Abstract of Special Research Report (RR-86)

National Institute of Occupational Safety and Health, Japan

Fatigue Crack Growth Behavior and Fracture Toughness of Structural Steels under Low Temperatures

RR-86-1

Yoshio KITSUNAI

:As structures or machines in cold climate are subjected to variable amplitude loadings, it is an important problem to prevent an unstable fracture caused by fatigue crack because of reduced fracture toughness of the materials. Therefore reliable data on the fatigue crack growth behavior and fracture toughness at low temperatures are required for safely designing or determining the inspection period of such structures. In this study, the fatigue crack growth rates and fracture toughness of structural steels are examined at temperatures of 123, 153 and 293 K, and the influences of specimen geometry and thickness, and stress ratio and frequency on the formation of cyclic cleavage during the fatigue crack growth, and the conditions of the final failure are evaluated. Moreover, the relation between fracture toughness, K_c, and maximum stress intensity factor, K_max at which the first cleavage appears is examined.
    The materials used in this study are JIS SM50 and HT60 steels which are widely used in structures. Two types of specimens, center-cracked tension (CCT) specimens with 4 and 8 mm thickness and 100 mm width, and compact tension (CT) specimens, with 4, 8 and 14 mm thickness, respectively and 100 mm width were machined out from the each steel for the fatigue crack growth tests. The CT specimens for fracture toughness tests were prepared with 12. 7 mm thickness and 51 mm width. The tests were carried out using a 196 kN closed-loop servo hydraulic fatigue testing machine. The test temperatures were controlled by changing the flow volume of liquid nitrogen into a refrigeration chamber. The fatigue crack growth tests were performed at stress ratios, R, of 0.05, 0.5, and 0.7, respectively. The load was varied sinusoidally and the frequency was 0.8 - 20 Hz. An electron fractographic analysis was employed to determine the mechanism of fatigue crack growth.
    The main results obtained in this study are as follows:
  (1) At 123 K, the fatigue crack growth rate was considerably affected by the stress ratio, and was a marked acceleration to high ΔK region near the final failure because of the occurrence of a cleavage during striation formation.
  (2) The fatigue crack growth rates at temperatures of 153 and 293 K were roughly the same and the mechanism of crack growth was dominated by striation formation.
  (3) At 123 K, the cyclic cleavage appeared at value of K_max above 20 MPa√m during the fatigue crack growth, independently of material, specimen type, specimen thickness and frequency.
  (4) The value of K_ci at which the first cleavage appeared increased with an increasing K_in defined as K_max at which a precrack was introduced in the specimen. However, a threshold of K_ci appeared as K_in decreased below 18 MPa√m.
  (5) The cyclic cleavage was found to form due to local deterioration of the material caused by cyclic straining ahead of the fatigue crack tip.
  (6) The value of fatigue fracture toughness, K_fc, at which the final fracture occurs was found to increase as the specimen thickness did not satisfy the size requirement for small scale yieldings.
  (7) Considering the scatter of K_c and frequency effect on K_c, it will be possible to prevent the initiation of an unstable fracture originated from fatigue cracks, if fracture toughness of the material used in cold climate is estimated at about 70% of K_c determined under monotonic load.

Evaluation of the Effect of Residual Stresses on the Fatigue Crack Growth Behavior of Welded Joins in Steel

