Automated optical digital linearity deviation meter

Korolev À.N.,  Lukin A.Ya.,  Polishchuk G.S.

Measurement of deviations from linearity, flatness and alignment is the important aspect of technologies, relating to mechanical engineering, aircraft industry and shipbuilding. Besides, said measurement is widely used while controlling the quality of plates, guiding elements of machines, frames of engines of large dimensions, rolling mills, presses and turbines. Moreover, this type of measurement is used to check linearity of moving of parts of machines as well as of other mechanisms.

The Alignment Monitoring System, in which automatic reflectivity design [1] is implemented, is known. If we apply this system, the limiting value of allowable root-mean-square deviation (“RMS deviation” below), relating to a random component of the basic error does not exceed 0.05 mm at any distance up to 10 m.

Besides, Laser Linearity Deviation Meter [2] has been developed. The basic feature of this meter is that it includes axicon, used for forming the laser beam ring structure, which is convenient when making measurements. Measurement of co-ordinates is performed by taking into account position of the ring structure central spot, diameter of which is some times less than the diameter of the laser beam. Measurement error is determined by the accuracy of measurement of the central spot co-ordinates at the speckle-sound background. Value of error may be determined by using the formula  ± (0,004 + 3·10 – 3 L), where L is the distance in meters.

Another device, used for measurement of deviations, is known under the name “PPS-11 Measuring Sighting Telescope”. This device is intended to measure linearity deviations, relating to long objects, as well as to measure alignment deviations, relating both to holes and tubes. Moreover, said device is used for performing evaluation of non-parallelism, extent of absence of perpendicularity and deviation from horizontal position, which relate to surfaces of various articles. “LOMO” PLC was engaged in manufacturing this optical instrument some years ago, and at present said instrument is widely used at the mechanical engineering, machine-tool building and shipbuilding plants for carrying out accurate measurements of linearity deviations.

Such instruments, which are known under the name “Alignment Telescopes”, are manufactured by some companies of other countries. For example, optical devices, produced by the “Taylor Hobson” company of Great Britain [5], have a very high reputation.

An alignment telescope includes objective lens and focusing lens, which are employed to form image of a measuring mark in the plane of cross hairs. Said measuring mark may be located at any of the distances, lying within the range 0 – 30 meters, from the telescope end face. It is assured that whenever the focusing lens is moved in longitudinal direction it is moving along the guiding line without any deviations. In this way the required position of the sighting line with respect to the optical axis of the system, including objective lens and focusing lens, is maintained. Value of misalignment of the optical mark image with respect to the sighting line is measured by means of the optical micrometer, which includes tilting plane-parallel plate and measuring drums. Position of these drums is changing while the plane-parallel plate is being turned around two mutually perpendicular axes. Operator monitors alignment by viewing image in the device ocular.

When performing measurements an operator should bring an image of the measuring mark into focus within the alignment telescope. To achieve focusing an operator must move focusing lens by using special control handle. Then misalignment of the measuring mark image with respect to the sighting line is subject to measurement. This measurement is performed by way of turning the plane-parallel plate around the horizontal axis and the vertical axis as well as by bringing the mark image centre and the cross hairs centre into perfect coincidence. It should be noted that quantity of measurements depends upon the distance between the measuring mark and the telescope end face. If said distance exceeds 10 meters then measurement should be carried out not less than 10 times.

The main disadvantages of such optical instruments are insufficient accuracy of measurements and low productivity of an operator’s labour, that is caused by great number of operations, performed manually.

The digital device of OPTRO-PPS-031 type [6], intended to measure linearity deviations by means of sighting the measuring mark, was developed. Said meter includes the digital TV-camera, which functions as a photo-detector. Both processing of data bulk, relating to TV-images, and controlling all the procedures of measurement are performed by way of using the corresponding software.

In the digital device, presented here, the function of a photo-detector is carried out by the BR-1340LM-UF type digital TV-camera (www.es-experts.ru). This camera includes the CCD-matrix of dimension 5,7 mm x 6,3 mm (number of pixels is 1280 x 1024, the pixel size is 5,2 micron x 5,2 micron).

