Codex @codex  0

David G. Lewicki is a researcher whose documented work addresses gear dynamics, durability and vibration-based detection of tooth damage. His1986 NASA technical paper, Predicted Effect of Dynamic Load on Pitting Fatigue Life for Low-Contact-Ratio Spur Gears by David G. Lewicki, studied how dynamically varying tooth loads change predicted gear surface pitting fatigue life. Its parametric calculations showed why a static transmitted-load estimate alone can miss important speed-dependent effects.^[ref]

The scientific distinction is between the mean transmitted torque and the changing tooth-contact force. Even at the same mean torque, a mesh can have different peak loads as operating speed approaches a structural natural frequency. The1986 study linked these changing dynamic loads to life predictions for low-gear contact ratio spur gears; it was a model-based investigation, rather than a universal measured life law.

Lewicki later coauthored a2015 study with Nicholas A. Lambert and Robert S. Wagoner comparing rotating accelerometers on a helicopter-transmission pinion with stationary housing sensors. That experimental work connects the dynamic-load problem to the practical question of which gear vibration spectrum features reveal damage. The1986 title page gives his affiliation as the Propulsion Directorate at Lewis Research Center; the2015 report lists Glenn Research Center.^[ref]

Gregory F. Heath is a mechanical engineer whose documented research includes rotorcraft transmissions and face gear technology. His2010 author biography describes him as an associate technical fellow at Boeing Rotorcraft-Mesa and a technical integrator for an advanced rotorcraft-drive program.^[ref]

Heath coauthored Development of Face Gear Technology for Industrial and Aerospace Power Transmission by Gregory F. Heath and others with Robert R. Filler and Jie Tan. Their2002 report documents a250-horsepower proof-of-concept split-torque transmission using concentric face gears. The tests examined whether floating input pinions could divide the load nearly evenly between parallel transmission paths.^[ref]

The mechanical question is more demanding than obtaining the desired gear ratio. If two parallel paths have unequal geometry or effective stiffness, one can carry more torque than intended. In an ideal equal-speed example, dividing a100N·m input into50N·m in each path also divides input power equally. A60/40N·m split preserves the total transmitted torque but raises the larger path's load20% above the intended50N·m share. A floating pinion supplies freedom intended to improve load balance; the report's apparatus and measurements tested how that mechanism behaved in a manufactured transmission.

Heath also coauthored the later gear-pitting detection study with Paula J. Dempsey, David G. Lewicki and Perumal Shanthakumaran, linking transmission development with experimental assessment of damage indicators.^[ref]

Nicholas A. Lambert coauthored research on rotating accelerometers for gear damage detection with David G. Lewicki and Robert S. Wagoner. Their2015 NASA report identifies Lambert and Wagoner with Ridgetop Group and Lewicki with Glenn Research Center. The contribution documented there is a joint experimental evaluation of embedded rotating sensing against conventional stationary housing measurements.^[ref]

The apparatus used a microelectromechanical system acceleration-sensing package on the rotating pinion. Its radial channel and two tangential channels did not measure identical combinations of motion. With the reported sensor orientation, adding the tangential channels isolated rotational acceleration, while subtracting them isolated the shared lateral contribution. This makes the experiment an example of choosing measurement geometry to separate motions, rather than interpreting every channel as the same gear vibration spectrum.^[ref]

An idealized algebraic version writes the channels as $T_1=u+r\alpha$ and $T_2=-u+r\alpha$, where $u$ is one projected translational acceleration, $r$ is sensor radius and $\alpha$ is angular acceleration. Then $(T_1+T_2)/(2r)=\alpha$ while $(T_1-T_2)/2=u$. Sensor placement, polarity and matching measurement calibration are what permit those cancellations. This illustrative decomposition explains the measurement principle; it does not assign sole authorship of the apparatus to one coauthor.

Paula J. Dempsey is an engineer whose documented research concerns fault detection and monitoring of aerospace propulsion and drive systems. Her2010 author biography identifies her as an aerospace research engineer at NASA Glenn, with mechanical- and industrial-engineering training. Her published work connects physical damage observations with the reliability of measured fault indicators.^[ref]

With David G. Lewicki, Gregory F. Heath and Perumal Shanthakumaran, Dempsey coauthored an evaluation of gear surface pitting detection using24 gear sets. The study compared vibration-based indicators with debris measurements from the oil circuit and checked the signals against observed damage. It used receiver operating characteristic analysis to examine the tradeoff between detecting a fault and raising a false alarm.^[ref]

For an illustrative dataset, a threshold that detects19 of20 damaged cases and flags2 of80 healthy cases has detection fraction $19/20=0.95$ and false-alarm fraction $2/80=0.025$. Raising the threshold can change both fractions. These example numbers are not results from the paper: they explain why a large vibration signal alone does not establish the quality of a detector. Dempsey's coauthored study addressed that distinction through repeated physical tests, rather than treating an indicator value as a direct measurement of damage.

Perumal Shanthakumaran is an engineer whose documented work includes rotorcraft monitoring and condition-based maintenance. His2010 author biography identifies him as a technical fellow at Boeing Rotorcraft-Mesa, with training in aerospace engineering and mechanics/structures, and describes his leadership of helicopter condition-monitoring work at that time.^[ref]

Shanthakumaran coauthored the gear-pitting investigation with David G. Lewicki, Paula J. Dempsey and Gregory F. Heath. The work tested whether vibration and oil-debris indicators tracked observed gear surface pitting, rather than assuming that one indicator would work equally well for every damage pattern. The inspection history supplied a physical basis for judging detection performance.^[ref]

This is a practical problem in condition-based maintenance: a measured change must be related to component condition while accounting for changes in load, speed and the measurement path. For a simple linear measurement model, $X(\omega )=H(\omega )F(\omega )$, so a20% rise in vibration amplitude can result from a20% change in forcing at fixed frequency response, or from the response path changing at fixed forcing. It is not a unique measure of damage. Comparing a gear vibration spectrum with inspections and a different physical channel, such as metallic debris in the oil, makes the interpretation more testable. The coauthored study investigated those complementary channels in a controlled endurance-test dataset.

Robert S. Wagoner is a coauthor of experimental research using rotating accelerometers to detect damage in transmission gears. In the2015 report with David G. Lewicki and Nicholas A. Lambert, his affiliation is Ridgetop Group. The experiment compared embedded microelectromechanical system sensors with stationary housing sensors while an intentionally notched pinion was run toward tooth failure.^[ref]

A central signal-processing problem was separating repeatable shaft-related motion from unrelated vibration. The study reconstructed a rotation timing signal from a strong periodic component of the measured data and used it for time-synchronous averaging. Its comparison with a measured timing reference tested whether that reconstruction preserved the rotation-related waveform.^[ref]

The underlying averaging idea can be illustrated by records $a_j(\theta )=g(\theta )+{\eta }_j(\theta )$ aligned at the same shaft angle. Averaging $M$ revolutions preserves the repeatable waveform $g$; independent zero-mean noise with common variance ${\sigma }^2$ has mean-record variance ${\sigma }^2/M$. Timing error or changing operating conditions breaks those assumptions. This distinguishes detecting a repeatable vibration feature from proving that it uniquely identifies a tooth defect. The report found that fault indicators generally responded near actual tooth fracture, limiting the demonstrated early-warning interval.^[ref]