Abstract of Special Research Report (RR-96)

National Institute of Occupational Safety and Health, Japan

A Study on the Compliant Actuators by Means of Electro-Rheological Fluids (ERF)

RR-96-1

Noboru SUGIMOTO

: As an effective safety measure, many sets of industrial robots and manipulators have been introduced in place of workers who have been suffering from dangerous or detrimental work. Because of their powerful dangerous action, however, industrial robots should be guarded in a remote work area. Recently, computers and other micro-electronics which have been progressing rapid strides, accelerate the progress in automatic control technology. Robots that take care of patients, physically handicapped persons and very old persons will soon be realized, and many people will work together with robots in factories, offices and houses. But, the most important thing must be safe when a robot and human are in contact each other.
   This paper considers the problem of actuators for safe conduct of robot in performing cooperation work with human. Compliant manipulation fundamentally requires that controlled robot can not only follow up the input trajectory exactly in free motion space, but also move adaptively in constrained space while making compliant contact with its dynamic environment. However it is difficult to achieve compliant actuator by means of conventional linear control means.
   Electro-rheological phenomena is defined as an essentially instantaneous reversible change in apparent viscosity when fluids are subjected to an externally applied electric field. Drive clutches and hydraulic valves for this Electro-Rheological Fluids (ERF) can control the torque and the pressure respectively only by application of high voltage.
   This study evaluates the non-linearity and passive function of an ER element from a standpoint of control and pursuits motion control strategies suitable for non-linearity of ER elements. An antagonistic rotary actuator system using ERF is proposed and various control strategies are attempted. The system is so structured that output cylinder is set in drums composed of a positive and a negative rotation drum driven by the electric motor and the ERF is filled in a gap. As a result of experiments, proposed control method realizes both force-adaptation (compliance control) and a fine positioning (stiffness) with changing a shear stress of ERF. This ability to regulate compliance/stiffness by means of such a property as changeable plasticity is essential for a safe robot to operate in cooperation with humans.

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

RR-96-2

Yoshimi SUZUKI

: In order to investigate information structure of descriptive statement in the "Occupational accident report" concerning construction work, 2131 accident cases are selected and their descriptive statements of the "Accident Situation and Circumstances" are examined by statistical quantification analytical methods. In this paper, statistical quantification analysis is made according to four types of accident cases; 1 153 cases of falling accidents, 183 soil collapse accidents, 271 crane accidents and 524 construction machinery accidents, considering various factors in construction work sites.
    Procedures in this study are summarized as follows.
  1) All free-terms are extracted from descriptive statements appeared in the item "Accident Situation and Circumstances" by means of the utility program in the "Occupational Accident Report Database System" developed in the National Institute of Industrial Safety (NIIS).
  2) The frequency distribution of these extracted free-terms are examined.
  3) High frequency free-terms are selected as key-words for arranging or characterizing the information concerning various factors of accidents in construction work sites. In this step, the relationship between these key-words and accident cases is investigated by the statistical quantification analytical method (quantification method of "third type").
  4) Investigation by cluster analysis on extracted free-terms is made from viewpoint of the relationship between these free-terms and some factors in construction work sites such as type of accident or kind of construction work etc.
  5) Using these free-terms, the possibility for exploration of accident cases concerning various factors such as kinds of construction work, kind of task of worker and so on, are also investigated by discriminant analysis.
     Main results of these investigations are as follows;
  1) From the descriptive statements item of the "Accident Situation and Circumstances" of 271 occupational accidents reports concerning cranes, 3,745 kinds of free-term in total are extracted. As regards other types of accidents, 6,151 kinds of free-term from 524 construction machinery accidents, 10,408 kinds of free-terms from 1,153 falling accidents, and 2,381 kinds of free-terms from 183 soil collapse accidents, are similarly extracted.
  2) The frequency of these extracted free-terms are distributed according to the Bradford's law or Zipf's law. There is no different tendency concerning each types of accidents.
  3) As a result of examination on actual frequencies of these free-terms, taking "accident type factor" into consideration, the existence of some unevenly distributed free-terms are estimated.
  4) As a result of the statistical quantification analytical method, scatter diagrams of accident cases(sample score diagrams) for each accident type are obtained. These diagrams indicate that the information about "Accident Situation and Circumstances" can be roughly arranged according to kind of construction work factor.
  5) Also, as a result of cluster analysis of free-terms, good relationship is obtained in dendrograms between these free-terms and kind of construction work factor.
  6) By discriminant analysis, using high-frequency free-terms which include about 3% or 5% ratio of cumulative frequency rank of free-terms, classification results tables about accident cases are obtained, and as regards of each kind of construction work factor, accident cases can be classified with about 80 or 90% estimation ratio.

