Abstract of Special Research Report (RR-93)

National Institute of Occupational Safety and Health, Japan

Effect of Thermal Aging on Degradation of Pressure Vessel Steel

RR-93-1

Yoshio KITSUNAI and Takashi HONDA

: In Japan, a number of plants which are composed of reactors, heat exchangers, pressure vessels and pipe lines had been constructed in chemistry and metallurgy industries during a decade of the 1960'. The plants have recently been operated more severely than 10 years ago to reduce the production price as much as possible. As a result, some plants have already reached their design life. During the operation, they have been exposed to elevated temperatures for long periods of time. Such exposure can alter the microstructure and bring changes in the material properties. Under such circumstances, failure or explosion of plants has happened due to degradation of the materials used for the plants. Understanding of the degree of materials degradation is of primary importance for quantitative evaluation and prediction of the residual life of plants. The aim of this study is to assess the influence of thermal aging on the degradation of plant materials.
A pressure vessel steel of a A533 Gr.B Cl.1 steel was selected for the research object and was manufactured to a 150 mm thick test plate to investigate factors and mechanisms of potential degradation. The chemical compositions of the base metal and weld wire were adjusted, referring to those of pressure vessels of Japanese older commercial plants constructed before 1970. A step cooling heat treatment was used to accelerate the degradation of the test plate. In this study, the degrees of material degradation were examined in room air by V-notch Charpy impact test, fracture toughness and fatigue crack growth behavior for the base metal and welded joint of A533B steel. The slow strain rate test (SSRT) in 288 °C pure water was also carried out in the range of stain rate between 1 × 10^-4 and 5 × 10^-2 mm/min. The cases of the degradation were examined by a scanning electron microscope and an Auger electron spectroscopy.
Based on the experimental results, it is concluded that the step-cooling heat treatment has little influence on the tensile strength, the fracture toughness, and the fatigue crack growth behavior of A533B steel. The transition temperature of the Charpy impact test for the base metal and welded joint rises approximately 30 °C. The rise of transition temperature is explained by the grain boundary segregation of phosphor or migration of carbon at around discontinuities. The strain rate dependence of the strength in SSRT test shows a minimum at a strain rate of 10^-3 mm/min, regardless of material and aging condition.

Experimental Stress Analysis of Gusset Welded Joints by Differential Infrared Thermography

RR-93-2

Etsuji YOSHIHISA

: A change of the state of elastic stress within a solid body produces small changes in temperature leading to the emission of infrared radiation (thermoelastic effect) and the emission can be monitored by the highly sensitive detector used in differential infrared thermography system. Stresses on the surface of the body are in two-dimensional state, and the method of stress measurement based on this technique gives the stress sums for the points on the surface. This method is useful in detecting highly stressed areas in the object body and in analyzing the positional change of the stress state. Some stress-concentrated areas generally exist in any actual structural component. Strength of the components, especially fatigue strength, is ruled by strength of the stress-concentrated areas. Stress concentration factor is an index for the strength of such area and stress intensity factor is that for the strength of the component containing fatigue cracks which frequently initiate in the stress-concentrated area.
A weld toe is considered as a typical stress-concentrated area in the component. In this study, stress pattern analyses by differential infrared thermography were conducted on specimens modeled after actual gusset welded joints, in order to estimate stress concentration factors and stress intensity factors. Measured values on the lines across the weld toes were curve-fitted by least squares method, and stress concentration factors were calculated from the value on the weld toes, given by the fitted curve. The results were 20% lower at the maximum, compared to the results estimated by using both strain gauge method and Heywood's equation derived from the results by photo-elastic experiments. It was considered that separating the stress sums into two principal stresses was necessary to improve the accuracy of stress concentration factor estimated by the thermoelastic method. The thermoelastic signal in the vicinity of a crack-tip is related to stress intensity factor K _I when the crack is under pure mode I loading. The results of K _I for the cracks initiated from the toes of gusset welded joints, obtained by using this relation, were compared to the value calculated by the equation developed by Maddox et al.. The results by these two methods showed good agreement for in-plane gusset joints. The equation is not necessarily applicable to K _I values for surface positions of the semi-elliptical surface cracks. Finite element analysis shall be required to evaluate the K _I value.

