how to set inital temperature new activitated element:abaqus 1

[p282h]

Hi:

I would like to know how to set initial temperature of new activated element each time in heat transfer analysis when it is carried out in multiple steps. (If the type of element used is DC3D20). I have tried to use *temperature in step modules when the new element is activated, but gave errors saying temperature cannot be used when temperature is the degree of freedom for elements. Then I tried using Boundary condition with Amplitude in tabular format (time=1, amp=1900, time=5, amp=0), where the total time period for the step is 5. The analysis kept aborting. The reason why I thought about it is to avoid if possible using the DFLUX subroutine for laser deposition.

 

[mrgoldthorpe]

*INITIAL CONDITIONS, TYPE=TEMPERATURE

 

[p282h]

Hey mrgoldthorpe:

I thought initial conditions can be set only once before all steps. I would like to set the nodes of each element that is activated to 1800C during each step. How can I do this?

 

[mrgoldthorpe]

So you want to set the temperature of all nodes on a particular set of elements to 1800 degC at the beginning of each step? If so, presumably this is the region where the laser is focused.

The only thing I can think of at the moment is a bit long-winded, as follows.

You replicate that element set n times, where 'n' is the number of analysis steps. Let's say n=5.

Use

*INITIAL CONDITIONS, TYPE=TEMPERATURE...

to set the initial temperature to 1800 degC of all nodes sets belonging to all these elements.

In an initial dummy step, remove all of these sets apart from one, let's say this is "ETEMPSET1".

so:
*STEP
*MODEL CHANGE, REMOVE, ELSET=ETEMPSET2
** down to
*MODEL CHANGE, REMOVE, ELSET=ETEMPSET5
...
*STEP
** do the first temperature transient
...
*STEP
** remove ETEMPSET1 and add in ETEMPSET2
*MODEL CHANGE, REMOVE, ELSET=ETEMPSET1
*MODEL CHANGE, ADD, ELSET=ETEMPSET2
...
*STEP
** do the second temperature transient
...

and so on.

You have to ensure:

* any heat fluxes, boundary conditions etc. applied to the relevant elements in each step would have to refer to the appropriate element/nodes sets used

* you'd have to use MPCs to connect the boundary nodes of the replica elements to the rest of the model

Someone might suggest a simpler method.

 

[p282h]

Hi mrgoldthorpe:

I have tried to do what you said to set the initial conditions of the nodes and deactivate and activate. It works, but the problem is in the vizualization module at the interface between the activated element set and the substrate which is at room temp (created a single part and partitioned), the nodes have two temps (300 & 2000 K), i.e.the elements are dead in the given step but some temp contours are seen like 2000 K on the top nodes, 1500 K in the middle and at the interface 300K. If you look at it in the dead state it seems there is conduction but obviously there is not until it is activated. How can I avoid this or make the dead elements disappear in the unactivated state and appear in activated?.

 

[mrgoldthorpe]

I'm not a CAE expert so I can't answer that question. However, I recall a thread here within the last year where this issue was discussed.

Otherwise try opening another thread along the lines of 'How to avoid display of inactive elements.'

 

[p282h]

Hi mrgoldthorpe:

I have been trying to simulate laser deposition and created a solid part (extrude-1) and extruded another solid from the same part. I partitioned the extrude-2 and created cells. I used model change to activated deactivate cells, and gave ICs 2000K to deposit and 300 K substrate. Meshed the part as single part. I see conduction occurring from substrate into deposit and not the reverse. I have not given any interactions, BCs,and loads but just ICs. Please tell me what is happening.

 

[p282h]

Hi mrgoldthorpe:


Can you give me any idea or skeleton on how to work on phase transformation using HETVAL?.

Thanks,