goal is to get geometry exactly where you need it to — to your mesher. The folks who really know simulation know how their geometry needs to look to make best use of their meshing and analysis software.

I wanted to finish my last thought about the integrations. I talked about the direct integrations. But to point out the second step that we see a lot —  MDAO tools.  Multidisciplinary Design Analysis and Optimization. These tools integrate with a number of different CAD and CAE systems to allow you to do just that. Some of the ones that are popular today… One is called “Optimus,” made by Noesis (www.noesissolutions.com). There’s another one called modeFRONTIER made by ESTECO (esteco.com). Of course, ANSYS Workbench is another product like this, that mainly integrates with ANSYS products. (ansys.com).

Right.

But with these tools — the Noesis product and the ESTECO product — work with SpaceClaim. So when customers have some special tool or maybe even an in-house tool, and they want to be able to do this multidisciplinary optimization on it, they can do that with SpaceClaim. That’s a way of integrating with many, many, many more CAE tools such as you were mentioning MSC or something from a CAD vendor. Where we may not have a direct partnership, like ABACUS.

So that’s a valuable path there.

How does that path work? You pass the geometry to these applications?

Well, those applications stand at the center and they can send parameters back-and-forth. They can read and interrogate geometry and act as the broker between all these different systems.

They typically have really beautiful visualization tools to map out Pareto frontiers and do sensitivity studies and that kind of thing.

I don’t know all the details of that, but you know, we have one of our case studies that involves someone using a SpaceClaim and Moldflow through Noesis Optimus. We didn’t have a direct integration. But those folks were able to wire it all together so they could have closed-loop optimization.

Well, what interface would you need? Wouldn’t you just export IGES or STEP to Moldflow and let them bang on it from there? Or what?

Absolutely. You can always do that. There’s always some route that way.

But if you want to have the parameters go back-and-forth and do optimizations and figure out what wall thickness you could have to minimize material and still have it pop out of the mold without getting too cool — or something like that… Then you’d want to have the full circle all managed in one environment. That’s what Optimus does.

Focusing on the market, it looks s like you’re having good success. Could you discuss that?

Sure. The answer is, “Yes.” We’re having good success in the market.

We don’t do absolute numbers. But I can tell you that… I probably told you that we’ve been more than tripling our revenue year-over-year for the past two years. Our licenses have been going up probably at around the same rate. There seems to be a lot of pent-up demand.

Do you have any idea what percentage of your sales would be in this market?

I think between quarter to a third, right now. But that’s a guess.

So you’re saying 1/4 and 1/3 of your new licenses are in the CAE analysis area?

Yes. It’s probably less than it used to be. But keep in mind that we have ANSYS selling SpaceClaim, so that’s a big part of it. And there are other CAE vendors who now sell SpaceClaim. So it’s a good market for us.

Yes. I’d say it’s a fantastic market, with those kinds of numbers.

We are also seeing a lot more sales now in concept modeling, bid modeling and manufacturing.

But yes, that is definitely where we found our stride with the Company. Also, it’s our inroad to getting to engineers. If we step back for a second and take a look at what we’re achieving as a company, it’s to make 3D as universal an idea as email or a spreadsheet. I think we can all agree that at some point in the future, every engineer is going to be able to scratch out basic problems in 3D.

As far as I’ve ever seen, SpaceClaim’s the first tool that really lets them do it without having to become real CAD-dedicated draftsmen.

We do this very effectively in large organizations where many engineers just haven’t been able to answer basic questions in 3D, before. And we have customers like Tyco, who have bought into this and deployed SpaceClaim far and wide to enable as many engineers as possible, just to be able to do a little basic thinking in 3D.

We’re not trying to reinvent the wheel. We’re not trying to be another detailed design tool. But we’re keeping SpaceClaim as simple and accessible and affordable as, in a TCO [total cost of ownership]. Affordable enough that SpaceClaim can really be broadly deployed throughout engineering organizations.

Does Tyco and Samsung represent a particular industry segment? Are they both one-up design companies. If so, they would be immediately attracted to a direct modeling system, one would think.

Well if you’re comparing, and if you look at sort of the pre-Creo or PTC rhetoric, they talked this way. They’ve changed their whole story right now, of course.

But it’s true that if you’re buying a really complicated, full-Monty drafting system for detailed design in manufacturing model creation, you might look at a direct model. Or one like the previous generation in a direct modeler. Like CoCreate, to do that, if you’re doing a lot of custom design. It didn’t make sense to invest in all the complexity of the design intent that you’d have to do to wire up a history based modeler correctly, like you would in Pro/E. That’s why you see certain industries have gravitated to direct modeling for their production CAD. And you do see that in the printer business and in some machine-design accounts.

It was probably the right decision not to go with feature-based modeling just because it’s so darned complex. If you’re only making one shape, that’s a reasonable way to go. But, we’re not pursuing that market. It’s true that SpaceClaim is a direct modeler, but the things that differentiate SpaceClaim are its intuitiveness and its low TCO. Now I think one of the things that we can lean on pretty strongly is, we’ve been so focused on doing it well. A big part of SpaceClaim’s simplicity comes from its ability to just work, rather than having to really put a lot of elbow-grease into making edits. I think you’ll find that for making edits to models or even better, editing parts, is something we’ve made much easier.

So although our modeling is industrial strength, the market opportunity for us is empowering engineers to answer questions in the process. And the business benefits that go along with what happens to an engineering organization where people don’t have to become dependent on the CAD team every time they want a basic answer about, “What if this,” or “What if that?”

That’s a real good summary. 

