Research

Research Areas
 

1) Modeling of Unsaturated Flow through Dual-Scale Porous Fiber Mats in Liquid Composite

2) Wicking in Porous Media

3) Flow Modeling in Green Composites

4) Evaporation in Porous Media: Multi-component Liquid Transport

5) Evaporation in Porous Media: Pore Network

6) Ensemble Averaging

7) Calibration of Permeability Evaluation Test Setups

8) Simulation of flow in pressure infiltration of dual-scale 3D woven mats with pure metal in fabrication of Metal Matrix Composites

9) Making Next- Generation Scaled CNF Nanocomposites using LCM Technology

10) Combustion Simulation

1) Modeling of Unsaturated Flow through Dual-Scale Porous Fiber Mats in Liquid Composite

Liquid Composite Molding (LCM) processes, including Resin Transfer Molding (RTM), Vacuum Assisted Resin Transfer Molding (VARTM), and Seemann Composites Resin Infusion Molding Process (SCRIMP), have become important methods for manufacturing polymer composites. One of the main activities of this research group is to investigate the physics of resin flow through a mass of reinforcing fibers in Liquid Composites Molding (LCM) processes used in the manufacturing of polymer composites. The experimental station shown on the main page is currently being used to study the physics of flow of thermosetting resin in a mold packed with reinforcing fibers in the form of woven fiber mats. Schematic of the computer aided data-acquisition and data-processing platform developed in the lab is as follows:

 

                                                                           Schematic of LCM setup

 

The data gathered using the above setup is being used to verify the theory of flow of a viscous liquid in a dual-scale porous medium under isothermal and non-isothermal conditions. Once thus tested, the theory will be used to develop an advanced mold-filling simulation which will be of immense help to the LCM mold designer in visualizing the flow of resin in a mold, and optimizing the placement of the inlet and outlet gates in the mold. Some other pictures of the setup and the experiments of the lab are available here.
To find out more about the details of this project, please visit here
 

2) Wicking in Porous Media

Wicking is the spontaneous transport of liquid into a porous medium through capillary suction. The capillary suction force arises as a result of the wetting of solid matrix by the invading liquid. We are studying the wicking into the standard polymer wicks. These kinds of wicks, made by sintering polymer beads, are often used for delivering active liquids such as fragrances and insecticides from a reservoir to ambient air or other regions of interest.
Factors affecting the wicking rate are also being studied here in this lab. Factors such as external pressure (hydraulic pressure) and structural changed (swelling or contraction) are under study in our research projects.
micrographic
                                                                           Micrographic photo of a wick

 

 

Spontaneous Imbibition of Liquid in Glass fiber Wicks

In this first part of our research, a theoretical model for wicking of a liquid into a glass-fiber wick is developed based on the sharp-front or plug-flow assumption with single-phase Darcy flow occurring behind a clearly-defined liquid front. Such an approach is superior to the traditional Washburn equation based approach derived from unidirectional capillary tubes as it allows one to model wicking in 2-D and 3-D geometries. The newness of the formulation lies in the fact that it was developed for wicking along fibers of a transversely-isotropic porous wick made from bundles of parallel glass fibers. The simplification of the permeability tensor for this system leads to a straight-forward development of a formula for the liquid-front height as a function of time. Later, the liquid-front height was shown to be directly obtainable from the mass of the liquid absorbed by the wick as a function of time.

Experimental setup for measuring the mass of liquid spontaneously imbibed by the wick using a sensitive balance

Typical optical micrographs of the cross-sections of wicks samples cast in epoxy for measuring porosity ( . The spotty dark spots represent fiber cross-sections while the lighter background is the epoxy matrix

 

The proposed theoretical model is tested against the results of a mass-gain experiment designed to study wicking of a liquid in vertical wicks. Eight different glass-fiber wicks, named from A to H, were used with a single wicking liquid (a hydrocarbon based oil) in the experiment. In order to avoid the usage of any fitting parameters, all the parameters involved in the problem were measured independently. They included the properties of the fibrous porous medium [porosity, axial permeability, fiber diameter], properties of the wetting liquid [viscosity, surface tension] and finally, the property related to the interaction between the solid and liquid phases [contact angle]. Estimation of the axial permeability, measured using the falling head permeameter, involved developing a novel formula using the flow formulation for a transversely-isotropic porous medium of the fiber-glass wicks.

