|
Thickness Variation in Seamless Tube (continued) |
|
|
|
Analysis of thickness variations
For the detailed analysis of thickness variations, the recording
from a single transducer on the rotary probe is selected and the
measurements obtained for each revolution of the transducer analysed as a
data set. The effect of eccentricity may be considered by taking an eccentric
tube and plotting the wall thickness against angle of rotation (Fig.3a); the
variation obtained is very close to sinusoidal at a frequency of one cycle
peer transducer rotation. |
|
The action of rotary forging tends to thin
the tube wall effectively at the top and bottom (deep in roll groove) but
leaves it thicker at the sides; this produces ovality (Fig.3b) which when
plotted approximates to a sinusoidal graph with two cycles per probe
rotation. |
To
eliminate ovality, the bloom is turned by about 90° between each blow from
the mill rolls. The small residual variation takes the form of a sinusoidal
thickness variation having four cycles per probe, rotation which is termed
‘squareness’ (see diagram to the right)
The above components, together with mean thickness, can be
extracted from each set of thickness data using Fourier series analysis.
This technique calculates an amplitude and phase angle for each component.
The amplitude is a measure of the size or importance of the component, and
the phase angle is the angle of rotation around the tube to the thickest
portion of the tube due to that component. |
Fig 3 |
|
Using the data analysis to improve performance |
|
The use of the
microprocessor-based unit enables the components of variation to be plotted
to a base of distance along the tube. In the example illustrated in Fig. 4,
mean thickness variation (Fig. 4a) is shown to be well under control except
at the very front of the pipe. Fig. 4b and 4c show the amplitude
and phase angle of the eccentricity component respectively. The amplitude is
fairly constant but the phase angle moves steadily through 180° along the
tube. A possible explanation of this variation is that the ingot was
eccentrically pierced along its full length and that the rotary elongator
has twisted the eccentrically pierced bloom by 180° along its length.
The sensitivity of the data analysis technique is such that even a
small and quite acceptable squareness effect can be detected and monitored
to ensure that it does not become excessive; Figs. Ed and 4c are examples of
the amplitude and phase angle of this effect. Variation in phase angle shows
that the twist given to the bloom between blows of the rolls is not 90° but
somewhat higher; this causes the pattern to spiral down the tube. (The
vertical breaks in phase angle from + 45° to - 45° are caused by the
four-corner pattern shown in Fig.3c which causes an ‘artificial’ jump as the
next corner comes nearest to zero angle of reference). From the rate at
which this spiral progresses down the tube, the twist per blow can be
calculated. This angle of twist, typically 110°, is carefully and
deliberately chosen and set up on the mill to minimise the squareness
effect.
|
Fig 4 |
|
Analysis of thickness variations of
seamless tubes by the technique described in this article has to be combined
with measurements and observations made on the rotary forge mill itself.
Further developments are in hand to enable the full potential of the rotary
forge method of seamless tube manufacture to be realised and to provide a
closer dimensional tolerance tube to cater for future market trends. |
|
|
| Previous
Page |
|
|
Author |
|
S. Prouten
BSc. |
|
|
Products Chinese News
Company Ultrasonics
Links |
