Best software to learn? 1

[ST111]

Hello. I am wondering if there is one software that is best to learn over others? Or is it really relevant? What I mean is would future employees be that bothered about not knowing a specific software, and more interested in the fact you understand finite element analysis and the underlying principles?

Thanks

 

[ESPcomposites]

To be the best engineer, you should first understand how to do classical analysis very well + be good at basic FEM knowledge. Learning the software is trivial by comparison.

However, some employers just look for keywords and translate their presence (often incorrectly) into the assumption that you have a solid background. Since that is the way it sometimes is, you should research the field you are looking for employment in and find out what software they use. In the aircraft industry, NASTRAN and ABAQUS are popular and so is PATRAN. However, for some industries, those software packages may not be used at all. There is no one size fits all answer to the "best" software for getting a job. The same could be said about CAD packages.

Brian

http://www.espcomposites.com

 

[ShellsPlatesMeshes]

Here is a book, supplied with FEA software (including complete source codes), which is based on systematic approach: from the mathematical fundamentals of FEA to practical implementation and testing:

http://www.amazon.com/Computational-Geometry-Surfaces-Application-Analysis/dp/0646594044

(the authors also have a separate webpage with more information:

http://members.ozemail.com.au/~comecau/quad_shell.htm

)

Hope this helps.

 

[IRstuff]

I would simply pick a software package that's reasonably well represented in the user community. The rationale would be that any seriously competitive package would have 90%+ of the features that the other packages have, so learning one is equivalent to learning 90% of any other package. Since all such packages all so about the same thing, their functionality and command structures will be similar, i.e., "form follows function."

Obviously, you should still have a strong understanding of what the basic math is supposed to be doing, so that you'll recognize what pitfalls you might encounter.

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529

 

[rb1957]

what is "standard" in your field ?

another day in paradise, or is paradise one day closer ?

 

[ThomasH]

Hi
To me this sounds like a variation of the "best software"-question. And it is hopeless to give a valid reply without a lot more information.

But regardless of what software you use, don't let it become a "black box". The software and the computer can handle a huge a amount of data at high speed. But you need to tell it how to handle the numbers.

Like "rb1957" said: What is tha standard in your field". That is usually a valid starting point.

And never ask the wonderful question," I have this problem, what software can solve it?"
Or am I being negative?

Thomas

 

[gravityandinertia]

The first step is to start.

If you read the link I created above, it is written for web developers, but it is true in every field. My advice is to find a free/cheap/open source FEA program and try to run an analysis. A real one. Check it with any sort of experimental data/ hand calculations that you can. Are you in the ballpark? Post specific questions on forums like these. Do your own research online, read books, take a training course.

Regardless of field, I always recommend starting simple and cheap. You need to learn the basics before you worry about fine details and advanced topics. It's like people in golf with a terrible swing who don't practice fidgeting with driver weighting systems. It's not going to make that much difference. 1. You have no repetition in your swing 2. You don't know what a good swing feels like yet. Give Tiger Woods a pair of clubs from Target and me the top-of-the-line, guaranteed he will still destroy me.

You're in the same position. You don't know what you don't know. It's time to get started, only then will learning start and truth be revealed.

P.S. I recommend doing some sort of real project. That's the best way to learn. Document it too.

 

[IceBreakerSours]

I agree with everything that has been said above. However, let me provide some context:

Depending on your interest, skill, stage in professional life, resources available (time/money/supervisor's blessings, etc.), you may fall somewhere in the spectrum ranging from a designer who wants to be able to communicate with an FE analyst, to a card carrying FE analyst, all the way to a developer. Designers are the closest to products (i.e., real world) and developers are the furthest (i.e., general abstract world); developers have the strongest hold on what is inside the 'black box' and the designers the least.

Think about these things, where you want to be in a given time-frame with your resource constraints you are operating under, and then consider the suggestions made above.

It is likely that you may not want to be a developer (otherwise, chances are, you would not be asking these questions). So, if you have resources, then register for an excellent FE course or two at a top school so you are 'forced' to write a mini-solver or an element. But that assumes you want to be an FE analyst. If you'd rather just be familiar with the basics, then self-reading from a resource like solidmechanics.org and watching Klaus Jurgen Bathe's videos on YouTube will be more than sufficient.

*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083

 

[ST111]

I want to be an analyst, I have worked on university projects in the past with FE software and industrial projects in various sectors (automotive and consumer products for example) so I understand the basics of the 'black box', I have experience with linear and non-linear analysis but I feel I need to put my theory into practice more!

 

[stressebookllc]

Personally, I believe you should get real good at modeling, meaning using the proper elements to capture the stiffness and response of the structure, how to interpret the loads along the load path etc. This really has nothing to do with a particular software, all reputable programs essentially are very good at what they do depending on the type of analysis.

