Cancel
Showing results for 
Search instead for 
Did you mean: 

FAILURE IN THE LANCZOS PROCEDURE

Experimenter
Experimenter

I have run a vibro-acoustic analysis on a FEM model of full vehicle that consists of structure and cavity fluid and the following error message is received. The model has 1 million elements. The frequency range of forced response simulation is between 10 to 200 Hz. The modes for both cavity and structure are calculated up to 400 Hz. I use the MAXSET parameter equal to 15. I use the “scr=yes”  command. The lower frequency limit in EIGRL card is left blank. I use the BEAT CAE ANSA for preprocessing and I run all NASTRAN checks before running.  What is the possible cause of this failure?  Thanks in advance

 

*** SYSTEM FATAL MESSAGE 3034 (LNNHERR)

     INTERNAL FAILURE IN THE LANCZOS PROCEDURE:

     M-ORTHOGONAL QR PROCEDURE FAILED TO CONVERGE.  PROBABLE CAUSE:

     MASS MATRIX IS INDEFINITE (MODES) OR STIFFNESS MATRIX IS INDEFINITE (BUCKLING).

     USER ACTION: CONTACT MSC CLIENT SUPPORT.

 *** SYSTEM FATAL MESSAGE 7340 (LNNHERR)

     WARNING REPORTED BY SUBROUTINE LNNDRVS  (IER=   728)

     USER INFORMATION:  GRAM-SCHMIDT DID NOT CONVERGE.

 *** SYSTEM WARNING MESSAGE 5407 (LNCCKS)

     INERTIA (STURM SEQUENCE) COUNT DISAGREES WITH THE NUMBER OF MODES ACTUALLY COMPUTED IN A (PARTIAL) INTERVAL

    FOUND    350 MODES BETWEEN -2.5535441E-03 HZ AND  1.5658104E+00 HZ BUT THERE SHOULD BE ONLY    318 MODES.

    USER ACTION:  THE FOLLOWING CHANGES TO THE EIGRL CARD MAY HELP TO SOLVE THIS PROBLEM

  1. INCREASE MAXSET.
  2. IF F1 IS GIVEN AS ZERO, LEAVE IT BLANK.
  3. IF F1 AND F2 ARE BOTH GIVEN, LEAVE F2 BLANK AND SPECIFY A VALUE FOR ND.
2 REPLIES

Re: FAILURE IN THE LANCZOS PROCEDURE

Phenom
Phenom

could you have to many rigid body modes? from the msg one reads that  there are 350 modes between 2E-3 and 1.56Hz ! Is that right?

Could it be a unit issue?

 

 

Production: NX9.0.3.4, NX10.0.2.6
Development: VB.NET (amateur level !)

Re: FAILURE IN THE LANCZOS PROCEDURE

Legend
Legend

You probably should 


Vahid_AIRIC wrote:

I have run a vibro-acoustic analysis on a FEM model of full vehicle that consists of structure and cavity fluid and the following error message is received. The model has 1 million elements. The frequency range of forced response simulation is between 10 to 200 Hz. The modes for both cavity and structure are calculated up to 400 Hz. I use the MAXSET parameter equal to 15. I use the “scr=yes”  command. The lower frequency limit in EIGRL card is left blank. I use the BEAT CAE ANSA for preprocessing and I run all NASTRAN checks before running.  What is the possible cause of this failure?  Thanks in advance

 

*** SYSTEM FATAL MESSAGE 3034 (LNNHERR)

     INTERNAL FAILURE IN THE LANCZOS PROCEDURE:

     M-ORTHOGONAL QR PROCEDURE FAILED TO CONVERGE.  PROBABLE CAUSE:

     MASS MATRIX IS INDEFINITE (MODES) OR STIFFNESS MATRIX IS INDEFINITE (BUCKLING).

     USER ACTION: CONTACT MSC CLIENT SUPPORT.

 *** SYSTEM FATAL MESSAGE 7340 (LNNHERR)

     WARNING REPORTED BY SUBROUTINE LNNDRVS  (IER=   728)

     USER INFORMATION:  GRAM-SCHMIDT DID NOT CONVERGE.

 *** SYSTEM WARNING MESSAGE 5407 (LNCCKS)

     INERTIA (STURM SEQUENCE) COUNT DISAGREES WITH THE NUMBER OF MODES ACTUALLY COMPUTED IN A (PARTIAL) INTERVAL

    FOUND    350 MODES BETWEEN -2.5535441E-03 HZ AND  1.5658104E+00 HZ BUT THERE SHOULD BE ONLY    318 MODES.

    USER ACTION:  THE FOLLOWING CHANGES TO THE EIGRL CARD MAY HELP TO SOLVE THIS PROBLEM

  1. INCREASE MAXSET.
  2. IF F1 IS GIVEN AS ZERO, LEAVE IT BLANK.
  3. IF F1 AND F2 ARE BOTH GIVEN, LEAVE F2 BLANK AND SPECIFY A VALUE FOR ND.

You probably should talk to MSC instead of Siemens & al., since this appears to be an MSC Nastran f06 (the hint being USER ACTION: CONTACT MSC CLIENT SUPPORT).  Unless of course, you've caught one of the lines of source code that escaped the search/replace MSC with EDS or Siemens Smiley Happy