DEVELOP3D - Technology for the product lifecycle

Basic construction of a Viking sailing ship. Boat design is one of the most elegant forms in the world of engineering. Simple, efficient and timeless (and when I say timeless, I mean, since 1500 BC). Image courtesy of the good folks at Vikingskip.com

Here’s something I was reminded of recently on a trip to see the team at Delcam (in their HQ quarters, a scant 15 miles from my home, rather than 8,000 miles away in Korea this time). If you’re not familiar with it’s solutions, Delcam has a huge range of technology which often solves real, live problems faced in the heady world of design and manufacture, rather than, as some vendors choose, creating solutions looking for problems. While there’s much elegance in many areas of Delcam’s offering, one thing lept out at me – and that’s how it’s flagship modelling system handles surface creation.

Surfacing is a complex business. From first principles, when you’re trying to create sculpted, complex forms, you’re looking at an inherently more complex workflow than when working with prismatic features. the geometry is more complex, so the creation of it is going to be more complex, right?

Traditionally, yes. Absolutely.

Surfacing requires that you first build a network of curves and the precise form of those are controlled by not only the form you want to create, but how you want to create it. There are many types of surfaces. Planar surfaces are flat and the simplest. Then you have four sided surfaces, n-sided, bi-rail surfaces, extrudes, lofted surfaces, swept surfaces, blends, flanges, fillets. Filleting in itself is a very complex art depending on your form requirements. If you’re working to corners, then you’re looking at trying to merge three or more surfaces converging on a single point and at that point, you might want different fillets, different set-back value.

All in all, its a complex and often daunting prospect – particularly for those that have learned their trade-craft using mainstream, solid modelling applications. Knowing what forms you’re aiming for is essential to create curves (often referred to as wires), before you even get to actually creating a surface.

Delcam’s PowerShape has been on the market for about ten years or so and the company has been through revision after revision to give its users a set of tools that allow you to work with complex geometry, fix it, prepare it for manufacture That’s given it a perspective that is only shared by a handful of vendors. Delcam has a set of tools that are used by a community that’s both a) demanding (as they need flawless data – which begets flawless tool forms) and b) very used to dealing with crappy third party data. These are the people that take crappy data and turn it into a manufacturable item – something that requires highly efficient tools.

Perhaps the perfect example of this is how PowerShape handles surface creation. As we’ve discussed, you’re often facing multiple decisions about what curves to create, then what exact type of surface you want to create, before you even start to think about creating any geometry. What Delcam has developed is Smart Surfacer and it takes many of these decisions out of your hands – or at least, gives you a helping hand.

Basically, you create the curve network you want, then invoke the Smart Surfacer command. This presents a simple dialog box. With this active, you then start to select the geometry, either from curves or from existing surface edges. The system inspects your selections, looks at the types of surfaces it can create, then presents you with the best guess is has for the best type of surface you could create based on that selection. As you add more geometry to the selection, it reevaluates the choice and switches the surface type and displays a preview.

1. Take this simple geometry set – two circles and a connecting arc.

2. Select the smaller circle and you get a planar fill surface.

3.Add in the connecting arc to the selection and it’ll switch to a drive Curve, to push the arc around the circle.

4. Adding in the large circle maintains a Drive Curve, but runs it between the two circles, using the arc as the Drive Curve.

Here’s another example

1: Rectangle, helix, circle. – selecting the rectangle gives you a planar surface.

2: Adding the helix into the selection gets you a drive curve that’s very similar to a swept feature.

3: Adding the circle in switches the Drive Curve to push between two forms, creating a smooth transition.

Of course, these are pretty simplistic demonstrations for the purposes of getting the concepts across, but the usefulness and the simplicity of the tool should be clear. Quite often you’re not dealing with singular surfaces such as these, but rather, dealing with the complexity of trying to finish up that set of surfaces, to squeeze that final last few in that tie together the whole form, the points where form quality is won or lost – and that’s exactly where this tool comes into its own. Rather you having to rework other surfaces in the set to patch in that final surface, the system can find the optimum solution and present it to you for inspection and fine tuning. there are also more manual tools avialable from the command, such as the Composite Curve creator, which can assist greatly when you have multiple, disjointed surfaces meeting at one area.

PowerShape’s Smart Surfacer is a perfect example of what I’m looking for in this search – it’s a deceptively simple tool that collects together best practice, knowledge and experience of dealing with some of the worse geometry known to man and presenting it in a tool that adds that intelligence in an unobtrusive manner, while giving you the freedom to dive in and edit things manually if needs be.

PowerShape-E is avialable for free, to play with at your leisure at www.powershape-e.com – I’d recommend doing so to anyone with a passing interest in complex shape description.

