CoresOnDemand: Helping Engineers Do Their Magic

CoresOnDemand-Logo-120hEngineers Do Magic

In the world of simulation there are two facts of life. First, the deadline of “yesterday would be good” is not too uncommon. Funding deadlines, product roll-out dates, as well as unexpected project requirements are all reliable sources for last minute changes. Engineers are required to do quality work and deliver reliable results in limited time and resources. In essence perform sorcery.

af-01

Second, the size and complexity of models can vary wildly. Anything from fasteners and gaskets to complete systems or structures can be in the pipeline. Engineers can be looking at any combination of hundreds of variables that impact the resources required for a successful simulation.

Required CPU cores, RAM per core, interconnect speeds, available disk space, operating system and ANSYS version all vary depending on the model files, simulation type, size, run-time and target date for the results.

Engineers usually do magic. But sometimes limited time or resources that are out of reach can delay on-time delivery of project tasks.

At PADT, We Can Help

PADT Inc. has been nostrils deep in engineering services and simulation products for over 20 years. We know engineering, we know how to simulate engineering and we know ANSYS very well. To address the challenges our customers are facing, in 2015 PADT introduced CoresOnDemand to the engineering community.

af-02

CoresOnDemand offers the combination of our proven CUBE cluster, ANSYS simulation tools and the PADT experience and support as an on demand simulation resource. By focusing on the specific needs of ANSYS users, CoresOnDemand was built to deliver performance and flexibility for the full range of applications. Specifics about the clusters and their configurations can be found at CoresOnDemand.com.

CoresOnDemand is a high performance computing environment purpose built to help customers address numerical simulation needs that require compute power that isn’t available or that is needed on a temporary basis.

Call Us We’re Nice

CoresOnDemand is a new service in the world of on-demand computing. Prospective customers just need to give us a call or send us an inquiry here to get all of their questions answered. The engineers behind CoresOnDemand have a deep understanding of the ANSYS tools and distributed computing and are able to asses and properly size a compute environment that matches the needed resources.

Call us we’re nice!

Two Halves of the Nutshell

The process for executing a lease on a CoresOnDemand cluster is quite straight forward. There are two parts to a lease:

PART 1: How many cores & how long is the lease for?

By working with the PADT engineers – and possibly benchmarking their models – customers can set a realistic estimate on how many cores are required and how long their models need to run on the CoresOnDemand clusters. Normally, leases are in one-week blocks with incentives for longer or regular lease requirements.

Clusters are leased in one-week blocks, but we’re flexible.

Part 2: How will ANSYS be licensed?

An ANSYS license is required in order to run on the CoresOnDemand environment.  A license lease can be generated by contacting any ANSYS channel partner. PADT can generate license leases in Arizona, Colorado, New Mexico, Utah & Nevada. Licenses can also be borrowed from the customer’s existing license pool.

An ANSYS license may be leased from an ANSYS channel partner or borrowed from customer’s existing license pool.

Using the Cluster

Once the CoresOnDemand team has completed the cluster setup and user creation (takes a couple of hours for most cases), customers can login and begin using the cluster. The CoresOnDemand clusters allow customers to use the connection method they are comfortable with. All connections to CoresOnDemand are encrypted and are protected by a firewall and an isolated network environment.

Step 1: Transfer files to the cluster:

Files can be transferred to the cluster using Secure Copy Protocol which creates an encrypted tunnel for copying files. A graphical tool is also available for Windows users (& it’s freeJ). Also, larger files can be loaded to the cluster manually by sending a DVD, Blu-ray disk or external storage device to PADT. The CoresOnDemand team will mount the volume and can assist in the copying of data.

Step 2: Connect to the cluster and start jobs

Customers can connect to the cluster through an SSH connection. This is the most basic interface where users can launch interactive or batch processing jobs on the cluster. SSH is secure, fast and very stable. The downside of SSH is that is has limited graphical capabilities.

Another option is to use the Nice Software Desktop Cloud Visualization (DCV) interface. DCV provides enhanced interactive 2D/3D access over a standard network. It enables users to access the cluster from anywhere on virtually any device with a screen and an internet connection. The main advantage of DCV is the ability to start interactive ANSYS jobs and monitor them without the need for a continuous connection. For example, a user can connect from his laptop to launch the job and later use his iPad to monitor the progress.

af-04

Figure 1. 12 Million cell model simulated on CoresOnDemand

The CoresOnDemand environment also has the Torque resource manager implemented where customers can submit multiple jobs to a job queue and run them in sequence without any manual intervention.

Customers can use SCP or ship external storage to get data on the cluster. SSH or DCV can be used to access the cluster. Batch, interactive or Torque scheduler can be used to submit and monitor jobs.

All Done?

Once the simulation runs are completed customers usually choose one of two methods to transfer data back. First is to download the results over the internet using SCP (mentioned earlier) or have external media shipped back (External media can be encrypted if needed).

After the customer receives the data and confirms that all useful data was recovered from the cluster, CoresOnDemand engineers re-image the cluster to remove all user data, user accounts and logs. This marks the end of the lease engagement and customers can rest assured that CoresOnDemand is available to help…and it’s pretty fast too.

At the end of the lease customers can download their data or have it shipped on external media. The cluster is later re-imaged and all user data, accounts & logs are also deleted in preparation for the next customer.

CoresOnDemand-Advert-Rect-360w

ANSYS Launches Free Student Version

ansys-student-1This week ANSYS, Inc. made a fantastic announcement that has been in the works for a while, and that we think will greatly benefit the simulation community:  A free ANSYS Student product.  This is an introductory product that is focused on students who are learning the fundamentals of simulation who also want to learn the full power and capability of the ANSYS product suite.  It includes ANSYS® Multiphysics™ , ANSYS® CFD™ , ANSYS® Autodyn®, ANSYS® Workbench™, ANSYS® DesignModeler™and ANSYS®DesignXplorer™

Yes you read that right, all of the flagship products for free. No features or capabilities are turned off. It is the exact same software as the commercial product, but the size of problems that you can solve is limited.  It runs on MS Windows. Perfect for students.

PADT is excited about this because it gives students access to the ability to learn FEA and CFD simulation with the world’s most popular and capable simulation tool, without running in to brick walls. Want to do a flat plate with a hole in it? No Problem. Want to model fluid-solid-interaction on a flexible membrane valve? No Problem.  Want to model explosive forming? No Problem.  Want to model combustion with complex turbulence? No problem.

