i need to simulating magnetic field around the stator of brushless motor and linking some physics to find out the changes of torque of the shaft by changing the input voltage. it mast be a great job but delicious. nice job
The AC/DC module is included in all Comsol versions. Induction heating is a multiphysics application so it requires both AC/DC module and Heat Transfer module.
Hello sir while creating same model i faced two problems. Please help me to solve them. The first one is i am unable to select frequency transient study. Second is in physics for the magnetic field interface i am unable to find coil in domain. What should i do to solve these problems?
After inserting the magnetic fields and heat transfer physics go to Add study. If you can't find the frequency transient solver it could be a licence issue. Same thing for the coil.
Thank you so much. Can you make another video of Rf/u-wave heat or induction heat (But instead of an solid iron bar, if it is a pipe of steel AISI 4043 with 1mm thickness) ?
Under global definitions, insert piecewise function and define the 2 segments of a single square wave. Then choose 'Periodic' from the 'Extrapolation' options. You can call this function in the physics using its tag name (ex: pw1(t) where t is time).
Thank you for these perfect videos. I wanna change the frequency from 500 Hz to 20 kHz to visualize how B[T] changes. can you please explain how to do that? As the frequency increases I expect B[T] to get increased. But don't know how to prove it.
When you do a frequency domain study you can change the input frequency. You can enter a range of frequencies so you can easily plot the results for different frequencies from the same solution. To show how B changes with frequency, plot the value of B at a point (like max point on specimen surface) against the frequencies in your solution.
Good evening, very good video, do you have any doubts about the equations of the 3 variables that you calculated, thermal conductivity, electrical conductivity and specific heat as a function of temperature?
@@JafScience Thank you for your kind reply. If the temperature decreases even though the distance between the specimen and the coil is closer, what mistake did I make to make that phenomenon? I already checked both thermal properties and natural convection.
Is the specimen within a rectangular coil? If the specimen gets closer to the coil from one side and further away from the other sides that could reduce the total energy in the specimen. Check the induced current in the specimen, it should increase if the specimen surface gets closer to the coil.
@@JafScience The coil is circular and computed as a Symmetry condition. Although changes in physical properties according to temperature have not been taken into account, it is not thought that this will affect the tendency. By any chance, if I send you the file, can you check it? If it's too much to ask, it's okay to refuse. I've already received a lot of help.
Define the amplitude of the input current using ramp function. Alternatively you can define a piecewise function with several steps if you want the increase to be discrete.
@@saleemakhter778 Under global definitions, you can find several types of functions. Select ramp. Then call this function in the physics to define your current input. You can check my tutorial on hysteresis to see how a function can be defined and used.
Hi Bro, I am doing an actual Physical Experiment of Induction Heating, and I am experimenting how, How does size of current supplied affect the time required to reach a equilibrium temperature(point) in the iron bar(workpiece)? As I read a bit up on how the equilibrium temperature is reached and a bit of explanation in other videos the basic reasoning is when the heat gained by the workpiece is equal to the heat loss to the surroundings. Can this be simulated in Comsol, given and using the appropriate and required parameters?
Of course, but to know the time to reach a steady state response it is necessary to do a transient simulation. And yes the steady state will be obtained when the heat loss is equal to the energy gain by the induction heating.
Hi, is it possible to use this method to heat a steel to 1250 degrees Celsius with a coil current of 10A. And where about in the comsol software can I view coil resistance and inductance?
I don't think so. 10 A is too low to achieve high temperatures (unless you ignore convection and radiation). The resistance and inductance are lumped variables (like a single global value) which can be viewed in the results section. They fall under the results category of the coil.
Winding losses consist of resistive heating and eddy current losses. Resistive heating is calculated in Comsol as the dot product of current density and electric field (variable mf.Qrh). Eddy current and core losses each have a specific formula which you need to define in Comsol. You can define your own variable names for these and then insert the formula in the expression field. If needed, you can include variables from the physics (ex: x component of current density is mf.Jx, y component of flux density is mf.By). You can also define operations like maximum, average, integration (these can be found under 'Definitions' - 'Nonlocal Couplings').
Start from the maximum value of induced current or magnetic field at the surface and measure the distance away from the surface at which the value of induced current or magnetic field decreases to 37% of the maximum surface value. You may want to use a 1D plot for this. You can also calculate the theoretical value of the skin depth using the formula: skin depth= 1/sqrt(pi*permeability*frequency*electrical conductivity.)
