upfront


15
Feb 10

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.


7
May 08

On-demand "remote" Upfront Simulation?!?

There has always been buzz in the CAE industry for users to take advantage of remote cluster computing. Essentially, you would set up your models locally on a machine and run it on a server hosted by a vendor. Some of the old timers, like Fluent, have been in this business for a number of years. I tried their version about 5 years ago. Cool in concept, but the interface and process took some getting used to. I found it difficult to track the model’s progress.
The main problem or limitation on this capability will be the bandwidth. As internet access becomes more readily available and connections speeds improve, I think there is a future in “cloud computing” for Upfront CAE.

I currently do a mocked up version of cloud computing myself. I have a laptop and 64-bit desktop. In CFdesign, I am able to choose which machine I want to solve on directly from within the interface. So, this allows me to very efficiently setup and review models on my laptop and solve and store them on my desktop. I can even do this similar process remote. Takes a bit more copying and pasting and uploading/downloading but it works for now. Still trying to get Remote Desktop to work reliably. In the meantime, I will stick with GOTOMYPC. I used this all week while in Amserdam. Even on a fairly slow connection, I was able to access my dekstop and review a model’s progress.
Ideally, I would love to improve this process. Here’s my wish list. For starters, I would like to build models on my laptop, set them up etc and hit “GO” and have the option of what computer to solve on, even when I am traveling. I would like to have a “favorites” list of computers to choose. Some of us have one or two machines to use, others have a bit more. The file exchange would have to be super fast and lightweight. I then want a “link” to be created that I can access from any browser where I can monitor performance – convergence, mesh quality, mass balance, max temps, max pressures… I want the interface to have the look and feel of a modern web page that I can access from any computer or my phone!. To make life easier, I would like to have the option to have a text message or email sent with the link so I can have immediate access to it. I also want the option to text/email me when the model is finished, encountered an error or simply update me on the progress.
Version 2.0 should have a lightweight preview of the model (similar to JT or Adobe 3d) that allows me to pan, zoom, rotate and view my results. If I could have acces to all of the results, that would be great. But a lightweight model seems to make the most sense for web browsing.
This whole process would be great if I could do this on my “other” machine. But what if I had the option to do this to a remote farm of computers hosted by the software vendor? This way, I could have access to unlimited computing power, unlimited storage capacity (with a fee of course) and unlimited access to technical support in the event that something went wrong. A premium support model would exist so that I can “collaborate” with a support engineer using a variety of medias (texting, iming, email, forum type posts).
Think of the info that could be provided to the vendor on performance, mesh times and sizes, solver times, solver performance, application statistics etc.. The pluses are endless for all.
If I could add version 3.0 to my wish list – it would include driving optimization via this web interface. Cloud computing will only be appropriate for a select group of engineers. But I think the concept can give us endless possibilities. I am looking forward to seeing where the on-demand market takes the CAE industry.

12
Apr 08

Upfront CFD for Consumer Products: A Wake-Up Call

The more I do an unofficial market analysis, the more I think the potential for Upfront CFD is endless. There are all of the obvious applications for fluid flow and thermal simulation, but I find that often engineers do not even realize that they are having issues.

I was at an event a few years ago manning our booth. People were stopping by to see what all of the buzz was about. I gave him my 30 second pitch to what we do. I then said, “so do you have any flow or thermal issues?” He gave me the standard shrug and said, “no, not really. I don’t see any application for us.” Forever being stubborn, I said, “what does your company make?”. “Vacuum cleaners!”, he responded proudly. It was a defining moment for me. It is not in his mind-set to identify with “flow & thermal issues”. It is my job to ask, “what are your design challenges?” Once I asked this question, the flood gates opened and they are now a great customer.