Omission Sensitivity Study on the Reliability Analysis of Crack Growth Fatigue Life

RR-93-3

Tetsuya SASAKI

: Fatigue crack propagation is a major factor that must be considered in the design and life prediction of industrial machines and structures. However, one finds that much of fatigue crack growth life data obtained using deterministic loadings in even tightly controlled laboratory setting exhibit a large amount of variability, and in these cases, deterministic approaches do not adequately evaluate crack growth fatigue life. Thus, the need to use probabilistic methods to predict fatigue crack growth in structures becomes evident, and in recent years, considerable attention has been given to the reliability analysis of crack growth life. In order to conduct the reliability analysis of crack growth fatigue life, the statistical properties of probabilistic factors such as initial crack size, crack growth rate, critical crack size etc. are needed. It is not realistic, however, to get so many statistical data because material tests often require so much time.
In this paper, omission sensitivity study is performed concerning probabilistic factors of material which appear in the reliability analysis of crack growth fatigue life. The omission sensitivity factors, proposed by Madsen, are formulated using the first order approximation technique. Comparing the approximated omission sensitivity factors with the accurate ones which are calculated using the computer reliability analysis code, the adequacy of this approximation is confirmed. The parametric studies on formulated omission sensitivity factors are also performed. It is shown that the factor which should be mainly treated as the probabilistic variable is initial crack size.

Simulation of Stress Spectrum Derived from the Load Lifting by a Mobile Crane

RR-93-4

Yutaka MAEDA

: Most of mobile cranes have been designed, manufactured and used with insufficient consideration of fatigue strength of structural components such as jibs and outriggers. In recent years, standards on fatigue strength of cranes are discussed internationally, and fundamental data are requested on the repeated load which may affect fatigue strength of cranes.
    The present work aims to develop an algorithm which enables the analysis and calculation of stress spectrum for any cyclic lifting of load from a given load distribution or the load spectrum.
    As a first step, a mobile crane, composed of a jib, a wire rope and a load, was modelled by a simple system of a spring and a dashpot. This model was supposed to give a similar behavior to a real crane: that is, after the load is lifted, stress values oscillate with damping. Using this model and introducing mathematical treatments, the number and the magnitude of peaks and troughs were counted for one cycle of lifting up and down the load. It was found that stress spectrum could be shown by a linear function on logarithmic scales. In other words, the spectrum for this type of oscillation could be calculated with ease.
    Secondly the method to obtain stress spectrum at a point near jib foot was discussed. As the stress of a jib was independent of slewing, the peak and the trough of the stress appeared always at the beginning or the end of load lifting, irrespective of the course of slewing. Stress spectrum in this case was obtained by calculating the stresses only at the loading and landing points.
    By use of this algorithm, static stress fluctuation data including dead weight components were recognized by calculating peak values, then dynamic stress or damped oscillation data were added to the static spectrum to obtain integrated spectrum.
    In the analysis of the spectrum for an outrigger, the stress values were found to depend on the slewing angle. Thus, peak and trough data had to be calculated for every points of the load carrying course.