All in the same interface as students will use when they enter the work force or do research at University.

This is great news and we can’t wait to see what schools and students do with this access.

How to Get It – The New Academic Web Pages

The previous Student Portal is being replaced with an Academic Web area on the ansys.com site: ansys.com/academic.

Go to the ANSYS Student site to learn more about ANSYS Student and how to download your copy. These same pages will have resources to help you learn and understand the product.

The “Pictures”

Let me state categorically that PADT was not consulted on the image that ANSYS, Inc. used for the “student” user that was so happy to find out that there is now a free version of the ANSYS software suite.  Here is their picture:

ANSYS-student-version We would have preferred something like this:

huge.1.7907

 

Just kidding. We were happy to see this product come out and thought the picture was hilarious.  In all seriousness, we will also plug the  recent #ilooklikeanengineer twitter hash tag , highlighting the diversity of female engineers. that was awesome and we would love to see more chances for engineers to show their true selves.

 

Major Enhancements in FLOWNEX 2015: Combustors, Importers, and Pipes

FlownexLogo_OfficialSimulation has revolutionized flow and heat transfer dependent systems over the past decades by minimizing costly physical testing and accelerating time to operation around the world. But for many companies, such simulation has largely focused on components and proved to be very time consuming. The technology advancements delivered by Flownex SE now offer a fast, reliable, and accurate total system and subsystem approach to simulation.

FLOWNEX-2015-ICONS

With the release of FLOWNEX 2015, users now have access to advanced combustor system level modeling and they can interact with more system and component simulation tools. This is on top of the already considerable capabilities found in the  tool

Gas Turbine Combustor Heat Transfer Library

During the Preliminary design phase or when considering modifications to existing combustor designs it’s essential to make realistic predictions of  mass flow splits through the  various air admission holes, total pressure losses liner temperatures along the length of the combustor etc.

FLOWNEX-2015-combustor-simulationAlthough very powerful, 3D CFD solutions of combustors are specialized, time consuming processes and therefore are seldom exclusively used during initial sizing of a combustor.

It has been demonstrated that 1D/2D network tools, like Flownex, are capable of predicting with reasonable accuracy the same trends as more detailed numerical models.

The advantage, however, is Flownex’s rapid execution, which allows design modifications and parametric studies to be conducted more simply than ever before. The ease of use and incredible speed of Flownex allows 1000s of preliminary designs to be evaluated under all modes of operation for steady state and dynamic cases. Furthermore, the data obtained from the one-dimensional analysis can be used as boundary conditions for a more detailed three-dimensional model, ultimately supplementing a typical combustor design work flow.

While the simulation of combustor systems was previously possible in the Flownex environment, much of the work of implementing industry standard heat transfer correlations was left to the user through scripting .Now in Flownex SE 2015 it’s all been built in to the tool, while maintaining the flexibility required to model any combustor configuration.

New components include

  • Film convection component
  • Fluid radiation component
  • Jet impingement heat transfer component

To sum up Flownex allows more accurate initial designs, less time is spent on advanced 3D combustor simulations and rig tests, thus reducing development time and cost.

Here is a Video that shows off these features:

Added importers and integration features

AFT Fathom/Impulse/Arrow importer

An importer was added to import the file formats of AFT products. The importer imports all the diameters, loss factors heights, etc. so 90% of the effort is done, and in some cases the networks solve without any modifications.

ROHR2 Integration (pipe stress analysis software)

Flownex has the ability to calculate forces during dynamic simulations. This is very useful in pipe stress analysis for surge or water hammer cases. The ability to import complete geometries from ROHR2 and export results in the format that ROHR2 expects natively has been added. This means a user can perform these combined analysis now with ROHR2 with the minimum of effort.

Fluid Importers

An Importer was added to import liquid and gas properties from CoolProp an open source fluid property library. The existing Aspen/Hysys fluid importer was changed to be a generic Cape-Open compliant importer. This means that fluid properties can now be imported from any Cape-Open compliant server software.

FLOWNEX-2015-turbine-engine

Donny Don’t – Thin Sweep Meshing

It’s not a series of articles until there’s at least 3, so here’s the second article in my series of ‘what not to do’ in ANSYS…

Just in case you’re not familiar with thin sweep meshing, here’s an older article that goes over the basics.  Long story short, the thing sweep mesher allows you to use multiple source faces to generate a hex mesh.  It does this by essentially ‘destroying’ the backside topology.  Here’s a dummy board with imprints on the top and bottom surface:

image

If I use the automatic thin sweep mesher, I let the mesher pick which topology to use as the source mesh, and which topology to ‘destroy’.  A picture might make this easier to understand…

image

As you can see, the bottom (right picture) topology now lines up with the mesh, but when I look at the top (left picture) the topology does not line up with the mesh.  If I want to apply boundary conditions to the top of the board (left picture), I will get some very odd behavior:

image

I’ve fixed three sides of the board (why 3?  because I meant to do 4 but missed one and was too lazy to go back and re-run the analysis to explain for some of future deflection plots…sorry, that’s what you get in a free publication) and then applied a pressure to all of those faces.  When I look at the results:

image

Only one spot on the surface has been loaded.  If you go back to the mesh-with-lines picture, you’ll see that there is only a single element face fully contained in the outline of the red lines.  That is the face that gets loaded.  Looking at the input deck, we can see that the only surface effect element (how pressure loads are applied to the underlying solid) is on the one fully-contained element face:

image

If I go back and change my thin sweep to use the top surface topology, things make sense:

image

The top left image shows the thin sweep source definition.  Top right shows the new mesh where the top topology is kept.  Bottom left shows the same boundary conditions.  Bottom right shows the deformation contour.

The same problem occurs if you have contact between the top and bottom of a thin-meshed part.  I’ll switch the model above to a modal analysis and include parts on the top and bottom, with contact regions already imprinted.

image

I’ll leave the thin sweeping meshing control in place and fix three sides of the board (see previous laziness disclosure).  I hit solve and nothing happens:

image

Ah, the dreaded empty contact message.  I’ll set the variable to run just to see what’s going on.  Pro Tip:  If you don’t want to use that variable then you would have to write out the input deck, it will stop writing once it gets to the empty contact set.  Then go back and correlate the contact pair ID with the naming convection in the Connections branch.