I tell people that the applications for Upfront CFD are endless. I woke up this morning and within a few minutes, I identified 15 applications. My mobile phone alarm buzzed early this am. As I rolled out of bed, I heard the roar of a public transportation bus (valves, hvac, hoses, radiators, electronics, hydraulics) outside my window. I walked into the bathroom, turned on the faucet, looked in the mirror and saw the washer and dryer in the corner, debated about a shower and went downstairs. I clicked on my XM radio player, walked into the kitchen for some coffee and noticed the refrigerator, coffee pot, blender, oven, microwave etc. Strapped on my shoes and ipod to walk the dogs as I heard the heater kick on. Oh yeah, I forgot to check my email on the laptop. Point is begin to think about devices and products and and ask yourself is there a fluid flowing in or around it or does it need to heat up or cool down? If so, you can probably optimize the design using Upfront CFD.

The above images are only a small percentage of the 1000s of applications out there in the world. CFD has been around for along time, but has really been used predominantly by specialists. Often people in the Consumer Product world, thought CFD was too complicated or was over-kill for their needs. Times have changed — go see for yourself.


5
Apr 08

The "Avg Joe" Relates to Upfront CFD

I have been bicycle shopping over the last few weeks. I have hit a handful of the local bike shops. It’s becoming clear that the bike offerings are all similar – brands vary a bit, but in general, they are all very similar. Well, at least at the level of bike that I am looking for. The real difference is in the shop. In other words, its the support of the local shop that makes all the difference.

It is great to be educated by guys that are passionate about what they do. I sure learned alot in the last few weeks. One of the local shops sells Trek bikes. He was telling me stories about how focused Trek is on performance, not only on the structural integrity, but aerodynamics as well. Wow, finally we were now talking my language. I casually mentioned that co-workers of mine closely work with the Trek engineers to perform fluid flow simulations to optimize the performance of their bikes.

The tables were turned a bit, as now, the bike shop guy had a bunch of questions. After a few minutes of discussion, he said, “so, your software allows the engineers to do a bunch of ‘what-if” scenarios before testing them out in the wind tunnel? That’s really cool, they probably have the liberty to try all kinds of things.” It was really cool to see a guy just get it. He related right away to something that he knew very well and was able to see the benefits of doing simulation early in the design process.


22
Mar 08

Upfront CFD is Much More than Just a Pressure Drop

One of the major benefits of my job is visiting companies all over the world from various industries. Although I am excited about our product and services, my favorite part of the meeting is when I am granted a tour of their facility. It never gets old to walk around and get a view of the inner workings of a prospective customer. It’s like being a kid in a candy shop. Fellow engineers are super proud of what they do and get really excited showing off their designs. I get to ask a 1000 questions and learn a great deal about the challenges that companies face. The challenges are all very similar from a fluid flow and heat transfer perspective, even though the industries are quite different.

We typically get back to the conference room and now its our turn to show off how we can help their business. I love showing off our software. It is super visual and we can create really cool images that show some valuable insight that is impossible to capture in a test facility. The audience is almost always super receptive to the demo and usually have some really interesting questions. So, we go through and  show all of the various ways to understand the fluid flow and thermal characteristics of a representative model.

After a round of high-fives (not really), we get down to business and discuss how to move forward and implement Upfront CFD into their current process. A common question I ask is, "if you had an Upfront CFD tool right now, what is the first thing that you would want to simulate?" More often than not, a common response is, "we have this new design and we would love to know the pressure drop (deltaP)". This is usually followed by an awkward silent pause. Hmm…all the cool stuff that I just showed and you want a deltaP. I think it has to do with a certain comfort level. The deltaP is a common thing they get from physical testing, so they can relate to that value. We can absolutely provide an accurate pressure drop on your device. 

But once you determine the deltaP, then what? There seems to be a disconnect between the fluid flow characteristics and the pressure drop. But in fact, these two are obviously dependent on one another. If you optimize the flow characteristics, you can optimize the pressure drop. So, what does that mean? If you eliminate recirculation regions, reduce separation points and control the impingement regions, you can reduce the pressure drop. How can you do this with physical testing alone? Once you compute the pressure drop, how do you know what to change?