RR-86-2

Yoshio KITSUNAI and Etsuji YOSHIHISA

: Various kinds of defects, such as lack of fusion or poor penetration, are sometimes introduced into welded structures during the welding operation, though the best welding techniques are made available today. Therefore, it is said that the fatigue life of weldments is almost entirely controlled by crack from pre-existing defects. Reliable data on fatigue crack growth rates in welded joints are required for assessment of the integrity or determining the inspection period of weldments. Considerable work has been carried out on the fatigue crack growth rates in welded joints under various conditions. However, a large variation is found in the results, because the fatigue crack growth rates are affected by many factors such as welding conditions or mechanical properties of materials. In particular, distributions and magnitudes of welding residual stresses seem to be a most important factor dominating the crack growth.
    In this study, the fatigue crack growth behavior in welded joints of STS 42 steel was evaluated using compact type (CT), center-cracked tension (CCT) and single edge-cracked (SEC) specimens, and effects of specimen configuration, thickness, width, crack orientation with respect to the weld line, location of weld line in the specimen and pre-loaded history on the crack growth rates in the welded joints were analyzed by the fracture mechanics approaches. The fatigue crack growth tests were performed using a closed-loop servo-controlled hydraulic testing machine under constant load and constant stress intensity factor range, ΔK, as control parameters, respectively. Welding residual stresses were measured by the sectioning method using rosette type strain gauges mounted on the specimens. Redistributions of residual stresses were measured with monitoring the change of released strains due to extension of saw cut notches.
    The main results obtained in this study are summarized as follows :
  (1) As a fatigue crack propagated from a weld line with tensile residual stress to compressive residual stress field, the crack growth rate was roughly the same as that of base metal or slightly higher than the base metal regardless of specimen types or specimen sizes (thickness and width), because the growing fatigue crack tip is always opened even when the crack propagated into the compressive residual stress field.
  (2) As a fatigue crack propagated from the compressive residual stress field to the tensile residual stress field, the fatigue crack growth rate decreased with increasing specimen sizes, because the welding residual stresses were a little released when the specimen sizes increased.
  (3) When the growing fatigue crack is within a region of initial compressive residual stress field, the overall residual stress distribution does not change appreciably.
  (4) The fatigue crack growth rates in the parallel CT specimens of which the crack growed in the weld metal to the direction parallel to the weld line decreased as compared with those in the base metal, because the crack tips in the parallel CT specimens were always existed in the field of compressive residual stresses.
  (5) Static pre-load history was found to have a reduction effect on the welding residual stresses. The fatigue crack growth rates of the CCT specimens which were pre-loaded above 50% of yield stress of the base metal showed a tendency to approach the crack growth behavior of the base metal.
  (6) The influence of residual stresses on the fatigue crack growth rates of welded joints was found to be the same effect of stress ratio on the crack growth of the base metal.
  (7) The fatigue crack growth rates in the welded joints were correlated with effective stress intensity factor range, ΔK_effR, which was estimated by superposition of respective stress intensity factor for the residual stress field and for applied stress regardless of specimen types, specimen sizes and difference of residual stresses in the specimens.

Study on Auditory Warning Alarms of Automated Guided Vehicles

RR-86-3

Yoshiyuki EGAWA

: The purpose of this study is to establish useful auditory warning alarm criteria (hereinafter referred to as "alarm") for automated guided vehicles. First, the author suggests 4 groups of alarming wave form for classification of warning alarms. The application of this classification allows a variety of alarms in the factory for proper classification. The 4 groups are termed "Continuous Sound with Periodicity (hereinafter referred to as C.S.P.) ", "Continuous Sound with Stability(C.S.S.)", "ON-OFF Sound with Periodicity(O.S.P.)" and "ON-OFF Sound with Stability(O.S.S.)".
    Enquete and tapes in respect of alarm are sent to the manufacturing industries of automated guided vehicles and collected to analyze. Problems to be solved are found through the analysis of these data. The problems are as follows.
    The alarm of automated guided vehicle has come into use in two cases ; they are in running and at an emergency stop. In case of running, it should be the type of C.S.P., because it is necessary to ring continuously while vehicle's running. The pattern of each periodical form should be the type of "slow rising up" for giving little jarring to the ear. But as a result of the analysis on site, 35% of the alarms do not satisfy the desirable condition.
    In case of emergency stop, the alarm should be the type other than C.S.P. The rising form of wave should be fast rising for operator's attention, but 33% of the alarms in emergency stop aren't the type of "fast rising up".
    For recommendations about the periodicity of running-alarm, two experiments are carried out. One is the experiment for evaluating the periodical duration by counting cycles and the other is the experiment of sensory detecting abilities for digital & analog sound pressures. In the result of these experiments, the periodicity of C.S.P. type alarm should be designed as the duration of about 900 msec.
    In conclusion, the recommendation of auditory warning alarm for automated guided vehicles is given as follows :
    "Continuous Sound with Periodicity" should be adopted as warning alarm while in running, and kinds of waving form other than C.S.P. type are preferable when the vehicle stops on emergency. The periodicity of running-alarm should be about 900 msec.

