[Trade Journal]
Publication: Journal of the American Institute Of Electrical Engineers
New York, NY, United States
p. 470-478, col. 1-2
Modern Production of Suspension Insulators
BY EDWIN H. FRITZ
Ceramic Engineer, Pittsburgh High-Voltage Insulator Co.
AND
GEORGE I. GILCHREST
Westinghouse Electric & Mfg. Co.
This paper pictures the progress made during the past few years in the production of electrical porcelain. The information covers: First: The engineering and works organization. Second: The manufacture. Third: Design and test.
THE DEVELOPMENT of transmission line networks has progressed even more rapidly than most of the pioneers in transmission engineering anticipated. No doubt, the development has been materially affected by the increased cost of fuel, which has encouraged the engineers to develop available water power sites. Perhaps, one of the most influential factors has been the necessity of irrigating the fertile lands of the Pacific Coast States. Electricity from water power sources can be generated and transmitted to the farming districts and can be economically employed to pump water, to heat the houses, and to operate the household appliances.
The quality of electrical porcelain, although sufficient to properly insulate low-tension lines, proved entirely unsatisfactory when it became necessary to increase the transmission line voltages. The first thought of the engineer was to emphasize the mechanical strength, but in so doing, he sacrificed other characteristics. He believed the insulators should withstand handling, impact blows from rifle shooting, etc. He believed suspension insulators must be mechanically strong to prevent dropping the line in service.
There are numerous reasons why the quality of electrical porcelain did not keep pace with the progress in transmission engineering.
First: There were few transmission networks when electrical porcelain was first applied to high voltages.
Second: The demand for electrical porcelain was extremely variable. When a transmission project was under consideration, it was necessary to supply a large number of insulators in a short time. After the material for this project was manufactured and supplied, the manufacturer could not keep a continuous production of high-tension electrical porcelain in his factory. He must again manufacture low-voltage insulators and dry process material such as knobs, tubes, cleats, etc.
Third: The type of labor from which the manufacturer drew his supply had never been trained to appreciate the advantages that can be derived from labor saving devices. They had been very adverse to accepting any new type of machinery, perhaps, more so than the usual workmen.
The most regrettable and fundamental reason of the slow development of electrical porcelain has been the lack of cooperation between the ceramic and electrical engineers. The ceramic engineer was occupied with manufacturing problems only. The electrical engineer, on the other hand, was usually a consulting engineer because the transmission company itself could not afford to assign an electrical engineer to the application of electrical porcelain. The design recommended by him would incorporate his ideas and opinions. He would probably not consult with the ceramic engineer to determine the limitations in manufacture. The varied line of designs which every insulator company carried a few years ago is indicative of this condition.
ENGINEERING ORGANIZATION
The increased demand for electrical porcelain has enabled the manufacturer to command men of greater engineering ability. These men have sufficient training to appreciate the advantages that are gained by close contact between the manufacturer, the engineer, and the consumer. The organization of the modern electrical porcelain manufacturer is built on close cooperation between engineers who have supervision of the works and the application of the finished product.
The ceramic research laboratory operated as a separate department, is of vital importance. Its function is to investigate the present commercial raw materials; new sources of supply; the proportioning of ingredients; the glaze, etc. This laboratory has complete equipment, a portion of which constitutes a miniature clay working plant. This enables the engineer in charge of the laboratory to produce experimental bodies and determine such properties as dielectric strength, mechanical strength, both tensile and impact, resistance to temperature changes, firing range, and shrinkage. Fig. 1 will give a general idea of the completeness of this laboratory.
In connection with this, the proper selection of materials is of extreme importance. For example, we find different grades of clays in the same deposit, some of which will give considerably better results than others which may be only a few feet removed from them in the same strata. The ball clays which are now used contain a considerable amount of lignite and organic matter. This type of clay vitrifies at a much lower temperature than some of the cleaner ball clays and at the same time is a very tough and strong material, such as is necessary to overcome the severe stresses which are encountered in manufacturing the complex insulator shapes. The low vitrification temperature is especially desirable because it assists the feldspar in the vitrification of the product and insures a greater firing range with less danger of underfired or over-fired material.
