1. Kinetics of Infiltration Processing Bryan Bals Adv: Prof Trumble

2. Infiltration  Molten material passes through the pores of a preform  Common method of creating metal-matrix composites  Wide variety of applications due to attractive properties  Spontaneous infiltration – requires no pressure to infiltrate

3. Capillarity  Contact angle – dependent upon surface forces  Liquid will rise if angle is acute, fall if obtuse  Due to balancing surface forces and gravity   P =  gh = 2  cos  / R

4. Kinetic Equations?  Washburn Equation  Assumes straight cylindrical tubes  l = [t*r*  *cos  /2  ]  Semlak and Rhines  Assumes chain of semicircular pores  Slightly more accurate  Darcy’s Law  Q = k/  (dP/l)  Used mainly in geology There is no perfect equation for kinetics due to the infinitely complex nature of the pores 1/2 Muscat, Daniel and Drew, Robin. “Modeling the Infiltration Kinetics of Molten Aluminum into Porous Titanium Carbide.” Metallurgical and Materials Transactions. Vol 25, p 2357, 1994.

5. Kinetics Concerns  Contact angle  Viscosity  Pore sizes and shapes  Surface tension  Temperature  Shape of infiltration front  Oxidation and other surface reactions

6. Purpose  Design an experimental procedure to quickly and easily test the kinetics of infiltration  Must be able to determine weight gain per time  Must be able to convert to length per time  Must be able to freeze the process in order to look at infiltration front  Must be able to change variables to look at their effect

7. Tungsten Preform  Major discrepancies in preform density between submersion data and counting the pores  Saturating preform with water revealed preforms to be 70-80% dense  Counting pores gave 31 and 33% for two samples, 51% for the third

8. Possible Reasons  Preform widely varied  Clear differences in preforms used during rate experiments  Damage done to preform during cutting and polishing  Samples weren’t perfectly infiltrated with epoxy  Sample that had higher density was polished further

9. Pictures

10. Liquid: Glycerin  Propane triol (C 3 H 8 O 3 )  Very viscous – 950 cP at 25 C  Miscible with water  Very slow to infiltrate  End result – Able to change viscosity by diluting glycerin with water (0.89 cP), providing a good test for the experimental setup

11. Viscosity Measurements

12. Experiment Setup

13. Results

14. Linearized

15. Problems  Not solely dependant on viscosity  Surface force of water is 74 dyne/cm for water, 63.4 dyne/cm for glycerin*  Different contact angles, although they were not measured  Three different preforms used, all three had significantly different densities  Reproducible? *CRC Handbook of Chemistry and Physics

16. Difference in preform density

17. Repeated water infiltration

18. Reproducibility  Appears problem lies in preform, not procedure  Major differences in preform  Tests done right after each other were closer together than tests done days apart  Preform losing weight as time goes on – changing the nature of the pores?  Differences are not significant for this experiment

19. Analysis  Used Semlak-Rhines Equation  Assume surface force is linear with concentration  0.063 N/m for glycerin, 0.074 N/m for water  Assume contact angle is 30 degrees  R = ½ * f/N = 5.01  m  f = pore density  N = number of interfaces between pore and preform per unit length

20. Results  Semlak-Rhines equation predicts rate to be 2-4 times faster than actual results  Aiming for less than a factor of 10, so this is still ok  Preform density most likely culprit  No correlations in terms of accuracy vs viscosity

21. Recommendations  Need a final preform density value  Different preform – more uniform  More exact cleaning methods – acid?  Better readings at initial contact – two people  Measure surface forces and contact angle

22. Acknowledgements  Dr Trumble  National Science Foundation  Dave Roberts  Lindsay Martin  Patti Metcalf

23. Questions?