>Safety Assessment of Automated Production Systems Using Microelectoronics (4th Report) --Evaluation of Hazard Controllability of an Industrial Robot (Part 1)--

RR-86-4

Yoshinobu SATO

: An actual unmanned manufacturing station consisting of an NC machine and an industrial robot is assumed (Fig.1). The main hazards of the robot to a human, which occur during troubleshooting in an automatic operation mode, are enumerated using "Action-Chain models" (Fig.2). Fault Trees of accidents caused by the hazards are given as subsets of "Logic Models for the Analysis of Accident-causation-Mechanisms" (Fig.3), and minimal cut structures are represented by Priority-AND Gates to consider sequential properties of basic events (Fig.4). Furthermore, a manual safety mechanism, an automatic one, and a combination of the two, which cut off the power sources of the robot when a human enters a hazardous zone, are evaluated by comparing expected numbers of occurrences of the top events (Fig.5).

Considerations of the Safety Assessment of Construction Work under Compressed Air Condition --Determination of Coefficients in the Formula for Danger Potential and Development of the Computer Interactive Safety Evaluation System--

RR-86-5

Yoshimi SUZUKI and Ikuo MAE

: Detailed analysis of labour accidents on the construction sites indicates that one of the major causes responsible for these accidents is inadequate preparation for planning of the execution of construction work. In other words, if necessary measures for safety and health were established in the execution plan and carried out in accordance with it, many of the accidents could have been avoided.
    Judging from this point of view, the Ministry of Labour has presented guidelines which give the simple method to evaluate the possible dangers occurring during execution of construction work, and to make sure of the necessary safety measures at the designing or planning stage. This notion is called "Safety assessment of construction work". As one of these guidelines, "Safety assessment of construction work under compressed air condition" was discussed in a committee and was made public in 1985.
    Basic concept of this safety assessment is a comprehensive check system for setting up precautions at the stage of work planning prior to the initiation of work. The procedure of this safety assessment is to check the necessary safety measures against all anticipated accidents by predicting potential risks inherent to the site's characteristics of construction work under compressed air condition. This procedure consists of five steps (1st ; Collection of basic data, 2nd ; Check of basic matters, 3rd ; Ranking of potential risks of the anticipated accidents, 4th ; Check of the safety measures against inherent accidents, 5th ; Numerical expression of results of the assessment). In the 3rd step, the simple formula are adopted in order to estimate promptly and briefly the potential risks of accident.
    This paper deals with, firstly, outlines of the "Safety assessment of construction work under compressed air condition". (Fig.1 - 2, Table 1 - 4)
    Secondly, the results and considerations about questionnaires which were executed in order to determine coefficients in the formula for the danger potential. By these formula, the risk of a potentially serious, or more frequent accident specific to the construction work under compressed air condition in question (referred to as "specific accident" in this paper) is evaluated. This evaluation of the potential risk had to be desirably quantified in a scientific method for assessment. So authors proposed the investigation by questionnaires to technical experts engaging in construction work under compressed air condition. By means of the Delphi method, the results of these questionnaires were adjusted and the coefficients in the formula for danger potential were obtained. (Table 5 - 9)
    Thirdly, this paper presented the outline of "The computer interactive safety evaluation system" substantiated in this study. Taking into account the progress of CAD (computer aided design) or CAM (computer aided management) in the construction industry, the guidelines mentioned above and its procedures are considered to be in need of the computer interactive support. The safety assessment would be effective to prevent labour accidents, if its procedures were combined with these CAM or CAD system in the planning stage or the management stage of execution work. From this point of view, the authors attempt to develop "The computer interactive safety evaluation system" as one of the prototype of sub-program of these systems. By this substantiated program, precautions against the labour accidents will be pointed out systematically and facilely according with site's characteristics.
    The contents of this system consist of four main operation modes in accordance with the steps of the assessment procedures. These modes of operations can be choiced by the menu screens which were prepared beforehand. The current of operation can be continued or finished by selecting the sub-operation modes indicated on the sub-menu-screens which serve according to the datailed procedures of the assessment. The actual operation can be simply executed by inputting one character (alphabetical or numerical) on the menu-screens or the sub-menu screens. (Fig.3 - 11)