The model solves and I get a bunch of 0-Hz (or near-0) modes, indicating rigid body motion:

image

Looking at some of those modes, I can see that the components on one side of my board are not connected:

image

The missing contacts are on the bottom of the board, where there are three surface mounted components (makes sense…I get 18 rigid body modes, or 6 modes per body).  The first ‘correct’ mode is in the bottom right image above, where it’s a flapping motion of a top-mounted component.

So…why don’t we get any contact defined on the bottom surface?  It’s because of the thin meshing.  The faces that were used to define the contact pair were ‘destroyed’ by the meshing:

image

Great…so what’s the take-away from this?  Thin sweep meshing is great, but if  you need to apply loads, constraints, define contact…basically interact with ANYTHING on both sides of the part, you may want to use a different meshing technique.  You’ve got several different options…

  1. Use the tet mesher.  Hey, 2001 called and wants its model size limits back.  The HPC capabilities of ANSYS make it pretty painless to create larger models and use additional cores and GPUs (if you have a solve-capable GPU).  I used to be worried if my model size was above 200k nodes when I first started using ANSYS…now I don’t flinch until it’s over 1.5M
    image
    Look ma, no 0-Hz modes!
  2. Use the multi-zone mesher.  With each release the mutli-zone mesher has gotten better, but for most practical applications you need to manually specify the source faces and possibly define a smaller mesh size in order to handle all the surface blocking features.
    image
    Look pa, no 0-Hz modes!Full disclosure…the multi-zone mesher did an adequate job but didn’t exactly capture all of the details of my contact patches.  It did well enough with a body sizing and manual source definition in order to ‘mostly’ bond each component to the board.
  3. Use the hex-dominant mesher.  Wow, that was hard for me to say.  I’m a bit of a meshing snob, and the hex dominant mesher was immature when it was released way back when.  There were a few instances when it was good, but for the most part, it typically created a good surface mesh and a nightmare volume mesh.  People have been telling me to give it another shot, and for the most part…they’re right.  It’s much, much better.  However, for this model, it has a hard time because of the aspect ratio.  I get the following message when I apply a hex dominant control:

    image
  4. The warning is right…the mesh looks decent on the surface but upon further investigation I get some skewed tets/pyramids.  If I reduce the element size I can significantly reduce the amount of poorly formed elements:image
  5. That’s going on the refrigerator door tonight!
    image
    And…no 0-Hz modes!
  • Lastly…go back to DesignModeler or SpaceClaim and slice/dice the model and use a multi-body part.image
    3 operations, ~2 minutes of work (I was eating at the same time)

    image
    Modify the connection group to search/sort across parts

    image

    That’s a purdy mesh!  (Note:  most of the lower-quality elements, .5 and under, are because there are 2-elements through thickness, reducing the element size or using a single element thru-thickness would fix that right up)

    image
    And…no 0-Hz modes.

Phew…this was a long one.  Sorry about that.  Get me talking about meshing and look what happens.  Again, the take-away from all of this should be that the thin sweeper is a great tool.  Just be aware of its limitations and you’ll be able to avoid some of these ‘odd’ behaviors (it’s not all that odd when you understand what happens behind the scenes).

Tech Tips and Videos for Electromechanical Simulation with ANSYS Products

ansys_free_techtipsWe just recieved a new tech tip bundle from ANSYS, Inc on Electromechanical Simulation.  You may remember when we published the Mechanical and Fluids ANSYS tech tips a few weeks ago.  This latest kit continues with information for people making devices and systems that have mechanical and electrical systems.  The focus of the kit is the application of ANSYS Maxwell and Simplorer – Maxwell to model low frequency electromagnetics and Simplorer to model systems.

Here is a link to “The Electromechanical Simulation Productivity Kit ” here. The kit includes:

  • ANSYS Maxwell Automation and Customization Application Brief
  • ANSYS Maxwell Magnetic Field Formulation Application Brief
  • Electric Machine Design Methodology Whitepaper
  • Electromagnetics And Thermal Multiphysics Analysis Webinar
  • Rechargeable Lithium Ion Battery Whitepaper
  • Robust Electric Machine Design – ANSYS Advantage Article

We also have a collection of videos that are a great introduction to the tool set and how to use it. Check out the overview and the video on the washing machine at a minimum.  Even if you have a simple EMAG or do hand calcs, you need to look at Maxwell and Simplorer.

Peeling Away the *VMASK

vmask-icon2One way to really unleash the power of APDL is to become familiar, and ultimately fluent, with array parameters. The APDL student will quickly learn that array manipulation involves heavy use of the *V commands, which are used to operate on vectors (single columns of an array). These commands can be used to add two vectors together, find the standard deviation of a column of data, and so on. *V commands consist of what I like to refer to as “action” commands and “setting” commands. The action commands, such as *VOPER, *VFILL, and *VFUN * have their own default behaviors, but these defaults may be overridden by a preceding setting command, such as *VABS, *VLEN, or *VMASK.

*VMASK is one of the most useful, but one of the most difficult to understand, *V command. At its essence it is a setting command that directs the following action command to a “masking” array of true/false values. Only cells corresponding to “true” values in the masking array are considered for the array being operated on in the subsequent action command.

For example, a frequently used application of *VMASK is in the compression of an array—for instance to only include data for selected entities. The array to be compressed would consist of data for all entities, such as element numbers, x-locations for all nodes, etc. The masking array would consist of values indicating the select status for the entities of interest: 1 for selected, –1 for unselected, and 0 for not even in the model to begin with. Only array cells corresponding to a masking array value of 1 would be included in the compression operation, with those corresponding to a value or 0 or –1 being thrown out. Here is a slide from our APDL training class that I hope illustrates things a little better.

image

Take the class or buy the manual (and review it at Amazon, please!)

What we’ve learned so far is that the masking array contains a list of true/false (or not true) values to refer to when performing its vector operation. But actually, it’s much more general than 1, 0, and –1. What *VMASK does is include cells corresponding to all positive numbers in the masking array (not just +1) and exclude cells corresponding to all values less than or equal to zero in the masking array (not just 0 and            -1), which broadens the possibilities for how *VMASK can be handy.