English ball clay has always satisfied these characteristics in the past to the greatest possible extent. Laboratory investigations in recent years have discovered some promising American clays and at the present time experiments are being made on some clays in an undeveloped field in this country which seem to be even superior to the English clays. The laboratory is, therefore, continually investigating new materials and with the continual progress that has been made, it would not be too optimistic to predict that we may at some time find materials which will improve the present product. The proportioning of feldspar, flint and clay has been given very careful consideration in the research laboratories. The present commercial body is based on the results of this work. It is possible to obtain porcelain bodies from these materials in which certain characteristics predominate. However, since the application of electrical porcelain is very wide, it is necessary to sacrifice high mechanical strength or very high di-electric strength in order to obtain a body which will be satisfactory in all applications. (1)
A second ceramic laboratory is maintained at the works; the function of this department being to properly apply information obtained in the research laboratory to the commercial product and to see that all departments of the works function properly, i. e., that the proportioning and mixing of the ingredients is uniformly performed; that the material is properly prepared for the manufacturing processes; that the speed of the machine in forming and trimming the ware is correct; that the firing conditions are uniform,, etc. There is very close contact, at all times between the two departments and by the continual exchange of ideas each is informed of the other's activities.
One of the most important activities of the works department is the testing of all raw materials as they are received, in order to insure uniformity of the material and the finished product. Feldspar has become so variable in recent years that the old method of assuming that the feldspar will always be uniform in quality is now a hazard. Each shipment must be tested especially in regard to fusibility, for it is this property which has become most variable. A sample is obtained from the car as soon as it is received; the flux value established and if not in accordance with purchasing specifications, it is rejected before coming in contact with any of the material in the bins. The remaining materials are tested in a similar manner. Other routine tests are the determination of the moisture content and non-clay substance in the clays from day to day so that this variable can be adjusted for on the scales and consequent uniform mixture produced.
The electrical engineering department functions in a similar manner. The consulting staff of the company is available when problems arise covering the design and application of the finished product. The Engineering Department at the works supervise routine testing and the application to the product of suggestions that are made by the consulting staff. It is also a function of the Engineering Department at the works to give careful thought to any suggestions from the field which are based on the inherent design of the product.
MANUFACTURE
The modern electrical porcelain works are the result of the changed conditions and indicate what closer cooperation and a more scientific organization have accomplished. This is most forcibly illustrated in the slip house where the materials are mixed. This department was formerly found in the most dilapidated part of the works and gave an appearance which was repulsive to the intelligent class of workmen. In other words, it was located and operated under conditions which indicated that it was an unimportant department.
Today we find the reverse of these conditions. This department is now considered one of the most important if not the most important, part of the works and the general design and operating conditions are carried out accordingly. It contains the latest type of machinery, with special attention given to labor saving devices in order to make the processes more mechanical and eliminate the human factor wherever possible, with resultant greater uniformity. The department is well lighted and presents a sanitary appearance which is in keeping with the kind of work which is carried on.
The raw materials upon arrival after tests have been completed are unloaded in an efficient manner by means of suitable conveying equipment and stored in large bins to prevent contamination with undesirable substances which are carried about in the air in industrial centers. (Figs. 2 and 3.) These bins are directly adjacent to the mixing department and the same conveying equipment can be used to bring the material into a location where it can be conveniently used in the process. (Fig. 4.) The materials are weighed with the minimum amount of manual labor and in such a way that the workmen cannot easily make an error of any consequence. The apparatus is built compactly so that the little transportation which is necessary is performed mechanically. (Figs. 5 and 6.)
The feldspar and flint which were formerly mixed together with the clays in the blungers are now first ground in ball mills in the wet state sufficiently to produce a fineness which has been found to be necessary and which cannot be obtained in the dry state which is the means employed by the producers. (Fig. 7). The cost is large, but the results that are obtained more than repay the manufacturer for this extra operation. The finer structure of the feldspar and flint gives a better mechanical mixture which makes these materials more active in their properties and effects a better performance of the body throughout the manufacturing process, especially in the firing operation. Vitrification begins at a lower temperature and a longer firing range is, therefore, obtained, or in other words, greater variation of temperature in the kilns is permissible without detrimental effect on the fired product. It is obvious that the danger of underfired ware is, therefore, materially lessened and on account of the improved methods of kiln firing, an underfired piece of electrical porcelain in the factory is indeed rare.
Porcelain made from ball milled feldspar and flint is more homogeneous in structure and gives a very smooth fracture. It has a higher dielectric and mechanical strength, although in the case of the latter there is a limit to the fineness which will increase the value of this property. In general it has been very definitely proved that this operation is one of the greatest progressive steps in recent manufacturing developments.
