TN 295 : kH No. 9110 1 *.' .' mv ■ 1 1 ■ 1 1 ^i ■ > ■ ■ ' I ^ A* 9 >•< * A>«* V'^'V a, av» • <, *'. V. ,1V * ^ v-O^ ° .0' ^ •.,,•* ^V -o, :-. ^. a* %«v* /J Ac ' f ° v/*., •^6* ^ / ^ v^ .a^% ^ v^ % 1 .' A^, o " " » ) Q Y ° ° V v> \^ ^ .A ,f ^^ * N < l * c ° " ° » "^ . . s • ,0 '■^ --J P*.*l^,*« VV .1* w • <> *^T«' ,0^ V '^ -.'.».■ A o V f^ -: ^. -?'..«.&*■ V -0.. :1° .0* #J ++0* *P^. P v >'^L% °o ^9= c ° " ° v <*> ^ ^ •£ ^ A." "^ -•-■-• V " A.0 4.°"^ V<^ v v^ *t ^.^ ^i^ v^:as\v/4vi\ ^.i v^ ^^ ■v^ v .* f V »b& .- c -> ^^ ^» • g^ * 'd- ■# ^' ... ^<6 ' r *' ^ G ^ '""' A <& ''JsL' * G° y ^^-\/ %.^ # .o' X:Wff\j> V r*> '.'OP. 4 J? &,. »yjlW* aV -^ .^iS* «5> ^ °«WW. a^ ^ . v** v . "%^ /^^fc %^ '' A i ' 4 < C f A 'K. »"&■ *">, V<^B5?y .^^ > # <^ v X&fcX * >Xi'« ^ s^fc * *° »•••'•. * 0- >.•••% c ♦ifeX ./^/t A^X ^,«:,V /,-^X V,, X/ , c o«o„ <& 4^ c°"°» <*. .5^v. ^/ /j§|^ x o / ,;<^* "b^ V^ *' ^ .. °* *^' ^ X *^\/ v^- V X -^ v • . . « * .6 o, *o . » * A ' ^feX /• *^feX » 4* ♦. 3l u T, " v .* .^ ICJ 9110 Bureau of Mines Information Circular/1986 Development of an Automated Breathing and Metabolic Simulator By Nicholas Kyriazi UNITED STATES DEPARTMENT OF THE INTERIOR (AXwi^ ■fefokv* , £iA/u^c^ 'khswJ Information Circular/ 9H0 Development of an Automated Breathing and Metabolic Simulator By Nicholas Kyriazi UNITED STATES DEPARTMENT OF THE INTERIOR Donald Paul Hodel, Secretary BUREAU OF MINES Robert C. Horton, Director Library of Congress Cataloging in Publication Data: Kyriazi, Nicholas Development of an automated breathing and metabolic simulator. (Information circular ; 9110) Bibliography. Supt. of Docs, no.: I 28.27: 9110. 1. Respiration- Simulation methods. 2. Metabolism -Simulation methods. 3. Respirators -Testing. I. Title. II. Series: Information circular (United States. Bureau of Mines) ; 9110. TN295.U4 [QP177] 622 s [681'. 761] 86-600287 CONTENTS Page Abstract. 1 Introduction 2 Past developments 3 IBM ABMS 3 Reimers MBMS 5 Bellows flexibility 8 N 2 fidelity 8 System response time 8 The hypoxia scenario 8 Reimers ABMS 9 DEEC Inc. ABMS 11 Conclusions 15 Appendix. — Average inhaled gas values 16 ILLUSTRATIONS 1 . IBM ABMS photo 3 2. IBM breathing and metabolic systems schematic 4 3. IBM breathing-simulation system schematic 4 4. IBM temperature and humidity system schematic 5 5 . IBM electric furnace 5 6. Reimers MBMS photo 6 7 . Reimers MBMS schematic 7 8. Reimers ABMS photo 9 9. Reimers ABMS breathing-simulation system schematic 10 10. Reimers ABMS temperature and humidity system schematic 11 11. Reimers ABMS metabolic-simulation system schematic 12 12. DEEC Inc. ABMS photo 13 13. DEEC Inc. ABMS schematic 14 14. Comparison of CO2 breath waveforms 15 A-l. C0 2 and 2 breath waveforms 16 A-2. Single inhalation air column 16 A-3. Minimum C0 2 value vs average inhaled C0 2 value 17 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT h hour mln minute L liter ms millisecond L/min liter per minute yr year lb pound DEVELOPMENT OF AN AUTOMATED BREATHING AND METABOLIC SIMULATOR By Nicholas Kyriazi 1 ABSTRACT The Bureau of Mines has been developing breathing and metabolic simu- lator technology since 1970. Breathing simulation has been widely achieved throughout the world and used in the testing of open-circuit breathing apparatus, but satisfactory metabolism simulation has not been achieved. This situation required that the testing of closed-circuit breathing apparatus, which are the only type used in mines, be done us- ing human test subjects. The goal was a machine that could accurately simulate both the breathing and the metabolic functions of a human being for testing of closed-circuit breathing apparatus. The advantages of using such a machine instead of a human being for testing respiratory protective devices lie in its ability to quantify metabolic input, its repeatability, and the lack of a need to deal with the vagaries of human subjects. This report will describe the breathing and metabolic simu- lators that have been developed and used by the Bureau over the past 15 yr. "Biomedical engineer, Pittsburgh Research Center, Bureau of Mines, Pittsburgh, PA. INTRODUCTION Closed-circuit rescue breathing appara- tus have been used for over 75 yr in mine rescue and recovery missions after fires or explosions have made the mine atmos- phere irrespirable. Since 1981, closed- circuit escape breathing apparatus have been legislatively mandated for every person going into an underground coal mine in the United States. At present, testing of newly developed and exist- ing apparatus both by manufacturers and by the certifying agency, the National Institute for Occupational Safety and