About Fred Vogel
Fred Vogel was born in Bangor, Maine in 1893. He graduated from MIT in 1915. During WWI he served both as a trial crew member on a Navy submarine and a ship inspector. After the war, Vogel joined the Westinghouse Company's Electrical Testing Laboratories, where his work in transformer design initially focused on research and development of insulation materials. Vogel was also a professor at the Illinois Institute of Technology.
In the interview, Vogel offers an overview of some of the early developments in transformer design. He discusses obstacles to design caused by lack of research on transient voltages and switching in the early 1920s. Much of the interview focuses on test programs, especially dialectric tests and the eventual establishment of impulse tests and standard values.
Vogel describe the problems of transformer design in the 1930s, including difficulties encountered with cores and end frames. He summarizes the improvements in arrestor performance, which led to the capability of establishing the levels of switching surges.
About the Interview
Fred Vogel: An Interview Conducted by Kenneth Van Tassel, IEEE History Center, 1974
Interview # 016 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, 39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.
It is recommended that this oral history be cited as follows:
Fred Vogel, an oral history conducted in 1974 by Kenneth Van Tassel, IEEE History Center, New Brunswick, NJ, USA
Interview: Fred Vogel Interviewer: Kenneth Van Tassel Date: 1974
Fred Vogel has been with Westinghouse and Allys Chalmers. He is a designer of transformers and was a professor at the Illinois Institute of Technology.
In 1919, the fundamentals of the electrical circuit for transformers were well known, and formulas for the resistance losses, reactance, regulation and other electrical characteristics had been derived and were in common use. The design, however, covers many other things because the properties of materials, the choice of materials, and the mechanical design are all factors. In comparison to today, several observations as to the progress and the art of transformer construction seem pertinent. The early transformers were of a dry type and if used outdoors were placed in a case. The materials were varnished papers, cloths, and mica. The only tests were for temperature rise, losses, resistance, and dielectric. The dielectric test consisted of a double induced voltage and an applied test of double or more than line voltage, plus a constant when oil was added. In some cases the dielectric test was still applied without the oil. For both tests, the clearances used were based on volts per inch — under creepage over insulating surfaces — jump from live point in oil or air, and puncture from one live part to another through an insulating material. Insofar as I could see, they were based primarily on experience. Since the transient voltages, switching and so forth had not been researched, the test voltages themselves were based on limited experience, and the practice of giving extra insulation in some parts was common. This was based on service experience in the field. One could go on in this vein regarding the core, short circuit strength, service conditions and so forth.
In 1920 and 1921, the possible increase in system voltages led to considerable research in the properties of air and oil, and in combination with solid insulation materials. The work in oil was done in tanks, where the test samples were merely set in oil and not vacuum filled. It was apparent that single tests varied very much, and a procedure to attempt to get the lowest values was instituted. A specific test program served to do this. Two programs however were more commonly used and they were: successive one-minute holes raising the voltage in ten percent steps, or using three- to five-minute holes with rest between tests. It was found that curves could be drawn from the data so obtained and reasonable laws applied so that the data could be used in design. This took place in the early 1930s and was reported in the literature around 1938. Some of the early deviations were found to be due to the presence of air bubbles, which were completely displaced by low-voltage seasoning. However, it was long after that before it was found that vacuum filling decreased the variations in tests. The strength between parts of windings was a mystery. In the period around 1915, the voltage distribution within windings was derived and an equivalent Fourier series calculated. That showed how a surge or steep fronted wave distributed itself within the winding and the oscillations that resulted. In the late 1920s, the effect of lightning in causing equipment failures became of great concern. At the same time, the cold cathode oscilloscope became available. It took ten or more years for the early mathematics, of 1915, on voltage distribution to become useful. The initial voltage distribution and the oscillations could then be seen with the oscilloscope and calculated with empirical inductance inductive values added.
But what was of greater importance, the impulse strength of various dielectric fields could be determined, and this took place in the early 1930s. It also led to the establishment of impulse tests and standard values, which endured for many years. However, although the sample tests showed practical ratios between low frequency and impulse voltages, the failure rate was still too high. The use of vacuum filling in apparatus eliminated gas pockets and bubbles and increased the dielectric strength, and became common practice. It also marked the decrease in the use of treated materials in most vanished or resin-impregnated materials and oil-immersed equipment. This did not come easily as it increased the cost of tanks, and it was only later that the reduction in cost of the material and the cores and coil assemblies compensated for it in cost reliability and margin of safety.
