Hi you ask the question on the ''How to split the simulation across more machines/cores in your computer''. You can do this by going into tools-> options -> HPC and analysis options. Here you can specify how to use your machine, be it your machine or an external compute node. You can specify to split the simulation over various cores in a machine. To do this you will need a HPC license. Each HPC license will allow you to split the simulation across up to 8 cores. Double click the machine under the ''Available configurations'' This will allow you to specify the number of cores enabled and the number of tasks to split across each core. You will need a good deal of RAM to complete this on one machine, but if you have the option you may split the simulation across multiple external machines on the network if you know their IP address / machine name.
Phil Aherne thanks phil. True. I was looking for distributed simulations. But that has the same path more or less. I have got the HPC license from ANSYS and they are pretty good.
thank you for showing the Boundary. The Boundary were not clear in part [2/3]. 1) Do I have to contact the terminals of the coils with the Boundary faces? 2)Is the current excitation DC or AC? 3)I saw that you are used in part [2/3] Magnetostatic solver, this means that you are used DC current. How can the power transmit from the first coil to the second coil? 4) Pleas, show us the setup for the ( Analysis setup) for the next tutorial videos. Thanks for your fantastic work.
Hi, 1)Yes 2)consider that AC (valid for low frequencies) 3)As I said it is on operationg point of 5A. the coil inductance is a function of current. in fact if you go into detail you will realize that the gebnerated matrix ca be at operating point or can be the average. it is like small signal rds of mosfet and large signal rout=vds/ids. 3) I was going to but it did not captured, nevertheless you can find them in my previous tutorials exactly the same...
Hello, this video is very helpful to me！Thank you very much, but i have a question. I wonder if Maxwell can simulate how to power on the transmitting coil and then get the current received by the receiving coil? For example, only the transmitting coil is excited in Maxwell-3D, because of the existence of mutual inductance, the receiving coil outputs the induced current. Can i obtain the induced current data through Maxwell?(such as use Maxwell circuit?)
as we know, wireless power transfer has to be done at higher frequency like from 10kHz-100kHz. but in solution type-magnetostatic we cannot change frequency, however in solution type- eddy current we can change and analyse results for different frequency. can we simulate this model in eddy current solution type ?? , i tried but it is taking 30-40mins and finally getting error message.
Hey, If you run the results from this simulation in simplorer you will get different efficiencies in different frequencies. that means the passed results from maxwell are frequency dependent. I think they are naming it magnetostatic to differentiate it from radiation.
***** okk, actually i dont have simplorer so i am just simulating maxwell. 1.) can u tell me is it possible to run your model shown in tutorial in eddy current solution type (because i tried it shows no error but when i simulate it takes 45mins and finally shows an error) ?? 2.) how much time did it take for you to simulate this model ?? 3.) for me it is taking too long, is there any chance to reduce it ?
the eddy current solution is one waty to go but you will get different parameters as results. you have to convert them by yourself. depends, mine takes about 10 min max. yes , try to make the air smaller, also try to make the distance between any two non-touching metals larger. fine structures add more mesh density and consequently more point to solve the maxwell EQ , hope you got the answer
Hello Sir, thank you very much for your tutorial. It really helps me understand how maxwell works. I have a question on the self-inductance results. The mutual inductance in your simulation is always a positive value. Why the self-inductance decreased first and get an increase in the end at 8:40?
I tried circular shape not rectangular one for the winding here, but when I changed it to circular shape, it just doesn't work and keep telling me insufficient memory, why circular shape doesn't work?
[info] A variation (dist='4mm') has been requested using the following machines: Local Machine. (12:15:53 PM Apr 13, 2020) [error] Find conduction path: Multilump bodies found unconnected, please separate them. Possible error bodies are 'rx_ter2'. (12:16:03 PM Apr 13, 2020) [error] Find conduction path: 'rx1 (Face_13554)' is not on any conduction path(s). Please make sure that the it is applied to a conducting object. (12:16:03 PM Apr 13, 2020) [error] Find conduction path: 'rx2 (Face_13620)' is not on any conduction path(s). Please make sure that the it is applied to a conducting object. (12:16:03 PM Apr 13, 2020) Hi this is the error im getting when i analyze all ... can anyone PLEASE help me with this?
Thank you very much for your tutorial video. Do you know how to determine the value of Magnetic filed (B) at a point (x,y,z) once we finish to draw the magnetic flux for whole space as shown in your video? Thank you.