The proposed theoretical model for predicting the height of liquid-front as a function of time during wicking. The model is compared with the experimental data for the wicks A and H.

The proposed theoretical model for predicting the height of liquid-front as a function of time during wicking. The model is compared with the experimental data for the wicks C and E.

The proposed theoretical model for predicting the height of liquid-front as a function of time during wicking. The model is compared with the experimental data for the wicks F and G

The proposed theoretical model for predicting the height of liquid-front as a function of time during wicking. The model is compared with the experimental data for wicks D and B.

From the comparison of the experimental results with the theoretical predictions, it was observed that the proposed theoretical model is able to predict wicking most satisfactorily only in wicks A, D and H. For the other remaining wicks (B, C, E, F, G), it is not as effective. The reason for this discrepancy could be traced to the lower contact angles seen in the former wicks that translate into higher capillary suction pressures, and which could possibly help in creating a sharper liquid-front during wicking. The general reasons for the differences between experiments and theory could also be traced to the lack of homogeneity in the wicks due to the clustering of fibers seen in the wick cross-section micrographs. The lack of homogeneity may also cause fudging of the sharp liquid-air interface, thereby necessitating the use of an alternative Richard’s equation type approach to model wicking in the glass-fiber wicks. Despite the observed weaknesses of the proposed theoretical model, it was heartening to note that, even for the cases where the model was not particular successful, it was able to predict the wicked height (or liquid-mass absorbed) for some duration during the wicking experiment and provide a satisfactory upper bound for the experimental observations. Our proposed model is powerful as it clearly and explicitly explains the dependence of liquid rise in wicks to the fluid and porous media properties.

In second part of our research, the Richard’s equation is used to predict liquid saturations inside the glass-fiber wicks made of aligned fibers. In order to find the saturation levels of the liquid in a transversely-isotropic porous medium with a principal direction of the permeability tensor aligned with the flow direction as a function of time and location, we developed and solved an expression based on Richard’s equation. In addition, in order to have a precise comparison between experimental results and numerical outcomes, we solved the Richard’s equation as a partial differential equation numerically in 2-D using COMSOL and analytically in 1-D using Mathematica.  All the related properties such as capillary pressure and relative permeability equations as a function of saturation were calculated directly for our particular fibrous porous medium using the state-of-the-art software GeoDict while solving for pore-scale flows within the wick microstructure. The obtained numerical solutions of Richard’s equation were compared with the experimental results which were obtained by measuring saturations (or liquid content) of the wicks at different axial locations.

Photograph of a wicking experiment where a glass-fiber wick was dipped in a liquid dyed blue. One can clearly see the saturation gradually dropping from 1 at the wick bottom to 0 at the wick top, thus forming a partially saturated region in the wick above the liquid level. Note the vertical streaks of blue, which indicate the presence of vertical channels for liquids inside the wick.

 

In our experiments, in order to measure the saturations at different locations, two different types of wicks made of glass fibers, were used. Comparing of the numerical and experimental results indicated that there is a satisfying correlation between the measured saturation levels and the numerical predictions. This good agreement allows us to repose confidence in the saturations predicted using Richard’s equation for unsaturated flows in porous media.

Experimental procedure for measuring saturation distribution along a wick axis

(a) The glass wick is marked wick into equal parts, (b) weighing the dry wick before immersing into the wicking liquid and achieving steady state, (c) submerging the wetted wick into liquid nitrogen to ‘lock’ the liquid, (d) cutting the frozen wick into the marked equal parts, (e) weighing each cut part of the frozen wick

 

It is clear that there is a reasonable match between the numerical solutions and empirical data.

