Discussion Forum

Posted: 1 decade ago 2011年6月27日 GMT-4 08:37

Can any body help me out please............??????????

Posted: 1 decade ago 2011年6月27日 GMT-4 17:13

[QUOTE] Can any body help me out please............?????????? [/QUOTE] When I opened the file you posted, I noticed that the port BC had no selection. That is one problem. I changed the file so that it selected BC 1. You also need to define domain 1 to be something. Even though you have its outer BC's selected as copper and applied an IBC to it, that domain needs something. I added air as another material and put domain 1 in it. The file then ran ok but I have no idea on correctness of results. It took 300s to run on my machine.

Attachments:

• mwheating3_7_40a_solved.mph

Posted: 1 decade ago 2011年7月28日 GMT-4 04:05

Thanks a lot Dennis.

Posted: 1 decade ago 2011年7月28日 GMT-4 04:19

Thanks for your help! But still it is showing only one color i.e. green without showing the temperature distribution. I was expecting the temperature distribution in colored form as we are using the microwaves generator as our source which provides heat to the material. Secondly, as long as the boundary conditions and definition of microwaves source are concerned, are both of them o.k? please find attached the file for your reference. Any help will be highly appreciated. Thanks in advance.

Attachments:

• mwheating3_7_40a_solved.mph

Posted: 1 decade ago 2011年7月28日 GMT-4 09:09

A video tutorial that you should check out is here: http://www.comsol.com/products/tutorials/waveguide_electromagnetic_simulation/ This could help you with setup of your problem. I did take a look at your file and I really did not see anything wrong with it. There is probably just a detail not setup correctly. The video should help. There is the microwave oven model that is fairly similar to what you are doing so check out the physics setup. http://www.comsol.com/showroom/documentation/model/1424/ You might want to do a frequency-stationary simulation first and make sure the temp you get is ok before going to the transient simulation, since that one probably takes longer to run. That is your call though. I myself am currently coming up to speed on the mh physics. Maybe going back and forth here will help both of us to sort it all out. In your file, under "Microwave Heating Model 1", you have all domains selected. Actually COMSOL did this for you automatically. Based on the video tutorial, you may not need to do this. If you just want to know the temp in one area you may only need to have that one or two domain(s) selected.

Posted: 1 decade ago 2011年9月19日 GMT-4 03:31

Thanks Dennis! I am still unable to diagnose the issue behind simulation of the file.Would you please share any information of setting up model and simulating it by microwaves using COMSOL Multiphysics? Specifically I have concerns regarding setting up of boundary conditions and definition of microwave source. Any help will be highly appreciated. Thanks in advance.

Posted: 1 decade ago 2011年9月21日 GMT-4 08:25

Can you post your latest file? It it is already above , just verfiy it is still valid as is.

Posted: 1 decade ago 2011年9月21日 GMT-4 13:41

Yes, it is the same as you sent me after updating it. Anyhow please find it attached. Thanks in advance.

Attachments:

• mwheating3_7_40a_solved.mph

Posted: 1 decade ago 2011年9月21日 GMT-4 14:15

A couple of questions: 1) Are you heating in all domains or do you only care about a certain one? Are you only interested in solving for the fields in a certain area? While you can solve all everywhere, it may be easier to focus in on a certain area first. 2) In your geometry node, at the end you had selected form an assembly. I think this should be form a union. Getting the physics setup correctly is important here. I also might just do a frequency stationary analysis since it may run faster than a transient. You will end up with the final temperature but you will not know how long it takes to get there. You can then go back and change it. 3) Do you have any ideas of what you are expecting for temperatures?

Posted: 1 decade ago 2011年9月22日 GMT-4 02:38

Thanks a lot Dennis. 1)-In this model, I want to heat only in the forward direction from source towards right. 2)-I will also try frequency stationary analysis with "form a union". 3)- I am expecting the color bands inside the box representing different temperature distributions. Thank you once again.

Posted: 1 decade ago 2011年9月26日 GMT-4 09:47

I just have not had enough time to work on your problem as I have been buried on my end for the last couple of days. Your problem is quite similar to the video here as I posted above. Just pay attention and you should be fine. http://www.comsol.com/products/tutorials/waveguide_electromagnetic_simulation/ If I get some time later this week I will take a deeper look at this.

