Unlubricated frictional behavior
Firstly, the most important conclusions of friction tests conducted
on rubber sliding on dry, apparently smooth harder surfaces
are reviewed. According to the aim of this study, most of
them refer to nitrile rubber.
From the friction test results analyzed in [1] it can be concluded
that when the rubber block slides on dry and smooth glass surface
with a velocity lower than 1 mm/s (negligible frictional heat
generation) the kinetic friction coefficient is dominated by adhesion
and usually load dependent i.e. decreases with increasing
applied normal pressure (nominal contact pressure). At the same
time, rubber frictional resistance to sliding on dry, smooth track
Fig. 1. Test configuration used in [6]. can be reduced by using rougher rubber specimens (smaller real
T.J. Goda / Tribology International 93 (2016) 142–150 143
contact area at low loads). Friction reducing effect of rubber
roughening, however, completely disappears at high loads
(entirely flattened rubber asperities).
In one part of the dynamic friction tests of Grosch (see [2]), flat
NBR specimens (containing various amount of carbon black) of
25.4 25.4 6.3 mm3 (length width thickness) were pressed
against dry, polished stainless steel track moving with constant
velocity under controlled ambient temperature (varied between
50 and 100 °C). The velocity ranged from 105 to 30 mm/s while
the highest normal pressure was 0.054 MPa. Grosch came to the
conclusion that the dry friction of NBR specimens on a smooth
surface is due to interfacial adhesion. He pointed out that under
constant load and temperature the coefficient of friction first
increases then decreases with increasing sliding velocity. Additionally
he was able to identify experimentally the sliding velocity
at which coefficient of friction reached its maximum value
(adhesion hump). The highest coefficient of friction at T¼20 °C
was about 2.4 which appeared at a sliding velocity of E10 mm/s
in case of NBR containing no carbon black. Increasing carbon black
content resulted in reducing coefficient of friction (at 50 phr carbon
black the peak value was half as high as at zero carbon black
content) and a slight shift of adhesion peak towards lower velocities
(from 1 to 0.1 mm/s) in case of smooth glass countersurface.
This tendency was explained by the fact that the addition of carbon
black to NBR shifts slightly the peak value of the loss modulus
towards lower frequencies and flattens the loss curves (loss
modulus and loss factor). At the same time, on increasing carbon
black content the hysteretic friction appeared to decrease more
rapidly than the adhesion component. When replacing polished
steel surface with a rough silicon carbide paper (particle
sizeE0.1 mm, average spacing between abrasive particles
E0.14 mm) a hysteresis peak (hysteresis hump) appeared beside
the adhesion peak (adhesion hump) in the coefficient of friction vs.
sliding velocity (master) curve. However the dusting of rough
surface with magnesium oxide powder eliminated the adhesion
peak and reduced the hysteresis peak. The adhesion-reducing
powder resulted in adhesion component being almost independent
of velocity and temperature. The frequency at which the
maxima of loss modulus (E”) of NBR occurred was 2U106 Hz at
temperature of 30 °C. It is worth to mention that both the magnitude
and the location of adhesion peak in coefficient of friction
vs. sliding velocity curve were about the same for NBR/polished
steel and NBR/wavy glass sliding pair. The waviness on the glass
was needed to enhance the reproducibility of measurement
results. The coefficient of friction of NBR (with 0, 20 and 50 phr
carbon black content) measured on dry wavy glass surface at
T¼20 °C and sliding velocity of 10 mm/s was 2.3, 1.9 and 1,
respectively. Increasing the temperature from 20 to 80 °C
decreased the adhesion coefficient of friction from 2.3 to 1.5 for
unfilled NBR sliding on wavy glass surface at a sliding velocity of
10 mm/s. When NBR specimens rubbed against polished steel
counter surface no hysteresis peak appeared in Grosch’s coefficient
of friction vs. sliding velocity master curve. It is due to the fact that
the wavelength of the roughness component which would be able
to excite the rubber surface with frequency of 2 10^6 Hz (frequency
belonging to the maxima of loss modulus) is unrealistically
low i.e. is in the nanometer range. At the same time the longer
wavelength roughness components are unable to generate considerable
hysteresis due to the lower loss modulus.
Friction behavior of NBR (with 76.1 Shore A hardness and
unknown carbon black content) paired with steel surfaces of different
surface roughness was studied at T¼22 °C and under unidirectional
dry sliding conditions by Mofidi and Prakash [13].
