powerplay123
Industrial
- Oct 7, 2010
- 19
Hi Everybody,
I'm currently working on a simulation model based on ANSYS for a workpiece (beam) in a woodworking machine. Since this is my first ANSYS project I need some help to find some mistakes and optimize my code.
You can find attached my current model.
The model (geometry) is relatively easy to understand.
0. Has anybody ever worked with orthotropic material like wood? For orthotropic material it is assumed that only two elasticity directions (parallel and perpendicular to the grain direction) exist. If x parallel to grain, do my y and z direcitions assume the same value or do I have to differentiate?
However, I was having some problems with the contact elements. First of all, I'm not sure about the right stiffness parameters (KN, KS, REDFACT) of the CONTAC52 Contact Elements.
1. Except for the beam everything else is supposed to be rigid and function as bearings. Any comments on the parameters?
Another problem was that I had to look up all the node-to-node contact partners manually. This requires a repeat of the whole process as soon as a small geometry change occurs that changes the node numbers which is very disturbing.
2. Is there a way to select nodes of an element for specific x, y, and z coordinates?
Transient Analysis
/SOLUTION
/TITLE,Transient Analysis of a clamped wooden beam in a planing machine
ANTYPE,TRANS,NEW
TRNOPT,FULL
nlgeom,on !Geometric Nonlinearities
sstif,on
OUTRES,ALL,ALL
f = 388.52 ! Frequency of time-varying forcing function
f_max = 432.66 ! Highest frequency of interest; needed for calculation of ITS
cycle = 1 ! Simulates for exactly 1sec
nFAP = 4*cycle*f ! Number of force application points
F_AMP = 200 ! 200N - Amplitude of time-varying force
F_AX = 100 ! 100N - Axial base displacement
dt = (1/4)*(1/f)
INIT = 1e-5 ! Initial starting time (must be > 0, convergence criterion)
ITS = 1/(20*f_max)
DELTIM,ITS
AUTOTS,ON ! Automatic Time Stepping is on
TM_START = INIT
TM_END = 1 + INIT ! Ending time of the transient
TM_INCR = dt ! Time increment
*DIM,TFORCE,TABLE,nFAP+1 ! Defining a (nFAP+1)x1x1 table,
*DO,i,1,4*f+1,1
t = ((i-1)/(4*f))+INIT
TFORCE(i,1) = F_AMP*sin(f*t+INIT) ! Force values in column 1
TFORCE(i,0) = t ! Corresponding time values in column 0
*ENDDO
*DO,TM,TM_START,TM_END,TM_INCR
TIME,TM ! Time value
NSEL,S, , ,11807,11811 ! Select last part of beam for load application
F,ALL,FZ,-TFORCE(TM) ! Time-varying force at selected nodes
F,ALL,FX,-F_AX ! Constant force at selected nodes
SOLVE
*ENDDO
SAVE
Other than just modeling, the main goal is to conduct an transient dynamic analysis. The beam is supposed to get excited by a time-varying force in z-direction. Also there should be a constant force in x-direction.
All in all, the transient analysis is not working very well. The first run was canceled because there were more than 10000 warnings, then I increased the number of allowed numbers and now the calculation is taking forever.
So the model should be optimized to make the calculations more efficient.
3. Does anybody have any suggestion regarding a more efficient performance of the transient analysis?
The typical warnings were:
"
*** WARNING *** SUPPRESSED MESSAGE CP = 457.250 TIME= 02:02:06
Node 9985 on element 4356 is unselected."
"
*** WARNING *** SUPPRESSED MESSAGE CP = 457.250 TIME= 02:02:06
Node 5563 on element 4356 is unselected."
"
*** WARNING *** SUPPRESSED MESSAGE CP = 424.125 TIME= 01:35:13
Constraint from area 14 is not transferred to unselected node 575."
ETC.
4. What is wrong with the model. Why do all these warnings occur and how can I get rid of them?
Post-processing and time-history plotting: In the model the beam is excited with a resonance frequency.
5. Therefore I'd like to plot (and animate) the general mode shape and have the amplitudes (displacements) etc. in a text file.
6. Besides I dont know for which node I should try to get the time-history results.
7. Can anybody provide me with an example code? I'm really not fimiliar with Time-History and General Postprocessor commands.
By the way the transient Analysis is running for 4.5 hrs now and is still not done :-(
You see I have a lot of questions and really need some expertise.