Development of Numerical Simulation Method for the Ground Flow Caused by Seismic Liquefaction

RR-96-3

Satoshi TAMATE

: The ground flow induced by seismic liquefaction is known to cause significant damage to lifeline networks and structures. The ground flow is not only caused by the dynamic forces induced by seismic acceleration but also by the static gravity force induced by the topography of the ground. Ground flow may sometimes continue after the seismic loading and finally the ground is significantly deformed and fails. Although the ground surface may keep its shape after an earthquake, differential movement inside tends to occur. However, this phenomenon has been researched only recently, and its mechanism is not well understood.
There have not been many reports of disasters induced by liquefaction in construction sites. However, as the construction works in such a deep excavation generally take a long period, temporary structures in those sites may be affected by the ground flow induced by seismic liquefaction. Therefore, both effects and safety against earthquake must be taken into account. A numerical simulation method which can predict the magnitude and direction of ground movement with reasonable accuracy is needed to prevent disasters.
This paper investigates by finite element simulation the ground flow that could occur when soil liquefaction continues for a sufficiently long period. It is considered that liquefied soil behaves like a viscous liquid, and hence, ground flow is governed by the minimum energy principle. In the calculation, liquefied sand is assumed to be a viscous liquid that deforms in the undrained condition and whose volume remains constant. To consider the geometric non-linearity due to the large deformations caused by ground flow, the updated Lagrangian method is used to calculate the equation of motion. To calculate the time history of the ground motion, the Newmark method is used. Finally, a simulation using this calculation method shows that the proposed method gives reasonable results for the conditions indicated.

SADT Values of Solid Organic Chemicals of the Pseud-autocatalytic Decomposition Type

