Steven E. Guffey Curriculum Vitae (resume)

2007-present Professor and Industrial Hygiene Program Director, Department of Industrial Management and Systems Engineering, West Virginia University, Morgantown, WV.

2000-present Associate Professor and Industrial Hygiene Program Director, Department of Industrial Management and Systems Engineering, West Virginia University, Morgantown, WV.

1994-present Associate Professor, Department of Environmental Health, Industrial Hygiene and Safety, University of Washington, Seattle, WA.

1987-1994 Assistant Professor, Department of Environmental Health, Industrial Hygiene and Safety, University of Washington, Seattle, WA.

1979-1983 Head, New Directions Program at the University of North Carolina Occupational Safety and Health Educational Resource Center, Chapel Hill, NC. Provided multi-level training for industrial supervisors, workers, and occupational health and safety professionals. Coordinated and supervised the activities of outside consultants.

1976 - 1979 Consultant to industry for the North Carolina Department of Human Resources, Occupational Health Branch, Winston-Salem, NC. Sampled exposures and designed engineering controls for nearly 300 industrial companies.

1973-1976 Industrial Hygiene Field Investigator for the North Carolina Occupational Safety and Health Administration, Raleigh, NC. Completed OSHA industrial hygiene inspections at over 220 industrial companies.

1983 - present Provide industrial hygiene consultation and ventilation system evaluation and design for industrial operations.

1981 - present Sole instructor for 3-day course on ventilation design and 2-day course on redesign of ventilation systems, each taught annually since 1982 at the University of Wisconsin and the University of Toledo.

"Best Paper", Presented by the Engineering Committee of the American Industrial Hygiene Association (best paper for the year in the AIHAJ and the AOEH), 1994

"Best Paper", Presented by the Engineering Committee of the American Industrial Hygiene Association (best paper for the year in the AIHAJ and the AOEH), 1993

Certification: Diplomate, American Academy of Industrial Hygiene (Comprehensive Practice)

1. Guffey SE, McKinney R and Woodcock R. Lead Exposure and Its Control in a Stained Glass Studio. Stained Glass, 74(1) (Spring 1979).

2. Guffey SE. The Four Most Important Principles of Hood Design. Ind. Hyg. News Report 29:3 (1986).

1. Curran JC. Use of Power Balance to Model Converging Flows in Bilateral Junctions for Ventilation Systems (1990)

2. McLoone HE. Effects of Hood Shape and Air Velocity on Pressure Loss of Suction Hoods (1990)

3. Barnea N. Effect of Face Velocity, Flanges, and Mannikin Position on the Effectiveness of an Enclosing Hood (1991)

4. Rafnsdottir H.* Effect of Face Velocity and Cross-drafts on the Effectiveness of an Enclosing Hood (1992) *Mechanical Engineering student, Co-Chaired with Ashley F. Emery, PhD.

5. Rogers C. Comparison of Effectiveness of Benchtop Enclosing and Capturing Hoods For Various Face and Cross-Draft Velocities (1992)

6. Flanagan ME. Comparison of Tracer Gas Concentrations Sampled Simultaneously at the Lapel, Ear, and Nose (June 1993)

7. Colvin S. Experimental Validation of the Efficacy of Power Loss Coefficients in Detecting Ventilation System Alterations and in Predicting New Airflow Levels and Pressures (August 1993)

8. Spann J. Experimental Validation of Assumptions Concerning Power Loss Coefficients (August 1993)

9. Carrell T. Validation of Power Loss Modeling In Predicting the Effects of Removing One or More Branches From a Five Branch Ventilation System (December 1993)

10. Gahn, JA. A Comparison of Tracer Gas Concentrations Taken in a Wind Tunnel at the Mouth of a Mannikin to Values Found at its Lapel and Cheek (June 1994)

11. Hoppe, J. AEmpirical Determination of the Error in the ACGIH Method of Predicting Airflow Distribution in Two Industrial Ventilation Systems (June 1995)

12. Pinsky, A. Comparison of Efficacies of Current Methods for Troubleshooting Industrial Exhaust Ventilation Systems to a Proposed New Method (1996)

13. Moody, D. Comparison of the efficacies of Troubleshooting Methodologies for Ventilation Systems - A Field Study (1996)

14. Yu, D. Effectiveness of a Downdraft Ventilation Hood in Protecting Human Subjects from Exposures to a Tracer Gas (1996)