Z Corp have released the latest addition to it’s range of 3Dprinter products with the launch of the ZPrinter 350. As many readers will know, Z Corp is one of the leaders in the 3D printing world, where speed and low-cost are absolutely key to support the product development process. While the company always grabs headlines with its colour printers, there’s still a big market for monochrome machines. Running costs are lower, the machines cost less (due to the reduction in complexity) and for many, the ability to quick create a series of prototypes, discuss them around a table and progress design is all that they want.

What this brings to the product line-up is an advancement of the existing 310 monochrome product, adding in the integrated post processing capabilities of machines like the 450 and 650 (which we took a look at a while back), to give you a system that builds quickly, provides you with the tools to break out the model from the build chamber, recycle the unused material and post process the material. It also takes advantage of Z Corp’s most recent build powder (ZP150) which gives you a much whiter model (which is ideal for concept development and for architectural users, is ideal) and a much more robust green model (green refers to the state before you infiltrate the model to ‘fix’ it).

Build volume is a very usable 203 x 254 x 203 mm, it builds just under Z Corp’s benchmark 1” per hour (they quote 0.8” per hour), with layers in the 300 x 450 dpi resolution range (no mention of layer thickness). One thing I did find interesting was the discussion of the affordable nature of the machine.

The ZPrinter 350 costs around $25,900.

While that’s a cheap machine by historical standards, there are much lower cost commercial machines from traditional vendors on the market (the Solido machine and Dimensions’ uPrint spring to mind). The 350 pulls things back in cost of consumables and a greater build volume, but there’s changes afoot in the RP market. One of the other low-cost hopefuls, Desktop Factory, got into financial trouble recently and the assets got picked up by 3D Systems – the results of which still seem uncertain.

Alongside this, there’s the homebrew market that is gaining huge interest amongst many users, purely because of the ability to create parts with very low cost hardware, often self built. Take the MakerBot, the RepRap project (which is now on its second generation).

I’m not for a minute suggesting that professional designers and engineers are going to foresake investment in professional level technology that solves a serious requirement, but there’s a home brew enthusiasm for this type of technology which is now 30 years old in many areas.

Another thing to consider is that many of the original patents are now starting to expire and that always means that the technology can be freed from the stranglehold (a morally correct one I might add) that the originators have on it.

There are interesting times to come for 3D printing. Very interesting indeed.

Phillipe Starck’s 1980 Juicy Saliff for Alessi. Often seen as an iconic design. Is it elegant? As a form, yes, undoubtedly, it’s lines are clean and refreshing like the lemons that it juices. Is it an elegant product? No. It’s rubbish. It spills juice everywhere, skids all over the worktop and generally annoys the living crap out of almost everyone that buys one, unless they’re just putting it on a shelf. Now let’s look at something useful.

We’ve already talked about the concept I’m trying to get across here and taken a look at what Siemens has been up to with NX 7.0 and HD3D. For this post, I want to look at a much different vendor from Siemens, namely, DriveWorks. One of the benefits that I’ve had, doing this strange job that I do, is I’ve had the chance to meet a lot of people over the years and seen them develop new tools, new ideas and grow their businesses from the very beginning. One of those that always springs to mind is DriveWorks.

The British company is a provider of design automation tools the SolidWorks community. I believe I first met co-founder and CEO Glen Smith when he worked for a now-defunct SolidWorks reseller, back in the late nineties. The occasion of our meeting was a trip with him to visit a customer of that reseller who had adopted an automation system that Glen had developed for them to automate the design of some very complex automotive products (I won’t mention the name as many years have passed and they might have changed their strategy). I got to see the company get a presentation of the barebones of what would become DriveWorks, based on Access databases, Excel spreadsheets and a whole host of custom API programming done by Glen himself. That was over ten years ago now and the company DriveWorks, now headed up by Glen alongside co-founder and Vice President, Maria Sarkar, has been through all manner of changes, buy outs and strategic decisions that have brought the company to its current position as providing an integral part of SolidWorks’ offering (DriveWorksXpress) as well as it’s own products that are sold by resellers across the globe. It’s been a true delight to see a company grow and become highly successful from very humble beginnings.

Only a few months ago, I got a call to come up and see Glen, Maria and the team to talk about something they had brewing. Not even a Swine Flu scare kept me away (even if we all agreed not to bother with the usual handshake or hug), this team always have something interesting to say and always something interesting to show.

What they had to show was DriveWorks Solo, a system that bridges the gap between the DriveWorksXpress product that almost all SolidWorks users have as part of their solution, and the high-end, web-based DriveWorks Pro system. DriveWorks Solo is meant to find that sweet spot where an organisation can make heavy use of automation of its products, but doesn’t need all the bells and whistles. The product is sold on the web, supporting digitally and while it’s early days indeed, seats have been sold within days of its launch. So, how does DriveWorks Solo fit into this series of articles?