Proof is in the pudding -let’s look at an example. We’ll keep it simple.

 

CAD

We take the model directly into CFdesign and apply "what we know", let the software automate the rest and click GO! Let’s see how we can optimize the design.

cfd1

iso

cut1

Okay, so I am have a great time with Snag-It. But the point is that Upfront CFD gives you so much more insight to your designs than just a pressure drop value. The key to Upfront CFD is that you get the Pressure drop value, but more importantly you are now in a position to improve the design based on the knowledge that you have gained. Upfront CFD allows you to innovate, testing is there to validate. Based on what we have learned above, we need to reduce the pressure drop and attempt to "even out" the flow field in the area of the valve stem. The original design had 4 cut-outs, "what-if" we increased that to 8, 12, 16 – even 24? Let’s make the change and find out.

design2

But what impact does the 24-slot design have on the overall Pressure Drop? Let’s compare.

dp

We can plot the pressure drop and make an apples to apples comparison. We see that the final design satisfies all of our criteria. We were able to reduce the force on the valve, ensure that the flow is uniform in the vicinity of the valve stem and reduce the overall pressure drop of my system. I was able to do this in the course of a couple hours today. I ran 5 different "what-if" scenarios and compare. Upfront CFD not only gave me my pressure drop, but it allowed me to see from every angle what impact my design change was having on the performance.


11
Mar 08

Big Moutains to Climb in Germany

I just returned from a trip to Garmisch-Partenkirchen, Germany. It is about an hour outside of Munich. Great little ski town. They take great pride in the fact that they were the host of the 1936 Olympics. Seems like along time ago to me, but nonetheless, they are proud.

I spent most of the time in meetings, but was fortunate enough to walk around for a bit. We also had some fantastic Bavarian meals in some of the authentic restaurants in town. I also snapped a few pics. Check out the image below. I was in awe of the shear size of the moutains. Sure, we have the Rockies in the US, but they seemed small in comparison. I found it interesting that some of the people in our meetings shared the same sentiment about the moutains, others seemed mildly impressed. I guess it all depends on what you are used to?

It is great to have the opportunity to meet German companies and understand their processes and challenges. They do business slightly different and definitely take great pride in their engineering. But many German engineers relate to the benefits of Upfront simulation. It is exciting to show people how to approach fluid flow and thermal challenges in an Upfront way. I am looking forward to my next trip to learn even more about this fascinating place.


9
Mar 08

You Must be Committed to Implementing Upfront CFD

We in the vendor world are committed to our products and services. We take great pride in our capabilities and we strive to provide first class support and training to ensure that our users are successful. Let’s face it, a major part of our day is focused on ensuring that our customers are gaining a return on their investment. I take full responsibility on transferring the knowledge it takes to implement Upfront CFD into your current process.

So, if we assume that all vendors have a similar passion in their products and services, why do so many engineers have so many CAE products sitting on their shelves? I live by the statement, that “the customer is always right”. But I do expect that the customer is willing to first identify a problem and be open to working together to formulate an implementation plan, but most importantly, the customer has to be committed to the implementation of the plan. Seems like a simple concept, right?

It has been my experience that the customer and vendor have to have a solid relationship and absolutely have to be on the same page regarding the implementation. This is where selecting the correct vendor is essential. Ease of use and the laundry list of features are a given. But as you shop around, you will quickly see that everyone will claim to have essentially the same “laundry list”.

What you need to ask yourself, “do I have confidence in the team that I am about to partner with”? Do they have the support infrastructure to help me when I am stuck? Do they understand my business? Do they understand my problem? But most importantly, do they know how to teach me to solve my problem?

Being committed to the implementation is your job. Formulating a clear implementation plan that teaches you to solve problems with our tool is our job. Don’t be afraid to ask the tough questions. Whether or not a vendor can match your test data is really irrelevant. Given the correct inputs, every vendor will stick their “Ace” user on it and present really fancy images. The real question is how will the vendor transfer the knowledge to you the user so that you can seamlessly begin solving problems?