Hey man, I know it's a few years late but for some reason the current doesn't go all the way through the coil. it shows in simulation only at the polyline terminals. Do you have any ideas on what could cause the problem ?
Hi Kamyar. Thanks for uploading it. I have a question that whether several turns of each transmitter or receiver coil can be replaced by a "big" turn only? The purpose of doing it is to accelerate simulation time. Thank you.
+binh vu yes. Please follow some of my tutorials that are using a solid coil replacing all turns with number of turns parameter in the maxwell. Watch them and let me know if you still have a question.
Thank you again for your videos. I have a question: " what do you think when i put the ferrite plate to coils systems, whether it can improve the coupling factor and efficiency (and how much it increases)? when i use air-core, the coupling is very weak.
Hi, sorry for late response, definitely it will increase the intensity of the magnetic field (if you add ferrite magnetic material) but note that when the field is omni directional you will have magnetic field in all directions. does it answer your question
Dear Kamyar K, When I want to draw air core coil in transient simulation, do i need to draw the coil separately and then try to connect each other? If so if I want to add 2500 layers of air core coil, i think it is impossible to draw each coil. Then I think I draw two cylinders and then subtract and I was planning to define the number of conductor 2500 in that cylinder but unfortunately it is a closed surface so I can't use transient simulation. Can you give me some advice how can I easily define the number of layers for air core coil?
Susan For the 3d air cored coil you can assign excitation on a radial section of the coil build. You need to create this radial section by selecting the coil >surface >section.
nice tutorial, I wanna know that is there any way to find out the parasitic capacitance of this coil and the value of AC resistance including the skin and proximity effect. thanks
I explained that before in comments section of one the videos. In brief, if you do not consider the effects of the eddy current (by not simulating them) you are very close to the Litz wire simulation results, especially in low frequencies.
Excuse me , now i encounter a problem of my simulation : Maxwell3DDesign1 (Magnetostatic) [error] Adaptive passp, process 3dnms : Insufficient memory (5:18:26 下午 四月 05, 2016) [error] Simulation completed with execution error on server: Local Machine. (5:18:26 下午 四月 05, 2016) How can I solve this ~~ ? About the error of memory , is that the problem of my computer ?? And there is always the problem of local machine , what does that mean ?? does the local machine represent my computer ?? Thank you ~
Hello, I`m trying to make a simulation almost like this one...but the problem is: "An external terminal must border the edge of the problem region and coincides with the surface of a 3D object" Yes, I'm drawing a 'Region' defined as vacuum and the 4 terminals(I have 2 coils) are bordering the face of this 'Region' , but I still get this message. Do you know how to make it work?
Paulo M. Hey, I think the reason is you are probably not using snap. either make sure that the face snaps to the region face, or draw with numbers. I am sure if you the same you would not get any errors;) regards
***** But I actually just create the 'Region'. When we put 0% of offset, Maxwell just border the faces automatically, right? Look at this design: goo.gl/7Fw5si All of the 4 current leads are bordering the the face of the 'Region'.
hi guys , i have a problem , i can not finalize my simulation because of this problem An external terminal must border the edge of the problem region and coincides with the surface of a 3D object" any one could help me
Sorry i should supplement more details to you , These are the comprehensive error reports : 2 (C:/Users/Jacky/Desktop/) Maxwell3DDesign1 (Magnetostatic) [info] User Defined Primitive: End edge position: (11:56:40 上午 四月 06, 2016) [info] User Defined Primitive: Point0: x = 192.265576731707, y = 0.00750507783471477, z = 3 (11:56:40 上午 四月 06, 2016) [info] User Defined Primitive: End edge position: (11:56:40 上午 四月 06, 2016) [info] User Defined Primitive: Point0: x = 149.927302062808, y = 0.00738402820029904, z = 3 (11:56:40 上午 四月 06, 2016) [error] Failed to copy local file C:\Users\Jacky\AppData\Local\Temp\maxwell_Jacky-PC_292332.pjt/initial.hyd to file C:/Users/Jacky/Desktop/2.mxwlresults/Maxwell3DDesign1.results/DV398_V410.imesh/initial.hyd. Error: ???????? . (12:29:26 下午 四月 06, 2016) [error] Unable to save current mesh data for simulation: Setup1 (12:29:26 下午 四月 06, 2016) [error] Unable to save current mesh data for simulation: %1 (12:29:26 下午 四月 06, 2016) [info] Parametric Analysis is done. (12:29:26 下午 四月 06, 2016) [error] Process '3dnms' terminated abnormally. It may have run out of memory or could have been killed by the user. (12:37:18 下午 四月 06, 2016) [error] Simulation completed with execution error on server: Local Machine. (12:37:18 下午 四月 06, 2016) [info] Parametric Analysis on ParametricSetup2 has been started. (12:37:19 下午 四月 06, 2016) [info] A variation (dis='150mm') has been requested on machine Local Machine (12:38:57 下午 四月 06, 2016) [info] User Defined Primitive: End edge position: (12:40:17 下午 四月 06, 2016) [info] User Defined Primitive: Point0: x = 192.265576731707, y = 0.00750507783471477, z = 3 (12:40:17 下午 四月 06, 2016) [info] User Defined Primitive: End edge position: (12:40:17 下午 四月 06, 2016) [info] User Defined Primitive: Point0: x = 149.927302062808, y = 0.00738402820029904, z = 3 (12:40:18 下午 四月 06, 2016) [error] Error closing file: %1. (1:19:27 下午 四月 06, 2016) [info] Parametric Analysis is done. (1:19:29 下午 四月 06, 2016)
Hi you ask the question on the ''How to split the simulation across more machines/cores in your computer''. You can do this by going into tools-> options -> HPC and analysis options. Here you can specify how to use your machine, be it your machine or an external compute node.