The numerical solutions of Richard’s equation (obtained after including the gravity effects) in comparison with the experimental data for the first wick

The numerical solutions of Richard’s equation (obtained after including gravity) in comparison with the experimental data for the second wick

3) Flow Modeling in Green Composites

Green composites are those composites that both resin and reinforcing fibermats are bio-based, degradable materials. Recently, natural fibers are being increasingly used to substitute artificial glass and carbon fibers in polymer composites. However, not much is known about the flow of bio-resins through preforms made from natural fibers. The swelling of natural fibers, due to liquid absorption, adds a new dimension to the conventional flow modeling in the Liquid Composite Molding (LCM). Swelling of natural fibers causes the permeability and porosity of LCM fibermats to be variable (not constant as assumed for the artificial carbon or glass fibers) during the mold-filling process. We study the flow of water-based liquids, such as bio-resins, in natural fibermats, such as wood fibers. This research contains experimental, theoretical, and numerical flow modeling. A Finite Element Code is under development, based on Pore-Flow software, that has capability to include the swelling effect in fluid flow prediction.
 

4) Evaporation in Porous Media: Multi-component Liquid Transport

Evaporation and transport of multi-component liquid mixtures is one of the most challenging problems in porous media studies. Slow evaporation of a non-dilute multi-component liquid mixture from a cylindrical porous wick made of sintered polymer beads is studied in this project. The application of this system is in consumer products such as Air fresheners and insect repellents. A volume-averaged model based on the work by Professors Quintard and Whitaker is employed for the nonlinear species-transport equation inside the fully-saturated porous medium. The Darcy velocity inside the wick is determined from the evaporation rate at the wick-top, which in turn is a function of mole-fractions of all components at the gas-liquid interface. A flux-balance condition, based on the overall mole-balance for each component, is employed to develop a mole-fractions boundary-condition at the wick-bottom interface with a liquid container. Results of this simulation are compared with experiments for a mixture of three alkanes, i.e., Decane, Dodecane and Hexadecane. A reasonable agreement is observed for the evaporation rates at the wick top as well as the composition along the wick. To find out more about the details of this project, please visithere
 

5) Evaporation in Porous Media: Pore Network

 
a) Drying of single-component liquid in irregular networks
Evaporation of single-component liquids in porous media is modelled by using the technique of Pore Network Simulation. A novel numerical algorithm is developed a to simulate the drying profile of porous materials with completely random internal structure. In this approach, the porous medium is represented by a set of pores which are connected by narrow channels. position of these pores as well as size of the pores and throats are completely arbitrary. Micrographs of cross-sections of polymer can be used to generate the network based on the actual geometry.
 
b) Dual-scale porous media
This method has been used in another computational-experimental project in order to study the evaporation from dual scale porous materials. This was a joint project with Institut de Mecanique des Fluides de Toulouse (IMFT) and the results were published in Journal of Heat and mass transfer. To see the abstract of this work, please see here:Figure below shows two intermediate steps of a typical evolution of drying in a 100X100 dual-porosity network with small pores in the bottom exposed layer. Evaporation happens at the bottom edge of the square; the other edges are sealed. The left picture shows the state of pore saturation at the (breakthrough):
To see more pictures of the results of the numerical simulation of this project together with the schematic of experiment setup please visit here.
 image-001 image-003

6) Ensemble Averaging

Most of the studies in the area of transport in porous media are done using a Representative Elementary Volume (REV), which is a suitable volume which the analysis is being done. By considering enough number of REVs, one can find the flow parameters in the Porous medium. In this project, the Ensemble averaging theory on porous media is introduced which is a new method for calculation of the desired parameters such as permeability without any concern about the size of the REV. Using this strategy, the existence and size of the well known concept of REV can be quantified. The main goal of this study is to introduce a new concept for the study of flow and heat transfer in porous media, independent of the REV size which is varies spatially in the problems geometry. Using the concept of ensemble averaging in statistical mechanics, the mean of a quantity that is a function of the systems micro-state is called the ensemble of possible states based on the systems distributions on its micro-states. Each ensemble average of parameter which is called a realization is depending on the ensemble (random) chosen. Therefore, the mathematical expression varies from realization to realization. However, the mean of all realizations given for a physical quantity is independent on the ensemble chosen at the thermodynamic limit.Figures below show two different realizations of the same problem:
 
