Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

[Chem. Eng. Module, Comsol 3.5a] Time dependent turbolent flow simulation

Please login with a confirmed email address before reporting spam

Hi to all,
I'm simulating with the Chemical Engineering Module of Comsol the fluid flow in a pipe with a lenght of about 6m and radius of 0.5m.
As preliminar approach, I'm working on a 2d model of such a problem, where the pipe is just a rectangle.
I'm interested in transient behavior of the fluid flow, therefore I'm using the time-dependent segregated solver since fluid flow is turbolent.
Fluid is water and inlet velocity is about 1m/s. Outlet pressure is set to 0.
The other two boundaries have a wall BC where the wall condition is "logaritmic wall function".
Therefore my model settings are quite similar to the ones proposed by comsol as default.
Since I usually work in the field of microsystems I'm not an expert of "macro" flows and I have never employed so far the segregated time dependent solver.
My problem is that any time dependent simulation fails, usually with the error:

*************************
Failed to evaluate expression
- Expression: d((((-rho_chns*Cmu_chns^0.25*exp(0.5*logk)*u/(log(abs(dwplus_chns))/kappa_chns+Cplus_chns)-nx_chns*p)*test(u))-(0))*(dvol),{test@1})

Log of negative
- Function: log

Segregated group X1
**************************

The group x1 contains the usual Navier Stokes variables u v p

When this not happens, time dependent solver fails simply with: "inconsistent initial values"

I'm wondering what is the problem.
I have tried to approach the time depenent problem first solving the steady state one (which runs correctly) and then looking at the order of magnitude of all the physical quantities in order to tune tolerance parameters in the time dependent solver settings. But time dependent simulations still fail.
To help time dependent solver to converge I have also used a smoothed heaviside function to make inlet velocity start from 0 and reach in a sufficient time its steady state value. But nothing gets better.
I'm wondering which is the most critical physical parameter in turbolent flow simualtions. And generally if you have some hints to tell me in order to get convergence.
Thanks a lot in advance for your help.

Alessandro

21 Replies Last Post Nov 26, 2015, 12:04 p.m. EST
Jim Freels mechanical side of nuclear engineering, multiphysics analysis, COMSOL specialist

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 3, 2009, 10:11 p.m. EDT
When you started your model creation, did you pick the steady-state form of the equations or the transient form of the equations ? If you picked the steady-state form, then try to solve a transient with these equations, perhaps that is causing the problem.

Also, look at manual setting on the solver scaling in the advanced tab. I use this routinely to stabilize the turbulent NS equations. The scale factor I use is typically the maximum absolute value of the variable. You could obtain that from your steady-state solution.
When you started your model creation, did you pick the steady-state form of the equations or the transient form of the equations ? If you picked the steady-state form, then try to solve a transient with these equations, perhaps that is causing the problem. Also, look at manual setting on the solver scaling in the advanced tab. I use this routinely to stabilize the turbulent NS equations. The scale factor I use is typically the maximum absolute value of the variable. You could obtain that from your steady-state solution.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jan 21, 2010, 11:45 a.m. EST
Hi Alessandro,

I am trying some similar model and I get the same error as you. If I try the laminar model it works (at least up to some time), but when I try the turbulent k-eps model I get the same failure message as you right in the beggining of the simulation.

Have you come up with any way to solve this? I would appreciate any help.

Thanks,
Hi Alessandro, I am trying some similar model and I get the same error as you. If I try the laminar model it works (at least up to some time), but when I try the turbulent k-eps model I get the same failure message as you right in the beggining of the simulation. Have you come up with any way to solve this? I would appreciate any help. Thanks,

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 21, 2010, 2:52 p.m. EDT
Hi.

I am also facing the same problem on Comsol 4.0.
I am trying to solve unsteady fluid flow in a pipe with heat transfer, with constant inlet velocity.

When I solve using laminar model or with steady model it works. If I am using turbulent model, Comsol gives error "Failed to find consistent initial values".

