Abstract of Special Research Report (RR-91)

National Institute of Occupational Safety and Health, Japan

On the Stress of Bolts of the Slewing Circle of a Truck Crane Extracting Sheet Piles

RR-91-1

Yutaka MAEDA

: A 30 tonf telescopic jib crane was fractured twice by breaking the bolts connected to its revolving circle. To clarify the cause of the accidents, experiments to measure the stress of those bolts were carried out. This paper describes the circumstances of the accidents and the result of the experiments.
    The first accident occurred when the truck crane lifted a girder of a bridge crane as the weight of the load was about 1.5 times as much as the rated load. Although the main cause of this accident might be due to the overload, there were some evidences that the bolts were broken by fatigue failure, but the details were not clarified for lack of sufficient evidence.
    After four years of the first accident, the same crane was fractured once again when it was directly pulling out a steel sheet pile. In this time, most of the bolts were obviously broken by the cause of the fatigue failure, but the magnitude of the pull load was unknown for the sheet piles were fixed by the ground. After the first accident, the bolts were replaced from SCM435 to SNCM630 to raise the tensile strength, and were made longer by inserting collars to improve the fatigue strength.
    Experiments were carried out as follows: all the bolts installed in truck crane were exchanged to the new bolts for the measurement, and the sides of the bolts had been cut before strain gauges were glued. Then their initial stresses were set at 470 MPa. At first, static stresses were measured when the crane lifted various weights of load. Next, dynamic stresses were measured while the crane pulled a steel sheet pile with a vibro hammer, or extracted it directly, pulled it out till the outriggers detached from the ground, and furthermore, winched up and down to swing the crane itself to see the dynamic effect.
    The results of this measurement were: (1) Load concentrations to some bolts located near the web plate of the base carrier were observed both by the static and dynamic stress measurements. (2) The maximum stress was around 700 MPa with the static test. (3) When a vibro hammer was used, the maximum pull load was 1.5 times as much as the rated load, and the stress of bolt was 886 MPa. (4) By swinging the crane, the maximum pull load reached to 1.9 times as much as the rated load, and the stress of bolt reached to 988 MPa.

Fail-safe Combustion Control and Flame Detective Sensor for Small-sized Boiler

RR-91-2

Shoken SHIMIZU, Noboru SUGIMOTO, Hiroyasu IKEDA and Shigeo UMEZAKI

: As a function in a boiler becomes more complicated, a computer control system is applied for a boiler to advance its function. However, such advanced systems are not always safe, because of the complicated mechanisms. Small size boilers, which can be easily operated by workers without particular technique, have widely been used in small factories such as a laundry or a bakery. Though the majority of the small size boilers have the advanced control systems, serious accidents such as burst or explosion sometimes occur in the small size boilers, because of the complexity of control systems.
    The main cause of the accidents is often associated with explosion of unburned gas in the furnace of the boiler. To prevent such accidents, it is necessary to develop a new type of sensors which are able to detect the unburned gas or the flame of the combustion gas.
    The main aims of this study are ;
  (1) to develop the fail-safe sensors which can monitor the combustion state of fuel in the furnace,
  (2) to develop the fail-safe control systems for small size boilers,
  (3) to evaluate the practicality and limitations of the developed sensors and the control systems.
    Three types of flame detection sensors, that is, thermocouple, UV tube and ultrasonic sensors were developed by applying fail-safe mechanism.
    A new control system which is composed of an AND gate, an OFF delay element and a window comparator was constructed. The AND gate is a kind of a logical element which has a function to be able to emit an output signal when two input conditions are satisfied. In case of applying the AND gate for the boiler, the AND gate functions to emit the output signal which permits the operation of the boiler, only when the self-check signal for the control circuit and the flame detection signal are normal. The new fail-safe control system can be achieved by the AND gate.
    The OFF delay element functions to control the purge time and the fuel injection time in the furnace of the boiler.
    The window comparator functions to emit the output signal when a signal from the flame detection sensor exists between upper and lower limits (window) of the comparator. If the signal level is higher or lower than the limitations, the output signal becomes zero. As a result, a new control system including the window comparator cuts off the supply of fuel immediately and operation of the boiler is shut down automatically.
    The new control system developed enables to release fuel into the furnace and to ignite the fuel only when the whole control circuit is in normal state. Accordingly, the new control system makes possible to keep a boiler under safety conditions even when a worker misoperates.

