RU2490111C1 - Method of control over transverse feed cycle in grinding - Google Patents

Abstract

FIELD: process engineering.

SUBSTANCE: invention relates to machine building and may be used in automation of round-grinding, hole-grinding and chute-grinding machines. At grinding end final rate of allowance removal in main circuit is defined to determine part surface roughness at said time moment. Grinding process imbalance and beginning of static estimation are defined by small sample, that is, part average size and roughness scope at the end of machining. Mean value and part production and roughness error range are defined under static conditions of post operation control in additional circuit. Using roughness estimate parametric identification of part roughness identification at the end of machining is performed. Proceeding from production error estimate, sparking-out allowance is corrected in the main control system circuit.

EFFECT: lower labour input, higher accuracy of grinding imbalance.

3 dwg

Description

The invention relates to the field of mechanical engineering and can be used to automate circular grinding, internal grinding and grooving grinding machines in mass and large-scale production.

A known method of controlling grinding based on dual-circuit systems (Reshetov A.G., Shelemetyev V.D. Self-tuning combined active control system with an electronic statistical comparator in the second circuit. In the collection "Algorithmization and automation of technological processes and industrial plants", Kuibyshev, KuAI , 1984, p. 157-162), in which the feed switching V _{C of the} grinding support is carried out by the main circuit in the stock function S of the workpiece V _C (S).

Feedback is carried out using an additional circuit that controls the size of the machined part in manual or semi-automatic mode. When the part size deviates by ΔL, correction is performed, as a rule, of the nursing allowance ΔS _in , i.e.

$$\Delta S_{\mathrm{at}}=f(\Delta L).\left(\mathrm{one}\right)$$

Thus, the two-loop control system implements an adaptive control algorithm of the form V _C (S, P), where P is the adaptation parameter determined by the deviation of the size of the machined part from the nominal value.

The use of a dual-circuit system and an adaptive lateral feed control algorithm is due to the presence of disturbing factors that are of a random functional nature. Among these factors characteristic of the grinding process are, for example, wear and blunting of the grinding wheel, thermal and power deformations, wear of the measuring tips and others. The resulting effect of these factors determines, ultimately, the error of the machined part ΔL.

The control action is designed to compensate for the effect of the total disturbing factor due to the correction of the control algorithm, in this case, by changing the nursing allowance ΔSв.

As a rule, postoperative control devices with statistical processing of measurement information (A. Reshetov, Automation of grinding and dimensional control of parts. Polytechnic, S.-P., 2003, p. 124) play the role of an additional circuit in dual-circuit systems.

Statistical processing of control results is carried out in such systems according to a small sample of parts 3 ... 5 pieces in size. with an interval of 30 ... 60 minutes. The size of the samples and the frequency of their selection are of fundamental importance from the point of view of the stratification of the measurement information. The small sample size provides at the same time "instantness", and therefore, better detection capabilities of specific causes. Another circumstance taken into account when determining the sample size is the complexity of the control, performed mostly manually.

The frequency of sampling also affects in a similar way: short intervals between samples reduce the risk of undetected breakdown, but increase the complexity of control.

), использующий в качестве показателя качества размеры обработанных деталей, обладает значительным запаздыванием в реализации управляющего воздействия, что снижает эффективность статистического регулирования технологического процесса обработки деталей.The disadvantages of dual-circuit control systems with statistical processing of control results in the second circuit include the low sensitivity of the device to the detection of a particular cause in the early stages of its manifestation. As a result, statistical control and, in particular, the method of average values (map $$\overline{X}-R$$

), using the dimensions of the machined parts as a quality indicator, has a significant delay in the implementation of the control action, which reduces the effectiveness of the statistical regulation of the technological process of machining parts.

The best results can be obtained if the detection of the moment of breakdown and the appearance of a special reason will be performed based on the analysis of the high-frequency components included in the initial information signal using a known device for controlling the transverse feed duty cycle during grinding (prototype patent RU 2355556 C2, publ. 20.05 .2009, BI No. 14). A signal containing high-frequency components belongs to the roughness of the machined surface and other manifestations of the shape error (M. Nevelson, Automatic control of processing accuracy on metal-cutting machines. L .: Mashinostroenie, 1982, p. 18). Roughness and shape error are the first to respond to changing processing conditions, and in this regard, roughness can be used to detect a particular reason in a controlled technological process of machining. For this purpose, the parts processed on the machine are controlled by a postoperative roughness control device for compliance with R _a <R _{a add} , where R _{a add} is the maximum allowable roughness value. The sample size here may be equal to one part, and the sampling frequency remains the same as when controlling the size of parts. When R _a ≥R _{a is} detected, the _additional operator begins to carry out the usual statistical control of the dimensions of parts according to a small sample (3 ... 5 parts), followed by the construction of an XR control map and statistical control of the technological process.

However, this method has disadvantages:

1) increases the complexity of postoperative monitoring due to the need to monitor the quality indicator and the roughness and size of the part. As a result, the cost of the postoperative control system increases, which contains autonomous control structures and statistical analysis of two quality indicators - roughness and size of the part;

2) the delay in detecting a particular cause remains significant, since the roughness control is carried out with the same frequency as the part size control.