A study on the Time Intervals between Accidents (3)

RR-86-6

Shigeo HANAYASU

: This paper presents the statistical analysis of the time intervals between occupational accidents in order to discover if there is any significant tendency for changing accident situation in succeeding intervals of time.
    The main features of this paper are as follows :
  (1) The accident frequency rate has been widely used as the measurement of safety performance in a working place having a certain accident risk. Statistical inference, therefore, to find out the significant changes of accident situation can be performed by means of the testing hypotheses of the accident frequency rates using time intervals between occupational accidents.
  (2) Depending upon the several investigations of accidents in various industrial fields, the time intervals between occupational accidents were assumed as the exponential distribution and the gamma distribution as well. The testing hypotheses for the accident frequency rate can be achieved after connecting the parameter of these distributions to the accident frequency rate.
  (3) There are two kinds of statistical errors involved in a testing hypothesis for the accident frequency rate using time intervals between accidents. One is the error of rejecting the null hypothesis when the null hypothesis is true and the other is the error of accepting the null hypothesis when the alternative hypothesis is true.
  (4) In conducting the so-called test of significance to explore whether or not the time intervals between accidents is significant under the given accident frequency rate, it is unnecessary to employ the gamma distribution instead of the exponential distribution. Because this test of significance does not commit a type II error.
  (5) To evaluate the testing scheme, the operating characteristic function as well as the power function can be utilized. The critical region whose power function is always greater than others under the specified type I error is called the uniformly most powerful test. Testing hypotheses for the accident frequency rate using the exponential or gamma distribution can facilitate the uniformly most powerful critical region.
  (6) In order to satisfy the requirements on the tolerant risk imposed on the testing hypotheses, a considerable number of observational accidents are needed. This minimum number of accidents were analyzed corresponding to the types of test, discriminant ratios, and specified type I and II errors respectively.
  (7) Another functions characterizing the testing scheme are the average number of accidents to a decision and the expected time to reach a decision. These functions were calculated in accordance with the types of testing scheme.
  (8) Sequential probability ratio tests were also employed for the analyses of time intervals between accidents. This method significantly improves the number of accidents and the length of observational time to reach a decision compared with the aforementioned testing scheme.
  (9) According to the analyses of accident situation by testing hypotheses for the accident frequency rate, serious accidents involving more than three persons injured are significantly increasing in the tertiary industry such as the commerce and leisure industry sectors. So far as the types of accidents are concerned, combustible gas explosion as well as the falling accidents are increasing, while goods handling and crane & derrick accidents are decreasing.