Everything I’ve used *VMASK for up to this point in my career has involved array compression, and I figured I’d be put out on a sweep meshed ice floe into a sea of CFD velocity streamlines (that’s what happens to old CAE engineers; you didn’t know that?) before I came up with anything else. However, I recently ran into a situation where I needed to add up just the positive numbers in an array. I was about to construct an algorithm that would test each individual number in the array to see if it was positive and, if so, add it to the total. It would’ve been cumbersome. Then I came up with a much less cumbersome approach: use the array as it’s own masking array and then perform the addition operation. Let’s look at an example.

Take the following array:

image

The sum of all values in the array is 1.5 whereas the sum of just the positive values is 11.5. What’s the most efficient way to have APDL calculate each?

In the case of summing all values in the array, it’s easy, just simply execute

*VSCFUN,sum_total,SUM,sum_exmpl(1)

which gives you

image

But what about summing just the positive values? That’s easy, just use SUM_EXMPL as its own masking array so that only the positive values are included in the operation.

*VMASK,sum_exmpl(1)

*VSCFUN,sum_pos,SUM,sum_exmpl(1)

image

Boo yeah

Now why was I doing this? I had to create a macro to calculate total nodal loads for an unconstrained component in just the positive direction (to ignore the loads counteracting in the negative direction), and this was my approach. Feel free to download the macro: facelds.mac and try it out yourself.

Video Tips: Topology Optimization with ANSYS and GENESIS

This video will show you how you can optimize a part using Topology Optimization with GENESIS through ANSYS Mechanical with support from ANSYS SpaceClaim

Vibro-Acoustics Analysis in ANSYS Mechanical as Told by a Structures Guy

Vibro-Acoustics-ANSYS-iconWith the introduction of ACT, the ANSYS Workbench editors have gained capabilities and shortcuts at much faster rate than what can be introduced in a development cycle. One of first and most far-reaching extensions is the acoustics. Inevitably I was called on by one of our customers to show them how to do a vibro-acoustics analysis (harmonic with acoustic excitation), which I did. Since the need for this type of analysis is quite broad, I’ll share it here too.

There was an extra level of excitement with this, in that I’m a structures specialist with no prior acoustics experience. So, I did my own self-training on this topic. I have to give tons of credit to Sheldon Imaoka of ANSYS Inc., who took the time to thoroughly answer the questions I had. That being said, this article will be from the standpoint of a structures engineer who’s just recently learned acoustics.

The first thing you’ll need to do is download the Acoustics extension from the Downloads section at the ANSYS Customer Portal and install it in Workbench.

image

It’s at the very top, under ‘A’ for “Acoustics”

One thing you’ll notice when you unzip the Acoustics Extension package is that it contains and entire Acoustics training course. Take advantage of this freebie when learning acoustics analysis. I’ll note that, most of the process outlined in this article comes from the Submarine workshop in the acoustics training course.

Once you’ve installed and turned on the Acoustics extension, insert a Harmonic Analysis system into the project schematic, link to the solid geometry file, and specify the material properties for the solid. You’ll specify the properties for the acoustic region in Mechanical under the appropriate Acoustics extension objects.

image

Rename as you see fit

Assuming you just have the geometry for the solid and not the acoustics domain, create two acoustics regions around the solid. The first region, surrounding the solid, will function as the fluid region itself, through which the acoustic waves travel and interact with the structure. The second region, surrounding the first acoustics region, will function as the Perfectly Matched Layer (PML). The PML essentially acts as the infinite boundary of the system. (If you’re an electromagnetics expert, you already know this and I’m boring you.) You can easily create these domains using the enclosure tool in DesignModeler.

image

Acoustics Regions

Now we’re ready for the analysis. Open up Mechanical. Look at all those buttons on the Acoustics toolbar! Yikes! Fortunately we just need a few of them.

image

Here they are

Insert an Acoustic Body and scope it to the acoustic region surrounding the structural solid. In the Details, enter the density and speed of sound for the fluid. Also set the Acoustic-Structural Coupled Body Options to Coupled With Symmetric Algorithm.

image

image

image

Pay attention to the menu picks, Details, and geometry scoping here and in the rest of the image captures

“Coupled” refers to coupled-field behavior, i.e. the mutual interaction between the structure and the fluid. You’re probably familiar with this. You need that, otherwise the acoustic waves are just bouncing off the structure and the structure isn’t doing anything. Regarding the Symmetric Algorithm: The degrees of freedom for the acoustic system consists of both structural displacements and fluid pressures, giving you an asymmetric stiffness matrix. However, ANSYS has incorporated a symmetrization algorithm to convert the asymmetric stiffness matrix to a symmetric matrix, resulting in half as many equations that need to be solved and thus a faster solution time yadda yadda yadda, so go with that.

Now insert another Acoustic Body, this time scoped to the outer acoustic region (body). This is your Perfectly Matched Layer. Specify fluid density and speed of sound as before. This time, leave the Coupled Body Option as Uncoupled. But, set Perfectly Matched Layers to On.

 imageimage

Apply an Acoustic Pressure of zero to the outer faces of the PML body (Boundary Conditions > Acoustic Pressure). As you may have guessed from the menu pick, this is your acoustics boundary condition.

clip_image020imageimage

Now we’ll apply some acoustic wave excitation to this thing. From the Excitation menu, select Wave Sources (Harmonic). In the Details, set the Excitation Type to either Pressure or Velocity, set the Source Location and specify the excitation pressure or velocity value. In this example, I went with Pressure since that’s what MIL-STD-810 specifies, but this option will be based on your customer requirements. I also assumed an external acoustic source (hence, Outside the Model), but again, that will be based on your particular project. You also need to specify the vector of the wave source, via rotations about the Z and Y axes (f and q). In this case I chose 30 and 60 degrees, respectfully, to make it interesting. Once again, enter the density and speed of sound for the fluid.

clip_image026image

Insert Scattering Controls under the Analysis Settings menu and specify whether the Field Output should be Total or Scattered. Total gives you constant pressure waves that interact with the solid but not each other. Scattered gives you wave that interact and interfere with each other as well as the solid.

imageimage

Set up the Fluid-Structural Interaction boundary condition where the structural faces are “wetted” by the acoustic domain. The FSI Interface is found under the Boundary Conditions menu.

imageimage

Apply structural constraints and specify harmonic analysis settings just like you would with a standard harmonic analysis. Make sure you request Stresses under the Output Controls. Solve the model.