Development of Safety System for Positive Clutch Type Press

RR-93-5

Shoken SHIMIZU, Soichi KUMEKAWA and Shigeo UMEZAKI

: Accidents caused by power press operation reach about 3,500 cases per year in Japan. Half the number of these accidents is caused by positive clutch type presses, because this type press has a structural inappropriateness as follows.
  (1) A slide cannot be stopped at any position other than the top dead point in case of an emergency.
  (2) A foot pedal switch is often used as a starting device of the press, which may allow hands of an operator to move into hazardous area.
  (3) A slide often overruns beyond the top dead point when the function of a friction brake decreases or the clutch pin is in failure.
    In order to solve these problems, a sudden stop mechanism which can stop the slide at setting position other than the top dead point, a safety confirmation mechanism for starting operation and a mechanism to prevent the slide overrun are required.
    The new mechanisms developed in this study are described as follows.
(A) Sudden stop mechanism
    This mechanism is composed of wedges, a ring with ditches, a light ray sensor and a mechanical stopper. The ring is located between the flywheel and the boss. When an operator interrupt a light ray sensor, wedges fall down and the ring is fixed by wedges. The clutch pin, which moves with the boss, is slided out along the ring's guide. Then, the movement of slide is locked by the mechanical stopper, and the slide stops at crank angle of 90°.
(B) Safety confirmation mechanism for starting operation
    This mechanism is composed of a connecting solenoid, connecting rods and a light ray sensor. When an operator interrupt a light ray sensor, a foot pedal connecting solenoid cannot be activated. Then, the linkage between the foot pedal connecting rod and the clutch activating cam connecting rod is disconnected. The slide cannot move even if the operator steps on the pedal.
(C) One-stroke one-stop mechanism with the preventing function for the slide overrun
    This mechanism has a function of preventing the slide overrun when the ability of the friction brake decreases or the clutch pin is in failure. This function is based on the action of wedges and the mechanical stopper. The wedge falls down after the slide arrives at the crank angle 326° and the slide can be stopped at crank angle 90° by the mechanical stopper.
    To examine the performance of the new mechanism, the test was carried out by using the positive clutch type press with a capacity of 25 tons, 100 strokes per minutes and stroke length of 75 mm. It was found that the sudden stop mechanism had a sufficient ability to stop the slide at the crank angle 90°. The safety distance on this press could be shortened from 1000 mm to 219 mm by the new system.

On the Load Acts on Scaffolds Installed with a Pipe for Placing Concrete by a Pump

RR-93-6

Yoshimasa KAWAJIRI, Katsunori OGAWA and Katsutoshi OHDO

: In concrete placing by pump, occasionally, scaffolds are used as a structure to support a concrete pipe. In such cases, the scaffolds bear the loads caused by vibration of the pump and if strength of the scaffolds is insufficient for them, parts of the scaffold may break, and in the extreme case the whole may collapse. There are reports that the ledger of frame-scaffold was broken. To prevent these accidents, it is necessary to investigate such loads and to brace the scaffolds so as to bear them.
    Experiments were carried out to clear the loads acting on the frame-scaffold and to verity the strength of scaffold frame against them.
    From the results of these experiments, it was shown that repeated loads acting on the scaffold were considerable and they brought excessive stress on the ledger installed with the concrete pipe.
    Following proposals are made as safety countermeasures:
  (1) Concrete pipes shall not be attached to scaffolds. However, in unavoidable cases.
  (2) Frames installed with a concrete pipe and their next frames in both sides shall be braced for securing the columns against 2 tonf in addition to live load per column, as well as the ledgers against 2 tonf per connection.
  (3) To distribute the load, connections between scaffolds and a concrete pipe shall be provided as many as possible and screw jacks of columns or suspending chains shall be properly adjusted.

Safety Dimensions of Tread and Rise in Considering the Psychological Load while Descending Stairs

RR-93-7

Hisao NAGATA

: According to the author's analysis of occupational injury data related to falls while walking on stairs, pedestrians are liable to have accidents while descending stairs, and high-heels footwear is comparatively prone to incur more injuries. The accidental factors to trigger human erroneous motion in descending stairs should be considered for safety dimensions. A number of formula on tread and rise have been proposed. When designing stairs, architects applied the long-established formula; in other words the tread should be one step length in walking on a flat surface minus two times the height of risers. This popular formula originating in Europe is based on the physiological cost in ascending stairs, not in descending stairs. It is questionable in regard to safety on stairs that the formula produces narrower and more dangerous treads as the rise becomes greater. But the formula gives much influence on building regulations for stair dimensions not only in Europe but also in Japan.
The previous report on "Safety dimensions of tread and rise in considering visual perceptive errors" indicated the significant formula to decide required dimensions of tread. As the psychological load becomes smaller, the occurrence of erroneous step is considered to be decreased. The dimensions of tread and rise are considered to be evaluated by pedestrians themselves. Therefore, in this report, the psychological load while descending stairs was studied for obtaining optimum dimensions of tread and rise and a practical safety index for designing stair dimensions. The psycho physical tests in this paper put emphasis on a safe walk in descent, because most fall accidents occur while descending stairs. Ten young males (average age of 21.0), ten young females (average age of 19.7) and ten elderly subjects (average age of 71.2) assessed 42 combinations of tread and rise in descending stairs. The results of those judgements were analyzed by one of scaling methods, specifically by the method of successive category. The female group wore 4 different types of footwear, differing in the height of heels.
Dimensions of tread and rise with less difficulty in descending stairs existed in definite combinations, around 29-30 cm tread and 18 cm rise for lower heels, and around 30 cm tread and 15.5 cm rise for higher heels. The psychological load increased as measurements deviated from these combinations. Based on these psycho physical results, the practical formula for obtaining safety requirements for dimensions of tread and rise is provided.

