Modal Analysis

[Anonymous]

I would like to ask about model analysis,w e would like to know what mode shape or what period is governing on a certain structure, how will we know if that is the right mode shape and period for the model?

I attached a screenshot of the model with modal analysis.

Parameters and defeormation tab are the ones being seen on the structure. What are those numbers represented?

[Anonymous]

Hello. 

I've seem to have completed the model of the Telecomunications tower that i have been trying to simulate for the past week.

But i have not been able to completely understand the results that i get from the modal analysis.

Will any of you guys be so kind to help me get this clear?

I am goin to attach te results of my modal analysis.

1) First of all, i would appreciate if someone tells me what are the "Rel.mas" y "Cur.Mas". (The result that i was looking for is the list of frequencies and the mode shapes but i have to make a presentation and i need to have this results as clear as posible so i can answer question about the simulation)

2) I would also appreciate if any of you gives me a recommendation or could simply give me some information on the differences between "Mass consistent, Lumped with roations, lumped without rotations". I have already read all the posts in this and other threads para i dont have it clear. I tried all 3 of them and compared the results, dont tend to vary much but i need to understand the difference or at leats when it is that i have to use one method or the other.

Thank you very much, hope someone can help me. I have to present this results on thursday so I am trying to get everything as clear as possible.

[Anonymous]

Oh, i almost forgot. One of my colleagues was reading about modal analysis on this forum and so in this thread that some of you say that the governing modes are the ones with "Rel.Mas%" greater than 90%. Why is that? What happens with the other modes, the ones below 90% Rel. Mass? Arent they critical in case of an external force exciting the structure at those other frequencies where the Rel.Mas is below 90%?

Thank you very much for your help.

Hope someone can help me to understand this.

Thanks again.

[Mirko.Jurcevic]

Cur.mass columns are showing how much (which percent) of total mass in that direction is activated in current case.

For example, from your picture:

In mode No.5, in UY direction, only 4,41% of total mass is activated. Since total mass in UY direction is 2679,73kg, that's about 2679,73*4,41/100 = 118,17kg

Rel.mass columns are showing cumulative activated mass. It add's percent of activated mass from each period (from rel.mass column)

Example: As you can see, rel.mass for mode No.6 (for UX direction) is = 0,00 + 21,39 + 0,00 + 8,70 + 0,00 + 4,44 = 34,53%

Point is to calculate as much modes as you need to get Rel.mass columns for UX and UY direction having over 90%. But there is limitation, I've never calculated more than 30 shapes.

I've noticed that you are calculating telecomunications tower, and as I can see, it has relatively small mass (2679,73kg). I assume it's made of steel.

If you're planning to design it against seismic forces, I doubt that's the governing case (I think wind is).

Just thinking out loud...

[Mirko.Jurcevic]

You don't need to have over 90% activated (cumulative) mass, sometimes it's impossible to get.

Try to raise modal analysis number of modes, if you see that after raising them by 5-10, nothing significant happens (to activated mass), than go back to your initial number of modes.

For example;

you have calculated 10 modes, and your activated masses are 37% for X, and 38% for Y.

Than you calculate 20 modes, and your activaded masses are 38% for X, and 40% for Y.  <-- see, no significant change

go back to 10 modes and calculate your model with that.

Sometimes you just can't get over 90% of activated mass.

(you could if you calculate 1000 modes, which makes no sense)

[Rafal.Gaweda]

2) I would also appreciate if any of you gives me a recommendation or could simply give me some information on the differences between "Mass consistent, Lumped with roations, lumped without rotations". 

• Mass matrix type:
• Consistent (consistent matrix always with regard to the rotational degrees of freedom)
• Lumped with rotations (diagonal matrix with regard to rotational degrees of freedom)
• Lumped without rotations (diagonal matrix without rotational degrees of freedom).

http://docs.autodesk.com/RSAPRO/2014/ENU/filesROBOT/GUID-AD28B8C7-F614-4A51-9A8B-2E1FA66D13F6.htm

[Anonymous]

Thank you very much for your answer.

"Cur.mass columns are showing how much (which percent) of total mass in that direction is activated in current case.

For example, from your picture:

In mode No.5, in UY direction, only 4,41% of total mass is activated. Since total mass in UY direction is 2679,73kg, that's about 2679,73*4,41/100 = 118,17kg"

----> I would appreciate if you can explain to me what is the meaning of "activated mass". What is such thing as mass being "activated"?

"Point is to calculate as much modes as you need to get Rel.mass columns for UX and UY direction having over 90%. But there is limitation, I've never calculated more than 30 shapes"

----> Why is this the point? What represents this value over 90%? I think i wont get them to be that high.

