# Fem heat transfer **URL:** \ **Category:** General Questions\ **Created:** [September 23, 2026, 9:52am UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688 "2026-09-23T09:52:30Z")\ **Posts on this page:** 15\ **Page:** 1
**Author:** ![mehmed.celebi](https://avatars.discourse-cdn.com/v4/letter/m/898d66/32.png) [@mehmed.celebi](https://prepomax.discourse.group/u/mehmed.celebi)\ **Post date:** [September 23, 2026, 9:52am UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/1 "2026-09-23T09:52:30Z")
Please can any one help. This model all symmetric but why this region shows different temp value. Waht I did: `Merge Coincident Nodes` tie contact. I try to solve by AI but I am really confused. ![pre1](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/1/18ca4756ca619acb0b144277cdb490b64c07baa2.png) ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [September 23, 2026, 10:07am UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/2 "2026-09-23T10:07:27Z")
Please share the .pmx file here or via a hosting website. You may need to use File → Close All Results before saving the file, though (to reduce its size). ---
**Author:** ![mehmed.celebi](https://avatars.discourse-cdn.com/v4/letter/m/898d66/32.png) [@mehmed.celebi](https://prepomax.discourse.group/u/mehmed.celebi)\ **Post date:** [September 23, 2026, 2:05pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/3 "2026-09-23T14:05:39Z")
[piramit2.pmx](https://prepomax.discourse.group/uploads/short-url/599jYMLRjOnpxwk2nlKG2c5aLP1.pmx) (3.8 MB) this is my file. ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [September 23, 2026, 2:12pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/4 "2026-09-23T14:12:23Z")
It seems that you are unnecessarily using shells here while it should be a solid (not hollow) geometry: ![image](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/3/36a50ad5d64c320e5c41689b1b7b2b5ed53f6ef9.png) So you should try importing one solid part or stitching all these faces into such a solid. To use shells, you would need only one surface (midsurface or top/bottom surface) per thin part. Btw. when you have multiple parts, they are treated as disconnected. Creating a compound part out of them will ensure a continuous mesh and heat transfer through the whole component. ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [September 23, 2026, 2:25pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/5 "2026-09-23T14:25:57Z")
I mean that you should try with a geometry like this: [piramit2 solid.pmx](https://prepomax.discourse.group/uploads/short-url/fj7RxrNJrtj4EVg8rp4ar8EYrf4.pmx) (62.5 KB) Although it’s still very thin-walled and may actually require proper shell modeling or a different approach (e.g. some simplification/homogenization or local analysis). And I think that were some flaws in the final solid, but you should get the whole idea now. ---
**Author:** ![mehmed.celebi](https://avatars.discourse-cdn.com/v4/letter/m/898d66/32.png) [@mehmed.celebi](https://prepomax.discourse.group/u/mehmed.celebi)\ **Post date:** [September 23, 2026, 3:53pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/6 "2026-09-23T15:53:05Z")
Thanks in advance. But when I wanted to make it solid, the software didn’t accept this feature. I don’t understand why. FEAnalyst \<[notifications@prepomax.discoursemail.com](mailto:notifications@prepomax.discoursemail.com)\>, 23 Eyl 2026 Çar, 17:35 tarihinde şunu yazdı: ---
**Author:** ![mehmed.celebi](https://avatars.discourse-cdn.com/v4/letter/m/898d66/32.png) [@mehmed.celebi](https://prepomax.discourse.group/u/mehmed.celebi)\ **Post date:** [September 23, 2026, 3:58pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/7 "2026-09-23T15:58:29Z")
WARNING! NOT ALL FACES HAVE BEEN MESHED 09/23/26 18:57:36 SURFACE MESHING ERROR OCCURRED IN 107 FACES: Mesh finished with failure. mehmed çelebi \<[mehmed.celebi@gmail.com](mailto:mehmed.celebi@gmail.com)\>, 23 Eyl 2026 Çar, 18:53 tarihinde şunu yazdı: ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [September 23, 2026, 3:58pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/8 "2026-09-23T15:58:51Z")
How did you prepare the geometry ? In what CAD software ? ---
**Author:** ![mehmed.celebi](https://avatars.discourse-cdn.com/v4/letter/m/898d66/32.png) [@mehmed.celebi](https://prepomax.discourse.group/u/mehmed.celebi)\ **Post date:** [September 23, 2026, 4:26pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/9 "2026-09-23T16:26:58Z")
I convert to shell part. And then still some flaws: ![image.png](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/2/284bf6e49446a9592cc4af140338a855f8b1c4f1.png) mehmed çelebi \<[mehmed.celebi@gmail.com](mailto:mehmed.celebi@gmail.com)\>, 23 Eyl 2026 Çar, 18:58 tarihinde şunu yazdı: ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [September 23, 2026, 4:49pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/10 "2026-09-23T16:49:59Z")
