NAFEMS Registered Analyst 8

[waterice]

Hello Everyone,

I was planning on taking the NAFEMS Registered Analyst Scheme. I have recently graduated and intend to build a career in FEA.

I was looking for some advice on how recognized the Scheme is and any other experience from you that I should keep in mind?

Thank you,
Regards.

 

[corus]

I've heard of NAFEMS but never heard of this scheme to generate income for them before. I doubt employers would look at this when deciding whether or not to recruit you, but instead look at your real experience gained. Look at job adverts to see if anyone asks for a Registered Analyst, just as they would ask for, say, an engineer with chartered status (or PE in the USA). I doubt there is any at all.

Tata

 

[ESPcomposites]

Same here, I don't think employers would care much.

A bit off topic, but something to consider is that if you are just graduating, don't try to be a FEM expert right away (just my opinion). FEM is really only done well after you have mastered the fundamentals of the problems typical of your industry. This should be the way FEM is used, but the current trend at companies is to start young engineers with FEM too early.

I speak from personal experience and built some questionable quality FEM's early in my career. Almost every engineer I know who has jumped in too early has said the same. I see rampant abuse of it today's engineering world. I know it is tempting to generate great looking models and you feel like you can solve any problem, but the reality is usually different.

Good luck. FEM in the right hands can be a powerful tool. But if use it before you have mastered the industry basics, it has many negative side effects.

Brian

http://www.espcomposites.com

 

[rstupplebeen]

I personally would suggest the opposite. Get into FEA as soon as you can if you are interested in it. Make sure however that you land a job with a group of people willing to mentor you, provide training classes and attendance to conferences. I look back on my first position as being equally valuable to my under grad and graduate work. I hope this helps.

Rob Stupplebeen

 

[ESPcomposites]

Interesting difference of opinions. I can see your point as well. The trade that I usually see made is that young engineers do not learn the basics and use the FEM as a crutch instead. Learning to use a FEM software takes weeks, while learning to be a good analyst takes years. The temptation is to take the easy way out. If you CAN do both in parallel and you have a good mentor, then you are golden. I mentor young engineers on proper FEM use, making sure they understand classical solutions first. Then I start them on simple FEM and build them up. In reality, not every has a situation like this and engineers are told to run with a project or worse have a lead who does not understand good FEM practices.

What industry are you in? For space type applications, FEM seems more popular due to the unique problems. For aircraft, FEM is used for loads models and certain detail models, but classical solutions are usually a firm recommendation. The FAA does not seem to be very fond of FEM usage due to the inconsistent quality (i.e. user dependence).

Brian

http://www.espcomposites.com

 

[ESPcomposites]

Just thought I would mention another aspect. Recently, a space division was contracted to solve some problems. The group consisted of mostly brilliant people. The problem was they went automatically to FEM, without regard for classical ways to solve the problems. While their solutions were not "wrong", the suggested FEM methods were financially impractical. The approach was reworked with a classical approach and an increase in efficiency of at 10X was gained.

The point here is that even brilliant people can fall into this type the FEM trap. Just something to consider. I realize this type of discussion can go on and on. It is good to hear the opinions from multiple vantage points and industries.


Brian

http://www.espcomposites.com

 

[rstupplebeen]

I share your view of grounding in theory and hand calculation and that a bad analyst can still make a pretty rainbow plot.

I started with

http://www.materials-sciences.com/

working mainly with composites of naval vessels where I was one of many (15?) analysts. I was then the FEA and composites guy designing military helmets. Now I am the FEA, bio-mechanics and CAD guy designing ophthalmic products.

After my initial position I have utilized these forums to learn and teach to continue to hone my skills.

Thank you.

Rob Stupplebeen

 

[ESPcomposites]

Rob,

I think I see better why we initially differed in opinion. In your work, I don't think there are plentiful "hand calculations" for typical problems. Therefore, FEM may be the only way to solve many problems accurately.

If you are working on an aircraft, classical solutions are more powerful than FEM many times. But that is also because a tremendous amount of effort has been made to develop classical approaches.

When I work on a research project, I may simply have to use FEM as there is not prior solution. So the industry has a lot to do with how much FEM work you might do and if it is really warranted. Either way, you need a good foundation.