Health (NIOSH), is largely dependent upon human subject testing. Since metabolic demand varies with weight and condition of the human subject, and since even the same subject performing the same physical activity can have different metabolic demand levels depending upon posture and type of food recently eaten, it seems logical to question the fairness of quan- titative evaluations of breathing appara- tus based upon human-subject testing. This is the reasoning behind the research and development efforts by the Bureau of Mines in its quest for a simple, reli- able, and repeatable breathing and met- abolic simulator for the quantitative testing of breathing apparatus. The first concept utilized for metabo- lism simulation was that of burning pro- pane; this process simulated not only 2 consumption but also CO2 production. While achieving a measure of success, this concept proved to be very compli- cated and maintenance intensive. Also, there is always a degree of danger pre- sent when burning a combustible gas. The second concept was that of cyclical removal and replacement of inhaled gases. On a breath-by-breath basis, gas is re- moved from the system to simulate 2 consumption. Since simple removal of in- haled gas is not selective of 2 , some N 2 is also removed that must be re- placed on the reverse cycle. C0 2 is also added during the reverse cycle. A manual simulator of this type was used for 2 yr at the Bureau's Pittsburgh Research Cen- ter. The major problem with this concept lay in its extreme mechanical complexity. The automated version of this simulator also suffered from the same problem. In addition, the large internal volume of this second-generation simulator had the effect of slowing the system re- sponse to rapid changes in inhaled gas concentration. The presently used (third-generation) simulator utilizes the removal-replace- ment concept to simulate metabolism but does so on a continual basis rather than a cyclical one. In addition, all of the breathing and metabolic functions are controlled by a computer. The mechanical portion of the machine is simple; the complexity of the system lies in its com- puter software, which is not subject to maintenance problems. This concept was developed through a contract to the Noll Laboratory for Human Performance Research at the Pennsylvania State University and built by DEEC Inc. , a company formed by employees of the university who were involved with the contract. The third-generation automated breath- ing and metabolic simulator (ABMS) has the capability to vary over a wide range the following metabolic parameters: ven- tilation rate, 2 consumption rate, CO2 production rate, respiratory frequency, tidal volume, breathing waveform, and breathing gas temperature. Also, any number of work rates may be combined in any order to simulate various activities. The simulator monitors numerous param- eters including average inhaled levels of 2 and C0 2 , breathing resistance, and inhaled gas temperatures. The test results are stored on either a floppy or a hard disk and may be plotted in any manner desired. The use of the Bureau-developed ABMS enables quantitative testing of closed- circuit breathing apparatus that are used in the mining industry for both escape and rescue. As mentioned, such apparatus are tested and certified by NIOSH using human subjects of various weights. This makes apparatus design difficult as manu- facturers do not know the weight of the subject their apparatus will be tested on. Present in human-subject testing are many unknown variables such as flow rate, exact C0 2 production, and exact 2 con- sumption, all of which vary with the weight of the test subject. Therefore, it is necessary to overdesign the apparatus to accommodate even the worst- case (heaviest test subject) situation. The Bureau will soon formally propose to NIOSH that a simulator such as the third- generation ABMS be used in certification testing of breathing apparatus in order to correct the perceived deficiencies in the present methods. In the following sections, more de- tailed descriptions are provided of the succession of Bureau -developed breathing and metabolic simulators. PAST DEVELOPMENTS IBM ABMS This simulator was developed by the Bureau through a research and development contract with IBM completed in 1973. It was extensively modified by the Bureau over the following years. (See figure 1.) While it served its purpose during FIGURE 1.-IBM ABMS photo. To apparatus being tested Combustion chamber Water trap 4 t * v D— J ulse damper Flow sensors s — Q L> ®— X — Metering valves — - Y — ® Dump Pump speed control - Solenoid valves - From CO2 supply From propane supply ^Safety circuit FIGURE 2.