You can specify to split the simulation over various cores in a machine. To do this you will need a HPC license. Each HPC license will allow you to split the simulation across up to 8 cores.
Double click the machine under the ''Available configurations'' This will allow you to specify the number of cores enabled and the number of tasks to split across each core. You will need a good deal of RAM to complete this on one machine, but if you have the option you may split the simulation across multiple external machines on the network if you know their IP address / machine name.
Phil Aherne thanks phil.
True. I was looking for distributed simulations. But that has the same path more or less. I have got the HPC license from ANSYS and they are pretty good.
thank you for showing the Boundary. The Boundary were not clear in part [2/3].
1) Do I have to contact the terminals of the coils with the Boundary faces?
2)Is the current excitation DC or AC?
3)I saw that you are used in part [2/3] Magnetostatic solver, this means that you are used DC current. How can the power transmit from the first coil to the second coil?
4) Pleas, show us the setup for the ( Analysis setup) for the next tutorial videos.
Thanks for your fantastic work.
Hi,
1)Yes
2)consider that AC (valid for low frequencies)
3)As I said it is on operationg point of 5A. the coil inductance is a function of current. in fact if you go into detail you will realize that the gebnerated matrix ca be at operating point or can be the average. it is like small signal rds of mosfet and large signal rout=vds/ids.
3) I was going to but it did not captured, nevertheless you can find them in my previous tutorials exactly the same...
Pleas, for the next tutorial , we want to show the Analysis setup.
Sure,;)
thank you very much. simple, informative and to the point.
Hello, this video is very helpful to me！Thank you very much, but i have a question. I wonder if Maxwell can simulate how to power on the transmitting coil and then get the current received by the receiving coil? For example, only the transmitting coil is excited in Maxwell-3D, because of the existence of mutual inductance, the receiving coil outputs the induced current. Can i obtain the induced current data through Maxwell?(such as use Maxwell circuit?)
Hi, what does the current value means in the excitation , sir? Is it a DC value or peak of a ac current?
as we know, wireless power transfer has to be done at higher frequency like from 10kHz-100kHz. but in solution type-magnetostatic we cannot change frequency, however in solution type- eddy current we can change and analyse results for different frequency. can we simulate this model in eddy current solution type ?? , i tried but it is taking 30-40mins and finally getting error message.
Hey, If you run the results from this simulation in simplorer you will get different efficiencies in different frequencies. that means the passed results from maxwell are frequency dependent. I think they are naming it magnetostatic to differentiate it from radiation.
***** okk, actually i dont have simplorer so i am just simulating maxwell.
1.) can u tell me is it possible to run your model shown in tutorial in eddy current solution type (because i tried it shows no error but when i simulate it takes 45mins and finally shows an error) ??
2.) how much time did it take for you to simulate this model ??
3.) for me it is taking too long, is there any chance to reduce it ?
the eddy current solution is one waty to go but you will get different parameters as results. you have to convert them by yourself.
depends, mine takes about 10 min max.
yes , try to make the air smaller, also try to make the distance between any two non-touching metals larger. fine structures add more mesh density and consequently more point to solve the maxwell EQ ,
hope you got the answer
Hello Sir, thank you very much for your tutorial. It really helps me understand how maxwell works. I have a question on the self-inductance results. The mutual inductance in your simulation is always a positive value. Why the self-inductance decreased first and get an increase in the end at 8:40?