After going through forums here, I understood that I need to use step function. So I used inlet velocity condition as – ‘U_mean*step1(t)’ (Umean – mean velocity and step1() is a step function ).
This hasn’t worked yet.

Please let me know if there is a better option to define inlet condition for unsteady turbulent fluid flow problem.

Thanks,
Rohan
Hi. I am also facing the same problem on Comsol 4.0. I am trying to solve unsteady fluid flow in a pipe with heat transfer, with constant inlet velocity. When I solve using laminar model or with steady model it works. If I am using turbulent model, Comsol gives error "Failed to find consistent initial values". After going through forums here, I understood that I need to use step function. So I used inlet velocity condition as – ‘U_mean*step1(t)’ (Umean – mean velocity and step1() is a step function ). This hasn’t worked yet. Please let me know if there is a better option to define inlet condition for unsteady turbulent fluid flow problem. Thanks, Rohan

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 31, 2011, 1:48 a.m. EDT
Is it Really matter form where we are starting ?

Because It is given the option to witch over from stationary to time dependent and vice versa .

I am also facing the same problem as described above.

Ishant
Is it Really matter form where we are starting ? Because It is given the option to witch over from stationary to time dependent and vice versa . I am also facing the same problem as described above. Ishant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 31, 2011, 2:11 a.m. EDT
Hi,

I am having a lot of problems with turbulent flow too but I have been trying to simulate something similar to what you said and it works.

Could you please post your model?
Hi, I am having a lot of problems with turbulent flow too but I have been trying to simulate something similar to what you said and it works. Could you please post your model?

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 31, 2011, 2:29 a.m. EDT
PFA is the model .
PFA is the model .


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 31, 2011, 4:09 a.m. EDT
Some observations:

1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution.

2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high?

3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high?


PFA is the model .


Some observations: 1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution. 2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high? 3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high? [QUOTE] PFA is the model . [/QUOTE]

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 31, 2011, 2:45 p.m. EDT
Thanks for your prompt response.
Please note that :
1.Your first point is not clear regarding which step or what should be multiplied
2.First i was working on turbulent flow but not getting the right result so tried on laminar flow.
3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC
Please help me in this regard.

IShant










Some observations:

1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution.

2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high?

3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high?


PFA is the model .



Thanks for your prompt response. Please note that : 1.Your first point is not clear regarding which step or what should be multiplied 2.First i was working on turbulent flow but not getting the right result so tried on laminar flow. 3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC Please help me in this regard. IShant [QUOTE] Some observations: 1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution. 2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high? 3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high? [QUOTE] PFA is the model . [/QUOTE] [/QUOTE]

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 31, 2011, 2:46 p.m. EDT
Thanks for your prompt response.

Please note that :
1.Your first point is not clear regarding which step or what should be multiplied
2.First i was working on turbulent flow but not getting the right result so tried on laminar flow.
3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC

I have made the necessary changes according to you)except the velocity part) , but still showing error as :

Segregated group X1
Previous time step not stored
- Geometry: 1
- Domain: 1
Failed to evaluate expression.
- Expression: bdf(mod1.T2,2)
Failed to evaluate variable.
- Variable: mod1.T2t
- Defined as: bdf(mod1.T2,2)
Failed to evaluate expression.
- Expression: d((-2*mod1.nitf2.C_eff*mod1.T2t*test(mod1.T2)*pi*r)*(dvol),{test@20})


Please see the model attached.

IShant


Some observations:

1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution.

2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high?

3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high?


PFA is the model .




Thanks for your prompt response. Please note that : 1.Your first point is not clear regarding which step or what should be multiplied 2.First i was working on turbulent flow but not getting the right result so tried on laminar flow. 3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC I have made the necessary changes according to you)except the velocity part) , but still showing error as : Segregated group X1 Previous time step not stored - Geometry: 1 - Domain: 1 Failed to evaluate expression. - Expression: bdf(mod1.T2,2) Failed to evaluate variable. - Variable: mod1.T2t - Defined as: bdf(mod1.T2,2) Failed to evaluate expression. - Expression: d((-2*mod1.nitf2.C_eff*mod1.T2t*test(mod1.T2)*pi*r)*(dvol),{test@20}) Please see the model attached. IShant [QUOTE] Some observations: 1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution. 2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high? 3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high? [QUOTE] PFA is the model . [/QUOTE] [/QUOTE] [/QUOTE]


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 1, 2011, 5:13 a.m. EDT
1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example
Velocity:
11,1*step1(t)

You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded.