15. Wang, L. Repeatability of Velocity Pressure Traverses and Static Pressure Measurements in Five Working Ventilation Systems

16. Geiger, H. Test of the Accuracy of a Proposed Airflow Balancing Method in a Laboratory Ventilation System

17. Booth, D.* Comparison of Three Methods for Troubleshooting Ventilation Duct Systems Using Measured Pressures and Flows (1998) *Industrial Engineering PhD student, Co-Chaired with Tim Larson, PhD

18. Zhang, Xuemei. Comparison of Efficacies of the Pressure Ratio and Target Airflow Methods of Adjusting Ventilation System Airflows

Research Activities
No. Fish are not involved; I just like the drawing. --->

Drawings and stuff / Back to top of page

I have four lines of research: (1) pressure and flow modeling, (2) investigation of hood design and operating parameters and their relationship to measures of hood effectiveness and efficiency, (3) investigation of exposure sampling accuracy, and (4) use of open path Fourier transform infrared analysis for personal sampling. The first is a very long standing interest that pre-dates my first paper on the subject in 1983 (publication 2) and continues in active research now. Items 2 and 3 are also of long-standing interest, but my active research on them began in 1989 and 1992, respectively. The FTIR work began in 1995 as collaboration with a colleague, Michael Yost.

Although one would not suspect it from reading most ventilation texts, engineers and industrial hygienists spend far more time attempting to improve installed systems than in designing new ones. Specific tools for troubleshooting and redesign have been conspicuously absent. I am working toward a goal: to change ventilation redesign from trial and error art to predictable engineering. My vehicle is a comprehensive "power model" of flows and pressures across complex systems that can employ measured pressures and flows (publications 12 and 13). With the power model, one can use measured pressures and flows to (1) detect, locate and quantify changes to any portion of the system independently of conditions in other portions of the system, and (2) predict future system performance when desired changes are made to the system.

To realize this ambition, it was necessary first to model individual junctions much more fundamentally and accurately than had been done in the past. To that end, a simplified expression for pressures and flows across junctions was derived from the Navier-Stokes equations (publication 3). Analysis of the resulting equations indicated that pressures across a junction of fluid flows are best described by potential, kinetic and dissipated ("lost") power and that Bernoulli's equation is often misapplied across junctions.
Using the system of derived equations ("power model"), it is possible to determine observed losses across diverging and converging flow junctions (publication 4). For a predictive model of losses, losses across a junction were assumed to be a linear combination of incoming and outgoing kinetic powers. The efficacy of this "linear kinetic model of losses" was demonstrated from experimental data on a wide variety of single-lateral converging flow junctions (R2>0.99) and on the single bilateral junction tested (publication 11). An equivalent model should apply to diverging flow junctions as well.

Airflow distribution and energy losses are quite different issues; both are critical to both new design and to redesign. I developed a model to predict distribution of airflows among competing pathways, a crucial factor in ventilation design (publications 6 and 8). I found that existing models fit the data poorly, while the proposed model has acceptable prediction accuracy (R2>0.92 for realistic systems).

With the theoretical groundwork in place and verified, the next goal was to model ventilation systems as collections of equivalent serial volumes represented by power loss coefficients (X), which represent the ratio of the power dissipate in a volume to the kinetic power exiting that volume. The first use of X values is "quantitative troubleshooting:" If the value of X for a given portion of the system is largely independent of airflow level and changes to other portions of the system, then X values can be used to detect and locate changes to any part of a system (publication 15). Traditional troubleshooting procedures produce so many false positives that practitioners tend to investigate only the grossest changes to the system. The second use of X values is in predicting the effects of alterations to a system: If a value of X represents a portion of a system, then it is possible to model the system using value of X, only, drastically reducing the information necessary to predict the effects of changes to the system. The traditional alternative to the latter is to model the installed system using published velocity pressure coefficients as if the system had never been built ó a daunting prospect given the unavailability of coefficients for worn, damaged, leaking, and non-standard components.

Data collected by Master's student, Jeff Spann, demonstrated that X values, in fact, did vary very little with gross changes in airflow (<3% for doubling of airflow) and that each value of X was independent of changes to other portions of the system. Furthermore, he showed that ratios of static pressures within branches had similar properties, making them good troubleshooting tools as well. He also modeled a system with X values and predicted the static pressure ratios that should exist if all branch dampers were adjusted to achieve target airflow levels. When the dampers were adjusted so that those pressure ratios were achieved, branch airflows were all within 2% of target levels. As predicted in publication 13, only one adjustment was required per damper even when the fan was initially set to deliver grossly incorrect total airflows.