"I've noticed that you are calculating telecomunications tower, and as I can see, it has relatively small mass (2679,73kg). I assume it's made of steel.

If you're planning to design it against seismic forces, I doubt that's the governing case (I think wind is)"

----> Yes indeed, it es made of S275 Steel. The tower is already built, the thing is that we are working on a green energy project, especifically wind energy. We designed an aerogenerator and now we are going to put it in this tower. This modal analysis is for us to have the mode shapes and natural frequencies of of oscillation so we can desing the control system (Knowing which frequencies are to be , let's cal it, dangeroues) to make sure the vibrating frequencies of the aerogenerator when spinning do not make the whole system to enter in resonance, which could cause structural damage.

Again, thank you very much for your help. This is my first time using Robot Structural Analysis and just getting out of college so excuse me if my questions are to basic haha.

[Mirko.Jurcevic]

Now I understand that you're looking for natural oscillations which you'll compare to those of the vibrating equpement (to avoid resonance).

For that you'll need to turn on periods and frequencies columns in table (right click > "Table columns"), and ignore all of the rest. You should compare those frequencies/periods with frequencies/periods of an equipement and they mustn't match (if you want to avoid resonance). Remember: Period = 1 / Frequency

My previous post was written with seismic analysis in mind... I didn't follow what you've be trying to do.

Basically, activated mass means percent of total structure mass being accelerated during earthquake in particular oscilation mode.

90% of activated mass means that you'll get almost maximum seismic force in observed direction when you combine all of the modes.

...just to answer your questions, but it is all about seismic analysis which you don't need in your case.

Anyway, I recommend you to study some literature about modal analysis (structural dynamics).

[Anonymous]

Thank you, i was just curious about this and wanted to have it all clear even thoug i was almost sure that the only data that matters are the frequencies and periods.

One last thing. Now that you know a little more about what i am trying to do, would you happen to know which option is best to select between "Consistent Mass", "lumped with rotations" and "lumped without rotations". I read what said about it in the RSA Autodesk tutorial but it was no clear when to use which of those options. I tried them all and compared the results. Results are almost identical (variations in the order of hundredths of hz) except for a coumple of frequencies.

Regards,

[Mirko.Jurcevic]

You should use "Consistent Mass" or "Lumped with rotations". Difference is in mass matrix assembly type. Results should be equal.

And maybe you should turn "Sturm check" option on to look for skipped modes:

• Sturm check - Sets the algorithm for detecting the skipped eigenvibrations. If a skipped mode is found, an iteration will be run again.

http://docs.autodesk.com/RSAPRO/2014/ENU/index.html?url=filesROBOT/GUID-AD28B8C7-F614-4A51-9A8B-2E1F...

[Anonymous]

"Your goal is to find as much modes as you need to get, for example, 90% of seismic mass participated in the seismic force. You will not be able to get this in every structure. Doing 100 modes doesn't make sense."

Hi!

I need little help with my master's theesis. I'm making seismic analysis for 8 floor precast concerete buliding. I can't get 90%, only 83% and 75%. I can't get 90% even if make calculation with 600 modes. I reached 94% in both directions when I made all floor connections rigid. But I dont think it nessecary in low seismic area. Can you please tell me what should I do? Can I make seismic calculations with 83% and 75% or I have to change connections rigidity untill I reach 90%.  

[Mirko.Jurcevic]

Hi!
Sometimes it is impossible to get 90% even if you try hundreds of modes.

Especially because the higher the mode, less mass is activated.

90% is the general rule according to Eurocode, but if I remember correctly it includes sum (square root of sum of squares of complete quadratic combinations) of masses oscilating at all 3 directions x,y and z.

Most of people exclude vertical (z) oscilations.. me too.

I am sorry, I should check the theorie and norms again.. it's been a while for me.

[Anonymous]

Hi!

Yes, Eurocode 8 point 4.3.3.3.1 says:

− the sum of the effective modal masses for the modes taken into account amounts to at least 90% of the total mass of the structure;

− all modes with effective modal masses greater than 5% of the total mass are taken into account.

And point 4.3.3.5.2 says: "If avg is greater than 0,25 g (2,5 m/s2) the vertical component of the seismic action, as defined in 3.2.2.3, should be taken into account..."

So I have excluded z direction too... 

But if sum of effective modal masses in x direction is 85% and y direction is 75% and it is impossibile to get 90%, then can I continue seismic calculations with 85% and 75% despite of Eurocode? What should I do? How I should go on?

