Design and Manufacturing solutions through Digital Prototyping and Interoperability

Category Archives: Manufacturing

The Future of Making Things at AU

Future – Making Things

The theme for manufacturing and mechanical design at Autodesk University 2016 was “The Future of Making Things.” Autodesk showcased how in the future making things will utilize technologies like generative design, augmented and virtual reality, robotics, and additive manufacturing.

The exhibit hall included an industrial robotic 3D printing, a brick-laying construction robot, the first-ever generatively designed office building, an open-sourced 3D printed drone, a virtual reality experience to configure a full-scale Ford Mustang, and augmented reality construction hard hats. Autodesk expects these types of technologies to be mainstream in the next one to four years but wanted to showcase their availability today for those “cutting-edge innovators who are propelling us into the future“.

“We’re all in on the cloud. Access to infinite computing power and the ability to work together effortlessly is completely changing how design and engineering are done,” – Carl Bass, CEO of Autodesk.

Virtual and Augmented Reality

I am no longer a Virtual Reality virgin. The picture does not do justice my initial experiences with VR… I was like a kid in a candy store, giggly, and excited.

AU16 MRT VR1

“Thanks to game engine technology and innovative hardware, people can now be immersed in any world you want to create”

The Immersive Learning booth utilized technology by NVIDIA, Lenovo, and HTC. You were placed in a room and able to interact with multiple objects in this “Learning Museum”. This included stacking bricks, playing musical instruments, and turning off gravity and hitting objects around the room. The goal was to show you different VR interactive experiences. I did not catch which software was used to create the VR experience.

AU16 Virtual Reality 1

Orange County Choppers

Autodesk commissioned the OCC guys to create an electric motorbike. The bike was developed completely in Fusion 360 and is 100% electric.

AU16 Volta Orange County Choppers Fusion 360

Hack Rod

Hack Rod started a research project to investigate using new technologies in building a performance car. This project evolved into Hack Rod creating the world’s first vehicle chassis engineered by artificial intelligence.

Future of making things pavilion_hack-rod-2

Hack Rod started by wiring both the car and the driver and putting both through a “punishing” series of test drives in the California desert. The result was 20+ million data points about the car’s structure and the forces on it.

Hack Rod and Autodesk fed the data collected into Project Dreamcatcher. The output from Dreamcatcher was applied to the existing 3D model of the chassis.

Hackrod Generative Design Future of Making Things

3D printing is now the plan for Hack Rob to fabricate the critical parts due to the complex forms created by the generative design.

Here’s the teaser video regarding the Generative Design exhibits at Autodesk University 2016

Additive Manufacturing

Additive Manufacturing is over 30-years old but only recently have the barriers been removed to allow this technology to move from just prototyping into production.

Autodesk equipped a six-axis industrial robot with a polymer (plastic) extruder controlling it with patented software. Autodesk is now printing complex structures at “surprising” speeds. By using the robot they are not limited to a single plane and can print layers along any path.

“Robotics is rapidly closing the gap between the digital and physical world, opening methods of building and manufacturing that were previously unimaginable. Intuitive software has made robots easier to program, more useful, and more versatile. This allows new industries to emerge using robots in novel and creative ways. “

future of making things pavilion_robots

By using the robot they are not limited to a single plane and can print layers along any path. It is also able to create complex geometry without the use of support structures.

Future of Making Things AU Additive Manufacturing 1

With the addition of thermal cameras and vision systems, Autodesk is pushing the use of polymers into scenarios where it is superior to metals. Printing with polymers uses significantly less energy compared to metal and comes at a much lower cost.

Future of Making Things AU Additive Manufacturing 2

 

AU Additive Manufacturing 3

“What if designing with infinite complexity was not a constraint, but a platform for brilliance?”

Kloner3D 240 Twin

Autodesk’s Project Escher software and control technology make it possible for multiple print heads to work together to produce large scale industrial parts at very reasonable speeds. Showcased in the AU2016 Exhibit Hall, the KLONER3D 240 TWIN printer produces parts quickly because of its collaborative printing capability.

AU16 Kloner3D 240 TWIN

UA Architect

This March, Under Armour released the world’s first completely 3D-printed training shoe. It was developed using generative design and advanced additive manufacturing.

AU16 Under Armour UA Architect

The key to the Under Armour shoe is that it provides both cushioning and support for even the “most strenuous of workouts

AU16 Under Armour Architect 3DP Phenom

Under Armour used selective laser sintering (SLS) to 3D print the flexible (yet durable) lattice structure.