The answer is something like this.

Automation is something that, when you strip it back, makes a huge amount of sense for many design and engineering based organisations. While most won’t be able to automate everything, there are a great deal of organisations that have design and engineering resources tied up in repetitive work. Standardisation is something that many organisations took to heart ten years ago and the ability to create custom solutions for customers, based on a set of standard components, can give you a real advantage. Custom solution, but without having to redesign everything.

Parameter and input capture is relatively simple, as is rules definition.

While that’s true, being able to do that digitally is somewhat difficult. Using a standard parametric modelling system to try and automate the design of even the most modest sub-system or modular product can be very difficult. Its theoretically possible to do, but once you start to do it in anger, you’re generating one hell of a lot of data that can very quickly become messy.

What you need is a set of tools that allow you to do that, but in a sensible and efficient manner. Hence, Design automation technology. Many of the higher-end solutions (such as NX, Catia, Pro/E to some extent) have knowledge-based automation tool, but even these are incredibly complex processes and not attuned to the needs of the mainstream.

On the other hand, DriveWorks Solo most certainly is.

The team has completely reworked the interface to build a system that runs within the SolidWorks UI. It steps you through the process of capturing a starting assembly, identifying the parameters and rules that drive the automation, then building a user interface on top of it, making it possible for anyone to jump into the system and create a customised product or sub-system, using performance and customer inputs, and have the system generate not only the 3D description, but also the supporting documentation in terms of drawings. None of the custom programming, none of the consultancy, none of the painstaking rework of existing products. It’s designed to be done by the designer or engineer and maintained by the same.

DriveWorks Solo gives you the ability to create a UI for automation, making it much easier to deploy and make use of.

The benefits that an organisation can derive from this are many fold. By automating design (without impacting other work), you have the chance to both remove the drudgery of repetitive work by your team, and can have them working on the ‘non-automatable’ (yes, I made that word up) parts of a design project, finding new areas for exploration and new potential. It’s done easily, cost effectively and could bring huge benefits to many SolidWorks users.

To my mind, that’s an incredibly elegant solution. It’s easy to use assuming you understand how your products are defined, it’s affordable and can pay real benefits. Benefits to both your personal productivity – who doesn’t prefer working on the challenging product, rather than the crappy-same-soup-reheated-work. And of course, for business in terms of more productivity, greater potential for innovation.

There you go, another example of how things should be. I wonder what’s next. Stay tuned to find out…

NB: We took an in-depth look at DriveWorks Solo in the September issue, which is available here.

Consider the single speed bicycle – it’s a seemingly complex arrangement of steel tubes, mechanical linkages, spoken wheels, gearing, pneumatic types, valves, rubber tyres and bearings. but when time optimises that combination, you reach something that provides transportation for millions, cheaply and efficiently.

I thought I’d start off with this first in my series of Elegance posts, as it’s fresh off the wire and brand new and all shiny. If you look at Siemens’ product portfolio, there’s an interesting dynamic. Solid Edge gains all of the marketing noise, all of the marketing push – after all, it’s the company’s mainstream offering to market. But behind this, there’s always NX, something I personally find much more appealing and interesting. Why?

Because it’s one of those systems that’s been around for so long that it does almost everything you could ever imagine. What Siemens are facing is the simple fact that while NX’s power lies in its long history, it’s problems also find their root cause in the same.

That said, there’s been a concerted effort that predates the Siemens acquisition, to bring NX up to speed. The work done on the user interface in NX2 (I think) has been progressed over the last four or five years and the system looks and feels fresh, but with all that technology behind it that allows you to define some very complex products. NX users spread across a huge range of industry sectors, but product complexity (in terms of assembly size), process complexity (in terms of development and manufacture) and form complexity (in terms of the shapes the users are defining) seem to unit them.

The big news for the NX 7.0 release is split into two areas. Of course, there’s extension of the Synchronous Technology that has been the big headline grabber for Siemens PLM over the past two years. Within NX, Sync Tech’s impact has been much lower key when you look at the grand scheme of things – much of this is down to the fact that with NX, you don’t have the bifurcating decision to go history-based or non-history-based as you do within Edge. NX is more refined and more flexible than that.

Yes, you can completely blow the history away or you can choose to use those more freeform modelling tools as part of a history-based part, recording each action, so its entirely traceable, repeatable and editable (this, I suspect, is how Edge will end up too – but that’s purely speculation). That’s a good example of this elegance we’re talking about. Sync tech isn’t a panacea for everything. It’s a technology that is best suited to solving specific design or more accurately, geometry creation and editing problems. Whether that’s down to working with dumb native data or working with highly complex feature-based parts, where you need to execute a seemingly small tweak, but doing so will require huge amounts of rework.

