×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

Test results vs FEA results.
5

Test results vs FEA results.

Test results vs FEA results.

(OP)
We wanted to study the variation of a gusset/rib in a clamp. We tested a sample and it was bend out of shape for a load of 200lbs. I need to redesign the geometry to improve the load using FEA.
When i modelled the geometry , FEA has predicting the stress to cross the yeild stress at 10 lbs  at one point. and the stresses in the neightbouring regions gradually decreased to value much lower than yeild strength.
In such a case what should be criteria on stresses to decide the load for making the clamp to bend.
Is it I need to take the section at the highest stress point and take that load that makes the least stress in the cross section cross the yeild stress. I want to make a decission based on linear static analyis only as i donot have stress strain for doing non linear anlysis
Regards
Kamalakar.

RE: Test results vs FEA results.

(OP)
The stress i talked above is not a stress due to singularity.

RE: Test results vs FEA results.

2
I'd almost give you a star, at long last somebody new is getting on the bandwagon - real world test results are very hard to duplicate in FEA.

Sadly I can't give you an answer, because I don't do stress analysis.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

RE: Test results vs FEA results.

2
kamal11,

You will have trouble duplicating this test using purely linear, static analysis.  As this object yields locally, the stress will redistribute.  Linear stress will not properly redistribute the stress.  If you look for "net section failure" (where the lowest stress across a section is yield), you will likely be conservative, but there is no guarantee.

You do describe

Quote:

the yeild stress at 10 lbs  at one point
  If you refine the mesh in that area, does the stress level change much?  Is the stress located at or very near a boundary condition?  Is the yield point that you have in the model at or near the real-world yield point?  There are many things that effect FEA results.
I have duplicated real world tests on many occassions...including non-linear situations.  It isn't always easy, but it certainly isn't impossible.

Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.
Lower Alabama SolidWorks Users Group

RE: Test results vs FEA results.

I'm not sure why you need finite elements for this geometry. I assume that the gusset/rib is a T section. if you calculate the section properties then you can calculate the expected bending stresses by simple My/I, where y is the largest fibre distance. You may need FEA if the geometry is any more complicated and also to check local buckling of the part. If you're wanting to compare your FEA results with measured results then you'll need to use non-linear analysis methods if the section was permanently bent out of shape as you say. If you don't have the stress strain behaviour then simply use a bi-lienar curve with plasticity occuring at yield.

corus

RE: Test results vs FEA results.

i agree wiith corus, i think FEA will be too difficult.  i'd attack the problem from a different direction.  

you know where the bracket failed (and where it didn't).  why not add material where it failed (and remove material where it didn't, if you're picky about weight).

that, and a little (structural) commonsense should help.

btw, how much do you want it incease the strength of the part ? ... 10%, 100%.  also, remember that this part is a component of a structure ... maybe strengthening this part might not be all you have to do (maybe the failure will migrate to the other parts of the structure).  also, how well did the test fixture represent the rest of the structure ?  maybe it was too stiff, causing the loadpaths to be not representative.

good luck !

RE: Test results vs FEA results.

The first thing I would look at would be the displacement plot deformed. The animate it . Does the movement make sense?  The displacement plot can show poor constraints.

Finite Element Analysis is faster and more accurate than handbook methods. CosmosWorks and SolidWorks are easy to use. However, there is a learning curve.

Finite Element analysis is accurate. However, assigning boundary conditions can cause incorrect results.

When in doubt, mesh fine. The only penalty is run time.

Your problem sounds simple. A FEA consultant might solve it for a reasonable cost. It is worth a quote.

Hope this helps.   

RE: Test results vs FEA results.

Good grief.

"Finite Element Analysis is faster and more accurate than handbook methods"

Let's see, deflection of a cantilever. delta =W*L^3/(3*E*I).

I think we can assume that I'll have finished long before your FEA program has even launched.

"CosmosWorks and SolidWorks are easy to use." and like many easy to use tools are of limited use.

"Finite Element analysis is accurate." Yes, the solver accurately solves the problem you have set up. Whether that bears any resemblance to reality is another thing entirely.

"When in doubt, mesh fine. The only penalty is run time." No. The optimum mesh density is not infinite.

"Hope this helps.   " well it made me grin.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

RE: Test results vs FEA results.

Well Greg, you did assume a uniform beam and ignored shear deformation, so you rather weighted the "contest" in your favour!

But yes you are quite right, the statement "When in doubt, mesh fine. The only penalty is run time." is rather scary ! Throwing zillions of elements at a problem, doesn't automatically improve a simulation, or magically correct poor boundary conditions as the words of the previous poster seemed to imply!

RE: Test results vs FEA results.

How many people will do the convergence test on their meshing before they reach the conclusion that mesh is fine enough?  

RE: Test results vs FEA results.

I think that Gregg deserves a star for his post. I don't know the geometry of kamal11's hardware but a lot can be done without the reliance on FEA. Cheap FEA and PC's have only been around 15 years or so. We did accomplish many things before that.
I once asked a young ME how his analysis results compared to a hand calc estimate and he was dumb founded that anyone would even consider doing that. His professors (apparently anyway) never taught him to question the results. I don't believe in the old adage to question everything-I could never see how you could actually accomplish anything if you questioned everything-but FEA results need to be questioned.
Bob

RE: Test results vs FEA results.

PC based FEA started in 1980 and it is absolutely not perfect, but this is an FEA forum.  Instead of acting like the tool is useless, let's foster an understanding of how to properly use the tool to simulate reality.

I'm all for hand-calcing, but some structures require a significant amount of simplification or an incredible amount of time to produce an answer that may suffer from many of the same problems found in FEA.  I have a master's degree in structural analysis and love to analyze things, but I use FEA as a tool so that I can remain cost competitive.  To date, nothing that I've designed using FEA has catastrophically failed, and tests tend to fall within 10% of my predictions.

I've seen many of those dumb-founded looks, but instead of criticizing the wearer, I prefer to educate them.

RE: Test results vs FEA results.

In FEA I do dynamic analysis, not stress analysis. Sometimes I do static deflections as well, but again, not stress analysis.

My background is that I worked in FEA for a few months, then switched to running a modal analysis lab, which is used to provide the FE guys with their primary correlation data. Since then I've worked in both fields off and on for 20 years. So, yes, I have seen many horrible things done with FEA, but /I/ don't do FEA stress analysis. Nowadays part of my job is to prepare the loads for static and dynamic strength analysis for components, but again, someone else creates and runs the models.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

RE: Test results vs FEA results.

A few comments (in a thread that's already contentious!):

"CosmosWorks and SolidWorks are easy to use." and like many easy to use tools are of limited use.
Careful now...  If a tool is made easier to use, it is not to say that it is not capable.  However chances of potential misuse increase because of the broader user base.  I solved non-linear (material) problems with similar software which e.g. NEi will only be capable to handle in its next release (results compared favourable with Abaqus results).

"How many people will do the convergence test on their meshing before they reach the conclusion that mesh is fine enough?"
I'll dare to say I don't do this every time.  Simply because one does develop a feeling & experience and it also helps if you understand the particular element formulation.  Because of many other assumptions I also don't hunt the 3rd decimal.

I agree with most of GBor's comments.  And, yes, it is extremely important to ALWAYS do hand-calcs; at least as a sanity check!

Kamalakar, consider using a non-linear analysis with bi-linear material model (a la corus).  Alternatively linear calculated max stress values can be reviewed (lowered) using approximate plastic correction as described in the book "Practical Stress Analysis in Engineering Design" by Blake.  It involves hand-calcs and is an estimate only.

Cheers
Gert

RE: Test results vs FEA results.

Doing a numerical convergence test should be as necessary and routine as the hand calcs, IMO. Each step, the hand calcs (which aren't just at the beginning of the analysis; I find hand calcs useful throughout an analysis) and the numerical convergence tests are important parts of the entire design and analysis process.

So add my '1' to Garland's '1' for a grand total of '2' who are admitting they do convergence checks every time. Of course I use a (h)p-version FEA code, and it's a lot more efficient and easier, not to mention one gets better convergence properties, than anyone using an h-version FEA code.

RE: Test results vs FEA results.

Back to the original topic, one thing that always seems to hold me back is the lack of time to do a correlation of the experiment and analysis. The other major thing is lack of actual data, of course, which will always be a problem--you can never have enough detail in, say the photoelastic stress measurements, to make thorough correlation of the measured and analytical engineering data (strains, stresses, deflections).

RE: Test results vs FEA results.

I ask the question to do the convergence checks. Because some of my co-workers don't do that, they decrease the element sizes as much as possible, and let the computer run for the few days (unnecessary computer cost), in order to ensure it is convergent.

RE: Test results vs FEA results.

John,

"they decrease the element sizes as much as possible  .... in order to ensure it is convergent" - sorry but unless you actually do a convergence test it is unsafe to assume that your mesh meets the criteria. This is especially true with auto-sizing routines which I have seen throw 100,000's of nodes and elements unnecessarily into low stress regions, but still leave critical areas with too coarse a mesh, despite the massive size of the whole model.

Prost,

Very often it is not possible to locate a strain guage on the exact point of interest as indicated by FE. Photoelastic measurements can also have their problems (depending on the method used) in critical regions. For these reasons we also locate strain guages away from any obvious stress raisers in what we call field stress areas to assist in validating our FE models.

RE: Test results vs FEA results.

johnhors: because strain gages seem to work well only in areas of small to no strain gradients, and work not so well in the really interesting areas, the high strain gradient areas, then they would appear to be of limited utility for correlation with numerical analysis. I have only limited experience with correlation of strain gage readings with FEA, so perhaps others could enlighten the rest of us ignorants.

