PROdry Technology Inc.

PROdry image

Responses to User Feedback to the Infrared Imaging Forum

How does the Thermal Frequency Spectrum tell me what's wrong on my machine?

A TFS tells us what frequencies are present in a time-based thermal signal from the sheet or piece of equipment. This is analogous to the result of a vibration analysis, where velocity or acceleration data is collected at a specified sampling frequency, producing a "time-domain" signal. Theoretically, any time-domain signal is composed of an infinite number of simple sine waves of various frequency and amplitude. So the vibration signal can be transformed into a "frequency-domain" spectrum using a technique such as the Fast Fourier Transform. The peaks in the spectrum tell what the frequency of the vibration component is, and the amplitude tells how severe the vibration is. Similarly, in a Thermal Frequency Spectrum, a long time-domain signal from PROdry's high-speed thermal imager is converted into the frequency-domain for temperature. An example is at the right. Peaks can usually be related to a known rotation frequency of a component in the papermaking system. This might be the fan pump speed, the rotation frequency of the pressure roll, or the rotating speed of the Yankee dryer. Multiples of the rotating speed of a component have special meaning. For example, in the sample TFS shown, the high-amplitude peak is 6.22 Hz, which corresponds to four times the rotating speed of the Yankee dryer, which conveniently is the "pass frequency" of the four condensate removal headers in the dryer. Since there are many components in the drying system that can contribute to sheet moisture variation, the TFS can become quite "cluttered", but the dominant frequencies can usually be related to individual machine components. Once a source component is identified, steps can be taken to address and reduce the variation it produces. Ultimately, the goal is to eliminate as much thermal (i.e., moisture) variation as possible, producing the best quality sheet the machine can make.

Why can't I use stop action images in place of line scanning?

The problem with "stop action", or frame-by-frame analysis of moving objects is that there can be significant movement during the capture of a frame when the capture frequency is 30 Hz. There can also be movement that produces gaps in the thermal coverage of the item. This is particularly true if the field of view (in the direction of the motion) is small, such as would be the case in getting detailed information. The faster the speed of the object, the more critical it is to have high frequency line scanning available to avoid loss of data. For example, if a paper machine is running 3000 ft/minute, then 50 feet of paper goes by an observer each second. If we capture infrared images at 30 Hz, the paper moves 50/30 or 1.67 feet in each frame. This produces two problems: 1) if the field of view is less than 1.67 feet, it is not possible to get total coverage of the paper, and 2) since the paper moves in each frame, there is a "motion blur" introduced into the image. If the same paper is sampled by line scan at 3400 Hz, then only 0.0147 feet, or 0.176 inches, of paper passes between lines! The thermal image will be virtually frozen without blur or loss of coverage.

Our Yankee is surrounded by peripheral equipment. How can I get a decent IR shot of the dryer?

This is certainly a common problem with infrared imaging around the Yankee dryer! The solution when using full-frame imaging is to shoot very close to the dryer between pieces of equipment, or crawl under the dryer and try to shoot the bare surface. Both approaches leave something to be desired as it becomes difficult to tell what you're looking at without a steady view. Line scanning has the advantage of looking at only a single line in the picture, so if there is a "slot" to look through, then it is possible to extrude the thermal image of the dryer using line scan. With good software, it is also possible to piece together multiple bands to obtain a full cross-machine thermal profile. This works particularly well when the image collection is synchronized to the dryer rotation. Some of the examples in the "Services" section of this web site were produced in this manner.