Units in ABAQUS (N-mm-s) and Mass Scaling Factor???

[marabou]

When I use ABAQUS/Explicit to solve a fully coupled thermomechanical indentation problem (with frictional contact), it cannot converge and even goes wrong in most cases. I thought that this could likely be a unit problem. Although it seems simple, I still have problem in setting the appropriate units. Whether or not the following settings are right?

Units: N,mm
Length = mm
Force = N
Mass = Tonne (1.0E3kg)
Time = second
stress = MPa (N/mm^2)
Energy = mJ (1.0E-3J)
Density = Tonne/mm^3
Conductivity = J/(s.mm.C)
Specific Heat = J/(Tonne.C)

If right, it does not work properly. In addition, do you know how to set a MASS SCALING FACTOR? Any guideline?

Thank you for your help

 

[jhardy1]

marabou,

I think you have one error in your units - in consistent N/mm/s units, Work / Energy is in Joules, not mJ.

The rest of your units look OK.

I can't help you with the "Mass Scaling Factor".

Hope this helps.

 

[harry123456]

Hi,
You can set the mass scaling factor by a number of methods. I think the most common method is setting a fixed mass scaling factor on the elements desired or by increasing the density directly.
Look in the abaqus analysis manual for the section on mass scaling.
You will get all the options there.
harry

 

[marabou]

Many Thanks, Julian Hardy and Harry,

Units have confused me for a quite while since I use ABAQUS. I try to run an example problem directly from ABAQUS Example eManual as "cylbillet_x_cax4rt_slow.inp". It uses N-m-Kg and works perfectly. However, when I changed all units to N-mm-Tonne, it does not work properly. How can I deal with the units of W and J in thermal problem?

For Mass Scaling, the factor is 1.0E+10 in the original code. I tried from 1.0 to 1.0E+10 in my test.inp code, but all do not work. The manual gives a guideline but not specific.

Again, thank you for your kind helps!

marabou


cylbillet_x_cax4rt_slow.inp --- Original from ABAQUS eManual with units of N-m-Kg.

*HEADING
- AXISYMMETRIC UPSETTING PROBLEM
- WITH RIGID SURFACE
- SECTION CONTROLS USED (HOURGLASS=STIFFNESS)
- SLOW UPSETTING
*NODE
1,
13,.01
1201,0.,.015
1213,.01,.015
*NGEN,NSET=MIDDLE
1,13
*NGEN,NSET=TOP
1201,1213
*NFILL
MIDDLE,TOP,12,100
*NSET,NSET=AXIS,GEN
1,1201,100
*ELEMENT,TYPE=cax4rt,ELSET=BILLET
1,1,2,102,101
*ELGEN,ELSET=BILLET
1,12,1,1,12,100,100
*NODE, NSET=NRIGID
2003,0.01,.02
*SOLID SECTION,ELSET=BILLET,
MATERIAL=METAL,CONTROL=B
*SECTION CONTROLS, HOURGLASS=STIFFNESS, NAME=B
*MATERIAL,NAME=METAL
*ELASTIC
200.E9,.3
*PLASTIC
7.E8,0.00
3.7E9,10.0
*RATE DEPENDENT
40.,5.
*SPECIFIC HEAT
586.,
*DENSITY
7833.,
*CONDUCTIVITY
52.,
*EXPANSION
1.2E-5,
*INELASTIC HEAT FRACTION
0.9,
*BOUNDARY
MIDDLE,2
AXIS,1
2003,1
2003,3,6
*SURFACE,TYPE=ELEMENT,NAME=BILLET
TOP,S3
SIDE,S2
*SURFACE,NAME=RIGID,TYPE=SEGMENTS
START, 0.02,.015
LINE, 0.00,.015
*RIGID BODY, REF NODE=2003,
ANALYTICAL SURFACE =RIGID
*ELSET,ELSET=TOP,GEN
1101,1112,1
*ELSET,ELSET=SIDE,GEN
12,1112,100
*NSET, NSET=SIDE, GEN
13, 1213, 100
*NSET, NSET=QA_TEST
SIDE,
*NSET, NSET=QA_TEST_NRIGID
NRIGID,
*ELSET, ELSET=QA_TEST
SIDE,
*STEP
*DYNAMIC TEMPERATURE-DISPLACEMENT, EXPLICIT
,100.0
*FIXED MASS SCALING, ELSET=BILLET,FACTOR=1.E+10
*BOUNDARY,TYPE=VELOCITY
2003,2,,-9.e-5
*SURFACE INTERACTION,NAME=RIG_BILL
*FRICTION
1.0,
*CONTACT PAIR,INTERACTION=RIG_BILL
RIGID,BILLET
*MONITOR,NODE=2003,DOF=2
*FILE OUTPUT,NUM=1
*EL FILE, ELSET=QA_TEST
PEEQ,
*NODE FILE, NSET=NRIGID
U,RF
*NODE FILE, NSET=QA_TEST
NT,
*OUTPUT,FIELD,VAR=PRESELECT,NUM=5
*OUTPUT,HISTORY
*NODE OUTPUT,NSET=NRIGID
U2,RF2
*ENERGY OUTPUT
ALLIE,ALLKE,ALLAE,ALLPD
******
*OUTPUT,FIELD,NUM=1, TIME MARKS=YES
*ELEMENT OUTPUT, ELSET=QA_TEST
PEEQ,
*NODE OUTPUT, NSET=QA_TEST
U, NT
*NODE OUTPUT, NSET=QA_TEST_NRIGID
RF,
*******
*END STEP


