IMPROVEMENT OF CORRECTION ACCURACY IN FLANGE ROLLER LEVELER PROCESS FOR H-BEAM RYUMA TANIGUCHI *

Transcription

1 IMPROVEMENT OF CORRECTION ACCURACY IN FLANGE ROLLER LEVELER PROCESS FOR H-BEAM BY RYUMA TANIGUCHI * SYNOPSIS Flange squareness of H-beam is corrected by Flange roller leveler in JFE Steel West Japan Works Shape department. There was a problem that the flange aquareness after online correction had been still out of acceptable range. Three actions have been carried out to improve correction accuracy. (1) Setting on Guiding-rollers at entry-side of the flange roller leveler () Installing of the software which considers deformation state in correction (3) Optimization of the coefficient for control-delayed system After these improvements, the ratio out of acceptable range about flange squareness has been reduced from 7.5% to 4.6%. Keywords: H-beam, Flange Squareness, Online Correct, Correction Accuracy, Deformation State, FEM, Feed Forward Regulation, Flange Roller Leveler 1/15

2 1. Shape Mill Layout and Overview of Production Fig. 1.1 shows the layout of JFE Kurashiki Shape Mill. The main products of this mill are H beams, steel beams with an H-shaped section as shown in Fig. 1., and steel sheet piles that are used at bank locations to protect the banks. Approximately 60 to 75 percent of the mill s total production volume is H beams. The material from the upper process is heated to a designated temperature in the heating furnace, and it then undergoes a rough rolling process at the BD mill, an intermediate molding press process at the UR E1 mill, and finally a finish molding press process at the UF mill. The mold-pressed material is cut into slices with a product length by a hot saw (HS) and cooled down to 40 or below on the cooling bed (CB). After cooling, the H beam is corrected with the leveler and the square correction instrument and placed on the inspection table for the final inspection. Products ranked as top quality are shipped after sorting, while products-in-process are shipped after the remaining work (press/finishing) is finished. Heating furnace Roughing mill (BD) Tongs saw Finishing mill (UF) Hot saw 1Zone3Zone Intermediate mill (UR) OLAC Cooling bed Rotary straightener Fig. 1.1 Layout of JFE Kurashiki Shape Mill Frange Section H shapes Sheet piling Fig. 1. Product of JFE Kurashiki Shape Mill /15

3 . Flange Squareness There are three types of flange squareness. The categories are shown Table.1. Different types of squareness have different reference locations for different tolerance ranges. For example, the tolerance range for Squareness I refers to the level difference between the flange ends. Squareness II refers to the level of inward deformation of the flange, with a tolerance range of the level difference between the flange end and flange center. Flange sweep refers to the gap between the imaginary line connecting both flange ends and the flange center. Table.1 shows a flange with a positive (+) sweep. A negative (-) flange sweep means a flange with ends that are deformed inward. In reality, many flanges have complicated shapes that are a mix of Squareness I and Flange Sweep. Table.1 The sort of Square tolerance levels (JIS). S O R T S H A PE T O LER A N C E LEV ELS R A N G E T O LER A N C E R A N G E S quareness Ⅰ Ⅰ S ection height or below 300m m S ection height greater than 300m m 1.0% of Flange w idth (m inim um 1.5m m ) 1.% of Flange w idth (m inim um 1.5m m ) S quareness Ⅱ Ⅱ Flange w idth or below 400m m Flange w idth greater than 400m m 1.5% of half Flange w idth (m inim um 1.5m m ) 0.6% of Flange w idth Flange sw eep V 1.% of half Flange w idth (m inim um 1.5m m ) 3/15

4 3. Overview of Flange Squareness Correction and Actual Out-of-Squareness 3.1 Squareness Correction Method for Flanges Correction instrument for flanges To address the unevenness of flange squareness, we use a flange square correction instrument to correct their squareness and sweep. As shown in Fig. 3.1, the flange is corrected by motor-driven vertical rolls as it is conveyed while its web is secured under the pressure of the screw-down cylinder of the horizontal rolls. Fig. 3. illustrates that flange squareness is corrected under compressive stress by the vertical rolls and tilting motion, which is powered by a jack connected to a titling motor and applied in the opposite direction of the deformation. Product Horizontal roll Vertical roll OP Fig. 3.1 Image of correction method (-) (+) (+) (-) 垂直ロール傾動による矯正垂直ロール圧力による矯正 Fig. 3. Tilt and pressure Correction 4/15

5 3.1. Correction mechanism for tilted flanges (1) During the correction of flange tilting, the web is thought to be deformed as shown in Fig When a bending moment is acted on this deformation against the cantilever, the bending moment MC acting on the web by the per-unit-length bending load on the flange F1 is given by B M C F1 (3.1). The amount of bending of flange δ is given by t1 y B 6Eμ F 1 6F1 3 t1 y 1 1 B μ L B μ L 0 3 9F t 4 1 y 0 (3.). The amount of springback after deformation δ is the sum of the web and flange s springbacks, which is calculated by (δ = δ1 + δ ). Each springback is given by 3F L B F 1B Et1 Et (3.3) Horizontal roll δ F 1 l 0 l 1 B t 1 Fig. 3.3 Deformation behavior of correction 5/15

6 3.1.3 Overview of continuous control The square correction instrument is continuously controlled by a feed forward with squareness values that are measured by the square instrument placed at the entry-side of the correction instrument. Fig. 3.4 outlines continuous control. The squareness measurement values from the square instrument are directly transmitted to the square correction instrument DDC as an analog signal at an interval of 100 mm. These transmitted squareness values are used to determine the tilting setting value by calculation with a control coefficient that is set to P/C. This means that the tilting setting value is calculated for every interval of 100 mm, which is transmitted to the square correction instrument as an analog signal for continuous control. Actual measured value (Analog signal) DR OP P/C Correction DDC Feed forward regulation (analog signal) continuous control DR OP Rotary straightener Square instrument Square correction Fig. 3.4 Tilt continuous control 6/15

7 Amount of change (mm) 4. Issues on Square Correction and Results of Countermeasures Some flanges were not successfully corrected within the tolerance, even after the square correction. This occurred either when the square correction instrument failed to address the unevenness of flange shapes (at the end), or when over or under-correction resulted in a flange with squareness outside of the tolerance for the entire length. By defining the correction capacity as Correction Amount capability Amount of of change Entry squareness Delivery squareness motion Entry squareness Amount of tilt (4.1) we investigated the difference in correction capacity among locations of correction. Fig. 4.1 shows the difference in correction capacity between TOP and MID for the area of This found that the correction capacity at TOP was as low as 0.1, which was significantly lower than 0.50 of the correction capacity at MID. 4 TOP Correction 矯正能力 capability (y=0.06x-0.54) MID Correction 矯正能力 capability (y=0.5x+0.13) (y=0.5x-0.13) 変化量 mm Amount of motion (mm) 動作量 mm Fig. 4.1 Difference of Correction capability at correction point This may have been caused by either the product shape at the end or the limitations of the facility. The former arose from products with a large deflection of web, while the latter resulted from areas where correction was insufficient. 7/15

8 4.1 Insufficient Correction at Ends Insufficient correction due to product shape (web sweep) As shown in Fig. 4.(a), there is a type of web sweep with the concave side up, which is expressed by web sweep (+), and with the concave side down, which is expressed by web sweep (-). In reality, the shape of the flange with web sweep (+) is combined with Squareness I as shown in (b). This is true for web sweep (-), which appears to be a similarly combined shape. (-) (+) (a) (b) (c) Fig. 4. Type of section sweep To correct these combined web sweeps, the flange is secured on the correction instrument, and pressure of 110 t is applied by the horizontal roll pressing cylinder. As shown in Fig. 4.3, the web shape is deformed within its elasticity range as if it had no sweep. When measuring squareness with the square instrument, the measured squareness has been affected by the web sweep. Because this affected value is converted to the tilting value at DDC, the tilting does not include deformed squareness within its elasticity range, making this tilting inappropriate for the ideal correction. Before Correction After Fig. 4.3 Shape deformation of during correction [Improvement of H beams with a web sweep] Effects of web sweep on the control coefficient In correcting products with a web sweep, we assume that the sweep is peaked at the web center and the web is perpendicular to the flange. Considering the change of squareness due to web sweep, the flange shape is deformed as shown in Fig Triangle at the web side and Triangle at the flange have a scaling relationship. The relationship between the web sweep δ and Squareness I T is given by Flange width T section height (4.1) 8/15