At the end of the day, we can agree that the important thing to keep in mind is that the FEM is only a software tool. If you don't understand the fundamentals of your problem, you cannot expect to get a good result from the FEM. Just be careful to not use FEM as a crutch, but rather as a tool to increase your your capability. Too many times it is used as the former.

Brian

http://www.espcomposites.com

 

[ThomasH]

Hi

I am working with FEM/FEA and have been doing it for quite some time. Regarding the concept NAFEMS Registered Analyst I have only seen it mentioned on the NAFEMS site. I have never seen in used as a reference or qualification requirement in the industry.

Don't get me wrong, I'm not saying it is bad. But I have never seen it actually being used in a resume or similair.

Regards

Thomas

 

[johnhors]

Brian,

I agree with your wise comments 100%

15, 20, 25 or more years ago FEA was a specialised task performed by a relatively small number of experts using unfriendly software running on hardware with very limited power and viewing results (if you were lucky) on a tektronix storage scope (green lines on a black background).

These experts were very resourceful in ekeing out the best solution possible given the limitations of the day, making full use of symmetry and sub-modelling techniques. A lot of thought went into a FEA analysis, and also a lot of checking since mistakes were very costly with long run times and the expense of computer run time in those far off days. These experts fully understood how to apply correct boundary conditions to a model.


Now this has been turned almost completely on its head with large numbers of users with questionable credentials driving analyses on overly friendly software, gaining what seems like almost instant results (compared to just a few years ago). Very often little or no thought is given to understanding the problem and ensuring correct boundary conditions are applied, instead millions of elements are blasted into a model in a mistaken belief that it gives a better result!

I guess the Nafems RA is an attempt to try and distinguish true analysts from the rest.

http://www.Roshaz.com

 

[waterice]

Hello Everyone,

Thank you all for your comments, thoughts, experiences and suggestions. I really appreciate it and feel that I should probably postpone the idea of becoming a Registered Analyst till it gains recognition.

But would you recommend that I get any kind of certification that would boost my chances of landing a good job in FEA?

Looking forward to your suggestions.

Best regards,
Sid.

 

[SAITAETGrad]

I have looked at the NAFEMS RA process in the past. Part of the program is submitting your work from real projects for peer review. Well, that seems like a great benefit.

A lot of employers won't recognize it but as NAFEMS gets more known (especially with underwriters, certification authorities, etc.) you can expect that discerning employers will see a lot of value in it.

There is no reason to put the FEA away...but always start out with hand calculations. You may find that hand calculations are also the place to end. Also, realize that validation of your FEM -especially with regards to constraints- requires heuristic knowledge of structures which you can only get through experience. Read everything you can find written by guys like Ian Taig.

Taig he has many great axioms.

"We are building up a vast capability for generating plausible nonsense faster and more convincingly than ever before." -1988.

"Nodes or Know-How?"

 

[ESPcomposites]

John,

Some excellent points there. Today's FEM codes present an interesting dilemma. It is so easy to use, which seems great. But this may attract those who are not good with a pencil and calculator. They may be more interested generating a stress plot. This is a shame because it hurts those who are truly qualified and damages the reputation of FEM analysis. The FEM software industry is also somewhat guilty because of attempts at "push button" analysis, which I have a strong distaste for.

As it stands, I am very suspect of any FEM analysis because of the generalized misuse. I demand to see complimentary hand calculations that demonstrate a certain degree of correlation. On the aircraft side, you might use a FEM to enhance your accuracy, but you should demonstrate the results are reasonable via hand calcs. At the very least, you will gain credibility and impress your lead.

The goal should be to call yourself an analyst who is good at FEM. Not a FEM analyst.

Brian

http://www.espcomposites.com

 

[chris9]

I became a Nafems Registered Analyst in 2004. Most employers don't know what it is or indeed ask for it in job advertisements. However, I think it has some credibility from people who have already heard of Nafems.

If you are an FEA jockey and don't understand how to do traditional hand calculations I think it is unlikley that you will be awarded this qualification anyway. You have to build up some formal training in FEA theory and practice (not FEA software training for ANSYS, Nastran, Abaqus etc).