Data collected in the Spring of 1993 demonstrated the efficacy of the "Power Model" in modeling systems where branches are added or removed (publication 21) and where specific components are substituted one for another (publication 20). The prediction errors in both studies were generally less than 3%. The near invariance of power loss coefficients with changes to other parts of the system make them extremely useful in troubleshooting, balancing with dampers, and predicting the effects of almost any change to a ventilation system, including adding or removing whole branches.

Under the NIOSH Ventilation Troubleshooting Grant, graduate student Anne Pinsky and I measured the performances of 4 ventilation systems at Boeing and employed 3 methods to discover modifications to the systems: (1) the ACGIH Industrial Ventilation method, (2) the hood static pressure method, and (3) the power method described in my papers and in the grant proposal.

Master's student Jeanne Hoppe measured the performances of two other ventilation systems to add to the data collected for the Troubleshooting grant and to determine how well existing new design methodologies actually predict installed systems' performances. She found that the predicted values deviated from observed values by more than 25% in over half of the branches, a startling finding.

Master's student Doug Moody measured the performances of two ventilation systems used to reduce exposures to carbide metal grinding. He found that the proposed equivalent resistance method and the proposed static pressure ratio method were both highly sensitive at low false positive rates. The method described in Industrial Ventilation and the hood static pressure method commonly employed both produce poor sensitivities even at high false positive rates.

Engineering PhD student Derrick Booth and I measured pressures and flows in 5 ventilation systems and compared changes in static pressures, pressures ratios and X-values to the incidence of obstructions, which we ranked subjectively by apparent degree of blockage of flow. We then compared the actual occurrences of blockage with the predictions made by 5 different troubleshooting methods for a given threshold value. We then repeated the analysis for a broad range of thresholds for each method. Thus we could determine the false positive rate and the sensitivity of each method for each threshold value. The area under the curve when one plots the false positive rate versus the sensitivy over all of the threshold for a given method gives a measure of the efficacy of that method. Using bootstrapping methods to allow estimation of error, we found that the proposed pressure ratio and X-value methods were far superior to any of the current methods, including an idealized version of the methods present in Industrial Ventilation.

With Master's student Holly Geiger, found that the proposed damper adjustment procedure produced less than 4% error in an 8-branch system whose airflows were varied by as much as 40% using dampers. With Master's student Xuemei Zhang, found that current methods were much slower and less accurate than the proposed method.

Other items related to pressure and flow modeling
The major impediment to applying the troubleshooting and redesign procedures in the field is the tediousness and time required for airflow determination (i.e., velocity traverses). To that end, I invented a traversing device that greatly improves the speed and accuracy of Pitot tube use in the laboratory and field (publication 5). I also developed control logic and software that allow very quick Pitot traverses in the laboratory (publication 7). The software could be embedded in the microelectronics of pressure transducer devices to enable very quick Pitot traverses in the field when used with the traversing device.

Master's student Metzger and I demonstrated that Pitot traverses are accurate to within 4% (usually within 2%) even when measurements must be taken at locations within two duct diameters distance of upstream elbows and junction fittings.

Master's student Lena Wang and I found that various shortcut methods in determining the average duct velocity produced substantial errors when applied to measurements in the field. We also found that single traverses were sufficient (<3% for 90% of readings) if the pipe factor were between 0.8 and 1.0, thus providing useful guidance to practitioners in the field.

Good hood design is crucial to the protection offered by ventilation systems, yet only capturing hoods and laboratory hoods have received much research attention. I am investigating the factors that affect performance of enclosing hoods other than laboratory hoods (whose sashes put them in a different category). In the two years, my students and I completed three tracer gas studies of hood performance, where relative performance was determined from breathing zone concentrations for a mannikin standing in front of a hood. Variables tested included mannikin posture (standing erect with arms straight versus leaning into the hood with hands placed on either side of the source), hood face velocity, cross-draft velocity, flange type, and type of hood (enclosure versus capturing hood).