imageimage

Plot your structural results as you would for a typical harmonic analysis. Acoustic Pressure wave results may be found under the Results menu in the Acoustics toolbar. If you used Total field output for the scattering option, you can verify your wave source direction by looking at the Acoustic Pressure Contours. Keep in mind that the contours will be orthogonal to the axis of the sine wave; you may need to put some extra spatial thought into it to fully understand what’s going on.

imageimage

image

Acoustic Pressures: Field Output = Total

image

Acoustic Pressures: Field Output = Scattered

image

Von-Mises Stresses, Max Over Phase: Field Output = Scattered

As you’ll note in the training course, there are a number of design questions that can be answered with acoustics analysis. In this article, I’ve addressed what I thought would be one of the more popular applications of acoustics simulation. If the demand is there, I’ll research and compose more articles on various acoustics applications in the future. For instance, another area I’ve examined is natural frequencies of a structure that’s submerged in a fluid. If there’s another acoustics topic you’d like us to write about, please let us know in the comments.

Instructions for Installing and Configuring ANSYS MAXWELL and PExprt, Versions 16.X

ANSYS_pexpert_maxwell-1ANSYS PExpert is a fantastic tool for the design, modeling, and analysis of transformers and inductors. Using a combination of classical and finite element analysis (FEA) techniques, ANSYS PExprt determines the core size and shape, air gaps, and winding strategy for a given power converter topology. What we and our customers have found very useful is the ability to then evaluate the magnetic design in ANSYS Maxwell to view such things as flux density in the core and current density distribution in the windings. Powerful stuff.

The first step in implementing ANSYS PExprt with ANSYS Maxwell is installing and configuring them correctly.  We created a step-by-step guild for our ANSYS customers here in the Southwest, and thought others would find it useful.

ansys-maxwell-pexprt-install-image

Download: InstallingMaxwellandPExprt16.pdf

As always, feel free to contact us if you have any questions or need more information. Also, even if you are not in our sales area, please consider using PADT for consulting or training.

 

 

 

 

 

 

 

3D Printing Users Lunched & Learned about Dealing with Scanned, Repaired, and Legacy Geometry

PADT-Geometry-Scanning-PartnersThis Thursday we had the first of seven free seminars on how to deal with geometry created with 3D scanning, how to repair faceted geometry, and how to deal with old CAD geometry.  Don’t panic, we have six more scheduled. Scroll down to see the schedule and register for upcoming versions of this seminar. The inaugural session was held in PADT’s Tempe office and engineers from several departments across the company shared the tools we use in consulting and the lessons we have learned over the years to a pack room full with customers that represented everything from the home inventor to engineers from some of Arizona’s largest aerospace and electronics companies.

badgeometry

As more and more companies do 3D printing we are finding that they struggle with imperfect geometry. Whether it was scanned, from another CAD system, or an STL (3D Printer) file from someone else, when it came time to print parts people were having difficulty getting valid geometry.  So we created a road show to go over the tools we use here to 1) get good scan geometry in the first place, 2) convert scan geometry into something useful, and 3) repair bad STL and CAD files.

Things got kicked off with a presentation on the various ways you can scan 3D geometry.  Our scanning engineer, Ademola, also demonstrated our Geomagic Capture and Steinbichler scanner on some real parts.

padt-scanning-spaceclaim-seminar-tempe-3-2015_05_14

After some food, we moved on to looking at Geomagic Design X.  This is the tool we use to convert our scan data to a fully usable and clean CAD model.  If you have tried to go from scan to CAD without this tool, you know how much work it is.  padt-scanning-spaceclaim-seminar-tempe-2015_05_14

Next we looked that the tool that we use to import, modify, and clean existing geometry: SpaceClaim.  As the presenter Tyler Smith said “No matter the source of geometry, SpaceClaim is the tool to help”

padt-scanning-spaceclaim-seminar-tempe-2-2015_05_14

We finished up with topological optimization. Where we spent most of the event talking about how to get good geometry, in this last presentation we talked about how to make the geometry better by using simulation to optimize the shape of your parts.

padt-scanning-spaceclaim-seminar-tempe-5-2015_05_14

It was a great crowd with the kind of questions you hope for when doing a seminar.  If you are in the Southwest, there is still time to attend one of these lunch & learns being held in other locations. Click on the event you want to register.

 
 
 
 
 
 

Or you can contact PADT directly to learn more about the products and services we covered, which included:

 

Five Ways CoresOnDemand is Different than the Cloud

CoresOnDemand-Logo-120hIn a recent press release, PADT Inc. announced the launch of CoresOnDemand.com. CoresOnDemand offers CUBE simulation clusters for customers’ ANSYS numerical simulation needs. The clusters are designed from the ground up for running ANSYS numerical simulation codes and are tested and proven to deliver performance results.

CoresOnDemand_CFD-Valve-1

POWERFUL CLUSTER INFRASTRUCTURE

The current clusters available as part of the CoresOnDemand offering are:
1- CoresOnDemand – Paris:

80-Core Intel based cluster. Based on the Intel Xeon E5-2667 v.2 3.30GHz CPU’s, the cluster utilizes a 56Gbps InfiniBand Interconnect and is running a modified version of CentOS 6.6.

CoresOnDemand-Paris-Cluster-Figure

2- CoresOnDemand – Athena:

544-Core AMD based cluster. Based on the AMD Opteron 6380 2.50GHz CPU’s the cluster utilizes a 40Gbps InfiniBand Interconnect and is running a modified version of CentOS 6.6.

CoresOnDemand-Athena-Cluster-Figure

Five Key Differentiators

The things that make CoresOnDemand different than most other cloud computing providers are:

  1. CoresOnDemand is a non-traditional cloud. It is not an instance based cluster. There is no hypervisor or any virtualization layer. Users know what resources are assigned exclusively to them every time. No layers, no emulation, no delay and no surprises.
  2. CoresOnDemand utilizes all of the standard software designed to maximize the full use of hardware features and interconnect. There are no layers between the hardware and operating system.
  3. CoresOnDemand utilizes hardware that is purpose built and benchmarked to maximize performance of simulation tools instead of a general purpose server on caffeine.
  4. CoresOnDemand provides the ability to complete high performance runs on the compute specialized nodes and later performing post processing on a post-processing appropriate node.
  5. CoresOnDemand is a way to lease compute nodes completely and exclusively for the specified duration including software licenses, compute power and hardware interconnect.