I’m saying some things here that I think could be taken as a little controversial. And I’ve been sort of quoted out of context in talking about things like how CAD users can’t do simulation and something like that. That’s not what I’m trying to say.

But I do think that gets people’s attention, because it’s so different from what they’re hearing from a lot of CAD vendors who have this really self-serving strategy of selling products without thinking about whether they’re the right products for whom they’re selling them to.

I think one of the things we’ve seen in this business is, the CAD vendors have really strong marketing organizations that can say whatever they want. And the folks that really understand engineering tend to just be more focused on getting the job done rather than listening to vendor claims. I try to do what I can to rectify that!

CAD vendors have always had their successes in large integrated, single-team design jobs like automotive and aerospace. Do you see any penetration in those areas?

Our biggest markets are automotive, aerospace and defense. Next would be medical devices.

I’ll bet that that concerns the big vendors!

You know, I’ve heard stories of… Obviously I’ve been in this industry for a while. I have a lot of good friends in CAD companies, and every once in a while, I hear stories of high-blood-pressure moments of certain CAD vendors, because of our presence in their accounts.

I think that’s unfortunate. But they’re used to thinking about CAD in an old-fashioned kind of way.

Yes. Well, they’re all thinking about what other upstart companies that were seemingly point solutions did to them in the past.

Yes, and I understand that. We have a great solution for concept modeling. Everyone hearkens back to when PTC played the concept-modeling card on Computervision and those guys. They did find a different market.

We actually think we have the right tool for this. I haven’t talked to anyone who doesn’t have the vision that in 5 to 10 years, every engineer wouldn’t be doing some basic work in 3D.

What happens if we change vendors? What happens if we change materials? What are we going to do for the next program? Hey — let’s make a little mockup that we can send to our customers to make sure that this is going to work for them. They can try it out in their assembly.

It makes sense. It wouldn’t make sense to be investing in a heavy-duty CAD model for this kind of work. Of course, this is the way people should work. We are the closest that anyone’s come to realizing that vision. That’s why we see these great expansions.

But when you look at our presence in these big, stalwart accounts, you will see us there on the simulation user’s desk. Not the CAD guys, as much. Although some CAD guys do like us for getting the concepts right before they put all the design in to make a real model.

Yes. Well, I’ve been beating on the CAD vendors for years to do this kind of concept modeling. And they don’t either seem to be able to do it or aren’t interested in it.

I’ve seen the mentality in CAD vendors. Take any serious CAD product — any one of the big-4 flagship products. Go to one of their application engineers or someone on their product-management team and says, “Hey! We saw a demo of SpaceClaim. Can you do this change in your software?” It’ll be like, “Yes. Give me a second.” They come back later and it’s like, “See? I can make the change!”

It took a couple minutes of thinking, and someone had to make a new model that was able to exclusively make that change. Because that’s the way feature based modelers work.

And it’s true. You can do absolutely anything in the big CAD systems. But no one dares measure how smart or knowledgeable in that CAD system you had to be to do it. Or, how expert in that CAD system you’d have to be to be able to do it. But once that box is checked off, I think the executives of the CAD company sort of say, “Hey. But we can do that! So we don’t need to do what SpaceClaim can do.”

Then we show them how, with SpaceClaim, with minimal training and with minimal effort, we enable users to answer questions in 3D, when they otherwise wouldn’t have had the experience and training required to do in a CAD system. Then we make a sale.

I think it’s that type of rationalization that leaves this market wide open to us.

Just to editorialize here a bit. As an example, take a look at what PTC appears to be doing with Creo. While I haven’t seen exactly what they’re going to deliver, it seems to me that they’ve changed the names, but as far as I can tell, they are putting forth the idea that you could have feature modeling and direct modeling so that CAD models can go back to Pro/E and survive the round trip, with all the features and parameters intact. If that’s really the vision they’re selling, then they don’t understand what direct modeling’s all about. When you chop up a model into 50 different pieces and rearrange the pieces and then merge them all back together in some unique way, there’s no sensible analogue in a feature modeler for doing that.

Isn’t that what Inventor Fusion does?

Yes, and if you look at reviews from other analysts that tried to do anything other than the most trivial little tweak with Inventor — they said that it didn’t work. Because you’re changing the design intent so fundamentally, there’s no way of putting that back into the old feature model. You need to build a new one.

And again, it’s a nice story, and I can see how some marketing person that hasn’t used CAD for 10 years could cook that up and say, “Hey. This is going to be better than the SpaceClaim version of the Round-Trip story.” But benchmark it! Try it on a real-world design change, and it falls apart.

I think that’s unfortunate, in summary, because it creates a chilling effect against direct modeling. But luckily, most of the people who buy SpaceClaim are smart enough to take a look at the big picture and say, “Wait a minute. We’re able to do an iteration in under a day, where we used to do five iterations before, and each one took about a week. Oh, my goodness! Isn’t this a huge process improvement even if it involves a little remodeling.” Not that the competition had anything other than remodeling to offer in the first place.

Yes. Interesting. Okay.

But yes, there’s certainly a lot of FUD out there. All I can say is, “This is the year!” There’s been a lot of talk, but PTC’s cards are on the table and they’re about to turn them over at their user event.

Autodesk has said where they’re going. I think they’re probably the sanest of all the CAD vendors in terms of product strategy and what they’re doing with Fusion and the bundling.

The other guys have some sort of strategy, as well. The Siemens guys have said that for them, synchronous technology isn’t about growing the market, but just about providing direct-modeling tools to their CAD users. Which I think is shockingly honest.

Well actually, I was at the Siemens conference a few weeks ago. I said to a number of the people who really knew what the technology was… “Tell me more about how synchronous technology enables direct modeling.” They, to a man, said, “It’s not direct modeling, at all.”