While these materials are being milled the clays are mixed in blungers. The design of the blunger has been changed in recent years and machines are now on the market which are much more efficient and rapid in their performance. Complete slaking of the clays is assured in this type of mill, so that lumps of unmixed clay are not found in the blunger when it is discharged, with consequent variation of the mixture. This has been accomplished by means of double rotating mixers which prevent any centrifugal force action and consequent collection of the particles at the edge of the tank which is so commonly found in the old type of blungers. Fig. 8). The feldspar and flint after being milled are added to this blunged clay slip and the entire mixture is then blunged sufficiently to produce as perfect a mechanical mixture as is possible.
The amount of water which is added to the clay flint and feldspar in the blungers and ball mills is carefully controlled by means of self-regulating water tanks above these machines. The amount of water in each tank is set from day to day in order to make proper adjustment of the water contained in the materials themselves. At the same time the water is heated to the proper temperature and kept constant by means of automatic temperature regulators. In this way the liquid in the ball mills and the blungers is always of the same density and temperature which is essential in pumping uniform filter cakes and in keeping the plastic clay body at the proper temperature. (Fig. 9).
The clay slip is passed over a double set of lawns, the first lawn being of a coarser mesh than the second. This distributes the amount of residue on each lawn and lessens the danger of breaking the lawn and con-sequent passing through of coarse materials. After screening, the material accumulates in cisterns, which are now built of considerable size and number. The object in view is to provide storage for the clay slip after it has completely passed through the mixing process where the air which is contained can be eliminated, giving a more homogeneous clay slip.
The pumps which force the material from the cisterns to the filter presses are designed to preserve the uniformity of the clay slip. The pumping action of the piston is now transmitted upon a diaphragm in order to isolate the clay slip from the pumping action. Mixing of air in the slip at this point is, therefore, impossible and at the same time it does not get in contact with the oil in the cylinders. If the air is removed in the cisterns the filter cakes which will be produced at the press will be solid and contain no blebs of any kind. Furthermore, the pumps will not produce a higher pressure than what they are set for, so that the filter cakes are always uniform in water content and working qualities.
With such filter cakes the beneficial effect of aging is largely reduced and it has actually come to a point where equal results can be obtained from the manufacture of insulators from clay directly from the filter presses. Furthermore, ball clays have already passed through considerable natural aging and weathering conditions, and since they compose the major part of the clay content, the body is not noticeably improved by a small amount of artificial aging. For this reason, aging would have to extend over a period of at least one month before any noticeable results would be obtained in the product and this, of course, necessitates a storage capacity which is not commercial in this country.
In pugmilling the material in preparation for the moulding process, the more modern and scientific methods of mixing have overcome many of the troubles which formerly were traced to the pugmill. It is now much simpler to produce a satisfactory material from the pugmill and with the improvements in this machine which have been obtained, the ever-present trouble with laminations has been largely overcome. If the material leaves the pugmill free of laminations and with no air contained, the success of the moulding of the insulators is practically assured.
In forming the insulators machine operations are used entirely. This method has proved to be superior to the jiggering process, because of the greater pressure which is obtainable and the elimination of the human factor with the consequent possibility of a non-uniform product. A greater production can be obtained from the machines with an improvement in quality. (Fig. 10.)
By means of modern conditioning dryers, the insulators can be rapidly and uniformly dried to the stage where they can be removed from the moulds without distortion. (Fig. 11.) They are then ready for trimming which perfects the insulators into their final form. In drying the product to the bone-dry stage the modern tunnel dryer has given the manufacturer a means of drying under carefully controlled conditions as to temperature and humidity. This is of primary importance in drying the material with the minimum amount of strain and at the same time provides a method which is labor saving. A truck carries the insulators through the dryers and they are then taken directly to the glazing department where the glaze and sand coatings are applied. The color of the glaze has proved of some assistance in indicating the degree of heat treatment received in the kilns. This method, however, has its limitations and is not completely satisfactory.
Recent experimental work in the laboratory has shown that there is a possibility of having the glaze indicate firing treatment by means of its luster. If this glaze can be made satisfactory to the trade, there is no doubt that it will provide a very accurate means of determining the degree of heat the insulator has received in the kiln.
In firing the material, various steps are taken to insure a uniform product such as the location of numerous cone plaques in all parts of the kilns and the use of the electric pyrometer with recording chart which is operated according to a standard firing curve in order to standardize the entire firing cycle. The cone plaques are marked according to their position in the kiln, so that upon removal if any show that insufficient heat has been received, the insulators which are immediately adjacent to such a cone plaque can be segregated and refired. At the same time all of the cone plaques from one kiln are gathered together and assembled in a rack which is photographed together with the pyrometer chart for the kiln. This gives a complete and permanent record of the firing of each kiln and also determines the efficiency of each kiln and of the kiln firemen. (Fig. 12.)
