Nodal Modal Mass

[GGB.BHB]

I am trying to use the nodal modal masses to understand how the dynamic mass is distributed for the various analysed mode shapes.  I have a simple truss model with 10 nodes.  When I look at the tables of results, the sum of the nodal modal masses (for the 10 nodes) does not add up to the total modal mass (see attached file).  Please can you explain what I am missing?

Many thanks

[mustafahesenow]

Hi @GGB.BHB

Can you send us the model please.

[GGB.BHB]

Attached is the file. 

I have already noticed that the self weight of the structure is double counted (disregard density button not checked) but thought I would send the model that matches the excel file posted earlier.

Thanks

[GGB.BHB]

Can anyone out there answer this query?

Many thanks

[Anonymous]

Hi,

from what I can see the "issue" is tied to the self weight. If you don't take it into account the results are identical.

What I think is that it consider the modal mass of all the elements but then not all of them are excited the same way. I would try to set the self weight manually and see if it gives you the same results. 

[GGB.BHB]

Thank you for your reply.  I think you are correct that it has something to do with the self weight. 

Firstly I amended the model so that the self weight wasn't double counted.

I then ran it for the Combination Case 22 which has 1*LC1 Self weight + 1*LC2 Super imposed dead + 0.1*LC5 Live Load.   (All the results are shown on the attached excel file)

These results show a similar discrepancy between the sum of the total nodal modal masses and the total modal mass.

Next I ran it for LC2 Super Imposed Dead load only.  For this case the nodal modal masses DID add up to the total modal mass 🙂

Finally I ran it for LC1 Self weight only.  Again, as for the combination case containing the self weight, the nodal modal masses did NOT add up to the total modal mass.  (I double checked this last option by using the disregard density button in the modal analysis tab, with no load cases in load to mass conversion. The results (as they should be) were identical to using LC1 which is just selfweight.)

Why if you include the self weight in the modal analysis (which you almost always do) does the sum of all the nodal modal mass not equal the total modal mass?

[GGB.BHB]

OK

I think the issue is with how the loading is applied.  If point loads at the nodes are used then the sum of all the nodal masses equals the total modal mass.  However if UDLs are used (e.g. self weight or 1000kN/m which I just tried) then the sum of the nodal masses is less than the total modal mass.

Why is this?  Is it because RSA creates pseudo intermediate nodes and puts some of the mass at these positions for the modal analysis?

[Anonymous]

Are you sure with linear loads has the same issue?

Could you please send the model through? 

[GGB.BHB]

No trouble. 

[Anonymous]

Hi,

is not connected with linear load but with releases.

every time there is a release, the mass on the element is not reported to the node. I think is because of the torsion on the element, but not sure. 

[Artur.Kosakowski]

Hi @Anonymous

Thank you very much for looking into this situation. I really do appreciate your help. Could you check if switching use of DSC on/off for releases makes any difference? 

[Anonymous · Accepted solution]

Hi all,

attached the result with DCS on (the project sent by @GGB.BHB), with DCS off, and without releases (with DCS still on but doesn't change as there are not releases).

So without releases the sum is the same but results are different. 

With DSC off results are similar and the sum is the same. 

Could you please explain this topic more in depth @Artur.Kosakowski?

I mean, if you know why with the default the results are different, and mainly if we should work with DSC off (and when turn it on) due to the fact that there is a warning when turned off.

[Artur.Kosakowski · Accepted solution]

Hi @Anonymous and @GGB.BHB

I mean, if you know why with the default the results are different, and mainly if we should work with DSC off (and when turn it on) due to the fact that there is a warning when turned off.

According to my knowledge DSC should prevent the situation when you have some local modes of vibrations not found especially for not divided (with no nodes along their lengths) bars with releases. After reading your posts I'd suggest adding nodes along bars and switching off DSC when you need nodal modal masses. Will this work for the model you are discussing? 

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you. 

[GGB.BHB]

Thank you @Anonymous and @Artur.Kosakowski

The model I sent was just sent as simple example, to enquire about the principle.  I normally do break up the bars with nodes (as I have found that otherwise potentially important modes can be missed) but with large models it does tend to slow things down. 

In the future I will bear in mind what you have said re how the DSC works particularly with nodal modal masses and turn it on/off accordingly.

Thanks again for resolving this.