The Effect of Air Velocity on Minimum Ignition Energy for Flowing Dust-Air Mixture in a Tube

RR-86-7

Toei MATSUDA

: In pneumatic transport system of combustible dusts, a dust explosion could be propagated extensively throughout the system by the presence of conveying air. Safety concerns are needed in the design and operation of the system.
    Experiments have been conducted to investigate the effect of conveying air speed on minimum energy requirements for the electric spark ignition in air flow of cork dust, ABC-resin powder and some other industrial types of dust in a small tube. The dust-air mixtures flowing at air velocities between 10 and 35 m/s were formed in the 4.2 or 5.3 cm in diameter by 13 m long horizontal steel tube of a suction-type laboratory pneumatic system. Dust was supplied at the air inlet of the tube by a vibratory feeder and optical monitors were used to follow uniformity of the dust concentrations through the tube. The optical transmission data at the cross-section of the 4.2 cm diameter tube, showed for the flow to be well homogeneously dispersed mixtures of dust at the concentrations up to 0.8 kg/m³, at which the dust probe showed an upper detecting limit. Therefore, the smaller diameter tube was more suitable for use in these tests.
    Ignition was produced with capacitance spark whose energy and duration could be varied independently. At a distance of 6.5 m downstream from the dust inlet, the spark electrodes of pointed steel bars of 2 mm dia. were situated vertical to the flow. A very wide range of phenomena is involved in two-phase flow. With reference to the available data on velocity gradient between particles and gases, tip of the lower electrode was fixed at a height of one-third of the tube diameter from the bottom wall.
    First in the preliminary ignition tests, the optimum spark duration and electrode gap width were determined at various dust concentrations and air velocities. The reverse effect of air velocity was found here both for the optimum spark duration and for the optimum spark gap width, since the reported data by other investigators suggest that optimum spark duration reduces and optimum gap width increases with increase in air velocity for gas or liquid spray fuels. The flow condition in the present test were different from that of the earlier studies. It will be also responsible for the effect that the relatively larger chemical time scale for the case of dust flames than gas flames is required.
    At a given velocity, measurements of minimum ignition energy versus dust concentrations were performed using predetermined optimum values of spark gap and spark duration.
    The ignition energy decreased rapidly as the dust concentration increased, until it reached the minimum for all the dust. After that, at the lower air velocities the ignition energy increased fairly slowly, but at the higher air velocities it raised sharply with increasing dust concentrations. The optimum dust concentration, at which the minimum ignition energy was given, slightly shifted its value to a lower side with increasing air velocities. Increase in the air-flow velocity markedly elevated the minimum ignition energy of flowing dust-air mixtures in the tube. A more detailed investigation of the flow parameters having an influence on minimum ignition energy in flowing dust mixtures will be required to obtain a fuller understanding of the dust ignition processes in the flows.
    From the relation between the ignition energy and the dust concentration, spark ignitability curves are drawn which separates the domain of concentrations and conveying air velocities that are spark-ignitable at an energy level from the domain that is not. Spark ignitability limits were narrowed by an increase in conveying air velocity. At an ignition energy level, the lower ignitability limit was almost not affected, but the upper one extremely shifted its value to a lower side with an increase in air speeds. Eventually they reached a critical extinction air velocity above which flame was no longer propagated at any dust concentrations at that given ignition energy for the dust-air flows in the tube. These ignitability curves validate the previously reported data, which have been obtained using a full-scale pneumatic transport system (7.5 or 10 cm in dia. and 86.6 m long pipes).

Characteristics of Gaseous Explosions in a Vessel with a Complex Inner Structure, and Pressure Venting by Rupture Disks