29
Feb 08

Engineering Grad School: MBA vs. MSE?

Great article in this month’s Mechanical Engineering magazine. It gets into the heart of making the decision about continuing to grad school after engineering school. It is a really good article. It gives an unbiased view of the pluses and negatives of two separate paths. Definitely a must read for new graduates.

I have been out of grad school for a number of years and found the article very interesting. I look back on my decision to take the master’s in engineering path as one of the smarter decisions that I made early in life. At the time, quite frankly, I was just looking to stay in school as long as possible. I was having a great time in engineering school while paying the bills bartending all through college. The thought that I would be able to sleep until noon, attend grad classes in the late afternoon and not have to pay for it, was a no-brainer for me.
I didn’t appreciate it at the time, but graduate school taught me some of the advanced fundamentals of fluids and heat transfer, but most importantly, it taught me how to take a conceptual project from idea to completion. It was also my first exposure to CFD. I dabbled in small development projects, tons and tons of lab testing and a fair share of “traditional” CFD. Although, I am not sure that at the time, it was fair to call it traditional. It was simply cutting edge at the time, but way out-dated by today’s standards.
It really gave me a great appreciation for “the process” of simulation. I really believe it kick started my career path into Upfront CFD. I am looking forward to what the future brings both for me and for things to come in the Upfront Simulation movement.
I always thought about what would have happened if I decided on an MBA instead? Looking forward to hearing others experiences.

26
Feb 08

An Underground Subway Stop is a Great Platform for Upfront CFD

HVAC applications keep popping up more and more. Anything from new construction of “green” buildings to retrofitting 100 year old subway systems.

Here is a model I worked on recently that allowed us to optimize the airflow inside an existing subway station. We were able to try a variety of what-if scenarios — with and without cars, with and without people as well as various heat loads (best and worse case scenario).


10
Feb 08

Efficient Design is the Key to Success

I mentioned Desktop Engineering Magazine in my last post. There is a great article by Amy Rowell that discusses the importance of design. The point is that it is extremely important to push the envelope on design regardless of the industry you serve. Pushing the envelope can mean many things. For some, it may involve the industrial design department, others may need to tweak a % or two to gain extra efficiency and others may need to ensure that their thermal management can accomodate the latest and greatest coming from the EE’s. No better example of a company that pushes the envelope than Apple. They have raised the bar so we now expect them to “wow” us. Difficult position to be in for a company, but they are certainly reaping the benefits.

We speak to many engineering manager’s all over the world everyday that acknowledge the fact that competition is tough and they would like to expand their foothold on the market. Unfortunately, they are already stretched thin and are struggling to keep up with the demand of their existing products. It is doubtful that hiring additional engineers is on the shortlist. So, doesn’t it make sense to optimize the efficiency of your existing team? I know people are cringing as this is the perfect time for a sales pitch. Bear with me.

Today’s engineer is multi-tasking. Time management is probably the biggest obstacle. Personally, I go through an exercise of jotting down where I spend my time each day. No fancy tools, just pen and notebook. I try to see where I am wasting time, not necessarily spending the most time. I am trying to identify tasks or areas that can be eliminated or automated.

So, where does Upfront CAE come into play? Begin to look at your process. Are you spending time, building and testing? Is it a fast and inexpensive process? I doubt the answer is yes. Are you spending time supporting customers after the fact because your products are failing in the field? Are there areas of your new designs that you really think would have an impact, but there just isn’t enough time in the day to try them? If the answer is not an definitive, NO, then you should at least talk to somone about Upfront CAE.

To start, my advice, ask the sales guy how he envisions incoporating Upfront CAE into “your” process? Often, we get very tied up in how software works click for click. This can be a huge time sink. It is important to understand how it can be implemented first. The rest, usually falls into place.