In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one.

2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos.

I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate.



Thanks for your prompt response.
Please note that :
1.Your first point is not clear regarding which step or what should be multiplied
2.First i was working on turbulent flow but not getting the right result so tried on laminar flow.
3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC
Please help me in this regard.

IShant










Some observations:

1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution.

2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high?

3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high?


PFA is the model .




1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example Velocity: 11,1*step1(t) You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded. In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one. 2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos. I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate. [QUOTE] Thanks for your prompt response. Please note that : 1.Your first point is not clear regarding which step or what should be multiplied 2.First i was working on turbulent flow but not getting the right result so tried on laminar flow. 3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC Please help me in this regard. IShant [QUOTE] Some observations: 1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution. 2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high? 3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high? [QUOTE] PFA is the model . [/QUOTE] [/QUOTE] [/QUOTE]

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 1, 2011, 6:21 a.m. EDT
I have made the necessary changes in the model attached below. Please check it .

I am here explaining you the whole problem:
A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter,

The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage.

For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as :
Error 1: Syntax error in expression.
- Expression: 11.76 * Step 1 (t)
- Subexpression: 1 (t)
- Position: 14

Error 2: Failed to find a solution.
Maximum number of segregated iterations reached.
Returned solution has not converged.
Since i have increased the no of time steps but it is still showing the same error.

Please look inti it and reply asap.

Ishant


1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example
Velocity:
11,1*step1(t)

You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded.

In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one.

2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos.

I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate.



Thanks for your prompt response.
Please note that :
1.Your first point is not clear regarding which step or what should be multiplied
2.First i was working on turbulent flow but not getting the right result so tried on laminar flow.
3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC
Please help me in this regard.

IShant










Some observations:

1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution.

2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high?

3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high?


PFA is the model .





I have made the necessary changes in the model attached below. Please check it . I am here explaining you the whole problem: A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter, The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage. For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as : Error 1: Syntax error in expression. - Expression: 11.76 * Step 1 (t) - Subexpression: 1 (t) - Position: 14 Error 2: Failed to find a solution. Maximum number of segregated iterations reached. Returned solution has not converged. Since i have increased the no of time steps but it is still showing the same error. Please look inti it and reply asap. Ishant [QUOTE] 1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example Velocity: 11,1*step1(t) You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded. In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one. 2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos. I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate. [QUOTE] Thanks for your prompt response. Please note that : 1.Your first point is not clear regarding which step or what should be multiplied 2.First i was working on turbulent flow but not getting the right result so tried on laminar flow. 3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC Please help me in this regard. IShant [QUOTE] Some observations: 1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution. 2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high? 3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high? [QUOTE] PFA is the model . [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE]


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 1, 2011, 7:49 a.m. EDT
In the model you attached there is no trace of what you wrote.

However, I think I can gather what you did. Let us go over the instructions again:

1) right click on the Definitions node and select Functions and the Step. This will create a function called step1(t), now define the proper mid-point and smoothing.

2) in the Inlet node multiply the velocity by the step1(t) function you defined beforehand.

3) in the Step 1: Time Dependent node uncheck the box near the Non-Isothermal Flow 2 physics so that the solver wll work on the Non-Isothermal Flow physics only.

For the rest you ought to begin with a smaller velocity, but it is up to you whether to do it.


I have made the necessary changes in the model attached below. Please check it .

I am here explaining you the whole problem:
A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter,

The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage.

For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as :
Error 1: Syntax error in expression.
- Expression: 11.76 * Step 1 (t)
- Subexpression: 1 (t)
- Position: 14

Error 2: Failed to find a solution.
Maximum number of segregated iterations reached.
Returned solution has not converged.
Since i have increased the no of time steps but it is still showing the same error.