For conditions with no cross-draft, Nir Barnea (Master's student) and I studied the interactive effects of position, face velocity, and flange angle on exposure to a mannikin standing at the face of an enclosing hood (publication 14). The most important variables were face velocity and standing position. High velocities "forgave" otherwise poor conditions; low velocities produced high exposures for almost all conditions of other variables. Leaning into the hood was associated with much higher exposures than standing stiffly at the face of the hood. Flanges had significant but far weaker effects. The effect of flanges was confounded with distance of the source from the face of the hood since acute-angle flanges pushed the mannikin back from the face.

A mannikin inside a 6'x6'x16' wind tunnel was positioned leaning into the enclosing hood in a "working" posture. Herdis Rafnsdottir's (Master's student) research showed that even modest cross-draft velocities (15 fpm) had great effects on a mannikin's breathing zone exposures (publication 17). Surprisingly, higher cross-draft velocities (30 to 55 fpm) showed less effect than 15 fpm while still producing higher exposures than did zero cross-draft conditions. Higher face velocities (>100 fpm) counter-acted higher cross-draft velocities to a large degree ó but not totally. Based on our findings, one would hesitate to recommend costly steps intended to reduce cross-draft velocities to moderate levels. Moreover, since it is quite likely that cross-draft velocities routinely exceed 15 fpm (a level too low to measure with a thermoanemometer), it is advisable to increase face velocities to 120 fpm where the hazard potential is serious.

For a mannikin inside the 6'x6'x16' wind tunnel, Cynthia Roger's (Master's student) research compared exposure levels when the mannikin stood leaning into the enclosure to exposure levels for the mannikin "working" at an equivalent capturing hood (publication 18, in preparation). Our results were surprising: capturing hoods allowed negligible exposures even for extremely high cross-drafts (55 fpm). Enclosures allowed low ó but not negligible ó exposures. The reasons for the surprising results became clear in the exposure sampling study discussed next.

David Yu (Master's Student) tested the effectiveness of 3'x3' downdraft hood in controlling exposures to 4 human subjects engaged in a simulated painting task that generated tracer gas emissions. All exposures occurred in a 18'x8'x38' wind tunnel I designed and installed. He found that the hood was highly effective for all tests when the downdraft velocity was 100 ft/min and whenever the work was done directly on the tabletop. It was ineffective at 50 ft/min if the task was done at chest height.

Exposure sampling
"Personal" industrial hygiene samples for airborne contaminants are nearly always taken at the lapel. We questioned whether lapel samples are valid representations of inhaled concentrations. Master's student Mary Ellen Flanagan sampled simultaneously from three chest locations, at the ear, and at the nose of the mannikin while varying cross-draft velocity, orientation to the wind, and degree of mannikin motion (rotating back and forth through a 35 arc). For all tests, the mannikin held the source in its hands. We found that chest samples grossly overestimated exposures, while samples at the ear grossly underestimated exposures. Cross-draft velocity and orientation profoundly affected concentration levels and ratios of concentrations at the various locations (publication 16).

Master's student Julie Gahn extended the study design to include source location and validation of a the cheek as a surrogate for the mouth. The source was fixed at one of several locations in a grid in front of the body instead of in the mannikin's hands. The grid extended from 2 inches in front of the mannikin to 36 inches at heights at the level of the mannikin's nose, sternum, and waist. She found that the ratio of lapel to nose concentrations varied from about 0.3 to 2, depending on the location of the source and the cross-draft velocity. For most locations not very near the body, the ratios ranged from 0.7 to 1.3. The cheek samples were indistinguishable from the nose samples.

Recent research focuses on three important areas of exhaust ventilation design: (1) modeling pressure and flow relationships, (2) tracer gas studies of hood design parameters, and (3) errors from use of lapel sampling locations to represent inhaled samples. The open-path FTIR research is new, so we have no findings to describe as yet.

Developed apparatus and software for pressure and flow data acquisition
Have developed and tested a fully computerized data acquisition system at the Union Bay ventilation laboratory (publication 7). Also perfected Pitot traverse holding device previously invented (publications 5 and 7). Data supporting the use of Pitot tubes as primary measurement devices was collected (publication 11).

Developed and validated model of junction pressure/flow behavior
From empirical evidence, produced a model to predict losses across junction fittings in ventilation systems and to predict distribution of airflows among competing pathways, a crucial factor in ventilation design (publications 3, 4, 6, and 8). Modified my computer program (Heavent) to model system pressures according to existing procedures and procedures developed from my research, allowing comparison of experimental data versus competing models.

