Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations SSS148 on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Risa-3D Frequency Check 1

Status
Not open for further replies.

jetboat

Structural
Jan 7, 2009
19
I am Wondering how others use Risa for doing frequency checks? I compared Risa frequency with hand calc frequency and the two were not comparable. I have tried a few different modeling techniques as listed below.

1. One model I used a sub meshed plate and assigned each node a spring constant calculated from the sub grade modulus. Also, on each of the corner nodes I placed a restraints ineach of the plan directions for stability issues.

2. Second model was same as first model but instead of restraints on each corner node I placed spring constants on all nodes simulating the soil to concrete frictional resistance in the plan directions.

3. The third and fourth models were similar to models one and two but instead of using plates I used solids.

What have others done or what is the best way to obtain realistic frequencies from Risa? Has anyone been able to compare Risa Calculated frequencies with their own?
 
Replies continue below

Recommended for you

I asked RISA about this a few months back. Below are my questions and answers:

Questions

Tech Support,

I have a slab on grade that is supporting a gas fired turbine. I need to ensure that the frequency of the support is outside of the operating speed of the turbine. I have a Modulus of Subgrade Reaction that was provided by the project geotechnical engineer. Here is my question: If a create a model made up of plates and use springs to represent the soil and then run an eigensolution will the results be valid? I understand that compression only springs will not work, but I manually changed the spring code from “CS” to “S” and added some lateral supports for stability. I got an eigensolution, but is the solution valid?


Answers

Very short answers to your two questions are, "yes" and "depends on engineering judgment-No definite answer".

At low forces, when the structure is still behaving in the elastic zone, two way springs represent soil behavior very closely and you will get valid results. Although at higher forces, may be caused by turbine in your case, structure softens considerably. At that point, who knows how many and which compression only springs would be active.
 
When RISA is solving for the natural frequencies of a structure we are merely solving for the eigenvalues that will cause the following to be true:

Det |[K] - w^2[M]| = 0

K is the stiffness matrix. M is the Mass matrix and w is the eigen value.

RISA should do this correctly every time.

Now, in Jetboat's case, the question is really how are you arriving at your hand calc solutions? There may be assumptions built into those hand calculations that are difficult to model in FEA, or which are not modeled properly in RISA.
 
Steve1, JoshPlum

How would you model a slab/footing in risa (spring constants,plan restraints, plates vs solids....) The models I've been comparing using a 2' thick square slab just don't seem to compare. What I am trying to do is eliminate doing the check by hand (save time) but first need to figuire out the most correct way of modeling in risa. Then I will be able to compare Risa results with hand calcs.
The equation i've used in my hand calcs for freq is 3.13*sqrt(k/w) which I havn't had any problems with from any past designs. Any suggestions/comments are appreciated!
 
Boat,

Reread my post. You must model the soil springs as "S" and add lateral restraints. However, as the tech support stated, this will only work for "lightly loaded" mats.

Steve
 
The equation I always use is:

Period = 2*pi * sqrt (m/k)

Where M = mass = w/g, and k = stiffness.

That ends up being the same thing as your formula. So, we're good there. Of course, that formula is only going to find the pure vertical mode and is not going to be very good finding the "rocking" modes.

I imagine that the rocking modes will end up controlling most of the time, but that is just a guess.

I would use your formula to confirm the pure vertical mode. If that works, then you have essentially validated the elastic response of that model.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor