## impact-users — Help, Comments &amp; Discussions regarding Impact

You can subscribe to this list here.

 2005 2006 2007 2008 2010 2011 Jan Feb (2) Mar Apr (7) May (6) Jun (5) Jul Aug (3) Sep (22) Oct (13) Nov (6) Dec (2) Jan (10) Feb (11) Mar (8) Apr (5) May (6) Jun (14) Jul (3) Aug Sep (8) Oct (5) Nov Dec (6) Jan (6) Feb Mar (16) Apr (1) May (9) Jun (2) Jul Aug (6) Sep Oct Nov (1) Dec Jan Feb (2) Mar Apr (11) May (3) Jun (3) Jul (6) Aug (1) Sep Oct (3) Nov Dec (5) Jan (2) Feb Mar Apr May Jun Jul (2) Aug Sep Oct Nov Dec Jan Feb (1) Mar Apr May Jun Jul Aug Sep Oct Nov Dec
S M T W T F S

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15
(3)
16

17
(1)
18

19

20

21

22

23

24

25

26

27

28

29

30
(3)
31
(2)

Showing 9 results of 9

 Re: [Impact-users] can the thickness of a contact triangle affect matrix singularity in contact problem? From: Jonas Forssell - 2007-05-31 16:49:43 ```Alexander, The reason for your error can be many. Usually, you can get a feeling for if it is a contact problem by checking the last printed timestep. If the model looks like it is "exploding" it might be the contact which is causing it. Usually, a node slips by detection and is detected so late that it is already deeply penetrated within the contact triangle "thickness". When it is detected, it is repelled by a force. Since it is deep in the contact, the force will be very large and the node will be "shot away", bringning the element with it and the model looks like exploding. The contact triangle works by checking if a node is within its surface. If a node is just along the edge of the surface, this might be detected or not. To fix this, I usually use contact line elements along the sides of the contact triangle. The contact line elements will detect the node and make sure no penetration occurs. The diameter of the contact line element should then be the same as the contact triangle thickness. Another way is to slightly offset one of the objects so that the nodes does not align. The likelyhood of the contact triangle element detecting the nodes on its surface will then increase. This is less robust than the first method, but is easy to implement. /Jonas Låt vårkänslorna spira! Hitta din tvillingsjäl på Yahoo! Dejting: http://se.meetic.yahoo.net ```
 Re: [Impact-users] can the thickness of a contact triangle affect matrix singularity in contact problem? From: Jens Bomholt - 2007-05-31 15:43:35 Attachments: Message as HTML ```Hey Alexander My advice for your problem would bee: make a very simple model (10 -100 elements) and then try some cases with different thickness, time step etc. And then make an article on the wiki with the result. Regards JB On 5/30/07, Alexander Rabtchevich wrote: > > Thank you for pointing me to the document. I really forgot about it. > > So, if I understood correctly, 2 conditions should be fulfilled: the > time of sound wave propagation within any element should be larger than > the time step and the time of element moving through contact element > thickness should be larger than time stamp too. As I programmed in Gmsh, > the smallest characteristic length is 0.01, so direct calculation gives > time step 1.9e-6. As some elements could be even smaller, I hope I > understood the reason of such a small time step requirement. > > I still have several questions :). > > If there any reasonable rule to choose contact elements thickness? At > least the built-in help doesn't explain the reason, only the geometry > and syntaxes. I thought that if I calculate overall contact > displacements below 50 micrometers, 5 micrometers should be enough. Was > I right? > > When I set initial