The Study of Wear 



331 



fehlergrenzen ubcrein. Damit wird die Annahnic, 

 daB der Staub den Konverter ini metallischen Zu- 

 stand verliiBt, stark gestiitzt. 



Die genannte Bcweisfiihrung ermoglicht es, die 

 Staubemission bei Stahlfriscliprozessen in einfachstcr 

 Weise durch Extinktionsmessungen triigheitslos zu 

 verfolgen. Voraussetzung dafLir ist die Bestimmung 

 der KorngroBenverteilungen des Staubauswurl's mit 

 der eingangs geschilderten Zahimethodik. 



AuBerdem ist die nunmehr mtigliche schnelle Er- 

 mittiung der KorngroBenverteilung von Interesse bei 

 der Neuentwicklung von Frischverfahren. 



Bild 8 zeigt den Staub eines neuen Frischverfah- 

 rens (des Graef-Rotor-Prozesses), bei dem es ge- 

 lungen ist, entstaubungstechnisch giinstiggrobe Kor- 

 ner zu ziichten. Aus Bild 9 gehen die in der Hutten- 

 werk Oberhausen AG. ermittelten Grenzen der Korn- 

 groBenverteilungskurven fiir verschiedene Prozesse, 

 einschlieBlich des Graef-Rotor-Prozesses, hervor. 



Ein groBer Teil der vorliegenden Arbeit basiert auf 

 Untersuchungcn, die im Rahmen einer Diplomarbcit 

 NOP. Herrn H. D. Pantke aiisgefiihrt wurden, wofur ihm 

 auch an dieser Stelle gedankt sei. Der Werksleitung der 

 Hiittenvverk Oberhausen AG. und Herrn Professor E. 

 Ruska gebiihrt unser Dank fiir die Erkuibnis zur Durch- 



fUhriing und Publikalion der Unlersuchung. Weiterhin 

 dankcn wir Herrn Dipl. Math. K. Schniek fiir wertvolle 

 Hilfe bei der kingwicrigen Erreehnung der Extinktions- 

 konstanten. 



LiTtRATUR 



1. Kfhlik, H., Koch, A., unci Tissik, K., Z. h/.v.v. Mikro- 



skop. 62, 521 (1956). 



2. KOSMIDIR, H., NEUHAU.S, H., und Kkatzenstein, H., 



Sicihl II. fuscn 74, 1045 (1954). 



3. Landolt-Bornsiiin, Piiys.-chcm. Tabcilcn, 5. Aiifl. 



Hauplwerk, II. Teil, 907. 



4. MhLDAU, R. unci Laii Mi'i I II , D., Arch, niscnhiiitt'nw. 



27, 149 11956). 



5. Mm, G., Ann. Pliyxik IV, 25, 377 ( 1908). 



6. Pi piTRHorr, W., Opiik 8, 354 (1951). 



7. Pi ppiKiiorr, W. und Zikm, F\, Arch. Ei.senhiitwnw. 11, 



295 (1951). 



8. RO.SSLER, P., Opiik 10, 531 II (1953). 



9. ScHLUGE, H., Vortrag aus der Elektronenoptikertagung 



1955, Munstcr. 



10. SiLiNKOPF, A., Unlcrsucluingcn libcr Konvcrtcrrauch im 



Hinblick auf die spektrale Uberwachung desThomas- 

 prozcsses. Dissertation der Rhcinischen Friedrich- 

 Wiiiiclms-Universitat zu Bonn. 1954. 



1 1 . WtvER, F., Koch, W., HorrRMANN, H., Sii inkoi'f, B. A., 



Knuppel, H., Mayer, K. E., und WiEruoFP, C, 

 Stahl u. Eisen 75, 549 (1955). 



The Electron Microscope in the Study of Wear 



D. Scott and H. M. Scott 



Mechanical Engineering Research Laln)ratory, Lubrication and Wear Divi.sion, I horntonliall, (jla.sgow 



Wear may be defined as the undesired displacement 

 or removal of solid material from rubbing surfaces, 

 and since the initiation of damage can be expected 

 to occur over a sub-microscopic area the electron 

 microscope is a useful tool in this field. 



There is a considerable diversity of electron-optical 

 techniques of value in surface examination and the 

 choice of method depends very much on the nature 

 of the problem to be solved. Direct surface examina- 

 tion as shown by Halliday (4) and Menter (5) can 

 be of great use but is limited to the examination of 

 specimens capable of being accommodated in the 

 specimen holder of the microscope and thus involves 

 the construction of special wear machines using 

 suitably sized specimens. To examine specimens from 

 wear tests carried out with large specimens in already 

 established machines involves sectioning of the speci- 

 men for direct observation or the use of replica 

 techniques. 



A replica to be suitable for electron microscopy of 

 metal surfaces should be a faithful reproduction of the sur- 

 face contour, possess sufficient contrast to permit ready 

 interpretation of the surface features and be structureless 

 so that any structure seen will be that of the specimen 

 surface from which the replica was taken. 



Metal surfaces from wear experiments were examined 



with a transmission t\pe Metropolitan-Vickers E.M.3 

 electron microscope using form\ar replicas made from 

 a I ",, solution of fornnar in chloroform. \\ lien backed 

 with a thick collodion film it is possible to remo\e thin 

 formvar films of high contrast from metal surfaces 

 without tearing or without any evidence of strain in the 

 formvar. The replicas were shadowed with gold palladium 

 at an angle of 45 to impro\e the contrast and to render 

 visible a great deal of fine detail without confusing the 

 interpretation of the main structure. 



Various two-stage replica techniques have been de- 

 scribed and used to produce a positive replica of a surface 

 and in this respect the high resolution carbon replica 

 technique deseloped h> Bradley (3) has been found to be 

 particularly etVective. Howe\er, most of the two-stage 

 replica techniques are somewhat lime absorbing and 

 electron micrographs of single stage formvar replicas 

 can be photographically re\erse printed in order to fa- 

 cilitate interpretation of the correct perspecti\e of the 

 original specimen. Reverse prints are made from a 

 positi\e plate, contact printed from the negati\e. 



Replicas from the worn surfaces of specimens, 

 taken from a wear machine in which upper and lower 

 annular specimens are rubbed together under con- 

 trolled conditions of speed, load and lubrication, 

 showed clearly the abrasions of the original surface 

 together with deformation, scoring, and build up of 

 material caused b\ the rubbing action. 



