upfront CFD is bridging the gap between the design and analysis communities

The word “upfront” seems to be the buzz word of many CAE companies. It is a word that can have a variety of interpretations. Many refer to it as “conceptual”, “over simplified”, “very early stages”. Upfront can have a much broader meaning than conceptual. It can certainly encompass the conceptual phase. But, there are numerous companies that do not have the luxury of doing napkin-sketch concepts. For these folks, the word upfront equates to the “design phase” or even the “re-design phase”. In other words, any phase prior to production- the earlier, the better.

proverb

If we look at upfront CFD, it was born with the vision to empower mechanical design teams with flow and thermal simulation tools. It is a methodology to help make the design process more efficient and unleash the power of design exploration. The concept of upfront CFD is to perform a ‘design study’, which is a design exploration of various what-if scenarios in order to explore design alternatives and compare these ideas side by side directly on your desktop. If the design is conceptual, great. If the design, is a re-design, that’s fine. If the design is a retro-fit, that’s ok too. How is this different than traditional CFD?

In the early days, those that were doing CFD, tended to be the large scale aerospace and automotive companies. In simple terms, it was for those that were pushing the forefront of large scale innovative design and needed a better way. They had very costly flow and thermal design challenges that needed to be solved and were close to impossible (size & cost) to solve in a physical prototyping environment. So, for them, their design process needed to be improved. But their process was vastly different from those in the SMB market. The big guys were dealing in a different $$$ bracket and could afford to be early adopters. Unfortunately, the CFD technology, computing power, adoption of 3D and life in computer-aided engineering was in its infancy. The breakthroughs were rapid, but it isolated the community of users that would even consider CFD.

The end result was a fracture between those that design and those that analyze. There are exceptional companies out there, that were able to develop a process where designers and analysts collaborate efficiently. But, in general, these are vastly different worlds. The experience and education is often different as well as the tools and methodologies.

2011 game 3 image thermal iso

Companies began to step out of this mold and multi-tasking design engineering became more common. Engineers/designers wore hats of all sorts. They were driving CAD, doing the testing, spec-ing materials etc. This community of engineers were the early adopters of upfront CFD. They felt “the pain” of not having the support of an analysis group, they realized physical prototyping was killing the bottom line and they were leaving innovated ideas on the white boards of abandoned brainstorming sessions. The multi-tasking engineering folks had a brand new way of thinking about CFD. Upfront CFD wasn’t simply making traditional CFD easier, it was changing the way engineers viewed and interacted with CFD. It was focused on helping improve the process of design. Allowing engineers to leverage their CAD models and receive design decision making information in minutes and hours rather than days/weeks or not at all.

In the early days, it made sense that multi-tasking engineers gravitated to upfront CFD. But, the reality is that all engineering organizations are different. Some subscribe to the philosophy of multi-tasking, others believe that individual groups is the way to go. The problem is not whether there are separate groups, the issue is about the process in which these groups approach design.

There has to be synergy in the entire process. Design procedures need to be established so that all members of the team can work as efficiently as possible. CAD models need to be modeled in a way that makes simulation a natural part of the process. Whether a CAD designer actually clicks the “solve” button or whether the provide the necessary design “variations” to the analyst. They have to be involved in the process and the education. Everyone needs to be onboard and approach the design process as a team.

Choke Valve 4 side by side with vector

Upfront CFD is about helping defining process within a single group, across multiple groups or within an entire organization. Eliminating redundant steps, such as the application of common boundary conditions and material properties is fundamental. Leveraging the rich information from the CAD model should be a rhetorical step. Getting you to the point that you can focus on design decision making should be the goal, not all of the steps to get you there.

CFdesign 2011 has tools in place to allow engineering organizations, to develop standards and best practices and can be deployed across the entire organization and allows sharing of these tools via the web. Sharing of process automation tools such as customized material libraries, design study templates, design study rules and even 3D interactive results can be securely accessed from the CFdesign Customer Portal. Upfront CFD is much more than simple conceptual models. Upfront CFD is an overall design philosophy than allows engineers to get more done in 24 hours than any other method.