Bilateral junction loss coefficients
With graduate student Jim Curran, tested power model on a bilateral junction. Model showed very good fit to empirical data, demonstrating robustness of the model (publication 11).

Hood pressure loss coefficients
With graduate student Hugh McLoone, investigated hood entry loss versus aspect and area ratios. Results (publication 10) confirmed Brandt's coefficients, providing the only published experimental corroboration to hood loss coefficients used in every ventilation manual, and extended them.

Velocity measurement error
With graduate student Don Metzger, investigated velocity measurement error versus indicators of velocity profile skewness. Measured at 1D (duct diameter length), 2D, 4D, 10D, and 50D downstream from obstructions (e.g., one elbow, two adjacent elbows twisted in and out of plane, and junction fittings). Surprisingly, we found less than 4% error if two-diameter traverses were taken and the sampling cross-section was two or more duct diameters downstream of the obstruction.
With Master's student Lena Wang, compared duct velocity estimates using various shortcuts to values found with two perpendicular Pitot traverses. Found that pipe factor commonly produced deviations of 20% or more. Estimating velocity pressure with repeated centerline velocity pressures made no difference. Single Pitot traverses deviated by less than 5% in the 90% of the cases.

Studies of pressure/flow models
Data collected by Master's student, Jeff Spann, demonstrated that power loss coefficients and static pressure ratios varied by less than 3% even with doubling of fan airflows. In addition, power loss coefficients and static pressure ratio for a given branch were independent of alterations made elsewhere in the system.

With graduate student Ted Carrell, investigated efficacy of modeling equations in predicting the airflow distribution that results from adding or removing branches from existing systems. Prediction errors were generally less than 3%.
With graduate student Scott Colvin, investigated efficacy of modeling equations in detecting alterations to existing systems and predicting their effects on airflow distribution. Prediction errors were generally less than 2%.
With graduate student Jeanne Hoppe, investigated accuracy of the ACGIH static pressure calculation/ventilation design method using 3 heavily employed exhaust ventilation systems at Seattle Central Community College.
With Master's student Holly Geiger, found that the proposed damper adjustment procedure produced less than 4% error in an 8-branch system whose airflows were varied by as much as 40% using dampers.

Studies of troubleshooting
With graduate students Anne Pinsky and Douglas Moody, found that static pressure ratios and power loss coefficients showed selectivity and specificity superior to the currently employed troubleshooting parameters. The test systems were five operating systems at the Boeing Company, a saw re-manufacturer and a local community college's woodworking systems.
With engineering PhD student Derrick Booth, found that static pressure ratios and power loss coefficients showed superior areas under the curve for receiver operator characteristics plots for 5 ventilation systems. Current methods performed very poorly. Economic analysis also showed that the proposed methods were far more economical than current methods under reasonable assumptions of costs and benefits.

Exposures to a mannikin in the absence of cross-drafts
With graduate student Nir Barnea, completed a study of exposures to a mannikin standing at an enclosing hood. Standing position, face velocity, and flange angle were statistically significant. The most important variable was face velocity. High velocities "forgave" otherwise poor conditions; low velocities produced high exposures for almost all conditions of other variables (publication 14).

Interactive effects of cross-draft velocity and hood face velocity
With Herdis Rafnsdottir, a Mechanical Engineering Master's student, investigated the interactive effects of cross-draft velocity and hood face velocity on exposure to a mannikin. Found an apparent threshold effect for cross-draft velocities at about 20 ft/min. Surprisingly, found that effect was U-shaped: highest exposure occurred for modest cross-draft (20 ft/min), an extremely important finding. This finding was the same for all three sets of replications.

Performance of enclosing hoods compared to capturing hoods under the same conditions
With graduate student Cynthia Rogers, found that, surprisingly, benchtop capturing hoods provided better protection than did enclosing hoods for every cross-draft velocity and hood velocity, an extremely important finding.

Factors affecting performance of a downdraft hood
With graduate student David Yu, investigating effects of downdraft velocity, cross-draft velocity, and worker position on exposures to human subjects performing various tasks at a downdraft hood located in a wind tunnel.

