| Author | Message | |||||
     fmpgomes Member Username: fmpgomes Post Number: 7 Registered: 04-2010 |
Dear Mr. Nelson, Thank you for your previous explanations. In my last question (07//07/2010) about the numeric problem when calculating the tension, it was sugested to me to add arbitrarily 1000 to the value of the elasticity modulus, which varies from zero to 30 thousand. I would like to ask if there is no other option, because the phenomenum I'm studying is exactly on the early ages. What is interesting is that the numeric problem did not occur when I used the non-linear heat equation, which means, with a internal heat generation rate depending upon temperature. Somehow, the Flex PDE became more carefull (accurate) in the calculus. I'd like to take this oportunity to say that I'm very close to developing something really interesting with the FlexPDE and the company in which I work is interested in acquiring two licenses of the software, once the one we have is not with me (I use the student version). Thank you for your time, Flávio mamede
| |||||
| rgnelson Moderator Username: rgnelson Post Number: 1402 Registered: 06-2003 |
You'll have to be clearer about your problem. Which script fails? At what problem time does it fail? What is the character of the failure? I ran SeparateTens for 275 timesteps and it seemed to be happy. | |||||
| fmpgomes Member Username: fmpgomes Post Number: 8 Registered: 04-2010 |
SeparateTens shows a really big oscilation in stress plots, how you can see in EXCEL plots. I don't know the character of the failure. Maybe an insuficient mesh because I use student version... But I will try with professional version of. prof. Gitirana. Thank you. | |||||
| jerrybrown11743 Member Username: jerrybrown11743 Post Number: 64 Registered: 03-2004 |
I found the same thing as Robert. If I prevented E from starting at zero by adding 1000 to the expression for E in region 2, I saw no problems. When E is zero, the displacement PDEs collapse to zero = zero. I can't imagine why you would want to solve an elastic solid problem with zero elasticity. It's meaningless. | |||||
| jerrybrown11743 Member Username: jerrybrown11743 Post Number: 65 Registered: 03-2004 |
By the way, I'm running version 5.1. | |||||
| fmpgomes Member Username: fmpgomes Post Number: 9 Registered: 04-2010 |
Hi, Mrs. Nelson & Jerry . E = 0 to 30000 MPa because material is a cement composite. Water-cement mixture become an exothermic reaction. Inicially is a single fluid (in general, a Bingham's fluid), after few hours a viscoelastic solid (or viscoelastoplastic). After few days or months, most of cases is sufficient consider the concrete an elastic solid. I posted part 2 fo anterior question because: 1) the problem was not correctly answered in part 1; the problem are not E=0, but variable 'temperature' error 2) the solution E'=E+1000 I didn't like I try run this problem with professional version and run ok, both scritps. FlexPDE worked very well with E ramp funtion from 0 to 30000. Thank you everyway. Flávio Mamede |
Térmico
Diference between two models