THE UNIVERSITY OF MICHIGAN College of Engineering Mechanical Engineering Department Cavitation and Multiphase Flow Laboratory Report No. UMICH 01357-35-I CLOSURE to discussion by K.S. Janaki Ram, S.R. Sasikanth, and B. C. Syamala Rao of "Interrupted Jet Water Gun Impact-Erosion Studies on Metallic Alloys, " 1974 ASMIE Polyphase Flow Forum, by F. G. Hammitt, J. B. Hwang, Linh N. Do by F. G. Hammitt February 1975 Financial Support Provided by: Nat ional Science Foundation Grant No. GK-ll89

CLOSURE to Discussion by K. S. Janaki Ram, S.R. Sasikanth, and B. C. Syamala Rao of "Interrupted Jet Water Gun Impact-Erosion Studies on Metallic Alloys", 1974 ASME Polyphase Flow Forum, by F. G. Hammitt, J. B. Hwang, and Linh N. Do (01357-22-I) The authors would like to thank the discussors for their comments, which certainly add to the understanding and value of our paper. In general we feel that their comments are very well taken. We will now attempt to reply to the various points which they raise. We agree that air entrainment in the jet may influence the results obtained to some extent. However, we believe that there are other factors, probably more important, which cause the jet shape to change with velocity. The mechanism of jet formation in this automated water-gun device is certainly very complex, and it is no doubt not "linear" over the range of operation, in that application of additional kinetic energy to the striking hammer through additional compression of the driving spring (the method by which the liquid jet velocity is increased) does not necessarily produce a proportionate increase in the kinetic energy of the striking portion of the jet, since complex shock wave interaction effects in the liquid reservoir from which the jet is produced are involved. We have made no attempt to study this non-steady complex flow phenomenon in detail, so we really cannot at this time pinpoint the causes of changes in jet shape with velocity. We agree that a correction to the effective impact area should be made for non-perpendicular impacts, to further explain reduction of damage for oblique impacts. We are not familiar with the paper by Homma and Sakamoto on air entrainment effects in reducing the area of the mostly liquid portion of a free jet, but certainly this effect may be involved in the results here obtained. The curve in Fig. 3 for nickel at 600 m/s does not indicate damage at zero impacts, as suggested by the discussors. The abscissa is in units of 103 impacts, and the curve is carried down to only about 500 impacts, where some damage was measured. The results cannot evidently be safely extrapolated to still lower numbers of impacts than those tested. The fact that impact at 600 m/s produces less damage in this particular device with this material than that at 400 m/s can only be ascribed in our opinion to the aforementioned change of jet shape with velocity. We agree that additional experimental data would be desirable to support the conclusions indicated by Fig. 4, but unfortunately financial and time limitations prevented the attainment of this admittedly desirable goal.

.7i.:, a-C.:.i! on on'' o 2>''-r "tor rt rrupntd JUt,>ter C-un T-L- ct- l- 0sion 5tudies On i.iet.llic Ai3V.s" by P. G. Iarimitt, J.I. vJan: rnJd Linh if. eo., bby 1. Kb.o.Jnaki iirrm, Research Sclbolar, 2) Th;..R*JasikCnttlh,'eSntific Officer, 3) Dr. E. C. Sysmala Reo, Asst. rorfes or..Departm;.ent of Civil.Sngineerirn, Indian ins:-titute of 2icience, Ba-1nga-lore-56O1012, Indi a. We -would lik!e to thank rrcf. F.G.ismz-itt for the c ortesy of s.endi, us a prep;rin-t of tlhe.bove paper. There ere - fcv eoinlts thl v;e -:ould like to discuss. It is statzd thcat the existence of r mrx-: r.,. in the dr:ne vs. velocity curve is C. result of the chllcne in shape of thIe jet lec-ding edge as velocity is varied. Is this trenrd due to then r' ntmi;Cnt in the liquid jet or are there any other factors?,,:egardingm th n de in cossttion d the rn corosi. -e s ction l area of the liauid jet 2s the imrnact area it is 3felt th;t th2is -v-Il hold good only for as erpendicular (90~) inmact. correction may have to be applied for the area to be considered in case of inclined jets. The change in the area for the various angles of attack is obvi.ously the reason for the decrease in the dramage, as shown by the experiments. For choosing the cross-sectional area of the liquid jet as the impact crea for evaeluating the mean depth of penetration, one of the reasons presented is that it is the area on the specimen actually subjected to maximum surface load. But due to air entrainment the area of the jet consisting'dense liquid may be much smaller than the area of the jet opening. This point has been ex}gerimncntally investigated by Homma and 3akaroto1 i.ioreover, when the area of the d2amag-ed portion is 1. LI.Homma, and T.Sakramato,'Breaking up of e circular shaped stre.eer of falling w.ater,' roc. IAHR, 8th 5Congress, Ilontreal, 24-29. Aug. 1959, pp. 2-D-1 to 2-D-14.

- 2 - larger thal the area of the jet, the dcf!:tet~ jet v:ould be acting on larger area than the jet. This deflected jet may shatter away some of the protrusions in the damaged area of the material, thus resulting in an increased area of surface load. The reason for the damage in the cse of nickel to be more. at 400 m/s than at 600 m/s velocity may need som.e explanation. iMoreover, the curve for 610 mn/s in Fig.3 indicates somr.e dcmage even at zero number of implcts. It is felt that the curve needs Qome modification. It would have been better if more xperimentel data v:ere collect.: (for various other velocities) to su.pport the conclusions based on Fig.4, in view of the varizability of damage for the same condlitions of operations.