— IBM breathing and metabolic systems schematic. those years, it was also very complicated and required continual maintenance. The metabolism simulation by this ma- chine was effected by the burning of propane that consumed 2 and produced C0 2 . (See figure 2.) The breathing sim- ulation was controlled by a variable- speed motor that was connected to a cylindrical, metal piston through linkage of a crankshaft, lever, and fulcrum. (See figure 3. ) The original lung was a flexible bellows that was replaced by the nonflexible, sliding-seal piston. The tidal volume could be changed by mov- ing the fulcrum. The functional residual capacity (FRC) was also adjustable. (See figures 4 and 5. ) One of the assets of this machine was its anatomically appropriate arrangement and functioning of components such as the simulated trachea with bidirectional flow that provided dead space. Also humanlike was the continuous simulated metabolism process. In addition, this simulator was FIGURE 3.— IBM breathing-simulation system schematic. Sample To and from apparatus being tested Humidifier- Gas analyzers / Return it \ =Jchec! valve Reservoir Expire rV- Metabolic system iv V ' ' • • • • • ' ' • • • • • • • • • > > > > > > I > > I I -r V J r ' * * 'i[ \\ Heaters O c^ Heater Sponges Water FIGURE 4.— IBM temperature and humidity system schematic. Inspire check valve Breathing piston Controller thermocouple port Asbestos end plate \ End-plate bolt^^m \ \ Moldable fiberglass insulation Ceramically potted - electric heater Catalyst beads Asbestos . end plate Combustion tube cap Thermocouple Electric heater terminal Sleeve and cap weld Electric heater terminal Combustion tube cap FIGURE 5.— IBM electric furnace. more of a true closed-loop system (like a human being) than simulators that simply remove and replace gases to effect metab- olism, but this was accomplished not without penalty. Control of the furnace ignition of the propane was not easy. Also, the system was very complicated with many modifications such that only the person who used and modified it understood exactly how it worked. When that person left the Bureau, the IBM simulator was retired. Only peak values of inhaled gases could be measured with this system as opposed to average inhaled values. For more information regarding the functioning of this simulator, refer to Bureau RI 8496. 2 REIMERS MBMS This machine was bought from Reimers Consultants of Falls Church, VA, in the spring of 1981 as a supply contract (S0308126). It was used in Bureau re- search and testing for 2 yr, then it was transferred to the NIOSH facility in Mor- gantown, WV, for its research projects. (See figure 6.) The Reimers manual breathing and meta- bolic simulator (MBMS) effected metabo- lism simulation through cyclical removal of breathing circuit gas and replacement with C0 2 and N 2 . The N2 replacement was made necessary by the unavoidable removal ^Sparks, A. w. , R. L. Stein, and J. W. Stengel. A Breathing Metabolic Simulator for Testing Respiratory Protective Equip- ment. BuMines RI 8496, 1980, 18 pp. <® - » o I f» f> WlHllllli ill wm FIGURE 6.— Reimers MBMS photo. of some N 2 in the 2 removal process. A schematic of the Reimers MBMS is shown in figure 7. The flow loop is unidirec- tional and contains approximately 7 L of breathing circuit gas. Air is inhaled from the inhalation port and then mixed in the inhalation mixing box from which inhalation gas is sampled. The major portion of the gas is then inhaled into the main bellows. Upon exhalation, the gas is forced out of the main bellows into the humidifying chamber where mois- ture is added to the gas. From this chamber, the gas goes into an after- heater and an exhalation mixing box from which exhalation gas is sampled. The motor-driven main bellows is controlled by a series of slide potentiometers that enable the breath waveform to be shaped. Part of the inhaled air is drawn into a smaller bellows, called the removal bel- lows, to simulate O2 consumption. The quantity removed depends upon the concen- tration of O2 in the inhaled gas. A quantity of gas equal to that removed is replaced by the supply bellows consisting of both CO2 and the makeup N 2 . An addi- tional bellows, called the balance bel- lows, is utilized to ensure that the operations of the removal and supply bel- lows do not add to the desired tidal volume. The balance bellows thus serves as a volume compensator. The metabolism simulation, then, is controlled by the supply and removal bellows. The quantity of gas exchanged through the operations of the supply and removal bellows is dependent upon the Vent to atmosphere Inhale port -=. RH indicator Supply bellows . COg (constant flow) plus N 2 (demand valve) 7777777/ Movable To drive fulcrum unit FIGURE 7.