Human blood cells build a percolating physical gel all over the sample
when at rest. This gel is progressively broken when it is continuousl
y sheared in the bulk. It can slip at the wall, a phenomenon that depe
nds on the roughness of the surface. Hence, smooth and rough walls wer
e used td investigate the rheometrical shear properties of blood. A Co
uette-type rheometer with cylindrical walls allowed the shear rate to
be varied in the range 10(-3)-10 s(-1) and the hematocrit in the range
0.53-0.95. Calibration was performed with standard silicon oil. The s
tress measured at low shear rates with rough walls seemed, indeed, to
tend to a constant yield stress value. The value of the shear stress a
t the shear rate of 10(-3) s(-1) was taken as a realistic approximatio
n of the yield stress of blood. This yield stress was measured for dif
ferent values of cell concentration. Data were fitted and showed to be
proportional to the cube of the concentration over the range of hemat
ocrit studied. However, variability between the donors was observed. T
his may have relevance in the case of complete cessation of flow in vi
vo. The higher the yield stress of the static blood, the higher the dr
iving pressure needed to get the flow started again. (C) 1998 The Soci
ety of Rheology.