 

Metal Printing

For 3D printing to work at a large scale the system needs to be able to do three things: work with high-performance materials, cover large areas, and adapt to variable conditions.

future of making things pavilion_robots-2

Utilizing six-axis robot arms and welding machines, printing structures in traditional metals like stainless steel and aluminum is a possibility.

AU16 Additive Manufacturing 3

The key is a closed loop feedback system employing cameras and computer vision algorithms to facilitate real-time corrections of unanticipated variations during the process. The feedback also enables automatic calibration and registration of multiple robots working together.

AU16 Additive Manufacturing 5

Future Making Things MX3D

The MX3D robots are capable of printing molten metal in mid-air. They can produce thin lines the diameter of a pencil or tubes of variable diameter. The examples presented at Autodesk University showcased the ability of these robots to create 3D objects in almost any size or shape.

AU16 MX3D Future Making Things 1

AU16 MX3D Future Making Things 2

You can read more at www.autode.sk/robotics

Technology In Motion

Here’s another trailer video from Autodesk…

Solid Edge Cloud Licensing and Storage

Siemens PLM decided to move Sold Edge a bit further into the cloud with some good options integrated into ST9. Integrated Solid Edge cloud licensing and flexible storage options should prove to be a welcomed addition to Solid Edge’s proven performance.

Cloud Licensing

Sold Edge cloud licensing and storage is available for version ST9, and is optional. This option allows users to check out their ST9 license, and use it on whichever machine they are working on. Home or office – you can work anywhere you need to with a single license.

Siemens Solid Edge ST9 Cloud Licensing

The License Manager handles the Cloud license option. Users may switch to cloud-licensing by using the “I need to visit the Siemens Licensing website option”, and follow the cloud licensing links. Some restrictions I picked up on include:

  • Users must call Siemens customer support to revert their cloud licenses back to node-locked licenses.
  • Solid Edge requires and internet connection to check out the license and start. Once it’s running the internet is no longer required.
  • Group licensing is required to be consistent. You cannot mix and match node-locked and cloud based licenses.

Cloud Storage and Management

Solid Edge ST9 will now store it’s data on any platform you like, including locally, vault solutions, and even the cloud. Siemens PLM understands that companies are using a wide array of storage solutions. ST9 was released with this in mind. Solid Edge can be set to incorporate various cloud solutions, including Dropbox, OneDrive, etc. Come to think of it, Kenesto Drive might also be a good fit.

I wanted to get some feedback on cloud storage functionality. Bill McClure, VP of Strategic Initiatives at Siemens PLM, was kind enough to tell me about his experiences. I asked the Solid Edge team how multiple users worked in this environment. McClure said that a form of lock files were now being incorporated, and that Solid Edge would recognize when the files were opened by another user. When asked about performance, he said that he was happy with the Solid Edge cloud pairing, using a Dropbox subscription. McClure noted that Dropbox has been stable and reliable (I have always said cloud syncing should “just work, like Dropbox”). He went on to describe the Dropbox bit-level detection and replacement algorithm [Dropbox calls this a “binary diff”]. It senses where the files have changed, and only replace the modified sections, not the entire files. This is apparently how Dropbox updates so quickly.

Thoughts

Siemens PLM has spent a great deal of effort trying too keep their customers happy. Licensing in the past has followed the company’s large, and somewhat rigid structure. However, the company is clearly trying to be more flexible about these issues. There is something for everyone:

Perpetual licensing – The company continues to respect their very-loyal Solid Edge user community by maintaining this option

Rentals and annual subscription – Some like a bit more flexibility and the most updated software

cloud / floating network / node-locked options

I think the cloud-license option is a great step in the right direction. I will probably do this myself.  Some of my associates remain unsure about using the option when the internet is unavailable, and then not being able to start Solid Edge. As the Solid Edge team points out, cloud licensing will not fit everyone’s situation. There are numerous options, and no one option is perfect.

I’ll be following up on the new data management features integrated into ST9. I hope that incorporating these into the cloud storage scenarios will produce a versatile fit for numerous companies. New integrated data management, along with Solid Edge’s particular strengths, have the capacity to fill the needs of a well-rounded CAD product, in a larger part of the CAD market.

Solid Edge Free to New Start-Up Companies

Siemens PLM announced that they will be furnishing Solid Edge free to New Start-Up Companies. This surprise was delivered last week during the General Assembly of Solid Edge University 2016, held in Indianapolis. The company said that they want to help empower new companies, and make design software part of the solution, not the problem.