before-the-collision velocity of one of the bodies, > should the period before OFF be larger than the time step or it doesn't > matter? > > > > Jens Bomholt wrote: > > Hi Alexander > > > > Try to read this article: > > http://impactprogram.wikispaces.com/AboutTimesteps > > > > regards > > JB > > > > > > > > > > On 5/30/07, *Alexander Rabtchevich* > > > > wrote: > > > > I'm solving a contact problem of intrusion of rigid (elastic) > indenter > > into a steel plate. I set contact element thicknesses to 0.005 (mm) > due > > to calculating micro meter displacements. The solver quits due > matrix > > singularity after some calculations all the time. So I have several > > questions. > > 1. Does the contact triangle thickness matter? > > 2. Why should I set time step to such a small increments? I use > default > > mm units. If I set the time step bigger than 0.5e-6 milliseconds, > the > > solver quits very soon. The strains are not too large even for much > > bigger steps. > > > > > > -- > > With respect > > Alexander Rabtchevich > > > -- > With respect > Alexander Rabtchevich > ```
 Re: [Impact-users] can the thickness of a contact triangle affect matrix singularity in contact problem? From: Alexander Rabtchevich - 2007-05-30 15:09:31 ```Thank you for pointing me to the document. I really forgot about it. So, if I understood correctly, 2 conditions should be fulfilled: the time of sound wave propagation within any element should be larger than the time step and the time of element moving through contact element thickness should be larger than time stamp too. As I programmed in Gmsh, the smallest characteristic length is 0.01, so direct calculation gives time step 1.9e-6. As some elements could be even smaller, I hope I understood the reason of such a small time step requirement. I still have several questions :). If there any reasonable rule to choose contact elements thickness? At least the built-in help doesn't explain the reason, only the geometry and syntaxes. I thought that if I calculate overall contact displacements below 50 micrometers, 5 micrometers should be enough. Was I right? When I set initial before-the-collision velocity of one of the bodies, should the period before OFF be larger than the time step or it doesn't matter? Jens Bomholt wrote: > Hi Alexander > > Try to read this article: > http://impactprogram.wikispaces.com/AboutTimesteps > > regards > JB > > > > > On 5/30/07, *Alexander Rabtchevich* > > wrote: > > I'm solving a contact problem of intrusion of rigid (elastic) indenter > into a steel plate. I set contact element thicknesses to 0.005 (mm) due > to calculating micro meter displacements. The solver quits due matrix > singularity after some calculations all the time. So I have several > questions. > 1. Does the contact triangle thickness matter? > 2. Why should I set time step to such a small increments? I use default > mm units. If I set the time step bigger than 0.5e-6 milliseconds, the > solver quits very soon. The strains are not too large even for much > bigger steps. > > > -- > With respect > Alexander Rabtchevich -- With respect Alexander Rabtchevich ```