Hmm. One of the things that I really respect about the Siemens guys is, they are really honest and practical. Unlike some of the other companies, they’re not going to succumb to ridiculous marketing hype. And they understand their customers and they’re going to give them good tools.

I think that’s part of the reason why we see NX customers, in particular, so enchanted with SpaceClaim.

I think another thing, to Siemens’ credit, is that they’ve done a wonderful job with JT. They’ve really made it open. JT and Teamcenter work incredibly well together. I see it in a lot of different accounts. Even if NX isn’t the CAD system of choice, of course, SpaceClaim is a beautiful JT editor. We can read and write JT natively with Teamcenter.

So Siemens has created a beautiful ecosystem that’s created, more than any other vendor, a level playing field. And I think that’s why we have so many very happy mutual customers.

Interesting.

Because many users will be importing CAD models, it seems you have a bevy of data translators. Right?

Yes. We offer separate bundles. They work really well. They’re a technology that we developed with our partners. They range from about $550 for a package of some of the more basic ones like SolidWorks and Pro/E and range up to $1200, with the top price for a CATIA translator that can read from and write to CATIA. SpaceClaim includes a lot of translators, such as STEP, IGES, ANSYS, Rhino, PDF, and ECAD. You can then add on others, such as Data-Exchange Package 1, which includes Pro/E, Inventor, CATIA V4, and VDA, for only $555. Data-Exchange Package 2 include Parasolid, SolidWorks, and NX, priced the same.

These are very reasonable prices.

It looks like SpaceClaim is re-defining portions of the engineering market. It’s definitely a different approach. Congratulations. It sounds like things are going well for SpaceClaim.

Yes. We’re happy! We’re having fun, and it’s probably the most interesting year the industry has seen in some time. So we’ll see how it shakes out.

Blake, thank you so much for being so cooperative and responsive. I really appreciate it.

It’s always a pleasure. Anytime.

Blake Courter is a co-founder of SpaceClaim Corporation (www.spaceclaim.com), where he helps product development organizations make 3D more accessible to all engineers. Blake started his career at PTC, where he held a range of product management and business development positions. He received a Bachelor’s degree in Mechanical Engineering from Princeton University in 1996.

Notes from Siemens PLM Connection 2011

A few weeks ago I attended the Siemens PLM Connection 2011 conference in Las Vegas. Since returning I have been busy with all kinds of matters, but I wanted to get some of my thoughts on the conference to you, my readers. Siemens promised me copies of the slide presentations, which I have recently received. As any of you who have attended similar conferences, the slides go by so fast that there is little time to write the key points down.

Arriving at the hotel on Sunday, May 1st, the event started with a small cocktail party of the media with some Siemens executives. I had a chance to meet with a briefly speak with some executives before going to my room exhausted from traveling and standing on my feet for three hours. Among the execs I had a chance to speak with were Eric Sterling, Dave Shook, Kris Kasprzak, Tony Affuso, and Dan Staples. The mood was decidedly upbeat and all were excited about their performance the previous year and so far in 2011. I was unable to get specific details, as is Siemens’ custom.

At the kickoff keynote the next morning Dave Shook, Sr. VP Americas, opened the session and introduced keynote speakers from Microsoft and IBM. These were the usual pitches from cooperating vendors. Instead both companies discussed interesting technology challenges; Microsoft discussing how unstructured data can be managed, and IBM discussing the challenge of hyper digitization and the need to close an innovation gap with the velocity of change accelerating.

Tony Affuso, Chairman of Siemens PLM followed with highlights of the company’s year. Basically Siemens PLM is on track after five quarters of growth, now with 68,500 customers, 7.2 million seats, and double-digit license revenue growth. This compares to Dassault Systemes (DS) real growth last year of 7%, less the temporary revenue bump from the IBM PLM acquisition. The company has a good backlog and is executing across the board and winning substantial accounts. Following Tony was Claus Oesterschulze, describing Siemens extensive effort to internalize the use of their own software (NX, Teamcenter, and associated apps). No easy task, as in most customers, organizational changes are required along with process changes. Convinced that “IT is a big lever for business” Siemens is aggressively moving ahead and focusing on how to manage complexity. Below are a few slide from the Oesterschulze presentation.

 

Later in the morning Chuck Grindstaff, President and CTO of Siemens PLM Software presented his “technology vision.” He viewed where they are today in fully integrating Teamcenter across an enterprise to fulfill its PLM needs: view models fully and anywhere, check a model against its requirements and be able to trace these on the products; review and initiate simulations against specs directly in the CAD environment and to ask highly complex questions; ability to optimize the design for performance including cost and sustainability; using Tecnomatix to insure that manufacturing has correct models, processes and allowable variations thus enabling what-if manufacturing studies. Thus his focus on a fully integrated system, much of which is already in place today, with more coming shortly, as evidenced by the NX, Teamcenter, and Tecnomatix presentation we saw later in the week. Grindstaff feels they are far ahead of their competitors and that customers can now see that. A strong point in his direction is the ability to implement continuous changes in the product rather that the discontinuous changes that the competition [Dassault Systemes] makes. An interesting point about his thinking was his comment that “If we [Siemens PLM] can formulate the value proposition correctly then customers will listen.” His key investment areas are in intelligently integrated information, continuing their future proof architecture using SOA and XML that isolates and allows ready integration of new software, expanding their TC HD experience, systems engineering expansion, integration of domains such as MCAD, ECAD, software, plants, etc. and continued openness.

