Nastran in CAD BOOK GUIDE -info

[Anonymous]

Hi @John_Holtz  , hi @wasim_younis  , hi Nastran In CAD team and Nastran in CAD users,

I have bought the @wasim_younis  "Up and Running with Autodesk Inventor Nastran 2020".

I like it and I have the Inventor series Up and Running too.

There are a lots of little errors in it, and I think it wasn't been reviewed very well, and it's a shame because there are interesting cases!

There is one error (in my opinion) that is pretty big (the model meets or doesn't meet the design criteria).

The following is for the ones that have the book and the authors.>>>>>

______________________

It'a about the ABS tank.

Page 212: it's written (about doing the mesh): "Select Continuos Meshing"

With that setting the author has:

Von Mises 3.7 MPa
and a FOS=5,4 >5 and meets the design criteria (the tank is ok!)

And that it's not true.
With that setting the readers and the users will have:

Von Mises about 6 MPa
and a FOS=3,33 <5 and that DOESN'T MEET the design criteria (the tank is NOT ok!)

Also the stress distribuition that the readers will have it will be like this:

While they will see in the book this:

That it completely different.

If the readers will not activate the "Continuos Meshing", as said incorrectly in the book, they will have exactly the results in the book.

So in this case the error is not like a typo, in this case we have a successful design or a not sucessful design.

So I am asking here to, or whoever has read the book:

Which is the correct setting? With or without Continuos Meshing?
And in which way this setting will affect the analysis?

These are not simple things for all the readers as topics, if the indications in the book are not correct or the images are differents, the readers will be very confused and insicure about the software and the tutorials.

I believe that it is an Autodesk interest to give to readers a good official guide.
I thought it was an official guide, there is the name of @John_Holtz  that I see here is the responsable for Autodesk as Technical Support Senior for Simulation and Nastran, as said in the book and here.
I see a little more attention in the official Solidworks simulation book.

Please improve the guides for the learners who want to learn.

Thanks

[John_Holtz]

Hi @Anonymous 

For those of us who do not have the book (my copy is at the office) and do not have the model, can you at least attach the model so that others can look at it? Hopefully the model can be understood without the book.

Also, which version of Nastran are you using? 

[Anonymous]

Hi @John_Holtz , sure!

And thanks for the help!

Attached the files, already setup (as in the book and without that setting)

Version Nastran 2020

In addiction I have some additionl requests..

1)I have seen your great class here:

https://www.autodesk.com/autodesk-university/class/Diagnosing-Bad-Models-Autodesk-Nastran-CAD-2018

Can I have the dataset?

There is a link inside the pdf:

https://myshare.autodesk.com/:u:/g/personal/david_weinberg_autodesk_com/EcJlcdLsNCdKnrD_VqU3FKUBEw65...

But it doesn't work anymore.

_____________

2) Here another class:

https://www.autodesk.com/autodesk-university/class/Durability-101-Dont-get-tired-Fatigue-2016

The shaft "done as setup" in the download, as it is, I receive an error in Nastran.

The shat (it's the same in the Younis book) is the same but , again, not so clear in the results or in the description.

Especially compared to yours.

Can I have a , definitive, corrected version of the file?

I am quite pissed off about the carelessness in the book, there are nice examples, but it's incredible that the readers find quite BIG difference in the results (same version) and so many errors (big and little ones)

Thanks!

[John_Holtz · Accepted solution]

Hi @Anonymous 

It should be clear from the results, especially by looking at the deformed shape, that the results when using Continuous Meshing are incorrect. (I assume that the beams are welded to the tank walls along the entire length. The results show the beams are only connected at certain locations.)

Continuous meshing is only required in assemblies or when there are multiple surface bodies. In this example, there is only one part, one surface body. In theory, continuous meshing should have no effect on the analysis. However, the following articles indicate that there are problems with continuous meshing when things are applied to edges. See "Rigid connector applied to edge of shell is missing some nodes in Nastran" and "Load applied to edge of shell is missing from analysis in Nastran". The same problem appears to affect beam elements, too.

For the fatigue question, see the recently updated article "How to understand Fatigue analysis in Nastran". The attachment to that article includes an updated PDF with correct instructions for the setup of the shaft model, and some version of the Inventor file. (It may be a 2017 version of the Inventor file, but it does not matter since it is the model only.) I will create a new post for your question about "diagnosing bad models" since that question is not related to this topic.

In all fairness to @wasim_younis , he is only responsible for his book. He is not responsible for the Autodesk software and any discrepancies created by issues with the software. I have no doubt that the software produced the results shown in the book when Mr. Younis wrote that exercise. Something changed in the software (either in Inventor or in Nastran, either in versions 2020.0 or one of the later versions of 2020) between the time the book was written and the version of the software that you are using (which probably is not 2020.0 of Inventor and Inventor Nastran unless you have not applied any updates).

[Anonymous]

Thank you @John_Holtz !

The fact is that that this setting suggested:

determines if the design is ok or not ok.

Thanks for clear this for me (and all readers).

I like your stuff and tutorials, do you have more  Holtz Nastran Sources 😀?

Thanks!