Study on Fail-Safe Mechanism for Power Press Control

RR-91-3

Shigeo UMEZAKI, Hiroyasu IKEDA, Shoken SHIMIZU, Soichi KUMEKAWA and Noboru SUGIMOTO

: Even today, nearly 5,000 labor accidents are caused each year by power press operation. Particularly in operating friction clutch type power presses, many accidents are caused by the failure of electric control systems including light beam type safety sensors. To prevent these accidents, periodical inspection have been stipulated by industrial regulations. The electric control system should have an inherent fail-safe structure to prevent any accident due to failure.
    The conventional power press control systems contains fail-safe circuits formed with relay contacts to achieve a fail-safe mechanism. A problem with this relay contact type circuit is that the structure is very complicated, in order to counter the possible relay failures at both the ON and OFF state. Recently, for improving the reliability by simplifying the structure by not using any relay contacts, the conventional relay contact type circuits have been substituted by electronic circuits. It should be noted that, unless safety is considered, the electronic circuits could cause serious accidents. To give a solution to this problem, this paper proposes a power press control circuit having a highly reliable and fail-safe structure by using fail-safe logical elements.
    There are two cases of the emergency stop of the slide in power press control where deterministic control is required to prevent an accident. They are (1) human hands may come into the die and (2) an error to the stop at the upper dead point (causing overrun). This paper deals with case (2). In the proposed power press control, a confirmation segment is provided in the transition segment from the slide up process to the slide down process. In this sub-segment, the brake ability is checked (i.e., overrun monitoring), and, if the sufficient brake ability is confirmed, the slide down process is permitted.
    It is also indicated in this paper that, to confirm that the start push-button switch is in the OFF position, an interlock mechanism is required in addition to the slide stop command signal and the slide stop confirmation signal. Furthermore, it is indicated that, to realize this mechanism, a self-hold circuit is configured with fail-safe logical elements, and that the slide down process can be permitted only when the output signal from this self-hold circuit is generated.

Statistical Analysis on the Damage Consequences of Occupational Accidents

RR-91-4

Shigeo HANAYASU

: This paper deals with the statistical analyses of occupational accidents to explore the basic statistical characteristics of the damage consequences due to occupational accidents. Emphasis was placed upon the probabilistic and statistical analyses to clarify, in particular, the relationships between frequency of occupational accidents and their consequences.
    In this paper, the damage consequences due to accidents were classified into two categories such as the number of days charged which required injured workers away from the work duties, and the number of injured workers involved in one accident. The workdays lost due to accidents indicates the severity or hardness of accidents imposed on the injured workers. The number of injured workers involved in one accident, on the other hand, implies the magnitude of accidents.
    The occupational accident under investigation in this study were those taken place during the tunnel construction works for the New Sanyo-Shinkansen and New Joetsu-Shinkansen railways. Also, accident data concerning the workdays lost due to permanent disability accidents were collected. These data were utilized mainly for the analysis of the accident severity and frequency relations.
    In addition to these accidents, serious occupational accidents involving three or more injured workers within an accident from the year of 1977 to 1990 were collected. These accident data were used for the magnitude-frequency relationship analysis.
    The main results obtained in this study can be summarized as follows:
  1) The relationship between damage consequences due to occupational accident such as the lost workdays and the number of injured workers involved in an accident and their frequencies can be represented by a simple power function which indicates the log-log linear relation.
  2) The average workdays lost due to two Shinkansen tunnel construction accidents were estimated between 80 to 90 days and no significant difference was found between two project's accidents.
  3) There was a difference between industrial sectors and types of accident in terms of the accident magnitude. Agriculture and forestry are the most dangerous industries. While, explosion and fire is the most serious types of accident, followed by traffic accident.

Analysis on Properties of Information Concerning Labour Accident in Tunnel Construction Work

RR-91-5

Yoshimi SUZUKI

: In this study, information obtained from the Database of Labour Accident Report concerning tunnel construction work was applied and examined for the analysis of labour accident in tunnel construction work. 3070 kinds of free keywords which were used in the Japanese sentence item applied in "Situation about occurrence of labour accidents" of "Labour accident report", were separated by means of the utility in the Database. These free keywords were distributed by Bradford's law.
Using the 156 free keywords of high frequencies, the relationship between these keywords were analysed by the quantification method of third type. As a result, labour accident information concerning tunnelling work was clustered into three groups according to their tunnelling methods.
After the quantification method of second type, each accident case can be discriminated according to its tunnelling method or its work-stage by using these keywords.

