The authors have provided an important addition to the current knowledge of the problems of high torques in steel mills.Their crucial contributions are twofold. Firstly,they have demonstrated, in Fig. 8, that backlash in a mill can have a serious effect .Secondly, they have shown, possibly better than anyone else so far, that stepwise integration procedures can predict torques accurately in the right circumstances. However, much more understanding of the mechanism is required before this serious problem can be successfully attacked. Permit the discusser to
illustrate this by elaborating on some of the authors' conclusions.21963
The authors point out that clearance prior to impact and nonsynchronous
propagation of the shock front lead to large torque amplification in systems with
backlash. There are other mechanisms.Suppose, for example, that at impact all the
clearance is taken up. The system is then a linear system. If it is shock sensitive it will
respond violently and develop reversing torques. The clearances will then Open and
further internal impulses will occur. Calculations on such systems predict very high
torque amplifications irrespective of the initial state of clearance.
The authors found that reducing the rate of increase of torque—in their case by
not cropping the bar—reduces torque amplification as theory predicts it should. It
follows that the time of entry of the workpiece should always be considered. In some
rolling mills the time for a squarely cropped ingot to enter the rolls is long compared
to the period of the most strongly excited natural vibration of the mill. In such cases
the torque amplification can be small even though the mill is somewhat shock-
sensitive.Ideally one would use the actual force versus time function if it were known.
Because it is not known one must use a ramp function but not a linear ramp function
because that can introduce spurious mathematical results. However, since the shock
sensitivity is primarily a function of the mill other fair approximations to the time
force curve should give accurate predictions of high torque amplifications.
The authors and their associates undertook a very difficult task here and the
discusser should be indebted to them for their work.
 
Journal of Engineering for Industry
E. I. Pollard4
    This excellent paper should prove to be of great value to designers of rolling mill
systems and to analysts of mill vibration problems. The authors and their employers
have been most generous in publishing both the idealized torsional system constants
as well as much detailed and obviously costly test data.
     Correlation between unusually high transient torques and degree of coupling
backlash, by analysis, is particularly important.
    This paper is another example of the ubiquitous problem faced by vibration
analysts. We possess sophisticated mathematical and computational tools for
analyzing vibration of systems, but often can only guess at the magnitude and type of
forcing function responsible for the vibration. We then must resort to the tactic used
by the authors, of assuming a forcing function and adjusting it, and the damping of
components of the system; until calculations agree with test results, if possible.     This paper attacks the same basic problem as the discusser's paper 71-Vibr-99. It
is therefore of interest to compare calculations by the discusser's approximate method
with the authors' test results.
    In doing so the discusser neglected backlash and the branch in the system, using
the sum of the respective component inertias and stiffnesses in the two branches.
Motor torque is also neglected since it is applied gradually and is a minor cause of 轧钢机的英文文献及中文翻译:http://www.youerw.com/fanyi/lunwen_14408.html