Recoverable error in residual, but solver could not recover.
What is the meaning of this error ?
What is the meaning of this error ?
↧
Recoverable error in residual, but solver could not recover.
↧
Best solver for non-linear, time-dependent superconductor model
Until now, I have used Comsol 3.5a for modelling highly non-linear, time-dependent superconductor problems. The solver I used for this was UMFPACK, which performed best out of all the available options in that version of Comsol. I have now upgraded to Comsol 4.3 and this solver seems to have disappeared.
Does anyone have any advice as to the best solver for this type of problem? And what might be the most suitable substitute for the UMFPACK solver that worked so well previously? I am using the General Form PDE in the Multiphysics module to solve the model, for which the superconductor's electromagnetic properties are modelled using a highly non-linear E-J power law.
Does anyone have any advice as to the best solver for this type of problem? And what might be the most suitable substitute for the UMFPACK solver that worked so well previously? I am using the General Form PDE in the Multiphysics module to solve the model, for which the superconductor's electromagnetic properties are modelled using a highly non-linear E-J power law.
↧
↧
dialysis
Hi,
I have got the attached model on dialysis process at COMSOL site. I have some questions regarding the solver. As you can see, under the "Stationary Solver 1" there are two nodes- Fully Coupled and Direct. As mentioned in the COMSOL Dynamic Help section that Fully Coupled uses the damped version of Newton's method and it uses damping factor to solve equation. Also under the Fully Coupled node, at method and termination, it suggests that it is Non-linear method. I am just wondering isn't in Newton's Method there is supposed to be an initial guess and then it iterates? I mean, here all I can see is damping factors for different cases. So how does it work?
There is also Direct node which according to the Dynamic help is used for linear systems. It uses the algorithm of MUMPS.
So actually which part of my simulation is solved by Fully Coupled and which part is solved by Direct node?
Thanks,
Md Shihamul Islam
I have got the attached model on dialysis process at COMSOL site. I have some questions regarding the solver. As you can see, under the "Stationary Solver 1" there are two nodes- Fully Coupled and Direct. As mentioned in the COMSOL Dynamic Help section that Fully Coupled uses the damped version of Newton's method and it uses damping factor to solve equation. Also under the Fully Coupled node, at method and termination, it suggests that it is Non-linear method. I am just wondering isn't in Newton's Method there is supposed to be an initial guess and then it iterates? I mean, here all I can see is damping factors for different cases. So how does it work?
There is also Direct node which according to the Dynamic help is used for linear systems. It uses the algorithm of MUMPS.
So actually which part of my simulation is solved by Fully Coupled and which part is solved by Direct node?
Thanks,
Md Shihamul Islam
↧
Solver Error: Divergence cleaning failed for linear System
I have the following situation:
- Model with two permanent Magnets (not touching each other)
- Study with Parametric Sweep to move one magnet along the other
- After some steps of the sweep, the solver gives the following error: "Divergece cleaning failed for linear System"
- In the messages window I get for each step some warnings: "Avoided 236 inverted elements by using linear geometry shape order"
As soon as I get the solver errror, the computation is aborted. How do i resolve this issue?
- Model with two permanent Magnets (not touching each other)
- Study with Parametric Sweep to move one magnet along the other
- After some steps of the sweep, the solver gives the following error: "Divergece cleaning failed for linear System"
- In the messages window I get for each step some warnings: "Avoided 236 inverted elements by using linear geometry shape order"
As soon as I get the solver errror, the computation is aborted. How do i resolve this issue?
↧
Time dependent solver of unsteady Navier Stokes
Hi,
I'm simulating flow of incompressible fluid in periodic media.
I have implemented dimensionless Navier Stokes equations with Comsol PDE module.
1- I resolve my problem with a steady solver (for high Reynolds number which leads to unsteady flow). I get a good solution.
2- I use this steady solution as initial values for my unsteady solver. The problem is that the unsteady solver is giving me a steady solution (for high Re !). The solution is still the same as the steady solver solution.