Stability of Slope Reinforced with Geogrid

RR-91-6

Yasuo TOYOSAWA and Satoshi TAMATE

: Reinforcement of slope with Geogrid is very effective in improving slope stability. There are many different kinds of reinforcement materials, among these materials, the Geogrid is becoming more popular. In the near future, the Geogrid for reinforcement will come into wider use, especially in construction works with steep slope. However the failure mechanism of a slope reinforced with Geogrid is not well known. The characteristic of the Geogrid used in this report shows that its strength and rigidity are nearly equal to iron, but its weight is much lighter than iron, (see Table 3)
    The series of dynamic centrifuge model tests were conducted in a 50g (50 times gravity) field. The horizontal acceleration applied to the model was approximately 20 g. Table 4 is the summary of the test conditions and results. The schematic test model are illustrated in Fig.3. The model grounds used in those experiments were composed of Kanto-roam and Toyoura sand.
    The purpose of the centrifuge experiments was to investigate the deflection behavior and dynamic characteristics of slopes reinforced with Geogrid. The responses of these reinforced slopes and un-reinforced slopes in earthquake simulations were also measured by accelerometers and displacement transducers. The displacement of soil and Geogrid were observed by X-ray photography, (see Fig.4)
    The centrifuge tests support the following results;
  (1) Reinforcement with Geogrid improved the stability of slope. It was also effective in improving the stability even during an earthquake.
  (2) Geogrid in the ground could have the effect of interlocking the soil and sands around it, then it sustained the strain of the soil small. So the modules of shear rigidity at reinforced area became higher than the other area.
  (3) As a result of the earthquake simulation, un-reinforced slopes collapsed during shaking, with the response of acceleration increased at upper part of embankment slope. On the other hand, the reinforced slopes with Geogrid didn't collapse. As the consequence of these data, it is implied that Geogrid can reduce the decline of shear rigidity due to the increase of strain of soil by cyclic shear stress.

Dust Explosion Characteristics of Rare Earth Metal Alloys

RR-91-7

Marek WOLINSKI and Toshihiro HAYASHI

: Rare earth metal alloys are now commonly used in such high-technology industries as manufacturing electronic devices and magnetic components for special use. Rare earth metals are known to be easily subjected to oxidation and their alloys are often utilised as fine powders. Thus their powders are ready to explode in industrial processes when dispersed adequately into air and ignited. Yet no sufficient information is available on dust explosion characteristics of fine powders of those new materials.
    This paper describes general behaviours of dust explosion in air for five commercial alloys containing rare earth metals, together with such parameters of explosion as maximum pressure, rate of pressure rise and lower limit of inflammability. Inhibitory effect of Baton 1301(CF₃Br) or nitrogen on explosions of those metal alloy powders was also studied.
    Explosion tests were carried out in an enclosed cylindrical chamber of 5.3 liter capacity; alloy powders were blown up, from the bottom, into the chamber by means of pressurized air, and then ignited by an electric arc discharged at the center of the chamber. Pressure of air for blowing up powder was varied according to the mass of the powder to be tested. Ignition timing and other experimental parameters were predetermined by the help of a transparent plastic vessel, which had exactly the same dimensions as the test chamber.
    Two Nd-Fe alloy powders of different compositions gave moderate explosions; maximum explosion pressures were nearly equal to one half of those reported for most of flammable dusts, yet the rates of pressure rise were relatively slow. Explosions of Co-Ce-Fe alloy powder, which contains large portion of Ce and named here as powder "D", gave slightly higher pressure and pressure rise rate than those for Nd-Fe alloys. Most remarkable feature of powder "D" was that it often exploded in the course of dispersion by blowing air before activation of the ignition source; in such explosions, much higher pressure rise rates were recorded than those obtained in tests by the ordinary procedure. Other alloy powders, Co-Sm-Fe (powder "A") and Co-Ce-Sm-Fe (powder "E" ) gave no explosion, in so far as the ignition source was applied after their dispersion; however both of them could be ignited and exploded when they were blown onto electric arc. Powders "A" and "E" were thus flammable, even if no explosion parameters could be determined by the ordinary procedure in the present work. It must be stressed that all powders under test were capable of smoldering.