If you are a new start-up, or are considering a new design or manufacturing endeavor, this great opportunity is for you. Siemens PLM intends to furnish one seat of Solid Edge Premium per engineer at the company. The license will be for a term of one year. The program includes training and co-marketing opportunities (more details to follow when available).

Who is eligible for Solid Edge Free to New Start-Ups?

The criteria for being considered for the program are quite simple, and are as follows:

  • The company must be a legally registered company, less than 3 years old.
  • The company must have less than 1 million U.S. dollars of funding
  • The Company must have less than 10 million U.S. dollars in annual revenue
  • The Company must be formed for the purpose of design and/or manufacturing

Application Process

The process is simple, and there is no fee. Fill out this Solid Edge Free for Start-ups online application. Once completed, the company will contact you with more information.

 

 

Fusion 360 | The Effects of Combine Order

I noticed something about the order I use to join solids in Fusion 360. Until recently, I didn’t consider the effects of Combine Order; It does matter. This is what happened and what I learned.

Combine Order

I had initially exported two solid bodies and four fluid bodies from Fusion 360 to a STEP file. This was imported into Autodesk CFD without a hitch. I realized that changes needed to be made, and returned to Fusion 360. I altered the design, exported again, and started a new study in CFD. This time things were not so good. Instead of two solid components, I had approximately 550. There was no way I wanted to troubleshoot the mesh with that kind of overlapping complexity. I tried a few Fusion 360 alterations, but nothing mattered until I adjusted the combine order of operations.

What doesn’t work

You can create one singe solid body from many, by using the Combine tool. This works well as you would expect. One option within the tool, create a new component, will create a new component comprised of the newly joined/combined bodies, as a single body. All seems well in Fusion 360. However when exported to a STEP file, the model remains as numerous separate pieces (as if you imported a block into AutoCAD, and then exploded it). Not what I wanted.

What does work

Combine Tool in Fusion 360    Combine Tool in Fusion 360

I  recommend combining the bodies first, then separately create the new component. In the image above notice that the “New Component” option is not selected.

Create New Component in Fusion 360    Create New Component in Fusion 360

Afterward, Adding the joined, single-body into the component will produce a single body in the STEP file.

NOTE: if you add an additional body into the component after one has been added, either by dragging or using the combine tool, you may get multiple bodies in the exported STEP file.

Fusion 360 Combine before Component    Fusion 360 Component After Combine 

Notice the history bar at the bottom of each image, and the components at the bottom of the Browser Tree at the left side. In the first operation, the components are combined into the single hub body (highlighted). In the second operation, that single-body is added to a new component on-the-fly, called HPC Rotor. Inside that component is only the single hub body. This will come out in the STEP file properly.

The key is to join all bodies in one step as a single body in the history list. Then add that single-body to the component later. If you realize later that you need to add more bodies, back-up the history list and include the bodies in the single combine operation first. While these symptoms might sound suspiciously like saving your tool bodies, or having multiple bodies evident in your components, this is not the case in this situation. Only one body was evident in each component, viewed in Fusion 360.

Fusion 360: Compressor Mock-up Easy or not?

I have reached the point in my overall turbofan design where I can do some preliminary Computational Fluid Dynamics (CFD) modeling. For this I need a good CAD model that I can beat up a little. I had always wanted to see if Fusion 360 would be able to handle a compressor model, and so I decided to give it a whirl. This way I could get the fluid body I needed, test Fusion 360, and evaluate the design direction prior to settling in on a specific CAD modeler. In this article I will not focus on the CFD, but on the experience I had using Fusion 360.

Is the Fusion 360 modeler a platform worthy of marketing to engineers?

Is the modeling space robust enough for serious design?

Sketching – does it suck? (Some of you are laughing at the directness of this question, I am. However, that is what you really want to know, right?)

Parametrics – Are they user friendly?

and so on…

Fusion 360 Modeling

 

Sketching in Fusion 360

Sketching was surprisingly good

I found sketching to be rather straight forward. The usual array of curves, lines, arcs, and so forth were available from the sketching panel. What is surprising is the smoothness that the system handles projections. I know we’re not talking rocket science here, but it was quite nice. The ability to turn on and off bodies, sketches, and constructions at their header while remaining in sketch mode is also much appreciated. Make your projections, turn off the features, and continue with a clean slate. Scott points out that Fusion 360 will handle most constrained projections automatically.