Natural gas, coal and more recently oil are used as fuels. Natural gas is undoubtedly the most desirable fuel because of the ease of control and comparatively low temperatures in the fire boxes. The works are located in the natural gas fields where the supply is abundant. However, protective measures by the gas producers have limited the supply during very severe weather. It, therefore, became necessary to resort to other fuels which has brought about the use of oil. This fuel has proved to be much more satisfactory than coal, because its control can be as easily manipulated as that of gas and the only difficulty encountered is the excessive local heat of the oil flame. This has been overcome by the use of superior refractories so that equal results can be obtained with oil, the only difference being in the cost. Coal is, of course, cheaper than oil as far as cost of fuel itself is concerned, but the ease of control and the more uniform and satisfactory results which are obtained from the oil with consequent better quality of finished product has proved conclusively that the final cost of production with oil is cheaper than coal. Combination oil and gas burners are used so that gas can be used whenever available and the change to oil can be made at any time without loss of heat in the kiln. (Fig. 13.)
ROUTINE TESTING
The porcelain parts are carefully inspected as they are drawn from the kilns. Parts having cracks or other defects that can be detected visually are immediately scrapped. Thereafter the parts are subjected to a 60-cycle flash-over test. The characteristics of the transformer are such as to give a 60-cycle voltage with a superimposed high-frequency voltage. Fig. 14 is indicative of the test.
The insulators having cemented hardware are then assembled with neat Portland cement and placed in a closed chamber where they are subjected to an atmosphere of steam during the initial set of the cement. The insulators are then given a routine mechanical test at a load dependent upon the inherent design of the insulator and its application. After the routine mechanical test the assembled insulators are again subjected to a flash-over test, similar in characteristics to that applied to the porcelain parts. It is not necessary, of course, to give the suspension insulators not assembled with hardware a second electrical test.
DESIGN TESTS
It is very difficult to provide tests which will duplicate service conditions since it is impossible to reproduce in short periods of time in the laboratory the temperature cyclic changes which will occur in line service. Although engineers who are familiar with the testing of electrical porcelain do not feel satisfied with laboratory tests, nevertheless all agree that the design which passes severe laboratory tests has apparent merits.
Many articles have been published in the engineering periodicals during the last few years discussing types and causes of failures of electrical porcelain. These articles have been presented by men in the field and by representatives of the manufacturers. In general, engineers agree that the failure of suspension insulators having cemented hardware is largely due to two causes. First: Porosity. Second:. Mechanical stresses.
No doubt, porosity was one of the vital factors in causing the failure of the suspension insulators manufactured prior to about 1914. It is not necessary to go into a discussion of the causes of porosity. A brief statement that the porous insulator is an insulator that is underfired is perhaps sufficient. Of course, if we consider the problems of the manufacturer it may be that the material is not properly fired, that the ingredients are not properly proportioned or ground, etc. The subject of porosity affords sufficient information for an article. Now that due consideration is given to all of these factors, the more progressive manufacturers have practically eliminated porous ware.
Second: Mechanical stresses have always been instrumental in the failure of suspension insulators having cemented hardware. Although porosity may have been a vital factor prior to about 1914, mechanical stresses have continued to give trouble and 'are more difficult to eliminate. The subject of mechanical stress has also been very generally discussed in the engineering periodicals and everyone is familiar with the present assembly methods and the very gratifying results obtained. The elimination of the mechanical stresses caused by temperature changes without doubt, depends upon first, the design; second, the assembly of the hardware and porcelain; third, the setting of the cement under temperature and moisture conditions, etc.
It is impossible to give in any detail an analysis of the design tests which our engineers have made during the past few months. In order to give a general picture of the problems involved, the various lines of research are indicated by the following paragraphs.
ELECTROSTATIC FIELD
The cap and pin suspension insulator or the inter-linked type insulator do not represent ideal electrical designs from a consideration of the electrostatic field. However, it is necessary in nearly all commercial designs to sacrifice some efficiency of electrical design to produce a commercial unit which will be economical under the average, conditions, i. e., a unit which will perform satisfactorily in dry climates, or in humid climates, indoors, outdoors, etc. The design of insulators based on theoretical principles is discussed in considerable detail in a paper presented before the American Institute in' June 1918. (2) It is not necessary to go into a detailed discussion of the electrostatic design of suspension insulators since everyone is somewhat familiar with the theoretical principles.