Realization 1
Realization 2
 

7) Calibration of Permeability Evaluation Test Setups

The permeability of a material is a parameter critical to the accurate design of molds for LCM, RTM, VARTM, and SCRIMP manufacturing processes. For flow simulation software to adequately predict flow characteristics in a material, the permeability of the material must be known very accurately. One-dimensional flow and radial flow experimental setups are commonly used to evaluate the permeability of a material, though there exists no standard dictating the characteristics of these or any other related experimental setups. Consequently, permeability evaluations of a material can vary significantly between different experimental setups. High variances in permeability evaluations decrease the effectiveness of using flow simulation software to design molds. Currently, a reference material created using stereolithography methods is being investigated as a means to calibrate lab equipment. The material is composed of unit cells, and through patterning these unit cells, different experimental setups can be accommodated. The permeability of this reference material is being determined through flow simulations using commercial software, and this value will be compared against permeability values obtained experimentally. Also, a design for a composite reference specimen that can be used in both radial and 1-dimensional flow setup is being developed. This increases the versatility of the stereolithographic specimens while avoiding the (very high) cost of constructing two individual specimens.
 

8) Simulation of flow in pressure infiltration of dual-scale 3D woven mats with pure metal in fabrication of Metal Matrix Composites

The pressure infiltration process (PIP) is an important technique for making MMCs where a liquid metal or alloy is injected and solidified in a mold packed with woven or stitched fiber mats. In this project, the computer simulation PORE-FLOW, which has been developed to model mold-filling in the resin transfer process used for making polymer matrix composites, will be adapted for optimizing the mold design in Pressure Infiltration Process. The flow physics developed to model resin flow in dual-scale fiber mats used in polymer matrix composites will be modified using the volume averaging method to serve as the basis for the simulation. The finite element/control volume algorithm will be implemented to simulate the PIP mold-filling process marked by the flow, solidification and remelting of the metal at the macroscopic inter-tow gap level and the microscopic intra-tow fiber level. The simulation will be validated by comparing its prediction for flow variables and grain-size distributions with the results from a PIP experiment.

PIP

9) Making Next- Generation Scaled CNF Nanocomposites using LCM Technology
What is CNF?

In recent years, demand for biodegradable, plant-based, composite, also called “green composites” or “biocomposites”, is increasing in various sectors especially the automotive sector of industry. Various kinds of natural fibers have been used in green composites; such reinforcements can be classified into two categories of plant and animal fibers. Plant fibers may be produced from different parts of the plants, such as seed, leave, skin, bast, and fruit. Examples of such fibers are cotton, jute, kenaf, coir, and flax. Examples of animal fibers include silk, sinew, wool, catgut, angora, mohair and alpaca. Natural fibers have lower mechanical properties as compared to carbon and glass fibers. As a result, the current breed of two biocomposites suffers from the twin drawbacks of lower strength and fatigue properties compared with the carbon or glass-fiber based polymer composites.

Cellulose nanofibers (CNFs), a new type of nanofibers made purely of cellulose molecules, have very good mechanical property compared to other natural fibers, and even carbon or glass fibers. Recent studies showed that the films or “nanopaper” of CNFs are the strongest man-made, cellulose-based materials. However, the full reinforcing potential of these materials has yet to be realized partly because of issues related to manufacturing processes. Recently, cellulose nanofibers have begun receiving serious consideration as potential reinforcement materials. The big problem with using CNF fibers was the energy requirements for breaking down cellulose fibers in nanofibers were prohibitively high until recently. Recent advances in chemical and mechanical technologies have drastically reduced the energy requirements for producing cellulose nanofibers.

cnf1cnf2

Fig 1. Figure of CNF

 