RR-86-8

Toshihiro HAYASHI and Hidenori MATSUI

: It has been known that explosion venting is one of the useful methods of protecting equipment or a vessel in which a gaseous or dust explosion may occur. Although many papers have been reported on explosion venting, most of them are concerned with explosions in vessels with simple shape and simple inner structure. Therefore, there always arises a doubt if the design criteria based on experiments with such a model vessel can be applied safely to protect real equipment used in hazardous situations. As is well known, the behavior of gaseous explosion depend largely on such factors of a vessel, in which a flame propagates, as the size, the shape and the inner contents (i.e. the existence of obstacles which may induce any turbulence resulting in the acceleration of flame speed). However, at present, no one can predict quantitatively how such factors affect on maximum pressure attained or other explosion characteristics.
    This report describes the behavior of gaseous explosions in a complicated vessel and also the venting of explosion pressure when the vessel is equipped with rupture disks.
    The test vessel, as shown in Fig.1, is a miniature model of a certain reactor, and is composed of outer and inner cylindrical shells, rod obstacles and two series of coils resembled to a heat exchanger. Each of the four openings at the top of the vessel is to be covered with a rupture disk or a blank flange.
    A 70% methane - 30% hydrogen mixture is used as a fuel gas. The pre-mixed gas of the fuel with air is prepared in a storage tank and is fed into the test vessel, following gas chromatographical analysis. The fuel content in the test gas mixture ranges from 6.5% (φ= 0.5) to 12% (φ= 1.0), where φ is a corresponding stoichiometric ratio. Initial pressure of the test gas mixture is varied from atmospheric up to 3.5 kg/cm² G.
    Ignition is made, by a nichrome-wire heater, at the bottom center of the vessel. A flame originated in the lower vacant space of the inner shell propagates upwards, at first, through the rod obstacles. After passing through the peripheral clearance, with a width of 11 mm, at the upper part of the vessel the flame propagates downwards through the coil obstacles. The whole length of flame path is about 2 meters.
    Explosion pressure-time histories at six different locations are recorded simultaneously in each test, by using a strain gauge-type pressure transducer, strain amplifier and electromagnetic oscillograph. Six measuring points are shown in Fig.1 as encircled numbers (1) - (6).
    In order to find out general behavior and to characterize explosions in this test vessel, a first series of experiments are carried out with a totally closed vessel, varying the fuel content and initial pressure of the test gas mixture. With mixtures of lower fuel content, the rate of pressure rise just before attaining the maximum pressure (P _e) is assumed to be constant (Fig.2). Subsequently the average rate of pressure rise (V _p) is read out on a pressure-time record, and the time to attain the maximum pressure (t ) is calculated by the equation t = ΔP /V _p or V _p = (P _e - P ₀) It, where P ₀ is an initial gauge pressure and ΔP is a difference between P _e and P ₀. With mixtures of 10% and 12% fuel content, pressure-time records suggest the transition of deflagration into detonation in the course of flame propagation. Explosion characteristics (P _e, t and V _p) in the closed vessel are shown in Figs.3, 4 and 5. The well-known explosion hazard index K _G, which is calculated from maximum rate of pressure rise (dP /dt )_max determined by igniting the quiescent flammable mixture in a vessel of an internal volume (V ) by the equation K _G= (dP /dt )_max·V^1/3 , is compared between typical flammable gases and the test gas mixtures in the present work (Tables 2 and 3), in order to predict the effect of the existence of obstacles on the flame path or the effect of turbulence induced by those obstacles.
    In a second series of tests, the effect of rupture disk on explosion pressure relief is determined. Two types of rupture disk with different static bursting pressure (P rupture) are used (Table 1). Effective diameter for each type of disk is defined here as an average diameter of the opening of the disk after explosion test, and the total vent area (S ) is calculated from this diameter and the number of disks provided for the vessel. Results of experiments are shown in Figs.6 - 11 and in Photos 1 and 2. From the viewpoint of practical application or the scaling up of the test results, discussions are also made on the use of non-dimensional vent area (A ) defined by A = S / V ^2/3 .
    Main results obtained are summarised as follows :
  1) The rate of pressure rise of an explosion in a complicated vessel can be far higher than that observed in a simple vacant one. Under suitable conditions, depending on the complexity of the inner structure of the vessel and the composition of flammable mixture, a transition of a deflagration into a detonation may occur.
  2) For explosions of lower pressure rise rate (less than 100 kg/cm² /s), the only one rupture disk can reduce the vented pressure (i.e. maximum pressure attained) to the static bursting pressure of the disk, provided that the effective diameter of the disk is adequately selected (i.e. non-dimensional vent area shall be larger than 0.025).
  3) For explosions of higher pressure rise (e.g. several thousands kg/cm² /s, but for detonation), the vented pressure may be reduced to a pressure higher by several kg/cm² than the static bursting pressure of the rupture disk, provided that the number and/or the effective diameter of the disk used is satisfactorily large (i.e. non-dimensional vent area shall be larger than 0.12). The provision of rupture disk for a vessel is almost useless for venting pressures where a detonation occurs in an early stage of flame propagation.
  4) Scaling-up effect of the size of complicated vessel has to be studied furthermore. But, the authors make bold to consider that the vented pressures for different vessels will be substantially equal where the vessel size is scaled-up with the same value of non-dimensional vent area.
  5) From the viewpoint of explosion pressure venting, no difference is found between tension loading type and reverse buckling type rupture disk.