These days, half the CV's out there have FEA on them so anything that reinforces your position as an FE Analyst can only be a good thing.

I say go for it!

Chris

http://www.value-design-consulting.co.uk

 

[loki3000]

<<Just thought I would mention another aspect. Recently, a space division was contracted to solve some problems. The group consisted of mostly brilliant people. The problem was they went automatically to FEM, without regard for classical ways to solve the problems. While their solutions were not "wrong", the suggested FEM methods were financially impractical. The approach was reworked with a classical approach and an increase in efficiency of at 10X was gained.

--------------------------
what was the problem about, if you don't mind me asking?

 

[chris9]

Loki3000,
I agree with you, I have seen this scenario many times. Sometimes it's cheaper to use FEA and sometimes hand calcs.

There are some very good engineers who have a wealth of experience but have very limited knowledge of FEA or they once had a bad experience with FEA and don't trust it. On the other hand there are people who are knocking out FEA results who have not yet grasped the basics of stress analysis.

In general, hand calculations are less accurate than FEA but on the plus side are easier to check and are more repeatable. Hand calculations tend to be favoured for safety critical work for this reason.

FEA on the other hand can give very accurate stress and is often the only practical method of finding stress for fatigue assessments.

It is important to have the ability to check FEA with hand calculations, also you need at least 2 years full time experience of driving an FEA software package to be productive.

The Nafems Registered Analyst qualification attempts to address these things.

Chris

http://www.value-design-consulting.co.uk

 

[4Pipes]

I am with Brian on the need to gain a wide and practical experience in your industry before you are locked away in an FE dept. I have seen several promising careers stall. FEA is a tool to do something else. A lot of folk treat for as a tool in for its own sake - it is not.

With regard to NAFEMS RA, you could say the same about many qualifications esp in the UK. However, it does not reduce the personal benefit you get by learning and going through the process. You still need the knowledge and confidence whatever you call yourself. Or then again, too much knowledge would in many cases put you at a disadvantage to the keyboard jockeys who can make everything work as if by magic.

 

[gwolf2]

I'm all for inexperienced people diving in with FE and getting garbage results which they believe. It is creating a huge range of products with serious fatigue problems which are just starting to emerge. I see this as a growth area for me over the coming years fixing designs.

(It is depressing to watch bad FE though, especially when you kindly point out that a certain FE approach will not work and you are ignored)

C'est la vie

 

[corus]

Another aspect is that finite elements is essentially a mathematical tool used by engineers that has a wide field of use. You may be an expert in carrying out thermal or modal analysis, but may never have used FE for structural analysis, for example. To say you are a Registered Analayst doesn't really have much meaning then, other than to say you're proficient in some aspect of the use of finite element methods. Do they also have a Boundary Element Registered Analyst scheme or a Finite Difference Registered Analyst scheme too?
I'd agree with some scheme that recognised the proficiency of people using FE but it needs to be more specific to have any use.

Tata

 

[ESPcomposites]

Something else to consider:

FEM can only give you stresses, strains, eigenvalues, etc. But it does not tell you what to do with these results. Even if the FEM is done correctly, the results can easily be misinterpreted.

For example, consider the local stresses around a hole. For a ductile material (typical aerospace metallic), the static capability is not significantly affected by the local peak stresses (not addressing fatigue at this point). However, I have seen structure sized (incorrectly) to these local stresses. Nothing was wrong with the FEM.

Another example is for a composite material. What do you do with the local stresses? If you use the peak, you may be too conservative. But you must consider the local effects as well, just not perhaps the peak stress (i.e. characteristic dimension approach). No FEM can tell you that. It can only report the stresses. If you were not familiar with the point stress criterion (or the like), you may not produce an acceptable result.

There are problems that rely on semi-empirical models (test data) to be accurate. Since the test data is integral to the mathematical formulation, a FEM usually cannot capture these effects. Usually semi-empirical approaches indicate a problem where the theoretical solutions do not yield satisfactory engineering results. So by the very nature of the problem, the FEM (on its own), won't properly solve the problem.

Without good understanding of the classical solutions, you would likely fail at producing an acceptable result for these scenarios. I can think of many more, but I don't want to bore everyone.

Brian

http://www.espcomposites.com