Simulating convective impingement heating in HASPIF Shubhamkar Kulkarni Advised by: Dr. Gregory Mocko

Introduction Single point incremental forming is a process for dieless forming of sheets Localized heating has shown to improve the formability (HASPIF) Empirical knowledge but limited theoretical understanding; need for simulation

Research Overview Objective: Simulation of HASPIF for improved understanding and planning Validation

Lit review Two possible approaches to model the moving heated spot: I. Using Computational Fluid Dynamics: CFD simulations techniques exists in ANSYS but computationally expensive Issues with updating the domain of the fluid II. Using thermal boundary conditions: 1. Modelling the heated spot as a moving heat flux Has been used for modelling welding and laser assisted SPIF of metals Not suitable as heat flux depends on the temperature difference 2. Modelling the heated spot as a temperature load Applying temperatures to different nodes 3. Modelling the heated spot as a convective/impingement jet (HTC distribution) Empirical correlations exist for Nusselt number distributions based on Reynolds number and some geometrical dimensionless numbers Valid for steady state conditions and comparatively high Reynolds number (>6000) Valid for stationary nozzle(s) without structural hindrance introduced due to the tool Requirement: model for moving heat source, transient heat transfer and low Reynolds number flow

Old setup Structural support: Hot air gun Issues with stiffness Experimental setup Old setup Structural support: Hot air gun Issues with stiffness New setup Structural support: Collet Additional chute to channel air Circular blank for symmetrical boundary conditions

Data Acquisition FluxTeq PHS 01-e heat flux sensor MS Excel Postprocessing LabView and NI 93210 Record the analog signals FluxTeq PHS 01-e heat flux sensor Measure heat flux and temperature

Theoretical background Equations used: Heat transfer through convection 𝑄=ℎ𝐴∆𝑇 Heat Flux 𝐻𝐹= 𝑄 𝐴 =ℎ∆𝑇 Convective heat transfer coefficient (HTC) 𝐻𝐹 ∆𝑇 =ℎ Record the heat flux and temperature, post process to calculate HTC

Test 1: Stationary nozzle test For quantifying boundary conditions required for version 1.0 of model Assumed the value of HTC to be equal in discrete concentric regions Similar to the data presented in the literature Procedure: Position the tool axis at a certain (X,Y) position with the tool barely touching the sensor Start recording data Switch on the hot air gun Stop recording after 60 seconds (approx.)

Test 1: Stationary nozzle test It was observed that the HTC value varied Literature shows similar results for water impingement cooling Since the duration of heat transfer is small, the variation of HTC value cannot be ignored Complexity is increased because the tool is also moving; the path of travel of the nozzle relative to a point’s location will result in different profile of HTC variation

Test 2: constant velocity nozzle movement test For quantifying boundary conditions for version 2.0 of the model The heated spot is split into two regions representing the leading and trailing regions relative to the tool motion Procedure: Position the centroid of the sensor at (0,0) Position the tool axis at a certain (-1.5,0) position Start recording data and switch on the hot air gun simultaneously Move the hot air gun with velocity equal to the feed used for forming Stop recording the data once the tool reaches (1.5,0) Direction of tool travel

Test 2: constant velocity nozzle movement test During post processing, the timestep associated with each recorded data point was used to identify the location of the tool axis The region was discretized into regions depending upon the value of the X coordinate The HTC values associated with all data points were averaged for each region

Test 2: constant velocity nozzle movement test   Upper bound Lower Bound ℎ 1 1.5 1 ℎ 2 0.5 ℎ 3 ℎ′ 3 -0.5 ℎ′ 2 -1 ℎ′ 1 -1.5

Test 2: constant velocity nozzle movement test Assumptions: The sensor is assumed to be a point object, located at its centroid. The values are essentially average values of flux on the sensor The manual error associated with the starting the data recording and the heat gun simultaneously is neglected The delay in response of the sensor and the delay to start the motion of the tool is neglected The value of HTC has been assumed to be constant in different discretized regions. In reality, it varies depending upon the location The HTC distribution is assumed to be independent of the deformation of the blank

Simulation workflow

Implementing in ANSYS

Simulation results

Validation Miniature thermocouples will be used Record temperature versus time data for validating experimental results Tests: Linear test, spiral test and if possible actual forming test Linear Spiral

Conclusion A model for simulating the transient convective heat transfer encountered in HASPIF is presented Validation studies are ongoing

References

Thank you

Outline Intro Lit review Theory Primary testing Proposed modelling Simulation results Validation