Systematization of Spontaneous Ignition Temperatures of Organic Compounds (2nd Report) --Spontaneous Ignition Temperatures of Alkyl Esters and Alkyl Acids--

RR-86-9

Shozo YAGYU

: Ignition temperatures of combustible substances have ever been determined mainly by experiments. However, only a limited number of substances has been considered in this connection in the past, among a wide variety of combustible substances. Therefore, still more sufficient number of data should be accumulated from the practical safety engineering point of view.
    The author has intended since some years ago to systematize ignition temperatures of combustible substances, mainly organic compounds, in order to estimate the ignition temperature of a compound without actual experiments on the bench. For this aim, it may be necessary in the first place to investigate the relation between ignition temperatures and molecular structures of various organic compounds.
    M.G.Zabetakis ¹⁾ reported a relationship between the ignition temperature and average carbon chain length for the paraffin hydrocarbon homologues.
    In the previous report ⁴⁾, the application and improvement of his method to alkyl alcohol homologues was described. In this report, the method is applied to the alkyl esters and alkyl acids. As the result, similar correlations to those of the alcohols and hydrocarbons are given, as shown in Figs.11 to 13 which show that the relation between the spontaneous ignition temperature and average carbon chain length (index-number) is plotted on a reverse-S pattern curve. The estimation of the spontaneous ignition temperatures of the alkyl esters and acids would now be possible with aids of those figures.

Study on the Relationship between Chemical Structure and Thermal Stability of Reactive Chemicals (1st Report) --Isomeric Effect on Decomposition Characteristics (Part 1)--

RR-86-10

Takayuki ANDO and Shigeru MORISAKI

: In manufacturing, transporting, and storage of reactive chemicals in recent years, the hazard evaluations have been extremely important. In the chemical industry, the most concerns are focused on thermal hazards such as runaway reactions and thermal decompositions, which are mostly governed by thermodynamics and reaction kinetics of these reactive chemicals in the system.
    The purpose of this work is to establish the method of estimating the thermal hazards of reactive chemicals by investigating the relation between the chemical structure and the decomposition characteristics of these chemicals.
    In this paper, initial decomposition temperatures and decomposition energies of the three respective isomers of fourteen nitrobenzene derivatives were measured to clarify the isomeric effect on the decomposition of the isomers by thermal analysis. In the course of the studies of the decomposition mechanism for the isomers, the nitrophenylacetic acid showed a distinctive decomposition feature, and these three isomers were investigated by such methods as the pressure TG, evolved gas analysis with autoclave-MS, and the adiabatic calorimetry (ARC : Accelerating Rate Calorimeter).
    The results are summarized as follows ;
  (1) Except for the tendency of o-isomers to show the lowest initial decomposition temperatures and decomposition energies, there may be no systematic variation in the decomposition characteristics for the respective three isomers of the substances investigated here.
  (2) Nitrobenaldehyde, nitrobenzhydrazide, and nitrophenylacetic acid show two or more distinct exothermic peaks in the DSC curves, suggesting that they decompose in multistage reactions.
  (3) While ortho and para isomers of nitrophenylacetic acid show three distinct exothermic peaks, meta isomer shows a single peak in the DSC curves. But, as the gaseous species evolved at the first stage is mostly carbon dioxide for the three isomers, at least the initial step of the decompositions seems to be due to the same reaction mechanism, and the difference in DSC curves seems to result from the different reactivity of decomposition reactions following the initial step.