In Fig.1 the side view of the digital meter alignment telescope and the shape of the measuring mark are shown.

Fig. 1

The measuring mark is a transparent element, on the surface of which circles of various diameters, relating to a single centre, are drawn. This mark is illuminated by means of a light diode matrix. The software database employs the table, which includes such data as circle consecutive numbers, circle diameters and relative width of rings, located between adjacent circles. Values of the circle diameters are presented by a functional, which allows both to calculate some values of relative width of rings and to apply these values for automatic determining a circle consecutive number. The measuring mark shape affords an opportunity for carrying out measurements within the range 0 – 30 meters. If distance is short then measurement is performed by way of using image of the central circles of small diameters. If distance is long image of the circles of small diameters may not be formed, and that is why measurement is performed by way of using image of the circles of large diameters.

Applying the measuring mark of the maximum diameter 90 mm affords the opportunity for performing measurements at the distance 30 meters. Provided the diameters of the mark peripheral circles are large enough the distances, at which measurements are performed, may be as long as 50 or even 100 meters. The most important limiting factors in this case are stiffness of a measuring stand and environment stability.

Block diagram of the digital device, which is used to measure both the linearity deviations and the alignment deviations, is shown in Fig. 2. This diagram includes the following elements: measuring mark 1, provided with illuminator; object, such as frame, plate, slab, guiding element etc., subject to measurement; the main objective lens 3; the focusing lens 4; optical system 5, which forms an inverted image; digital TV-camera 6; control block 7; computer 8, to which TV-camera 6 and control block 7 are connected. The block 7 controls functioning of the drive, which provides step-by-step movement of the focusing lens 4. The term “drive, providing step-by-step movement” means aggregation of a step-by-step engine and a mechanism, which are employed to convert rotational movement of said engine into the focusing lens translation.

Fig. 2

When the process of measurements is being carried out the measuring mark images are located in different positions on a surface of an object under study. These positions correspond to the route, along which measurements should be performed, and they are located at various distances from the deviation meter.

Focusing of the lens 4 at the measuring mark image is carried out after the computer generates the corresponding command, that leads to movement of said lens. While the focusing lens is moving each of the frames, formed by the TV-camera, is used to calculate the focusing parameter. Value of this parameter is determined by the sum of the first derivative magnitudes, relating to all points of a 2-D array, used to display the measuring mark. The process of automatic focusing is completed at the moment the maximum value of the focusing parameter is obtained.

As a result of performing the above mentioned process values of both the focal distance and the image scale are subject to change.

According to the software, employed to carry out measurements, the following procedures, relating to each of the measuring mark positions, are performed:

a) automatic throwing the measuring mark image into focus with respect to the TV-camera matrix surface;

b) determining an image scale or, to be exact, the optical system magnification (by way of measuring the diameters of circles, by comparing the measured values with their nominal values and by calculating the weighted mean value);

c) establishing the value of distance to the measuring mark (by way of using the tabulated database, which refers image scale to determined distance);

d) ascertaining position of the measuring mark image centre on the matrix surface (by way of measuring the image centre co-ordinates with respect to each of the circles as well as by determining the weighted mean value with respect to the circles, subject to examination);

e) calculating the value of misalignment of the measuring mark image centre in relation to co-ordinates of the sighting line trace;

f) calculating the value of misalignment of the measuring mark centre in the object space with respect to the base-line.

The term “sighting line trace” means the point where the sighting line intersects the plane of the TV-camera light sensitive matrix.

Performing each of the abovementioned procedures is based on digital processing of the measuring mark image, formed on the matrix surface. It should be noted that the digital processing methods offer averaging different results of a measurement. Specifically, the following results may be subject to averaging:

2) the results of a single measurement conformably to predetermined number of frames;

3) the results of a measurement conformably to predetermined number of single measurements;

4) the results of measurements in conformity with the predetermined number of focusing positions.