1-Click Simulation in CFdesign 2011

One of my favorite new features in CFdesign 2011, is the ability to setup a CFD simulation in a single click. Yeah, that’s right, a single click. No smoke and mirrors. We built automation tools into 2011 that allow engineers to leverage the rich information that exists in their CAD models, to help drive their CFD design studies.

We are talking about the info that we, as engineers, deal with everyday –part names, material names. We have leveraged this information in the past, but we are making the concept of Rules that much easier and accessible. We are taking “set it, forget” it to the next level. You want to create a rule, simply right click. We built a Rule Manager that makes controlling the what rules get applied where, simple.

If you define your CAD materials in your CAD system, the materials inside your CFD system should automatically be applied, right? Suppose you do not apply material properties in CAD? Or you want to give a name to a component, like an inlet or an outlet. We implemented CAD-entity Groups. Simple tool on the CFdesign toolbar in your CAD system that allows you to name simulation specific information and allow Rules to take over in CFdesign.

 

One of the goals we were after in 2011 was to make upfront CFD a natural part of your design process. Reducing any barriers that you might see in order to allow you to perform design studies every project, every time.

Why simulation (FEA/CFD) is being slowly adopted…

If you ask 100 engineering organizations if and how they use simulation, you will get 101 answers. If you ask those that don’t use it, you will get 102 reasons why not. Seems odd that something that seems so closely tied to saving time, money and has the potential to create better products has such a wide array of success and standardization.

My opinion–simulation hasn’t been adopted in organizations due to lack of pain and vision. Many people spend a great deal of time talking themselves out of investing in new technology, rather than looking at their existing process and considering, “how could we improve our current process?”

If you aren’t feeling the pain currently or lack the vision to step outside the box and look to improve, you most likely will toe the line of status quo.

The sad part is your competitors aren’t doing the same.

Engineers should not fear “the cloud”

I’m going to go on record to state that I’m a huge fan of “the cloud”. They are two words that seem to really send people into a bit of rage lately. People want a definition, they want clarity What is the cloud? Call it what you want, I don’t care, I know what you mean and I want it. I want it now.

I think my friend, “the cloud”, has gotten a bad rap. It definitely took a ribbing at SolidWorks World 2010, where engineers were trying to make a drinking game out of it all. :) My sense is that its easy to make fun of something that you either don’t quite understand or that you don’t exactly see the value. I suspect its the latter in most cases.

If you have read anything on The Virtual Engineer, you will know that I am a huge fan boy of the good ole “cloud”. What is my definition of the cloud? In very simple terms it is web-enabled, web-based, web-hosted, web-stored, web-collaborated computing. Common theme here, “the web”. We all use it, many of us depend on it and some of are are even slightly addicted to it.

upfront cfd external flow simulaiton

So, why does using the word cloud cause anxiety? I’ve been thinking about this for quite some time, because I clearly don’t get it. Am I the minority? Not sure. But, it dawned on me the other day that I have felt the pain of “the old way”. If you haven’t, I’m not sure you will appreciate the value. The old way is a collage of bloated hard drives, sorry-ass USB external drives, various copies of files, handfuls of thumb drives full of stuff. I’m tired of being a VPNing, remote desktoping, begging for more horse power, tired of ftp-ing slave to status quo. I want a new way. I want to have access to my files, wherever, from whatever- laptop, netbook, iphone. If I want to share my files, I want to be able to send a link to someone with highly encrypted password protected safety. I want to be able to, at a minimum, preview my files in a browser. How can I be the minority? Doesn’t everyone want this?

But, the above is just the tip of it. I’m a simulation driven, upfront CAE doing, design guy. I am drinking a Big Gulp size of the the Kool-Aid. But, you want to talk about pain. I’m not sure there is a better reason today for engineers to openly embrace the cloud than in the simulation community. But not just the simulation community, the “upfront cfd” simulation community. Why?

  1. Design engineers tend to want to try numerous what-if scenarios, resulting in numerous back to back runs that ideally could run around the clock
  2. CFD models are getting bigger and bigger and are requiring more and more power
  3. Every design engineering department on the planet is trying to reduce cost of hardware, minimize capital investment- so paying for what you need, when you need it is a no-brainer
  4. More and more engineers find that they are on the go, out of the office and need access to their data, models, computing power from anywhere

I think we are on the cusp of a major break through in the way we engineer. There are a ton of things that need to be figured out. Security, licensing, bandwidth, hardware etc are simply just the basics that need to be worked out. Exciting to see some of the most talented people in the world working on it.