The objective of the invention is to reduce the complexity of control and improve the accuracy of determining the moment of disruption and the emergence of a special reason. To this end, a method is proposed for determining the quality indicator of a part — surface microgeometry — based on measuring the rate of metal removal at the end of processing and removal of the grinding wheel.

Indeed, microgeometry (roughness) of a part’s surface at the time of machining completion, ceteris paribus, is a function of the final metal removal rate Vmk (Mikhelkevich VN Automatic grinding control. M: Mashinostroenie, 1975, p. 24).

$$R_a=C_{\mathrm{one}}{V}_{m\mathrm{to}}^n,\left(2\right)$$

where C ₁ , n are empirical coefficients.

Since during machining the workpiece is affected by numerous disturbing factors, the microgeometry of the surface of the part is formed randomly and therefore its quantitative assessment is carried out by statistical methods for a small sample (3 ... 5 parts) or the moving average method

$${\overline{R}}_a=C_2{\overline{V}}_{m\mathrm{to}}^{};\left(3\right)$$

$$R_{Ra}=C_2(V_{m\mathrm{to}\max }-V_{m\mathrm{to}\min }).\left(\mathrm{four}\right)$$

The technical result consists in the control of the microgeometry (roughness) of the part directly in the part processing cycle without the participation of the machine operator by measuring an indirect parameter - the rate of stock removal at the end of processing.

The problem is solved in that in the proposed method of controlling the grinding machine, including switching the feed of the grinding support as a function of the current allowance controlled by the main contour of the control system, and determining the allowance for nursing the part at the stage of postoperative control, carried out on the basis of statistical estimates of the average value and small scale samples of parts controlled by an additional circuit of the control system, at the end of processing determine the final speed tiya allowance. On the basis of the final speed of stock removal, the surface roughness of the part in dynamics is determined at the time the processing is completed. A small sample determines the average value and the magnitude of the roughness at the end of processing, the deviation of which determines the moment of debugging of the grinding process and the moment of the beginning of the statistical assessment - the average value and the magnitude of the error in the manufacture and roughness of the part, determined under static conditions of postoperative control. Based on the roughness assessment in static conditions, parametric identification of the roughness at the end of processing is carried out. Based on the assessment of manufacturing errors, a corrective effect is performed by changing the nursing allowance in the main loop of the control system.

The method of controlling the duty cycle is illustrated by the graphs of figure 1, figure 3 and the block diagram of figure 2.

по малой выборке; (в) - методом скользящей средней. На фиг.3, г представлена карта средних значений для погрешности размеров $$\Delta \overline{L}(n)$$

, определяемых в статических условиях послеоперационного контроля до и после обнаружения особой причины. Момент появления особой причины обозначен на диаграмме линией А-А.Figure 1 shows the most common in industrial practice three-interval cycle control the transverse feed V _C (S) implemented by the main circuit. Figure 2 presents a block diagram of a dual-circuit control system. Figure 3 (a, b, c) shows the roughness diagrams R _a , determined in various ways: (a) - in accordance with (2) at the end of the processing of the part; (b) through the average value $${\overline{R}}_a$$

on a small sample; (c) - the moving average method. Figure 3, g presents a map of average values for the size error $$\Delta \overline{L}(n)$$

determined in static conditions of postoperative control before and after the discovery of a particular cause. The moment of occurrence of a special cause is indicated on the diagram by line AA.

The processing of the part on the machine is carried out, as shown in Fig. 1, according to a three-interval control cycle 1 of transverse feed of the form V _C (S). The phase trajectory V _m (S) reflects the main characteristics of the grinding process.

We assume at the initial stage, for example, after straightening the circle, that the action of the disturbing factor (blunting the circle) is absent or insignificant, and the part is machined along trajectory 2 (S _n -V _m1 -V _m2 -V _{m opt} ). The value of the final velocity V _mk = V _{mk opt} will be considered optimal for the allowed range of final velocities designated as V _{mk min} and V _{mk max} , which in turn uniquely determine the range of roughness variations ΔR _a from the nominal value.

. При этом фазовая траектория 3 пройдет через точки S_н-V_м3-V_м4-V_{мк оpt} и положение конечной точки фазовой траектории сохранит свое значение V_мк=V_{мк оpt}.At subsequent stages of the work, the action of the perturbing factor will increase, and to compensate for it, it will be necessary to change the nursing allowance from S ₂ to $$S_2^{\text{'}\text{'}}$$

. In this case, the phase trajectory 3 passes through the points S _n -V _m3 -V _m4 -V _{mk opt} and the position of the end point of the phase trajectory will retain its value V _mk = V _{mk opt} .

The value of the nursing allowance is determined in accordance with the expression (1) in the additional circuit 4 of the active control system using the postoperative control device 5. The specified device, using sensors 6 and 7, which control the error in the size of the part and roughness, respectively, performs statistical processing of information, such as presented in the graphs of figure 3.

The most important point in the statistical control of product quality is the earliest possible detection of the occurrence of a special reason leading to the disorder of the technological process.