In aerospace for example, if you are hell bent on knowing about software, Patran is a big dog for pre-post, and nastran is a very big dog for the solver.

Maybe that is a good start. Other than that you need to get really good at free body diagrams, and classical hand calculations in aerospace, which will help you avoid costly mistakes people relying heavily on software alone may make.

Aerospace Stress Analysis and FEA Courses

http://www.stressebook.com

Stressing Stresslessly!

 

[ThomasH]

Hi
One possible option is to look at job advertisements and see if a possible future employer has any preference regarding software. I have never seen anything like "Experience in ABAQUS/Nastran/Ansys/? is required", only things like "Experience in ABAQUS/Nastran/Ansys/? is valuable".

If you want to learn a software on your own time I think that may be a problem even if I like the initiative. You can probabry find something like a "Student Version" or "Evaluation Version" for free or at a low cost. But the full version is usually to expensive to buy without having the ambition to use i tprofessionally.

Good Luck

Thomas

 

[stressebookllc]

I have seen jobs that require experience with certain FEA programs..

Aerospace Stress Analysis and FEA Courses

http://www.stressebook.com

Stressing Stresslessly!

 

[IceBreakerSours]

If I were in a job interview I would want to know IF/WHY the hiring manager is hung up on facility with a particular UI. If they do not make serious sense to me, then I'd know a little about the maturity level of computational modeling as a strategic investment at the company and will make an informed decision about the offer, if it is made.


*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083

 

[stressebookllc]

It is quite common in many aerospace stress engineering jobs to ask for experience with PATRAN/Nastran for example, its a critical skill set.

Aerospace Stress Analysis and FEA Courses

http://www.stressebook.com

Stressing Stresslessly!

 

[ThomasH]

Yes, and if an aerospace company planed to hire an experienced analyst they might expect him/her to know PATRAN/Nastran. But if they hire somebody without experience, would they require a particular software? The OP wants to be an analyst, so far not in a specified industry. At least as I understand it.

Last time I got the offer for a new job they wanted my knowledge in analysis and FEM, not in a specific GUI. Where I work we have a few different softwares. I only use one but there are a few options.

Best Regards

Thomas

 

[LRJ]

gravityandinertia wrote: 'My advice is to find a free/cheap/open source FEA program and try to run an analysis.'

Could you - or anyone else, for that matter - recommend some free FEA programs that could be used for validation? If there is a list someone already with some reviews that would be great to 'sticky' in this sub-forum.

 

[gravityandinertia]

LRJ,

I have a couple ideas for you. Cheap is relative while free and open source are pretty cut and clear. For cheap I would recommend SOLIDWORKS Simulation. It is in the thousands of dollars but it isn't in the tens of thousands of dollars like the big boys of Abaqus, LS-Dyna, ANSYS.

Depending on your CAD software, there are different recommendations. If you have SOLIDWORKS, their Simulation product is the easiest to get started I have seen anywhere. I do work for a company that sells it though, so I'm warning you I'm biased. Bias aside, it removes a lot of the details (or options) that you aren't likely to understand until you have some experience with FEA anyway making it a good tool to start with. If your company has a budget for stuff like this, I recommend investigating it with your local SOLIDWORKS Reseller.

A couple other options are SOLIDEDGE with FEMAP. There is a 45 day trial available for both if you register on the website. Though to use it for projects it's going to be expensive when it comes time to purchase. Abaqus has a node limited version as well I believe.

Besides those, you can find many options at [link feacompare.com]FEAcompare.com[/url]

 

[IceBreakerSours]

Do you want a code that is cheap OR a code that is cheap, easy to use, offers plenty of support (documentation, examples, online forums, etc.), something you'd like to put in your CV because of job postings you've seen, .. ?



*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083

 

[ESPcomposites]

If you are looking to get a job as an analyst (and not as a designer), I am not sure using a CAD based FEA tool is attractive to employers. In my experience, it is not. This is not to say they don't have their place, but an employer wants you to demonstrate knowledge of a fully featured FEA tool (and the ability to work with an input file/deck and the cards). As mentioned, CAD based FEA has a low learning curve so that skill can be learned OTJ if needed. Its not something that stands out on a resume. Skill sets with steeper learning curves, such as fully featured FEA software (ABAQUS, NASTRAN, ANSYS), are going to be more attractive to employers. Those skill sets are also relatively interchangeable (i.e. knowledge of one of those greatly reduces the learning curve for another).

Brian

http://www.espcomposites.com

 

[LRJ]

Thanks for the recommendations. I'll check those out.