A Study on Some Causal Factors of Action Slip and Mistake

RR-93-8

Shinnosuke USUI

: Human error was classified into two categories, action slip and mistake, on the basis of "the seven-stage process of action" proposed by Norman (1988). According to his theory, slip is the error that occurs when a person forms an appropriate goal but does an action that is not intended, and mistake results from the choice of inappropriate goal.
    The aim of this study was to detect the factors of human error, particularly found in mistake that makes workers misunderstand the electric live line and action slip that causes drivers to lock their keys in their cars unintentionally (key lock-in trouble).
    The main results obtained in this study can be summarized as follows:
  1) To clarify the causal factors of falling accident by electric shock, interview and questionnaire were administered to workers who were engaged in the same job as sufferer.
  2) Analysis showed some reasons why workers mistook the electric live line and revealed that there were some factors that made workers neither detect nor correct the error, and some other factors that justified the mistake at that time.
  3) A questionnaire was administered to drivers who had experienced lock-in trouble. Then the driver's sequential behavior and the trouble-generating processes were analyzed. The most frequently marked items under the environmental and psychological situations were hurrying, doing other things before getting out of the car, and thinking about other things.
  4) In factor analysis, six factors contributing to key lock-in trouble were extracted: interruption of chain behavior, thinking about upcoming matters, thinking about other things, hurrying to get out, hurrying on business, and having to deal with a crowded situation.
  5) From the study of slip, some solutions were drawn by two different kinds of design, one for preventing slips before they occurred and one for detecting and correcting them when they did occur. From the study of mistake, some countermeasures were suggested for each of three aspects according to the seven stages of action: the stage of goal, execution and evaluation.
  6) To prevent the accidents, this paper discussed on clarifying the importance of human factors which consist of not only personal factors but also inter-personal ones, socio-organizational and life environmental factors.

Statistical Quantification Analysis on Free-terms Extracted from Descriptive Statements of the Occupational Accident Report Concerning Construction Works

RR-93-9

Yoshimi SUZUKI

: This study has been made for exploring practical use of descriptive statement of the "Occupational Accident Report". In this study, using information retrieval accident report database system developed in the Research Institute of Industrial Safety (RIIS), 3377 accidents occurred in construction works are selected for statistical quantification analysis on information structure concerning occupational accident.
    Procedure of analysis in this study are briefly as follows. Firstly, all free-terms are extracted from the descriptive statements appeared in the item "Accident Situation and Circumstances" by means of the utility program in the database system of RIIS. Secondly, the frequency distribution of these extracted free-terms are examined. Thirdly, high frequency free-terms are selected as key words for arranging/characterizing information concerning various factors of accidents occurred in construction work sites. In this step, the relationship between these key-words and accident cases is investigated by the statistical quantification analyzing method (quantification method of third type), and using these keywords the possibility for information retrieval on accident cases is also investigated by discriminant analysis. And finally, investigation by cluster analysis on extracted free-terms is made from viewpoint of the relation between these free-terms and other factors such as type of accident and kind of construction work.
    Main results of investigations are as follows;
  1) For all occupational accidents, 39,813 kinds of free-term in total are extracted from the descriptive statements of the item "Accident Situation and Circumstances".
  2) The frequency of these extracted free-terms are distributed according to the Bradford's law or Zipf's law.
  3) As a result of examination on actual frequencies of these free-terms, taking "accident type factor" into consideration, the existence of some unevenly distributed free-terms is estimated.
  4) As a result of cluster analysis of free-terms, good relationship is obtained in dendrograms between these free-terms and accident type factor or kind of construction work factor.