Errors from Use of Lapel Samples to Represent Inhaled Samples
With graduate student Mary Ellen Flanagan (publication 16), sampled simultaneously from three lapel locations, at the ear, and at the nose of the mannikin while varying cross-draft velocity, orientation to the wind, and degree of mannikin motion (rotating back and forth through a 35 arc). For all tests, the mannikin held the source in its hands. Found that lapel samples grossly overestimated exposures, while samples at the ear grossly underestimated exposures. Cross-draft velocity and orientation profoundly affected concentration levels and ratios of concentrations at the various locations.

With graduate student Julie Gahn, investigated use of accuracy of various sampling locations (ear, cheek, lapel, and neck). Study extends work of Flanagan by including varying vertical and horizontal locations for the source and varying degrees of source dispersion.

With graduate student Janice Varr found erratic, moderate (typically around 20% differences) changes in the ratio of lapel concentrations to concentrations at the mouths of human subjects working at two very different tasks.

Associate Director for Student Affairs of the Industrial Hygiene Program, June 1997 present.

Chair, Curriculum Committee: Department of Environmental Health, Sept 1995 present. (Member Jan. 1993 present).

Admissions Committee: Department of Environmental Health, Fall 1987 Fall 1988, Fall 1997 present.

Fiberglass Task Force (task force put together by EH&S to investigate whether fiberglass should be removed from some or all ducts in University HVAC systems), WinterSpring 1997.

Arts Building Committee (task force appointed by the Provost to investigate complaints about chemical exposures in the University Arts Department), SpringSummer 1991.

Promotions, Tenure, and Policy Committee: Department of Environmental Health, 1988 1997.

Industrial Hygiene Research Committee: Department of Environmental Health, 1989 June 1992.

Funding History
A. Funded Research:

Recent

funded by the U.W. Industrial Hygiene Research Committee. Principal Investigator. $6000.

July 1988 - June 1989: "Power Balance Model Applied to the Redesign of Existing Ventilation Systems," funded by the U.W. Graduate School Research Fund. Principal Investigator. $6100.

October 1989 - September 1990: "Pressure Losses of Air Flowing into Ventilation Hoods," funded by the U.W. Industrial Hygiene Research Committee. Principal Investigator. $815.

October 1989 - September 1990: "The Effect of Flanges on the Collection Efficiency of Industrial Enclosing Hoods," funded by the U.W. Industrial Hygiene Research Committee. Principal Investigator. $1030.

September 1990 - January 1990: "Sharp-Edged Orifice For Flow Measurement in Ventilation Design Research," funded by Biomedical Research Support Grant. Principal Investigator. $4000.

October 1991 - September 1992: "The Effect of Cross-Draft Velocity on the Effectiveness and Efficiency of Simple Industrial Enclosing Hoods," funded by the U.W. Industrial Hygiene Research Committee. Principal Investigator. $4800.

August 1993 - August 1994: "A New Personal Air Sampler to Accurately Sample the True Breathing Zone," funded by U.S. Coast Guard. Principal Investigator. $5430.

September 1994 - August 1995: "Theory and Evaluation of a Workplace OP-FTIR Monitor." Co-Investigator, 10% effort. $110,032 (subcontract).
March 1995 - June 1997: Consortium for Risk Evaluation with Stakeholder Participation (CRESP), DOE Cooperative agreement # DE-FCO1-95EW55084. Co-Investigator, 10% effort. $2,979,113 (subcontract).

April 1993 - March 1996: "Field Validation of Power Coefficients and Static Pressure Ratios as Tools for Troubleshooting Existing Industrial Exhaust Ventilation," funded by Centers for Disease Control (CDC), NIOSH RO1 OH 03165. Principal Investigator, 50% effort. $211,229.

September 1995 - August 1999: "Theory and Evaluation of a Workplace OP-FTIR Monitor," NIOSH RO1 OH 02666-06. Co-Investigator, 15% effort. $459,543.

September 1999 - August 2001: "Evaluation of Dust Control Technologies in Construction NIOSH 1RO1 OH04039-01. Co-investigator 10% effort. $153,119.

Current

August 2001 - July 2004: "Investigation of Sampling Errors at the Lapel Using CFD with Experimental Verification Using Human Subjects." NIOSH 1R01OH07578-01. Principal investigator 50% effort. $575,000.

July 2001 - June 2002: NIOSH Training Grant for OHOS. T01/CCT310450-08. Principle investigator. 11% Effort. $58,038.

Funding History A. Funded Research:

Recent

Current

Funding History
A. Funded Research: