What do you guys think about the recent announcement that Ansys will buy CFX from AEA?

From my perspective it must be good - Ansys is a profitable company and integration between Ansys, CFX and perhaps ICEM CFD's tools would really make it more easy to do coupled simulations etc. I'm not a great fan of Ansys' user interface though.

Perhaps this new ANSYS/ICEM-CFD/CFX troijka will be able to give Fluent a match for the top position in the CFD sector?

Ansys has a very strong position in many important CFD markets, for example, in aerospace. This will surely be good for CFX in the long term. Or what do you think?

I agree. ANSYS is producing many new products but still remaining focused on quality analysis tools. Becoming a top player in the CFD market is a natural choice. The combination of ICEM and CFX will make for tough competition.

I also agree, the ICEM capability will enhance CFX. Other markets should open up for the code, especially if the junction is as good as the CFX / Tascflow merger. It would be good to see Fluent put under some more pressure. CFX5 has progressed very quickly and very positivly over the last few years (it needed to to keep up!). I think the code is already showing great prospects as it develops into a more mature state.

My only worry is that development does not progress as it has done up until now and that the excellent support that is offered by the staff at CFX is reduced. You only have to speak to people who use multiple codes and they will all tell you (from what I have found) that CFX's support is the best. Lets hope it remains that way !! Bob

I am interested in hearing more about CFX's support. In my opinion, having a superior code is somewhat expected but superior support is usually harder to come by. In your opinion, what does CFX do specifically that makes their support better than others?

The answer to that question could be worth a lot of money ))

Ansys is focused on one thing - making money (not that that is so bad). They have not developed much of anything new themselves since John Swanson left-they buy it all. I don't think they will sink the type of development money into CFX that it needs to stay up with Fluent or STAR-CD. My guess is that they will try hardest to sell this to where they are making the most money with Ansys - low end Windows based users. I'm sure that they will do some interesting things to try to integrate with DesignSpace but I would be surprised if they advance the the state of the art of CFD. Its certainly not what they do on the FEA side.

I see your point, but I meant it more in a general sense. What does good support entail? What are customer's expectations in regards to contacting a vendor for support? Some feel as though support should cover everything from licensing questions, to debugging models to suggesting approaches to a problem. Where is the dividing line between what is considered support and what should be considered consulting?

cheers,

derrek

Hi Derrek, I'd say CFX do all of what you suggested. There is an element of When is support consultancy. However they are never shy of offering good advice on an approach to a problem, and have always been very quick to help with a tricky problem. The support people do expect that you have tried to solve the problem yourself using their advice. However if that does not solve the problem, then they always (in my case) will look at the model for you. I do know what you mean though, and this very good setup that CFX have and provide should not be abused. I personally always try hard to solve a problem first, then ask CFX for their advice, then only when that is drawing a blank do I send the model. Even then I always ask if there are things that I can be trying in parallel to what they may be doing. Aas a result I feel I have a very good relationship with the support personel who are very willing and able to help.

As for the Ansys selling statement made by steve, may be you're right ? I don't know ? however does Fluent not sell its code very agressively ?

Company policy for quality support is intrinsic to CFX. Support is very central to their main concerns. Support procedure are fairly comprehensive and undergo regular customer and internal reviews I understand from friends at CFX. Consultancy is also a big business, and needs good support and expert staff.

There is nothing wrong at with aggressively selling a code. That is how you stay in business. But my point was that Ansys's recent history says they don't invest heavily in state-of-the-art numerics. The recent additions to Ansys have to do mostly with somewhat peripheral items. They are not investing much in faster solvers or distributed computing or non-linear stress analysis. They are doing things mostly to make Ansys more attractive to low end users. I don't know what their plans are, but I would be surprised if they handle CFX any differently. I don't think you can be that complacent with CFD in the long run. Its a much younger and less mature field than FEA. Look what they did with Flotran all these years. Nothing.

Steve, I used Flotran back in 1995, is it still as bad ?? No I agree with you 100%, the CFD market is very cut throat. CFX have almost completed a successful (in my mind) come back from a position wher they were definitely lagging behind the competition. Lets hope for all our sakes that what you suggest does not come true !

Steve:

I totally disagree with your perception of ANSYS. They are continuously investing back into R&D, improving element technology and continuously improving the solvers. Look at Design Space, for example, the initial motivation was to expose designers to FEA technology. They then took that one step further and built this tool inside their own revolutionary application toolkit, Workbench. Workbench is literally changing the way people view analysis tools.

Look at all of the new products coming out of ANSYS. DesignXplorer, AGP, FEMXplorer. I don't know of any other company that is offering these tpyes of tools.