Description of LCM Technology for making polymer composites

The liquid composite molding (LCM) processes, including RTM, VARTM and SCRIMP, form an important class of environment-friendly technologies for manufacturing the polymer composites, well known for their light weight, strength and toughness. In LCMs, a resin undergoing the thermosetting reaction is injected into a hollow mold cavity packed with reinforcing carbon or glass fibers. Mold-filling simulations are important approach to optimize the placement of processing parameters to ensure its complete filling. Most of these simulations treat the resin flow as the flow of a liquid through a porous medium where it is assumed that the medium is fully saturated behind the resin flow front in an LCM mold. Recently, researchers discovered that this assumption of full saturation is incorrect for a majority of cases where the reinforcement is in the form of woven, stitched or braided mats made from fiber tows (these fiber mats are often referred to as dual-scale porous media due to the presence of two length scales of pore size in such a porous medium). A distinct region of partial saturation was discovered behind the front where the flow can be characterized as unsaturated flow as shown in the picture. It was discovered that the presence of such a region altered the flow so significantly that it could not be predicted by the typical physics used for LCM simulations.

1Dflow

Fig 2. The experimental setup

schematic

 Fig 3. The 1D mold

Current Efforts to make CNF performes using Freeze- Drying Process

For extra information about current efforts in making CNF perform please visit here.

Current efforts to make scaled CNF nanocomposite using LCM

Cellulose nanofibers are one class of natural fibers that have resulted in structures with remarkable mechanical properties. In this study, the cellulose nanofibers are used as reinforcements in the forms of layered films in a bio-derived resin. Assessment of the tension and fracture behavior is performed for each sample. Crack resistance behavior is compared to glass fiber systems and strategies for improving the fracture toughness of CNF composites are discussed. A bio-based epoxy is used to reinforce the CNF/Epoxy specimens. The epoxy used in all these experiments is a Super Sap 100/1000 made by Entropy Bio-Resins Co. The bio-based 160 epoxy is an epoxy resin that is made from 37% bio-content obtained as co-products of other green industries including wood pulp and bio-fuels production. The resin is classified as a USDA (United States Department of Agriculture) Bio Preferred SM Product using ASTM D6866. The resin has a total calculated biomass of 50%. Using the cellulose nanofiber films as reinforcements, the reinforcements were integrated into the composites.

The testing performed in this section consisted of tension and fracture specimens based on standard specimen geometries. The standards do not directly address nano-cellulose composites, so the specimens were adapted accordingly as noted. The fracture specimens were as described in ASTM D5528 with the following differences. Two sets of specimens are examined; in one case Fibreglast E-glass (Saertex) are used as reinforcements. The glass fibers are primarily in unidirectional form and had a dry areal weight of 955 g/m2. The resin pass through the reinforcement and comes out from two upper connections and goes to resin trap. The samples are ready after one day curring process in room temperature. The pictures of set-up and mold are as follow:

image-017
MOLD

Fig 4. Figures of the mold and experimental setup

For extra information about testing of CNF composites please visit here.

Latest development in making CNF

We are making composite parts that qualify as ASTM standard tensile test specimens. Our goal is to study the resin flow and ultimately the added tensile and flexural strength when using freeze dried, foam, cellulose nano fibers as reinforcement material for Super Sap Resin. These materials are expected to be an important advancement in bio based technology throughout the automotive and packing industries. Our LCM process is based on a conventional injection molding operation. Instead of using positive pressure to inject resin, we use negative vacuum pressure at one atmosphere to pull the resin through the fibers. We have designed our mold using aluminum and polycarbonate because these materials are easily machinable. The mold consists of a base slab of aluminum with removable inserts for tensile and flexurual test samples. A polycarbonate sheet is used completes the air seal and acts as a window to view resin movement. The mold was originally designed for the resin inlet to be in the center of the specimen and the outlets were at both longitudinal ends of the specimen. We’ve found that using this technique creates race tracking down the center of the part because the resin tends to flow around the outer edges of the fibers. This happens because the fiber foam mat is manufactured in such a way that the upper and lower faces of the material have very low porosity and thus low permeability from top to bottom. For better resin flow we inject the resin from one side face and allow it to travel horizontally through the fibers. By using this technique we achieved better surface finish and wetting. The CNF fibers are very light weight and only allow for us to create samples with approximately one percent volume fraction. Using a higher volume fraction stressed our vacuum pump because when the fibers are packed tightly their permeability decreases substantially. This makes it difficult to compare the added strength of the CNF fibers to that of fiberglass of carbon fiber matting because alternate materials such as these have a much higher density. Currently, we are testing our samples for tensile strength. The resin we are using creates issues with an industrial sized tensile testing device because the infusion based resin has a very low viscosity making it very brittle when cured. We have found that conventional testing devices can easily crack the samples when clamped and causes fractures at the grip locations rather than the gauge area.

