Hitoshi Akami

Robot hand with sensor for handling cloth

A robot hand to handle cloth is presented. In the study carried out, the authors tried to pick up a piece of cloth from a stack of cloth, to judge its thickness and to control the tension of it by the robot hand with a strain gauge sensor. The robot hand has two fingers with two degrees of freedom. One fingers tip was made from balsa wood, the other was made from balsa wood and phosphorus bronze plate with strain gauge sensor. The robot hand was attached to a robot arm and picked up a piece of cloth in cooperative motion with the robot arm. The hand was able to pick up a piece of wool cloth 0.4 mm in thickness. And it was able to pick up a piece of rayon cloth 0.2 mm in thickness in some trials by an automatic adjusting software. Sensor output was nearly in proportion to the thickness of cloth picked up. And sensor output was nearly in proportion to the tension of the cloth too.<<ETX>>

Designing a Prototype of a Computerized Patterning System on Jacquard Weaving Process

In order to modernize patterning works of figured cloth by jacquard loom, a prototype of DDC weaving system for figured cloth has been developed. The system is consisted of the following equipments.1. Pattern Analyser, 2 . Paper Tape Reader 3 . Mini Computer, 4 . Actuator for Selecting Needles of a Jacquard Machine, 5 . Jacquard LoomThe signal processing and weaving in the system are performed as follows : (1) Colors and patterns in a sketch design are recognized by the pattern analyser, conver-ted into color pattern signal at run length and then punched out on a paper tape.(2) On weaving, color pattern signals read out from the sketch design are read into a mini computer and are converted into pattern weavingsignals, corresponding to on-off bit pattern of the pattern card, by a stored program.(3) The signals are sent to the actuators by a request signal from the jacquard loom and needles of a jacquard machine are selected by the signal synchronously with cycle of the loom motion.Thus, a figured cloth is automatically woven according to a sketch design. Since highly skil-lful craftmens techniques for production of pattern cards can be stored as a program, it is easy to weave a highly complicated figured cloth only by changing a color pattern tape andby resetting a design of fabric. The performance of the prototype shows that it seems possible to realize labor-saving and speed-up of the patterning work of jacquard weaving process requiring many workers.

MATRIX REPRESENTATION OF BLOWING PROCESS IN COTTON SPINNING SYSTEM

In this paper calculation of size distribution (by weight) of fiber stocks in Blowing Process is discussed by matrix method*. By using histgram or cumultive frequency, it is easy to decide inqut and output size distribution of the process by measuring the weight of fiber stocks.When the size distribution of input and output are represented by vector form, an element of the vector might be given by the ordinate (either as percentage or fraction) at which a size (abscissa) intercepts the smooth curve drawn from the size analysis, that is, these vector elements are culculated from smooth curve of the histograms or cumultive frequencies by substituting an arbitrary unit to the abscissa.Then, input vector F and output vector F′ can be explained as fallows, where it is suitably chosen enough to show each distributions. When it is considered that F is changed to F′ by the process operation, it is reasonable to introduce a transfer matrix B, which characterizes the process; F′=BF Taking a probability of vector undergone the process action, F must be manipulated by the probability π;F′=B(πF)+(I-π)F Then open circuit breakage system can be stated by the above equation.Since transfer matrix B is introduced to show the state of transition of vector components to each size after a cycle of breakage, column of B must give a discrete probability function, but is so difficult to decide the function either theoretically or experimentally that Poison distribution is supposed.Then an element of B, aij show the probability of transition of fi from size i to j after the breakage. where it is difficult for more than two stocks to combine; then B becomes triangular matrix.Elements remained are derived from Poisen distribution table or the equations; by using two conditions, that is, parameter K is decided from aij=0 i=j x=n-i i=jEexperiments were done between Hopper Opener and Hopper Feeder with Overflow, and Hooper Feeder and Lattice Feeder.Cumultive frequency curves were completed using more than 500 data, which were smoothed by Least Square Mean Method in order to decide veotors easily.To find the influence of vectors dimensions to the accuracy of culculation by the equation, they were changed and critical weight of fiber stocks, at which the condition of πij becomes πij≤1, were compared with each other.

FUNDAMENTAL STUDIES ON THE AUTOMATIC CONTROL OF SLIVER THICKNESS

Although there are many studies on the automatic control of sliver thickness, contorolable variations are generally of “long period” and the control of “short period” variation is left intact.In case of controlling “short period” variation it becomes impossible to ignore the influence of the fiber length in slivers and improcessing this factor would have important becomings are obtained.in this study, the theoretical process character of roller-drafting is made and the following results:-(1) The differential equation in the unsteady roller-drafting is;(2) When the variation is small, the transfer function between the surface velocity of front roller and produced sliver thickness is;G1(s)=k1/(1+T1s)(3) The transfer function between the thickness of the feed and produced sliver isG2(s)=k2(1-T2s)/(1+T1s)(4) Each quantity in roller-drafting coresponds to the one in automatic control in the following way:-Potential sliver thicknessFlow Fiber flowCapacitance Fiber quantity of taft-diagram per unit sliver thickness.Resistance Reciprocal of the surface velocity of roller(5) The transfer function of front roller is 1/(1+T1s) and of back roller (1-T2s)