3D COMSOL Multiphysics Model of a Plate Heat Exchanger to ...3D COMSOL Multiphysics® Model of a...

1
3D COMSOL Multiphysics ® Model of a Plate Heat Exchanger to Support a Laboratory Teaching Environment William M. Clark and Nicholas J. Medeiros Chemical Engineering Department, Worcester Polytechnic Institute, Worcester, MA, USA Introduction: Chemical engineering students and practitioners need an understanding of the flow and heat transfer inside heat exchangers in order to analyze, design, and utilize them efficiently. Because the flow pattern inside of plate type heat exchangers is difficult to visualize, we are developing COMSOL models of a plate heat exchanger for students to study alongside a physical heat exchanger in a laboratory setting. With COMSOL simulations, students can easily adjust flow rates, flow direction, and thermal fluid types as well as internal corrugations to study the effects these parameters have on heat transfer. Post processing capabilities allow students to visualize the velocity, pressure, and temperature profiles inside the exchanger. Conclusions: Simulations allow students to investigate quickly a wide range of operating conditions and learn their effects on heat exchanger performance. Visualizing the velocity, pressure, and temperature profiles inside the exchanger provides a deeper understanding of the process. In particular, temperature profiles illustrate why countercurrent is preferred over co-current operation and velocity profiles illustrate why heat transfer coefficients increase with increasing flow rate and increasing internal corrugations. Figure 3. Simplified four channel models with and without internal corrugations; velocity and temperature plots with 1 L/min water in countercurrent flow. Figure 4. Temperature profiles in adjacent channels near center of 4 channel heat exchanger in co-current and countercurrent flow at 1 L/min water both streams. Figure 1. Plate heat exchanger experiment, cut away view of exchanger, and SolidWorks model. Computational Methods: The heat exchanger geometry was built to manufacturing specifications using the CAD package SolidWorks. Incompressible laminar flow physics was used for both the hot fluid and the cold fluid flow modeling with assigned inlet flow rates and pressure outlet boundary conditions. Heat transfer in fluids physics was used together with the results from the laminar flow studies to calculate the heat transfer and temperature profiles in the exchanger. Models were studied with water in both flow streams as well as with water in the cold stream and ethylene glycol in the hot stream. The stainless steel plates were modeled using the thin thermally resistive layer boundary condition. To provide for faster solution and allow more studies to identify expected trends during a typical lab setting, simplified 4 channel models with different internal plate geometries were built in addition to the realistic 10 plate model with chevron shaped internals shown in Figures 1 and 2. Co-current Countercurrent Direction Plate Area Flow rate ΔP Tc,out Th,out Q U Efficiency cm 2 L/min Pa K K W W/m 2 K % Co Straight 382 1 56 288.5 304.6 580 618 25.2 Counter Straight 382 1 56 290.0 303.1 689 783 30.0 Counter Straight 382 2 154 287.5 305.8 1018 1036 22.1 Counter Straight 382 3 289 286.1 307.0 1244 1209 18.0 Counter Corrugated 423 1 414 297.3 295.9 1195 1788 52.0 Counter Corrugated 423 3 2780 297.0 296.3 3525 5160 51.1 Figure 2. Complex flow pattern inside 10 plate exchanger with chevron shaped internals. Results: Figure 3 shows velocity and temperature result plots for the simplified four channel models with water at 1 L/min. Internal corrugations increase the heat transfer by increasing the surface area and the local velocity. Table 1 shows how flow pattern, internal corrugations, and flow rate affect the performance of the heat exchanger. The heat transfer coefficient increases with increasing velocity and increased corrugations, but at the expense of increased pressure drop. Figure 4 shows the temperature profile for adjacent channels near the center of the four channel exchanger without corrugations for both co-current and countercurrent flow. This clearly illustrates why countercurrent operation is preferred due to higher average driving force. Table 1. Results for 4 channel exchanger with water in both streams, with and without internal corrugations. Tc,in = 280.15, Th,in = 313.15 K. Future Work: Our plans for this work include using the models to develop and test correlations for heat transfer coefficients and exchanger efficiency with different internal geometries and an extension to turbulent flow. It is also of interest to match the model and experimental results more accurately to illustrate the value of simulation in predicting quantitative results in complex systems. Excerpt from the Proceedings of the 2014 COMSOL Conference in Boston

Transcript of 3D COMSOL Multiphysics Model of a Plate Heat Exchanger to ...3D COMSOL Multiphysics® Model of a...