9 Amount of deformation (mm) By compensating for the tilting value with this T, the elastic deformation can be included in the correction. Fig. 4.4 Pattern diagram of section sweep Fig. 4.5 shows the test results using the actual instrument. The dotted line represents a normal product with no web sweep, while the solid line represents a deformed product with web sweep for linear approximation. The symbol represents a corrected web sweep after compensating for the tilting value. This compensation resulted in an increase of the tilting values (from -.0 to 6.5 mm), solving the issue of insufficient correction Add motion for offset Fill a gap of correction Product of sweeping section (n=1) -8.0 Product of non sweeping section(n=5) Product of sweeping offset (n=) Amount of motion (mm) Fig. 4.5 Real machine test of offset effect 4.1. Insufficient correction due to facility limitations In addition to the difference in diameter between the vertical roll (100 mm) and the horizontal roll (1000 mm), the correction instrument sets the vertical roll interval as the web height minus.0 mm and the horizontal roll interval as the web thickness minus 0.5 mm. For this reason, when the product is fed to the instrument, it is first touched by the vertical roll and corrected by it without being secured by the horizontal roll at the end of the product. This reduces the correction capacity for these parts of the product. To address this issue, we have installed a web guide roll at the entry-side of the correction instrument. This roll was added to secure the web before the horizontal roll secures the product to address the issue of reduced correction capacity at the ends. The dimensions of the web guide roll gap are shown in the 9/15

10 Fig Because the limit gap was not known, we analyzed the relationship between the guide roll gap and the correction capacity using ABAQUS. Guiding roll Horizontal roll D g D h Fixing zone for only vertical roll Gap = D g - D h Fig. 4.6 Pattern diagram of correction and guiding roll drawing dimension [Improvement of end correction] FEM analysis to identify the optimum value of the web guide roll gap Analysis conditions To determine the best location for the web guide roll gap, we evaluated the effects of the amount of web guide roll gap on the correction capacity. Analysis model: 1/ model targeting the x axis All rolls are defined as having a rigid body. Size targeted for analysis: /0 Material standard: SM490A Product shape: A shape with Squareness I minus 3.0 mm for the entire length Each roll condition Horizontal roll gap: Web thickness minus 0.5 mm Vertical roll gap: Web height minus.0 mm Vertical roll tilt: -5.8 mm (1 ) We conducted analysis under these conditions. Fig. 4.7 illustrates the analysis model. 10/15

11 -3.0mm Step: Step- Basic variable: U, U1 Deformation variable: U Deformation scaling factor: e+00 (a)product shape (b) Simulation model Fig. 4.7 Simulation model [Analysis result] Determination of the optimum gap by analysis Fig. 4.8 shows the analysis result, which is the relationship between the amount of flange deformation and the distance from the entry-side of the correction instrument with varied gaps. The analysis found that 1.5 mm of the gap resulted in larger deformation at the end of the entry side than at the central part. It appeared that an 8.0-mm gap had no effect on the guide roll, resulting in smaller deformation at the end of entry side. Fig. 4.8 Difference amount of deformation from gap adjustment Fig. 4.9 shows the effects of the V roll on preliminary deformation at the correction end, where we found that the pattern of deformation varies depending on the gap. When the gap was 0.0 mm, there was no preliminary deformation; only correction part was deformed. This implies that the constraints of the web guide roll were so strong that the subsequent preliminary deformation of flanges was inhibited. From the amount of flange (preliminary) deformation in the x direction, we have obtained the pattern diagram shown in (d). Increasing the gap from 0.0 mm resulted in larger preliminary deformation, with the largest effect on the wide range at the gap of 1.5 mm. 11/15