Can I know which type of iron you have been used? There is three type ...and when i didi all steps the temprature stay at 20 and no change like your result? ....thanks in advance for your response
I did not use a specific type of iron. The permeability value I used is approximate for what is generally found in Iron. If your temperature doesn't change then there is no coupling. Make sure the inputs are specified and that the multiphysics coupling is added.
I am copying the video step by step but when I run the simulation, I am getting this error "- Feature: Time-Dependent Solver 1 (sol1/t1) Unknown function or operator. - Name: sigma" Can you please guide?
You're using a function or parameter named "sigma" somewhere in your model but you haven't defined it. Either define it or remove it from the physics/materials.
Brother i try to simulate a nanoparticle model based on magnetic field. Temperature rises due to magnetic field from Nanoparticles. Can you help me about that?
If the magnetic field is alternating it will induce eddy currents, which will generate heat. You can use the built in multiphysics to compute this heat.
@@JafScience Don't get properly what you talk about!. How i can get built in Multiphysics. If you want, may we talk in person via What's app or any communicating apps?????
@@JafScience thanks for your fast replying, I am working on 2D electromagnet simulation using iron core and homogenized multiturn coil, for the geometry I represent the iron core by drawing rectangle and the coil by a square relying at the middle of iron rectangle, I hope that you can imagine the geometry because I cannot put an image here. what's matter here that when I saw the result the magnetic field lines are not make sense at all since they are not similar to the actual magnetic field lines like what you obtained in the tutorial, so my question where is the problem with my simulation? is it in the geometry of coil or in the configuration I set?
second question, is it appropriate to use electric current physic with magnetic field to simulate electromagnetic field? I did not see any one use it maybe because they do not need it while running the simulation in 3D or even 2D axisymmetric, but in 2D the problem is that I cannot understand how the program will realize the current direction and ,as consequence, the shape and direction of magnetic field.
In 2D the homogenized multiturn coil makes the current flow into or out of the screen so the magnetic field direction is unlikely to be wrong. If your coil does not overlap with the iron core geometry then the geometry is also fine. Make sure the permeability of the iron core is set correctly.
@@JafScience It makes sense to me now why the magnetic field lines appeared silly; since the geometry of the coil is a square that overlap with iron core, but what is the proper geometry from your point of view? are there any suggestions that you can give to me in order to handle tis situation of coil geometry?
And my second question is about the skin effect, why didn’t you try to build the mesh layers around the coil instead of the workpiece to demonstrate the skin effect? Best regards…
In induction heaters the coil is usually made of a small wall thickness so the skin effect is not substantial. Instead of creating a very small wall thickness (which would complicate the mesh) I used the homogenous coil feature to ignore the skin effect. However, the specimen is thick enough for the skin effect to be observed at high frequencies and so the fine mesh layers are required to accurately capture the magnitude of this effect.
@@iam_mayurpatil If the temperature increase is significant enough as it is in this video, the skin depth will increase. You can plot the induced current at different time values and check.
Sir, How can I calculate the efficiency of induction heating in COMSOL??, I tried evaluating "mf.Qh" but it is giving an error as "undefined variable: freq"
That would be the heat energy generated in the bar divided by the total power supplied to the coil. You need to integrate mf.Qh to get the total heat energy. Coil power in frequency domain is 0.5*mf.ICoil_1^2*mf.ZCoil_1. mf.Qh should be defined in the bar. Note that mf.PCoil_1 gives the real power only (excludes reactive power).
@@JafScience Can I simulate the same problem instead of fixed iron bar, I want a m0ving iron wire with some speed around which EM field is created.Can I do this in comsol
Yes it can be done but it might be computationally very demanding depending on the operating frequency and the magnitude of bar displacement. You need to use a moving mesh to specify the velocity or displacement of the wire.
Don't use physics-controlled mesh. At high frequency the induced current will be concentrated near the surface and that's where heat is generated. If you ignore skin effect then the EM field will penetrate deep into the sample and heat will be generated everywhere. If the temperature is constant then no heat is being generated. Make sure the physics coupling is applied.
Thank you, it was helpful!
congratulation for 100 subscribers
Hello Sir, thanks for very nice tutorial. I am missing, where did you specify the current frequency in the coils?
Thank you, I appreciate it!