Historical Review on the Statistical Analysis of Occupational Accidents

RR-93-10

Shigeo HANAYASU

: Statistical analysis of occupational accidents plays an important role in establishing measures against accidents and for evaluating safety performance in work places. The first statistical/probabilistic analysis of occupational accidents was carried out by Greenwood, Wood and Yule during the World War I. Since this research work, numerous numbers of research work have been conducted, which attempted to verify the existence of the concept of "accident proneness". Also researches have been developed in the fields of systems safety analysis and risk analysis to analyze the characteristics of occupational accidents.
This paper presents a brief historical review of these research activities with particular emphasis on the statistical/probabilistic analysis of occupational accidents. Based on this review, further development needs in relation to the statistical/probabilistic research on occupational accidents in the near future are discussed.

Decomposition of Chlorofluorocarbons by Use of Gaseous Explosion

RR-93-11

Hidenori MATSUI

: It is scheduled to prohibit the production of CFCs (chlorofluorocarbons) by the end of 1995 under the international agreements, because of their depleting effects on ozone layer. Thereby, CFCs used in industry and civil life must be recovered and decomposed into harmless substances. There have been proposed many CFCs decomposition techniques, but definitive method has not been established.
    This research aimed to develop a safe and economical new technology for CFCs decomposition utilizing high temperature and pressure of gaseous explosion. Since CFCs are not flammable even in oxygen by itself, for achieving combustion of CFCs, propane was used as an additional fuel and oxygen as an oxidant. The following three methods were experimentally tested and evaluated. CFC12 (CCl₂F₂) was mainly used as a test gas in this study.
  (1) Decomposition of CFCs by a detonation induced shock wave in a vertical detonation tube (25 mmφ, 1.07m in length). In this method, stoichiometric propane-oxygen mixture and CFC12 were separately introduced into the detonation tube, then the upper detonable mixture was initiated into detonation and the shock wave from the detonation was induced into the lower CFC12 volume.
  (2) Decomposition of CFCs by a gaseous detonation in the premixed system n CFC12 + (C₃H₈ + 5O₂) in a detonation tube.
  (3) Decomposition of CFCs by a constant volume combustion in the premixed system same as (2) in a 2L spherical vessel.
These new methods shall be categorized into "thermal oxidation decomposition method".
    The following results were obtained from each experiment.
  (1) CFC12 was partially decomposed in the shock wave method. The decomposed percentage of CFC12 was less than 70% even when the CFC12 concentration ratio n (= CCl₂F₂ / C₃H₈) was smaller than 1. This means other reactants must be necessary for complete decomposition of CFC. It was suggested that the thermal decomposition only by shock heating was not satisfactory for practical application.
  (2) CFC12 was almost completely decomposed (> 99.7%) in gaseous detonation. In the range of n > 4, no detonation was obtained and in this deflagration range, less decomposition percentage (nearly equal 98%) was observed. Detonation parameters showed good agreement with the calculated values.
  (3) CFC12 was also almost completely decomposed in constant volume combustion of n CFC12 + (C₃H₈ + 5O₂) system. In this system, flame propagation was observed in the range of n < 10 and decomposed percentage was higher than 99% in the range of n < 6. Observed explosion overpressure was comparable to the calculated value.
    Decomposition techniques for CFCs by propane- and oxygen-assisted gaseous detonation or deflagration showed high CFC decomposition efficiency. These techniques can be developed into new automated continuous system of practical use.

