Abstract of Special Research Report (RR-95)

National Institute of Occupational Safety and Health, Japan

Low Velocity Control of a Pneumatic Cylinder --The Effect of Nonlinear Friction on Pneumatic Cylinder Motion and the Compensatory Method Using ERF--

RR-95-1

Hiroyasu IKEDA and Noboru SUGIMOTO

: With all the advantages over other actuators, such as high power/weight ratio and intrinsic compliance, the use of a pneumatic cylinder is limited to simple tasks at present. This is because the pneumatic cylinder has stick-slip motions at low velocities caused by a nonlinear friction force due to the mechanical scaling parts. Stick-slip motion has so far been considered not only difficult to control the position or low velocity of the pneumatic cylinder, but potential cause of an accident due to unexpected action of the cylinder.
    The electro rheological fluid (ERF) has lately been attracting a great deal of attention as a new functional material. ERF can increase its shearing stress when subjected to an electric field. Therefore, by use of this characteristic, a functional damping device capable of controlling the damping force by controlling the electric signal can be developed.
    This paper centers on a development of a compensatory method capable of achieving a constant and slow motion of a pneumatic cylinder by using an ERF damper. This method aims to overcome any decrease in the nonlinear friction force, i.e. the friction force against the increase in piston velocity, which may cause stick-slip motions. In order to achieve this compensatory method, the authors have used ERF of micro-cellulose particles dispersed into silicone oil that can rapidly generate shearing stress in proportion to the approximate square of the applied electric field.
    The ER-actuator developed for the study is united with the pneumatic cylinder and the ERF damper fulled with this ERF. The nonlinear friction characteristics of the piston cylinder is analyzed by means of lubricant models, and the above compensatory method is examined.
    As a result of this study, applying the electric field proportional to the piston velocity as a feedback realizes constant slow motion of the piston, which has been considered hard to realize by usual PD controllers. This compensatory method is proved to be applicable to the pneumatic cylinders in a wide range of velocity controls.

Fundamental Structure of Safety Control System --Principles of Safety Control and Configulation Method for Fail-safe System--

RR-95-2

Noboru SUGIMOTO, Shigeo UMEZAKI, Hiroyasu IKEDA, Soichi KUMEKAWA and Kiyoshi FUKAYA

: Recently, the reliability of industrial machines has been remarkably improved. In reality, however, there still have been a large number of accidents caused by industrial machines mostly due to failures in safety securing means. On the other hand, the evolution of mechatronic machines has brought new safety problems, such as the runaway of machines caused by electromagnetic noises. Therefore, the safety securing means of a machine should be so fail-safe as to exactly prevent any accident by stopping the machine in case of machine failure.
    In view of the above consideration, this paper describes a theory of securing safety for such a system that enable men and machines to cooperate to work each other, and examines the validity of the system configuration theory proposed for a fail-safe system to achieve such a system.
    Chapter 2 treats the safety promotion process through many experiences of accident. Such a conventional safety system that people accept the occurrence of a hazardous failure owing to the probabilistic reason are discussed. This study is based on the concept of fail-safe that never accepts the failures of hazardous side.
    In chapter 3, such logically unite relation that outputs energy only when safety is confirmed is discussed. Energy used for the industrial purpose is basically different from the energy that causes natural disaster. Industrial energy should be normally controlled and any disaster should be avoided in deterministic (not probabilistic) manner by means of fail-safe control technology. The reason for deterministic safety based on logically unate relation is explained by a thermodynamic model.
    A logical structure of safety securing system is discussed in chapter 4. An accident is analyzed, and the accident process is expressed by binary logic. The safety defined as the denial of accident can be indicated as a confirmation of safety by means of safety signal. Such uncertain condition that safety is not confirmed should be judged as hazardous for a reason that safety must be guaranteed by all means. This truth is concluded as the "principle of safety confirmation".
    In Chapter 5, an evaluation of safety is newly proposed, and in chapter 6, a model of the safety man-machine operation system attained through formulation is actually applied to the working site where men and machines are collaborating each other, and the validity of such model is demonstrated.

Error Detection by Coded Microprocessor and Its Evaluation

RR-95-3

Hiroyasu IKEDA

: A microprocessor is indispensable to complex control system such as railways, aircraft and atomic power plants. However, standard mono-microprocessors are obviously not designed fail safe, and they require protection against programming errors and hardware failures when used in safety-related systems or critical control fields.
    In order to realize such protection, multiple microprocessors with a voter (for example, a majority element for 2-out-of-3) have been widely used to give high dependability, which equates to high safety, high-security, high-availability and high-reliability. However, the redundant hardware system may become complex and be expensive, thus this kind of system would imply being used only in a limited process.
    On the other hand, various coding techniques of information sources and channels have been developed in communication systems to improve transmission reliability or fidelity. A redundant arithmetic coding has been specially considered for vital data processing. It is required reducing hardware complexity to improve speed and to ensure reliability in computation. Above all, redundant residual codes have been used as error-detecting or error-correcting codes in data processing.
    The coded microprocessor, which produces and deals with such residual codes to protect against hardware failures, has redundancy not of hardware and software but only of information. This unique innovative technique has been designed to detect errors occurring in data processing.
    In this paper, the architecture of the coded microprocessor to detect any potential error during the processing of data was proposed by means of the coding principle. Further, the error detectability of the coded microprocessor was discussed.
    The coded microprocessor is a standard mono-microprocessor running one program with external fail-safe monitoring or checking hardware (such as code generator, clock generator, dynamic comparator), and uses all coded redundant information (data) and added signatures based on residual arithmetic (remainder theorem). Therefore, it has the safety level that depends on only information redundancy, not on the total number of redundant codes and the processor or its reliability, provided that the global signature streams generated by the microprocessor by comparison with a predefined sequence.
    The error detection capabilities is presented as well as the assessment of the non-detection probability of errors. For example, such error detectability can be set up 10^-12 in terms of probabilities, if the check word length is 40 bits. The safety demonstration of the coded microprocessor has been investigated and accepted so that any error which may occur in the processor is detectable by the code and the probability of non-detection of error is defined by choice in a random manner.

Strength of Softwood-Plywood Used for Scaffold Planks

RR-95-4

Yoshimasa KAWAJIRI and Katsutoshi OHDO

: Thirty years have passed since the plywood scaffold planks of tropical hardwood began to be produced and be used in Japan. Recently their demand and output are decreasing compared to that of steel or aluminum scaffold boards, but they are still used in many construction sites.
    On the other hand, the conservation of forests in the torrid zone as one of environmental problems have been advocated for more than ten years. Most of plywood panels are produced and used in Japan are made of tropical hardwood. Therefore, the plywood makers and the authorities concerned are making many experiments or trials to reduce of consumption of the tropical hardwood, or to change the tropical hardwood into the softwood produced in plantations managed continuously. Then we studied on the strength of the plywood scaffold planks made of softwood, so as to examine the possibility of their practical use. In the experiments, tested were order-made softwood plywood planks faced with the Southeast Asian hardwood veneers and commercial plywood planks made of Southeast Asian hardwood veneers only. They were subjected to bending and gluing tests, after 3-24 months exposure to weather.
    Test results are summarized as follows :
  (1) The bending performance (the modules of rupture and the modules of elasticity) of the softwood plywood planks was lower than that of the commercial plywood planks, but was higher when compared statistically with that added the data for the commercial plywood in this tests and previous tests. The bending performance of the softwood plywood planks with the face veneers jointed by scarf-joints was poor and less than the values of the constructional requirement in the Ordinance on Industrial Safety and Health.
  (2) The ratio of the reduction of bending performance of the softwood plywood planks caused by out-door exposure was nearly equal to that of the commercial plywood planks.
  (3) The gluing strength of the softwood plywood planks was lower than that of the commercial plywood planks, but in combination with the wood failure it satisfied the criteria of the Japanese Agricultural Standard for structural plywood.
  (4) The expected life of the softwood plywood planks was predicted statistically to be about four years, but taking into account practical conditions of use in the construction sites, it would be less than three years.
  (5) High correlation was found between the modules of rupture and the modules of elasticity for the softwood plywood planks. Therefore, the non-destructive bending test for the used scaffold planks are effective in predicting their bending strength.
    In conclusion, the softwood plywood scaffold planks faced with the Southeast Asian hardwood veneers, except those with face veneers jointed by scarf-joints, would be used as the scaffold planks.