For general information and interest, Figs. 15 and 16 are included. Fig. 15 pictures the electrostatic field of the cap and pin suspension insulator; Fig. 16 pictures that of the interlinked insulator. To obtain the most efficient electrical characteristics the lines of force in the area about the insulator should be either parallel or perpendicular to the surface of the insulating material. It is interesting to note the difference in the field about the two suspension insulators. As a matter of fact, the distribution of the stress in the air about the interlinked type insulator is materially better than the distribution about the cap and pin type. This perhaps, explains why the interlinked type insulator has a flash-over comparable to that of the cap and pin type, although there is more corona formation around the interlinking hardware of the inter-linked insulator below flash-over than about the cap and pin. It is probable, however, that the corona formation does not build up as rapidly over the surface of the insulator sheds as in the case of the cap and pin because of the more favorable electrostatic field.
Several authors have discussed the advisability of obtaining an insulating material having a low dielectric constant or a high dielectric constant. Obviously it is possible to divert further from the theoretical principles in the design of the insulator if the material has a dielectric constant approaching that of air. Electrical porcelain has a dielectric constant of approximately five compared to the dielectric constant of one of air. It is impossible to vary this constant to any great extent by changing the composition of electrical porcelain.
To the practical engineer it may appear that investigations of the field form of insulators are of no particular consequence. However, the determination of field form of a separate unit or assembled units is particularly useful in determining the concentration of stress. Although experimental methods indicated in the figures do not give the distribution of stress, nevertheless when obtaining the results the investigator can determine the sections of high local stress by the manner in which the particles of material act during the experiment. This method of investigation is applicable to strings of units as well as to separate units.
MECHANICAL STRESS
It is, perhaps, more difficult to provide tests to indicate the comparative resistance of insulators to temperature changes. We have followed a number of lines of investigation and will very briefly discuss each line.
SPECIAL DESIGN OF EYEBOLT
A theoretical analysis of the mechanical stress transmitted to the porcelain by the cemented hardware indicates clearly that the expansion of the eyebolt is probably the most serious factor. To definitely determine this in the laboratory, porcelains assembled with caps alone, with eyebolts alone, and without hardware, were tested by alternate immersions in hot and cold water baths. The porcelain parts assembled with eyebolts alone failed under the same severity of tests as the assembled insulators. The porcelain parts without hardware and with caps alone did not fail under any of the temperature change tests.
Several modifications of the solid eyebolt were then considered and two special types were produced and tested. These were assembled with identical porcelain parts and metal caps and under the same conditions. The two modifications consisted of (A) an eyebolt having two drop forgings assembled with a porcelain sleeve. (B) eyebolt having a one-piece drop forging with a metal sleeve into which the eyebolt could be turned.
These two special designs were compared directly with standard design having the solid eyebolt. The insulators were tested by immersing them alternately in water baths maintained at zero and 100 degrees centigrade. The periods of immersion were. varied from one minute at the start to ten minutes at the end.
The insulators were not under mechanical load during the tests. Briefly the results of this were as in Table I.
From an analysis of this table it is very apparent that the insulators having a separable metal sleeve resisted the mechanical stresses from temperature changes more successfully than the insulators having the solid eyebolt or the eyebolt with the porcelain sleeve. Also the insulators having the eyebolt with a porcelain sleeve resisted the mechanical stress better than the insulators with the solid eyebolt.
The inherent design of these insulators is such that the ultimate strength and the combined mechanical and electrical strength are practically identical. Under combined electrical and mechanical test the type with the solid eyebolt gave the highest results averaging between 10,000 and 11,000 pounds. The insulator with the eyebolt having the porcelain sleeve averaged between 9,000 and 10,000 pounds the insulator having the separable metal sleeve averaged between 8000 and 10,000 pounds. The lower strength of the two special types was due largely to the lower strength of the eyebolts. Later, additional samples were made which gave practically the same combined mechanical and electrical test as the solid eyebolt type.
After making these preliminary tests with insulators not subjected to mechanical load, it was thought particularly desirable to have similar tests made with the insulators under tension. The three designs were again tested under a series of temperature changes at 4000 and 5000 pounds load. During the immersions, the insulators were assembled in strings in series with a dynamometer as indicated in Fig. 17.
The load on the insulators shifted somewhat during the transfer from the hot water bath to the cold water. This amount of change varied for the different units and is, perhaps, somewhat indicative of the rigidity of the assembly. It is obvious that the insulator showing the least change of load must have the greatest resilience in its assembly.