Study on the Relationship between Chemical Structure and Thermal Stability of Reactive Chemicals (2nd Report) --Thermal Decomposition of Nitrophenylacetic Acid Isomers--

RR-91-8

Takayuki ANDO

: 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 our previous report, the decomposition characteristics of the three isoiners of nitrophenylacetic acid (NPA) were measured with Differential Scanning Calorimeter (DSC) to clarify the isomeric effect on the decomposition of the isomers, and gaseous decomposition products of the isomers were analyzed with autoclave-Mass Spectorometer (MS) to investigate the isomeric effect on the decomposition mechanism for the isomers.
    In this paper, the three isomers of NPA were allowed to decompose with DSC and Curie Point Pyrolyser, and the decomposition products were analysed with GC (Gaschromatograph)-MS to get more detailed information on the decomposition mechanism for the isomers.
    The result are summarized as follows:
(1) The first exothermic peaks in the DSC curves for ortho and para isomers of NPA are mainly due to decarboriation, and carbon dioxide and nitrotoluenes are formed.
(2) The second exothermic peaks in the DSC curves for ortho and para isomers of NPA are mainly due to the decomposition of nitrotoluenes formed in the first step of decomposition.
(3) As m-nitrotoluene is not found in the decomposition products of m-NPA, the decomposition reaction of m-NPA is not simple decarbonation, but seems to be spontaneous formation of carbon dioxide and m-nitrobenzaldehyde.

Consideration of Hazard Test Methods for Reactive Chemicals

RR-91-9

Haruhiko ITAGAKI, Yasuhiro FUJIMOTO, Takayuki ANDO, Toei MATSUDA and Takashi KOTOYORI

: In chemical industries, various tests have been carried out to evaluate potential hazard of the reactive chemicals which easily decompose or ignite by heat or mechanical shock. Usually, those testing methods have been used pure chemicals. These reactive chemicals, however, are often contaminated with residual reactants or impurities such as rust in manufacturing processes. Therefore, the evaluation for the contaminated reactive chemicals may be different from those for the pure chemicals.
    In this report, the following three items were investigated to evaluate the hazards of reactive chemicals with a view of handling in the manufacturing processes, using for an example benzoyl peroxide(BPO).
  (a) Influence of thermal stability by contamination.
  (b) Hazard of dust explosion.
  (c) Decomposition hazard resulting from ignition of contaminated flammable liquid.
    The results showed that some of the impurities in BPO such as residual reactant of benzoyl chloride or contaminated substances of sodium hydroxide, ferric chloride or ferric oxide decreased the thermal stability.
    From the result of BPO dust explosion test, the minimum ignition energy was around 10 mJ and the concentration of explosion lower limit was presumed to be less than 15 g/m^3.
    The residual methanol in BPO gave a great effect in the explosion hazard, showing that the ignition of methanol vapour easily propagated to the ignition and explosion of the BPO powder clung on the wall of vessel. The accumulated BPO on the bottom of the vessel, however, would not ignite by the methanol flame if fine BPO were not existent.

Observations on Motility of Rabbit in Exposure to 2.5 V/m Electric Field in Water

RR-91-10

Eiki YAMANO and Tatsuo MOTOYAMA

: The electric shock accident is also possible for divers, since electricity is used in under water work. In the accident, if a diver is motile (able to move voluntarily), he will be able to swim out of danger by himself.
    This paper, which is supplementary to the last (1989), reports observations on the motility of rabbits and the degree of shock effects in the exposure for 100 min to the electric field of 2.5 V/m in water. The 2.5 V/m is a limit for no stiffness which was set in a previous investigation and we supposed that the limit might also allow voluntary motion. In this experiment, it was intended to prove that 2.5 V/m could be used as an allowable limit for motility.
    In the 1st pattern of tests, a rabbit was bound on a stand to observe respiration and EGG (discretely). And in the 2nd pattern, a rabbit was free and able to walk on the bottom. The electric field was uniform and the commercial a.c. of 50 Hz. Water in test tank was 3.5% salt water, with a temperature of 30 °C and a conductivity of 6 S/m.
    From the observations, we found the following results.
  (1) Relative frequency to be motile is considerably great in 2.5 V/m. However, convulsions (extensions) of forelegs were observed in 1 test out of 28 tests. The rabbit recovered motility after all convulsions in several minutes. The value of 2.5 V/m is not considered to be proper for motile limit.
  (2) It is considered that the electric field of 2.5 V/m has no lethal effect, excluding the secondary effects of convulsion such as a drowning.
  (3) Convulsion in hind legs occurs with high probability in the electric field of 5.3 V/m. The rabbits, however, bent legs again in many cases. The rabbits were not killed in the exposure because they were bound and protected from drowning.
  (4) Low temperature water (e.g. 10°C) is harmful to not only motility but also respiration and heart beat.