The current frequency is specified in the study because the magnetic fields physics is solved in the frequency-domain.
i need to simulating magnetic field around the stator of brushless motor and linking some physics to find out the changes of torque of the shaft by changing the input voltage. it mast be a great job but delicious. nice job
Does Comsol Multiphysics 5.6. includes AC/DC module for induction heating simulations? Thanks for your reply in advance.
The AC/DC module is included in all Comsol versions. Induction heating is a multiphysics application so it requires both AC/DC module and Heat Transfer module.
Sir, where do you get cast iron temperature dependent properties? Please do let me know ?
From a research paper.
i appreciate for great tutorial. would you please give any example about application builder. thank you.
I'll add this topic to my list.
Hello sir while creating same model i faced two problems. Please help me to solve them.
The first one is i am unable to select frequency transient study. Second is in physics for the magnetic field interface i am unable to find coil in domain. What should i do to solve these problems?
After inserting the magnetic fields and heat transfer physics go to Add study. If you can't find the frequency transient solver it could be a licence issue. Same thing for the coil.
Thank you so much. Can you make another video of Rf/u-wave heat or induction heat (But instead of an solid iron bar, if it is a pipe of steel AISI 4043 with 1mm thickness) ?
Hey, I follow the same step but I didn’t get temperature across work piece.can you help me to figure out this?
Does the temperature remain the same?
Hi how can i increase the coil temperature to 1300 degree celcius. Which input parameters do i need to change?
You can increase the input current.
how can I model magnetocaloric effects using Comsol Multiphysics?
I'm not familiar with magnetocaloric effects so I can't help with this. Sorry.
Hello Sir !
How can we excited a transformer coil by square wave in comsol?
Under global definitions, insert piecewise function and define the 2 segments of a single square wave. Then choose 'Periodic' from the 'Extrapolation' options. You can call this function in the physics using its tag name (ex: pw1(t) where t is time).
Thank you for these perfect videos. I wanna change the frequency from 500 Hz to 20 kHz to visualize how B[T] changes. can you please explain how to do that? As the frequency increases I expect B[T] to get increased. But don't know how to prove it.
When you do a frequency domain study you can change the input frequency. You can enter a range of frequencies so you can easily plot the results for different frequencies from the same solution. To show how B changes with frequency, plot the value of B at a point (like max point on specimen surface) against the frequencies in your solution.
@@JafScience Thank you very much for your kind answer, I could solve it.
Thank you so much
Good evening, very good video, do you have any doubts about the equations of the 3 variables that you calculated, thermal conductivity, electrical conductivity and specific heat as a function of temperature?
What do you mean by doubts about the equations?
Thank you for your video. If I use a rectangular coil, should I model it as 3D?
Yes it should be done in 3D. 2D axisymmetric is only for geometries that have an axis of revolution.
@@JafScience Thank you for your kind reply. If the temperature decreases even though the distance between the specimen and the coil is closer, what mistake did I make to make that phenomenon? I already checked both thermal properties and natural convection.
Is the specimen within a rectangular coil? If the specimen gets closer to the coil from one side and further away from the other sides that could reduce the total energy in the specimen. Check the induced current in the specimen, it should increase if the specimen surface gets closer to the coil.
@@JafScience The coil is circular and computed as a Symmetry condition. Although changes in physical properties according to temperature have not been taken into account, it is not thought that this will affect the tendency. By any chance, if I send you the file, can you check it? If it's too much to ask, it's okay to refuse. I've already received a lot of help.
You can attach a link here.
Hello Sir!
how can we change electric with time i.e. every 2 secs there must be an increase of 50nA in circuit. Please help.
Define the amplitude of the input current using ramp function. Alternatively you can define a piecewise function with several steps if you want the increase to be discrete.
@@JafScience sir i am not getting your point please elaborate it so that i can get it easily
@@saleemakhter778 Under global definitions, you can find several types of functions. Select ramp. Then call this function in the physics to define your current input. You can check my tutorial on hysteresis to see how a function can be defined and used.
Hi Bro, I am doing an actual Physical Experiment of Induction Heating, and I am experimenting how,
How does size of current supplied affect the time required to reach a equilibrium temperature(point) in the iron bar(workpiece)?
As I read a bit up on how the equilibrium temperature is reached and a bit of explanation in other videos the basic reasoning is when the heat gained by the workpiece is equal to the heat loss to the surroundings.
Can this be simulated in Comsol, given and using the appropriate and required parameters?
Of course, but to know the time to reach a steady state response it is necessary to do a transient simulation. And yes the steady state will be obtained when the heat loss is equal to the energy gain by the induction heating.
can we please have files with models just to make learning easier. please . . .
