《AA-ATS-1-1994 9.99.pdf》由会员分享,可在线阅读,更多相关《AA-ATS-1-1994 9.99.pdf(575页珍藏版)》请在三一文库上搜索。
1、COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services Table of Contents Processinq “Modern Pusher-Type Furnaces for Reheating and Homogenizing Aluminum Ingots at Alcan Oswe
2、go, NY and Logan Russellville, KY” . 1 Karl Gdovka, Alcan Rolled Products company Gerhard Richter, Ebner-lndustrieofenbau Ges.m.b.H. B.Ren, University of Kentucky J.G. Morris, University of Kentucky W.Y. Lu, University of Kentucky “Ultrasonic Determination of Texture in AA 31 04 Sheet Materials” 23
3、“A New Technology for the Cooling and Quenching of Hot-Rolled Aluminium Plates” . 41 Malcolm Hill, Alusuisse Aluminium Suisse S.A Miroslaw Plata, Alusuisse-Lonza Services Ltd. Ren Von Kaenel, Alusuisse-Lonza Services Ltd. Rolling “Model Supported Profile and Flatness Control Systems” 57 Achim Schnei
4、der, Mannesmann Demag Corp. Paul Kern, Mannesmann Demag Corp. Martin Steffens, Mannesmann Demag Corp. Wolfgang Rohde, SMS Hans Georg Hartung, SMS Peter Sudau, SMS Gerard Collette, Pechiney-Rhenaulu Neuf-Brisach Patrick Malewicz, Pechiney-Rhenaulu Neuf-Brisach Jean-Philippe Guillerault, Pechiney-Vore
5、ppe Research Center Michel Morel, Davy-Clecim G.D. Holmesmith, Kaiser Aluminum the ingot handling equipment was not to come into contact with the scalped surface of the ingots. the heating chamber was to be kept glass bead-free to prevent surface contamination operation of the pusher-type furnace fa
6、cilities had to be fully automated. the productivity and energy consumption had to be superior to traditional reheating/homogenizing methods. the furnace design had to meet or exceed the strictest environmental and OSHAs safety regulations demanded by national and state codes. . 5. SPECIAL FEATURES
7、(EBNER) ALCANs special requirements listed above were achieved by utilizing the following special EBNER design features. a) clean ingot surfaces: 0 the ingot handling equipment is designed to handle the ingots on the edges only (see fig. 8) special skid bars (EBNER patent) eliminate the need for lub
8、ricants in the furnace (see fig. 9) O special metallic inner casing and insulation provide a glass bead-free furnace atmosphere (see fig. 10) EBNERIALCAN Paper; Ric.GdovlcalAlumitechB4 USAnO.Wam 9 GUWl?.vTA COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services CO
9、PYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services 12 fig. 8: ingots are handled on edges only (example: downender - LOGAN Russellville, KY) fig. 9: patented skid bars EBNEWALCAN P a m Ric,Gdovl c E i= 3 Fig. 8 Local heat-transfer coefficient versus plate surfac
10、e temperature COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services 52 500 450 400 350 300 250 200 150 1 O 0 50 O I I I I I I I I I II I 10 20 30 Time SI Fig. 9 Typical pla
11、te surface temperature curve measured in the laboratory simulator 385 390 395 400 405 41 O 41 5 420 425 430 435 440 445 450 Fig. 10 Surface temperature distribution 5 seconds after the start of cooling COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services COPYRIG
12、HT The Aluminum Association, Incorporated Licensed by Information Handling Services 335 340 345 350 355 360 365 370 375 380 385 390 395 Fig. 11 Surface temperature distribution 10 seconds after the start of cooling 285 290 295 300 305 31 O 31 5 320 325 330 335 340 345 350 53 Fig. 12 Surface temperat
13、ure distribution after 15 seconds after the start of cooling COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services 54 450 415 I 380 345 31 O 275 240 205 170 135 100 5 10 15
14、 20 Time s 25 30 Fig. 13 Evolution of temperature at the middle of the plate from the start to the end of the cooling + vapours exhaust v water evacuation Fig. 14 Illustration of the laboratory simulator COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services COPYR
15、IGHT The Aluminum Association, Incorporated Licensed by Information Handling Services 55 O 200 400 600 800 400 300 200 1 O0 O O 1 O00 O i o 20 30 Time Is 40 50 Fig. 15 Typical simulator cooling curves for two types of nozzle and various water flow-rates 150 300 Plaie width mm 450 Fig. 16 The sagging
16、 of the plate from the horizontal before and after cooling COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services 57 Model Supported Profile and Flatness Control Systems Ach
17、im Schneider Paul Kern Martin Steffens MDS Mannesmann Demag Sack GmbH, RatingedGermany ABSTRACT In the design of modem rolling systems, equipment with an effective crown and flatness control device is a normal part of process automation today. Therefore model-supported adjusting commands must reflec
18、t the characteristics of the controlling elements and the physical aspects of the rolling process with sufficient accuracy. 1 . APFC -AUTOMATIC PROFILE AND FLATNESS CONTROL FOR HOT ROLLING MILLS Introduction Ever increasing demands on the quality of finished rolled strips are an every day requiremen