Posted: 1 decade ago 2011年10月21日 GMT-4 13:54

Hi Dennis, Sorry to disturb you. If you have time can you please look at the model attached? Water is lying as a layer below Oil and Perfectly Matched Layer has been set as boundary condition for Water since it is high absorber of microwaves. I simulated it in 2D as well as 3D.For 2D simulation, I obtained some results without error but the same model for 3D gives me error: FAILED TO FIND CONSISTENT INITIAL VALUES UNDEFINED VALUE FOUND NAN OR INF FOUND WHEN SOLVING USING SOR Initially I was thinking that it could be a problem of mesh, but the problem persists even after making the mesh fine.

Attachments:

• mw_oil_water_3D1.mph

Posted: 1 decade ago 2011年10月21日 GMT-4 16:26

Not distubing me at all. I have just had some busy days here for a couple weeks. Just looking at this real quick I see that your PML does not look setup correctly. You have a mixture of both SBC's and PML's at the same location. Not an expert on PMLs but all the cases I have seen them used are for antennas where the PML is an air domain. You may be ok with having another material as a PML though. The SBC and PML boundary conds need more refinement. You have no items selected as PECs. Is that correct? And you just want to know the heat in domain 2 only. The mh seems ok. If you have a 2D model that works, I would go with that one because you can get finer meshes and maybe better accuracy.

Posted: 1 decade ago 2011年10月21日 GMT-4 22:43

Thanks a lot Dennis for your clarifications! If PML setup is wrong, why it is working in 2D? You are right, for PECs( Perefect Electric Conductors) I have not selected any domain. What might be defined as PEC in this model? What is your suggestion regarding SBC and PML boundary conditions? Yes I want to see the effect in domain 2 only. Sorry, because of the file size(18Mb), I could not attach the 2D model but still wondering how it is working if the same model in 3D gives errors? Thanks a lot for your time!

Posted: 1 decade ago 2011年10月24日 GMT-4 09:50

[QUOTE] Thanks a lot Dennis for your clarifications! If PML setup is wrong, why it is working in 2D? You are right, for PECs( Perefect Electric Conductors) I have not selected any domain. What might be defined as PEC in this model? What is your suggestion regarding SBC and PML boundary conditions? Yes I want to see the effect in domain 2 only. Sorry, because of the file size(18Mb), I could not attach the 2D model but still wondering how it is working if the same model in 3D gives errors? Thanks a lot for your time! [/QUOTE] I am looking at the a little closer but it does look like that maybe you do not need any PECs. Seems like you just have a RF launch point that has copper BCs but the rest of the model is just other materials. OK fine. Your IBCs looks ok, it is just for the waveguide BCs. I think it is the SBCs that are wrong. For a PML, most cases I have seen have a second surrounding geometry. You would have a cube within a cube, sphere within a sphere, etc. Read the docs and it should help. Search for keyword "antenna" on the forums . You are bound to get an example of PMLs being used. I am pretty sure that there is an example by COMSOL as well in their library as well. I would be curious to see your 2D file. Because you are using 4.0a, you will have to go and manually delete your results to clear the solutions. Save a copy of the file and rename it, then right click all solution files under results -> data sets (I assume you just have 1 solution file) and then delete. Resave your file and check file size again. It should drop a lot. I would think your file unsolved with be about 1Meg or so in size. You should not have to delete any plots but with 4.0a I can't remember if will delete the plots after you delete your solution node.

Posted: 1 decade ago 2011年10月25日 GMT-4 09:33

Thanks Dennis for your response. In fact I am even getting very high temperature distribution with 2D model (please see the attached model). In reality, temperature should have increased up to 300-400 degree centigrade for the defined conditions but actually it is reaching a few thousands of centigrade for one day microwaves heating, which is very strange for me. I sit because of the properties I defined are constants and do not incorporate the phase change of oil with the temperature rise? Secondly, I am getting the temperature distribution in the area just touching the microwaves source but in fact it should cover almost all the block (2meters length) after one day of heating, isn't it? Waiting for your educated response. Thanks in advance.

Attachments:

• mw_oil_water_2D3.mph