Hi,
can you tell us where you applied the thermal insulation 1 bc?
The thermal insulation BC is added by default to the outer edges of the bar. But it was overridden by the heat and radiation BCs which I added.
Hi, is it possible to use this method to heat a steel to 1250 degrees Celsius with a coil current of 10A. And where about in the comsol software can I view coil resistance and inductance?
I don't think so. 10 A is too low to achieve high temperatures (unless you ignore convection and radiation).
The resistance and inductance are lumped variables (like a single global value) which can be viewed in the results section. They fall under the results category of the coil.
A very nice video. Could you provide the information on the paper you used for the Iron Materials data?
Here's a link to the paper: www.emerald.com/insight/content/doi/10.1108/03321640910918940/full/html
@@JafScience Appreciate it!
❤@@JafScience
@@JafSciencethere's no direct data in that research paper,could you tell me how you gain it❤❤❤❤thank u vvvvvery much
How can I find core and winding losses of transformer in comsol ?
Winding losses consist of resistive heating and eddy current losses. Resistive heating is calculated in Comsol as the dot product of current density and electric field (variable mf.Qrh). Eddy current and core losses each have a specific formula which you need to define in Comsol. You can define your own variable names for these and then insert the formula in the expression field. If needed, you can include variables from the physics (ex: x component of current density is mf.Jx, y component of flux density is mf.By). You can also define operations like maximum, average, integration (these can be found under 'Definitions' - 'Nonlocal Couplings').
@@JafScience Bundle of Thanks
Sir ! Can you make a tutorial on that topic ?
Actually I am a beginner !
I can make a tutorial for this but probably in the coming months.
May I ask you to explain how to measure the value of skin depth (delta) for the bar?
Start from the maximum value of induced current or magnetic field at the surface and measure the distance away from the surface at which the value of induced current or magnetic field decreases to 37% of the maximum surface value. You may want to use a 1D plot for this.
You can also calculate the theoretical value of the skin depth using the formula: skin depth= 1/sqrt(pi*permeability*frequency*electrical conductivity.)
@@JafScience Thanks a bunch🙏🙏
You're welcome!
Can I know which type of iron you have been used? There is three type ...and when i didi all steps the temprature stay at 20 and no change like your result? ....thanks in advance for your response
I also have problem with layer boundray ...i try to build them after i reach to the result of constant temprature but i cant did them dont know why ?
I did not use a specific type of iron. The permeability value I used is approximate for what is generally found in Iron.
If your temperature doesn't change then there is no coupling. Make sure the inputs are specified and that the multiphysics coupling is added.
@@JafScience i follow all the steps , but dont know where the problem ....which input you mean ? The frequency and current or what ?
Both. And make sure you're using the frequency-transient study. If all steps are replicated then it should work.
@@JafScience thanks alot for your response ....i will try to fix them
I am copying the video step by step but when I run the simulation, I am getting this error "- Feature: Time-Dependent Solver 1 (sol1/t1)
Unknown function or operator.
- Name: sigma"
Can you please guide?
You're using a function or parameter named "sigma" somewhere in your model but you haven't defined it. Either define it or remove it from the physics/materials.
Brother i try to simulate a nanoparticle model based on magnetic field. Temperature rises due to magnetic field from Nanoparticles. Can you help me about that?
Unfortunately, I don't have experience with nanoparticles, sorry.
@@JafScience have you any idea about heat generation from magnetic field excitation?
If the magnetic field is alternating it will induce eddy currents, which will generate heat. You can use the built in multiphysics to compute this heat.
@@JafScience Don't get properly what you talk about!. How i can get built in Multiphysics. If you want, may we talk in person via What's app or any communicating apps?????
There is an interface for electromagnetic induction. It's in this video.
sir , why do not specify the material of the coil
Because the only property needed is electrical conductivity and that is defined in the coil feature under physics.
@@JafScience one more question , if the coil is helical does it make any difference in result
@m.s4477 yes but insignificant, as long as the number of coil turns is fixed.
@@JafScience if i work in 3d model , and icreate helical coil and workpiece ،will it be the same ?
Almost similar for practical purposes.
Really interested in the COMSOL how can you help me to learn it very quick because it is highly needed in my research urgent?
What features do you need to learn about beyond what's covered in my channel?
brother I have a problem in simulating an electromagnetic field in 2D, where can we contact to ask you multiple questions?