Yes, they do acquire companies to expand their capabilities, but I consider that a natural progression. THey are not just acquiring companies to expand their sales potential, like MSC. They are not late to the game offering designer level tools, like Abaqus.

I am interested in hearing your thoughts.

Derrel

Let's face it, Flotran has always taken somewhat of a backseat at ANSYS. It was always considered as a complement to ANSYS, not necessarily a stand alone CFD tool. It is used by companies that do some CFD work and have ANSYS for their structural/thermal/emag etc needs. Great tool for the price. But never intended to be a competitor to the big three, and should never be compared to them. I think ANSYS' acquiring of cfx shows ANSYS committment to entering the CFD market with full force and taken on Fluent and CD.

derrek

Didn't Ansys also buy the CFD code from Centric several years ago? What ever happened to that?

Derrek, I don't mean to say that Ansys does no development - but all the programs you list are not particularly things that advance the state of the art of FEA. Its true that they may have brought linear elastic analysis down to the designer, and even done that well, but people have been doing linear elastic analysis for 30 years. I have been an Ansys user for more than 20 years and I have not seen all that much change in the last couple of years for what I do - and I don't do design. All I am saying is that they are focusing their efforts on lower end users and apparently making a lot of money doing it. They have a business model that works well for them, but I don't think it will play the same way with CFX. Perhaps they won't take the same track with CFD. I am just taking my best guess.

Integrating CFD with FEA hardly sounds like a move geared to low-end users. Robin

Steve:

I apprecitate what you are saying but just don't agree. With the exception of Desing Space, where is the focus on lower end users? A new interface, along with a toolkit to customize the interface and give you the ability to build your own isn't geared towards to the lower ended user by any means. The expansion series technology and other optimization tools are definitely not lower ended? But hey, everyone is entitled to their opinion.

derrek

Derrek, Actually, I disagree. The toolkit is just meant to make it easier for companies to integrate their own processes into Ansys or the reverse, whichever way you like. I am not against this at all, as it makes sense for many places. The focus seems to me to be on integrating what already exists more into a client's design process than pushing the envelope on FEA. Certainly a low-end user is not going to use the toolkit - but what will happen is the one good programmer or analyst at a company will use the toolkit to extend his knowledge to lower end users. However, its just not the same as doing cutting edge work on plasticity or coming out with highly parallelized solvers. Maybe FEA is so mature that you don't have to do this kind of thing anymore? I don't know and I certainly have no idea what Ansys' plans for CFX are either - I'm just guessing. Steve

Steve,

I suggest you read the writing on the wall. ANSYS now owns ICEM-CFD and CFX. What do you really think will happen? The top 2 guys at ICEM have left the company and all of ICEMs consulting group use Star/Fluent at the moment. ICEM has made a habit about not aligning themselves with a flow solver in the past. Now they are aligned whether they like it or not. This really smells like some coming, very positive, change to me.

I think that over the next few years you will see ever tighter integration with CFX and ICEM products which will eventually turn them into the leading CFD package on the market. Between the CFX-5 mesher, tetra and hexa they have some of the best meshing technolgy, they have the best flow solver, and the CFX-5 post processor is nothing to turn a nose up to.

I don't know much about the FEA technology in ANSYS but certainly the CFD technology in CFX-5 is the most state of the art on the market today. If ANSYS wants to continue the mometum that CFX-5 is showing then certainly they will have to maintain the investment levels that were maintained by AEA in the past few years. So, from that standpoint I am not worried at all about what ANSYS will do with CFX and ICEM, I'm quite optimistic.

Neale

Neale,

I think you are reading the wrong wall. First of all, ICEM supports something like 100 formats and CFX users are only one (and I'm sure a small minority) of all their users. Do you think Ansys is going to kiss off the revenues from more than 50% (probably 80%) of ICEM users just to focus ICEM on CFX? I don't. Certainly not any time soon since those revenues are paying for what it cost to buy ICEM.

Next, do you think that ANSYS is going to deliver CFX and ICEM together at the same discounted prices that CFX alone has been selling at for the last two years? I hope you are sitting down when you get your first renewal bill. ANSYS does not discount.

I have no experience to know if CFX is a good solver or bad. I have a little bit with ICEM and they are OK. But I have been an ANSYS customer for a long time and I do know how that company operates. They are led by business people, not engineers, and they are driven to make money. If selling to the high end makes it they will, but if selling to the low makes more they will do that. The thing is, at least in FEA, there are many more low-end than high-end customers around. Maybe they will approach CFD the same way or maybe not.

Perhaps this discussion should continue after they have had CFX for a year. Then none of us will have to play guessing games.

Steve