Critical Temperatures for the Thermal Explosion of Liquid Organic Peroxides

RR-93-12

Takashi KOTOYORI

: It is possible to calculate, in accordance with the self-ignition condition presented by Semenov, the critical temperature for thermal explosion (hereafter symbolized as T _C) of an arbitrary quantity of liquid placed in an arbitrary container by measuring experimentally the rate of heat generation of the sample as well as the rate of heat transfer from the container in temperature increments of 1.25K.
    The T _C values of ten liquid organic peroxides are then calculated in accordance with the above procedure for the practical situation arising when the liquids are either in a Dewar vessel which is used in the BAM (Bundesanstalt für Materialforschung und -prüning, Berlin), heat-accumulation storage test or in a 5- or a 10-liters polyethylene container which is used by the peroxide manufacturing company for carrying 5- or 10-kg quantity of liquid organic peroxides to the users. For the Dewar vessel, the corresponding measured values were available, and they showed good agreement with the calculated ones in this study.
    The results of this study suggest that the application of Semenov's self-ignition condition to the calculation of the T _C value of a liquid is possible without stirring, provided that both the rate of heat generation and the rate of heat transfer are measured during the early stages of the self-heating process.
    The Frank-Kamenetskii model is usually assumed to be valid exclusively for solids, while the Semenov model is originally assumed valid only for fluids. However, the procedure introduced herein is applicable also for the calculation of the T _C value of a solid as measured in the BAM heat-accumulation storage test, because the Semenov model, that is, the thermal conductivity of the contents is much greater than the overall heat transfer coefficient of the container, holds well for the Dewar vessel used in the BAM test, even when a solid powder is contained in it.
    In other words, there is a possibility that one can dispense hereafter with the BAM heat-accumulation storage test, for solids as well as for liquids; in addition, for liquids, there is also such a prospect that any experimental measurement of the T _C value of an arbitrary bulk quantity of liquid placed in an arbitrary container will become unnecessary hereafter.

Anisotropic Properties for Dielectric Characteristics on Composite Insulating Materials

RR-93-13

Kenji ICHIKAWA and Tatsuo MOTOYAMA

: Epoxy resin is one of organic polymers which have been widely known in the world, but there is almost no example of using only epoxy resin itself as insulating material, and it is usually used as composite insulating material in combination with other insulating material. The electrical and mechanical properties of such composite materials are affected by the internal interfaces which are boundary layers between matrix and insulating base materials. Therefore, in order to improve the durability and elucidate the mechanism of deterioration and destruction, it is important to study the properties of interfacial layer.
    There are many kinds of organic composite insulating materials which are generally called FRP (Fiber Reinforced Plastics) having epoxy resin as matrix and glass fibers as other insulating base material. In this paper, for two kinds of FRP rods which are manufactured with pultruding unidirectionally, effect of anisotropic properties on dielectric characteristics, especially electrical breakdown voltage (BDV) and dieletric loss tangent (tanδ), were studied. The principal results obtained from the experiments are as follows :
  1) The dielectric characterises of the voidless FRP which was pultruded under pressurized condition, P2, is superior to that of the FRP including void, P1, which was pultruded under normal pressure condition.
  2) Electrical breakdown goes along the glass fibers direction, that is, interfacal layer excepting the case of the interface angle θ= 90°, therefore the BDV of FRP with θ= 0-60° increases as interface angle is larger, because breakdown path is longer. The fact that electrical breakdown goes along the interfacal layer is not affected by the voltage waveforms of ac, dc, and impulse.
  3) The BDV of FRP with θ= 90° is very higher than those of FRP with θ= 0-60°. It means that electrical breakdown by penetrating between glass fibers is difficult. Therefore, it is desirable that FRP material is positioned in the way that the interface direction is perpendicular to the electric field direction.
  4) When FRP absorbed water or was heated to high temperature, the dielectric characteristics such as tanδ obviously shows the degradation tendency, but BDV hardly decreases. The reason might be ascribed to the conditions between electrodes and FRP specimen tested, but the further research is necessary to explain it.
  5) Electrical properties of interface are very important elements to decide those of FRP material. Therefore, tanδ of the interfacal layer is calculated from a three layer model which consists of epoxy resin, glass fiber, and interfacal layer arranged in parallel. The results indicated that tanδ at the interfacal layer is larger than that of other two portions. And the fact proved that the interfacal layers are electrical weak points.