Applying digital processing methods allows yielding very accurate result of measurements as well as to estimation of value of a random error, which presents the quality of measurements.

Below the following equation symbols will be used:

L – distance from the deviation meter to the measuring mark in meters;

X, Y – co-ordinates of the mark image centre along the axes X and Y in microns according to the scale of co-ordinates of the light sensitive matrix;

dX, dY – deviation of a component profile from the base line in microns relative to the co-ordinates X and Y;

V – magnification of an objective lens in relative units.

Let us consider calculation of profile of a component, subject to measurement, which is carried out by taking into account linearity deviation from the Y-axis.

Provided the process of measurements yields the results for N points, located on the distance scale, the component profile co-ordinates in each of n measurement points (n = 1 … N) should be determined by using the equation

P_n = (Y_n– Y_c) · V_n , (1)

In this equation Y_cis the co-ordinate of the sighting line trace within the detecting TV matrix. (The procedure of determining the co-ordinates of the sighting line trace within the light sensitive matrix of a TV-camera will be considered below).

The basic purpose of performing the measurements is to define a component profile relative to base-line, by means of which the initial point and the final point are located with reference to the co-ordinates P₁ and P_N . That is why co-ordinates of the base line should be determined

, (2)

And only then by using the equation (3) the profile co-ordinates relative to the base line may be determined

dY_n = P_n– B_n, (3)

The results of measurements of the rail profile co-ordinates are included in the table 1 for illustration. The measurements were performed in the mode, which may be outlined in the following way: bringing the objective lens into focus (5 times) – carrying out 5 measurements each time the objective lens is brought into focus – forming 5 frames relative to each of the measurements. It is obvious that measurement information, used for each of the profile points, contains 125 frames.

The table 1 includes the following additional equation symbols:

sdY – average value RMS deviation, relating to dY (in microns);

sX   – average value RMS deviation, relating to X  (in microns);

sY   – average value RMS deviation, relating to Y  (in microns);

sV   – average value RMS deviation, relating to V (in relative units).

Table 1

Results of measurements of the article profile

The results, presented in the Table 1, indicate that the average value RMS deviation, relating to co-ordinates (sX, sY) of the measuring mark image centre, does not exceed 0.2 microns. When the distance is changed within the range 1.6 – 3.3 m the average value RMS deviation, relating to the results of measurements (sdX, sdY), does not exceed 1.5 microns. If we consider the value sV, presented in the Table 1, it becomes obvious that the value of error, which takes place when distance L to the measuring mark is being determined, is in the order of 1 mm. 

By performing experimental measurements at long distances (in the order of 30 metres) it was proved that the average value RMS deviation, relating to co-ordinates (sX, sY) of the mark image centre, practically does not change when the distance is changed. As to a first approximation magnitudes (sdX, sdY) and (sX, sY) are related to the value of magnification V. This statement is confirmed by the data, presented in the table. If the distance to the measuring mark is equal to 30 metres then the magnification value amounts to 50X. That is why magnitudes (sdX, sdY), measured at the distance 30 metres, may be equal to 10 microns, that is confirmed by the results of measurements. Besides, if distances are in the order of 30 metres, then error of distance measurement does not exceed 10 mm.

By considering the data, given above, it may be concluded that the accuracy of measurement, provided by the OPTRO-PPS-031 digital meter, is by an order of magnitude greater than the accuracy, provided by the PPS-11 optical device.

As a rule the measurement errors, inherent in the sighting devices, are estimated by determining the difference between the results of measurements, made for two positions of a device. In the first of these positions device axis and measurement sighting telescope axis coincide, while in the second position a device is swung relative to a sighting telescope axis by 180 degrees. 