— Reimers MBMS schematic. 3_A- Exchange u ratio control concentration of 2 in the breathing circuit. If an O2 removal rate of 2 L/min is desired and the O2 concentration is 100%, one simply removes 2 L/ min from the circuit. If, however, the 2 concen- tration is only 50%, 4 L/min of circuit gas must be removed in order to remove the 2 L of 2 . In a much simpler process, the C0 2 flows into the supply bellows at a con- stant rate and is independent of the gas exchange processes. The N 2 flows into the supply bellows elicited from a demand valve as needed to complete the exchange. Thus, at an 2 concentration of 50% (as above), and a C0 2 flow rate of 2 L/min, 2 L/min of N 2 must be supplied from the demand valve to equal the 4 L/min of gas being withdrawn via the removal bellows. The quantity of gas exchanged in the metabolism process is controlled manually by a knob referred to as the exchange- ratio controller. The ratio used is the removal bellows volume divided by the main bellows volume. Thus, the higher the 2 concentration, the lower the ex- change ratio. Since this is a manually controlled operation, a human monitor must carefully observe the 2 concentra- tion (measured by the analyzer) of the inhaled gases in the inhalation mixing box and then adjust the exchange ratio accordingly. The advantages of this simulator over the IBM machine were that it did not use combustion of flammable gases to simulate metabolism, and that it enabled us to measure average inhaled gas concentra- tions. This was accomplished mechani- cally through its design by using a unidirectional flow loop that drew all of the inhaled gas from a breathing apparatus into the inhalation mixing box where the gas concentrations were mea- sured. With the IBM simulator, and in human subject testing labs, even though continual monitoring of gases may be per- formed, only minimal and maximal concen- trations are utilized. The contribution of apparatus dead space to inhaled gas concentrations is not calculated if only peaks of gas concentrations are noted; measuring minimal values of C0 2 , for ex- ample, tells only how well the C0 2 scrubber is working and not how much C0 2 is actually being inhaled. Average in- haled values of gases tell us what con- centration of gases a person would actu- ally inhale. One of the major drawbacks of this system was its mechanical complexity. Also, since control of the exchange ratio was manual, the user was forced to always be present during a test because 2 con- centration was continually changing. The manual exchange ratio also made it prac- tically impossible to simulate more than one metabolic state. Further problems with, the MBMS are next described in detail. Bellows Flexibility Because of the flexible nature of the bellows, even though it was reinforced with wire, the volume fidelity of the system was not always good. If a breathing apparatus being tested had significantly higher exhalation resist- ance than inhalation resistance, for ex- ample, the simulator might not be able to force the appropriate volume of gas back into the apparatus while continuing to extract the appropriate volume. This would have the effect, in the case of closed-circuit breathing apparatus, of drawing the breathing bag flat and de- manding more 2 than desired. N 2 Fidelity It is an assumption by design of the MBMS that N 2 in the correct quantity will be drawn into the circuit to balance the 2 removal and CC^ addition processes. This, however, was not necessarily the case. Again because of the flexibility of the bellows, if the inhalation resist- ance of the apparatus were high and the exhalation resistance low, for example, more N 2 would be forced into the breath- ing circuit. Also, upon every exhala- tion, pressure would increase in the sup- ply bellows causing the CC^ flow to slow; this would reduce the quantity of CO^ being added to the system. Since the supply bellows demanded that a certain quantity of gas be added to the system, more N 2 would then be added to make up the required volume. Both of these fac- tors had the effect of causing (^ concen- trations to decrease because the 0^ was being diluted by the t^ . System Response Time Because of the large internal flow-loop volume (approximately 7 L) and the uni- directional flow pattern, the system re- sponse time to a change in inhaled gas concentration was longer than that of a human being. This had the effect of com- promising its simulation. This