RR-96-4

Takashi KOTOYORI

: Chemicals which are solid at the normal temperature are classified into the following three groups according to the difference in their self-heating behaviour;
    Group I. Solids to which the Frank-Kamenetskii's equation is applicable to calculate their critical temperatures for the thermal explosion, T _C values. When exposed to the surroundings maintained at a temperature equivalent to the starting temperature T _s, a solid chemical of this kind warms up to the T _s within one hour. It then continues to self-heat as a solid with a rate varying as a function of the ambient temperature in accordance with the Arrhenius equation;
    Group II. Solids to which the Semenov's equation is applicable to calculate their T _C values. When exposed to the surroundings maintained at a temperature equivalent to T _s, a solid chemical of this kind fuses at once (if the Ts is above the melting point of the chemical). The resultant liquid then warms up to the Ts within one hour, then continues to self-heat with a rate varying as a function of the ambient temperature in accordance with the Arrhenius equation;
    Group III. Solids to which neither of the two equations of the thermal explosion theory is applicable to calculate their T _C values. When exposed to the surroundings maintained at a temperature equivalent to T _s, a solid chemical of this kind warms slowly up to the T _s, remains at the T _s over a long period of time, in due time begins to fuse little by little, and commences a rapid self-heating, once the fusion has been accomplished.
    Solids in Groups I and II are referred to as the thermal decomposition or TD type. Both groups have the same characteristic that their decompositions are not accompanied by any phase changes, so that their rates of thermal decomposition, which are observed as adiabatic temperature rises of chemicals, vary as functions of ambient temperature in accordance with the Arrhenius equation. Therefore, the application of either equation used in the thermal explosion theory becomes possible.
    On the other hand, solids in Group III are called the pseud-autocatalytic decomposition or pseud-AC type, because their self-heating behaviours resemble closely those of chemicals of the true AC type. The characteristic of chemicals of this type is that their decomposions are accompanied by a phase change, fusion. That is, a molecule of solid chemicals of this type begins to decompose exothermically once it is released from the fixation composed by the surrounding crystal structure by fusion. Therefore, their apparent rates of thermal decomposition, which is observed as an adiabatic temperature rise of the system, is not in accordance with the Arrhenius equation. Accordingly, the T _c of solid chemicals of the pseud-AC type is not calculable by applying either equation of the theory.
    Instead, it is known that, when a solid chemical of this type is exposed to surroundings maintained at a temperature equivalent to T s, an empirical formula, ln Δt = a /T _s + b, holds between the the elapsed time, Δt, from the start of the exposure till the commencement of the self-heating of the chemical and T _s, where a and b are constants.
    In this work, the above empirical formula was established for chemicals of the pseud-AC type, using an isothermal storage testing device equipped with an aluminum block. The accuracy of temperature control of the block is ± 0.05 - 0.08 K. A glass closed cell, the reference cell, containing a definite quantity of an inactive reference material, silica (silicon dioxide) powder, is first placed on the bottom of a hole bored in an aluminium block maintained at a temperature T _s. Then, the T _s value of each run is determined as an arithmetic mean of digital outputs of the T _s recorded for two days. After the reference cell has been removed, another closed glass cell, the sample cell, in which a known quantity of sample is contained, is inserted into the hole, while the aluminium block is maintained at the same temperature, T _s. Then, the time, Δt, is measured. The measurement is automatically interrupted once the sample temperature has risen by 0.3K above the T _s to prevent the cell exploding.
    The Self-Accelerating Decomposition Temperature or SADT is then calculated as a T _s value corresponding to a Δt value of 7 days, based on the empirical formula established above.
   SADT values thus calculated for five solid chemicals of the pseud-AC type showed, in general, rather good agreement with a few corresponding reported values. However, these SADT values are in most cases 20 - 30K lower than the corresponding temperature values such as the decomposition temperature, the melting points listed in some catalogs of fine chemicals, or the exothermic onset temperatures measured by thermal analysis. This fact indicates that the latter temperature values are not adequate as criteria for the thermal instability of, at least, solid chemicals of the pseud-AC type.

Influence of Stirring on Exothermic Rate of Liquid-liquid Reactions (2) --Polymerization of Epichlorohydrin--

RR-96-5

Yasuhiro FUJIMOTO

: In the previous paper, nitration of benzene and neutralization were reported with reference to the effect of mode and rate of stirring on the thermal behaviors of exothermic chemical reaction. A reaction calorimeter RC1 was used to evaluate reaction heat and temperature rise under different operating conditions of stirring rate, cooling capacity and dosing rate.
   This paper reports influence of stirring rate on the exothermic rate of epichlorohydrin polymerization using the reaction calorimeter RC1.
   The polymerization runs under the condition that 60% hydrogen peroxychloride aqueous solution (60% HClO₄) is fed into the reactor supplied with the mixture of epichlorohydrin and water either with isothermal mode or with adiabatic mode both with several stirring rates (from 40 rpm to 120 rpm).
    In case of isothermal mode (reaction temperature control mode), as the stirring rate becomes slower, the heat generation while dosing HClO₄ decreases. And the heat generation after the dosing has the maximum value for 60 rpm of stirring rate. Higher concentration of epichlorohydrin is detected in the reactor for 60 rpm of stirring rate than for 120 rpm.
    When the control mode is switched into adiabatic mode after dosing HClO₄, for 120 rpm of stirring rate, reaction temperature rises up to 95°:C within 20 minutes after dosing. Yet the exothermic reaction can be controlled safely by changing the control mode from adiabatic mode to reaction temperature control mode again when the temperature goes up to over 95°:C. For 60 rpm of stirring rate in the same procedure, reaction temperature rises up to 95 °:C within 15 minutes after dosing, and the reactants cannot be cooled down by changing the temperature control mode. In spite of changing the control mode, the reaction temperature goes up to over 100°:C, The reaction is quenched only by adding 300g of water.