Please look inti it and reply asap.

Ishant


1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example
Velocity:
11,1*step1(t)

You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded.

In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one.

2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos.

I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate.



Thanks for your prompt response.
Please note that :
1.Your first point is not clear regarding which step or what should be multiplied
2.First i was working on turbulent flow but not getting the right result so tried on laminar flow.
3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC
Please help me in this regard.

IShant










Some observations:

1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution.

2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high?

3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high?


PFA is the model .






In the model you attached there is no trace of what you wrote. However, I think I can gather what you did. Let us go over the instructions again: 1) right click on the Definitions node and select Functions and the Step. This will create a function called step1(t), now define the proper mid-point and smoothing. 2) in the Inlet node multiply the velocity by the step1(t) function you defined beforehand. 3) in the Step 1: Time Dependent node uncheck the box near the Non-Isothermal Flow 2 physics so that the solver wll work on the Non-Isothermal Flow physics only. For the rest you ought to begin with a smaller velocity, but it is up to you whether to do it. [QUOTE] I have made the necessary changes in the model attached below. Please check it . I am here explaining you the whole problem: A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter, The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage. For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as : Error 1: Syntax error in expression. - Expression: 11.76 * Step 1 (t) - Subexpression: 1 (t) - Position: 14 Error 2: Failed to find a solution. Maximum number of segregated iterations reached. Returned solution has not converged. Since i have increased the no of time steps but it is still showing the same error. Please look inti it and reply asap. Ishant [QUOTE] 1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example Velocity: 11,1*step1(t) You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded. In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one. 2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos. I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate. [QUOTE] Thanks for your prompt response. Please note that : 1.Your first point is not clear regarding which step or what should be multiplied 2.First i was working on turbulent flow but not getting the right result so tried on laminar flow. 3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC Please help me in this regard. IShant [QUOTE] Some observations: 1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution. 2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high? 3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high? [QUOTE] PFA is the model . [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE]

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 1, 2011, 8:17 a.m. EDT
Thanks again ,

Could you please share you contact detail or add me on skype : ishantj or gtalk : ishantmiet@gmail.com

Ishant


In the model you attached there is no trace of what you wrote.

However, I think I can gather what you did. Let us go over the instructions again:

1) right click on the Definitions node and select Functions and the Step. This will create a function called step1(t), now define the proper mid-point and smoothing.

2) in the Inlet node multiply the velocity by the step1(t) function you defined beforehand.

3) in the Step 1: Time Dependent node uncheck the box near the Non-Isothermal Flow 2 physics so that the solver wll work on the Non-Isothermal Flow physics only.

For the rest you ought to begin with a smaller velocity, but it is up to you whether to do it.


I have made the necessary changes in the model attached below. Please check it .

I am here explaining you the whole problem:
A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter,

The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage.

For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as :
Error 1: Syntax error in expression.
- Expression: 11.76 * Step 1 (t)
- Subexpression: 1 (t)
- Position: 14

Error 2: Failed to find a solution.
Maximum number of segregated iterations reached.
Returned solution has not converged.
Since i have increased the no of time steps but it is still showing the same error.

Please look inti it and reply asap.

Ishant


1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example
Velocity:
11,1*step1(t)

You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded.

In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one.

2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos.

I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate.



Thanks for your prompt response.
Please note that :
1.Your first point is not clear regarding which step or what should be multiplied
2.First i was working on turbulent flow but not getting the right result so tried on laminar flow.
3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC
Please help me in this regard.

IShant










Some observations:

1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution.

2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high?

3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high?


PFA is the model .







Thanks again , Could you please share you contact detail or add me on skype : ishantj or gtalk : ishantmiet@gmail.com Ishant [QUOTE] In the model you attached there is no trace of what you wrote. However, I think I can gather what you did. Let us go over the instructions again: 1) right click on the Definitions node and select Functions and the Step. This will create a function called step1(t), now define the proper mid-point and smoothing. 2) in the Inlet node multiply the velocity by the step1(t) function you defined beforehand. 3) in the Step 1: Time Dependent node uncheck the box near the Non-Isothermal Flow 2 physics so that the solver wll work on the Non-Isothermal Flow physics only. For the rest you ought to begin with a smaller velocity, but it is up to you whether to do it. [QUOTE] I have made the necessary changes in the model attached below. Please check it . I am here explaining you the whole problem: A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter, The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage. For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as : Error 1: Syntax error in expression. - Expression: 11.76 * Step 1 (t) - Subexpression: 1 (t) - Position: 14 Error 2: Failed to find a solution. Maximum number of segregated iterations reached. Returned solution has not converged. Since i have increased the no of time steps but it is still showing the same error. Please look inti it and reply asap. Ishant [QUOTE] 1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example Velocity: 11,1*step1(t) You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded. In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one. 2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos. I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate. [QUOTE] Thanks for your prompt response. Please note that : 1.Your first point is not clear regarding which step or what should be multiplied 2.First i was working on turbulent flow but not getting the right result so tried on laminar flow. 3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC Please help me in this regard. IShant [QUOTE] Some observations: 1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution. 2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high? 3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high? [QUOTE] PFA is the model . [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE]

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 1, 2011, 8:39 a.m. EDT
I have searched for the mid-point and smoothening in documentation still not been able to get their significance into the same.

Please see the attached model :
Now it is showing error as:

Segregated group X1
System matrix is zero.

Sorry to bother you , please let me know from where I can read well about this step function. It would be better if you please share your contact detail .

Ishant


Thanks again ,

Could you please share you contact detail or add me on skype : ishantj or gtalk : ishantmiet@gmail.com

Ishant


In the model you attached there is no trace of what you wrote.

However, I think I can gather what you did. Let us go over the instructions again:

1) right click on the Definitions node and select Functions and the Step. This will create a function called step1(t), now define the proper mid-point and smoothing.

2) in the Inlet node multiply the velocity by the step1(t) function you defined beforehand.

3) in the Step 1: Time Dependent node uncheck the box near the Non-Isothermal Flow 2 physics so that the solver wll work on the Non-Isothermal Flow physics only.

For the rest you ought to begin with a smaller velocity, but it is up to you whether to do it.


I have made the necessary changes in the model attached below. Please check it .

I am here explaining you the whole problem:
A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter,

The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage.

For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as :
Error 1: Syntax error in expression.
- Expression: 11.76 * Step 1 (t)
- Subexpression: 1 (t)
- Position: 14

Error 2: Failed to find a solution.
Maximum number of segregated iterations reached.
Returned solution has not converged.
Since i have increased the no of time steps but it is still showing the same error.

Please look inti it and reply asap.

Ishant


1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example
Velocity:
11,1*step1(t)

You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded.

In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one.

2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos.

I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate.



Thanks for your prompt response.
Please note that :
1.Your first point is not clear regarding which step or what should be multiplied
2.First i was working on turbulent flow but not getting the right result so tried on laminar flow.
3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC
Please help me in this regard.

IShant










Some observations:

1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution.

2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high?

3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high?


PFA is the model .








I have searched for the mid-point and smoothening in documentation still not been able to get their significance into the same. Please see the attached model : Now it is showing error as: Segregated group X1 System matrix is zero. Sorry to bother you , please let me know from where I can read well about this step function. It would be better if you please share your contact detail . Ishant [QUOTE] Thanks again , Could you please share you contact detail or add me on skype : ishantj or gtalk : ishantmiet@gmail.com Ishant [QUOTE] In the model you attached there is no trace of what you wrote. However, I think I can gather what you did. Let us go over the instructions again: 1) right click on the Definitions node and select Functions and the Step. This will create a function called step1(t), now define the proper mid-point and smoothing. 2) in the Inlet node multiply the velocity by the step1(t) function you defined beforehand. 3) in the Step 1: Time Dependent node uncheck the box near the Non-Isothermal Flow 2 physics so that the solver wll work on the Non-Isothermal Flow physics only. For the rest you ought to begin with a smaller velocity, but it is up to you whether to do it. [QUOTE] I have made the necessary changes in the model attached below. Please check it . I am here explaining you the whole problem: A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter, The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage. For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as : Error 1: Syntax error in expression. - Expression: 11.76 * Step 1 (t) - Subexpression: 1 (t) - Position: 14 Error 2: Failed to find a solution. Maximum number of segregated iterations reached. Returned solution has not converged. Since i have increased the no of time steps but it is still showing the same error. Please look inti it and reply asap. Ishant [QUOTE] 1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example Velocity: 11,1*step1(t) You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded. In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one. 2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos. I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate. [QUOTE] Thanks for your prompt response. Please note that : 1.Your first point is not clear regarding which step or what should be multiplied 2.First i was working on turbulent flow but not getting the right result so tried on laminar flow. 3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC Please help me in this regard. IShant [QUOTE] Some observations: 1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution. 2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high? 3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high? [QUOTE] PFA is the model . [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE]

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 4, 2011, 8:30 a.m. EDT
It would be kind favor if you please respond asap.

Ishant

It would be kind favor if you please respond asap. Ishant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 24, 2011, 2:25 p.m. EDT
Dear Friend,

My problem is almost solved , need a small help. Please find attached is the model and try to run it : the error this time is segregated group.

Please respond asap.

Ishant.





I have searched for the mid-point and smoothening in documentation still not been able to get their significance into the same.

Please see the attached model :
Now it is showing error as:

Segregated group X1
System matrix is zero.

Sorry to bother you , please let me know from where I can read well about this step function. It would be better if you please share your contact detail .

Ishant


Thanks again ,

Could you please share you contact detail or add me on skype : ishantj or gtalk : ishantmiet@gmail.com

Ishant


In the model you attached there is no trace of what you wrote.

However, I think I can gather what you did. Let us go over the instructions again:

1) right click on the Definitions node and select Functions and the Step. This will create a function called step1(t), now define the proper mid-point and smoothing.

2) in the Inlet node multiply the velocity by the step1(t) function you defined beforehand.

3) in the Step 1: Time Dependent node uncheck the box near the Non-Isothermal Flow 2 physics so that the solver wll work on the Non-Isothermal Flow physics only.

For the rest you ought to begin with a smaller velocity, but it is up to you whether to do it.


I have made the necessary changes in the model attached below. Please check it .

I am here explaining you the whole problem:
A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter,

The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage.

For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as :
Error 1: Syntax error in expression.
- Expression: 11.76 * Step 1 (t)
- Subexpression: 1 (t)
- Position: 14

Error 2: Failed to find a solution.
Maximum number of segregated iterations reached.
Returned solution has not converged.
Since i have increased the no of time steps but it is still showing the same error.

Please look inti it and reply asap.

Ishant


1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example
Velocity:
11,1*step1(t)

You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded.

In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one.

2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos.

I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate.



Thanks for your prompt response.
Please note that :
1.Your first point is not clear regarding which step or what should be multiplied
2.First i was working on turbulent flow but not getting the right result so tried on laminar flow.
3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC
Please help me in this regard.

IShant










Some observations:

1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution.

2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high?

3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high?


PFA is the model .









Dear Friend, My problem is almost solved , need a small help. Please find attached is the model and try to run it : the error this time is segregated group. Please respond asap. Ishant. [QUOTE] I have searched for the mid-point and smoothening in documentation still not been able to get their significance into the same. Please see the attached model : Now it is showing error as: Segregated group X1 System matrix is zero. Sorry to bother you , please let me know from where I can read well about this step function. It would be better if you please share your contact detail . Ishant [QUOTE] Thanks again , Could you please share you contact detail or add me on skype : ishantj or gtalk : ishantmiet@gmail.com Ishant [QUOTE] In the model you attached there is no trace of what you wrote. However, I think I can gather what you did. Let us go over the instructions again: 1) right click on the Definitions node and select Functions and the Step. This will create a function called step1(t), now define the proper mid-point and smoothing. 2) in the Inlet node multiply the velocity by the step1(t) function you defined beforehand. 3) in the Step 1: Time Dependent node uncheck the box near the Non-Isothermal Flow 2 physics so that the solver wll work on the Non-Isothermal Flow physics only. For the rest you ought to begin with a smaller velocity, but it is up to you whether to do it. [QUOTE] I have made the necessary changes in the model attached below. Please check it . I am here explaining you the whole problem: A producer gs is flowing with a flow rate of 11.76m/s through a small inlet of 20cm diameter( shown as 10cm due to asymmetric model) of a cylinder shape vessel with two outlet for the model of 10 cm diameter, The cylinder is lined with refractory lining 0f 30cm thickness ans covered with 10cm steel cage. For unsteady case : I have implemented what you suggested in your earlier mail but it is showing error as : Error 1: Syntax error in expression. - Expression: 11.76 * Step 1 (t) - Subexpression: 1 (t) - Position: 14 Error 2: Failed to find a solution. Maximum number of segregated iterations reached. Returned solution has not converged. Since i have increased the no of time steps but it is still showing the same error. Please look inti it and reply asap. Ishant [QUOTE] 1) in the Inlet node you are prompted to choose a type of boundary condition, if you choose velocity then you must write down the value/expression of the velocity. That very value/expression must be multiplied by a time function which grows smoothly from zero to your desired final value, an example Velocity: 11,1*step1(t) You need to define the step1 function in the Definitions node of your model. You didn't do in the last model you uploaded. In the Time Step node you are solving both the physics of your model, you need to check only the physics you are interested in. It seems the error is due to the fact that both Non-Isothermal Flow and Non-Isothermal Flow 2 boxes are checked, I would deselect the second one. 2) there is no way you can solve Re=220.000 with a laminar mode, what type of errors did you get with the turbulent one? I have been havign several issues too with the turbulent model I am quite interested in having some more infos. I think it would better first look at these two issues before thinking about the high temperature and the long span of time you want to simulate. [QUOTE] Thanks for your prompt response. Please note that : 1.Your first point is not clear regarding which step or what should be multiplied 2.First i was working on turbulent flow but not getting the right result so tried on laminar flow. 3.Inlet temp i.e Fluid temperature that is flowing inside with certain velocity have the temperature of 2200 deg C, It will heat up the refractory lining ( Brick of magnesite) upto a temp rage of 1600-1700 , so the usual temp range for the steel shell would be around 300- 500 degC Please help me in this regard. IShant [QUOTE] Some observations: 1) Your inlet velocity needs to be multiplied to a smoothly increasing function like 'step' so that the software can begin to solve from a 0 value and then reach a solution. 2) Your inlet velocity is 11 m/s and the inlet is 10 cm, coupled with the dynamic viscosity of your fluid, the corresponding Reynolds number is 220.000, i.e. fully turbulent flow, are you sure the velocity is that high? 3) Your inlet temperature is above 2200°C, steel is going to melt at that temperature, are you sure it needs to be that high? [QUOTE] PFA is the model . [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE] [/QUOTE]


Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 25, 2011, 4:40 a.m. EDT
Hi

An advice if you turn off the "repeat message" the threads are easier to understand, in my view ;)

I had a quick look at your model:

1) you have no velocity at inlet (probably selected/changed the wrong inlet type) or this could be a translation issue because I'm using latest patch of 4.1 today, I must Select Normal inflow to get a reasonable inlet velocity
2) You define a variable "t", but that one is already reserved and defined by COMSOL from the moment you call a transient solver.
You can at most define a Parameter t=0[s] to help debug a multi solver case when the time dependent one only comes in 2nd or later position, anyhow by default at start t=0 when you call a transient solver.
3) I'm not sure the settings of the solid initial conditions in the turbulent section are OK, I would have cleared them, but it could be that COMSOL does it internally, not sure as there is no turbulence in the "solid" ;)

The set up of the complex muti-segregated solver nodes was an issue in the early v4.0, fixed in v4. anyhow CFD is/was also announced not really ready before 4.2, but I find quite OK in 4.1, was poor in v4.0. So you might need to study it in details and update, change it by hand.