problem would also be carried over to the auto- mated version of this design. The Hypoxia Scenario Also because of the large internal flow-loop volume, a problem surfaced with the compressed O2 apparatus when 0^ con- sumption rate was greater than the con- stant C^ flow of the apparatus. Upon first attachment of the apparatus, even if the simulator were inhaling 100% O2 , it would exhale ambient air back into the apparatus since the simulator flow loop was full of ambient air. This would have the effect of diluting the 0^ concen- tration and filling the apparatus with mostly N2 . Since the 0^ removal rate was higher than the 0^ supply rate, the 0^ concentration would fall at a constant rate until too low for life support. The apparatus demand valve would not be trig- gered since the large quantity of N 2 kept the breathing bag inflated. REIMERS ABMS This machine was a result of a 3-yr development contract with the Bureau. This simulator was to automate the design of the Reimers MBMS, which was then in use. The Reimers ABMS was delivered to the Bureau in February 1984. (See figure 8.) This simulator was also to remedy the problems the Bureau had identified with the Reimers MBMS. The bellows were replaced by flexible rolling-seal pis- tons that were stretched to a taut condi- tion by pulling a vacuum on their nonsys- tem side. It was felt that this design change would solve the N 2 fidelity problems, but it added another system to an already complicated device. See fig- ures 9 through 11 for schematics of this ABMS. This simulator is described in more detail by Reimers. 3 A supervisory computer was used to con- trol the functions of breathing simula- tion and to automate the exchange-ratio control in order to remove the correct amount of gas to simulate 2 consumption. Temperature, humidity, and C0 2 flow were input and then controlled by computer. The advantages of this machine over its manual version were its automated exchange-ratio control that freed the •^Reimers, S. D. The Development of a New Automated Breathing Metabolic Simu- lator. J. Int. Soc. Respir. Protect. , v. 2, No. 1 , 1984, p. 170. FIGURE 8.— Reimers ABMS photo. 10 t I /i Rolling seal ^ piston assembly Piston Ldc r K i a y: 3 a Jl A Vacuum pump Rotary absolute shaft encoder Vacuum switch A, Waveform generator w Zero- backlash ball screw drive II W i |i |i|i|i|| i i iii M ii ||i|i|i| i| i [ i|i| i |ffiM Point and scale -^ assembly for position «/ L Amplifier Velocity transducer ♦ Readout Readout for position FIGURE 9.— Reimers ABMS breathing-simulation system schematic. user from being present at all times during a test, and its capability to simulate more than one metabolic state per test. The major problem with the Reimers ABMS was that it never worked for any length of time. At some point during a test, for some inexplicable reason, the piston would attempt to move beyond range limits and would trigger the limit switch, thus shutting down the system. The mechani- cal system and the computer system were designed by different persons, with the result that effective control of the mechanical system was not achieved by the computer system. The cause of the problem could not be isolated by either Reimers Consultants or by the person who created the com- puter program through contract to Rei- mers. The Bureau then decided to remove the computer-control system and run the mechanical system from the in-house computer. Also, the servo-motor was replaced with a stepper motor. The Bu- reau is currently attempting to bring the 11 Electronic indicator RO ^1 External < readouts ' 4 1 -s*- Inhale mixing box RO i u_ "T 1^7^-RHr *= n A-L Test manikin *=<«= T 1 Drain H I Gas sample lines =5*= A/D-D/A \*-^-°-\ ""T^RhI* converter ' ^ , converter and supervisory SP computer RO I ^_ Afterheater- >, JH _*Lt£# T to- from analysis module - Afterheater assembly • Exhale mixing box On -off I ^""^7" " — i 120 Vac H^ t- To all temperature and relative humidity system electronics Fan Bubble chamber— «■ -L_ SP RO 11 ^ H 2 temperature limit switch Fill-drain line ^ To AP test location " Heater Liquid level sight line Indicator-controller Main air cylinder KEY RO Readout SP Set point RH Relative humidity T Temperature AP Pressure FIGURE 10.— Reimers ABMS temperature and humidity system schematic. 12 Regulated diluent supply (user supplied) CO2 volume compensator From CO 2 add system Manual t_ —■-I FIGURE 11.