| OCR Text |
Show was probably dominant, started digging into what the physicists had to say about gas radiation, and had my approach to the problem well along by 1926 when I discovere1 th~t Shn~k at the W~rmestelle in Dusseldorf had been through the same thinking and had published in 1924. I was depressed, felt Shack had said most of what I was about to say, but not all, presented my first paper on gas radiation in furnaces in the Spri~ of '27, and started on a program which, later at MIT, made experimental measurements on gas radiation that are good by modern standards. Almost a decade after my Lackawanna days Dr. John Eberhardt, ;later Bethlehem Steel's Director of Research and one of the founding members of the American Flame Research Committee, worked with me on an Sc.D. thesis that included in-situ measurements on radiation from billet reheating furnace gases. Our present recommendations on furnace-gas radiation are based on a weighted mean of early total-radiation measurements combined with much later spectral measurements, interpreted with a sound model. Though extrapolation outside the temperature range of the early work calls for significant correction of the latter and though band-broadening corrections have changed, the early work in both the U.S. and Germany(Schmidt, Eckert) has helped formulate the best present recommendations for engineering use. In the late 20's I got interested in petnr leum processing furnaces, particularly for thermal cracking of heavy oil to make gasoline. My experience does not include the earliest days of oil cracking, 1915 on, but I picked up anecdotes about the early headaches. The inventors thought flame on tubes carrying oil under high pressure was dangerous, and separation of combusti on from heat transfer was tri ed fi rst, wi th all the tubes putin the convection section. But the first row of tubes failed early. Furnaces with perforated arches over the convection section reduced the flux, but the design was expe~ sive and the arch bricks tended to fuse or fall out. Some tubes were put in the combustion section to protect the first convection row, but not enough, and they burned out faster than the first design. Lining the radiant or combustion secti on ·wi th enough tubes to protect each other was slow coming; establishing the balance between firing rate and fractional coverage of tre combustion chamber walls by the heat sink was a concept still in the future. Fluegas recirculation to moderate the gas leaving the burners was tried, and it worked well but was too expensive. One solution was burning in small parallel tunnels roofed with carborundum tiles, and with tubes mounted in the roof above the tiles. Same furnaces of that type were still in operation at Humble Oil in Baytown, Texas on my second trip there in about 1931. As the tiles failed, the remaining ones had been moved toward the burner end and, when I was there, the ~ furnaces were operating with only about one-third of the pa ra 11 e 1 flow passages covered wi til carborundllTl; 4 the roof tiles were surviving direct contact . with flame gases. By the early 30's the predl C-tion of overall performance of the radiant section of oil processing furnaces was not in large error, and it had become well known that the best furnaces were those that carried out most of the heat transfer in the radiant or combustion section. But the oil industry had not yet settled on any best pattern of chamber desig~, and any quantitative prediction of flux distr~bution over the radiant chamber walls was stlll far in the future. The largest cracking-coil furnaces of the time were s0,uare boxes thirty to forty feet on a side. I will skip a decade or so and come to the late 40's and the beginning of what is today tre International Flame Foundation and its American branch AFRC. Thring, chief physicist of the British Iron and Steel Research Association (whom I had met in England during the War when he was ar BCURA), de Graaf, director of research of Royal Dutch Steel, and Ma1cor, head of IRSID (the French Institute for Ironworks Research) got together and decided on a joint three-nation effort to improve industry's knowledge of f1mre in furnaces. The first efforts were on a barter basis; Royal Dutch Steel donated an available test furnace and working area, Dutch Shell Oil supplied fuel, and the British and French provided manpower, mostly from BISRA, BCURA,IRSID, Air Liquide, and Gaz de France. British, French, and Dutch national committees were set up, plus an international committee headed by Prof.Ribaud of the Sorbonne, also head of Gaz de France's Laboratoire des Hautes Temperatures. Med Thring of BISRA was made the general manager. A proposed program was circulated to industry, with an invitation to join and contribute financially. The proposal that went to Esso's British subsidiary was forwarded to Standard Oil Development in New Jersey, then on to me because I was a consultant for SOD. Esso's contribution was to send me to Europe in 1950 for discussions in London, Paris, Delft, Ijmuiden, plus a couple of lectures. When I aked if they wanted U. S. cooperation, the answer was "Not yet. We three nations still have to learn how to work together harmoniously." But the next year de Graaf came to an A.I.M.M.E. meeting in the U.S., visited me in Cambridge, and asked for U.S. support. Ralph Sherman of Battelle and I ultimately qot together a U.S. committee and it started contributing to the Ijmuiden experimental work. The early years were by hindsight mostly lear~ ing years, with measurements made that were i~ teresting but difficult to use. Sometime in the 60's I made a plea to the Foundation to better identify the heat sink and to include 2n steradian measurements of radiation, asserting that information that could not in some way be fed into a model to find its effect was almost useless. The biggest early effect of the Umuiden reports was their emphasis on the importance, to furnace performance, of momentum transport . You would be amazed at how poorly most early |