To use shells, you should have only one surface (typically midsurface) for each rib or plate. And merge all surfaces. Here, they are probably still disconnected in some places. This geometry is too complex; you should start with a simplified version. You can also analyze a quarter of this model (utilize symmetry), but the geometry still needs careful preparation. ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [September 23, 2026, 4:56pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/11 "2026-09-23T16:56:50Z")
Actually, if you take a closer look at this geometry, ribs on one side are not fully connected: ![image](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/b/bcf35641d6ab718257db990769ee534b8bdd1396.png) On the other side: ![image](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/9/93f9990da48671516d3ab6212efd753aa09b2322.png) This is the problematic side: ![image](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/a/a1f4302ea2a7cf4ed6ba5c214707df311e8a0fec.png) This explains the asymmetry. Also, aren’t the units off ? ![image](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/a/ac578d6d8429c3ea4cb2befeebbfce5e3186da4e.png) Make sure that you export and import geometry in the proper unit system. ---
**Author:** ![mehmed.celebi](https://avatars.discourse-cdn.com/v4/letter/m/898d66/32.png) [@mehmed.celebi](https://prepomax.discourse.group/u/mehmed.celebi)\ **Post date:** [September 24, 2026, 12:07pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/12 "2026-09-24T12:07:25Z")
Hi, How can I analyze a quarter of this model. Thanks. FEAnalyst \<[notifications@prepomax.discoursemail.com](mailto:notifications@prepomax.discoursemail.com)\>, 23 Eyl 2026 Çar, 20:06 tarihinde şunu yazdı: ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [September 24, 2026, 12:21pm UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/13 "2026-09-24T12:21:02Z")
You would have to cut it in CAD software in such a way that you keep only a quarter of the whole part: ![piramit quarter](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/5/5d344410cbca607cd248c366eac8ce42cb467a23.png) In thermal analyses, you don’t need any specific boundary conditions for the symmetry edges/surfaces because the default BC allows for that. ---
**Author:** ![mehmed.celebi](https://avatars.discourse-cdn.com/v4/letter/m/898d66/32.png) [@mehmed.celebi](https://prepomax.discourse.group/u/mehmed.celebi)\ **Post date:** [September 29, 2026, 10:42am UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/14 "2026-09-29T10:42:49Z")
How can i define a liquid in the software to make heat transfer? Especially between two solids. MC FEAnalyst \<[notifications@prepomax.discoursemail.com](mailto:notifications@prepomax.discoursemail.com)\>, 24 Eyl 2026 Per, 15:31 tarihinde şunu yazdı: ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [September 29, 2026, 10:49am UTC](https://prepomax.discourse.group/t/fem-heat-transfer/3688/15 "2026-09-29T10:49:50Z")
There are 3 ways: 1. Model the solid’s interaction with a surrounding fluid in an approximate way without modeling the fluid explicitly. Instead, you can define convective heat transfer load with a proper heat transfer coefficient depending on the type of fluid and flow conditions. This is the most common approach in FEA for heat exchangers. 2. Model the fluid as a solid part with your actual solid component embedded inside it. Just make sure that you have a cutout and define contact or tie constraints. Alternatively, you can use compound parts and continuous mesh. Just make sure that heat can flow properly between these materials. Of course, the disadvantage is that you can only account for conduction in a steady fluid this way and it’s rarely needed in practice (at least for external fluid domains and cases like this). 3. Run CHT (Conjugate Heat Transfer) analysis in a different software such as OpenFOAM. This will be the most accurate approach (since you also analyze the fluid flow), but also the most difficult. You could import temperature distribution from CFD to PrePoMax as well. Whichever approach you choose, make sure that your heat exchanger itself is properly modeled and can be correctly simulated on its own.