Joan Hirsch, VP of NX Products, and Paul brown, et al reviewed their thinking about the future of NX. Some of their goals include: building an effortless UI, improving complex product visibility by allowing viewing of multiple data sources and using HD3D visual reporting, integrating multiple design disciplines together, allowing front loading of best practices and knowledge, and requirements management and validation. The chart below summaries much of what they discussed.

What I found most impressive was their emphasis on non-disruptive improvements and Siemens focus on using making HD3D easy to use to “bring to life” the enormous reserves of data stored within TC. This is their 4th release of synchronous technology (ST) within NX and strides continue to be made in its use, some of which I hope to explore in more detail in an upcoming report.

Later Steve Bashada, VP of Teamcenter Products and Bill Boswell, Director, gave a Teamcenter update. New versions are due next February, with a “more visionary” version due next September 2012. Bashada viewed their key investment areas as the following: systems engineering, corporate social networks, cacheless search, massive model viewing, HD-PLM extensions, and thin client access. A Teamcenter mobility app for the iPad was announced. I downloaded it, but have yet to try it out, because I need to sign into their TC central demo app and database. I may report on this later. Think it’s complicated? Take a look at their portfolio list below.

Getting tired from furiously scribbling notes, I then attended a Velocity business update starring Karsten Newbury, SVP and GM of the unit. Newbury discussed their business momentum, noting that in 2010, their growth in licenses was 30%, with Solid Edge (SE) accounting for 50% of that. He is investing in an expanded presence by adding resources [people], channel enablement [his words], and a focus on community for a bigger academic presence and mode direct feedback events. ST3 was well received due to its ability to support both ordered and synchronous approaches.

Mario Joyal, of Matritech, a small company in Quebec, described his results using SE with ST3. A recent design took 50% less time and other users found it easy to use. Kris Kasprzak, product manager for SE, described ST4 which is due to release on June 15. We are allowed to discuss some info about the upcoming release of ST4. New tools and functions coming include those for: advanced machine design, expanded collaboration, simulation for sheet metal, and improvements to their already impressive drafting.

After a delightful dinner and cocktails for the press and analysts Monday evening on the top floor of the Rio hotel, we finally ended the day.

On Tuesday we heard a digital manufacturing (DM) strategy update from Ziyon Amran, VP of digital manufacturing software, followed by an amazing presentation by Gene Coffman of how Ford performs virtual manufacturing. Siemens still leads the industry in revenue by a wide margin, as shown in this chart.

Key manufacturing technology domains include assembly planning and validation, robotics and automation planning, part manufacturing, and plant design and optimization. Amran discussed many of the new capabilities in each of these areas. Amran summarized what Tecnomatix offers their customers

  1. Breadth and depth of offering addressing all DM functional Areas
  2. Focused industry and domain Solutions
  3. Integrated Knowledge Management through single source of Product & Manufacturing data as part of the Teamcenter PLM platform
  4. Advanced technologies for Machining, Robotics and Human Simulation and for Enterprise Process Authoring

Coffman’s presentation focused on how Ford makes virtual manufacturing work at Ford (it isn’t easy) and how it contributes to Ford’s profitability and design to manufacturing cycle reduction. Here is one example of the continuing benefits shown by Coffman.

And finally, to wrap things up I met with Al Hufstetler, VP Quality Planning and Validation, who took me to task on a comment I made in my blog about the DS acquisition of Intercim. I stated that “In shop floor analytics DS now has the edge.” Hufstetler pointed out that Intercim’s solution uses analytics to isolate issues, whereas, Siemens solution uses a better solution, a feedback system that not only can detect quality issues, but can actively correct them. For more details on this, contact Siemens PLM Software.

www.siemens.com/PLM

Disclosure: Siemens paid for my hotel and conference admission. TechniCom paid my other travel expenses.

TechniCom test-Part 8 shows how Inventor and SolidWorks compare for Mechatronics

Mechatronics

This blog series and the tests reported herein is designed to show some of the key differences between Autodesk Inventor Professional 2011 and SolidWorks Premium 2011 for digital prototyping workflows. This final part of our 8 part blog series examines Mechatronics – the ability to perform cable and harness design in an existing design from an imported electrical wiring diagram.

We test the ability of the mechanical CAD system (MCAD) to leverage data from an electrical CAD system (ECAD). The ECAD system specifies the appropriate connectors, wires, and their connection points while the MCAD system specifies the physical location of those wires and connectors within a product.

Electrical schematic to be imported into mechanical assembly on the right

Autodesk supplied an Inventor video of their solution, a net list in Excel format, a STEP file of the enclosure assembly, and a schematic drawing (.dwg) of the connections.

What’s Important in Mechatronics Design

  • Leverage the data stored in schematic drawing files to design wire harnesses in the mechanical system. Such data can be stored exported from an electrical design file using various techniques. At its most basic, the electrical design software sends a net list to the mechanical package containing connector information for each wire, wire types, and a list of pin-to-pin connections.
  • Generate correct wire lengths
  • Generate output to enable manufacturing of the wire harness
  • Not tested were two-way associativity between the electrical and mechanical software, nor were any tests designed to simulate electromechanical interconnections such as activating switches or sensors based on mechanical actions.

Autodesk supplied us with an Inventor video of their solution, a net list in Excel format, a STEP file of the enclosure assembly, and a schematic drawing (.dwg) of the connections.

What we found out

The two software packages (Inventor and SolidWorks) are comparable. Inventor has a tight connection to AutoCAD Electrical with the xml file transfer. SolidWorks has similar tight coupling with some third party software such as Zuken’s E3. Both systems use added cost electrical software to generate the net-list. SolidWorks was not able to read the AutoCAD Electrical generated xml list, and instead used an Excel file with similar data that needed manual cleanup in Excel.