Thermal Deterioration of Natural Rubber Insulating Material and its Evaluation

RR-91-11

Kenji ICHIKAWA

: In order to keep good insulation performance of insulating personal protective equipment and device for linemen, it is necessary to develop a method evaluating the insulation performance with electrical non-destructive test. In that case, it is important to know deterioration features of insulating materials being used in them and to establish the measuring technique.
    In this paper, natural rubber material was adopted as insulating material being used in electrical rubber gloves for high voltage live line working. The material was artifically deteriorated by the application of heat and its physical properties such as chemiluminescence amounts per second (refered to as "CL amounts") and dielectric loss tangent (refered to as "tanδ") were measured with the examination of the relations between those properties and thermal deteriration (heated days).
    Principal conclusions obtained from the experiments are as follows:
  (1) CL amounts depend on the measuring temperature over 60°C and reach certain saturation after about 5 minutes. The saturated CL amounts were increased as the material was deteriorated, showing that there was a good linear correlation between CL amounts and heated days.
  (2) The activation energy, one of parameters to estimate the thermal endurance life of material, was calculated from the saturated CL amounts and the reciprocal temperature, and giving 47.3 kJ/mol.
  (3) The tanδ was measured when high voltage, whose frequencies were changed from 70 to 400 Hz, was applied to material. A smooth increasing tendency of tanδ was found until the applied voltage of 3 kV. Also, tanδ was increased as the material was deteriorated, and it was observed that there was a good correlation between tanδ and heated days.
  (4) The tanδ was also measured when the measuring frequencies with low voltage were changed from 30 Hz to 3 MHz. The values were similarly increased as the frequencies became higher, but a difference due to heated days was not found at frequencies of 400 Hz and less showing that it was smaller than tanδ measured with high voltage. However, the difference appeared clearly at frequencies from 1 kHz to 300 kHz.
  (5) The tanδ was measured at the low voltage of 1 kHz when the measuring temperature was changed from -60°C to +80°C. The values were greatly changed with increase of the temperature, but a difference due to heated days was only found at the temperature range of -20°C and +20°C.
(6) Relative dielectric loss factors were calculated from conductance and capacitance of material obtained with experiment and it had also a good correlation with heated days.

A Study on Electrostatic Coagulation of Bipolarly Charged Mists

RR-91-12

Shigeo YAMAGUMA, Tsutomu KODAMA and Yasuyuki TABATA

: In some industrial processes such as rolling and cutting, a large amount of mists can be generated from machinery by the vaporization of cooling oils, lubricants etc. and are often dispersed in environment for a long time. Some mists are inflammable or toxic. Hence, they may cause serious labor accidents such as fires and explosions or aggravate workplace environments. At present, electrostatic precipitators, bug filters and scrubbers are widely used to cope with such hazardous particulate materials. However, particles of several micrometers or below one micrometer are difficult to be treated because of their small size. For example, those small particles have insufficient electrification capabilities. So it is inevitably important to enlarge the size of mist particles in order to efficiently treat them with electrostatic methods.
    In this study, a method in which oil mists are divided into two portions, where positive and negative charges are separately given and then mixed together again, was employed. In this method, rapid coagulations of particles were confirmed through observation of the size distribution changes. We carried out both experiments and computer simulations to clarify basic characteristics of charged mists, the mechanism of electrostatic coagulation and important parameters to enhance coagulations.
    The results are summarized as follows:
  (1) After mixing bipolarly charged mists, their distribution shapes change from the original "one-peak" type to "multi-peak" type as time passes. The median diameters also become larger as the charging currents of the corona chargers increase. For example, when ±35μA were applied, the median diameter of mists almost doubled within two minutes -- from 0.86μm to 1.6μm.
  (2) The computer simulations show that electrostatic coagulations are caused mainly by coulomb forces exerted among charged particles. It is also revealed that the time needed for a certain number of coagulations to occur is inversely proportional to the square of the amount of initial charges of mists and directly proportional to the cube of initial distance between two adjacent particles.
  (3) According to the computer simulations, if mists are unequally charged, the number of particle which collide more than twice becomes much larger. This suggests that unequal charging may produce larger particles than equal charging.

JNIOSH