Experiments were conducted using a block-on-ring configuration
where a rubber block of 16 4 2 mm3 (length width thickness)
was pressed against the external cylindrical boundary of a
rotating stee
Unlubricated frictional behaviorFirstly, the most important conclusions of friction tests conductedon rubber sliding on dry, apparently smooth harder surfacesare reviewed. According to the aim of this study, most ofthem refer to nitrile rubber.From the friction test results analyzed in [1] it can be concludedthat when the rubber block slides on dry and smooth glass surfacewith a velocity lower than 1 mm/s (negligible frictional heatgeneration) the kinetic friction coefficient is dominated by adhesionand usually load dependent i.e. decreases with increasingapplied normal pressure (nominal contact pressure). At the sametime, rubber frictional resistance to sliding on dry, smooth trackFig. 1. Test configuration used in [6]. can be reduced by using rougher rubber specimens (smaller realT.J. Goda / Tribology International 93 (2016) 142–150 143contact area at low loads). Friction reducing effect of rubberroughening, however, completely disappears at high loads(entirely flattened rubber asperities).In one part of the dynamic friction tests of Grosch (see [2]), flatNBR specimens (containing various amount of carbon black) of25.4 25.4 6.3 mm3 (length width thickness) were pressedagainst dry, polished stainless steel track moving with constantvelocity under controlled ambient temperature (varied between50 and 100 °C). The velocity ranged from 105 to 30 mm/s whilethe highest normal pressure was 0.054 MPa. Grosch came to theconclusion that the dry friction of NBR specimens on a smooth
surface is due to interfacial adhesion. He pointed out that under
constant load and temperature the coefficient of friction first
increases then decreases with increasing sliding velocity. Additionally
he was able to identify experimentally the sliding velocity
at which coefficient of friction reached its maximum value
(adhesion hump). The highest coefficient of friction at T¼20 °C
was about 2.4 which appeared at a sliding velocity of E10 mm/s
in case of NBR containing no carbon black. Increasing carbon black
content resulted in reducing coefficient of friction (at 50 phr carbon
black the peak value was half as high as at zero carbon black
content) and a slight shift of adhesion peak towards lower velocities
(from 1 to 0.1 mm/s) in case of smooth glass countersurface.
This tendency was explained by the fact that the addition of carbon
black to NBR shifts slightly the peak value of the loss modulus
towards lower frequencies and flattens the loss curves (loss
modulus and loss factor). At the same time, on increasing carbon
black content the hysteretic friction appeared to decrease more
rapidly than the adhesion component. When replacing polished
steel surface with a rough silicon carbide paper (particle
sizeE0.1 mm, average spacing between abrasive particles
E0.14 mm) a hysteresis peak (hysteresis hump) appeared beside
the adhesion peak (adhesion hump) in the coefficient of friction vs.
sliding velocity (master) curve. However the dusting of rough
surface with magnesium oxide powder eliminated the adhesion
peak and reduced the hysteresis peak. The adhesion-reducing
powder resulted in adhesion component being almost independent
of velocity and temperature. The frequency at which the
maxima of loss modulus (E”) of NBR occurred was 2U106 Hz at
temperature of 30 °C. It is worth to mention that both the magnitude
and the location of adhesion peak in coefficient of friction
vs. sliding velocity curve were about the same for NBR/polished
steel and NBR/wavy glass sliding pair. The waviness on the glass
was needed to enhance the reproducibility of measurement
results. The coefficient of friction of NBR (with 0, 20 and 50 phr
carbon black content) measured on dry wavy glass surface at
T¼20 °C and sliding velocity of 10 mm/s was 2.3, 1.9 and 1,
respectively. Increasing the temperature from 20 to 80 °C
decreased the adhesion coefficient of friction from 2.3 to 1.5 for
unfilled NBR sliding on wavy glass surface at a sliding velocity of
10 mm/s. When NBR specimens rubbed against polished steel
counter surface no hysteresis peak appeared in Grosch’s coefficient
of friction vs. sliding velocity master curve. It is due to the fact that
the wavelength of the roughness component which would be able
to excite the rubber surface with frequency of 2 10^6 Hz (frequency
belonging to the maxima of loss modulus) is unrealistically
low i.e. is in the nanometer range. At the same time the longer
wavelength roughness components are unable to generate considerable
hysteresis due to the lower loss modulus.
Friction behavior of NBR (with 76.1 Shore A hardness and
unknown carbon black content) paired with steel surfaces of different
surface roughness was studied at T¼22 °C and under unidirectional
dry sliding conditions by Mofidi and Prakash [13].
Experiments were conducted using a block-on-ring configuration
where a rubber block of 16 4 2 mm3 (length width thickness)
was pressed against the external cylindrical boundary of a
rotating stee
การแปล กรุณารอสักครู่..
