EVALUATION OF HYSTERESIS IN MERCURY INTRUSION POROSIMETRY BY SECOND-INTRUSION METHOD

The theoretical basis of mercury intrusion porosimetry is reviewed, and the limitations of the Washburn equation as a means of converting the test data to a pore-size distribution curve are discussed. A procedure is suggested to distinguish the volume of pores with uniform radii from the volume of nonuniform of "ink-bottle" pores. The depressurization curve that results from reducing the pressure on mercury to allow retraction and ejection of mercury from the pores does not coincide with the pressurization curve. This hysteresis is attributed to the presence of ink-bottle or nonuniform pores. Physicochemical factors such as chemisorption and the difference in the contact angle of advancing and retreating mercury menisci contribute slightly to this hysteresis. Consideration of pore geometries demonstrates that the ejection pressure may not be equal to the intrusion pressure for a pore of given size. A comparison of the intrusion, depressurization, and reintrusion of two hypothetical samples illustrates the advantage of the second intrusion for interpreting pore-size distribution curves. The second intrusion gives the distribution of uniform pores and volumes of ink-bottle pores intruded at each entrance diameter. Finally the second-intrusion method is applied to portland cement pastes with a 0.4 water-cement ratio that were hydrated for 3 and 60 d. It is observed that 60 to 64 percent of the porosity is in uniform pores that have a size distribution curve similar to the first-intrusion curve. /Author/