The unsteady solver : Direct MUMPS solver, Times: range(0,0.002,1), Time stepping method: BDF, Steps taken by solver: Strict, Error estimation: Exclude Algebraic.
Can you advice me please how can I fix this problem ?
Thanks in advance.
Mehrez Agnaou
I'm simulating flow of incompressible fluid in periodic media.
I have implemented dimensionless Navier Stokes equations with Comsol PDE module.
1- I resolve my problem with a steady solver (for high Reynolds number which leads to unsteady flow). I get a good solution.
2- I use this steady solution as initial values for my unsteady solver. The problem is that the unsteady solver is giving me a steady solution (for high Re !). The solution is still the same as the steady solver solution.
The unsteady solver : Direct MUMPS solver, Times: range(0,0.002,1), Time stepping method: BDF, Steps taken by solver: Strict, Error estimation: Exclude Algebraic.
Can you advice me please how can I fix this problem ?
Thanks in advance.
Mehrez Agnaou
↧
↧
Coupling Solid-Beam in a Time-dependent simulation
Hi
I am working with a simplified railway model. I wish to couple a beam with the 3D model structure of the rail and to apply a moving (depends on time) force on a surface, by means of the command "Boundary Load".
I tried to connect the beam with the solid model with the "Prescribed displacement" command on the first node of the beam setting the average values of u,v,w of the facing surface of the 3D model structure and the rotation through the curl of displacement, as Ivar Kjelberg suggested.
If I try to simulate a stationary study, with a boundary load applied, it works.
Otherwise if I try to apply a moving force (by means of the definition of a square wave dependent on time), it does not work and the solver has big problems gets stuck.
(I used a fully coupled solver)
Thank you in advance
Emanuele
I am working with a simplified railway model. I wish to couple a beam with the 3D model structure of the rail and to apply a moving (depends on time) force on a surface, by means of the command "Boundary Load".
I tried to connect the beam with the solid model with the "Prescribed displacement" command on the first node of the beam setting the average values of u,v,w of the facing surface of the 3D model structure and the rotation through the curl of displacement, as Ivar Kjelberg suggested.
If I try to simulate a stationary study, with a boundary load applied, it works.
Otherwise if I try to apply a moving force (by means of the definition of a square wave dependent on time), it does not work and the solver has big problems gets stuck.
(I used a fully coupled solver)
Thank you in advance
Emanuele
↧
PDE solver with variables at different points
Dear All,
I am trying to solve a PDE where two terms contain variables at different points, something such as:
u"(x) + u'(L-x)=f(x)
This is a sample though, maybe there is no solution to this, but my point is how we can use the function u(x) at different points? The model is 1D with length equal to L.
I appreciate any help,
Best,
Hossein
I am trying to solve a PDE where two terms contain variables at different points, something such as:
u"(x) + u'(L-x)=f(x)
This is a sample though, maybe there is no solution to this, but my point is how we can use the function u(x) at different points? The model is 1D with length equal to L.
I appreciate any help,
Best,
Hossein
↧
solve an equation to find a point
Dear all,
I have solved the below system of PDEs
u"(x)+u'(x)=f(x)
v"(x)+v'(x)=g(x)
And this equation is solved. Now in each solved time step, for each point like (x), I'd like to find a point (y) which satisfies:
x+u(x)=v(y)+y
This is just an example though, suppose that I am sure of the uniqueness of that point. Any idea how to incorporate this?
I appreciate any help,
Thank you,
Hossein
I have solved the below system of PDEs
u"(x)+u'(x)=f(x)
v"(x)+v'(x)=g(x)
And this equation is solved. Now in each solved time step, for each point like (x), I'd like to find a point (y) which satisfies:
x+u(x)=v(y)+y
This is just an example though, suppose that I am sure of the uniqueness of that point. Any idea how to incorporate this?
I appreciate any help,
Thank you,
Hossein
↧
Discretization schemes of Comsol stationary solver
Hi,
I would like to ask you how can I see what are the discretization schemes that a Comsol solver uses.
