Xiaojun Wang’s research while affiliated with University at Buffalo, State University of New York and other places

Unidirectional continuous carbon fibre reinforced epoxy was found to be able to sense its own strain in the fibre direction, due to its longitudinal electrical resistance decreasing reversibly and its transverse resistance increasing reversibly upon longitudinal tension. The strain sensitivity (gauge factor) is from -35.7 to -37.6 and from +34.2 to +48.7 for the longitudinal and transverse resistances respectively. Both effects originate from resistivity changes associated with the increase in the degree of fibre alignment upon longitudinal tension. Either effect allows strain sensing. Slight irreversibility is associated with the resistance decreasing after the first strain cycle and stems from the decrease in the degree of neatness of the fibre arrangement.

Real-time monitoring of fatigue damage and dynamic strain in a continuous unidirectional carbon fiber polymer-matrix composite by longitudinal electrical resistance measurement was achieved. The resistance R decreased reversibly upon tensile loading in every cycle, thus providing dynamic strain monitoring. The peak R in a cycle irreversibly increased as fatigue damage occurred, due to fiber breakage. Fiber breakage started to occur at 50% of the fatigue life, but significant growth of the fraction of fibers broken did not start till 55% of the fatigue life. From 55% to 89% of the fatigue life, more than 1000 fibers broke at a time, but not in every cycle. From 89% to 99.9% of the fatigue life, fiber breakage occurred continuously in every cycle. Beyond 99.9% of the fatigue life, fiber breakage occurred rapidly, both continuously and in spurts, with the last spurt occurring at 99.99% of the fatigue life. Catastrophic failure occurred when 18% of the fibers were broken.

Self-monitoring of static/fatigue damage and dynamic strain in a continuous crossply [0/90] carbon fiber polymer-matrix composite by electrical resistance (R) measurement was achieved. With a static/cyclic tensile stress along the 0° direction, R in this direction and R perpendicular to the fiber layers were measured. Upon static tension to failure, R in the 0° direction first decreased (due to increase of degree of 0° fiber alignment and fiber residual compressive stress reduction) and then increased (due to 0° fiber breakage), while R perpendicular to the fiber layers increased monotonically (due to increase of degree of 0° fiber alignment and delamination). Upon cyclic tension, R (0° decreased reversibly, while R perpendicular to the fiber layers increased reversibly, though R in both directions changed irreversibly by a small amount after the first cycle. Upon fatigue testing at a maximum stress of 57% of the fracture stress, R (0°) irreversibly increased both in spurts and continuously, due to 0° fiber breakage, which started at 15% of the fatigue life, while R (perpendicular to the fiber layers) irreversibly increased both in spurts and continuously, due to delamination, which started at 33% of the fatigue life. The peak R (0°) in a cycle irreversibly decreased, while the minimum R (perpendicular to the fiber layers) at the end of a cycle irreversibly increased during the first 0.1% of the fatigue life, due to irreversible increase in the degree of 0° fiber alignment. R (0°) became noisy starting at 87% of the fatigue life, whereas R (perpendicular to the fiber layers) became noisy starting at 50% of the fatigue life. For a [90] unidirectional composite, R (0°) increased reversibly upon tension and decreased reversibly upon compression in the 0° direction, due to piezoresistivity.

Epoxy containing 10 vol.% short carbon fibers and applied as a coating (0.2 mm thick) directly on cement mortar was found to be an effective piezoresistive strain sensor for tensile strain up to 0.042% and compressive strain down to −0.11%. Exceeding these strain limits caused damage in the coating. The strain sensitivity (reversible fractional increase in electrical resistance per unit strain) was 94–97 and 20–24 when the strain was contraction and expansion, respectively. The irreversible fractional increase in resistance per unit strain was 0.5–2.3 and 15–69 when the strain was contraction and expansion, respectively; it increased with increasing magnitude of strain. The resistance change was almost totally reversible when the strain was contraction, but was only partly reversible when the strain was expansion. The sensor in coating form was similar to that in bulk form in the electromechanical behavior.

Electromechanical testing involving simultaneous electrical and mechanical measurements under load was used to study the fiber-matrix interface, the fiber residual compressive stress, and the degree of marcelling (fiber waviness) in carbon fiber composites. The interface study involved single fiber pull-out testing while the fiber-matrix contact electrical resistivity was measured. The residual stress study involved measuring the electrical resistance of a single fiber embedded in the matrix while the fiber was subjected to tension through its exposed ends. The marcelling study involved measuring the electrical resistance of a composite in the through-thickness direction while tension within the elastic regime was applied in the fiber direction.

Carbon fibers are widely used as a reinforcement in structural composites. The use of these fibers that are already in the composite for strain sensing eliminates the need for strain gages, optical fibers or other sensors, thus decreasing cost and improving durability. For this purpose, the electromechanical behavior (change in DC electrical resistivity upon strain) of the fibers was investigated. A single bare carbon fiber increases its electrical resistance (not resistivity) reversibly upon tension due to changes in dimensions, such that the gage factor (fractional change in resistance per unit strain) in +2. A single carbon fiber embedded in epoxy (a common matrix) has its volume electrical resistivity increased by 10% after curing at 180 degree(s)C and subsequent cooling of the epoxy, due to the compressive residual stress resulting from the thermal contraction mismatch between fiber and epoxy. Subsequent tension of the embedded fiber at its two exposed ends decreases the residual stress and causes the fiber resistivity to decrease back to its value before embedding, such that the resistivity decrease is reversible, with gage factor -17. Excessive tension causes the resistivity of the fiber to increase, due to damage. Thus, carbon fiber in epoxy is a sensitive piezoresistive strain sensor.

Delamination in a crossply [0/90] continuous carbon fiber polymer-matrix composite was sensed in real time during fatigue by measuring the electrical resistance of the composite in the through-thickness direction. Upon 0° tension-tension fatigue at a maximum stress of 57% of the fracture stress, the resistance irreversibly increased both in spurts and continuously, because of delamination, which started at 33% of the fatigue life. The resistance increased upon loading and decreased upon subsequent unloading in every cycle, thereby allowing strain sensing. The minimum resistance at the end of a cycle irreversibly increased during the first 0.1% of the fatigue life. The resistance became noisy starting at 62% of the fatigue life, at which delamination occurred rapidly and the fraction of laminate area delaminated reached 4.3%.

Electromechanical testing involving simultaneous electrical and mechanical measurements under load was used to study the fiber-matrix interface, fiber residual stress and marcelling (fiber waviness) in carbon fiber composites. The interface study involved single fiber pull-out while the fiber-matrix contact resistivity was measured. The residual stress study involved measuring the resistance of a single fiber embedded in the matrix while the fiber was tensioned at its exposed ends. The marcelling study involved measuring the resistance of a composite in the through-thickness direction while tension was applied in the fiber direction.

By measuring the electrical resistance of a continuous unidirectional carbon fiber epoxy-matrix composite along the fiber direction during loading in this direction, fiber breakage was progressively monitored in real time. Fiber breakage occurred in spurts involving 1000 fibers or more. It started at about half of the failure strain during static tensile loading and at about half of the fatigue life during tensiontension fatigue testing. Immediately before static failure, 35% of the fibers were broken. Immediately before fatigue failure, 18% of the fibers were broken. The fiber breakage was accompanied by decrease in modulus.