It appears that there are a few more interactions with SolidWorks, but this may be due to the operator-preferred method. Both systems effectively produced the required output. There appears to be no real operational advantage to either package when used with tightly integrated electrical schematics software. Since AutoCAD Electrical is one of the most widely used electrical schematic packages, the advantage goes to Inventor.

Observations

For this test, on the AutoCAD side, AutoCAD Electrical exports an XML file to Inventor. Inventor reads this file and generates the 3D wiring and, under user control, assigns wires to cables. It can then generates wire lengths, a flat wire harness diagram and a pin board for manufacturing.

Inventor opens the 3D model and then the xml file of the net-list from AutoCAD Electrical. This designates the pin-to-pin connections where the wires are to be placed. Different than SolidWorks, the Inventor user placed the harnesses in anticipation of the wiring to be imported. The wire import could also have been done first, as seen in the SolidWorks video. The names of the connectors and the number of pins on each connector are stored in coordinated libraries in both the electrical and mechanical systems.

Importing the wires in Inventor

Importing the wires in SolidWorks

After the import, the imported wires appear as direct point-to-point connections between the pins without using any harnesses. 19 wires were imported and identified as un-routed. Then Inventor asks for an auto-route of all un-routed wires. It then places all 19 wires into the predesigned harness, we guess by using closest entry and exit points. Then Inventor builds (and reports) a pin board payout of the harness showing the 3D derived wire lengths. The video below shows an Inventor user performing the test. 

SolidWorks takes a slightly different, albeit very similar approach. After importing the net-list, the operator builds a 3D representation of the harness and then places the wires into the harness, with the software computing the wire lengths. This took more manual interaction than the Inventor solution, but yielded the same end result. The video below shows a SolidWorks user performing the test. 

This is the final blog in this series. Users can review a summary of these tests, published as Part 1 of this series by clicking here. We have also published a pdf file of the complete report here. The pdf file does not contain any videos. To see them you have to revisit this blog series at raykurland.com.

About the author

Raymond Kurland is president of TechniCom Group LLC and its principal consultant and editor. His firm, founded in 1989, specializes in analyzing MCAD and PLM systems and has been involved in reviewing and comparing such software since 1987. Ray frequently consults with both vendors and users. Ray has degrees in Engineering from Rutgers University and from NYU. His career included stints with Bell Telephone Laboratories, IBM, and Dassault Systemes. Ray can be reached at rayk@technicom.com.

For more information about TechniCom Group and other software reviews please visit http://www.cad‑portal.com and Ray’s blog at www.raykurland.com. You can also follow Ray on twitter using the id technicom.

——

TechniCom tests Part 7 reveal key differences between Inventor and SolidWorks Design Automation Solutions

Design Automation

Continuing on with part 7 of our 8 part blog series dedicated to showing the differences between Autodesk Inventor Professional 2011 and SolidWorks Premium 2011 for digital prototyping workflows, we examine the ability to automate the design process by automating the creation of drawings for part families, creation of parts from parameters, and creating copies of an assembly constrained along a variable path.

This test looks at simplified automation examples, yet it provides a glimpse of this capability in both products.

  • Create a simple piece of stock lumber (2×4 board) and examine how a user can make that same part file represent several variations of lumber that could be used in a project.
  • Automate the variation of individual drawing views, scales, and annotations.
  • Automate assembly variations that vary by size and position.

Autodesk provided us with three movies showing Inventor completing the tasks. They also provided three STEP files of the frame, the assembly, and the curves to follow for the frame assembly resizing.

Key differences you will see in this test

SolidWorks has some of the same capabilities built into the core modeling system and by using DriveWorks Xpress, a 3rd party add-in delivered with SolidWorks. It can handle configuring a part when placed into an assembly, but updating it in the part model on the fly is not possible. It was able to create the multiple configurations of an assembly – although it required more steps than Inventor to achieve the same solution. SolidWorks Premium was unable to automate the creation of drawings for part families, which requires users to go through the manual process of creating a drawing for each instance of the family. SolidWorks’ third party partner, DriveWorks, offers software that can perform this process, although at additional cost. The no-charge version was not able to control the final drawings, as desired. We did not evaluate DriveWorks, although the extra cost versions of DriveWorks Solo and DriveWorks Pro appear able to perform this task, again, at added cost.

Autodesk Inventor includes iLogic and iCopy technologies that use rules to control the parameters of the part, assembly, or drawing and these capabilities were used to complete this test. Inventor created the lumber workflow, the frame resizing and the drawing scaling without flaws.

What’s Important for Design Automation

  • Engineers can capture design intelligence by using rules to embed design intelligence into parts, assemblies, and even drawings
  • Such design intelligence, in the case where repetitive designs or portions of repetitive designs are used, can radically reduce design time and produce more repeatable results.
  • What techniques are used to build the design intelligence (often programmatic)
  • How easy is it used to create new designs once the rules have been built
  • How such design intelligence is accessible and how it can be maintained in the future

Observations

Autodesk Inventor includes iLogic and iCopy technologies that use rules to control the parameters of the part, assembly, or drawing. Inventor controls the parameters of a part through a single dialog box that updates the model on the fly. It is also able to automate the process of creating unique assembly configurations by modifying the parts and sub-assemblies automatically for the user. iLogic also can use rules to drive drawings – from view placement to scales and annotations – for a family of parts or assemblies, which can save significant amounts of time in large scale projects.