Thanks,
PP
I'm currently working on a simulation model based on ANSYS for a workpiece (beam) in a woodworking machine. Since this is my first ANSYS project I need some help to find some mistakes and optimize my code.
You can find attached my current model.
The model (geometry) is relatively easy to understand.
0. Has anybody ever worked with orthotropic material like wood? For orthotropic material it is assumed that only two elasticity directions (parallel and perpendicular to the grain direction) exist. If x parallel to grain, do my y and z direcitions assume the same value or do I have to differentiate?
However, I was having some problems with the contact elements. First of all, I'm not sure about the right stiffness parameters (KN, KS, REDFACT) of the CONTAC52 Contact Elements.
1. Except for the beam everything else is supposed to be rigid and function as bearings. Any comments on the parameters?
Another problem was that I had to look up all the node-to-node contact partners manually. This requires a repeat of the whole process as soon as a small geometry change occurs that changes the node numbers which is very disturbing.
2. Is there a way to select nodes of an element for specific x, y, and z coordinates?
Transient Analysis
/SOLUTION
/TITLE,Transient Analysis of a clamped wooden beam in a planing machine
ANTYPE,TRANS,NEW
TRNOPT,FULL
nlgeom,on !Geometric Nonlinearities
sstif,on
OUTRES,ALL,ALL
f = 388.52 ! Frequency of time-varying forcing function
f_max = 432.66 ! Highest frequency of interest; needed for calculation of ITS
cycle = 1 ! Simulates for exactly 1sec
nFAP = 4*cycle*f ! Number of force application points
F_AMP = 200 ! 200N - Amplitude of time-varying force
F_AX = 100 ! 100N - Axial base displacement
dt = (1/4)*(1/f)
INIT = 1e-5 ! Initial starting time (must be > 0, convergence criterion)
ITS = 1/(20*f_max)
DELTIM,ITS
AUTOTS,ON ! Automatic Time Stepping is on
TM_START = INIT
TM_END = 1 + INIT ! Ending time of the transient
TM_INCR = dt ! Time increment
*DIM,TFORCE,TABLE,nFAP+1 ! Defining a (nFAP+1)x1x1 table,
*DO,i,1,4*f+1,1
t = ((i-1)/(4*f))+INIT
TFORCE(i,1) = F_AMP*sin(f*t+INIT) ! Force values in column 1
TFORCE(i,0) = t ! Corresponding time values in column 0
*ENDDO
*DO,TM,TM_START,TM_END,TM_INCR
TIME,TM ! Time value
NSEL,S, , ,11807,11811 ! Select last part of beam for load application
F,ALL,FZ,-TFORCE(TM) ! Time-varying force at selected nodes
F,ALL,FX,-F_AX ! Constant force at selected nodes
SOLVE
*ENDDO
SAVE
Other than just modeling, the main goal is to conduct an transient dynamic analysis. The beam is supposed to get excited by a time-varying force in z-direction. Also there should be a constant force in x-direction.
All in all, the transient analysis is not working very well. The first run was canceled because there were more than 10000 warnings, then I increased the number of allowed numbers and now the calculation is taking forever.
So the model should be optimized to make the calculations more efficient.
3. Does anybody have any suggestion regarding a more efficient performance of the transient analysis?
The typical warnings were:
"
*** WARNING *** SUPPRESSED MESSAGE CP = 457.250 TIME= 02:02:06
Node 9985 on element 4356 is unselected."
"
*** WARNING *** SUPPRESSED MESSAGE CP = 457.250 TIME= 02:02:06
Node 5563 on element 4356 is unselected."
"
*** WARNING *** SUPPRESSED MESSAGE CP = 424.125 TIME= 01:35:13
Constraint from area 14 is not transferred to unselected node 575."
ETC.
4. What is wrong with the model. Why do all these warnings occur and how can I get rid of them?
Post-processing and time-history plotting: In the model the beam is excited with a resonance frequency.
5. Therefore I'd like to plot (and animate) the general mode shape and have the amplitudes (displacements) etc. in a text file.
6. Besides I dont know for which node I should try to get the time-history results.
7. Can anybody provide me with an example code? I'm really not fimiliar with Time-History and General Postprocessor commands.
By the way the transient Analysis is running for 4.5 hrs now and is still not done :-(
You see I have a lot of questions and really need some expertise.
Thanks,
PP