Upfront CAE should help drive design decisions

If you are leveraging upfront CAE (fea, cfd & the like) than kudos to you. You are most likely a step ahead of your competitors. I’d love to hear your candid experiences, both good and bad. There was a big push for upfront CAE about 10 years ago. CosmosWorks, Design Space and others really gave some street credit to upfront structural and some thermal simulation. Hats off to them. They did a great job waking up the engineering community and making “upfront” a reality in the minds of engineers.

Unfortunately, they took the “top down” approach. They were both respectively veterans in the CAE space that had “analyst” tools that were successful in their own right, and now they were changing their tune and exposing their technology to the design community. They found themselves in a very difficult place. They were taking what they knew and tried putting it in different terms. So, the end result was making FEA “easy to use” and making structural analysis “possible for design engineers”. But, I bet if you poll engineers and ask them how many are looking for an “easy to use FEA tool” or an FEA tool created for “design engineers”. I suspect, you will get a mixed bag. Let’s face it, most engineers are not looking for an FEA tool or a CFD tool, for that matter.

Engineers and their managers are looking for a tool that allows them to make better design decisions, reduce the number or physical prototypes, provide information whether a design passed or failed based on some failure criteria. Right? Most engineers could care less that its FEA or CFD. They expect and in many cases, demand that its easy to use. They also should not have to deal with status quo. They need tools that help them make decisions.

Where am I going with this? Don’t be fooled when looking to improve your design process. Don’t just get in line and assume that the “traditional analysis” companies or even worse, the MCAD companies, really understand how to help you make better design decisions. Each one of them is focused 100% in what they do. And what they don’t do is upfront CAE.

Traditional FEA/CFD companies are focused on working with analysts. MCAD companies are focused on 3D design. Each has a “solution” for you or so they claim. But, the reality is they are not focused on what you need.

There are very few companies that know and own the “upfront” FEA and CFD markets. Go out there, bang on google, cream rises to the top.

Don’t take my word for it. See for yourself. But, don’t believe the hype. Make sure you are making the right decision.

Upfront CFD remains a few steps ahead of the rest

I read a blog post yesterday attempting to explain Upfront CFD and then comparing it to traditional CFD and concurrent CFD. Yea, Concurrent CFD was a new one for me too.  A fancy word for “CAD embedded” CFD. It’s been around for a number of years and when you first look at it, sounds great!! A CFD tool that lives and breathes right in my CAD tool. Awesome! Immediately, there is a comfort level that everything will be ok and it can get the job done. For very simple (geometric and physics related) problems, where comparing one design to the other isn’t important,  this can be true. But even for the simple problems, it can come up a little short. Let’s take a look.

The Problem

MCAD tools have been around for awhile, they were built from the ground up to design 3D parts, generate 2D prints and the like. They can be fairly intensive programs with tons of features. Many companies have laid an FEA program on top of this foundation and have been fairly successful. SolidWorks Simulation is super popular for structural simulation. It does a great job for design level simulation. Structural simulation is fairly straight forward these days. Users tend to plot deformation and von mises (others as well) stresses to determine if something passed or failed. Often, you can get away with a simple static image to compare across designs.

In CFD, we tend to look at “pressure drop”, “peak velocity”, “max temperature” etc. We use these values as our pass/fail criteria. But this static data isn’t enough to tell you why one design passed over the other. A static image often just doesn’t cut it. This is where Concurrent CFD begins to break down. The CFD system is laid on top of the heavyweight CAD system, so having the ability and flexibility to compare models quickly and easily in 3D becomes impossible. You are now spitting out images and laying them on your desk, really hard to see what’s going on “inside” the model. This is one of the fundamental differences between CFD and structural FEA. The action of structural FEA happens on the outside (surface of the model). With CFD, the magic, often happens on the inside. In other words, having the ability to slice/dice and COMPARE in 3D is essential.