Page 1: 3D COMSOL Multiphysics Model of a Plate Heat Exchanger to ...3D COMSOL Multiphysics® Model of a Plate Heat Exchanger to Support a Laboratory Teaching Environment William M. Clark

3D COMSOL Multiphysics® Model of a Plate Heat Exchanger to Support a Laboratory Teaching Environment

William M. Clark and Nicholas J. Medeiros Chemical Engineering Department, Worcester Polytechnic Institute, Worcester, MA, USA

Introduction: Chemical engineering students and practitionersneed an understanding of the flow and heat transfer inside heatexchangers in order to analyze, design, and utilize them efficiently.Because the flow pattern inside of plate type heat exchangers isdifficult to visualize, we are developing COMSOL models of aplate heat exchanger for students to study alongside a physicalheat exchanger in a laboratory setting. With COMSOLsimulations, students can easily adjust flow rates, flow direction,and thermal fluid types as well as internal corrugations to studythe effects these parameters have on heat transfer. Postprocessing capabilities allow students to visualize the velocity,pressure, and temperature profiles inside the exchanger.

Conclusions: Simulations allow students to investigate quicklya wide range of operating conditions and learn their effects onheat exchanger performance. Visualizing the velocity,pressure, and temperature profiles inside the exchangerprovides a deeper understanding of the process. In particular,temperature profiles illustrate why countercurrent is preferredover co-current operation and velocity profiles illustrate whyheat transfer coefficients increase with increasing flow rate andincreasing internal corrugations.

Figure 3. Simplified four channel models with and without internal corrugations; velocity and temperature plots with 1 L/min water in countercurrent flow.

Figure 4. Temperature profiles in adjacent channels near center of 4 channel heat exchanger in co-current and countercurrent flow at 1 L/min water both streams.

Figure 1. Plate heat exchanger experiment, cut away view of exchanger, and SolidWorks model.

Computational Methods: The heat exchanger geometry was builtto manufacturing specifications using the CAD packageSolidWorks. Incompressible laminar flow physics was used forboth the hot fluid and the cold fluid flow modeling with assignedinlet flow rates and pressure outlet boundary conditions. Heattransfer in fluids physics was used together with the results fromthe laminar flow studies to calculate the heat transfer andtemperature profiles in the exchanger. Models were studied withwater in both flow streams as well as with water in the cold streamand ethylene glycol in the hot stream. The stainless steel plateswere modeled using the thin thermally resistive layer boundarycondition. To provide for faster solution and allow more studies toidentify expected trends during a typical lab setting, simplified 4channel models with different internal plate geometries were builtin addition to the realistic 10 plate model with chevron shapedinternals shown in Figures 1 and 2.

Co-current Countercurrent

DirectionPlate Area

Flow rate

ΔP Tc,out Th,out Q U Efficiency

cm2 L/min Pa K K W W/m2K %

CoStraight

3821 56 288.5 304.6 580 618 25.2

CounterStraight

3821 56 290.0 303.1 689 783 30.0

CounterStraight

3822 154 287.5 305.8 1018 1036 22.1

CounterStraight

3823 289 286.1 307.0 1244 1209 18.0

CounterCorrugated

4231 414 297.3 295.9 1195 1788 52.0

CounterCorrugated

4233 2780 297.0 296.3 3525 5160 51.1

Figure 2. Complex flow pattern inside 10 plate exchanger with chevron shaped internals.

Results: Figure 3 shows velocity and temperature result plotsfor the simplified four channel models with water at 1 L/min.Internal corrugations increase the heat transfer by increasingthe surface area and the local velocity. Table 1 shows how flowpattern, internal corrugations, and flow rate affect theperformance of the heat exchanger. The heat transfercoefficient increases with increasing velocity and increasedcorrugations, but at the expense of increased pressure drop.Figure 4 shows the temperature profile for adjacent channelsnear the center of the four channel exchanger withoutcorrugations for both co-current and countercurrent flow. Thisclearly illustrates why countercurrent operation is preferred dueto higher average driving force.

Table 1. Results for 4 channel exchanger with water in both streams, with and without internal corrugations. Tc,in = 280.15, Th,in = 313.15 K.

Future Work: Our plans for this work include using the modelsto develop and test correlations for heat transfer coefficientsand exchanger efficiency with different internal geometries andan extension to turbulent flow. It is also of interest to matchthe model and experimental results more accurately toillustrate the value of simulation in predicting quantitativeresults in complex systems.

Excerpt from the Proceedings of the 2014 COMSOL Conference in Boston