Corrigenda and comments on the infiltration of fiber preforms
- PDF / 103,704 Bytes
- 1 Pages / 594 x 774 pts Page_size
- 32 Downloads / 224 Views
-v D
LK1
+
K2 J
i.e., with the term K deleted from the numerator. Another typographical error (by the authors) is in Eq. [1] of Reference 2. The equation should have been written: p = O'La K
i.e., with the term cos 0 deleted. Equation [4] of Reference 2 was derived by Carman t31 in 1941 using a force balance for perfectly wetting fluids and by White t41 in 1982 using an energy balance similar to that in Reference 2. We were unfortunately unaware of the existence of these two articles when Reference 2 was written. It is emphasized that Reference 2 provides bounds, not precise values, for the pressure necessary to fully infiltrate fiber preforms. The lower bound, Eq. [4] of Reference 2, is independent of fiber distribution, whereas the upper bound depends on fiber distribution and was shown to be infinite in some cases, for example, in the practically important case of parallel contacting fibers. Equation [4] of Reference 2 provides a lower bound on the infiltration pressure, because it is the capillary pressure for reversible infiltration. In reality, the infiltration process does not take place reversibly; among other reasons, because energy is lost whenever the infiltrating liquid encounters a constriction, from which it jumps irreversibly into a pore on the other side of the constriction. These jumps bear the name "Haines jumps "t5,61 and are one of the causes for the hysteretic nature of drainage/ imbibition curves of porous media.t5,7] Another cause for hysteresis in drainage/imbibition curves is surface roughness on the pore walls which leads to a difference between the macroscopically apparent advancing and receding contact angles of the fluids on the porous me-
A. MORTENSEN, ALCOA Assistant Professor, is with the Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139. Manuscript submitted March 16, 1990.
METALLURGICAL TRANSACTIONS A
dium. Surface roughness causes other more microscopic jumps as the invading fluid passes over asperities on the pore walls. The importance of irreversible energy losses during wetting of a fiber preform varies with the porous medium geometry, and capillary pressure generally cannot be predicted in ways other than prohibitively extensive characterization of pore geometry and distribution. Often, however, it is expected that Eq. [4] of Reference 2 will only slightly underestimate the pressure needed to fully infiltrate a preform. It was shown by Carman t31 to agree with capillary pressures measured in water within the pores of water-wetted soil. Morrow t51 reports experimental evidence that irreversible energy losses are below 50 pct of the energy for reversible infiltration of packed spheres. Furthermore, calculations of capillary pressures for the infiltration of perfectly regular fiber arrays t8,9,1~ indicate that Eq. [4] of Reference 2 will only underestimate the capillary pressure drop at the infdtration front by a factor on the order of 2. In these highly idealized configurations, most of the infiltration process ta
Data Loading...