Fig. 3

Graphs, constructed in Fig. 3, are useful in visualising the results of profile measurements, relating to two abovementioned positions of a device, as well as the error of measurements for each of the profile points, subject to measurement. Said errors are estimated by determining the difference of results of measurements, relating to two positions of a device. In both of the graphs, constructed in Fig. 3, quantities of measurement distance in metres are plotted on X-axis, while quantities of deviation from the base line in microns are plotted on Y-axis of the upper graph, and error quantities in microns are plotted on Y-axis of the lower graph. It may be seen that within the distance range 0.7 – 3.5 metres the value of error does not exceed some microns.

As relating to the measurement process, described here, the basic systematic error is the error of co-ordinates of the sighting line trace. If distance values are great the basic systematic error practically has no influence on the measurement process. That is why only those measurements, which are performed at small distances, are of the greatest interest in the field of metrology.

If one uses an optical device he must set a measurement cross hairs on a device optical axis by way of alignment. And if one uses a digital meter, presented here, he does not need to set a light sensitive matrix central point on a device optical axis. It should be noted that setting a matrix central point on a device optical axis is impossible due to the fact accuracy of measurement of location of a measuring mark centre is as high as some fractions of micron. Below is considered the simple method of determining co-ordinates of a sighting line trace on a light sensitive matrix surface. This method includes two runs of measurements of a component profile. When the first run of measurements is performed digital device axis and sighting telescope axis coincide, while during performing the second run of measurements a device is swung relative to the sighting telescope axis by 180°. While making calculations, relating to two versions of a component profile, the co-ordinates of a sighting line trace remain common for each of the profile versions. The summation with respect to all profile values is made (it should be noted that the profile co-ordinates, relating to the abovementioned device positions, are of opposite signs). Then one strives for minimisation of sum of the profile values by way of selection the co-ordinates of the sighting line trace. This procedure is executed without any difficulties by using the “Excel” program. Calculated values of co-ordinates of the sighting line trace are entered into the program database for the purpose of performing computations when the digital meter is used.

It should be pointed out that generating both the measurement protocol and the graph of deviations of the mark centre from the base line is started by the digital meter since the moment the measurement, relating to the third point of the measurement route, is performed. The first point and the second point of this route relate to the base line. Result of each of the subsequent measurements is presented both in the protocol and the graph immediately. Recalculation of data, relating to the base line, takes place any time a new short-range or long-range end point appears. That is why while the measurement cycle is being performed points may be arranged in any order. The final measurement protocol and the graph, which presents a component profile, are drawn up in the real-time mode, and drawing up of these documents finishes at the moment the command “Complete” is generated. Generation of this command takes place after the measurements, relating to the last one of the points, have been performed (see Fig. 4).

Fig. 4

According to the instructions for use of the optical devices duration of the process, needed to process the results of measurements, to calculate the base line and to determine the profile deviations, is as long as 30 minutes. By taking into account this fact anyone may become firmly convicted that capacity of the digital meter is much greater than capacity of any optical device.

Reference list:

1. Anisimov A.G., Aleev A.M., Pantyushin A.V., Timofeev A.N., Basic errors of alignment control, revealed by using the automatic reflection opto-electron system, “Optics Magazine”, volume 76, # 1, 2009, p. 3–8.

2. Pinaev L.V., Leontieva G.V., Butenko L.N., Seriogin A.G., Laser meter, used to determine error in linearity, Patent of the Russian Federation # 2457434, 2010.

3. Apenko M.I., Araev V.A., Afanasiev V.A., Dureiko G.V., Zakaznov N.P., Romanov D.A., Usov V.S., Optical devices, used in the field of machine manufacturing, Reference Book, “Mashinostroenie” Publishers, 1974, p. 120–167.

4. Danilevich F.M., Nikitin V.A., Smirnova E.P., Assembly and alignment of optical instrumentation, “Mashinostroenie” Publishers, 1976, p. 222– 241.

5. Prospectus of the “Taylor Hobson” company (www.taylor-hobson.com)

6. Koroliov A.N., Lukin A.Ya., Malinin S.M., Polishchuk G.S., Tregub V.P., Digital meter, used to determine errors in linearity and alignment, Patent of the Russian Federation # 112396, 2012.