— Reimers ABMS metabolic-simulation system schematic. Reimers ABMS to a working condition so that it may be evaluated. DEEC INC. ABMS This simulator was delivered to the Bureau in March 1985; it was modeled after its conceptual twin, still in use at the Noll Laboratory of Pennsylvania State University in State College, PA. The DEEC Inc. simulator is a commercially available item. The simulator at the Noll Laboratory was developed as a laboratory tool, part of a Bureau con- tract with Penn State. 4 (See figure 12.) ^Kamon, E., S. Deno, and M. Vercruys- sen. Physiological Responses of Miners to Emergency. PA State Univ. (contract J010092). Volume I — Self -Contained Breathing Apparatus Stressors. BuMines OFR 29(1) -85, 1984, 32 pp.; NTIS PB 85- 186831. Volume II — Appendices (contract J010092). BuMines OFR 29(2)-85, 1984, 181 pp.; NTIS PB 85-186849. The inventors have described this simula- tor in detail. 5 This ABMS is presently in use at the Bureau and is being continually evalu- ated. We have found that, due to its physical simplicity, there is inherently less that can go wrong with it. Its com- plexity is in the computer software. The breathing simulation is achieved through use of a piston attached to a stepper motor that is controlled by the computer. (See figure 13.) Any shape of waveform is capable of being reproduced. At pres- ent, one can choose from sine waveforms, Silverman waveforms, or waveforms devel- oped by Pennsylvania State University personnel through a Bureau contract, which, in the opinion of the creator of the ABMS, are more like those from human subjects. Water is circulated to a mixing chamber on top of the piston from a heated water 5 Volume II, page 120 of work cited in footnote 4. 13 FIGURE 12.— DEEC Inc. ABMS photo. reservoir. The heated water rains down on the piston, humidifying and heating the air. The water then drains out from the bottom of the piston and returns to the reservoir. Metabolism is simulated by continual withdrawal and replacement of gases through needle valves controlled by stepper motors, which are in turn con- trolled by the computer. System gas is withdrawn from a point in the trachea above the lung and through a needle valve by a vacuum pump in order to simulate 2 consumption. If the system gas is 100% 2 , and an O2 consumption rate of 1 L/min is desired, the stepper motor will open the needle valve to permit 1 L/min to be withdrawn. If the O2 concentration in the system is only 50%, 2 L/min of system gas must be withdrawn by the vacuum pump. The computer measures the system gas concentrations through gas analyzers and adjusts the needle valve with the stepper motor accordingly. 14 Heat exchanger Valve A <3T Valve B, | Motor -^6>— Motor -^$— Motor -A») C0 2 W ^ w Valve C. Vacuum pump J Mouth Rapid response O2 and CO2 analysis Temperature measurement (4 channels) Pressure measurement Humidity water reservoir Control data * Computer Disk (program and data storage) X Plotter Line printer Measurement data Console terminal Note: 1. All motors have direct total control from the computer. 2. Lung assembly is heated to 37° C. 3. All measurements are available as averages or instantaneous. FIGURE 13.— DEEC Inc. ABMS schematic. Since the system gas rarely reaches 100% O2 , some N 2 is removed in the pro- cess of 2 removal. This N 2 must be re- placed and is metered in through another needle valve that is also controlled by the computer. In addition, C0 2 is added to the system to simulate C0 2 production by the body through the third needle valve. The computer controls all these processes. A unique feature of this ABMS is its ability to electronically measure average inhaled gas concentrations. Whereas the two Reimers machines measured average inhaled gas concentrations by physically collecting the inhaled gas into an in- halation mixing box that could hold several breaths, this ABMS measures aver- age inhaled gas levels by integrating the area under the inhalation curve of the gas tracing, weighted by instantaneous flow rate, taking into account gas trans- port and analyzer response time. See the appendix for a more thorough explanation of this concept. Other features of the ABMS are 1. A humanlike bidirectional breathing flow path. 2. Variable rates of 2 consumption (0-7 L/min), C0 2 production (0-7 L/min), respiratory frequency (6-100 breaths/ min), and ventilation (0-130 L/min). 3. Metabolic rate changes of 4 per minute. 15 ' 4 - o o V 25 7 mln Vn= 1.0 '/. •DEEC INC. ABMS f Human min I ,/Mi Mixing box simulators (estimated) TIME.s FIGURE 14.— Comparison of C0 2 breath waveforms. 4. Continuous monitoring of average inhaled 2 and C0 2 , breathing resist- ances, and gas temperatures. 5. Computer programs for self- calibration. 6. Fast-response gas dryer and ana- lyzers for breath-by-breath, average in- haled gas concentrations. 7. Disk storage of complete tests. 