Photoelastic coatings work very well in many many cases. There is also a lot of research in thermoelastic measurement techniques that use infrared to detect tiny changes in temperature which sometimes can be correlated very well with strains and stresses.

Every measurement technique, as well as every numerical method, has problems, so I wouldn't discount any technique just because it has limitations, I would only make sure those limitations weren't distorting my results.

RE: Test results vs FEA results.

Prost, you may also enlighten me about photoelastics - I'm under the impression it is only useful in the elastic range of the photoelastic material?  So, in context of this thread which is about substantial plastic yielding of metal, will the photoelastic material typically have sufficient range to correlate in?

Thanks

RE: Test results vs FEA results.

photoelastic material need to be transparent? So. light can pass through.

RE: Test results vs FEA results.

Typically, I don't have time for runs to check for convergence. I check the "energy norm error." (the difference between nodal and elemental stresses)  My usual practice is to refine the mesh, with mesh controls, finer and finer until the meshing and run time slows.

"Finer the better" with H elements draws negative comments,  However, I have not heard a refutation of this practice.  In meshing finer and finer, I have seen a realistic appearing stress gradient  change to  a   stress field containing spots of high stress that make no sense. This occured with a  very fine mesh;  (Global mesh/100)

In one reply, and in reality, inexpensive FEA tools are criticized as simplistic or useless. I never hear why, just that these are "toys" .  Without naming names, I have seen excellent correlation between  a lower cost and a higher cost package in a complex linear stress analysis of an assembly.

In the world that I have worked in for 30 years, engineers do not have master's degrees in engineering mechanics,  and companies will not invest in the high end FEA packages.

Often, checking with hand calculations is not useful due to simplifying assumptions. Checking with a simple FEA model is one technique. Perhaps a more useful question than checking with hand calculation might be
"Did you do a free body diagram?"

There is a  value in using FEA  in that the problem must be defined.  The user must think about the boundary conditions.

RE: Test results vs FEA results.

Quote (FEAHappy):


Without naming names, I have seen excellent correlation between  a lower cost and a higher cost package in a complex linear stress analysis of an assembly.

It is not difficult to program the code for doing a linear stress analysis without meshing capabilities. Actually, there exist free FEA tools which are far of being simplistic.

RE: Test results vs FEA results.

"In one reply, and in reality, inexpensive FEA tools are criticized as simplistic or useless. I never hear why, just that these are "toys" .  Without naming names, I have seen excellent correlation between  a lower cost and a higher cost package in a complex linear stress analysis of an assembly.
"

No, I said of limited use. Don't misquote me. I did not say they were toys.

Tell me, if they were as useful as full powered products, why would people pay for the full powered products? All you have to do is compare the feature-list, typically, to see what the differences are. At least one of the bundled products cannot handle any complexity in boundary conditions, and makes dangerous assumptions about them, and I doubt any of them can deal with assemblies.

So far as 'toys' go FWIW at home I do the vast majority of my FEA in the 200 node demo version of nastran for windows. And get very good correlation- typically within 5% of the correct (measured) static stiffness. But, I have full control of the model.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

RE: Test results vs FEA results.

Quote:

Tell me, if they were as useful as full powered products, why would people pay for the full powered products?

I think it is generally support.  If you are only doing linear, static analysis, there are a variety of GREAT product ranging from free to $20k that will all do a great job, but if you have any questions about how to run them, the ones that cost generally come with better support.

Quote:

All you have to do is compare the feature-list, typically, to see what the differences are.

Glad you added the word "typically".

Quote:

At least one of the bundled products cannot handle any complexity in boundary conditions, and makes dangerous assumptions about them, and I doubt any of them can deal with assemblies.

There are a few that handle assemblies depending on what you mean by "bundled", what you are trying to do with the assembly, and how you create the CAD model.

I think the "moral" to this story is, "Prepare yourself with the knowledge of your own limitations and the limitations of your tools or those for whom you design your products may pay for it with their lives"

Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.
Lower Alabama SolidWorks Users Group

RE: Test results vs FEA results.

Hi,
Garland and Greg, I personally think that you both hit good points.
While it's true that in the linear elastic, static, domain (and not only, I'd say...) the FE solvers behave all almost the same way (hopefully!!!), be them hi-end ultra-expensive systems like Ansys Multiphysics (but then, why get the Multiphysics license level if it's only for linear static?) or 100%-free packages, when it comes to capabilities in handling "exotic" boundary conditions then the music changes: for single example (and only to name systems I directly know or knew), up to v.2004 (and perhaps v.2005) CosmosWorks couldn't handle non-holonomic boundary conditions, while this capability has been existing in ANSYS at least since v.5.3 with COMBIN37 element. And the list could be VERY longer... Of course, the model you have to analyze can be as simple as you want, but as Greg says if you need a particular boundary condition and your package doesn't handle it, you'll have no chance to get any realistic result, I think.

RE: Test results vs FEA results.

If you supply analysis work for certain (large) companies, you have no choice about what analysis software you use to do the job.

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources