How can I verify that FEA analysis(like autodesk inventor and solidworks) match real world? 1

[NewEngineer652]

Hello All,
I am learning how to do FEA analysis on autodesk inventor (using you tube videos) and I see that blue, yellow, red lines on the body, when I apply a force on a body and constrain it. For example, I have a wooden table on inventor, I fixed the the four legs to the floor and now when I apply the force on the top center, I see where the stresses are high and where they are low.

I have two questions here,
1. how does the stress values help us? how can we use these stress concentration points in real life?
2. can I somehow prove that the stress analysis match the real life situation? any simple experiment? maybe I can break something cheap?

Thanks,

 

[IRstuff]

1. See your second question. Your job, as an engineer, is to ensure that the stresses don't wind up breaking or fatiguing the part.
2. Technically, FEA doesn't "match the real life situation." FEA is an abstraction, an approximation, of real life. It's close enough that we can live with that. Nevertheless, breaking something is a great experimental approach. Another possibility is to use polarized materials that can reveal their stresses through changes in their polarization.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com:

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

 

[NewEngineer652]

Thanks IRstuff:
I am very new to this so I have more questions.

1. I can see that each color represent specific value of stress. Does the red color mean it will break at that point?
2. So next step would be to go to material properties, find the yeild strength, use some safety factor and compare the allowable stress to the stresses shown on the FEA model?
3. can you give an example of situation where FEA has helped you(or someone) significantly ? In other words, I want to know the real benefits of using it.

 

[GregLocock]

Here's ways we correlate our models

1) strain gages - if your predicted and actual strains match for the same loading you are on the right track

2) modal analysis - if the mode shapes are right and the natural frequencies are right it's a bleeding miracle the you have probably managed to get the masses and the stiffnesses right

3) dynamic response - if you can apply realistic inputs dynamically to your model and get realistic vibrations (this is an extension of 2) then you are doing very well

4) fatigue life - well in theory you might be able to predict fatigue life for a complex structure. Good luck

5) static stiffness - not a bad check, at the very least you know your boundary conditions are right.

as regards 2)- in about half a day I could generate enough modal data to invalidate the FEA of a car body or powertrain.




Cheers

Greg Locock


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

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

 

[mmar2087]

1. I can see that each color represent specific value of stress. Does the red color mean it will break at that point?
Of course not, it is just a visualisation of how stress developes in the part.
Often the autogenerated "Von Mises" stress plot sets the red band to the highest stress level in your result (rather often a singularity caused by bad fixtures)
You can set the color band to what value you want and also note that there are other stress plots also.

2. So next step would be to go to material properties, find the yeild strength, use some safety factor and compare the allowable stress to the stresses shown on the FEA model?
If you're really confident that your setup and your assumptions are correct, and yeild strength is the allowable criteria, then yes. But based on your questions I feel that you need to know a lot more about FEA to be confident in your setup. Best way of getting the confidence is to perform a couple of hand calculations and then setup and compare results from FEA.

3. can you give an example of situation where FEA has helped you(or someone) significantly ? In other words, I want to know the real benefits of using it.
Strange question. You see it everywhere in your ordinary life. Engineers use FEA and hand calculations to decide how cars, buildings, aircrafts, consumer products e.tc. are designed.
I just assisted our design deparment, in a development project for an external customer, with FE Analysis. They worked with weight reduction, on a high stressed bracket and with my analysis we could remove even more material, compared to their original propsal, and in the same time reduce the stresses in the critical areas of the part with 40%.

 

[rb1957]

personally, my first step would not be to use Inventor for stress analysis (as it is a very limited analysis program) and not to look at the colours and think "everything is green, so that must be ok".

how do we know FEA is a good tool to use ? 'cause we've understood the theory it's based on, and tested things enough to know (and to cautious) that we're (probably) ok.

how should you validate your model ? take some predictions from the model, test it and see.

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

 

[IRstuff]

"3. can you give an example of situation where FEA has helped you(or someone) significantly ? In other words, I want to know the real benefits of using it."

I suggest that you ask your own senior engineers, since that would be for whom you would be doing the FEA. But, let's reverse the question; if you didn't have FEA, how would you determine whether your design meets requirements without building it only to find that it doesn't work, for whatever reason.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com:

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

 

[GregLocock]

Also of course you can do hand calculations and check that the FEA gives the same answer. This is all crucial stuff, anybody who relies on FEA models for real design work and has no correlation method in mind is either very confident or ignorant or both.

Personally I've found it doesn't matter much which FEA package is being used, for linear FEA, the faults are almost invariably wetware issues.

Cheers

Greg Locock


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

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

 

[EHudson]

You can use brittle lacquer to visualize strain.