CoresOnDemand is backed up by over 20 years of PADT Inc. experience and engineering know-how. Looking at the differentiating features of CoresOnDemand, it becomes apparent that the performance and flexibility of this solution are great advantages for addressing numerical simulation requirements of any type.

To learn more visit www.coresondemand.com or fill out our request form.

Or contact our experts at coresondemand@padtinc.com or 480.813.4884 to schedule a demo or to discuss your requirements.

CoresOnDemand-ANSYS-CUBE-PADT-1

ANSYS Convergence in Chicago – Smart People Talking about Cool Stuff, and Only a Little Wind

chicago-clouds2Once ANSYS started doing more regional user group meetings, we here at PADT decided to stay out west where we felt comfortable. So we have only attended the California and Texas events in the past.  This year we decided to venture further East and go to the Convergence meeting in Chicago.  I have to say it was a great experience, different then the Santa Clara meeting a few weeks ago.

Being from Arizona, I was a bit worried about the weather. It was appropriately windy, and unfortunately overcast with low clouds so my pictures of the famous skyline was a bit stunted.

What was so great was that the same type of smart people who get simulation were there, the products they work on were very different.  From train locomotives to exercise equipment to automotive electronics, we were exposed to a variety of very unique and very cool applications.  And as usual, the people from ANSYS, Inc. had a lot to contribute and show off that was new or coming in various programs.

The event started off with a great presentation from Sin Min Yap, VP of Marketing at ANSYS, on how simulation can be used to turn good ideas in to great products.  Some great customer stories were shared and it really set a foundation as to why we do this thing called modeling and simulation.

However, the customer keynote address stole the show. It was from Jim Kennedy at Mars Corporate Innovation. No, there is no corporation on Mars (outside of SciFi movies). This is the Mars Corporation that is famous for their candy (M&M’s!!!!) and also does several well known pet foods, Wrigley chewing gum, a bunch of food brands, and drinks.  His talk was how the manufacturing of food can be improved with simulation.  Here is a bad picture of a great slide showing the modeling in FLUENT of their gum Kneaders.

gum-kneeders

And here is a model of forming Skittles:

skittles-forming

For an engineer, it doesn’t get much cooler than that. He had other great examples, and tied it all together to show how they do some very sophisticated simulation to improve their efficiency, product quality while reducing cost and minimizing their energy footprint.

Several of us sitting in the back were just amazed at the complex material models they must be using.  Candy, chocolate, gums – much more difficult than stainless steel for sure!

The next speaker talked about the Internet of Things (IoT) and the Industrial Internet, which is the IoT applied to the machines that are used to make things, and to monitor products in the field.  My key takeaway is that those of us who are responsible for designing new products have to start figuring out how we are going to make it all work. Simulation can be used to solve difficult packaging issues with batteries, antenna, and sensors the will soon be in most products we develop.  And ANSYS has the tools to do the simulation.

ANSYS also talked about their new ANSYS Enterprise Cloud solution. A very impressive effort to do a true Cloud solution for simulation… not just call time-sharing “cloud computing.”  Working with Amazon they have introduced a truly scalable, interactive, secure, and robust solution that sets the industry standard for Cloud based simulation. We also got a chance to play with it, because ANSYS’s Judd Kaiser was in the booth next to me.  It really is easy to implement and use.  I took a picture when Judd was not looking:

ansys_enterprise_cloud_demo

Renee Demay, the head of the ANSYS Customer Excellence team explained how ANSYS, Inc. is delivering a new more effective solution for customer support and services – focused as the name says on giving the customer and excellent experience.

And then the morning session finished on my favorite combined topic: Simulation and 3D Printing.  John Graham from ANSYS SpaceClaim gave a great talk on how SpaceClaim can be used to improve 3D Printing and serve as the bridge between scanning, 3D Printing, and simulation.

Here he is talking about the repair functionality in the tool. Something we use here at PADT all the time:

spaceclaim-stl-repair

That finished up the morning session, which was followed by a nice lunch where we were able to interact with people a lot. Several of you who read this blog stopped by to say hi. That really made my day.

PADT  had a booth:

IMG_6361

Several of our fiends and partners were also there, so I did a selfie with them all to say hi.  First stop was our good friends and fellow ANSYS Channel Partners SimuTech. They have a local office in Chicago:

IMG_6373

Right across from them were a team from VR&D, our favorite topological optimization tool:

IMG_6374

And NICE was there as well, showing of the remote visualization tool DCV that we use for CoresOnDemand.com and ANSYS uses in EKM and the ANSYS Enterprise Cloud:IMG_6379

The afternoon sessions were great. Lots of ANSYS and customer applications that showed the breadth and depth of usage of ANSYS products in the Midwest.

Then we had a reception, which for a Friday evening where everyone had a big commute ahead of them, was well attended.  Wine was drunk, HFSS models were shared, and the best strategy for disk array RAID configuration was debated… among other less interesting things.

A great trip, where I caught up with some old friends and made some new ones. I  look forward to exploring further east in the future!

And at least one of us will be in Houston, so if you are going make sure you stop by and say hi!

Personal Note – A Saturday of Frank Lloyd Wright

So instead of trying to red eye it back to Phoenix, I spent the night and on Saturday I went on a pilgrimage I’ve wanted to do some time: a visit to some of Frank Lloyd Wrights early creations. I won’t bore you all with my ineloquent ramblings on what a genius he was. Let me just say it was better than I expected.  His studio in his first home was nothing short of amazing (there I go, rambling…) anyhow, here are few shots I took from the outside. I didn’t take any inside pictures because: 1) I take crappy pictures, and 2) I wanted to look and explore instead of take selfies.

The first studio.

flw3

The house attached that the studio is attached to.  You should see the playroom on the top floor. Here is a link to a great blog posting about the house.

flw4

I then went to the Unity Temple which is just down the street. It was covered in scaffolding, but the inside, although worn and in need of repair, was so amazing:

flw2

Then I drove to the University of Chicago to see my second favorite Wright structure (Falling Water being my fav, duh), the Robbie House.  It did not disappoint:flw-1

A day well spent.

flw5

 

 

 

 

 

 

 

Have You Ever Dreamed in Color – 3D Color? 3D PDF Is Here with VCollab!

