I am fairly new to NX simulation but have done a few of the tutorials to get up to speed. Past experience has all been with Ansys workbench so I'm familiar with FEA and general workflow but need some guidance on specifics on NX Nastran.
I am trying to model a component method of manufacturing and in particular focus on distortion resulting from the below processes in order A-B-C-D.
A - Flat sheet metal laser cut to shape
B - Form sheet metal using press
C - Deposit boss using laser metal deposition
D - Heat treat (solution treatment)
Each process imparts some residual stresses and by the time we get to heat treatment it is distorted. The thinking is to hot set the shape of the part by restraining it during heat treatment. This requires fixtures properly designed for the part and heat treatment cycle which is ultimately the aim of this study.
Instead of modelling each process, I can assess the feasibility of hot setting using the following approach:
Attached images should make it easier to understand.
1 - Take flat sheet and apply a spatially variable displacement load to deliberately distort the part so that it looks similar to the part after process C, where flat part turns to saddle shape. I know the internal stresses will not be the same as if I had modelled laser deposition but I'm interested in effectiveness of hot setting for now.
This analysis will be transient elastic-plastic, non-linear and with use of displacement load, it wont require modelling of other bodies and therefore contact (SOL601).
2 - Apply a force on the fixture top and constrain the fixture base to flatten the part. Again transient elastic-plastic non-linear but including non-penetrating contact between fixture and part (SOL601).
3 - Apply the clamped sandwich assembly to a thermal solution (SOL159) and apply the heat treatment cycle (Vacuum furnace, 20C to 980C in 1 hour, hold 980C for 1 hr, argon quench and air cool to room temp - IN718 solution treatment for anyone interested). Then feed this to a structural (SOL601) to get final shape and stress.
I will assume the part uniformly heats up and soaks as this is applied to much bigger parts without issue.
I'm not sure what to do about the quench though. Perhaps all external surfaces of fixture and part can have a forced temperature to match the quench? This should still show the effect of different thermal masses at the boss and fixture intersections.
I have neglected the argon quench flow and its effect. My thinking is the part is small relative to furnace size so assuming uniform heating and cooling (1.58m diameter by 1m long furnace and part is 150mm x 150mm square).
I will use table fields to define temperature dependant properties up to 1000C and both fixture and part will be IN718. Im not sure how to factor creep in to this.
Im pretty confident with modelling, assembly meshing but the setup and linking of these seperate solutions is where I need advice.
My thinking is to have these as seperate solutions and map results from one to the next so structural > Thermal > Structural. Could I have this all as one fully defined simulation using condition sequences?
Any advice is appreciated, Thanks!