19、t for operators of rolling mills. The attainment of good strip flatness and of certain strip profiles, that means certain thickness contours over the width, at the same time is beeing given greater attention. The desire of minimize flatness defects means that the demands on hot rolling mills rise to
20、o. In particular reproducible variable strip profiles can basically only be obtained by the hot rolling process. T h i s means that the use of innovative technologies to optimize operation of a l l available actuators is inevitable. As shown in the following section, it is in fact the profile contro
21、l, which is a very complicated formation of physical effects and physical/mechanical restrictions (Fig. 1 ) . The complexity of the task of profile control during the hot rolling process can only be registrated and considered in its entirety by a model-based Set-up, such as the system, developed joi
22、ntly by MDS Mannesmann Demag Sack and Sidmar described in the following section. The number of influences affecting the exit strip profile and its flatness can scarcely be counted using statistic models. Especially in greater variations of rolling conditions with regard to final geometry andior ther
23、mal or geometrical conditions in the plant statistic models do have their limitations. APFC, the automatic profile and flatness control is based on the physical description of the strip profile and flatness effects. The APFC is, therefore, particularly suitable for meeting requirements for target va
24、lues and reproducibility. 1 COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services 58 Roll wear Flattening between rolls Grinding/shifting of the rolls Tension distribution
25、concave conmx : = % CyndrkiPI e however, to ensure analogy to the damped springs in the mechanical system, the damping constant D , , was split into the strip modulus CB and the damping time constant TDw. Clearly evident now is the analogy between a spring in the mill stand and the imaginary spring
26、CB (the strip modulus) resulting from the rolling process. It is not surprising anymore that a simulation of this total system shows that, if certain operating parameters are given, the system begins to chatter just due to its own instability and without any outside excitation (Fig. 4 ) . For this r
27、ecording, however, a setting close to the boundary of instability was selec- ted and the system excited by a step in the ingoing strip thick- ness. The 1st part of the figure as seen from top shows this situation. In the 2nd part figure, the rolling speed was reduced by 10 %. As a result, chattering
28、 is immediately damped. In the 3rd part figure, a nearly identical effect was achieved by reducing the entry tension by 10 %. The 4th figure shows the influence of the friction coefficient in the roll gap which results in a variation of the strip modulus CB and of the coefficient KT for influencing
29、the tension. The instability as existing under control-technological aspects can be proven by the frequency response (Fig. 5 ) . This frequency response was recorded for the control loop opened at point AZO and for the state close to the stability boundary. Clearly recogniz- able is the critical pha
30、se angle of -180 at the very point where the magnitude characteristic is less than O dB. Even though the correlations may be hardly comprehensible for the practician, the comparison of simulation and measurement surely should be convincing (Fig. 6 ) . COPYRIGHT The Aluminum Association, Incorporated
31、 Licensed by Information Handling Services COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services 76 - 6 - Shown here is the as-measured and the as-calculated situation of a multi-stand tandem mill with simultaneous chattering on two stands with 126 hertz. Observe
32、d was a beat of 0,8 hertz. The simulation showed a chattering frequency of 120 hertz with about the same beat frequency as observed in practical operation. This beat was caused by different roll set dimensions in the two stands under investigation. The dimensions of the rolls conformed to the actual
33、 ones in the mill. The slight difference between as- calculated and as-measured frequency may be due to the lineariza- tion of the elasticities acting between the rolls. As you know, flattening takes place according to an exponential function which however can be linearized with a sufficient approxi
34、mation. I want to emphasize how extremely difficult it is to predict the existence or nonexistence of chattering just by way of simulation. Even though the physical correlations are clear, it is not yet possible at this time to give any comprehensive rule or guideline for design or operation. The re
35、ason for this is the almost infi- nite number of working points in regard to strip modulus CB, influence of tension K , and damping factor TDw. On the other hand, clear statements can be made in regard to the measures suitable for eliminating any such chattering by changing the working point. Partic
36、ularly evidenced can be instabilities at different eigen- values. i.e. the 3rd and 5th eigenvalues. To sum up I want to emphasize once again that the foregoing explanation of the chattering problem must not be considered true for all potential cases. Experience shows that even though the mechanism j
37、ust described is active in most cases, it is far from instability. In such cases, chattering may also be caused through outside excitation (antifriction bearings, roll crown). COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services COPYRIGHT The Aluminum Associatio
38、n, Incorporated Licensed by Information Handling Services - 7 - In many cases, measures taken with regard to the work roll neck bearings and the grinding process therefore turned out successful. However, it remains of decisive importance that in any case the phenomenon described leads to a reduction
39、 of damping of the total system along with a corresponding increase of its sensitivity to outside excitation. Activities are under way to find measures suitable for counter- acting this phenomenon. ImDact of Entrv-Side Strb Deflection on Tor and Bottom Roll Toruues Each practician knows that a varia
40、tion of the entry angle results in different top and bottom roll torques. Especially the provision of a bridle for multiple strip deflection leads to surprisingly great torque differences. The impact shall now be investigated: It can be assumed generally known that a stationary rolling opera- tion f
41、eaturing uniformly distributed tensile stresses on entry and exit sides as well as identical angular velocities at top and bottom rolls produces identical neutral points at the rolls (Fig. 7 ) . The torque distribution between top roll and bottom roll can be expected to be 50 % to 50 %. When increas
42、ing the entry-side tensile stress, the neutral point of each of the two rolls is shifted towards the exit side (Fig. 8). The total torque goes up, but remains distributed 50 % to 50 % among top roll and bottom roll. 77 The symmetrical state and condition as described is however significantly unbalan
43、ced when the strip enters under a certain angle, for example due to forced downward deflection (Pig. 9). COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services COPYRIGHT The Aluminum Association, Incorporated Licensed by Information Handling Services 78 - 8 - Let
44、us assume that the originally uniformly distributed tensile stress is, owing to said deflection, superimposed by a bending stress which is linearly distributed over the strip thickness and which produces a stress increase on the top side and a stress decrease on the underside. Experience shows that
45、this in turn leads to a higher top roll torque and a correspondingly lower bottom roll torque. It has repeatedly been assumed that this phenomenon is due to an asymmetrical distribution of reduction. The truth however is that this phenomenon is caused by the non- uniform distribution of the tensile
46、stress over the strip thick- ness. To make this clearer, let us first take a look at the phenomenon as such. When neglecting, due to the presumably small deviation of the entry angle from the horizontal plane, the differently high reduc- tions and when moreover dividing the rolling process symmetric
47、ally relative to the strip center, it is safe to imagine two different rolling processes featuring differently high entry tensions (Pig. 1 0 ) . The upper half shows a rolling process at high tension which according to the recognitions reflected by Figs. 7 and 8 leads to a displacement of the neutra
48、l point towards the exit side. The forward slip reduces compared to the symmetrical rolling case. The torque goes up due to the higher strip tension. On the lower strip half, the opposite phenomena will develop. You may now argue that different forward slips between two strip halves cannot possibly
49、occur and that, consequently, a synchroni- zation is enforced at the two partition faces due to tangential stress. This obvious objection can however be evaded by increasing the speed of the top roll and reducing that of the bottom roll until the exit speeds are identical again though the neutral points are different. As a result one will find that the torque difference goes up as the stress difference between top side and underside increases (Fig. 11). The values shown on the figure are the result of an exact calculation for the rolling case r rn O Y) . 5 COPYR
链接地址:https://www.31doc.com/p-3727946.html