WYSIWYG in manufacturing: what you see (will soon be) what you get

While "what you see is what you get" is not quite there for injection molding, 3D printing, and composites, WYSIWYG in manufacturing is on its way.

WYSIWYG header rubber duckies

Andrew Anagnost

October 5, 2016

min read
WYSIWYG groundhog pops out of ground
With different 3D printing technologies, additive manufacturing may soon do for physical goods what WYSIWYG software did for web page creation.

Remember the early days of computers, before WYSIWYG (“what you see is what you get”) editors? There was often no correlation between the on-screen and print versions of a document, which led to hours of hair-pulling frustration.

What happened if you sent the document to a print shop like Kinko’s? You’d drive over to pick up your order, only to discover that it looked completely wonky—wasting your time and money in the process.

It was a little like the life of Bill Murray’s character in Groundhog Day: waking up repeatedly to Sonny & Cher’s “I Got You Babe” on the radio alarm clock and reliving the same painful scenarios over and over. But in the case of those early word-processing days, the scenario was guessing and re-guessing what changes you needed to make—ad nauseam—until you got your document right.

Then software companies created WYSIWYG editors to solve that problem. You could finally trust that what you saw on screen would be what you got—no matter what computer or printer (inkjet, laser, daisy wheel, dot matrix) you used.

Unfortunately, what you see is still not what you get in manufacturing. With injection molding, 3D printing, and composite materials, manufacturing still requires multiple iterations and prototypes, which can be a massive drain on time, resources, and your sanity.

Now imagine a world where what you see really is what you get for designed things: You could cut down on iterations; save huge amounts of time and money; and use those extra resources to innovate, explore better options, and create the ultimate design rather than something merely workable.

Break the mold

Twenty years ago, engineers used simulation to see what was wrong with an injection-mold design, but it was always post-catastrophe. It took a lot of trial and error for engineers to get to a workable design and prototype, which caused delays and ate into profits.

Today, rather than allowing an expensive problem to suck up your time, you can use simulation in the design phase and then start building a mold.

An example is the plastic injection-molded parts for the interior of a car: The design might look perfect on screen, but if there is uneven stress in the injection-molding process—such as melt-temperature, cavity-pressure, fill-rate, and cooling-rate variables—your part could significantly warp or shrink.

Your design drastically impacts what you get out of an injection mold, and if you mess up a mold, it can cost tens of thousands of dollars—especially if you don’t have control of the manufacturing facilities.

3D-Printed metal mania

WYSIWYG carbon-fiber composite process of a helicopter tailboom
The carbon-fiber composite process of a helicopter tailboom.  Image courtesy of Automated Dynamics.

Achieving WYSIWYG in additive manufacturing is even harder. Whether it’s plastic, metal, or even cement, the material used in 3D printing deforms during and after the printing process, so you need to apply inverse deformations to arrive at a product that comes close to what you see on screen.

Say you’re an aerospace engineer designing a turbine blade to be 3D-printed using an Inconel alloy: You could burn many days on the prototype-iteration loop and spend more than $10,000 per iteration. Because additive-manufactured parts are commonly produced in low volume, if it costs $100,000 to get to a reliable production process, that cost has to be amortized across the number of parts you need. What happens if you have to print 20 parts to get a return on the cost of the iterations? And what if you need only five parts? The inability to predict the final outcome of 3D printing is limiting the adoption of this revolutionary manufacturing process.

Mixed-medium materials

If you thought reaching WYSIWYG in 3D printing and injection molding was hard, imagine the world of laminated composites, where your manufacturing process is unique to the thing you design.

Carbon-fiber composite materials are superstiff, superstrong, lightweight, and impact resistant. They became revolutionary after being used on the Boeing 787 Dreamliner. The fuselage was made on a massive spindle with the carbon fiber wrapping around and around like threaded tape. As it progressed, the spindle would move accordingly to lay fibers in the right direction for each layer.

Generally, composites are made with threads of carbon laid into resin and forming plies in different orientations—unidirectional, woven, or chopped. You get different strength and warping properties depending on how you lay the plies.

Because it’s so complex, few experts know how to use composites with simulation. But if the computer could identify which composite will have the right performance characteristics—such as strength and weight properties—then you could pick the right material. And then the computer could tell you what manufacturing process would be required or how the design should be modified to make the part look and behave exactly as it should.

Fail fast—or don’t fail at all

Injection-mold design of an automotive speedo cluster, using Autodesk Moldflow
Injection-mold design of an automotive speedo cluster, using Autodesk Moldflow.

Autodesk is working on building these simulation capabilities—Moldflow for injection molding, Netfabb Simulation for additive manufacturing, and Helius PFA for composites—to solve the WYSIWYG problem for manufacturers.

The goal is to replicate the manufacturing process inside the computer so you can see the difference between what you design and what you’re going to get. If the result is not what you intended, the computer will tell you how to fix it so you’ll have fewer failures and arrive at a reliable product sooner.

Injection mold of an automotive speedo cluster
Injection mold of an automotive speedo cluster.

Instead of engaging in a lengthy prototype loop, you could send your design to a manufacturer and know that the product you get back will meet your requirements. Right now, no one would dare do a volume run immediately, but if you could guarantee the first test run was always right, it would remove iteration cycles. And that would enable more time and money to explore the possibilities of making a better product.

The nearfuture of manufacturing simulation is like the advent of WYSIWYG editors in word processing: Instead of sending a design back and forth 18 times before it looks the way you want it to, you will finally be able to print one or two and move on.

Right now, you may be accustomed to hearing something like an annoying refrain of “I Got You Babe” in your head over and over. But have faith: There will soon be a time when WYSIWYG moves the day forward for manufacturing.

Andrew Anagnost

About Andrew Anagnost

Andrew Anagnost is the president and chief executive officer of Autodesk.Dr. Anagnost’s career spans more than 25 years of product, business, and marketing experience focused on driving strategy, transformation, and product development — and includes positions at Autodesk, Lockheed Aeronautical Systems Company, and EXA Corporation. He also completed a doctorate degree at Stanford University and worked at NASA Ames Research Center as an NRC post-doctoral fellow.Anagnost began his career at Autodesk in 1997 and has held a wide range of roles in the areas of marketing, new business development, product management, and product development. Prior to becoming President and CEO in June 217, he served as Chief Marketing Officer and SVP of the Business Strategy & Marketing organization. In this role, Andrew served as architect and leader of Autodesk’s business model transition—moving the company to become a software-as-a-service (SaaS) solutions provider. Previously, Anagnost held various executive positions across Autodesk. Early in his Autodesk career, he led the development of the company’s manufacturing products and grew Autodesk Inventor revenue to over $500 million.Anagnost is a member of the Autodesk Board of Directors. He holds a Bachelor of Science Degree in Mechanical Engineering from California State University, Northridge (CSUN), and holds both an MS in Engineering Science and a PhD in Aeronautical Engineering and Computer Science from Stanford University.

Recommended for you