CNF samples

Fig 5. Figure of the CNF and Pure resin samples

cnf test

Fig 6. Figure of the CNF sample in Tensile test Machine

To see the resin filling process watch the movie as follow video.

Recent publications in CNF nanocomposite using LCM

1- “Characterization and Processing of Nanocellulose Thermosetting Composites” Ronald C. Sabo, Rani F. El-Hajjar, Craig M. Clemons, Krishna M. Pillai, Handbook of Polymer Nanocomposites. Processing, Performance and Application – Volume C: Polymer Nanocomposites of Cellulose Nanoparticles, Springer-Verlag Berlin Heidelberg 2015; pp. 265-295. (pdf)

2- “Mechanical characterization of cellulose nanofiber and bio-derived epoxy composite” R. Masoodi, R.E. Hajjar, K.M. Pillai, and R. Sabo, Materials and Design, v 36, p 570-576, April 2012. Abstract

3- “An experimental estimation of Liquid Absorption Coefficient for Cellulose Nano-Fiber Films” A. Javadi, K.M. Pillai and R. Sabo, Proceedings of 11th International Conference on Flow Processes in Composite Materials (FPCM11), Auckland, New Zealand, July 9-12, 2012. Abstract

4- “An Experimental Study on Swelling of Cellulose Nano-Fiber Films in Epoxy Resins and Water” R. Masoodi, R. Javadi, K.M. Pillai, R. Sabo, 2011 Spring SAMPE Technical Conference and Exhibition State of the Industry: Advanced Materials, Applications, and Processing Technology, International SAMPE Technical Conference, 2011. Abstract

5- “An Experimental Study on Crack Propagation in Green Composites made from Cellulose Nanofibers and Epoxy” R. Masoodi, R.E. El-Hajjar, K.M. Pillai, A. Javadi, and R. Sabo, 2011 Spring SAMPE Technical Conference and Exhibition State of the Industry: Advanced Materials, Applications, and Processing Technology, International SAMPE Technical Conference, 2011. Abstract

 6- “An Experimental Study on Mechanical Properties of Green Composites Made Using Cellulose Nanofiber Films” R. Masoodi, Benton J. Weibel, K.M. Pillai, R.E. Hajjar, and R. Sabo, 2011 ASME International Mechanical Engineering Congress and Exposition, Denver, Colorado, November 11-17, 2011.

7- “Single-Phase Flows in Swelling, Liquid-Absorbing Porous Media:A Derivation of Governing Equations using the Volume Averaging Method” K.M. Pillai, January 8, 2013. Abstract

10) Combustion Simulation

Our research in combustion has been started with the simulation flame over a glass fiber wick of a burning lamp. The first stage of our research started with the characterization of wick and fuel properties such as; porosity, permeability, contact angle, surface tension, etc. Then, we started doing several measurements such as fuel mass consumption and flame length based on combustion experiments. In the next stage based on the experiment, the combustion over the glass fiber wick like the lamp, dove, etc. has been investigated. At this stage, fluid flow and heat transfer studies were included. All the coupled physical problem was solved using CFD analysis. The effect of turbulence on the combustion, characteristic of flow and radiation models were analyzed. Also for the next step, the flow coupled with the combustion inside the wick will be investigated.

C24188A0-A35C-4837-ACE7-4D1382013230

Initial experiments in order to measure the flame length

Flame-CO2

CO2 concentration inside the flame

Flame-H2O

H2O concentration inside the flame

Flame-O2

O2 concentration inside the flame

 

Flame-T

Temperature distribution inside the flame over the wick