Effects of Electromagnetic Noises on Electronic Circuits and Electrostatic Protection of Semiconductor Devices

RR-96-6

Hajime TOMITA and Yasuyuki TABATA

:Electronic devices and equipment are prone to be malfunctioned by electromagnetic noises emitted by nearby industrial equipment, such as inverters and weldering machines. Such equipment are also affected by electric over stress and electrostatic discharge (EOS/ESD). EOS/ESD become the cause of semiconductor failure. Conductive materials are commonly used for shielding semiconductor devices from EOS/ESD.
In this paper, a fundamental study was conducted on electromagnetic noise in industrial environments, malfunction of digital circuits and the relationships between electrical properties of conductive materials and their field screening capabilities. In an industrial environment, motors and weldering machines are the main sources of electromagnetic noise. To investigate the malfunction of a digital circuit, such was tested by injecting a pseudo-noise onto a signal line of the circuit.
Malfunction was found to occur when the signal level changed from high to low or vice versa as a result of the composite effect of signal and electromagnetic noise. Considering from IEC standard, the noise level at the malfunction of a digital circuit was found to be lower than the severity level requested in industrial environment. With regards to conductive materials, which are used for the protection of semiconductor devices from EOS/ESD, their electric field shielding effectivenesses were investigated based on the standardized method of test, IS-5-A, and calculated using an equivalent circuit of a human body and conductive material. The calculated results nearly agreed with the experiments. Electric field shielding effectiveness was also calculated using the simple electrostatic discharge model. The electric field generated by the discharge was assumed to be the summation of plain waves. The comparison between the computational result obtained in this study and experimental results revealed that volume resistivity less than 1Ω·m is necessary to protect a semiconductor device from being affected by EOS/ESD. The computational method for assessing the electric field shielding effectiveness is applicable to designing conductive materials for electrostatic protection.

Electromagnetic Susceptibility of Photo-electric Safety Device for Power Press to Conductive Electromagnetic Disturbance

RR-96-7

Hajime TOMITA

: Recently, electronic equipment, such as industrial robot, is commonly used for enhancement of productivity and reduction of labour. Recent safety devices, such as photo-electric safety devices, have also ICs within. Electronic equipment is susceptible to electromagnetic disturbances because signal level is several volts. To protect electronic equipment against electromagnetic disturbances, many countermeasures are taken, such as ferrite core, line filter, noise cutout transformer and surge absorber. Standard methods for evaluating the susceptibility level of electronic equipment to electromagnetic disturbances are proposed by International Electrotechnical Commission(IEC) and International Special Committee on Radio Interference(CISPR). But electromagnetic susceptibility level on photo-electronic safety device for power press in Japan is not well understood.
    In this report, susceptibility level of seven photo-electric safety devices used for power press to conductive electromagnetic disturbances was measured in accordance with methods defined in IEC 1000-4-4, MIL-STD-462D and JIS C 1003. Some countermeasures against electromagnetic disturbances were evaluated by applying to photo-electric safety devices.
    The principal results are as follows:
  (1) Malfunction of a photo-electric safety device occurred at 1300V by the susceptibility test based on IEC 1000-4-4. The susceptibility level was lower than that in guideline which requires 2 kV in industrial environment.
  (2) All safety equipment tested worked normally when the sinusoidal current whose frequency is less than 5 MHz was injected into a power line. Susceptibility level ranged from 60 dBμA to 110 dBμA by virtual value.
  (3) Noise suppression devices could enhance the susceptibility level of photo-electric safety devices up to 2 kV by the susceptibility test based on JIS C 1003. The result also shows that the malfunction of photo-electric safety devices is mainly due to the noise whose frequency is more than several MHz.

JNIOSH