Price of FEA software?

[bongirs]

Hello,

I am considering starting my FEA consultancy.
So I would like to know the prices of FEA software packages.
It would be great to understand the fixed initial seat price as well as the yearly price of the packages.
I have tried searching a lot using Google but in wain...

 

[glass99]

rick: thanks, I know a bunch of folks that work with Ansys and love it. From my perspective Ansys, Nastran, and Abaqus are all pretty amazing. I guess I wish they were a bit cheaper so it wasn't such an investment to try them out. Really high end analysis where Strand7 chokes only comes up a few times a year. We used to use NeiNastran with FEMAP at my old firm, which was pretty good. But 20k/yr for software only makes sense if you are using it every day.

 

[bongirs]

Greg: Thanks for suggesting the idea of making specialised apps in addition to the FEA package. I gave a thought to it and now I have some more specialised areas like gears, cams, fasteners, etc. to ponder upon.

Rob: As mentioned previously, I have a long term plan of setting up a consultancy. So right now I am not looking at really accurate figures. Just rough estimates to understand how much I would have to invest and whether or not my business would be able to sustain. Overall I understand that FEA alone could possibly not sustain my business. I have to work on some related activities (Rapid Prototyping maybe?).

Rick: Yeah the software is going to be a huge investment. So I am seriously contemplating developing my own software. Maybe also provide ANSYS consultancy side-by-side. ANSYS is really famous compared to Abaqus here in India. Also I would never sustain the business with these high investments as people want everything cheap over here. FEA itself is not used much apart from the big MNCs and OEMs.

glass99: I am starting FEA activities alongside Product Engineering in my organisation, to support my teammates using Creo Simulate. I have been having a very steep learning curve with Creo Simulate especially after my experience with ANSYS in college. Recently it failed to deliver a very simple multi-body simulation despite having very short intervals for non-linear analysis. The mesh was mostly prismatic but not very fine. Anyways cant expect much from a CAD software to deliver extra-ordinary FEA results.

 

[rstupplebeen]

Unfortunately, I can't post prices on a userboard. Please email me directly and I can get you connected with the appropriate people.

rob at optimaldevice dot com

Rob Stupplebeen

http://www.optimaldevice.com

 

[Fvibtech]

I was looking for a cheap FEA software and they recommend me Heron rotordynamics in another forum. It's cost less than 100 usd per month! depends of the plan and it's online (Link or info(at)heronrd.com).

Regards.

 

[GregLocock]

Just before you get booted off the site, perhaps having a website that is not accessible without a password is not the best way to drum up business?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[SergioP]

Hi, I'm in charge of the FEA team in an automotive company, so we deal (and suffer) with that numbers, please send me a PM if you need the real numbers. Guess that is not possible to post it directly.

 

[4Pipes]

I'll just rock the boat a bit in the spirit of balance! I would agree with the general points about Autodesk Simulation. However, I would not write off Autodesk just yet. I purchased Algor at the height of the crash at the end of 08. It was as much as a toy as anything and because I could afford it. Abaqus and Ansys are total non-runners for individuals unless you have a near full time use which pays top dollar. Since then I have been very impressed with Autodesk and kept the licence paid up. Autodesk are still expanding and it now includes NEI Nastran from August 14. I suspect that Autodesk will get their break at some time soon. End users of FEA services will wake up at some time and realise that they do not need to pay for work done with Abaqus and Ansys. Neither guarantees that it will be correct. Big name consultants with big name software can and do use screwy boundary conditions with very pretty but equally screwy models to produce meaningless rubbish. The same rubbish can be produced with less famous software at a fraction of the price. If one wants to deal with clients who only want big name software and are willing to pay for it, then there is no point considering Autodesk Simulation just yet. However, there are other more price sensitive clients. You pays you money etc.

 

[GregLocock]

I am not quite sure why you'd need to pay for software to produce rubbish results, but it's your money.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[IDS]

I know Greg was joking, but I think 4Pipes makes a valid point. For the great majority of applications reasonably priced software will produce just as good results as the software from the big names. The results need to be independently verified by some means anyway; relying on the name of the software supplier doesn't sound like a great verification method.

I'd like to see more price competition from all software suppliers, and easily accessible fixed prices would be a good start. Whether Autodesk will be useful in driving prices down in the long term, I doubt.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[GregLocock]

Sure for simple structures. I'm a great believer in simple FEA models, and have correlated three 100-300 node models very successfully, at least to a level that demonstrated their usefulness - that is where static deflection was an issue we got good agreement on spring rate, and for dynamics we got good modal results within the frequency range of interest.

Most people don't correlate their models to any serious extent, they just want pretty pictures.






Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[bongirs]

Just want to make it clear that if it is illegal to discuss the actual price of the software then I would not want to know the actual figures.
I don't think there is any crime to discuss rough estimates as done above and am satisfied with the information.

Regarding the reputation of the software I believe only one principle: Garbage in, Garbage out.
A tool is only as good as the operator is.

 

[glass99]

greg: 100% agree about simplicity in FEA. The 100,000 to 10,000,000 node model is for confirming the 200 node model, not for first instance design.

I also 100% agree about testing/correlation, though that can be harder to get through the project management folks. Question always arises: why am I paying you to do a fancy model if you have to do testing too?

 

[4Pipes]

I was just being tongue in cheek. I just get a bit cynical about the way folk talk about the big names as though these were the only things that could produce decent results - certainly at the level that covers +80% of the market. I have tried to reverse engineer calcs in the past - basic stuff so nothing really fancy. The only one real difference was that well a known user of Abaqus was not able to run one of my linear buckling models because it was too big. In fairness, I am not totally sure it was the fault of Abaqus. Maybe Abaqus has small and large versions or the user made a mistake - I don't know.

In another case, I was already on a full time contract at the company but not doing FEA. I would have done the FEA within my existing rate so giving them the cost of Autodesk licence just to be able to get them as a reference. Internal politics and software snobbery ended up with the work being subcontracted to experts. The end cost ended up at nearly $0.5m. Not one model was geometrically correct and none had the correct boundary conditions. I was happy with what was built because I had designed it and checked it with Autodesk Simulation first. The client was ripped off by nearly a factor of 10 on price even if it had been done properly. Frustrating but such is life.

 

[4Pipes]

Bongirs, Sorry, I got a bit off piste. The price you want is best obtained from a software supplier. Get to know them and get budget prices. Then ask for the basic 30 day licence to be extended to give more time for review. Suppliers are desparate to get market share. They are not going to charge top prices for small users. If they think you might be a good bet and are moving into markets that they are interested in, then you will get a good price. Keep them guessing - in the nicest possible way. They are only trying to do a job. List prices or prices charged to larger organisations could be meaningless.