Often, users find that the honeymoon of CFD living inside their CAD system lasts for a very short time. Many of the CAD embedded programs are developed by 3rd parties which requires them to add functionality on top of the CAD framework. The result is that you have 20+ brand new dialog boxes that are popping up on the screen. So are you really in a familiar environment anymore? I’d argue that all of the familiarity and comfort of flying around in your CAD tool goes out the window. You are often locked down by wizards and forced to follow a specific recipe. This works fantastic day 1 of training, but many of us want to take off the training wheels and do it our way.

The Solution

Upfront CFD isn’t a new term to learn. It’s been around for 18 years. It has been developed from day 1 to address one simple purpose – to empower design engineers to solve fluid flow and thermal design challenges early and throughout the entire design process. Upfront CFD has all of the comforts of your CAD system- mouse operations are the same, layout is native and clean (works exactly the way you expect), CAD materials and attributes are read directly from the native CAD system as well as being fully associative. But the gem of Upfront CFD is the ability to properly do what-if scenarios.

The real impact of Upfront CFD is the ability to conduct multi-scenario design studies in a single environment. The environment is completely associative, you can clone designs or scenarios and update any change made to the CAD model. These clones are extremely lightweight, not simply entire copies of the previous, and give you the flexibility and power to do numerous what-if scenarios. Automation is one of the fundamental concepts of Upfront CFD. The data can be reused intelligently from one scenario to the next with a simple click of a button as well as being reused for future simulations.

multi-scenario design study environment

Once you have your defined results, whether they are critical values such as max temperature or pressure drop or 3D results, you can compare the data side by side instantly. Imagine having 4 designs that you want to see the flow behavior or temperature profile in 3D side by side? Click of a button in Upfront CFD, impossible in Concurrent or Traditional CFD. That’s the difference!

The Disclaimer

Many of you know this, but fair to state that I’m the Product Manager for CFdesign, so this may come across as grand standing. Not my intention. I am just trying to lay out some of the facts and dispel the myths. There is a place in the world for traditional CFD- PhD-type research problems, extremely massive models that require compute farms to crank out and when comparison is not important. This is a fairly niche market, but valid nonetheless. Same goes for Concurrent CFD. The market is even a bit more niche as it caters to very simple problems, that are fairly small in size and complexity and comparison is of little value.

Don’t take my word for it- go out there and check it out. Google it all- the facts are there.

Big Props to Pump Design Engineers

Started my Sunday, pretty typical, hanging with the kids etc. Thunderstorms had kicked in and we had one of the hardest rains in Philly in a very long time. Having a fairly old house by US standards comes with flooding wet basements from time to time. As you can imagine the clean up side of the story was less than fun.

I purchased a submersible pump (Water Ace Model R6S) model from Lowe’s or Home Depot a few years ago for just such an occasion. Fortunately, I have only needed to use it only a few times and it has saved me a ton of aggravation. It is amazing the volume of water that can accumulate in a very short time. I bet I pumped a few 100 gallons in less than an hour. It got me thinking of all the pump engineers I have spoken to over the years and it was really cool to use a device where some engineer somewhere in the world designed it so that it just worked. Maybe, just maybe CFdesign (or some CFD software) played a contributing factor on the design.

<span class=

Check out the pump above in action. Got me also thinking, about my expectations as a consumer of what appears to be a simple device. I want to just stick it in the watery/muddy mess and turn it on and miraculously it just works. Fact is, that is basically what has happened, every time I used it. But in reality, what appears to be a simple device is pretty sophisticated. It has moving parts internally, tight clearances and is expected to prime itself and pump “whatever”. Obviously the rain water gets mixed with the sand and dirt from my cellar floor as well as all kinds of debris it picks up along the way. There is a very simple screen on the suction side that allows the debris to collect in a pile instead of being suctioned into the impeller etc. But, watching the discharge end its pretty obvious that this “simple” pump is working it and pumping all kinds of stuff.

pump_pressure

Not trivial stuff to design, obviously. We have worked with the folks at Pentair Pumps (parent company of Water Ace folks) for a few years as well as many other pump manufactures.

Take, Cornell Pumps shown above for example, we have helped them reduce the amount of prototyping. There is a long list of things Upfront CFD can do to help such as providing flow rate, pressure head, performance curve data, efficiencies and torque to name a few of the basics. But where we really help is seeing inside the device, which simply can’t be seen in a testing environment.