Introduction of Cold-Rolled Steel and Changes in Core Construction
In the thirties also, the use of cold- rolled steel led to an increase in flux density, which was from thirteen or fourteen thousand gauss to sixteen five hundred or seventeen thousand gauss. The results of this reduced core weight and the insulation clearances due to vacuum filling led to marked changes in the ratings of transformers versus their size and weight, and incidentally cost. The change in construction of the core from butt and lap to mitre or to wound cores for smaller units was necessary to obtain the full properties of the material.
Impulse Strength and Switching Surges
The transformer is a part of the system. The effect the system has on the transformer can be illustrated by the following comments. Transformers are subject to short-circuit, and very early it was considered in the design. A rather crude calculation was devised and design changes made. The service records then became satisfactory, and the systems did not deliver currents sufficient to make short-circuit failures a serious consideration. Very few tests were made to see whether the design short-circuit strength was met. Those that were made on distribution transformers. In the late 1960s, power systems became stiffer and tests had been devised to cover intermediate sizes, but they are still not commonly applied. Lightning arresters and breakers have been improved. The original transformer designs, tested for impulse strength, were based on system trenches as limited by the line insulation or by air gaps. Arrestor improvements led to the belief that the impulse strength could be based on arrestor performance. Also arrestors limited switching surges, as also did the changes in breaker designs. Hence, switching surge levels could also be established. This change is now in progress since the test procedures and test values are being changed. In the mid fifties, the possibility was noted that reduction in low-frequency tests and impulse tests might result in units that would fail in normal frequency and service. Indeed this happened.
IRV and Corona Tests
This either led to IRV tests or a corona test. This is still controversial. A more basic method, proposed fifteen years ago, has not been analyzed as it should be, and probably will be. Stray losses were not a serious condition in the early days except when it resulted in cover or tank heating. Imperial rules, alloy flanges, and aluminum shields or tank inserts corrected the troubles. In the late 1930s, core troubles and later end-frame troubles appeared, and some designers studied the earlier literature and found out that stray losses in metal could be calculated, checked, and then in general the trouble could be corrected. However, it is possible to forget or overlook possibilities, and service failures have occurred and probably still will. The same applies to switching surges, where present tests may not be sufficient or where the use of shielding to meet impulse tests and produce clearances may not be adequate. This latter can be a cause of corona, a loss or a local occurrence that can cause trouble. This subject gives some idea as to future changes and standards and tests, and how sophisticated transformer manufactures has become.
Effect of Load Cycle and Materials on Life of Transformer
The life of transformers is affected by the load cycle and the materials. It was found in the 1930s that paper and oil had better characteristics than varnished ones, and there were many investigations made. The early 55 degree rise, which is the basis of rating, was for dry-type insulation and not too firmly grounded. The aging process was a function of temperature, oxygen, water, and other miscellaneous materials present. Around 1925 the use of nitrogen above the oil level, then sealed tanks for smaller units, and later a diaphragm between the oil and air in the conservator, or tank, prevented oxidation. The use of fuel resonance materials reduced the number of chemical reactions. The result was that a law of deterioration versus time and temperature could be established based on material laboratory tests, but has not been completely tested on assemblies accept for distribution transformers. The rules are, however, applied to even the large transformers. A guide for loading to give recommended values for emergency loads, some cyclic loads, and the procedure for other loads was recently approved by the Power Engineering Society. But as systems change and typical loads vary, there are always new things to take into consideration. As a further comment, it is to be noted that as with all material things, they may be arranged in various and different ways. These different design practices may accentuate one weakness and may eliminate others, and they are all different. Tests to cover all of these possibilities may be a very desirable objective to purchasers but may result in prohibitive time and expense to the builder and to the industry. It is to be presumed that this subject always will be a live one.
Family Background and Education
Could we have a little of your background, when you were born and so forth?
I was born in Bangor, Maine in 1893, and I moved to Cambridge in Massachusetts eventually, where I went to MIT and graduated in 1915. I worked for the Navy department as a ship inspector for submarines and in a trial crew in the first World War. Later I worked for the Westinghouse Company, starting in 1919. I forget how long it was that I worked for the Electrical Testing Laboratories. I knew Mr. F.M. Farmer, who was one of the early lights of the IEEE in 1916. In the Westinghouse Company I was interested in the insulation work largely because it was a thing that was in the development stage. It wasn't electrical, primarily, it was mechanical, chemical, and manufacturing and of course the insulation development work itself what you would call Physics.
[Editor's note: It is possible that there was once another tape continuing this interview, but no tape and no record of it can be found.]