SolidWorks has some of the same capabilities built into the modeler. It can handle configuring a part when placed into an assembly, but updating it in the part model on the fly is not possible. It was also able to create the multiple configurations of an assembly – although it required more steps. Without using extra cost third party software, SolidWorks is unable to automate the creation of drawings for part families, which requires users to create a drawing for each instance of the family.

For the stock lumber workflow, both Inventor and SolidWorks were able to capture all the variations within a single part file.

A family of parts or assemblies also requires a family of drawings to document their design intent. Recreating essentially the same drawing, which only varies by a few critical dimensions wastes time and effort. Inventor allows the user to easily automate drawings using iLogic functionality. Inventor drawings can be set up to automatically vary view placement, scale, and annotations for a family of parts or assemblies. SolidWorks, without extra cost third party software, is unable to automate the creation of drawings for part families.

Creating copies of a frame along a path using Inventor

Creating copies of a frame along a path using SolidWorks

For the frame variation example, Inventor allows the user to automate complete assemblies that vary along specified paths. SolidWorks was able to manually model the frames along a path, but took substantially longer.

See how an Inventor user solved the problems in this 3 part video series:

  

See how a SolidWorks user solved the problems in this 2 part video series – automating the drawing was unable to be done without additional cost software:

 

—-

The next, and final, blog in this series will examine Mechatronics – the ability to perform cable and harness design from an imported electrical wiring diagram. Stay tuned or sign up to be notified of my blog updates.

—-

Ray Kurland interviews Blake Courter, co-founder of SpaceClaim, about the LG Electronics win

On 22 Feb 2011 SpaceClaim announced that LG Electronics “is optimizing their engineering processes by using SpaceClaim to modify and manipulate CAD models.” This piqued my interest because LG Electronics is one of the largest electronics consumer companies in the world. I immediately placed a call to Blake Courter, co-founder of SpaceClaim, to find out more details. I was warned by their PR firm that they could probably not discuss any more details about the win, but I went ahead with the interview anyway to get Courter’s opinion of what drove the win and where SpaceClaim is finding success.

Q. SpaceClaim seems to have invented, or at least popularized direct modeling. Now everyone is trying to get in the game.

A. SpaceClaim did not invent direct modeling. It predated feature based modeling. We just realized that if we were going to get the rest of the world using CAD, it was not going to be with feature based CAD. We recognized that there was a market for the right tools [direct modeling] for specific users in specific markets and that was the way to go.

Direct modeling is the right way for most of the world to do 3D. Clearly the vendors have bought into this. Every one has a direct modeling strategy.

It’s great for all users to have these capabilities that are much easier to use than a feature based CAD system. For engineers doing certain jobs that do not require a history based or feature based system, such as simulation, manufacturing, sales engineers doing bid modeling, we have the best tool for the job. Rational companies are using SpaceClaim because we figured out how to make the right tool for this class of users. We are looking forward to having the opportunity to have that conversation with more customers as more users are introduced to direct modeling technology. This will be great for us.

Q. Clearly the market is agreeing with you and the vendors as well, considering that they are all introducing products in this area. Why did LG select SpaceClaim and where in their business do they intend to use it?

A. Can’t speak to anything outside the details shown in the press release, but I understand that they are using it for general use cases around manufacturing model preparation and manufacturing engineering.

Q. Was it a reseller closing the sale? Why can you not discuss any more details?

A. It was a direct sale. LG only agreed to let SpaceClaim disclose what was stated in the release.

Q. What is their current CAD system?

A. They have a CAD system; I think they have several and obviously have had and used CAD systems for a long time. For certain use cases they claim that SpaceClaim is a better tool for the job.

Q. Do you foresee that after they manipulate the models they would be brought back into the CAD model?

A. I cannot speak about LG, but can discuss how in general how customers use SpaceClaim as compared to how they use a detailed design CAD system. In general we see two workflows. One is doing concept models up front of traditional CAD. This uses existing CAD data plus new requirements. Users would bring all this together to create a new design. This is hard using history based systems because feature based constraints limit design flexibility. In SpaceClaim it’s a piece of cake. Users can sketch in SpaceClaim’s 3D environment to get a clear idea of what the new design might look like. Then they will release the SpaceClaim model to detailed design to develop all of the final details using a feature based design system.

Q Do you expect that detailed CAD design might directly read the SpaceClaim model or remodel the design?

A. It depends on the use case. For example, if the new design is heavily based on an existing design, SpaceClaim provides what might be needed for redesign or remodel. It is up to the user to make that decision. Dumb solids (non history based) are not so dumb anymore. Many detailed CAD systems can use non-history based solids as a starting point, can recognize features, and can even directly edit these models within the CAD system. In addition, many companies now outsource detailed designs, and thus do not even work with the detailed CAD model. Thus SpaceClaim is easier to use in many cases.

Q. What was the size of the LG order? Was it sizeable or just a handful of seats?

A. I cannot comment, except to say that is was a decent sized order. Not just one seat. In general, companies that we do announce have done a pilot and plan to use it SpaceClaim substantially in the future. We are not claiming that such customers plan to replace all CAD; instead the seats are additive for new users. For manufacturing engineering, SpaceClaim is a better tool than a detailed CAD system.

There are a lot of big name companies that use SpaceClaim software and there is a consistent pattern of where they see SpaceClaim playing a role. The usage is a different model than the traditional CAD usage we are used to. I can point to Tyco as an example. Customers like this have done pilots and intend to deploy. We are seeing a different world than the tightly integrated CAD system world, where it is  difficult to send data from one system to the other. We saw a need for all users and their supply chain to use the same system. Not having a history based system enables easy transfer of models. Most users want to visualize models, take measurements, do some what-ifs, investigate alternate components, and change some variables in the design. Feature based models are not right because they do not allow these questions to easily be answered.

