COMSOL Forums: "Matrix has zero on diagonal" (AC/DC in 3D)

Re: "Matrix has zero on diagonal" (AC/DC in 3D)

Thu, 09 Sep 2010 19:25:09 +0000

With the current solver, u dont have to have zero conductivity, try to change solver or set a non-zero conductivity. when u reached the convergence try to reduce sigma.

Re: "Matrix has zero on diagonal" (AC/DC in 3D)

Mon, 16 Aug 2010 05:00:24 +0000

I have also the same problem with the same task (AC/DC, Quasi-static). In more details i have the cable model putted in the air and connected to the pspice circuit. Ich have perfectly done the similar task in 2D but in 3d mode - no chance:
"Failed to find consistent initial values. Matrix has zero on diagonal. Last time step is not converged". I have tried all what i know - changed in wide range the step size; different solving methods; pspice impulse begins not from null:
"I_IN 1 0 PULSE(1 100000 1e-7 10e-6 700e-6 0.1e-6)". And i dont know what can i do more.. ill be obliged to everyone help me.

Re: "Matrix has zero on diagonal" (AC/DC in 3D)

Tue, 06 Jul 2010 15:42:15 +0000

I faced with the same problem and I don't know what to do.

"Matrix has zero on diagonal" (AC/DC in 3D)

Mon, 31 Aug 2009 02:56:48 +0000

Has anyone encountered this error in 3D ac/dc quasi-static model?

I found this error can occur even with a single elementary shape without any physics setting.

Is this a bug or did I do anything wrong?

Thanks.