I have seen this question posted here before, but without a definite answer...I am wondering if it is possible to stop an analysis in Abaqus when the LPF increment is getting negative i.e. if the LPF has already reached a maxium. I know you can configure the Riks step so that the analysis stops either when you reach a specified value for the maximum LPF or a maximum displacement. The maximum LPF however is unknown (the value that will be extracted) and the maximum displacement is also unknown. As an alternative, I know of course that one can purely prescribe a maximum displacement considerably larger than the one expected and extract the maximum LPF after the analysis is complete, but this can be rather time consuming.
It is difficult to do so in Abaqus CAE. You could use python to control the command window and read your .sta file every 60 s, and if there is a drop in the load increment, terminate the analysis by commanding 'Abaqus terminate job = jobname'. The python code below will do the job.
#########################################
command = 'abaqus job=yourjobname'
system(command)
LPFv = 10 #set an initial value for load increment
for count_time in list(numpy.arange(1.0,100,1.0)):
time.sleep(60) # at 60 s interval
filepath = 'yourjobname'
fid = open(filepath+'.sta', 'r', 0)
for line in fid:
fields = line.split()
if len(fields) > 0:
if fields[0] == '1':
LPFv = float(fields[7])
else:
break
fid.close()
if LPFv < 0.000001:
system('abaqus terminate job=yourjobname'')
break
#########################################
Hi,
Try using the previous LPF when you have seen it as positive . using the that LPF value set your coefficient of tolerance so that the when the riks method reaches this tolerance value it will stop and your LPF will remain positive.
Hi, thanks for the answer, Bala.
The only Riks step parameters I see control the maximum / minimum arc length increment. Is it possible to control the LPF increment somehow? Or could I tie the arc length increment to the LPF / LPF increment?
hi
i run the riks for a SS plate under unix com load.
but the ultimate load result is very smaller than Pcr,
why?
It is difficult to do so in Abaqus CAE. You could use python to control the command window and read your .sta file every 60 s, and if there is a drop in the load increment, terminate the analysis by commanding 'Abaqus terminate job = jobname'. The python code below will do the job.
#########################################
command = 'abaqus job=yourjobname'
system(command)
LPFv = 10 #set an initial value for load increment
for count_time in list(numpy.arange(1.0,100,1.0)):
time.sleep(60) # at 60 s interval
filepath = 'yourjobname'
fid = open(filepath+'.sta', 'r', 0)
for line in fid:
fields = line.split()
if len(fields) > 0:
if fields[0] == '1':
LPFv = float(fields[7])
else:
break
fid.close()
if LPFv < 0.000001:
system('abaqus terminate job=yourjobname'')
break
#########################################
Hi,
Jie Wang thanks for your code. I had to rewrite it a bit for my file, but it works properly .
Just a few minor comments:
One thing that I changed especially is the line:
for count_time in list(numpy.arange(1.0,100,1.0)):
a while loop is easier to use here and it will run even if you have more than 100 checks to run, so it might be a bit safer and you won't have to estimate the number of check that you will need for a successful simulation.
I had to change :
system('abaqus terminate job=Job-2')
to:
mdb.jobs['Job-2'].kill()
I don't now why I had this issue, it might be due to my usage of Linux.
Here is what works for me:
_____
def check_run(path):
.....#Initialise
.....run = True
.....LPFv=10
.....n=0
.....# check model
.....while run:
..........n+=1 # number of checks
..........print('-CHECK {}-'.format(n))
..........time.sleep(60) # wait for job to generate new output
..........fid=open('{}Job-2.sta'.format(path),'r',0)
..........for line in fid:
...............fields=line.split()
...............if len(fields) > 0:
....................if fields[0]=='1': # check when actual data starts
.........................LPFv=float(fields[7])
...............else:
....................break
....................fid.close()
...............if LPFv
- Badges
- Science topic