Monitoring of Insulation of Power Supply Lines Underwater and Detection of Earth-Leakage Current

RR-86-11

Tatsuo MOTOYAMA and Eiki YAMANO

: Working underwater such as in the sea or rivers and the use of electricity in such an environment are increasing in recent years. These situations may lead to the hazard of underwater electric shock to divers and other workers.
    This paper reports on the underwater earth-leakage detection (i.e. a detecting method by monitoring the insulation level of lines and a detector on the basis of the method) for the prevention of underwater electric shock.
    This detecting method have the following features :
  1) Monitoring the insulation level of lines,
  2) A certain degree of disturbance due to earth capacity of lines,
  3) The possibility of detecting a leak between lines.
    The results of the investigation are summarized as follows ;
  (1) A leak can't be detected by line insulation, but by d.c. earth-leakage current. Therefore, this method has somecases, where the leakage can't be detected between lines. In these cases, the area having the possibility of electric shock (i.e. an area more than 2.5 V/m) is not spacious.
For example, in the condition where leaking points are two spheres, the potential of which are 50 V and -50 V respectively, the line insulation resistance (R_x) is 20 kΩ and conductivity (σ) of water is 10^-2 (S/m), this area is included in two spheres which radii are 0.12 m (see 3.4).
  (2) The detector is connected to the resistance grounded neutral system, the neutral resistance of which is 20 kΩ, earth capacity of which is less than 1 μF, and the line voltage is 100 V. Then, d.c. 10 volts are superposed on the lines.
  (3) Where the threshold of line insulation resistance (R_x) is set to 20 kΩ, the time for detection of leak is less than 78 ms at R_x = 10 kΩ and 90 ms at R_x = 20 kΩ, respectively.
  (4) In a severe case of leak upper limit of Rx having possibility to detect the leak is greater than 50kΩ.

Considerations for Electrostatic Noise Caused by Electrostatic Discharge

RR-86-12

Yasuyuki TABATA and Hajime TOMITA

: It is well known that electrostatic discharge (ESD) and electromagnetic wave radiated with ESD cause various malfunctions of sophisticated electronic components and devices. For studies of electrical noises in respect of ESD, electromagnetic fields have been investigated to solve electromagnetic interference (EMI) problems for electronic systems. Electrostatic fields are, however, hardly considered.
    Electrostatic field produced around a charged material varies generally and is incoherent when ESD occurs. As a result, it seems that the incoherent electrostatic field affects the signal level of the electronic circuit assembled by components with high impedance. Therefore, electrostatic noises caused by the incoherent electrostatic fields have been examined theoretically and experimentally in this paper.
    First, the theoretical analyses have been developed, using a charged di-pole as an ESD model, both to identify the electrostatic noise caused by the induction of incoherent electrostatic field and to determine the optimum conditions of experiments. From these analyses, it has become clear that the maximum electrostatic noise generates at the moment ESD occurs and that it may cause the malfunctions of the electronic circuits with high impedance. These estimations about the electrostatics introduced from the analyses have been confirmed by experiments.
    The results obtained from both the analyses and the experiments are as follows :
  (1) ESD generally produces the incoherent electrostatic as well as electromagnetic fields around the charged material.
  (2) A region of the incoherent electrostatic field extends to the fairly far distance, a few meters, when ESD occurs from the charged body with a large capacitance, such as a charged human body, table and wagon.
  (3) A level of the electrostatic noises caused by the incoherent electrostatic field is generally high, a few volts in voltage level and comparable to the signal level of the general electronic circuits.
  (4) Electrostatic noise in respect of ESD causes malfunctions of the electronic circuit with high impedance, located within a few meters from a position of ESD.

JNIOSH