test.inp -- change the units from N-m-Kg to N-mm-Tonne

*HEADING
- AXISYMMETRIC UPSETTING PROBLEM
- WITH RIGID SURFACE
- SECTION CONTROLS USED (HOURGLASS=STIFFNESS)
- SLOW UPSETTING
- ADAPTIVE MESHING
*NODE
1, 0.0, 0.0
13, 10.0, 0.0
1201, 0.0, 15.0
1213, 10.0,15.0
*NGEN,NSET=MIDDLE
1,13
*NGEN,NSET=TOP
1201,1213
*NFILL
MIDDLE,TOP,12,100
*NSET,NSET=AXIS,GEN
1,1201,100
*ELEMENT,TYPE=CAX4RT,ELSET=BILLET
1,1,2,102,101
*ELGEN,ELSET=BILLET
1,12,1,1,12,100,100
*NODE, NSET=NRIGID
2003, 10.0, 20.0
*SOLID SECTION,ELSET=BILLET,MATERIAL=METAL,CONTROL=B
*SECTION CONTROLS, HOURGLASS=STIFFNESS, NAME=B
*MATERIAL,NAME=METAL
** *********************************************************
** Thermal Properties
** Use mild steel's thermal properties
** 1Joule=1Nm=1000Nmm ??
** 1W=1Nm/s=1000Nmm/s ??
** CONDUCTIVITY=52.0 W/(mK)
** =52.0 W/(1000mm.K)
** =52.0E-3 W/(mm.C)
** SPECIFIC HEAT=586.0 J/(Kg.C)
** =586.0 J/(10^-3.Tonne.C)
** =586.0E3 J/(Tonne.C)
**
** Other Units:
** Velocity = -0.9E-5 m/s = -0.9E-2 mm/s
** Density = 7833 Kg/m^3 = 7.8330E-9 Tonne/mm^3
** Young's Modulus = 210GPa = 210E9 N/m^2 = 210E3 N/mm^2
** *********************************************************
*ELASTIC
200.0E3, 0.3
*PLASTIC
7.E2, 0.00
3.7E3, 10.0
*RATE DEPENDENT
40.0, 5.0
*SPECIFIC HEAT
586.0E3,
*DENSITY
7.8330E-9,
*CONDUCTIVITY
52.0E-3,
*EXPANSION
1.2E-5,
*INELASTIC HEAT FRACTION
0.9,
*BOUNDARY
MIDDLE,2
AXIS,1
2003,1
2003,3,6
*SURFACE,TYPE=ELEMENT,NAME=BILLET
TOP,S3
SIDE,S2
*SURFACE,NAME=RIGID,TYPE=SEGMENTS
START, 20.0, 15.0
LINE, 0.00, 15.0
*RIGID BODY, REF NODE=2003, ANALYTICAL SURFACE =RIGID
*ELSET,ELSET=TOP,GEN
1101,1112,1
*ELSET,ELSET=SIDE,GEN
12,1112,100
*NSET, NSET=SIDE, GEN
13, 1213, 100
*NSET, NSET=QA_TEST
SIDE,
*NSET, NSET=QA_TEST_NRIGID
NRIGID,
*ELSET, ELSET=QA_TEST
SIDE,
*STEP
*DYNAMIC TEMPERATURE-DISPLACEMENT, EXPLICIT
,130.0
*FIXED MASS SCALING, ELSET=BILLET,FACTOR=1.0E+1
*ADAPTIVE MESH, ELSET=BILLET, FREQ=10,MESH SWEEPS=3
*BOUNDARY,TYPE=VELOCITY
2003, 2, , -9.E-2
*SURFACE INTERACTION,NAME=RIG_BILL
*FRICTION
0.2,
*CONTACT PAIR,INTERACTION=RIG_BILL
RIGID,BILLET
*MONITOR,NODE=2003,DOF=2
*FILE OUTPUT,NUM=1
*EL FILE, ELSET=QA_TEST
PEEQ,
*NODE FILE, NSET=NRIGID
U,RF
*NODE FILE, NSET=QA_TEST
NT,
*OUTPUT,FIELD,VAR=PRESELECT,NUM=5
*OUTPUT,HISTORY
*NODE OUTPUT,NSET=NRIGID
U2,RF2
*ENERGY OUTPUT
ALLIE,ALLKE,ALLAE,ALLPD
******
*OUTPUT,FIELD,NUM=1, TIME MARKS=YES
*ELEMENT OUTPUT, ELSET=QA_TEST
PEEQ,
*NODE OUTPUT, NSET=QA_TEST
U, NT
*NODE OUTPUT, NSET=QA_TEST_NRIGID
RF,
*******
*END STEP

 

[johnhors]

marabou,

As any analyst would (or should) know your troubles merely highlight the dangers of straying from the SI standard (kilogramme, metres, seconds) when working with metric units.

 

[ferro1001]

Hello,

I don´t know if this is your problem, but if you change units from N m Kg to N mm Tonne you have to scale your model in ABAQUS. For example, if you mesh a square of 3x3 meters in N m Kg units you have to draw a square 3x3 in ABAQUS; when you change loads and properties to N mm Tonnes you can´t use that square, you have to draw a square 3000x3000 if you want to analyse the same problem.

Hope this helps you.