Hey I am currently working on my Master thesis. To compare my results i need to use the latent heat feature of Abaqus.
O modeled a transient heat transfer analysis with a constant boundary of T = 1 at the one end. 
The rest of the beam (ord the rectangle) hast the temperature 0. 

I have a FEM software to calculate the Stefan Problem and validate it with the analytical solution of Jonsson (https://www.diva-portal.org/smash/get/diva2:647481/FULLTEXT01.pdf).

I expected that there ist no big difference between the 1d and 2d solution (round about numerical accuracy). In the attached diagram you can see the difference between the analytical solution and the one obtained from Abaqus.

The software i validate uses linear elements, so do i in Abaqus.

At the end is the input file withoud the nodes and elements.

Thank you for your effort best regards

Florian Ewald

The inut file for 1D looks like (for 2D the nodes get a second coordinate and the element type is DC2D4):

*Element, Type=DC1D2, ELSET=EALL

*Solid Section, elset=EALL, material=Mat

,

*End Part

*Assembly, name=Assembly

*Instance, name=Part-1-1, part=Part-1

*End Instance

*End Assembly

** MATERIALS

*MATERIAL, NAME=Mat

*DENSITY

1.0000000000

*SPECIFIC HEAT

1.0000000000

*Conductivity

0.0000100000

*Latent heat

0.1000000000, -0.001000, 0.001000

** PREDEFINED FIELDS

*Initial Conditions, type=TEMPERATURE

Part-1-1.NALL, 0.

*Time points, generate, name=timesteps

0.0000, 1000.0000, 1.0000

** STEP: Step-1

*Step, name=Step-1, nlgeom=No, inc= 100000

*Heat Transfer, end=PERIOD, deltmx=100.

1.000000, 1000.000000

** BOUNDARY CONDITIONS

** Name: BC-1 Type: Temperature

*Boundary

Part-1-1.Boun, 11, 11, 1.

** OUTPUT REQUESTS

*Restart, write, frequency=0

** FIELD OUTPUT: F-Output-1

*Output, field, time points=Timesteps

*Node Output

COORD, NT, RFL

*Element Output, directions=YES

HFL, TEMP

** HISTORY OUTPUT: H-Output-1

*Output, history, variable=PRESELECT

*End Step