Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations MintJulep on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Abaqus units?

Status
Not open for further replies.

MichEng211

Aerospace
Jan 30, 2011
8
I'm new to abaqus, and I've got some simulations run, though they seem outlandish, and I'm wondering about units that may make my numbers incorrect. For example, when inputting a material, abaqus asks for the young's modulus, but it gives no indication as to what units I should give it. GPa? Pa? This goes the same for when I input forces/pressures - it asks for a magnitude, without specifying units. My model is inputted as a mm IGES - does Abaqus inherit these units? Should everything be a combination of mN and mm instead of N and meter?

Thanks
 
Replies continue below

Recommended for you

Asking this question indicates you may not have been properly introduced to FEM? If so, you should be careful about GIGO (garbage in/garbage out). There is usually more than meets the eye in a good FEM analysis.

Anyway...to directly answer your question, there are no units in ABAQUS, which is standard in FEM programs. You just have to be consistent with the inputs.

For example, a typical unit system is IPS (inches, pounds (force), seconds). If your modulus is psi, then the loads should be in pounds. Deflections would be in inches and stresses in psi. Same goes with any other system you might choose.

*On a side note, mass is a function of the system you choose (or vice versa) in case you do anything that would involve a natural frequency, etc.

Brian
 
I have taken a course where we were introduced to NX-Ideas, though it was a bit ago. I suppose I've gotten too used to CFD applications where units are nearly always specified.

I'm not too concerned with GIGO since I'm doing rather simple statics analysis - and FEA is a little easier on the new user than CFD.

Okay, so units are inherited. Thanks!
 
Dear, dear. Another case of "dumbing down" FE?

Think of it this way. You're doing a calculation on paper, and you have a series of equations that you use to calculate a particular variable - that variable may be stress, displacement, force, etc. The terms (mass, length, density, acceleration...) in the equations you use have to have consistent units in order that the resulting variable is calculated in a particular SI (or other) unit. This unit might be Newtons for force or Pascal for stress for example.

Model units in FE software are no different. All units in your FE model have to be consistent in order to produce results in a particular unit system, and it's up to you to ensure that your unit system is consistent.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com
 
Not to bash FEA, but you must admit it's a fair bit more forgiving than CFD.

I understand the model must have consistent units, I was just wondering where the starting point was. By inheriting units, I meant that, if the file I input is in mm, the resulting output from FEA will be on that relative order of magnitude and unit convention (english/metric); mPa for pressure and mN for force.

Just wanted to be sure. Again, the CFD applications I use generally spell the units for different inputs for you.
 
"Not to bash FEA, but you must admit it's a fair bit more forgiving than CFD."

Perhaps, since FEA has been around longer and may be more mature. If you mean that FEA can give a result easier (i.e. push some buttons), yes it may be more "forgiving". But to get a result that meets the today's expectations is probably no more or less difficult. Sure, there are some problems that are fairly simple and give the illusion of being forgiving, but there are others that have "hidden" traps that also give the illusion of being correct and at the same time are completely wrong. I have no doubt this would give the impression to someone not trained in FEA (and more importantly classical structural analysis) the impression that it is more forgiving.

As far "inheriting" units goes, this is somewhat true, though it would only be the length. And even then, it is a simple matter of scaling the model to put it in the system you want.

Some FEA codes do have units, but they tend to be more of the "push button" types that are associated with a CAD program (Cosmo, Mechanica, etc). The programs usually used by structural analysts don't have units (ABAQUS, NASTRAN, ANSYS, etc).

Brian
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor