Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Learning Inventor Nastran................... 2
- Thread starter CAD2015
- Start date
- Thread starter
- #3
Are you experienced in FEA? Then Nastran is just a tool. If you're unfamiliar with FEA, then Nastran is a lethal weapon. Any FEA-software package is a lethal weapon for that matter.
Huub
- You never get what you expect, you only get what you inspect.
Huub
- You never get what you expect, you only get what you inspect.
-
1
- #5
geesaman.d
Mechanical
I picked up Inventor Nastran last year because its part of our corporate license package and was getting little use. I taught myself based on my knowledge of Creo Simulate (and years ago, Ansys Designspace connected to SolidEdge). With that background it was straightforward.
Big picture: I work for a company that was bought by a larger competitor. I spent 15 years using Creo and Creo Simulate. Now the corporate mandate is Inventor and Ansys for simulations. However, Ansys was a big jump for various reasons so I explored Inventor Nastran before Ansys.
The user needs to understand finite element analysis. None of these softwares are acceptable for use by users who can run CAD software but have not had robust solid mechanics training. All of them require some sense of 'good' and 'bad' results as well as how to construct different variations of the model to establish proper model behavior and convergence. Even Creo Simulate, with its excellent convergence algorithms, sometimes does not give a well-converged solution.
The real difference is that Inventor Nastran and Creo Simulate are parametrically connected to the geometry. When your design needs to evolve based on the simulation results, Inventor Nastran and Creo Simulate are the right tool. When there will be no evolution (only pass/fail) or highly advanced simulation methods are required, you need to use a standalone FEA tool. Inventor Nastran and Creo Simulate do limited nonlinear studies.
Big picture: I work for a company that was bought by a larger competitor. I spent 15 years using Creo and Creo Simulate. Now the corporate mandate is Inventor and Ansys for simulations. However, Ansys was a big jump for various reasons so I explored Inventor Nastran before Ansys.
The user needs to understand finite element analysis. None of these softwares are acceptable for use by users who can run CAD software but have not had robust solid mechanics training. All of them require some sense of 'good' and 'bad' results as well as how to construct different variations of the model to establish proper model behavior and convergence. Even Creo Simulate, with its excellent convergence algorithms, sometimes does not give a well-converged solution.
The real difference is that Inventor Nastran and Creo Simulate are parametrically connected to the geometry. When your design needs to evolve based on the simulation results, Inventor Nastran and Creo Simulate are the right tool. When there will be no evolution (only pass/fail) or highly advanced simulation methods are required, you need to use a standalone FEA tool. Inventor Nastran and Creo Simulate do limited nonlinear studies.
- Thread starter
- #6
CAD2015 said:
I don't really know what a 'designer' is in this context. A firm I do work for has a bunch of 'design draftspeople' who are generally excellent are the roles and have many but not all have a reasonable idea of suitable sizing. But FEA is a dangerous tool in their hands, and I recently discovered that one 'design draftsperson' took it upon himself to verify his design of a vacuum vessel. (These vacuum vessels are done commonly but this particular one was outside of normal design parameters. He calculated very low stresses, documented it thoroughly and I'm sure was quite proud of his efforts.
However he had no concept that the strength of such a vessel has little to do with the stresses in the material and everything to do with the buckling behaviour of a thin cylinder under pressure.
FEA is a great tool, I lean heavily on it. However you have to have a good understanding of what you are analysing otherwise you can readily get yourself into trouble. (Same as ANY engineering task really.)
- Thread starter
- #8
Feel free to play with FEA. If that is a way you learn then you can become moderately competent in using the software.
But 'learning FEA' without understanding or learning about mechanics is like learning how to drive a car on a race track without knowing any road rules or having had any experience interacting with other drivers on the road.
Sure you can use the tool. But you will be a danger on the roads as you won't always use the tool in the correct manner for the circumstances. If you want the whole picture you'll need to understand how materials behave to the extent that is relevant for your intended use as well as loading condition and boundary conditions.
Eg; Modelling a steel girder in FEA is simple. And it will give you accurate answers for the stresses and deflections. But you'll likely have to know how to model a pinned connection and also perform a buckling check with suitable restraints to even start to do the job properly. There may be a bunch of other factors too depending on the circumstances. (There is a reason why 3D FEA generally isn't used for structural steel.)
But 'learning FEA' without understanding or learning about mechanics is like learning how to drive a car on a race track without knowing any road rules or having had any experience interacting with other drivers on the road.
Sure you can use the tool. But you will be a danger on the roads as you won't always use the tool in the correct manner for the circumstances. If you want the whole picture you'll need to understand how materials behave to the extent that is relevant for your intended use as well as loading condition and boundary conditions.
Eg; Modelling a steel girder in FEA is simple. And it will give you accurate answers for the stresses and deflections. But you'll likely have to know how to model a pinned connection and also perform a buckling check with suitable restraints to even start to do the job properly. There may be a bunch of other factors too depending on the circumstances. (There is a reason why 3D FEA generally isn't used for structural steel.)
- Thread starter
- #10
-
1
- #11
That all depends on your finances, time, skill level and your intended purpose and ultimately your responsibility in decision making. The most time efficient manner might be paid courses in Nastran, relevant engineering courses and advanced FEA courses. But given you are asking this question I'm guessing a paid course is not a strong preference.
For me I learnt Inventor on the job and then I learnt Nastran on the job. Formal training wasn't financially available initially (nor did I chase it either self funded or employer funded). I'm sure if I requested from my employer a paid course that It would be covered now, but I'm now sufficiently comfortable with the software I'm not sure the benefit is clear. (I have attended a general FEA course relating to structural steel.)
I slowly expanded my use of Inventor FEA and Nastran until I was competent and comfortable in applying in a wide variety of circumstances. But whenever I use it for highly complex tasks I usually approach the same problem using traditional numerical methods to check that the output is what one should expect.
For me I learnt Inventor on the job and then I learnt Nastran on the job. Formal training wasn't financially available initially (nor did I chase it either self funded or employer funded). I'm sure if I requested from my employer a paid course that It would be covered now, but I'm now sufficiently comfortable with the software I'm not sure the benefit is clear. (I have attended a general FEA course relating to structural steel.)
I slowly expanded my use of Inventor FEA and Nastran until I was competent and comfortable in applying in a wide variety of circumstances. But whenever I use it for highly complex tasks I usually approach the same problem using traditional numerical methods to check that the output is what one should expect.
- Thread starter
- #12
- Status
Similar threads
- Locked
- Question