FEA mesh settings

[Anonymous]

HI,

For my master thesis i am doing a FEA on a propellor blade. Normally when doing a FEA i would refine my mesh until i see no more change in results (stress/displacement). However on this propellor blade i keep seeing an increase in stress at the root of the blade starting at 25mpa (+/- expected result) i refine my mesh further and it keeps increasing up till 400Mpa (a lot higher then expected). The root of the blade is fixed and an angular velocity of 12600 deg/s is apllied. I  am assuming this has something to do with my mesh? All advice and tips are welcome.

i added a .ipt file.

Thank you!

[admaiora]

If there is a stress singularity, you can ignore it. (but before you have to recognize it)

[Anonymous]

You can ignore singularities, here's a really nice explanation of how to interpret singularities with a video. It's from the Fusion 360 blog, but the same principles apply. https://www.autodesk.com/products/fusion-360/blog/quick-tip-singularities-and-simulation/

[imajar]

I respectfully disagree with the previous posts.  It is true that often times, you can safely ignore singularities, and I frequently do.  But sometimes you cannot!  It is important that you understand when you can and when you cannot, in your case, I do not think it would be wise.

For example:  Consider a sharp corner:  FEA predicts an infinite stress, and in real life, there IS is a stress concentration, but since most construction materials are ductile, they can locally yield at the stress concentration and allow the stress to redistribute.  You are safer to ignore singularities for ductile materials, and you should be more careful with brittle materials. In this context, consider whether the application is a one time load, or if it is a cyclic load and you are studying fatigue.  Sharp corners are great places for crack initiation - Treat singularities with caution in fatigue studies, which yours is!

Usually, I will ignore hotspots if they are not part of the area I am interested in analyzing (but had to include to represent global stiffness).  If you don't care about the base the blade, and you are not specifying the method of attachment then don't worry about that area!  (On the flip side, if you are, then you may want to do 2 separate analyses, one general analysis for the blade, and a second detailed analysis for the connection).

Another source of stress hotspots can be in contact analysis (not applicable to your analysis), and with skill and understanding, those can often times be ignored.

The moral of the story is:  Don't blindly ignore hotspots unless you understand why they exist and you can justify why you ignore them!

[johnsonshiue]

Hi! Another possibility is that the geometry isn't in ideal conditional. Turn on Zebra Analysis, Cross Section Analysis, and Curvature Analysis. Though some places are smooth, there are a lot of bad continuities and curvature changes on the faces. The bad shape can lead to bad function, resulting poor FEA result. If I were you, I would recreate the part using simplified curves.

Many thanks!