Let us consider sequentially the process of detecting a moment of disorder by a system and taking measures to form a control action using the graphs of FIG.

соответствует значению конечной скорости съема припуска V_мк=V_{мк оpt}.When processing parts 8 on the machine 9, the main circuit 10 of the active control system determines, in accordance with (3), the roughness value R _a at the end of processing. The resulting roughness values R _a are displayed on the main circuit display 11 in the form of a scatter plot 12 R _a (n). Average value $${\overline{R}}_a$$

corresponds to the value of the final _stock removal rate V _mk = V _{mk opt} .

The stratification of information in order to determine a particular reason is carried out by statistical processing of the sequence R _a (n), for example, using samples of a certain size (graph 13) or by the moving average method (graph 14). At the time indicated by straight line A-A in FIG. 3, an action of a special reason is detected, expressed in this case in the approximation of the technological process to the upper control limit of UСL _Ra .

From this moment, automatically or with the help of the operator, postoperative control of the dimensions of the machined parts 8 is carried out using the sensor 6 and the postoperative control device 5.

In Fig. 3, curve 15 on the control map of average values reflects the course of the observed technological process. Before the appearance of the special reason indicated by direct AA, the moments of sampling due to the absence of the need to detect a specific reason are quite rare and are determined by long-term reasons: analysis of the stability of the technological process in time, control of the accuracy of the equipment, improvement of the technological process, etc. After the emergence of a special reason, the sampling frequency increases and is determined by the measures of regulatory influence and the nature of the compensated disturbing factor.

Upon reaching the result of regulation, the sampling frequency returns to the original value, which was in effect until a special reason. This method of statistical control significantly reduces its complexity.

For parametric identification of the coefficient C ₂ required for model (3), roughness is periodically monitored using a sensor 7 and a postoperative control device 5.

In real practice, in order to save costs, the identification procedure with a stable nature of 12 is sufficient to be carried out periodically in the metrosal or OTC production.

The practical implementation of the claimed system was made on the basis of devices developed in the joint research and production laboratory "Automatic Control Systems" of Togliatti State University and OJSC "AvtoVAZ".

The main circuit 10 of the system is based on the model ASK2974 device, which includes an on-board computer and a visualization device for observing the technological process and constructing control cards (Fig. 3).

An additional circuit 4 contains a device for monitoring the size and roughness of a part of the model ASK1147. Brief technical information on the indicated devices is attached.

TECHNICAL CHARACTERISTICS OF THE ACTIVE CONTROL DEVICE MODEL ASK2974 1. Number of input channels 2; 2. Type of sensors inductive contact; 3. Type of display system TFT-LCD; 5.7 ” 4. Visual information provided by the indicating device - the current allowance of the workpiece in linear-discrete form in the range of 100-0-500 microns; - the current allowance of the workpiece in digital form with a resolution of 1.0 microns; - scatter plots of the error in the size of the part ΔL and roughness R _a , determined at the end of processing; - phase characteristics of the processing process in coordinates V _m -S; - tuning and service information. 5. The limit of the range of response of commands in the working range, microns 0.5; 6. The number of control teams up to 8; 7. Type of controlled surface smooth, intermittent; 8. The interface used for receiving and transmitting messages RS232; 9. Power supply 100 ... 240 V, 50 Hz; 10. Power consumption, VA thirty; 11. Overall dimensions, mm 340 × 285 × 135.

TECHNICAL CHARACTERISTICS OF THE CONTROL OF THE SIZES AND ROUGHNESS OF THE PARTS OF THE MODEL ASK1147 1. The method of obtaining measurement information pin in the measuring device; 2. Type of sensors used inductive, differential; 3. The number of input information channels up to 8 4. The used interface for receiving and transmitting information signals RS232; 5. Controlled parameters: ΔL, R _a deviation of the size of the part from the nominal value; 6. The calculated statistical parameters

среднее значениешероховатости $${\overline{R}}_a$$

, размах R_ΔL, стандартное отклонение S;average sample value $$\Delta \overline{L}$$

roughness average $${\overline{R}}_a$$

, range R _ΔL , standard deviation S; 7. Visual presentation of statistical information map of Shekhart; 8. Overall dimensions, mm 310 × 265 × 120; 9. Nutrition 100 ... 240 V: 50 Hz.

Claims (1)

A method of controlling the working cycle of the lateral feed during grinding, including switching the feed of the grinding support as a function of the current allowance controlled by the main contour of the control system, and determining the allowance for nursing the part at the postoperative control stage, based on statistical estimates of the average size of the parts and the size of the small sample of parts controlled by an additional circuit of the control system, characterized in that at the end of processing determine the final the allowance removal rate, on the basis of which the surface roughness of the part is determined at the time of processing and, according to a small sample, the average size and range of roughness at the end of processing, the deviation of which determines the grinding process debugging time and the start of the statistical estimation of the average size and size manufacturing errors and roughness of the part, determined in static conditions of postoperative control, and based on the roughness assessment in static conditions carry out the parametric identification of the roughness of the part at the time of processing completion, and based on the assessment of the manufacturing error, perform the corrective action by changing the nursing allowance in the main loop of the control system.

Images