# No solving with C3D10 elements

**URL:** <https://calculix.discourse.group/t/no-solving-with-c3d10-elements/1452>\
**Category:** Uncategorized\
**Created:** [January 22, 2023, 3:09am UTC](https://calculix.discourse.group/t/no-solving-with-c3d10-elements/1452 "2023-01-22T03:09:59Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![dichtstoff](https://avatars.discourse-cdn.com/v4/letter/d/dec6dc/32.png) [@dichtstoff](https://calculix.discourse.group/u/dichtstoff)\
**Post date:** [January 22, 2023, 3:09am UTC](https://calculix.discourse.group/t/no-solving-with-c3d10-elements/1452/1 "2023-01-22T03:09:59Z")

</div>

I have generated with netgen a mesh with first order elements C3D4.  
With these mesh I run an static linear calc. and everything works fine.  
So i want to switch to second order elements and created a second mesh,  
with netgen and element second order  
With the same script i get no result with ccx 2.20 win and bconverged ccx 2.10 version  
and with SOLVER=SPOOLES

STEP 1  
Static analysis was selected

> Exit code: 0

I created the change with cgx with “mids all gen” from C3D4 mesh, but the same result, still no result!  
and no error output from ccx.

i run a calc. with CL35-win64 and ccx2.12-stable and it works fine, but with only 1 cpu.  
Now i have created a small mesh with netgen, and it works without problems,  
so i have changed to the standard solver PastiX and it works,  
i have 338333 elements and 523984 nodes,  
here you can download the org. files  
what is the reason? Has someone simular experiences?  
wbr

> **[Ihr persönlicher Freigabelink | GMX Cloud](https://c.gmx.net/@324820432001700943/1GwUtUyrQdCPlVLdaBbNDg)**
>
> Mit diesem Link erhalten Sie Zugriff auf die mit Ihnen geteilten Daten.

Update:

there is a problem with mesh,  
I have two singularities, mabe the mesh is to fine!?  
is there a posibilty to check the mesh? or recreate with cgx?

 ![image](https://global.discourse-cdn.com/free1/uploads/calculix/original/2X/9/935d8733886f6f32818645025801616f8e9e0fa7.png)

---

<div class="post-metadata">

**Author:** ![vicmw](https://avatars.discourse-cdn.com/v4/letter/v/6de8d8/32.png) [@vicmw](https://calculix.discourse.group/u/vicmw)\
**Post date:** [January 22, 2023, 11:03am UTC](https://calculix.discourse.group/t/no-solving-with-c3d10-elements/1452/2 "2023-01-22T11:03:32Z")

</div>

Spooles can’t solve that many nodes, at least not without 8-byte integers. Limit is about 300000. You need Pardiso or Pastix.

---

<div class="post-metadata">

**Author:** ![dichtstoff](https://avatars.discourse-cdn.com/v4/letter/d/dec6dc/32.png) [@dichtstoff](https://calculix.discourse.group/u/dichtstoff)\
**Post date:** [January 22, 2023, 11:22am UTC](https://calculix.discourse.group/t/no-solving-with-c3d10-elements/1452/3 "2023-01-22T11:22:12Z")

</div>

ok, maybe problem is the mesh,  
if i use the command merg n all ,  
i get these:  
merge ongoing, please wait  
evaluated:1001 from:89270  
evaluated:10001 from:89270

what is the meaning of these?

 ![image](https://global.discourse-cdn.com/free1/uploads/calculix/original/2X/6/609324b9e8700e0cb2a8cc097698bb5b8c930eed.png)