Summary: |
In long distance transport of waste water to regional treatment plants, or in oil and gas drilling, amongst others, synthetic fluids, based on polymer additives, flow in the turbulent regime, exhibiting drag reductions as high as 80%. Drag reduction has motivated research to understand the phenomenon and extend the use of complex fluids to other industrial applications, as the recent effort aimed at locally producing biopolymers at the surfaces of ships hulls, to reduce drag or the formulation of new thermal fluids for district heating and cooling systems.
In all these cases the wealth of experimental data has allowed a comprehensive phenomenological description of drag reduction, but common engineering tools are unable to predict even the integral flow characteristics from the fluid rheology. An alternative solution is the use of turbulence models and CFD codes but the models recently developed are still very simple because they are based on simplified rheological constitutive equations. Indeed, such turbulence models, developed by our group, rely on a generalized Newtonian constitutive equation modified to include a strain-thickening Trouton ratio effect, since the enhancement of the extensional viscosity has been associated with drag reduction. However, this rheological equation has no memory and is not capable of capturing transient rheological effects that are certainly relevant for the fluid dynamic behavior in turbulent flow. Recent research on polymer behavior has clearly shown that FENE type models (Finitely Extensible Nonlinear Elastic) are quite adequate descriptors of fluid behavior of dilute and semi-dilute polymer solutions.
Using our acquired experience on turbulence model development for viscoelastic fluids, as well as data from recent direct numerical simulation (DNS) of turbulent flow of FENE type models in parallel plate flows, in particular for FENE-P, this project proposes to develop a new family of turbulence closures for polymer solutions |
Summary
In long distance transport of waste water to regional treatment plants, or in oil and gas drilling, amongst others, synthetic fluids, based on polymer additives, flow in the turbulent regime, exhibiting drag reductions as high as 80%. Drag reduction has motivated research to understand the phenomenon and extend the use of complex fluids to other industrial applications, as the recent effort aimed at locally producing biopolymers at the surfaces of ships hulls, to reduce drag or the formulation of new thermal fluids for district heating and cooling systems.
In all these cases the wealth of experimental data has allowed a comprehensive phenomenological description of drag reduction, but common engineering tools are unable to predict even the integral flow characteristics from the fluid rheology. An alternative solution is the use of turbulence models and CFD codes but the models recently developed are still very simple because they are based on simplified rheological constitutive equations. Indeed, such turbulence models, developed by our group, rely on a generalized Newtonian constitutive equation modified to include a strain-thickening Trouton ratio effect, since the enhancement of the extensional viscosity has been associated with drag reduction. However, this rheological equation has no memory and is not capable of capturing transient rheological effects that are certainly relevant for the fluid dynamic behavior in turbulent flow. Recent research on polymer behavior has clearly shown that FENE type models (Finitely Extensible Nonlinear Elastic) are quite adequate descriptors of fluid behavior of dilute and semi-dilute polymer solutions.
Using our acquired experience on turbulence model development for viscoelastic fluids, as well as data from recent direct numerical simulation (DNS) of turbulent flow of FENE type models in parallel plate flows, in particular for FENE-P, this project proposes to develop a new family of turbulence closures for polymer solutions based on FENE-type constitutive equations. From the rheological constitutive equation for FENE fluid, with a specific rheological closure (say FENE-P), the corresponding time-averaged transport equations of momentum, turbulent kinetic energy, its rate of dissipation, Reynolds stress and polymer stress-related quantities will be derived. This will be followed by an order of magnitude analysis relying on results of experiments, of DNS computations and on previous experience to identify terms in need of modeling for closure.
The performance of the new model will be assessed against experimental and numerical (DNS) data in the literature for viscoelastic fluid flows (fully-developed pipe and channel flows, jet flows and expansion flows with separation). These tests will use either small purpose-built numerical codes or our own extensively validated 3D computational rheology code. |
Results: |
Energy savings, limitations of emissions of pollutants or cost reduction in process industry are all related and are targets of a modern industrial society seeking sustainable development. These require not only direct research on energy systems, but also developments in predictive tools, such as these turbulence models, to be used in the design of fluid dynamical and heat exchanger processes of the industry operating with complex, polymeric or surfactant based fluids.
Such complex fluids are present in processes of the chemical and mechanical engineering, and potentially in future energy distribution systems. All these involve millions of Euros every year (examples being increasing pumping costs in waste water systems, and drilling in oil and natural gas wells) and are responsible for tonnes of emissions of CO2 so that even small savings are bound to have important impacts in both energy expenditure and emission of pollutants.
Until very recently there were no tools for predicting even integral flow quantities of engineering interest, such as the friction factor and heat transfer coefficient. The recent developments of simple turbulence models by our group (Pinho, 2003; Cruz and Pinho, 2003; Cruz et al, 2004) were important steps in the quest of closures for viscoelastic fluids, but have so far only been applied to turbulent pipe flow and have shown some weaknesses
due to inadequate physics in the rheological equation of state.
Based on our experience in the field and recently available DNS and experimental data, and confirmed in the works listed in the bibliography, we think the time is now right for a major effort to be made to develop a realistic turbulence model for RANS equations and based on adequate rheological constitutive equations for drag reducing fluids, such as the FENE-P model.
This will be the first such model and will have a widespread lasting impact across the chemical and processing industries, which everyday is pushing forward the use of complex synthetic fluids for specific and more efficient applications. Our recent performance in developing the above mentioned simple turbulence models, shows our leading world position in the field which should continue to be nurtured. |