This information is of great importance for any numerical study but Comsol doesn't show this !
Thank you in advance.
PS: references are welcome
Mehrez
I would like to ask you how can I see what are the discretization schemes that a Comsol solver uses.
This information is of great importance for any numerical study but Comsol doesn't show this !
Thank you in advance.
PS: references are welcome
Mehrez
↧
↧
How to set convergence criteria?
Hi everyone,
I'm working with the CFD and heat transfer modules and I'm not able to figure out how to set the convergence crtieria.
I'm performing a stationary study with a relative tolerance of 0.001 (defined in Stationary Solver --> General --> Relative Tolerance). I'm using a segregated solver with two steps, and each one of them uses an iterative solver.
I have got a couple of questions:
1) In the settings of the Segregated Solver, what is the difference between "Termination technique" and "Termination Criterion"? I mean, I know what termination technique means, I can't just figure out what termination criterion is, and why if I choose Solution or residual it asks me to define a residual factor.
2) In the settings of each segregated step, why does Comsol ask me to define again the termination technique and criterion? Shouldn't they be the same of the general segregated solver?
3) In the settings of each iterative solver, what does "Factor in error estimate" mean?
Thanks for any help you may provide.
- Davide
I'm working with the CFD and heat transfer modules and I'm not able to figure out how to set the convergence crtieria.
I'm performing a stationary study with a relative tolerance of 0.001 (defined in Stationary Solver --> General --> Relative Tolerance). I'm using a segregated solver with two steps, and each one of them uses an iterative solver.
I have got a couple of questions:
1) In the settings of the Segregated Solver, what is the difference between "Termination technique" and "Termination Criterion"? I mean, I know what termination technique means, I can't just figure out what termination criterion is, and why if I choose Solution or residual it asks me to define a residual factor.
2) In the settings of each segregated step, why does Comsol ask me to define again the termination technique and criterion? Shouldn't they be the same of the general segregated solver?
3) In the settings of each iterative solver, what does "Factor in error estimate" mean?
Thanks for any help you may provide.
- Davide
↧
Recoverable error in residual, but solver could not recover.
Recoverable error in residual, but solver could not recover.
What is the meaning of this error ?
What is the meaning of this error ?
↧
Best solver for non-linear, time-dependent superconductor model
Until now, I have used Comsol 3.5a for modelling highly non-linear, time-dependent superconductor problems. The solver I used for this was UMFPACK, which performed best out of all the available options in that version of Comsol. I have now upgraded to Comsol 4.3 and this solver seems to have disappeared.
Does anyone have any advice as to the best solver for this type of problem? And what might be the most suitable substitute for the UMFPACK solver that worked so well previously? I am using the General Form PDE in the Multiphysics module to solve the model, for which the superconductor's electromagnetic properties are modelled using a highly non-linear E-J power law.
Does anyone have any advice as to the best solver for this type of problem? And what might be the most suitable substitute for the UMFPACK solver that worked so well previously? I am using the General Form PDE in the Multiphysics module to solve the model, for which the superconductor's electromagnetic properties are modelled using a highly non-linear E-J power law.
↧
dialysis
Hi,
I have got the attached model on dialysis process at COMSOL site. I have some questions regarding the solver. As you can see, under the "Stationary Solver 1" there are two nodes- Fully Coupled and Direct. As mentioned in the COMSOL Dynamic Help section that Fully Coupled uses the damped version of Newton's method and it uses damping factor to solve equation. Also under the Fully Coupled node, at method and termination, it suggests that it is Non-linear method. I am just wondering isn't in Newton's Method there is supposed to be an initial guess and then it iterates? I mean, here all I can see is damping factors for different cases. So how does it work?
There is also Direct node which according to the Dynamic help is used for linear systems. It uses the algorithm of MUMPS.
So actually which part of my simulation is solved by Fully Coupled and which part is solved by Direct node?