VCollab_Shaded_Logo_FinalIf you have ever dreamed of, or at least had a need for a 3D .pdf file of your simulation results, the dream is now realized thanks to VCollab.  As Eric Miller mentioned in The Focus blog back in February, the latest release of VCollab software enables users to save their results in 3D .pdf format.

We’ve now had a chance to test out the capability here at PADT, and we find it quite useful.  We’ve talked about VCollab before, but it’s a software suite that enables virtual collaboration (hence the name) by reducing what may be huge simulation files to a much smaller size, enabling others in your organization or your customers to dynamically view simulation results as well as CAD data in a light-weight viewer.  The folks at Vcollab have gone one step beyond that now by supporting the 3D .pdf format that is viewable in the standard Adobe reader.

Vcollab works with ANSYS results as well as results from lots of other simulation tools.  The process is:

You can download the sample file used in the images below:

vcollab-3d-pdf-sample-bolted1.pdf.

This is what a typical 3D .pdf file created from an ANSYS Mechanical/MAPDL results file looks like, with using the mouse to rotate and zoom around within Adobe reader.

So, if you recognize value in being able to create 3D .pdf files like this, the Vcollab software suite is worth investigating.  Vcollab does lots of useful things besides writing 3D .pdf files, including the capability to be imbedded within the ANSYS Engineering Knowledge Manager (ANSYS EKM) tool.

The best way to see the power of this tool is to request a demo.  Just fill out this form and we will reach out and set one up, followed by a 30 day trial.  .

Or check out www.vcollab.com.

 

Announcing CoresOnDemand.com – Dedicated Compute Power when you Need It

CoresOnDemand-Logo-120hWe are pleased to announce a new service that we feel is remote solving for FEA and CFD done right: CoresOnDemand.com.  We have taken our   proven CUBE Simulation Computers and built a cluster that users can simply rent.  So you get fast hardware, you get it all to your self, and you receive fantastic support from the ANSYS experts at PADT.

It is not a time share system, it is not a true "cloud" solution.  You tell us how many nodes you need and for how long and we rent them to you. You can submit batch or you can configure the machines however you need them.  Submit on the command line, through a batch scheduler, or run interactive. And when you are done, you do not have to send your files back to your desktop. We've loaded NICE DCV so you can do graphics intense pre- and post-processing from work or home, over the internet to our head nodes.  You can even work through your iPad.

CUBE-HVPC-512-core-closeup3-1000h

If you visit our Blog page a lot, you may have noticed the gray cloud logo with a big question mark next to it. If you guessed that was a hint that we were working on a cloud solution for ANSYS users, you were correct. We've had it up and running for a while but we kept "testing" it with  benchmarks for people buying CUBE computers. Plus we kept tweaking the setup to get the best user experience possible.  With today's announcement we are going live.

We created this service for a simple reason. Customers kept calling or emailing and asking if they could rent time on our machines.  We got started with the hardware but also started surveying and talking to users. Everyone is talking about the cloud and HPC, but we found few providers understood how to deliver the horsepower people needed in a usable way, and that users were frustrated with the offerings they had available. So we took our time and built a service that we would want to use, a service we would find considerable value in.

simulation-hardware ansys-expertise dependability

You can learn more by visiting www.CoresOnDemand.com. Or by reading the official press release included below. To get your started, here are some key facts you should know:

  1. We are running PADT CUBE computers, hooked together with infiniband. They are fast, they are loaded with RAM, and they have a ton of disk space. Since we do this type of solving all the time, we know what is needed
  2. This is a Bring Your Own License (BYOL) service. You will need to lease the licenses you need from whoever you get your ANSYS from.  As an ANSYS Channel partner we can help that process go smoothly.
  3. You do not share the hardware.  If you reserve a node, it is your node. No one else but your company can log in.  You can rent by the week, or the day.
  4. When you are done, we save the data you want us to save and then wipe the machines.  If you want us to save your "image" we can do that for a fee so next time you use the service, we can restore it to right where you were last time.
  5. Right now we are focused on ANSYS software products only. We feel strongly about focusing on what we know and maximizing value to the customers.
  6. This service is backed by PADT's technical support and IT staff. You would be hard pressed to find any other HPC provider out there who knows more about how to run ANSYS Mechanical, ANSYS Mechanical APDL, ANSYS FLUENT, ANSYS CFX, ANSYS HFSS, ANSYS MAXWELL, ANSYS LS-DYNA, ANSYS AUTODYN, ICEM CFD, and much more.

To talk to our team about running your next big job on CoresOnDemand.com contact us at 480-813-4884 or email cod@padtinc.com

CoresOnDemand-ANSYS-CUBE-PADT-1

See the official Press Release here

Press Release:

CoresOnDemand.com Launches as Dedicated ANSYS Simulation
High Performance Cloud Compute Resource 

PADT launches CoresOnDemand.com, a dedicated resource for users who need to run ANSYS simulation software in the cloud on optimized high performance computers.

Tempe, AZ – April 29, 2015 – Phoenix Analysis & Design Technologies, Inc. (PADT), the Southwest’s largest provider of simulation, product development, and 3D Printing services and products, is pleased to announce the launch of a new dedicated high performance compute resource for users of ANSYS simulation software – CoresOnDemand.com.  The team at PADT used their own experience, and the experience of their customers, to develop this unique cloud-based solution that delivers exceptional performance and a superior user experience. Unlike most cloud solutions, CoresOnDemand.com does not use virtual machines, nor do users share compute nodes. With CoresOnDemand.com users reserve one or more nodes for a set amount of time, giving them exclusive access to the hardware, while allowing them to work interactively and to set up the environment the way they want it.

The cluster behind CoresOnDemand.com is built by PADT’s IT experts using their own CUBE Simulation Computers (http://www.padtinc.com/cube), systems that are optimized for solving numerical simulation problems quickly and efficiently. This advantage is coupled with support from PADT’s experienced team, recognized technical experts in all things ANSYS. As a certified ANSYS channel partner, PADT understands the product and licensing needs of users, a significant advantage over most cloud HPC solutions.

“We kept getting calls from people asking if they could rent time on our in-house cluster. So we took a look at what was out there and talked to users about their experiences with trying to do high-end simulation in the cloud,” commented Eric Miller, Co-Owner of PADT. “What we found was that almost everyone was disappointed with the pay-per-cpu-second model, with the lack of product understanding on the part of the providers, and mediocre performance.  They also complained about having to bring large files back to their desktops to post-process. We designed CoresOnDemand.com to solve those problems.”

