Laser based, noncontact speed sensor helps reduce breaks on high speed unwind, Solutions!, Onlines Exclusives, March 2004

LASER BASED, NONCONTACT SPEED SENSOR HELPS REDUCE BREAKS ON HIGH SPEED UNWIND

By Peter Nawfel, Polytec PI, Inc.

With high speed unwinds continually running faster, speed matching of the parent reel to the expiring reel has become more critical. At these high speeds, however, typical methods of calculating speed may not suffice. Measuring the initial reel diameter at slow speeds in conjunction with the RPM at high speed may not provide the reliability and repeatability required to meet process efficiency goals. Diameter expansion of the parent reel due to centrifugal force during acceleration, and potential errors in the initial diameter measurement with traditional sensors, can result in surface speed calculation errors. These errors can be large enough to create wide variations in tension, causing a break during splicing.

Figure 1. LSV Laser Surface Velocimeter for noncontact speed and length measurements

In February of 2002, a Polytec LSV5000, laser based, noncontact speed and length sensor was installed in conjunction with a high speed unwind (5200 ft/min) supplied by Metso Paper, in order to improve the speed matching of the parent reel with the expiring reel. The mill justified the project based on the fact that, on average, one break per day occurred with the existing unwind system, causing a significant loss to production efficiency. It was believed a high percentage of the breaks were due to mismatched speeds as indicated by the significant variation in tension during splicing. Online data from the LSV5000 supported this assumption and illustrated that speed matching becomes more critical as machine speeds increase.

LSV5000 sensor technology
The LSV5000 directly measures surface velocity without contact by using the proven technique of Doppler frequency shift. Laser light of known frequency will shift in frequency as it scatters off the surface of a moving material. The velocity is determined by measuring this frequency shift, which is directly proportional to the velocity of the material. By integrating this high-speed velocity data over time, the system can also calculate precision length measurements in real time.

The system offers a variety of outputs, such as, RS232/RS422, -10V to +10V, Quadrature Encoder pulse count, for easy interface of velocity and length data to process control systems. 24V logic I/O are also available to set up interlocks, for example laser on/off, open/close of external shutters, start/stop of the internal length measurement and other logic functions.

Figure 2. LSV5000 sensor technology

The LSV5000 measures on all types of materials and has been used in a variety of industries including steel and metals, plastics, high grade films, glass, textile, rubber, paper and more. The sensor is typically used when contact devices—including contact encoder wheels, tachometers and other devices—do not provide an accurate measurement of product speed or length due to slippage, bad contact, thermal expansion/contraction or when a contact device marks, scratches or deforms the product to be measured.

Paper mill application
The existing control loop was based on measuring the diameter of the parent reel at very slow speeds with an ultrasonic sensor. Final surface speed, at the time of splicing, was calculated based on this initial diameter measurement and the RPM of the reel at speed. However, in practice, the mill found that at higher speeds the actual diameter of the parent reel increased with increasing speed due to centrifugal force. The result was an error in calculated surface speed at the time of splicing. Due to the increasing diameter, the true surface speed was greater than the calculated speed.

The LSV5000 verified that errors in calculated speed were indeed occurring and were as great as 20 - 30 ft/min. Although the mill attempted online measurements of diameter at speed with various sensors, including ultrasound and lasers, they found that the readings were too unstable for controlling the drives.

Initially, the LSV5000 noncontact speed sensor was installed as a monitoring device to measure the true surface speed of the parent reel, with the quadrature encoder output integrated into the control system for future use as a possible feedback device. Review of the data after one week of operation showed a noticeable difference between the calculated speed and the true surface speed, demonstrating the correlation between surface speed and the variation in tension during splicing. The resulting data enabled the mill to predict when a break would occur due to mismatched speeds.

With this information, the mill immediately integrated the sensor into the control loop as a trim device to fine-tune the final speed of the parent reel before splicing, and has used this outer feedback loop ever since.

The new control loop measures the parent reel diameter at slow speeds and uses this value as the initial input for ramping the drive system roughly to the specified speed. However, once in range, the measurement from the LSV5000 is used to precisely match the parent reel with the expiring reel just before splicing. By integrating the surface speed signal into the control loop, the mill now consistently matches parent reel speeds to within 5 ft/min at about 5000 ft/min (0.1%). The control could be tighter, but this is sufficient for the particular process. The result was a significant reduction in tension variation during the splice, which eliminated breaks due to mismatched speeds.

A sound investment
In general, a missed splice can cost a mill from $500 to $8500 for various operations, processes, down time, and equipment. The number is based on the cost of the discarded chemicals, lost paper, possible slow down to the paper machine, rescheduling of sets at the winders and other issues. In addition to process and production losses, matching of the reel speeds also minimizes the mechanical shock and stress to the machinery resulting from the splicing event, which may reduce maintenance costs in the long run.

By analyzing the total cost per missed splice and the number of missed splices related to mismatched speeds or tension, a mill can determine if a justification exists for such an upgrade. Installation at the mill referenced above received a ROI in a few months.

The LSV5000 is typically used when traditional contact devices do not provide the appropriate measurement accuracy. Since it measures the true speed of the paper, directly, without contact, rather than a turning drum, wheel or motor, it eliminates errors due to slippage, bad contact or changing diameter. It also eliminates quality issues due to scratching or marking of the surface.

In regards to the parent reel on a flying splice, due to centrifugal force, the diameter will increase with increasing speed. A variation in diameter results in a variation of surface speed. Therefore, if the assumed diameter is incorrect, the calculated speed will be in error. This error could be great enough to cause a mismatch in speeds resulting in significant variations in tension causing a break during splicing. The LSV5000 measures surface speed directly thus providing an accurate, repeatable and reliable method of determining the true speed of the parent reel.

Other applications:

• Precision length verification at PM and winders
• Differential sheet speed for stretch or draw calculations
• CD speed variations
• FFT analysis of velocity variations on sheet, felts, drums, etc. for PPM applications

About the author:
Peter M. Nawfel is LSV sales/product manager at Polytec PI, Inc. in Auburn, Massachusetts, USA. Reach him at (tel.) 508-832-0501 x34, or by email at petern@polytecpi.com; or, for more information,contact Dan Blank, senior electrical engineer, Metso Paper, Appleton, Wisconsin, USA, (tel.) 920-733-7361, email dan.blank@metso.com