Frame Analysis - Instability of type 2

[karthur1]

Just trying to learn a little about the Frame Analysis. I first did the hand calcs for two simple supported cases.  First a continuous loaded beam and then for a point loaded case. Then I modeled it in Frame Analysis.  The results from the hand calcs and the FA is not the same.  This may be because of the errors that I get when I run the analysis, but I don't know how to clear the error.

Attached is the files for the Frame Analysis.

Thanks,

Kirk

Inventor 2017.4

[johnsonshiue]

Hi Kirk,

Just for the record, I don't claim I am an expert on Simulation. I wish I knew more than I do. Please bear with me. I notice that you used Pin constraints. This might be the reason you are getting instability warning. Pin constraint is more for a bolt in a hole. Also it assumes that the hole is strong enough and it would not deform, while the bolt would deform. This does not seem to apply to the frame member here.

I replace the Pin constraints with the Fixed constraints and the warnings are gone. Could you take a look and see if the result is closer to your expectation?

Many thanks!

[karthur1]

Johnson,

I had tried that, but the beam does not deflect like it should with the fixed constraint.  The ends of the beam should be free to rotate and move horizontally....but it can't move vertically.  That is why that I used the pin constraint.  

It seems like I need to use a combination of Pinned-Floating or Floating-Floating so the beam is not constrained in anyway on the ends. My hand calculations for this simple case shows the max deflection to be .02604 in. Seems pretty simple, but I can't get either to work.

Kirk

[vaclav_prchlik · Accepted solution]

Hi Kirk,

I downloaded your file and did the same exercise to do a cross-check calculation. Instead of paper I used Design Accelerator Beam Calculator. Picture bellow shows it step by step.

I have got very similar number from Frame Analysis and Beam Calculator. Please would you share your calculation so I would be able to understand the difference you see?

Thanks,
Vasek

[vaclav_prchlik · Accepted solution]

Hi Kirk,

One more comment to the Frame Generator model: I changed the constraints to "Floating Pinned" (that represents the end of the beam that can move and rotate) and to "Custom" with the only degree of freedom to rotate around Z axis. That may represent the most common scenario for a beam: one end absorbs axial forces and the second one is free for axial movement.

I hope it helps. If not please let me know. I would be happy to chat with you and find a solution.

Thanks

Vasek

[karthur1]

Vasek,

Creating the custom Constraint for the end condition was exactly what I needed to make the Frame Analysis cases work. I had already read through the help about the "Uplift None" condition, but it just did not register that it was saying it would all the rotation in the local "Z" direction.  When I did that constraint I get the exact same deflection that you show in your screen shot (0.2447in).

My hand calculations show the same values for deflection in both the continuous load and point load cases.

When I try to use the Design Accelerator, the problem I have is in step 2 of your slide.  I cannot find the values for the "Torsional Sectional Mod" and Torsional Rigidity Mod".

 Where did you get the values for Wk and Jk to enter into the DA?

Thanks,

Kirk

[vaclav_prchlik]

Hi Kirk,

Thanks for sharing your calculations. It is nice to hear that Frame Analysis works for you now.

I calculated it using the DA section (as on following picture). I have overwritten the other section characteristics from Inventor as they are more precise. DA provides simplifies beam section without filets.

However for this particular case Wk and Jk does not have any influence. You can use any number. Jk and Wk are used only if there is a torsional moment around the beam axis.

BTW: I gave you kudos for your pragmatical engineering approach. Doing a crosscheck hand calculation validation is IMHO the best way to make sure that software is used in the right way.

Thanks,
Vasek

[karthur1]

@vaclav_prchlik wrote:

Hi Kirk,

Thanks for sharing your calculations. It is nice to hear that Frame Analysis works for you now.

I calculated it using the DA section (as on following picture).

.

.

.

However for this particular case Wk and Jk does not have any influence. You can use any number. Jk and Wk are used only if there is a torsional moment around the beam axis.

---

Vasek,

When I setup my Design Assistant case, rather than manually entering the values for the section, I picked the beam object.  Since the beam object is a content center item, I thought that it would have all the necessary information for the calculations. Hopefully, the Wk and Jk can be added to the content center profiles to eliminate this issue.

I have always been confused as to WHY the DA even wanted those values. In all the hand beam calcs I have done, I have never used the Wk or Jk values and it frustrated me that the DA would not run the calculations without those entered. But, as you say, they are not needed in this case, which is exactly true.  I tested it by entering an arbitrary values for Wk and Jk which made no difference in the beam deflection.  Could Inventor be made smart enough to know that since there is no moment applied to the beam that the Wk and Jk values are not required input?

There is a wealth of information in the Design Assistant and Frame Analysis that is very beneficial to us engineers.  But in order for me to use it, I first have to understand how to use it and know that the results are accurate.

Thank you for your help in this. It is very much appreciated.

Kirk

[vaclav_prchlik]

Kirk,

I'm with you - having it as a section property in Content Center would make a lot of sense.

>>But in order for me to use it, I first have to understand how to use it and know that the results are accurate.

Absolutely. I'm doing the same thing in my hobby projects . There is one source of information that may actually help you with it: Autodesk Engineer's Handbook. It contains formulas used in Design Accelerators and sample hand calculations for a cross-check.

>>Could Inventor be made smart enough to know that since there is no moment applied to the beam that the Wk and Jk values are not required input?

Sure, there are many areas for improvements, no doubts.

Thanks and have a nice weekend,

Vasek

[karthur1]

Here is some information from the 2017 help that is either misleading or inaccurate.  This is why I thought that if I used the CC part, it would have the data needed for the calculation.

Kirk

[vaclav_prchlik]

Thanks a lot for identifying it. It is for sure an area of improvement.