I tried circular shape not rectangular one for the winding here, but when I changed it to circular shape, it just doesn't work and keep telling me insufficient memory, why circular shape doesn't work?
[info] A variation (dist='4mm') has been requested using the following machines: Local Machine. (12:15:53 PM Apr 13, 2020)
[error] Find conduction path: Multilump bodies found unconnected, please separate them. Possible error bodies are 'rx_ter2'. (12:16:03 PM Apr 13, 2020)
[error] Find conduction path: 'rx1 (Face_13554)' is not on any conduction path(s). Please make sure that the it is applied to a conducting object. (12:16:03 PM Apr 13, 2020)
[error] Find conduction path: 'rx2 (Face_13620)' is not on any conduction path(s). Please make sure that the it is applied to a conducting object. (12:16:03 PM Apr 13, 2020)
Hi this is the error im getting when i analyze all ... can anyone PLEASE help me with this?
Hi! How can i found the magnetic flux?
Thanks!
Thank you very much for your tutorial video. Do you know how to determine the value of Magnetic filed (B) at a point (x,y,z) once we finish to draw the magnetic flux for whole space as shown in your video? Thank you.
Thanks Kamyar.. It was a good tutorial and it was helpful.
Is it possible in Maxwell 3D to calculate Q factor and also to use ferrites on coil ?
Hey man, I know it's a few years late but for some reason the current doesn't go all the way through the coil. it shows in simulation only at the polyline terminals. Do you have any ideas on what could cause the problem ?
thanks kamyar...
Hi Kamyar. Thanks for uploading it. I have a question that whether several turns of each transmitter or receiver coil can be replaced by a "big" turn only? The purpose of doing it is to accelerate simulation time. Thank you.
+binh vu yes. Please follow some of my tutorials that are using a solid coil replacing all turns with number of turns parameter in the maxwell. Watch them and let me know if you still have a question.
Thank you again for your videos. I have a question: " what do you think when i put the ferrite plate to coils systems, whether it can improve the coupling factor and efficiency (and how much it increases)? when i use air-core, the coupling is very weak.
Hi, sorry for late response,
definitely it will increase the intensity of the magnetic field (if you add ferrite magnetic material) but note that when the field is omni directional you will have magnetic field in all directions. does it answer your question
Dear Kamyar K,
When I want to draw air core coil in transient simulation, do i need to draw the coil separately and then try to connect each other? If so if I want to add 2500 layers of air core coil, i think it is impossible to draw each coil.
Then I think I draw two cylinders and then subtract and I was planning to define the number of conductor 2500 in that cylinder but unfortunately it is a closed surface so I can't use transient simulation.
Can you give me some advice how can I easily define the number of layers for air core coil?
Susan
For the 3d air cored coil you can assign excitation on a radial section of the coil build. You need to create this radial section by selecting the coil >surface >section.
nice tutorial, I wanna know that is there any way to find out the parasitic capacitance of this coil and the value of AC resistance including the skin and proximity effect. thanks
+Ali Raza elsctroestatic engine. I have toturials for that.
i tried to find in that in the playlist but could not found can you pls paste the link of the video here. thanks
hello, i have same problem.
did you find solution?
Hi, I am getting negative value as Mutual Inductance what's probably going wrong?
Well depends how you define your excitation directions. Remember an induced current is always opposite of source current.
thnks for the tutorial
it is taking around half an hour to simulate,.
Thank you so much. please tell me how can I do a simulation for coils that consists of many isolated strands (Litz) conductor.
Thanks again
I explained that before in comments section of one the videos. In brief, if you do not consider the effects of the eddy current (by not simulating them) you are very close to the Litz wire simulation results, especially in low frequencies.
thanks a lot Kamyar...i am doing the exact simulation it giving my the error "simulation was terminated by license error"
that means there is an error in your model. Check your design
Hi Kamyar, great stuff ! Could you send me the file of this simulation if you still have it ? Thanks
Dear Kamyar K,
When ever I try to analyse all, my laptop hangs ( can't even move mouse ). How can I simulate? :( :( :(
I have already tried 5 times. And the simulation doesn't finished.
