Learning Forum

[Randel Ilumin]

Hello. I'm currently investigating the behavior of a heritage structure during an earthquake in APDL. Is it right to simulate the effect of earthquake load by accelerating the base using ground motion data (acceleration vs time)? How about the effect of gravity as well as the boundary conditions at the ground level? Right now, I already developed a command based on other forums that I read but I'm not really sure about this. Thank you. 

/SOLU

FILE = ‘accelvalidate1’

DT = 0.01

SKIP = 0

/INQUIRE,NUMLINES,LINES,FILE,TXT

READ = NUMLINES - SKIP

 

*DEL,ACCEL,,NOPR

*DIM, ACCEL,TABLE, READ - 1, 3

*TREAD,ACCEL,FILE,TXT,,SKIP

 

ANTYPE, TRANS

TRNOPT,FULL

TIMINT,ON,STRUC

ACEL,0,9.806,0

KBC,0

OUTRES,ALL,1

OUTPR,BASIC,ALL

 

NT = READ

NSUBST,1,,,1

 

*do,i,1,NT                                                                      

TIME,i*DT                                                                    

CMSEL,ALL                                                                

NSEL,S,LOC,Y,0                                                        

DDELE,ALL,ACCX

DDELE,ALL,ACCY

DDELE,ALL,ACCZ

D,ALL,ACCX,ACCEL(i,1)   

D,ALL,ACCY,ACCEL(i,2)   

D,ALL,ACCZ,ACCEL(i,3)                                       

ALLSEL                                                                        

SOLVE

*enddo    

 

[Ashish Khemka]

Hi Randel,

Yes, you will need to apply the base acceleration in a transient MSUP analysis. The gravity load should be defined in modal analysis. Please see if the following link helps:

5.4. Performing a Mode-Superposition Transient Dynamic Analysis (ansys.com)

Also, there is an Ansys Innovation Course on MSUP Transient analysis which might be of use to you:

Mode Superposition Transient Analysis | Ansys Courses

 

To access Ansys online help please refer to the following link: How to access the ANSYS Online Help

 

Regards,

Ashish Khemka

 

[Randel Ilumin]

Hello. Sorry, I'm confused but how am I going to apply boundary conditions and at the same time apply acceleration load (acceleration vs time record) on nodes at the ground level (say at y = 0)? Also, as far as I know, MSUP is limited to linear analysis and I need to consider nonlinearity in my case so I need to implement a full transient method right? Attached is my meshed structure. Thank you.

[Ashish Khemka]

Yes, you are right that non-linearity is not supported with MSUP transient, and my apologies for the negligence. Please see if the following link is of help:

 

How to perform nonlinear time history analyses for a Reinforced Concrete structure? (ansys.com)

 

Regards,

Ashish Khemka

[Randel Ilumin]

I see but if the accel command was used in this type of application then how to consider the weight or effect of gravity?

[Randel Ilumin]

Can I add instead the effect of gravity, say +9.81 to the y-component of acceleration?

[Ashish Khemka]

The effect of acceleration load will be opposite to that of gravity. If gravity is acting in -y direction then applying acceleration load in +y direction should be fine.

 

Regards,

Ashish Khemka

[Randel Ilumin]

Suppose that I have a ground motion record having three components (North-South, East-West, and Up-Down). From the other forum that you shared, it says that I need to apply this motion to the whole body itself (using the accel command) while restricting or fixing the nodes on the base level. However, gravity is also an upward-downward component too, so do you mean that it is right to combine or add this gravity (9.81) to the vertical component considering the opposite direction that you have mentioned? Thank you.