Post the list here and I'll try to answer.
@@JafScience thanks for your fast replying, I am working on 2D electromagnet simulation using iron core and homogenized multiturn coil, for the geometry I represent the iron core by drawing rectangle and the coil by a square relying at the middle of iron rectangle, I hope that you can imagine the geometry because I cannot put an image here. what's matter here that when I saw the result the magnetic field lines are not make sense at all since they are not similar to the actual magnetic field lines like what you obtained in the tutorial, so my question where is the problem with my simulation? is it in the geometry of coil or in the configuration I set?
second question, is it appropriate to use electric current physic with magnetic field to simulate electromagnetic field? I did not see any one use it maybe because they do not need it while running the simulation in 3D or even 2D axisymmetric, but in 2D the problem is that I cannot understand how the program will realize the current direction and ,as consequence, the shape and direction of magnetic field.
In 2D the homogenized multiturn coil makes the current flow into or out of the screen so the magnetic field direction is unlikely to be wrong. If your coil does not overlap with the iron core geometry then the geometry is also fine. Make sure the permeability of the iron core is set correctly.
@@JafScience
It makes sense to me now why the magnetic field lines appeared silly; since the geometry of the coil is a square that overlap with iron core, but what is the proper geometry from your point of view? are there any suggestions that you can give to me in order to handle tis situation of coil geometry?
And my second question is about the skin effect, why didn’t you try to build the mesh layers around the coil instead of the workpiece to demonstrate the skin effect?
Best regards…
In induction heaters the coil is usually made of a small wall thickness so the skin effect is not substantial. Instead of creating a very small wall thickness (which would complicate the mesh) I used the homogenous coil feature to ignore the skin effect. However, the specimen is thick enough for the skin effect to be observed at high frequencies and so the fine mesh layers are required to accurately capture the magnitude of this effect.
@@JafScience I understood it very well, Thanks alot.
hey can i have the project
Is the same model used for Eddy current thermography
This video is about heating a metal using eddy current. Thermography is not applicable.
@@JafScienceusing eddy current to crak inspection can you please do a tutorial related to this
sir, how can i make skin depth (delta) vary with temprature, how to plot that??
Skin depth does not vary with temperature.
@@JafScience electrical conductivity varies with temp then why skin depth does not increase with temp can you please clear this doubt of mine 🙏🙏
@@iam_mayurpatil If the temperature increase is significant enough as it is in this video, the skin depth will increase. You can plot the induced current at different time values and check.
@@JafScience Thnak you Sir😊
Can you do a simulation for transformer losses and draw Step by step
can i sumlate nanofilm of gold and terahertz radition shed on him ?
i want to measure the conductivity
I don't have knowledge of this application, sorry.
Sir, How can I calculate the efficiency of induction heating in COMSOL??, I tried evaluating "mf.Qh" but it is giving an error as "undefined variable: freq"
That would be the heat energy generated in the bar divided by the total power supplied to the coil. You need to integrate mf.Qh to get the total heat energy. Coil power in frequency domain is 0.5*mf.ICoil_1^2*mf.ZCoil_1. mf.Qh should be defined in the bar.
Note that mf.PCoil_1 gives the real power only (excludes reactive power).
@@JafScience Thank you Sir, i applied this and it works for me 🙏🙏
hi, do you teach this comsol?
I usually don't but I can answer questions if you need support.
What is the temperature in the air
can you send me your simulation file please
why can u get 3D temperature fingure
It's a revolution of the 2D diagram because the model is axisymmetric.
Can you please upload the actual comsol file you created in this video. I try to mimic them but couldn't run my simulation.Please help
What is the problem you are having? If you apply the same steps in the video it should work.
@@JafScience Can I simulate the same problem instead of fixed iron bar, I want a m0ving iron wire with some speed around which EM field is created.Can I do this in comsol
I have used physics controlled mesh and probably ignored skin effect through that. But I am getting constant temperature profile in the rod , why?
Yes it can be done but it might be computationally very demanding depending on the operating frequency and the magnitude of bar displacement. You need to use a moving mesh to specify the velocity or displacement of the wire.
Don't use physics-controlled mesh. At high frequency the induced current will be concentrated near the surface and that's where heat is generated. If you ignore skin effect then the EM field will penetrate deep into the sample and heat will be generated everywhere.
If the temperature is constant then no heat is being generated. Make sure the physics coupling is applied.