Imbalance of Human Standing Posture with High Heeled Shoes

RR-95-5

Hisao NAGATA

: According to the author's analysis of occupational injury data related to falls, young female employees wearing high or semi-high heeled shoes are relatively prone to fall especially while descending stairways. In this study the influence of the heel height of shoes on human standing posture was experimentally investigated by an originally devised linear accelerator to give subjects horizontally accelerated forces. Twelve young females of the age around 20 participated in a series of experiments. Each subject was asked to stand upright with either bare feet or four shoes with different heel-height on a movable cart of the linear accelerator. This was then operated at different rising acceleration to the point at which the subject fell. The patterns of acceleration were controlled like rectangular waves by use of a servo controller.
    From the experimental analysis, the following conclusions were obtained.
  (1) When giving a forward force to a standing subject, a significant difference between bare feet and high heeled shoes was found on the effect of critical acceleration to cause imbalance of human standing posture.
  (2) The critical acceleration for subjects with high heeled shoes (heel height 89 mm) was 38% less than that for subjects with lower heeled shoes (heel height 12 mm), when giving a forward force to a standing subject.
  (3) Heel height less than 30 mm was recommended from the safety viewpoint of attaining stable standing posture against acceleration disturbances.
  (4) The critical acceleration and the reciprocal duration time was linearly correlated.
  (5) There is a possibility that the dynamic method to give actual accelerated forces to a standing subject so as to lose their balance will cover the role of a force plate which is in general used in the medical field to measure the sway of gravity point of human standing body and to examine the sense of equilibrium.
  (6) Not only safety performance of heel height but also of other working conditions such as efficiency of handrail, the decline of physical performance by aging etc. could be numerically evaluated by the critical acceleration values to cause imbalance of human standing posture.

Influence of Weather on Work Efficiency in Construction of Tall Structures

RR-95-6

Katsutoshi OHDO

: Height of bridge pylons and buildings has been increased significantly in recent years. Tall buildings with the height of more than 500 m and long bridges with the span exceeding 2.5 km are now entered upon planning phases. In planning of such tall buildings and long bridges, their structural resistance after completion against the wind and earthquake loads has been extensively discussed, but constructability of such buildings and bridges is not considered in detail yet. In construction of such tall structures, the weather conditions such as wind, rainfall and temperature have more strong influence on work environment and efficiency than that on conventional structures. For constructing extremely tall structures in future, there is a strong need to identify the problems of the present construction techniques, and to develop technologies for safe and speedy construction under severe weather.
    In this study, for this purpose, major recent construction sites in Japan, 10 tall bridge pylons and 4 tall buildings, were selected for the investigation. Those sites were studied by investigating job schedules, interviewing supervisors and workers and sending out questionnaires to workers. This investigation was intended to survey how the weather affected work environment and efficiency, and to clarify what kind of weather-robust construction methods were needed.
    The results of the investigation are summarized as follows.
  (1) The weather influence estimation for construction schedules in building plans is more quantified and detailed than that in bridge plans.
  (2) For the highest bridge pylon and building in Japan with the height of about 300 m, the rates of non-workable days were as high as 15% respectively. In construction of the bridge pylons, pulling down work of scaffoldings was influenced mainly by wind, while welding work was more influenced by rainfall for the tall building.
  (3) More than 80% of the workers have experienced difficulty or felt danger by swing and rotation of the load in crane work under strong wind. Control systems to reduce swing and rotation of the heavy weight blocks need to be improved.
  (4) Based on the questionnaires addressed to the workers, construction work under extremely hot or cold weather was stopped only with around 10% probability, though the construction work under extremely strong wind or heavy rainfall was stopped with about 60% probability. The weather element such as temperature which affects the physical conditions of workers but has no direct influence on the structure's quality and risk is not considered as a reason to stop the construction.

Chemical Reaction Hazards of Organic Solvents (1st Report) --Thermal Stability of Epichlorohydrin-Dimethysulfoxide Mixture--

RR-95-7

Takayuki ANDO

: As most of organic solvents used commonly in the chemical industry are inflammable, the potential hazards of the solvents are usually discussed on fire or explosion hazard. However, some solvents show thermal hazards such as exothermic decomposition and/or exothermic reaction with the solute, which are found, for example, in distillation processes.
    Distillation is usually regarded as 'physical' process and is not considered so hazardous as other 'chemical' processes such as a reaction process. However, as heat is supplied continuously to distilland for compensating the heat of evaporation, distillation process is inherently with potential hazard which leads to runaway of exothermic reaction. The hazard is especially feasible when the distillating vessel was designed for vacuum processing, because the pressure in the vessel is not supposed to get far above a atmospheric pressure.
    In this paper, the exothermic properties of the mixture of epichlorohyrin (ECH) with dimethyl sulfoxide (DMSO) were measured by use of DSC(Differential Scanning Calorimeter) and ARC (Acceleration Rate Calorimeter), and the reaction products were analyzed with DSC-GC-MS (Gaschromatograph-Mass Spectrometer) so as to investigate the hazards of the vacuum distillation process of ECH often found in industry.
    The results are summarized as follows:
  (1) The thermal hazard of a mixture of ECH with DMSO is higher than either of ECH or DMSO.
  (2) The exothermic onset-temperature for a mixtures of ECH with DMSO does not depend so much on the content of ECH in the mixture.
  (3) The first exothermic peak in a DSC curve for a mixture of ECH with DMSO is mainly attributed to the formation of mono- and di- cholorohydrins.
  (4) The second exothermic peak in a DSC curve for a mixture of ECH with DMSO is mainly attributed to the decomposition of mono- and di- cholorohydrins formed in the reaction corresponding to the first exothermic peak.