We are seeing SpaceClaim used for bid modeling by sales engineers, concept modeling, simulation both up front and used with existing CAD data, and to cleanup and prepare models for manufacturing and to create tooling. That is the kind of use our customers are making of SpaceClaim.

I think it’s pretty clear that SpaceClaim’s direct modeling tools are the best tools for these uses. It does not matter whose CAD tool is in use at a company. If you want all these other people to be empowered with 3D then SpaceClaim is the best application than any other and I would welcome the chance to prove it. We are showing it over and over again to many types of companies in industries as diverse as: automotive, aerospace, defense, consumer electronics, and medical devices.

Q. Can you tell me anything about the company today so I can share how fast SpaceClaim has grown?

A. In 2009 the business was 3.5X the size in 2008, and in 2010 it tripled again. We are hiring and there are lots of job openings. However, we are a private company and are unwilling to release number of seats or revenue.

Q How many people are in the company?

A. We have about 50 in total, including development and are hiring for Concord, MA as well as some positions worldwide.

Q. Are there any aspects of recent SpaceClaim releases that drove this kind of business?

A. We have achieved a level of maturity now, with our recently shipped 7th release of SpaceClaim 2011. It’s now robust and stable – a very important factor for customers. We think we offer a more reliable system than traditional CAD vendors, which often crash a few times a day. Our goal is to be as reliable as Microsoft’s operating system.

We have several case studies available on our website that show how customers use SpaceClaim. Particularly interesting are the ones from Emhart Glass, Batelaan Plastics, and Schramm Inc. We believe that there is a high degree of pent up demand by customers looking for a tool that lets people get their jobs done without wrestling with the complexity of feature based CAD systems. This is the primary need driving the business.

Q. Do you see significant competition from CAD vendor direct modeling solutions?

A. These big beefy products are not what the customer wants. They are not appropriate for the customers we are going after. What does direct modeling in most CAD systems get you? You can edit imported parts better and easier. Is that useful for most users? Users will still need to use beefy CAD system to do their job along with a lightweight viewer. This is not the way I see things going; eventually people are going to be creating ideas in 3D, mailing them to each other and collaborating with customers and suppliers. To enable this, we are looking to remove all the CAD friction and help customers get products to market faster. SpaceClaim is the only tool able to do that – to enable really pervasive 3D. Maybe Autodesk might understand this, but other CAD vendors don’t seem to get it.

We figured out what most users want and made the best tool.

——

Click here to see the SpaceClaim – LG Electronics press release. Users can see more information on the SpaceClaim website, including case studies and self-running demos at www.spaceclaim.com.

——

TechniCom tests Part 6 show why Inventor’s digital prototyping outshines SolidWorks in Interoperability

Interoperability and Direct Modeling

Continuing on with part 6 of our 8 part blog series dedicated to showing the differences between Autodesk Inventor Professional 2011 and SolidWorks Premium 2011 for digital prototyping workflows, we examine the ability to import MCAD models from CATIA and to perform direct edits on the imported geometry. Finally we take a drawing off the final model.

As in the other blogs in this series, this blog includes videos of both systems being used to perform the test.

To examine interoperability, we tested the capabilities of the software by importing a CATIA part, modifying the imported part, and creating and validating the accuracy of a DWG drawing of the part for communication with vendors.

View of the bell housing used in this exercise

Autodesk provided a video of Inventor accomplishing this test, a DWG drawing of the bell housing, the bell housing in CATIA format and the bell housing in IGES format.

Key differences you will see in this test

Autodesk Inventor is able to import and export most common CAD formats as well as neutral formats. Working with imported data uses the direct modeling tools found in Inventor Fusion Technology Preview to make changes. Creating a fully associative drawing in DWG format requires no additional effort since Inventor uses native DWG as the file type for drawings created from the 3D model.

SolidWorks can also import and export from a variety of CAD formats but has no support for CATIA files, which must be translated into a neutral file format introducing opportunity for errors. It also has tools for modifying geometry with several functions like feature recognition and move face. Lastly, DWG drawings are not associative to the 3D model and may require a significant amount of time and effort to clean up translation errors prior to sending them to customers and vendors. In this test, the SolidWorks DWG associativity did not work, however, SolidWorks supported this capability in past releases. It did not work on TechniCom’s version of SW2011; it may work in other installations.

What’s Important in Interoperability

  • Directly reading the other systems data directly – in this case CATIA – rather than performing a multi-step and error prone process of intermediate data conversion.
  • Easily share design data with customers, vendors, suppliers, and other departments using different CAD systems.
  • Reading and writing native DWG files for production, and publishing designs in formats that customers can use in their own applications.

Observations

Importing CATIA Part

The desired result of this test was to import a CATIA V5 model into the software.

Autodesk Inventor read the CATIA data directly and was able to open the model with no issue.

SolidWorks was unable to read the model and requires a third party add-on at additional cost to import CATIA V5 models. To perform the later tests, an IGES format file was made available and was imported successfully. This is a major issue for automotive and aerospace suppliers and OEMs since there are many companies involved, many of which require data in native CATIA format! Oddly enough SolidWorks is owned by the same company as CATIA and yet cannot read the data directly.

Modifying Imported Geometry

This test examined the ability of each software system to make small modifications to the “dumb” solid created from the imported file.

Modifying the geometry in Inventor Fusion

Inventor made the necessary modifications using the free Inventor Fusion Technology Preview labs application. The changes were made successfully and then Change Manager was used to update the dumb solid in Inventor.