Recovering from lost references was smooth

Then next item was recovering from lost references. I should preface this by saying that I never once lost a projection in Fusion 360 by re-ordering the features. I am speaking about redefining sketch planes, and losing references during that process; a procedure that most CAD users will not move to lightly. Fusion 360 lost sight of the references, but maintained the remaining constraints. It permitted cleanup easily. In my case I use 3 constraints per stage to tie the sketch in place. After the sketch plane re-definition, I simply turned on my limiting construction planes, added the 2 projections, and then 3 constraints. The model snapped into place without any other adjustments or coaxing.

Copying sketch profiles – Why didn’t I think of it sooner?

I must say that after manually redrawing the same features in 16 sketches, you’d think I would have figured it out. You can copy sketch profiles from sketch to sketch, and all the constraints transpose in-tact! In the image below I simply copied constrained profiles from one sketch, and pasted into another. Then 3 constraints added to tie it to the relevant projections, and the sketch was in place; all the parameters were preset. It took days to model first 8 stages; it took half an hour to model the last 3.

So what about the parameters and dimension adjustments?

Parameter Organization

Parameters in Fusion 360 are awesome, and I mean it. Each parameter is organized by name: name of body, name of sketch, name of construction plane, etc. All of my control parameters are added as User Parameters. When I copy a sketch, Fusion 360 automatically groups the dimensional constraints together in the parameters under the new sketch name.

Parameter Organization in Fusion 360

Adjusting these parameters was simple: go to the new sketch section and select the User Parameters as needed. They are so easy to discern under this type of organization.

Parameter Auto-complete, I love you!

One of my favorite Parameter features is the auto-complete type lookup. When you pick over an existing parameter, or begin typing one, the system opens a pull-down with all possible relevant parameters. I was able to find a parameter that needed to be indexed, pick it, pull down to the new parameter, and go on the the next change; Soooooooooo nice.

 

Parameter Lookup in Fusion 360

Sketching was irritating

Huh, what happened to surprisingly good? Well, there were some issues. The only ones that are really worth mentioning in this edition were: inferred constraints and order of constraints.

Inferred Constraints

Sketching had all the constraints you’d hope to have. Horizontal and vertical were unfortunately in the same function, but not a deal killer. That said, the first issue that came up repeatedly was Fusion 360’s inferred constraints. Inferred constraints are important, and a powerful tool, but I tend to think of these like automatic osnaps in AutoCAD: always in the way. In Autodesk Inventor, these don’t really seem to be too much problem. However Fusion 360 is hell bent on snapping a constraint to any damn thing you pass or seem to pass. Obviously you can work around it with care, but it shanked me quite  a few times. I need to learn how to shut these off.

…Scott has informed me that the control key will toggle these. I meant I want them off-off.

Order of constraints needs some love

Order of constraints is one of those things that you look at the cost of the software and decide how bad the problem is. When I say order of constraints, I’m speaking about the order that you create geometric and dimensional constraints in the sketch. In Inventor, we try to get as many relevant geometric constraints down first, then add the dimensional scalars. If Fusion 360… probably not. There is a delicate balance that must be observed. If you apply too many geometric constraints first, when the dimensional changes are to be applied, often Fusion 360 will report that the sketch is over-constrained. If you go back and remove some constraints here and there to relax things, add the dimensions, Fusion 360 will often then allow the old geometric constraints.

Moreover, the more you constrain the sketch, the more edgy manual, drag type manipulations become. It’s almost like a resolution problem. Modelers with a better resolution don’t react so sharply when you drag a model with only one or two degrees of freedom (DOF). In Fusion 360 I’d want that single DOF to be quite linear, and well focused. I started to think in terms of constraints, and applying angular constraints carefully while the model was still very relaxed.

Modeling in Fusion 360

I didn’t go too deep into modeling. I lofted between profiles, performed revolutions, trimmed bodies, and so forth. In general, modeling was well mannered.

Filleting between bodies

You cannot fillet at intersections of separate bodies until they are joined. OK, I get that. However, the ease of filleting that used to be in Fusion R3 (if you have to ask, you probably don’t need to know) doesn’t seem to be there. A different model kernel is in use most likely, but still sucks a bit. You have to really watch the geometries when applying intricate fillets.