Detection of Electrostatic Field Diverged from Charged Cloud Formed by Fluidized Particles

RR-93-14

Yasuyuki TABATA and Tsutomu KODAMA

: Electrostatic activities of charged particles often encountered in powder processing or liquid atomizing must be measured to assess quantitatively electrostatic hazards in industry. Therefore, technologies for measuring electrostatic activities such as charge, electrostatic potential and electric field strength have been studied and some of them applied to instrumentation systems for static electricity. However, practical technologies for measuring electrostatic charge of particles but one per unit mass have been hardly found, because major difficulties such as an adhesion of particles on an electrostatic sensor have not been solved yet in the measurement of the charge of particles.
From the background described above, a study has been made on the measuring technology to develop an electrostatic detector applicable to the charged particles. Simple analyses and experiments were tried to investigate a few problems about the detection of the charge of particles: as a result, they indicated that the detection of electrostatic field strength gave available information for evaluating electrostatic hazards caused by charged particles since the strong electric field was produced by the charged cloud at the wall of a vessel in which the fluidized particles grew up in the charged cloud.
Consequently, a new electrostatic detector has been developed for measuring the electric field strength produced by charged cloud in the vessel such as a silo and fluidized bed. The new detector developed consists of an electric field sensor, electronic system for arranging signals picked up by sensor and air blow-off hole for avoiding the adhesion of particles and sweeping away them from the sensor. The new detector is also worked out with a design for the real time measurement of electric field strength and manufactured as a trial for performance tests. Tests were carried out on operating characteristics of the detector to examine a sensitivity, linearity and responsibility using a fluidized bed. It has become clear from various tests that the electrostatic detector developed surveys fairly steadily and accurately the electrostatic field strength produced by charged cloud in the vessel where behaviours of the charged particles are irregular. It has been also made sure through experiments that the detector is possible to be used for monitoring whether the charged cloud formed in the vessel exacerbates the electrostatic hazards or not. The specifications, structure of the detector developed and its operating characteristics obtained from tests are presented in this paper.

Study on Influence of Electromagnetic Noise Radiated by Electrostatic Discharge on Signal Transmission Lines of Electronic Equipments

RR-93-15

Hajime TOMITA and Yasuyuki TABATA

: Electronic equipments driven by microcomputer, which are commonly used for factory automation (FA) and computer integrated manufacturing (CIM), are susceptible to electromagnetic interference (EMI) due to electromagnetic noise, because their signal level is relatively low. Malfunctions of an electronic equipment induced by electromagnetic noise sometimes bring about industrial accidents and electrostatic discharge (ESD) is one of the main noise sources in industry. For example, ESD from a charged worker gives malfunctions of an electronic equipment in the vicinity. To prevent such accidents due to ESD, transient change of electric field, influence of discharging gap length on noise spectrum and electromagnetic energy of noise radiated by ESD have been studied by focusing mainly on ESD from a human body. On the other hand, typical capacitance of a conductive charged matter in industry ranges from several to a few hundred pF. Few are studied on the influence of capacitance of a charged matter on electromagnetic noise radiated by ESD.
    In this paper, the influence of capacitance of a conductive charged matter on the noise induced by ESD was studied experimentally with special reference to the effect on signal transmission lines of electronic equipments. A capacitive spark circuit was used as a model of ESD generation and a loop antenna for magnetic field detection was used as a model of a signal transmission line. The principal results obtained by this study are as follows;
  (1) The rate of change of discharge current was maximum for the capacitance of less than 3pF and the gap length of less or equal 0.5 mm.
  (2) The peak-to-peak voltage of noise induced in the loop antenna located 10 cm away from an ESD source increased in proportion to the capacitance up to 1000 pF for the gap length of 0.5 and 2 mm.
  (3) The peak-to-peak voltage induced in the loop antenna located 2 m away from an ESD source was maximum for the gap length of 0.3 mm and the capacitance of about 3 pF. This result approximately coincided with that for the effect of the rate of change of discharge current.
  (4) The dependence of the peak-to-peak voltage induced in the loop antenna located 2 m away from an ESD source on capacitance was small for the capacitance ranging from about 3 to 1000 pF.
  (5) The peak was observed on the spectrum of magnetic field measured at 2m away from an ESD source, and the frequency which gave maximum magnetic field intensity decreased with the square root of capacitance.

JNIOSH