Cooling and hardening during injection molding of field joint coatings for deep sea pipelines

Transcription

1 Cooling and hardening during injection molding of field joint coatings for deep sea pipelines Robrecht Verhelle 1, Luk Van Lokeren 1,2, Samir Loulidi 1, Guy Van Assche 1, Guido Ridolfi 3, Helen Boyd 3 1 Vrije Universiteit Brussel, Physical Chemistry and Polymer Science, Belgium 2 Materials Innovation Institute, The Netherlands 3 Heerema Marine Contractors, The Netherlands

2 Field joint coatings for deep sea pipelines 2 Field joint coating application process Pipe sections Welded together Surface cleaning Grit blasting FBE application Corrosive protection Injection moulding Thermal insulation Field joint Needs cooling

3 Field joint coatings for deep sea pipelines p. 3 Chamfer Overlap FJC FAC Steel Toe Challenges: - Insufficient bonding and cracks - Processing optimization Aim: 1. Understanding FJC process 2. Support optimization FJC process: Materials Process conditions Approach: 1. Build a FEM model 2. Acquire experimental data

4 Physicochemical aspects of the FJC process p. 4 Heat transfer Cooling rates Process conditions Thermal shrinkage Geometry Structural mechanics Visco-elastic properteries Stress calculation Crystallization Crystallization kinetics Crystallization shrinkage Multiphysics problem: Heat transfer Crystallization Structural mechanics

5 p. 5 Model: geometry and process steps Geometry FJC FAC FJC FAC Steel Field joint coating (FJC) Factory applied coating (FAC) Pretreatment Process steps Induction heating Cooling Infrared heating Crystallization and cooling Mold filling Cooling in mold Cooling in air Fast Slow Slow

6 Simulation pretreatment steps p. 6 Purpose: obtain realistic T-distributions for cooling and crystallisation 1. Induction heating 2. Cooling 3. Infrared heating 1. Induction heating Domain heat scource T max steel reached 260 C Natural convection: long horizontal cylinder Symmetric boundary FAC Steel Convective cooling Infinite element domain Temperature Potential degradation polymer at toe

7 Simulation pretreatment steps p. 7 Purpose: obtain realistic T-distributions for cooling and crystallisation 1. Induction heating 2. Cooling 3. Infrared heating 2. Cooling + FBE application FAC acts as heat sink T max steel 235 C Natural convection: long horizontal cylinder Symmetric boundary FAC Steel Convective cooling Infinite element domain Temperature Foto pijp voor FJ

8 Simulation pretreatment steps p. 8 Purpose: obtain realistic T-distributions for cooling and crystallisation 1. Induction heating 2. Cooling 3. Infrared heating 3. Infrared heating chamfers Domain heat scource Beer-Lambert law T max at middle chamfer 100 C Natural convection: long horizontal cylinder Symmetric boundary FAC Steel Convective cooling Infinite element domain Temperature

9 Simulation pretreatment steps vs. experimental data Temperature ( C) p. 9 Induction heating Cooling IR heating 250 TC1 TC2 TC3 A. Induction heating TC4 TC5 200 B. Cooling C. Infrared heating Locations thermocouples 0 Time (min) Experimental (Solid) Model (Dashed)

10 Rate p. 10 Cooling and crystallization: principles polymer crystallization Polymer crystal morphology Folded chain lamellar (single) crystals Lamellar thickness L lam : L lam 1/(T M -T C ) 1/DT Formation spherulites L lam Differential scanning calorimetry in cooling Polymer crystallization kinetics -10 K/min -0.5 K/min Temperature ( C)

11 Crystallization kinetics model: description Nucleation Haudin & Chenot model 1 : N number of potential nuclei N a number activated of potential nuclei Growth α relative crystallinity Aid functions Implemented in COMSOL using domain ODEs Domain heat source for crystallization heat: FJC FAC 21/10/ J.-M. Haudin, J.-M. Chenot, Intern. Polym. Process, XIX (2004) p. 11

12 Crystallization kinetics model: instabilities 21/10/2015 p. 12 Relative crystallinity α: Numerical calculation out of limits [0,1] Instability Avoided by transformation function Stable solution: b [-, + ] a [0,1]

13 Cooling and crystallization: in mold p. 13 Short injection time T mold =15 C T FJC =220 C Large ΔT Mesh refinement at boundaries DOF: Natural convection: long horizontal cylinder Symmetric boundary FJC Convective cooling Steel FAC Infinite element domain Temperature Crystallinity

14 p. 14 Cooling and crystallization: in air Short injection time T mold =15 C T FJC =220 C Large ΔT Mesh refinement at boundaries DOF: Demold after minutes After hours still liquid in center Thickness of the crystallized layer compares favorably to experiment Natural convection: long horizontal cylinder Symmetric boundary FJC Convective cooling Steel FAC Infinite element domain Temperature Crystallinity

15 Temperature ( C) p. 15 Cooling and crystallization: model vs. experimental Cooling in mold Cooling in air A. Cooling in mold TC1 TC2 TC3 TC4 TC5 150 B. Cooling in air Locations thermocouples 50 0 Time (min) Experimental (Solid) Model (Dashed)

16 Specific Volume (cm³/ g) 21/10/2015 Extension of the Model: Estimation Stresses p. 16 Heat transfer Thermal shrinkage Crystallization Crystallization shrinkage Structural mechanics Stress calculations Implementation Comsol 1.4 Thermal shrinkage Liquid Modules: - Heat transfer - Structural Mechanics Solid Thermal expansion Linear elastic material: Polymer: modulus T and α dependent Crystallization shrinkage Temperature ( C) Crystallization shrinkage via α dependent initial strain

17 Extension of the Model: Estimation Stresses p. 17 Crystallinity (top) and stress (bottom) distributions upon demolding Crystallinity (Top) and Stress (Bottom) Crystallinity Stress Crystallinity Stress Approximation: Linear elastic material Future Visco-elastic material

18 18 Extension of the model: reactive systems Curing polymer systems: polyurethanes dx dt Cure kinetics autocatalytic model with diffusion control k overall 1 x n kchemical k1 k2 k diffusion Z D with x m E C D exp RT C2 k k 1 1 overall i A T T T T i k g 1 chemical E i exp RT x x g k 1 diffusion Polyurethane FJC Polyurethane FAC Steel During cure During cure Cooling over 1 day Conversion Temperature Temperature Middle of the FJC (weld): steel pipe, IMPU inner skin (1 mm), center, IMPU outer skin (1 mm) Observation: (nearly) adiadiabatic temperature rise in center

19 Conclusions p. 19 Heat transfer and phase change model - Crystallization kinetics model T-dependent nucleation and growth Variable transformations to avoid numerical errors - Start of induced stress calculations Modelling opportunities Optimisation - Process - pretreatment steps - injection moulding - Material - Geometry

20 Acknowledgements p. 20