8. Close match between C0 2 exhalation waveform of a human subject and that of the DEEC Inc. ABMS (fig. 14). The metabolism and breathing simulation of this simulator overall have proved accurate to within 5% of desired values. Two possible weak points in the DEEC Co. design are the indirect control of metabolic flow rates and the sensitivity of the electronic calculation of average inhaled gas concentration. The metabolic flow needle valves are calibrated for flow during a 1-h self-calibration pro- cedure that is dependent upon steady in- let gas pressures for C0 2 and 2 and repeatable vacuum pump performance. If the inlet gas pressures change or the vacuum pump changes its characteristics, the metabolic flow rates will be in error. The calculation of average inhaled gas concentrations depends upon correct measurement of gas transport and response times that are used to delay the integra- tion of the area under the figurative gas concentration curves. If the gas trans- port or response times, which are mea- sured in ms , change due to turbulence in the sample lines, the measured values will be incorrect. Turbulence could be caused by discontinuity in the sample line if, for example, two lines joined by a butt connection become slightly sepa- rated and the inner diameter of the sam- ple line suddenly expands. Carelessness would permit these two weak points to be- come problems . After further evaluation and familiari- zation with this ABMS, the Bureau intends to develop recommended revisions to 30 CFR 11, which details requirements for approval of breathing apparatus by NIOSH and the Mine Safety and Health Adminis- tration (MSHA) . Quantitative evaluation of breathing apparatus based upon ABMS tests, rather than human subject tests, will be recommended. The performance of a breathing apparatus should be evaluated with a known, controlled input. It will be recommended that human-subject testing be used only for ergonomic evaluation. New stressor levels will be based upon recent physiological research. CONCLUSIONS After more than a decade of Bureau of Mines research in the simulation of breathing and metabolism, a system has been developed that meets the Bureau's goals. It can be used as a laboratory tool in both research and testing in the evaluation of breathing apparatus, espe- cially closed-circuit types that cannot effectively be evaluated by simple breathing machines. This automated breathing and metabolic simulator can be called upon to produce any waveform or metabolic demand that can be produced by a human subject, and do it in a repeat- able manner that is precisely controlled. This tool will be used to better and more fairly evaluate breathing apparatus in research and testing. 16 APPENDIX. —AVERAGE INHALED GAS VALUES Some elaboration on the concept of average inhaled gas concentrations is warranted. When monitoring the gas con- centrations sampled at a position close to the mouth of a user of a closed- circuit breathing apparatus, one would observe cyclical changes: high C0 2 and low O2 upon exhalation, and low C0 2 and high 2 upon Inhalation. A chart record- ing of such cycling is shown in figure A-l. What is not widely recognized is that the low reading of the C0 2 , for example, is not the concentration of C0 2 that is actually being inhaled. This is merely the lowest level of C0 2 escaping the C0 2 -absorbent canister. The column of air being inhaled from the breathing apparatus, as shown in fig- ure A-2, will contain some exhaled gas that resides in the dead space of the apparatus. The exhaled gas is high in C0 2 and low in 2 ; thus, even though 1 1 1 1 1 1 1 1 1 . a. 4 - ~ ■ O O 2 - - ^ . ^. • 1 1 1 1 1 1 1 1 1 58 a. 50 1 1 1 r I J I I I I I I I L I I 2 3 4 5 6 7 8 TIME, s FIGURE A-1.— C0 2 and 2 breath waveforms. 10 K0 2 High C0 2 , canister Low0 2 Dead space W Breathing bag Single inhalation air column FIGURE A-2.— Single Inhalation air column. inhalation has begun, the monitored gas concentrations will not change until this slug of exhaled air passes through the subject's mouth. This must be considered in order to determine what the subject is actually inhaling. A chart recording of the CO2 breath waveform, such as in figure A-l, does not indicate precisely where inhalation has begun. The drop in C0 2 , for example, in- dicates the point in time at which the air inspired from inside the CO^ scrubber and breathing bag has reached the gas sampling point, been transported to the gas analyzer, and registered on the chart recorder. At some point before this time on the chart recording, the inhalation cycle began. If we knew exactly where on