Influence of Stirring on Exothermic Rate of Liquid-liquid Reactions

RR-95-8

Yasuhiro FUJIMOTO

: This paper reports the effect of mode and rate of stirring on the thermal behaviors of exothermic chemical reaction. Nitration of benzene is tried first, which is well-known as a typical reaction capable of exothermic runaway. Neutralization reaction is also studied to compare with nitration. A reaction calorimeter RC1 is used to evaluate reaction heat and temperature rise under various operating conditions of stirring rate, cooling capacity and dosing rate.
    In the first series of tests, the equi-weight mixture of nitric acid and sulfuric acid is fed into the reactor supplied with benzene. Tow modes of stirring are applied: in mode (1), 120 rpm throughout feeding, and in mode (2), 30 rpm while dosing, then increased to 120 rpm. In the later mode, reactants arc expected to be separated into two phases before rating up of stirring. No test is tried on dosing benzene into acid mixture because of possible danger.
    Heat flow-time profiles are almost same for two modes of stirring. In the later mode, no significant heat generation is detected after increasing the rate, because the substantial reactants may be almost consumed before rating up of stirring.
    In the second series of tests, neutralization of aqueous solution of NaOH and acetic acid in benzene was tested under the same modes of stirring as nitration. In these tests, benzene was used as diluent of acetic acid, so as to realize two separated layers of reactants before rating of stirring.
    When NaOH is added into benzene solution of acetic acid in the reactor, heat flow profiles are same, as in nitration, for two modes of stirring. On the contrary, heat flow profiles are quite different depending on the stirring mode, if benzene solution of acetic acid is fed into NaOH solution: that is, abrupt and high peak of heat generation is found by increasing stirring rate after dosing whole amount of benzene solution of acetic acid. And the rate of heat generation is higher for slower rate of stirring while dosing.
    Those results prove that runaway of exothermic reaction will be possible in a chemical reactor, if the stirring is started or re-started after dosing reactants which may form two or more separated layers in a reactor depending on the difference of specific gravity of each reactant.

Thermal Reactivity of Neodymium-Iron alloy

RR-95-9

Teruhito OHTSUKA and Toshihiro HAYASHI

: Neodymium-Iron (Nd-Fe) alloys have been widely used in advanced industries as source materials for magnetic devices. Because Nd is a typical one of the rare earth metals with high chemical reactivity, Nd-Fe alloys are suspected to have potential hazards of unusual reaction, such as exothermic reaction with substances commonly misunderstood as "inert". There have been, in fact, reported accidents which might be attributed to ignition of Nd-Fe alloys.
    This paper describes the reactivity of Nd-Fe alloy of certain composition with such "inert" compounds as nitrogen (N₂), carbon dioxide (CO₂) and water (H₂O). Thermal analysis is tried first for Nd, Fe and Nd-Fe powders, by use of a calorimeter (Setaram C80D). Spontaneous ignition tests are then carried out for several grams of powders in an enclosed test vessel with substantial volume. Effect of CO₂ on inerting dust explosions of Nd-Fe alloy powder is also studied.
    Main results obtained from thermal analysis were;
  1) Pure Nd had exothermic peaks of oxidation around 300 °C in air as well as in oxygen (O₂), yet no reactions were suggested in N₂, CO₂ and, of course, in argon(Ar) atmosphere.
  2) For Nd-water mixture, the exothermic peak temperature was remarkably lowered to 100 °C in air, CO₂ and even in inert Ar. In N₂, much higher peak was observed at 130 °C.
  3) By ESCA analyses on solid products after calorimetric test on Nd-water mixtures, only Nd and
oxygen were found in products generated not only in N₂ and CO₂ but also in Ar atmosphere. Those facts suggested an exothermic oxidation between Nd and O₂ extracted from water.
  4) Nd-Fe alloy showed similar behaviours, but the exothermic rates were relatively small.
    Spontaneous ignition tests for Nd-Fe alloy powder gave following results;
  5) In the atmospheric air, non-flame burning (red-hot oxidation) was observed when the powders were heated above 170 °C. No reaction was observed in N₂ and CO₂ for temperatures up to 350 °C.
  6) When alloy powders mixed with water were heated, only found was the evaporation of water. On the contrary, several drops of water dosed onto alloy powders heated above 290 °C triggered a red-hot oxidation even in Ar. The increase of mass in solid phase was converted to the mass of oxygen, then compared to that of hydrogen generated in the gas phase, so as to find the ratio of both substances quite equal to the composition of water. The fact clearly proved the validity of an assumption made in item 3) above. Those results implied that Nd-Fe powders should not be brought into contact with water, especially when they are heated. Dust explosions of Nd-Fe alloy powder could be suppressed when more than 50% of atmospheric air was substituted by CO₂ .