12 Correction capability (%) (a) Gap: 1.5 mm (b) Gap: 8.0 mm Gap 1.5 mm Gap 8.0 mm Gap 0.0 mm Distance of correction point (c) Gap: 0.0 mm (d) Pattern diagram of preliminarily deformation Fig. 4.9 Effect of guide roll gap on preliminarily deformation Change in correction capacity with varied gaps Setting various guide roll gaps, we examined the changes in the correction capacity according to gap amount. Fig shows the results. The correction capacity was peaked with a value of 45.1% at a gap of 1.5 mm and rapidly decreased at a gap of 1.5 mm or lower with the minimum capacity of.5% at a gap of 0.0 mm. It is believed that this was caused by a decrease of preliminary deformation amount due to the constraints of the web guide roll. Stronger constraints may inhibit deformation itself, reducing the correction capacity Correction capability Guide roll gap (mm) Fig Effect of guide roll gap on correction capability 1/15

13 Fig shows the effects of correction at various gaps (0.0 mm, 1.5 mm, and 8.0 mm) on the Mises stress on the product. We found that the maximum Mises stress was on the location near the fillet of the web, which is adjustable. Fig Effect of guide roll gap on Mises Stess Adjustment of web guide roll gap with the actual instrument Fig. 4.1 shows an example of the change in the correction capacity before and after the guide roll gap adjustment. Major repair of the instrument improved the correlation capacity at both the OP side and the DR side at the TOP, and achieved a correction capacity of 35%, which is an increase of 6% from the previous capacity before the repair. This result implies that adjustment of the web guide roll gap is related to the correction capacity at the end. Choosing the right gap will increase the correction capacity. 50 Correction capability (%) TOP MID /3 50 k Fig. 4.1 Change of correction capability at the around the adjusting 13/15

14 Tilt (mm) 4. Inclusion of Tilting Delay in the Tilting Coefficient and Correction Factor After checking the correction settings and actual reference, we have found a control delay, as shown in Fig This delay, including the motor response delay and acceleration rate, was 4 seconds. Previously, tilting for correction was based only on squareness at the entryside without consideration for this delay, which caused failure of correction. Fig shows the correction results before and after the review of each coefficient. The tilting delay was successfully eliminated, which resulted in correction within the tolerance range along the entire length of product. Set value Actual value 4sec. Time (sec.) Fig Control delay of tilt Fig Correction after review 14/15

15 5. Conclusion By improving the correction capacity of the flange square correction instrument, we found the following: Tackling out-of-squareness at the end 1) We examined the effect of the web sweep on squareness through models and experimented with the actual instrument. As a result, we found that correction within the tolerance can be achieved by tilting which takes the web sweep into account. However, we also found that products with a larger curvature may be corrected in the shape of a web sweep even if they are secured by an H roll, as the effect of web sweep on the web curvature is significant when the web height is small. ) The FEM analysis on the effect of the facility on the failure of the correction at ends found that the optimum web guide roll gap at the entry-side of the correction instrument is 1.5 mm. We also found that the guide roll will not secure the product when the gap is 4.0 mm or more. 3) With the FEM analysis results, we decided to set the web guide roll gap of the actual instrument to 1.5 mm, achieving a 6 to 10 percent improvement of the correction capacity. Tackling out-of-squareness along the entire length 4) By investigating the limit operation speed of the tilting motor, we managed to reduce tilting delay and perform correction within the tolerance at BOT. 6. Reference (1) Examination of Mechanism of Flange Squareness Correction Instruments, Yoshii et al., 1986, Document for examining squareness correction instruments 15/15