Thanks,
Md Shihamul Islam
I have got the attached model on dialysis process at COMSOL site. I have some questions regarding the solver. As you can see, under the "Stationary Solver 1" there are two nodes- Fully Coupled and Direct. As mentioned in the COMSOL Dynamic Help section that Fully Coupled uses the damped version of Newton's method and it uses damping factor to solve equation. Also under the Fully Coupled node, at method and termination, it suggests that it is Non-linear method. I am just wondering isn't in Newton's Method there is supposed to be an initial guess and then it iterates? I mean, here all I can see is damping factors for different cases. So how does it work?
There is also Direct node which according to the Dynamic help is used for linear systems. It uses the algorithm of MUMPS.
So actually which part of my simulation is solved by Fully Coupled and which part is solved by Direct node?
Thanks,
Md Shihamul Islam
↧
↧
Solver Error: Divergence cleaning failed for linear System
I have the following situation:
- Model with two permanent Magnets (not touching each other)
- Study with Parametric Sweep to move one magnet along the other
- After some steps of the sweep, the solver gives the following error: "Divergece cleaning failed for linear System"
- In the messages window I get for each step some warnings: "Avoided 236 inverted elements by using linear geometry shape order"
As soon as I get the solver errror, the computation is aborted. How do i resolve this issue?
- Model with two permanent Magnets (not touching each other)
- Study with Parametric Sweep to move one magnet along the other
- After some steps of the sweep, the solver gives the following error: "Divergece cleaning failed for linear System"
- In the messages window I get for each step some warnings: "Avoided 236 inverted elements by using linear geometry shape order"
As soon as I get the solver errror, the computation is aborted. How do i resolve this issue?
↧
Time dependent solver of unsteady Navier Stokes
Hi,
I'm simulating flow of incompressible fluid in periodic media.
I have implemented dimensionless Navier Stokes equations with Comsol PDE module.
1- I resolve my problem with a steady solver (for high Reynolds number which leads to unsteady flow). I get a good solution.
2- I use this steady solution as initial values for my unsteady solver. The problem is that the unsteady solver is giving me a steady solution (for high Re !). The solution is still the same as the steady solver solution.
The unsteady solver : Direct MUMPS solver, Times: range(0,0.002,1), Time stepping method: BDF, Steps taken by solver: Strict, Error estimation: Exclude Algebraic.
Can you advice me please how can I fix this problem ?
Thanks in advance.
Mehrez Agnaou
I'm simulating flow of incompressible fluid in periodic media.
I have implemented dimensionless Navier Stokes equations with Comsol PDE module.
1- I resolve my problem with a steady solver (for high Reynolds number which leads to unsteady flow). I get a good solution.
2- I use this steady solution as initial values for my unsteady solver. The problem is that the unsteady solver is giving me a steady solution (for high Re !). The solution is still the same as the steady solver solution.
The unsteady solver : Direct MUMPS solver, Times: range(0,0.002,1), Time stepping method: BDF, Steps taken by solver: Strict, Error estimation: Exclude Algebraic.
Can you advice me please how can I fix this problem ?
Thanks in advance.
Mehrez Agnaou
↧
Coupling Solid-Beam in a Time-dependent simulation
Hi
I am working with a simplified railway model. I wish to couple a beam with the 3D model structure of the rail and to apply a moving (depends on time) force on a surface, by means of the command "Boundary Load".
I tried to connect the beam with the solid model with the "Prescribed displacement" command on the first node of the beam setting the average values of u,v,w of the facing surface of the 3D model structure and the rotation through the curl of displacement, as Ivar Kjelberg suggested.
If I try to simulate a stationary study, with a boundary load applied, it works.
Otherwise if I try to apply a moving force (by means of the definition of a square wave dependent on time), it does not work and the solver has big problems gets stuck.
(I used a fully coupled solver)
Thank you in advance
Emanuele
I am working with a simplified railway model. I wish to couple a beam with the 3D model structure of the rail and to apply a moving (depends on time) force on a surface, by means of the command "Boundary Load".
I tried to connect the beam with the solid model with the "Prescribed displacement" command on the first node of the beam setting the average values of u,v,w of the facing surface of the 3D model structure and the rotation through the curl of displacement, as Ivar Kjelberg suggested.