Suppression of Incendiary Electrostatic Discharges between Charged Liquid and Grounded Conductor

RR-95-10

Tsutomu KODAMA and Yasuyuki TABATA

: In a flammable liquid storage tank, equipment that protrude from the top mounting wall such as cleaning nozzles or high level probes may encourage the initiation of incendiary discharge to the approaching liquid surface and cause explosion or fire if the liquid being loaded is highly insulating and electrified. To prevent such a dangerous situation from happening, the authors have developed a new type of discharge suppression device consisting a PTFE (Teflon) disc and a metallic rod which can be attached to the equipment. The device was found to work as well as using an earthed wire linking both the equipment and the bottom of the tank. Experiments on the device were conducted using a small-scale plastic tank containing kerosene charged by using a DC corona ionizer. In the experiments charge transfer, peak value of discharge current and electrode distance were measured whenever a spike discharge occurs between the kerosene and a metallic electrode which was initially brought into the kerosene and then pulled vertically up at a constant speed. The surface potential was controlled in the range 10 to 60 kV. For the electrode, either sphere, cone or rod tip-shape electrodes, a metallic column with a wire attached and a high level probe-tip with the safety device attached were used. Discharge incendivity was evaluated by analyzing both charge transfer and peak current.
    The experimental results were as follows:
  (1) In the case of using the sphere, cone and rod electrodes the incendiary discharges generated in a surface potential range between 30-40 kV corresponded to a charge density range between 12-16μC/m³, respectively. The positions of the electrode tips were just above the kerosene surface.
  (2) The safety devices consisting Teflon disks of 60-80 mm in diameter and 20-30 mm in thickness and metallic rods of 16-20 mm in diameter and 35-55 mm in length suppressed incendiary discharges up to a surface potential of 50 kV which corresponded to a charge density of 20μC/m³ .

Influence of Ozone on Insulation Deterioration of Wiring Board Composed of Organic Materials

RR-95-11

Tatsuo MOTOYAMA and Kenji ICHIKAWA

: Ozone generated by switching discharge reacts with wiring board surface composed of organic materials and changes their hydrophobic characteristic to hydrophilic. So, the insulation between the tracks becomes deteriorated.
    This paper describes the insulation deterioration on wiring board surfaces composed of organic materials caused by ozone.
    Two kinds of wiring boards were tested. One was a glass-reinforced epoxy resin board with copper tracks and the other a phenolic resin board with filler and copper tracks. They were exposed to ozone at levels between 0.5 - 1.5 ppm, 10 - 17 ppm and 45 - 55 ppm, for 24, 120 and 240 hours and at a constant temperature of 20 °C. The contact angles of water of the wiring boards were measured in ambient condition and their surface resistances measured in a vessel at 65, 80 and 95% RH. The temperature was kept constant at 40 °C.
    The experimental results are summarized as follows:
  1) Contact angle of water (θ), concentration of ozone (C) and exposure time (H) were related by the following equation, provided that D is greater than zero.
        θ nearly equal - D·log(C × H)
  2) The surface resistances between the tracks decrease with increasing ozone concentration and exposure time. In the case of the glass-reinforced epoxy resin board, which was exposed to ozone at 45 - 55 ppm for 120 hours, its value of surface resistance measured at 40 °C/95% RH was approximately 100 times less than the value measured at 0 ppm ozone concentration.
  3) Surface resistance (R_s) between the tracks of the glass-reinforced epoxy resin board was approximated by the following equation, provided that F is greater than zero.
        R_s nealy equal F/(C × H)

JNIOSH