[Mirko.Jurcevic]

Of course you should. From my practice, I almost never had 90% of active masses. And in cases I had, it was around 90% in only one direction.

Usually I had 70-80% active mass, but sometimes even less than that.

And I calculated it as it is.

And, I was lucky to never had over 0,25g acceleration.

You must know that sometimes you will have cases when a large mass is underground (or significant portion of it). How can you make it move? You can't. All you can do is separate your model into two, one above the ground, one under the ground, and do modal analysis on the one above the ground... or, subtracrt underground mass from total active mass to see how much is really activated. I never had a case to calculate underground structure for seismic (if it is totally submerged into the ground).

And another thing, if you use apsolute rigid supports (ground supports), you will get highest seismic forces possible. That's how did it. I never used elastic ground or elastic supports because they have damping effect and lowering your forces. So that's why I didn't bother about those 90%.

So, in practice, I usually had two models, one on the elastic ground and one on the rigid ground for seismic analysis only.

Of course, I always used calculation spectar (using behaviour factor q into account).

[Artur.Kosakowski]

You may find this approch helpful.

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.

[Anonymous]

Thank you very much! It was really helpful.

[aimanalaawar]

Hi @Artur.Kosakowski,

I have done seismic analysis for a 5-story RC residential building using the UBC97 code and in the modal analysis parameters I have chosen Consistent for mass matrix. In the modal analysis the mass participation for X & Y directions is about 73% and the frequency is about 33Hz (no significant change in these results from mode number 30 to mode number 40).

When I changed the mass matrix from Consistent to Lumped without rotations the mass participation reached 90% at mode number 10 as I remember.

Lets say that I kept the mass matrix Consistent and for such case (frequency about 33Hz and mass participation greater than 70%) it is recommended using residual mode but it is not available for UBC97. I have defined a spectral analysis for UBC97 by plotting the graph of Acceleration vs Period and I defined two spectral cases one in the X direction and one in Y direction with residual mode activated.

My questions:

How to decide which mass matrix method to choose between Consistent, Lumped with rotations or Lumped without rotations and what is the Sturm check?

After doing spectral analysis with residual mode is it necessary to reach mass participation greater than 90%?

How to define an eccentricity of 5% in a spectral case?

Thank you

[Artur.Kosakowski]

Hi @aimanalaawar

I'm not an expert in seismic design so you may want to check my remarks with people having more experience in this field than I do but:

I have done seismic analysis for a 5-story RC residential building using the UBC97 code and in the modal analysis parameters I have chosen Consistent for mass matrix. In the modal analysis the mass participation for X & Y directions is about 73% and the frequency is about 33Hz (no significant change in these results from mode number 30 to mode number 40).

You may try to increase the number of modes further to see how many you really need.

When I changed the mass matrix from Consistent to Lumped without rotations the mass participation reached 90% at mode number 10 as I remember.

Do you really need the consistent mass matrix? Perhaps having the lumped one with additional nodes generated along bar elements will be just fine?

Lets say that I kept the mass matrix Consistent and for such case (frequency about 33Hz and mass participation greater than 70%) it is recommended using residual mode but it is not available for UBC97. I have defined a spectral analysis for UBC97 by plotting the graph of Acceleration vs Period and I defined two spectral cases one in the X direction and one in Y direction with residual mode activated.

My questions:

How to decide which mass matrix method to choose between Consistent, Lumped with rotations or Lumped without rotations

Depends how important the exact shapes of modes (especially local ones for single elements) are. Usually for seismic analysis thay are not so the lumped without rotations are in most situations )fr multi-story buildings) good enough.

and what is the Sturm check?

It is the search for missing eigenvalues - usually not needed for seismic/spectral case as related to local modes of vibrations with low mass participation anyway.

After doing spectral analysis with residual mode is it necessary to reach mass participation greater than 90%?

The aim of this approach is to 'compensate' for the missing mass participation when reaching the target value cannot be done with 'reasonable' number of modes. It is 'instead of' rather than to 'reach' option.

How to define an eccentricity of 5% in a spectral case?

You need to set it in the modal analysis.

If one or more of these posts answered your question, please click Accept as Solution on the posts that helped you so others in the community can find them easily.

[Anonymous]

I have a question, Do you know where it is mentioned in the Eurocode, how much I shall take as a percentage of the live load or snow load in the conversion of load to mass because I am not finding that if you have used the Eurocode.thanks

[Anonymous]

Can you clarify for me, What is the percentage of the live load and snow load coefficients that I can use in the model analysis, for example in the Eurocode 8, Is that mentioned how much I shall take from these loads?

[Simau]

@Anonymous 

Look at these paragraphs (EC8)