In addition to exclusive nodes, great hardware, and ANSYS expertise, CoresOnDemand.com adds another advantage by leveraging NICE Desktop Cloud Visualization (https://www.nice-software.com/products/dcv) to allow users to have true interactive connections to the cluster with real-time 3D graphics.  This avoids the need to download huge files or running blind in batch mode to review results. And as you would expect, the network connection and file transfer protocols available are industry standards and encrypted.

The initial cluster is configured with Intel and AMD-based CUBE Simulation nodes, connected through a high-speed Infiniband interconnect.  Each compute node has enough RAM and disk space to handle the most challenging FEA or CFD solves.  All ANSYS solvers and prep/post tools are available for use including: ANSYS Mechanical, ANSYS Mechanical APDL, ANSYS FLUENT, ANSYS CFX, ANSYS HFSS, ANSYS MAXWELL, ANSYS LS-DYNA, ANSYS AUTODYN, ICEM CFD, and much more. Users can serve their own licenses to CoresOnDemand.com or obtain a short-term lease, and PADT’s experts are on hand to help design the most effective licensing solution.

Pre-launch testing by PADT’s customers has shown that this model for remote on-demand solving works well.  Users were able to log in, configure their environment from their desktop at work or home, mesh, solve, and review results as if they had the same horsepower sitting right next to their desk.

To learn more about the CoresOnDemand: visit http://www.coresondemand.com, email cod@padtinc.com, or contact PADT at 480.813.4884. 

About Phoenix Analysis and Design Technologies

Phoenix Analysis and Design Technologies, Inc. (PADT) is an engineering product and services company that focuses on helping customers who develop physical products by providing Numerical Simulation, Product Development, and Rapid Prototyping solutions. PADT’s worldwide reputation for technical excellence and experienced staff is based on its proven record of building long term win-win partnerships with vendors and customers. Since its establishment in 1994, companies have relied on PADT because “We Make Innovation Work. “  With over 75 employees, PADT services customers from its headquarters at the Arizona State University Research Park in Tempe, Arizona, and from offices in Littleton, Colorado, Albuquerque, New Mexico, and Murray, Utah, as well as through staff members located around the country. More information on PADT can be found at http://www.PADTINC.com.

Taking NASTRAN Input Files Into ANSYS Mechanical via External Model in ANSYS 16.0

nastran-ansys-external-model-tnI found another very nice enhancement to version 16.0 of the ANSYS Workbench/ANSYS Mechanical toolset.  If you happen to have a NASTRAN input file (.dat, .nas, and .bdf) that you need to get into ANSYS Mechanical, no longer do you have to use FE Modeler in ANSYS Workbench to perform the translation.  In fact, not only can you move the NASTRAN model into ANSYS Mechanical, but you get the existing mesh as well as newly-created geometry that can be used for boundary condition application, etc.  As with most translations from one FE tool to another, you can’t expect everything will be translated.  However, this new technique can be an incredible time saver in addition to giving us capabilities to continue and augment simulations that were previously performed in NASTRAN, now in ANSYS.

Here is an example of this new procedure.  (Note that we don’t have NASTRAN here at PADT, so I couldn’t create a generic sample of a NASTRAN model in NASTRAN.  Instead, I created a model in ANSYS, then converted it into NASTRAN using ANSYS FE Modeler to get a NASTRAN input file for the purpose of this exercise.)

Once I have the NASTRAN input file that I need to convert into ANSYS Mechanical, I launch ANSYS Workbench 16.0 and insert an External Model branch.  I then click the … button to browse to the NASTRAN input file.  In this case, the file is NASTRAN.nas.

nastran-ansys-external-model-f1

Next, I drag and drop a new analysis type block into the Project Schematic.  In this case, it was a modal analysis.  Note that you can’t drop this onto the Setup cell in the External Model block as you might expect.  You set it up as a separate block and establish the link in the next step.

nastran-ansys-external-model-f2

Next, we drag and drop the Setup cell from the External Model block onto the Model cell of the Modal analysis block.  This establishes the link from the NASTRAN model to the new Modal analysis.

nastran-ansys-external-model-f3

We also need to right click on the Setup cell in the External Model block and select Update to get a green checkmark in that cell:

nastran-ansys-external-model-f4

Notice that there is no Geometry cell in the resulting Modal analysis block.  If all goes well, there will be geometry within the Mechanical model that can be used for selection purposes (in addition to the mesh that comes in from NASTRAN). 

Next we open the Mechanical editor by double clicking on one of the cells in the Modal analysis blocks (other than the Engineering Data cell).  It may take several minutes to bring in the NASTRAN model depending on the size of the NASTRAN model.  The Mechanical window doesn’t really let you know that it’s working, but if it’s sitting there with nothing being displayed, it’s probably churning away at bringing in the NASTRAN mesh and creating surface geometry on it.

Here is what the Mechanical window looks like after the mesh is read in and geometry is automatically created.  This is the mesh from the NASTRAN file, but in this case both solid and surface geometry is also present.  It’s not guaranteed that everything will come across.  I’ve seen contact elements come through for certain types of contact but not for other types of contact for example.

nastran-ansys-external-model-f5

The next image shows that geometry was created that can be used for the purposes of inserting fixed supports, just as if the geometry had come in from a CAD system.  Note that the NASTRAN input file had NO geometry, just finite element entities.  ANSYS is creating the geometry for use in Mechanical from the information in the NASTRAN input file.

nastran-ansys-external-model-f6

Finally, after manually creating a needed contact region, I was able to solve the modal analysis, demonstrating that further simulation can be performed in ANSYS Mechanical from this model which originally came from NASTRAN.

nastran-ansys-external-model-f7

So, the main take away here is that with version 16.0 of ANSYS, we can take a NASTRAN input file and through the use of the External Model block, go directly into ANSYS Mechanical.  Not only do we get the nodes and elements as well as other finite element entities from the NASTRAN model, but if all goes well we get geometry that facilitates further processing within ANSYS Mechanical.

We certainly hope this new capability makes it easier for you to perform additional simulations in ANSYS when the starting point is a NASTRAN model.  The other formats documented for version 16.0 are ABAQUS, Fluent input files, and ICEM CFD input files.