the C0 2 recording inhalation began, we would have a better idea how much CC^ from the dead space of the breathing hose was inhaled. Measurement of the instan- taneous flow rate would tell us where inhalation began, but the response times of the gas analyzers must also be known for correlation between the CO2 curve and the flow rate curve. Figure A-3 shows inhalation and exhala- tion cycles depicted in three ways: lung volume, instantaneous flow rate as mea- sured at the mouth, and instantaneous CC^ concentrations as measured at the mouth reflecting delays of gas transport time to the analyzer and analyzer response time. In this case, we know exactly when inhalation started. If we subtract the contributions of the gas transport time and the analyzer response time (300 ms) on the C^ curve, it would seem that all we need do is integrate the area under the curve from that point when inhalation begins until inhalation ends and exhala- tion begins. Exhalation can be assumed to begin 300 ms (gas transport and ana- lyzer response time) after the point in time indicated by the lung volume cycle. The further delay in CO2 rise, after the gas transport and analyzer response times have been accounted for, is due to the low C0 2 found in the tracheal dead space. Another factor must ever, before actual concentrations can be considered, how- average inhaled gas be determined, and 17 that is the fact that the breathing flow rate is changing all the while the gas is being sampled. If, for example, a high C0 2 concentration is registered over a 100 ms time period while the flow rate is high, it indicates a greater quantity of C0 2 being inhaled than the same con- centration over the same time interval while the flow rate is low. Therefore, each instantaneous CO2 reading must be weighted by multiplying it by its corresponding instantaneous flow rate; each of these products are then added together in order to determine how much gas is actually being inhaled. With a breathing and metabolic sim- ulator such as the DEEC Inc. ABMS, it is easily determined when inhalation begins. The computer knows when it gives directions to the stepper motor to start inhaling. The computer also knows what the instantaneous flow rate is because of the defined relationship between stepper- motor speed and the fixed-diameter pis- ton; this determines flow rate. During a calibration procedure performed before each test, the DEEC Inc. ABMS measures the gas transport and analyzer response times. Then, knowing what these time delays are, it multiplies the instan- taneous flow rates by the correlated instantaneous gas concentrations and adds them up over the inhalation cycle in order to determine the quantity of CO2 or O2 inhaled. This is divided by the calculated tidal volume to get a truly accurate average inhaled gas concentra- tion measurement. On the DEEC Inc. ABMS, these values are determined every two breaths. o Q. o o 10 ~8 6 4 2 1 Li 1 1 1 — H H-Gas sample transport and analyzer response time ■*- C0 2 due to apparatus dead space -HH*-C0 2 due to tracheal dead space U-C0 ? escaping CO? scrubber KEY Y7\ Av inhaled C0 2 '/>»>/?/>J»V>?l Minimal C0 2 value I [£2gzzzzzzzz2zzz! 2 4 TIME t s FIGURE A-3.— Minimum C0 2 value vs average inhaled C0 2 value. 8 U.S. GOVERNMENT PRINTING OFFICE: 1986—605-017/40.099 INT.-BU.0F MINES,PGH.,PA. 28372 « 29 U.S. Department of the Interior Bureau of Mines— Prod, and Oistr. Cochrans Mill Road P.O. Box 18070 Pittsburgh, Pa. 1S236 OFFICIAL BUSINESS PENALTY FOR PRIVATE USE, WOO ' ] Do not wish to receive this material, please remove from your mailing list* j J} Address change. Please correct as indicated* AN EQUAL OPPORTUNITY EMPLOYER °"Q \- o 4- "V o V av. -^ • A V ■ ? *v * V °o '. *bV" v^V %/ :MB%\ \\, ^°«fc> > */^k-\ aV -?U • ! * V s o».. ^ r ^o* tf ■^ V" 5= °° ' + O. -0 ^0* /,C^/°o O, »o . » ■» A o v ^r^ ^"^ ,* v ** • A V ^t. J ^ . » • A ^ ^-^i X ^^!^S y.-^akX f+makS ,//-°' 4°^ ■a? «<> » -"&\ %.^ .vase*. ^^ /Jfe\ v./ 4?^ V V"i J .-,- <*. *-.. -*> C> ,*" o v ^ .> .*«yK'. '^ n c>*' .: ^O 1 * v 6 »""» ** <,---^-->* V ,^ v . 4°^ V .* .•*'!• ^r ^^ ^ ^ G ... °^ , o i/^'* t*» A * > iA^^Ai*° ^tf a^* * vP V -^^ "^ A* .0* t - 1 , ' .'J^% ' c '/^^i'. °o .^..^w-.V c° \c^*:* °° 4 V ^ v -^ v-^f^ ^ r . °% A^ « v^^r-V ^0* «b^ ^ **i *'.0.\ "■' ,0* .-•.- % v<" .-^.°*o '-'o* *° . jpf^sy \/W?j? v^V '° \ * / °"" o °, ' /- W i- L ^r "*« % - AV 47 0" .W^- °o < o LP 0_ A^*^ K/ ^ "^ G* o *o . » ♦" A* "^ J .^ ^"^ ;* ^ ^ v -v. -.■ V^^V '° *\yW?s 4 ^^K*° ^/\. * 9 Wm§: «f\ : »^k* : «s^ -j ;jsfe-. \ /*sfe v« ^°^:X xsasX • ^s»X' >t^% ; 1 o V *bv* ,0° %> ip^v *o° to. '*77^ „*■*' ^ y % • J 1 c5- *^ & .* ■ *M I