SolidWorks had no problem with the direct modification of the imported part. Feature recognition capabilities were used to modify the plates and the holes as required.

Modifying the geometry in SolidWorks

In this case it was easier than Inventor, which required back and forth interaction with Inventor Fusion.

Creating DWG Drawing

This test involved creating a drawing in DWG format, opening the DWG in a 2D viewer, and making a change to the 3D model and updating the DWG.

Measuring the resultant DWG created by Inventor

Measuring the resultant DWG created by SolidWorks

Note the incorrectly scaled dimension in the SolidWorks created drawing.

Inventor created the drawing in DWG format so no translation was required. The file was opened in AutoCAD and presented exactly as it was in Inventor. After making the change to the 3D model, the DWG version of the drawing updated automatically.

SolidWorks could create DWG files for export to vendors. It lacked the ability to be fully associative with the SolidWorks 3D model. Adding dimensions or taking measurements in the scaled view in the resulting DWG drawing were not scaled correctly with the view. In this case SolidWorks added a dimension that showed as 64mm instead of the correct 32mm.

See how the Inventor engineer performed the test: 

See how our SolidWorks engineer performed the test: 

—-

The next blog in this series will examine design automation and creating drawings from the resulting design. Stay tuned or sign up to be notified of my blog updates.

—-

TechniCom tests (Part 5) – Inventor’s digital prototyping vs. SolidWorks for Export to BIM

Exporting BIM Ready Models

Continuing on with part 5 of our 8 part blog series dedicated to showing the differences between Autodesk Inventor Professional 2011 and SolidWorks Premium 2011 for digital prototyping workflows, we examine the ability to export MCAD models to BIM. Not just the models, but models that contain smart information needed by BIM systems. Besides the images we have videos below of how our engineers used each system to perform the test objectives.

In this test, we started with a complex assembly that had already been designed in the mechanical system. The goal was to export a simplified version of the assembly for inclusion in Autodesk Revit software (BIM software). The software should allow the user to reduce the level of detail in the exported file to protect proprietary design information, to indicate to the BIM software the category (window, door, HVAC, etc.) of the exported data, to provide types and locations for plumbing and electrical connections, and to be able to control the orientation of the exported part or assembly.

View of the chiller unit used in this test

Autodesk provided a video of the chiller and the workflow performing the desired steps. Also provided was an IGES file of the chiller with all the model details.

The key differences you will see demonstrated below

Autodesk Inventor has a dedicated workflow that communicates a lightweight version of the geometry with critical information like connection points and component types while maintaining all physical and visual aspects of the design in a file format that can be easily handled by the most common BIM applications.

SolidWorks does not have a dedicated workflow. SolidWorks 2011 added the ability to exchange simplified geometry data, but with limited amounts of component information due to the use of a neutral file format. It was not able to provide connection points or carry physical and visual properties over to the BIM system.

What’s Important in exporting BIM Ready Models

  • Provide lightweight BIM-ready models which can be directly incorporated into the building design process
  • Mechanical models should include unique architectural and construction information

Observations

Model Preparation

The first step was to simplify the model by suppressing proprietary and unnecessary parts from the assembly and removing details such as small holes, grilles, etc. This eliminates exposing intellectual property and also reduces the level of detail, avoiding large file sizes.

Inventor was able to suppress all unnecessary components easily. A shrinkwrap feature simplified the assembly with native and non-native data by removing small features and patching small holes, reducing the file size.

SolidWorks was able to suppress unnecessary components without issue. Because this was imported data, SolidWorks was unable to further simplify the designs, although had it been native data it may have been possible. Readers should note that the Inventor data, however, was native and thus a somewhat unfair comparison.

Orientation for Import

The second step was to orient the part properly for use in BIM software. Ensuring that models come in with the correct orientation removes the frustrating process of reorienting every time the product is inserted into a design.

Inventor allows the user to create and assign a custom local coordinate system that can be specified on export thus eliminating this issue.

SolidWorks can create a custom user coordinate system (UCS), but cannot use it when exporting IFC files (a BIM standard file format). A new assembly needs to be created with the product placed in the correct orientation. This workaround requires additional time to create the new assembly and creates additional data to manage.

Define Connection Points

The third step was to define connection points to the assembly with sizing and connection attributes. Mechanical, electrical, and plumping (MEP) engineers need to know the location, size, and type of connections required for the product when designing piping or wiring for the building.

MEP connection definition in Inventor

Inventor allows the user to specify the location of connections along with information about pipe size, wiring, connection method, system type (i.e. hot water/cold water/120v/240v/etc.), and flow direction.

SolidWorks is unable to assign connection properties, instead requiring the data to be manually communicated, a minimally acceptable alternative.

Data Export

The final step was to export the assembly to a file format that can be read into BIM software with the necessary attributes assigned. Inventor exports file formats that can be directly read into Revit or AutoCAD and can be included in Building Information Models.

Inventor model of the Chiller as it looks in Revit

Component types assigned in Inventor were carried over so no additional categorization was required. Additional properties are carried over, such as weight, size, and appearance.

SolidWorks exports IFC 2×3 neutral format files that can also be directly read into Revit or AutoCAD. Component type can be assigned on export and are carried over into Revit. BIM designers are required to manually assign additional properties.

SolidWorks model of the Chiller as it looks in Revit

Videos of each system performing the test

Watch the video of Inventor performing this test:

Watch the video of SolidWorks performing this test:

—-

The next blog in this series will examine interoperability and direct modeling of an imported model, focusing on importing a CATIA model, modifying it, and creating drawings. Stay tuned or sign up to be notified of my blog updates.

—-