 Re: [Impact-users] can the thickness of a contact triangle affect matrix singularity in contact problem? From: Jens Bomholt - 2007-05-30 13:36:48 Attachments: Message as HTML ```Hi Alexander Try to read this article: http://impactprogram.wikispaces.com/AboutTimesteps regards JB On 5/30/07, Alexander Rabtchevich wrote: > > I'm solving a contact problem of intrusion of rigid (elastic) indenter > into a steel plate. I set contact element thicknesses to 0.005 (mm) due > to calculating micro meter displacements. The solver quits due matrix > singularity after some calculations all the time. So I have several > questions. > 1. Does the contact triangle thickness matter? > 2. Why should I set time step to such a small increments? I use default > mm units. If I set the time step bigger than 0.5e-6 milliseconds, the > solver quits very soon. The strains are not too large even for much > bigger steps. > > > -- > With respect > Alexander Rabtchevich > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Impact-users mailing list > Impact-users@... > https://lists.sourceforge.net/lists/listinfo/impact-users > ```
 [Impact-users] can the thickness of a contact triangle affect matrix singularity in contact problem? From: Alexander Rabtchevich - 2007-05-30 08:02:46 ```I'm solving a contact problem of intrusion of rigid (elastic) indenter into a steel plate. I set contact element thicknesses to 0.005 (mm) due to calculating micro meter displacements. The solver quits due matrix singularity after some calculations all the time. So I have several questions. 1. Does the contact triangle thickness matter? 2. Why should I set time step to such a small increments? I use default mm units. If I set the time step bigger than 0.5e-6 milliseconds, the solver quits very soon. The strains are not too large even for much bigger steps. -- With respect Alexander Rabtchevich ```
 [Impact-users] Compiling Impact From: Jens Bomholt - 2007-05-17 11:10:29 Attachments: Message as HTML ```Hi All I try to compile Impact with ant and get this error: [javac] 588 problems (1 error, 587 warnings) BUILD FAILED /home/dkjeboje/Impact/build.xml:71: Compile failed; see the compiler error output for details. I a log file I see this: [javac] 1. ERROR in /home/dkjeboje/Impact/src/j3d/Canvas3DSW.java [javac] (at line 32) [javac] import com.sun.image.codec.jpeg.*; [javac] ^^^^^^^^^^^^^ [javac] The import com.sun.image cannot be resolved My system is Fedora 5 with sun java. The com.sun.image.codec.jpeg filed have I found in: /usr/java/j2sdk1.4.2_14/jre/lib/rt.jar and in: /usr/java/jdk1.5.0_10/jre/lib/rt.jar How can I point/link ant too see and use these libraries ? regards Jens Bomholt ```
 Re: [Impact-users] impact solver error java.lang.RuntimeException: Matrix is singular. From: Jens Bomholt - 2007-05-15 15:19:56 Attachments: Message as HTML ```Hi Alexander try to decrease the time step: CONTROLS run from 0.0 to 1.0 step 1e-6 print every 0.01 step Look in the console when you start Impact to see what time step size Impact calculate! and then make a new time step: .5 * (Impact sugstion ) regards Jens Bomholt On 5/15/07, Alexander Rabtchevich wrote: > > I created a mesh with Gmsh 2.0.7, saved it to a file. Then I wrote a > Delphi program to convert Gmsh *.msh into *.in for tetrahedral elements. > The result opened in Impact without any problems. I assigned constrains > and a force, saved the file and tried to solve the problem. But the > solver complained: impact solver error java.lang.RuntimeException: > Matrix is singular. > > I sent the cube1.in file to the Gmsh developers and was told the mesh is > perfect from Gmsh point of view. What is wrong with the file from the > Impact POV? > > -- > With respect > Alexander Rabtchevich > > # Created Impact Pre Processor > # Mon May 14 18:54:17 EEST 2007 > > CONTROLS > run from 0.0 to 1.0 > print every 0.01 step > > > MATERIALS OF TYPE ELASTOPLASTIC > STEEL_EP E = 210 RHO = 0.0000078 NU = 0.3 YIELD_STRESS = 0.180 EP = 0.1 > > CONSTRAINTS OF TYPE BOUNDARY_CONDITION > FIXED_ALL VX = 0 VY = 0 VZ = 0 VRX = 0 VRY = 0 VRZ = 0 AX = 0 AY = 0 AZ = > 0 ARX = 0 ARY = 0 ARZ = 0 > > LOADS > FORCE_Z FZ = -5 > > NODES > 1 X = 0.0 Y = 0.0 Z = 0.0 > 2 X = 0.0 Y = 1.0 Z = 0.0 > 3 X = 1.0 Y = 1.0 Z = 0.0 Load = FORCE_Z > 4 X = 1.0 Y = 0.0 Z = 0.0 > 5 X = 0.0 Y = 0.0 Z = -1.0 Constraint = > FIXED_ALL > 6 X = 1.0 Y = 1.0 Z = -1.0 Constraint = > FIXED_ALL > 7 X = 1.0 Y = 0.0 Z = -1.0 Constraint = > FIXED_ALL > 8 X = 0.0 Y = 1.0 Z = -1.0 Constraint = > FIXED_ALL > 9 X = 0.6444356 Y = 0.35574186 Z = -0.64425826 > 10 X = 0.25734907 Y = 0.49407724 Z = -0.50590336 > 11 X = 0.50594425 Y = 0.49411032 Z = -0.2573258 > 12 X = 0.5058657 Y = 0.74267584 Z = -0.5057951 > > > ELEMENTS OF TYPE Solid_Iso_4 > 1 nodes = [8,7,6,9] Material = STEEL_EP NIP = 1 > 2 nodes = [5,7,8,9] Material = STEEL_EP NIP = 1 > 3 nodes = [7,5,1,9] Material = STEEL_EP NIP = 1 > 4 nodes = [5,2,1,10] Material = STEEL_EP NIP = 1 > 5 nodes = [8,2,5,10] Material = STEEL_EP NIP = 1 > 6 nodes = [7,1,4,9] Material = STEEL_EP NIP = 1 > 7 nodes = [2,4,1,11] Material = STEEL_EP NIP = 1 > 8 nodes = [3,4,2,11] Material = STEEL_EP NIP = 1 > 9 nodes = [8,3,2,12] Material = STEEL_EP NIP = 1 > 10 nodes = [6,3,8,12] Material = STEEL_EP NIP = 1 > 11 nodes = [11,9,10,12] Material = STEEL_EP NIP = 1 > 12 nodes = [6,7,4,9] Material = STEEL_EP NIP = 1 > 13 nodes = [11,9,3,4] Material = STEEL_EP NIP = 1 > 14 nodes = [8,9,6,12] Material = STEEL_EP NIP = 1 > 15 nodes = [9,8,10,12] Material = STEEL_EP NIP = 1 > 16 nodes = [5,1,9,10] Material = STEEL_EP NIP = 1 > 17 nodes = [10,2,11,12] Material = STEEL_EP NIP = 1 > 18 nodes = [9,10,8,5] Material = STEEL_EP NIP = 1 > 19 nodes = [9,1,11,10] Material = STEEL_EP NIP = 1 > 20 nodes = [10,11,2,1] Material = STEEL_EP NIP = 1 > 21 nodes = [3,9,11,12] Material = STEEL_EP NIP = 1 > 22 nodes = [3,11,2,12] Material = STEEL_EP NIP = 1 > 23 nodes = [8,12,2,10] Material = STEEL_EP NIP = 1 > 24 nodes = [3,9,6,4] Material = STEEL_EP NIP = 1 > 25 nodes = [6,9,3,12] Material = STEEL_EP NIP = 1 > 26 nodes = [9,11,1,4] Material = STEEL_EP NIP = 1 > > > nodes = [8,12,2,10] Material = STEEL_EP NIP = 1 > 24 nodes = [3,9,6,4] Material = STEEL_EP NIP = 1 > 25 nodes = [6,9,3,12] Material = STEEL_EP NIP = 1 > 26 nodes = [9,11,1,4] Material = STEEL_EP NIP = 1 > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Impact-users mailing list > Impact-users@... > https://lists.sourceforge.net/lists/listinfo/impact-users > > ```
 Re: [Impact-users] impact solver error java.lang.RuntimeException: Matrix is singular. From: Alexander Rabtchevich - 2007-05-15 07:22:18 ```The last incorrect lines appeared after saving the file in the solver. The original file does not contain them. And the solver complains for the clean file. Alexander Rabtchevich wrote: > > I sent the cube1.in file to the Gmsh developers and was told the mesh is > perfect from Gmsh point of view. What is wrong with the file from the > Impact POV? > > > ------------------------------------------------------------------------ > -- With respect Alexander Rabtchevich ```
 [Impact-users] impact solver error java.lang.RuntimeException: Matrix is singular. From: Alexander Rabtchevich - 2007-05-15 06:56:18 Attachments: cube1.in ```I created a mesh with Gmsh 2.0.7, saved it to a file. Then I wrote a Delphi program to convert Gmsh *.msh into *.in for tetrahedral elements. The result opened in Impact without any problems. I assigned constrains and a force, saved the file and tried to solve the problem. But the solver complained: impact solver error java.lang.RuntimeException: Matrix is singular. I sent the cube1.in file to the Gmsh developers and was told the mesh is perfect from Gmsh point of view. What is wrong with the file from the Impact POV? -- With respect Alexander Rabtchevich ```

Showing 9 results of 9