If I try to simulate a stationary study, with a boundary load applied, it works.
Otherwise if I try to apply a moving force (by means of the definition of a square wave dependent on time), it does not work and the solver has big problems gets stuck.
(I used a fully coupled solver)
Thank you in advance
Emanuele
↧
PDE solver with variables at different points
Dear All,
I am trying to solve a PDE where two terms contain variables at different points, something such as:
u"(x) + u'(L-x)=f(x)
This is a sample though, maybe there is no solution to this, but my point is how we can use the function u(x) at different points? The model is 1D with length equal to L.
I appreciate any help,
Best,
Hossein
I am trying to solve a PDE where two terms contain variables at different points, something such as:
u"(x) + u'(L-x)=f(x)
This is a sample though, maybe there is no solution to this, but my point is how we can use the function u(x) at different points? The model is 1D with length equal to L.
I appreciate any help,
Best,
Hossein
↧
↧
solve an equation to find a point
Dear all,
I have solved the below system of PDEs
u"(x)+u'(x)=f(x)
v"(x)+v'(x)=g(x)
And this equation is solved. Now in each solved time step, for each point like (x), I'd like to find a point (y) which satisfies:
x+u(x)=v(y)+y
This is just an example though, suppose that I am sure of the uniqueness of that point. Any idea how to incorporate this?
I appreciate any help,
Thank you,
Hossein
I have solved the below system of PDEs
u"(x)+u'(x)=f(x)
v"(x)+v'(x)=g(x)
And this equation is solved. Now in each solved time step, for each point like (x), I'd like to find a point (y) which satisfies:
x+u(x)=v(y)+y
This is just an example though, suppose that I am sure of the uniqueness of that point. Any idea how to incorporate this?
I appreciate any help,
Thank you,
Hossein
↧
Discretization schemes of Comsol stationary solver
Hi,
I would like to ask you how can I see what are the discretization schemes that a Comsol solver uses.
This information is of great importance for any numerical study but Comsol doesn't show this !
Thank you in advance.
PS: references are welcome
Mehrez
I would like to ask you how can I see what are the discretization schemes that a Comsol solver uses.
This information is of great importance for any numerical study but Comsol doesn't show this !
Thank you in advance.
PS: references are welcome
Mehrez
↧
How to set convergence criteria?
Hi everyone,
I'm working with the CFD and heat transfer modules and I'm not able to figure out how to set the convergence crtieria.
I'm performing a stationary study with a relative tolerance of 0.001 (defined in Stationary Solver --> General --> Relative Tolerance). I'm using a segregated solver with two steps, and each one of them uses an iterative solver.
I have got a couple of questions:
1) In the settings of the Segregated Solver, what is the difference between "Termination technique" and "Termination Criterion"? I mean, I know what termination technique means, I can't just figure out what termination criterion is, and why if I choose Solution or residual it asks me to define a residual factor.
2) In the settings of each segregated step, why does Comsol ask me to define again the termination technique and criterion? Shouldn't they be the same of the general segregated solver?
3) In the settings of each iterative solver, what does "Factor in error estimate" mean?
Thanks for any help you may provide.
- Davide
I'm working with the CFD and heat transfer modules and I'm not able to figure out how to set the convergence crtieria.
I'm performing a stationary study with a relative tolerance of 0.001 (defined in Stationary Solver --> General --> Relative Tolerance). I'm using a segregated solver with two steps, and each one of them uses an iterative solver.
I have got a couple of questions:
1) In the settings of the Segregated Solver, what is the difference between "Termination technique" and "Termination Criterion"? I mean, I know what termination technique means, I can't just figure out what termination criterion is, and why if I choose Solution or residual it asks me to define a residual factor.
2) In the settings of each segregated step, why does Comsol ask me to define again the termination technique and criterion? Shouldn't they be the same of the general segregated solver?
3) In the settings of each iterative solver, what does "Factor in error estimate" mean?
Thanks for any help you may provide.
- Davide
↧