Moving features about in the history

I have to say that moving objects about in history was pretty smooth. In other modelers, when I move thing about, I cringe waiting for the rebuild, and hoping nothing will fall apart. The speed at which it remodeled and the ease of moving things was quite nice. Not being able to see the names during the move sucks, but I am still impressed.

Revolving  a surface – not suckin now that I know how to do it!!

I couldn’t determine how to model a parametrically controlled, updatable, revolved surface. You can perform a revolve of an open profile in the pushy-squishy modelling section, but none of that updates in history. Scott has informed me that it’s in the Patch area, and sure enough, it works. Just remember that the pushy-squishy (“Form”) version more obvious in the Model area will not update with your parameters.

Conclusion

This was but a small list of what Fusion 360 can do. As you can see from the images in this article, Fusion 360 can perform moderately complex geometry, and in most cases control it all parametrically. If you paid 10 large for it, you got robbed. If you paid the going rate for the subscription, I think you got a sweet deal.

Is the Fusion 360 modeler a platform worthy of marketing to engineers?

It all depends on what you are designing. I can’t say until I get further into the product, but for now I will say that it’s powerful for the price, and I’m hopeful of it’s continued maturity.

Is the modeling space robust enough for intricate design?

This kind of relates to the last question, but consider to this mock-up. I’m controlling the bend, aspect, twist, and taper angles of every blade set in every stage, independently. Not a task for a cheesy modeler.

Sketching – does it suck?

I answered that with specifics in the section above. That said, in general, it’s a pretty good sketching environment.

Parametrics – Are they user friendly?

Oh no doubt about it, they are nice.

Would I consider engineering with Fusion 360 full time as my modeler of choice?

If I had little money and was starting out – for the price I paid, hell yes. However I would be quite cautious of what I was designing. I will likely continue to model my preliminary engine structures in Fusion 360, as well as preparing my CFD models.  When it comes time to commit to a very expensive, very dangerous design, a more powerful platform will likely be in order.

I have seen Fusion 360 growing far beyond where I thought it would go. It’s tied into A360, easy to collaborate in, and packed with so many great features far beyond simple modeling. I think the rendering engine is nice, and with Simulation and CAM in the package, it’s already hard to resist; for the money, it’s unbeatable. I’m keenly interested how it will evolve in the next two years. I’ll follow up with more observation as I add more to this design.

HP Compressor in Autodesk Fusion 360

 

Modeling Limits: Top-Down or Bottom-Up?

As many of you know, I recently returned to school to complete my engineering degrees; and simultaneously began researching small footprint turbofan engine technologies. As none of this in theory has anything to do with modeling (or fun), I haven’t had a lot of time to write. Modeling this engine is now on my mind, and so is the question of Top-Down or Bottom-Up?

Top-Down or Bottom-Up?

We are nearing the end of the basic mathematical model. I have begun considering how I would bring this beast to life in a digital model; top-down or bottom-up. I am a controlled person and like designing within limits; top-down gives me precisely that. However, in this case I am not so sure that torturing myself by bringing an adaptable model to bear is a good idea, where the limits are still quite uncertain.

Modeling Turbine Blades

Image Courtesy of Stinging Eyes 

The Problem

Stage count is the fly in the ointment. For example, I think we will have 6 low-pressure (LP) fan stages (yes, 6 bloody stages), but am still trying to reduce it to 4. The High Pressure (HP) compressor is equally ugly, and the HP and LP turbines are dependent on the compressor models. If I do not know how many stages will be involved, I will have to constantly return to the master skeleton, and adjust the coefficients to adjust how much space I have to work with. Ugghh!!

However, if I let go of my notions about “how to design something properly”, I might find an easier way. Perhaps if I made my design sections as CAD models with named planes at either end, I can join the design sections end to end, in a bottom-up approach. It will be easier to keep all my section specific design constraints in the respective design sections. Later when I do the bypass and exterior skin sections, I can import key surfaces from each section to work from, and build outward.

Conclusion

Top-down or Bottom Up is one of the first questions I ask prior to modeling. I rarely decide against a master skeleton approach. Moreover building outward is my big fear; since this entire project involves a very small cross sectional dimension limit, it seems counterproductive to work outward. That said, it also seems insane at this point to try and monkey-fart the parameters about in a skeleton file to eventually fit sub-structures that can perform as intended. After much consideration, I think keeping the mathematical design limits tight, and then building the structure outward from the HP compressor will allow the greatest degree of flexibility, and the easiest model to alter, which we expect to do a lot until all aspects of the engine design have been played out.

Your thoughts?

 

Join the Community