Maybe your mesh is too dense. What’s your laptop RAM?
kamyar K Ram is 4GB.
that is not ideal bu should be ok. your total mesh numbers should be below 60,00 to see smooth simulation. is it the case ?
Excuse me , now i encounter a problem of my simulation :
Maxwell3DDesign1 (Magnetostatic)
[error] Adaptive passp, process 3dnms : Insufficient memory (5:18:26 下午 四月 05, 2016)
[error] Simulation completed with execution error on server: Local Machine. (5:18:26 下午 四月 05, 2016)
How can I solve this ~~ ?
About the error of memory , is that the problem of my computer ??
And there is always the problem of local machine , what does that mean ?? does the local machine represent my computer ??
Thank you ~
maybe you are not putting the air boundary , I was getting the same error
How did you fix this problem. I am facing the same problem??
Hello,
I`m trying to make a simulation almost like this one...but the problem is: "An external terminal must border the edge of the problem region and coincides with the surface of a 3D object"
Yes, I'm drawing a 'Region' defined as vacuum and the 4 terminals(I have 2 coils) are bordering the face of this 'Region' , but I still get this message. Do you know how to make it work?
Paulo M. As I see it, you 'hid' your region...but if I do it, I still get the same message.
Paulo M. Hey, I think the reason is you are probably not using snap. either make sure that the face snaps to the region face, or draw with numbers. I am sure if you the same you would not get any errors;)
regards
***** But I actually just create the 'Region'. When we put 0% of offset, Maxwell just border the faces automatically, right? Look at this design:
goo.gl/7Fw5si
All of the 4 current leads are bordering the the face of the 'Region'.
change the region with a air BOX. make sure the box is fitting and the faces are touching.
***** I'll try it out!
Thanks
hi guys , i have a problem , i can not finalize my simulation because of this problem
An external terminal must border the edge of the problem region and coincides with the surface of a 3D object"
any one could help me
+TheSooncomin please make sure that the terminal sheet is touching the region or the air box face.
Sorry i should supplement more details to you ,
These are the comprehensive error reports :
2 (C:/Users/Jacky/Desktop/)
Maxwell3DDesign1 (Magnetostatic)
[info] User Defined Primitive: End edge position: (11:56:40 上午 四月 06, 2016)
[info] User Defined Primitive: Point0: x = 192.265576731707, y = 0.00750507783471477, z = 3 (11:56:40 上午 四月 06, 2016)
[info] User Defined Primitive: End edge position: (11:56:40 上午 四月 06, 2016)
[info] User Defined Primitive: Point0: x = 149.927302062808, y = 0.00738402820029904, z = 3 (11:56:40 上午 四月 06, 2016)
[error] Failed to copy local file C:\Users\Jacky\AppData\Local\Temp\maxwell_Jacky-PC_292332.pjt/initial.hyd to file C:/Users/Jacky/Desktop/2.mxwlresults/Maxwell3DDesign1.results/DV398_V410.imesh/initial.hyd. Error: ???????? . (12:29:26 下午 四月 06, 2016)
[error] Unable to save current mesh data for simulation: Setup1 (12:29:26 下午 四月 06, 2016)
[error] Unable to save current mesh data for simulation: %1 (12:29:26 下午 四月 06, 2016)
[info] Parametric Analysis is done. (12:29:26 下午 四月 06, 2016)
[error] Process '3dnms' terminated abnormally. It may have run out of memory or could have been killed by the user. (12:37:18 下午 四月 06, 2016)
[error] Simulation completed with execution error on server: Local Machine. (12:37:18 下午 四月 06, 2016)
[info] Parametric Analysis on ParametricSetup2 has been started. (12:37:19 下午 四月 06, 2016)
[info] A variation (dis='150mm') has been requested on machine Local Machine (12:38:57 下午 四月 06, 2016)
[info] User Defined Primitive: End edge position: (12:40:17 下午 四月 06, 2016)
[info] User Defined Primitive: Point0: x = 192.265576731707, y = 0.00750507783471477, z = 3 (12:40:17 下午 四月 06, 2016)
[info] User Defined Primitive: End edge position: (12:40:17 下午 四月 06, 2016)
[info] User Defined Primitive: Point0: x = 149.927302062808, y = 0.00738402820029904, z = 3 (12:40:18 下午 四月 06, 2016)
[error] Error closing file: %1. (1:19:27 下午 四月 06, 2016)
[info] Parametric Analysis is done. (1:19:29 下午 四月 06, 2016)
do anyone have a free version or crack for this program please give me the link if you can
sâs