At the end of the day, the cleanup was no fun at all. But it always brings a smile to see products designed by mechanical engineers that just work as expected, easy to forget the blood, sweat and tears that was put into the design.

A New Way of Thinking for the AEC Community: Upfront CFD Simulation

There is a good chance that where you are sitting right now there is some sort of environment control system– fancy word for HVAC or simply a heating and cooling system. It is an ever expanding business. Architects and engineers are constantly being pushed to optimize the air flow distribution and temperature values. It can be a fairly sophisticated engineering problem involving humidity, comfort temps, solar radiation among other things. It’s not getting any easier as there is more and more pressure to be green and maximize efficiency. It’s not only being required from a budgetary perspective, new regulations are now requiring it. In the past, often the fix was to simply over size the unit, manually alter the ducts, and even more and more rigging to solve the problem.
One would think that a problem that involves air flow, ducts, dampers, diffusers, grills, filters, blowers, solar loading, ever increasing heat generating electronics- that computational fluid dynamics (CFD) would be a no brainer? Well, it’s not really the case. In fact, up until recently it was used sporadically at best. Some outfits used it all the time, others rarely used it at all. But as with everything- times are changing. Now AEC firms are being required to provide simulation proof that the design will work and that the energy consumption meets certain requirements. This is sending folks scrambling to outsource the work, which is fairly costly. Others are investigating specific vertical very HVAC simulation tools. The problem is neither of these solutions really fits into their current process. Check out the variety of problems we are solving.

Historically, folks in this industry have used AutoCAD or Microstation. Their entire world has been born to think in 2D and expand into 3D. But with tools like Autodesk Revit making traction in this world, 3D is clearly a step in the right direction. There is really no benefit to going to a 3D MCAD CAD system, like Autodesk Inventor. But Revit now opens up an entire new world for this community. It is the launching point to really embracing CFD. Check out the videos below by CFdesign guru, Parker Wright. Parker has been fortunate to be right in the thick of things in this world. Working in the heart of NYC, he is hearing the evidence first hand right from the firms themselves. Check out his approach to solving the problem.

The below video shows the basic modeling inside Revit.

Here is setting up the CFD model leveraging the Revit geometry inside CFdesign.

Finally, reviewing the flow and thermal results in CFdesign.


I think we will be seeing more and more usage in this world. We are literally at the calm before the storm. The problems are only going to get bigger and happen more often. The good news is the technology and integration to solve the problem is growing even faster.

Inventor 2010 & CFdesign rocks

Everyone in the world blogged and posted about this a few weeks ago. I have been burning at both ends and neglecting the old blog. But figured it was worth a quick me too post. Check out the video below, it highlights some of the coolness.

Exciting to see our friends at Autodesk, jumping full bore into the simulation world. They are doing what thet do- taking tactful, well thought out implementation of the technology. Some might say a bit slow, some might say just fast enough. They have a dedicated audience of fanatical customers that can truly use some of the FEA capabilities coming from the Plassotech acquisition and expand out into the Mechanical Event Simulation stuff in Algor.

Personally, we have seen a huge interest from the Inventor community in flow and thermal simulation. The future for the Autodesk community looks bright, lots of cool stuff. Looking forward to the day when I am running flow simulations in CFdesign while sculpting my Inventor models with some of the Alias technology and morphing and pulling with Fusion. This day is not too far in the future

What does Upfront CFD have to do with Halloween?

Well, probably more than you thought. Little bit of fun here. I modeled a pumpkin in SolidWorks, brought it into CFdesign and ran a flow and thermal simulation to predict the air flow behavior due to the candle burning.

Obviously, this post is mostly in fun. But it does point out a few things.

  1. Pumpkins are fairly complex geometry, modeling the “seams” etc. Many meshers would require extra care in capturing that level of detail.
  2. No need to model a “flame” when you are simply interested in flow effects, how to “vent” an enclosure with an extremely high temperature.
  3. There are more and more ways that Upfront CFD can be used to improve even the simplest of designs.
Happy Halloween, everyone! Be Safe!