Then you have a stationary solver and two time series ? For me you should start to delete all that and restart from a fresh default solver sequence, I get a simple time dviscrete solver case proposed as default in my version

--
Good luck
Ivar
Hi An advice if you turn off the "repeat message" the threads are easier to understand, in my view ;) I had a quick look at your model: 1) you have no velocity at inlet (probably selected/changed the wrong inlet type) or this could be a translation issue because I'm using latest patch of 4.1 today, I must Select Normal inflow to get a reasonable inlet velocity 2) You define a variable "t", but that one is already reserved and defined by COMSOL from the moment you call a transient solver. You can at most define a Parameter t=0[s] to help debug a multi solver case when the time dependent one only comes in 2nd or later position, anyhow by default at start t=0 when you call a transient solver. 3) I'm not sure the settings of the solid initial conditions in the turbulent section are OK, I would have cleared them, but it could be that COMSOL does it internally, not sure as there is no turbulence in the "solid" ;) The set up of the complex muti-segregated solver nodes was an issue in the early v4.0, fixed in v4. anyhow CFD is/was also announced not really ready before 4.2, but I find quite OK in 4.1, was poor in v4.0. So you might need to study it in details and update, change it by hand. Then you have a stationary solver and two time series ? For me you should start to delete all that and restart from a fresh default solver sequence, I get a simple time dviscrete solver case proposed as default in my version -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 4, 2012, 8:46 a.m. EDT
hey i am trying to simulate turbulent open channel flow. but every time it shows an error " Failed to find a solution.
Maximum number of segregated iterations reached.
Returned solution is not converged."

i tried refining the mesh so many times but same error comes again and again.

the dimensions are 50m X 2m X .5m
inlet velocity is 5 m/s
outlet boundary condition is " pressure = 0 "
top surface is taken to be free surface with boundary condition as " open boundary, having no viscous stress "

note: no volume force is used

please help me out if any one has solved this kind of model

Regards
hey i am trying to simulate turbulent open channel flow. but every time it shows an error " Failed to find a solution. Maximum number of segregated iterations reached. Returned solution is not converged." i tried refining the mesh so many times but same error comes again and again. the dimensions are 50m X 2m X .5m inlet velocity is 5 m/s outlet boundary condition is " pressure = 0 " top surface is taken to be free surface with boundary condition as " open boundary, having no viscous stress " note: no volume force is used please help me out if any one has solved this kind of model Regards

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 4, 2012, 8:53 a.m. EDT
Go to study >> Solver >> Segregated

Increase the number of iteration.



--
ishant.jain@tatasteel.com
Go to study >> Solver >> Segregated Increase the number of iteration. -- ishant.jain@tatasteel.com

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 5, 2012, 3:17 a.m. EDT
hey

i incerased the no. of iterations as specified but this time it shows different error.
"Failed to find a solution:
In segregated group 1: Very ill-conditioned preconditioner. The relative residual is more than 1000 times larger than the relative tolerance.
Returned solution has not converged."

i even tried increasing the relative tolerance from 1 to 1000 . but then the result obtained is not as desired.

hey i incerased the no. of iterations as specified but this time it shows different error. "Failed to find a solution: In segregated group 1: Very ill-conditioned preconditioner. The relative residual is more than 1000 times larger than the relative tolerance. Returned solution has not converged." i even tried increasing the relative tolerance from 1 to 1000 . but then the result obtained is not as desired.

Sebastián Gómez Cardona

Please login with a confirmed email address before reporting spam

Posted: 8 years ago Nov 26, 2015, 12:04 p.m. EST
Dear friend,

I have been having the both same problems in my model than you. I see that it was a long time ago however, I would like to know if you could solve the problem.

Regards
Dear friend, I have been having the both same problems in my model than you. I see that it was a long time ago however, I would like to know if you could solve the problem. Regards

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.