All of the failures occurred during the test under 5000 pounds load. In these tests the insulator having the eyebolt with a metal thimble again proved superior. A portion of the porcelain parts assembled with the solid eyebolt had corrugated gripping surfaces and a portion sanded surfaces. The insulators with the sanded surfaces proved superior in resisting temperature changes and gave less change in load during test, indicating that the sanded surface affords greater resilience than the corrugated surface.
Although the insulators with the special design of eyebolt proved superior to the type with the solid eyebolt, it is not possible to determine in the laboratory whether or not service results will be particularly better. We have, however, sufficient of each type now in transmission line service to determine, we believe, in the next two or three years, whether or not there is any advantage in these modifications. However, the improvements in design and assembly have resulted in insulators of solid eyebolt type, that will undergo these laboratory tests without failure.
SPECIAL CEMENT
A number of investigators have contended that Portland cement may be the cause of some of the line failures. They contend that the cement gradually crystallizes and that it may transmit mechanical stresses to the porcelain. Moreover, the coefficient of expansion of cement is greater than that of porcelain and this may increase the hazard. In an attempt to lower the coefficient of expansion of the cement ground porcelain was mixed with the cement. The materials were ground together in ball mills in the proportion of one part porcelain to three parts cement. Standard tensile briquettes made from this mixture gave an average of 327 pounds after 48 hours.
Suspension insulators assembled with this material gave the same ultimate mechanical strength as when assembled with neat Portland cement. The results of these initial tests are so encouraging that the investigation will be continued.
ABRASION TESTS
Probably porcelain that is most resistant to grinding will better resist weathering conditions, the action of acid, moisture, etc. Occasionally, sections from commercial and special insulators are tested to determine their comparative resistance to abrasion.
As an example of the variation, a test of pieces from three commercial insulators of different manufacture is given. The samples were ground under the same pressure and the same area subjected to grinding. The loss in grams per square inch of surface for each minute of grinding was:
No. 1--0.075
No. 2--0.080
No. 3--0.087
It is very apparent that there is a considerable variation in the mechanical structure. Undoubtedly, the individual insulators of any one manufacturer would vary, but the difference is also due to some extent to the materials used and to the factory processes.
FLASH-OVER TESTS
The electrical stress impressed on insulators in line service has been discussed in so many papers that it is entirely unnecessary to go into any detail at this time. The distribution of stress over suspension insulators and the flash-over characteristics of strings of insulators would require a separate paper. Without any special metal fittings to effect a better distribution, the line unit of a string of suspension insulators will assume from 20 to 40 per cent of the total line voltage depending upon the number of insulators in the string, upon the design of the insulator, etc. Many curves of distribution have been published and everyone is familiar with the general shape of these curves which indicate that the line unit or the two units next to the line have much greater than the average stresses impressed upon them.
The installation of transmission lines operating at 150 to 220,000 volts, makes it necessary to give careful thought to the concentration of stress upon individual units. The insulators will be subjected to extremely high stresses whenever a flash-over occurs or whenever surges are impressed upon the line from switching, lightning, etc. Fig. 18 is indicative of the stresses from flash-over. This figure shows wet flash-overs on a combination of the interlinked insulators and cap and pin insulators. The interlinked insulators form the V part on the string. It is interesting to note the parallel arcs which occur under wet flash-over. Fig. 19 shows a dry flash-over on a string of interlinked insulators. A study of these flash-ovens by means of a high-speed camera is giving good results.
CONCLUSIONS
We should at all times consider that porcelain is an extremely fragile material, and that very careful thought must be given to the assembly of porcelain with metal parts, especially if the assembled unit will be subject to severe temperature changes or to mechanical vibrations.
Rapid strides in electrical porcelain manufacture have been made during the past few years. Perhaps the greatest advancement is in the methods of production in the factory.
The manufacturers are very open to suggestions from the field and solicit the constructive criticisms which can, perhaps, be given to better advantage by men from the field than by men in the factory. This attitude will continue to react to the mutual advantage of all.
To be presented at the Annual and Pacific Coast Convention of the A. I. E. E., Salt Lake City, June 21-24, 1921.
(1.) "Experimental Investigation of Porcelain Mixes," G. I. Gilchrest and T. A. Klinefelter, The Electric Journal, March 1918, p. 77.
(2.) Application of Theory and Practise to the Design of Transmission Line Insulators, G. I. Gilchrest, TRANS. A. I. E. E., 1918.