Sync Tech aside, the thing I’m most fascinated by is the introduction of HD3D or to give it its full moniker, High Definition 3D and this is a truly elegant solution – not necessarily because it’s new technology (which it isn’t) but the manner in which the various component parts have been put together and implemented to solve a very complex issue relating to buried data (both geometric and metadata) and issue resolution workflow. Let’s look at those two very quickly.

Visual Reporting: Take the average design and engineering organisation that’s fully adopted both 3D design and PLM. Within those two technologies you now have the ability to fully document a product. Not only it’s form and function, but the full gamut of information that relates to its development and its manufacture and further into its use lifecycle and eventual retirement. In short, that’s a crap lot of data. The problem is that unless you have a very intimate knowledge of where that data is, who created it and what you’re actually looking for, it’s very hard to get an idea of where a product is at. the 3D datasets contain the form, then these are linked to the metadata attached to each part, sub-assembly, sub-system in the form of documents, text, spreadsheets, pdfs etc etc etc. and to find both, you need, typically, to use two different systems – in the case of Siemens, that’s NX (or a 3D viewing technology) and Teamcenter and despite all of their best intentions, the two don’t exactly work together.

Here a visual report has been created based on weight, using customised ranges to define different categories of parts and sub-systems.

What HD3D does is provide a framework that’s delivered in both NX and Teamcenter, that allows you to use the graphically rich nature of 3D data to gain access to the metadata that’s underneath it, to explore and visualise that data, delve into details where needed and to filter it to gain the information you need. Whether that’s a peep at where development efforts are concentrating (by filtering and visualising parts or sub-assemblies under work – in itself derived from change status), what parts are being outsourced, where costs or weight are found (by filtering for parts within specific cost or mass ranges).

Customisable searches and filters allow you to find the information you want and filter out that you don’t need. Here, the Visual Reporting allows you to colour code the assembly by supplier, while the dialog shows you more detail.

What Siemens has done is take its experience with the JT format (for lightweight viewing and data manipulation), some of the technology it’s had for large scale visualisation and its master of data management with Teamcenter and build a technology that allows this to be done visually, efficiently and very cleanly indeed. Searches can be saved (there are presets delivered), but its possible to create custom searches and reports based on whatever search criteria you require. This allows you to load up, gain an idea of where things are at, based on your focus areas, then get to work. it’s a combination of some quite complex technologies that have been reworked into a very slick environment. One thing that’s key is just that: getting on with the job. This is the other focus for this first HD3D release.

Issue Management: When you have highly complex products and geographically dispersed and outsourced input into the development of those products, you have serious issues with data management. Not in terms of storing them within a database, but rather around the workflow relating to ensure that data is conforming to company requirements, whether thats in terms of geometry quality, ensuring compliance with company, customer or international standards. What NX 7.0 introduces is a workflow, backed up with better use of existing technology, to solve these types of issues. Siemens has, for sometime, had the NX Check Mate product. This has performed just these types of checks for a while. What Siemens has done is integrate these checks into the HD3D environment. you load up the assembly, run the checks and get back a very visual list of issues that it finds. The combination of visualisation tools and reporting allow you to work through those issues, find the problems that need to be addressed (whether that’s small faces that don’t match FEA requirements or PMI formatting issues doesn’t really matter).
It’s done very interactively and very efficiently.

The issue management tools built into HD3D inspect your product models, find potential problems and present them in a graphically rich environment that not only presents the information, but allows you to very quickly gain an understanding of the context in which they occur

What’s impressive from my perspective is that there are then tools available to move these issues you discover into a workflow to resolve them and progress the project forward. Essentially, an issue is identified through automated checks, you assigning it to the person or team responsible for its resolution and that then kicks off a change request (handled by Teamcenter) to progress and resolve it.

The CoverFlow style issue browser (top right) allows you to flick through issues, explore further. Clicking a part or issue tag (shown in a small red icon with a white cross) brings up further details.

This is an initial release but the promise it holds is phenomenal. None of this technology is new. Check-Mate, JT, Teamcenter and NX itself are existing technologies that have been combined, rationalised and delivered to create a solution to real engineering problems. Data burial and retrieval is a constant problem for many organizations. The data is there, but how you get at it is anyone’s guess. By providing a combination of rich graphical visualization backed up with clean efficient search tools, you can get the information you want, almost instantly. On the flip-side the issue management and resolution tools again do the same, take existing tools and redeploy them to create an environment where fundamental bottlenecks can be first identified, but then progressed through to resolution in a fully traceable environment.

To my mind, it’s a perfect example of what I’m talking about with this Elegance idea. Stay tuned for the next part, when we look at the seemingly complex world of rules-based design automation and how DriveWorks Solo is changing how it’s done.