CN108279121B

CN108279121B - System and method for testing bottom contact force characteristics of large-cutting-depth lower rolling blade

Info

Publication number: CN108279121B
Application number: CN201810127399.8A
Authority: CN
Inventors: 张魁; 刘金刚; 张高峰
Original assignee: Crec Sunward Intelligent Equipment Co ltd
Current assignee: Crec Sunward Intelligent Equipment Co ltd; Hunan Hao'er Information Consulting Co ltd
Priority date: 2018-02-08
Filing date: 2018-02-08
Publication date: 2023-10-13
Anticipated expiration: 2038-02-08
Also published as: CN108279121A

Abstract

The invention particularly relates to a blade bottom contact force distribution characteristic test system and a blade bottom contact force distribution characteristic test method for rolling and crushing hard rock under large cutting depth of a TBM disc cutter. The system comprises a standard linear cutting experiment table, a data acquisition instrument, a strain gauge and an eddy current displacement sensor, and is characterized in that: a custom hob is adopted, the strain gauges are distributed circumferentially at a given radial distribution distance on a side vertical plane of the custom hob according to equal circumferential interval angles, and test points are correspondingly formed; the data acquisition instrument can acquire output signals of the three-way force sensor, the strain gauge and the eddy current displacement sensor in real time. A test method for use with the system, characterized by: and (3) utilizing the actual measured stress/strain value of each measuring point at any rolling rock breaking moment, combining with a calibration data sample, and reversely solving the contact force component of the moment edge bed knife rock by an influence function based on a truncated singular value regularization method. The invention has the advantages that: the test result is accurate and reliable, and the material waste and the cost increase are avoided.

Description

System and method for testing bottom contact force characteristics of large-cutting-depth lower rolling blade

Technical Field

The invention belongs to the crossing fields of full-face tunnel boring machine technology, signal analysis processing technology and rock breaking science, relates to a cutter blade bottom contact force distribution characteristic test system for cutting and stripping rock and a test method thereof, and particularly relates to a blade bottom contact force distribution characteristic test system for rolling and breaking hard rock under large cutting depth of a TBM disc cutter and a test method thereof.

Background

Along with the rapid development of economy and continuous improvement of urban level in China in the 21 st century, the available space on the ground is reduced and the development and utilization of underground space are enhanced, so that the development and utilization of the underground space become the necessary trend of modern construction of urban and urban at present. In the process of excavating the tunnel in the underground space, the full-face heading machine is gradually widely used by virtue of the characteristics of high excavation efficiency, excellent engineering quality, strong geological adaptability and the like. Full face heading machines can be divided into two types: a full face rock tunnel boring machine (Full Face Rock Tunnel Boring Machine, which is commonly and simply called TBM in China, hereinafter the same) is mainly used for rock stratum tunneling with certain self-stabilization capability, and is particularly suitable for long-distance tunnel tunneling (diversion tunnel, railway tunnel and the like) in the field; the other type is a full-section soft stratum tunnel boring machine (the domestic habit is called a shield machine, the following is the same), and is mainly used for the tunneling of urban underground engineering or river-crossing tunnel engineering with water stratum, weak unstable surrounding rock and strict sedimentation control requirements on the ground surface.

Disc cutter (hereinafter referred to as cutter) is a core rock breaking cutter of the full-face tunnel boring machine (the performance of the cutter is directly related to the excavation efficiency and engineering safety of the tunneling machine), and the cutter is arranged at different positions (fixed by cutter holders) on a cutter disc at the forefront end of the full-face tunnel boring machine. The hob generally comprises a hob ring, a hob hub, a hob shaft, a bearing and other parts. When the full-face tunnel boring machine works, the hob directly presses and wedges a rock face by means of the cutter ring under the action of strong cutter disc thrust, and revolves around the central axis of the cutter disc along with the rotation of the cutter disc; simultaneously, the rotary cutter can rotate around the axis of the rotary cutter to continuously roll and crush the rock. Because the working environment of the tunneling machine is extremely bad, and the working process of the cutter is extremely complex (such as the phenomenon that the contact force is highly concentrated in the theoretical contact area of the cutter bed cutter), the abnormal abrasion of the cutter ring, the breakage of the cutter ring and other failure accident frequency occur (such as engineering statistics shows that the cost consumed by the cutter of the tunneling machine accounts for about 30% -40% of the total construction cost in the process of the Qinling tunnel). The prior research results further show that: in particular, in the tunneling construction of hard rock or super hard rock, uneven contact force distribution on the contact interface of hob and rock is a main cause of abnormal failure of the cutter. Therefore, the cutter bottom contact force is tested and researched when the hob breaks the rock so as to master the distribution characteristic of the cutter bottom contact force in the hob breaking process, thereby being beneficial to guiding the structural design and selection of the hob, and being particularly beneficial to improving the hob breaking efficiency under the hard rock tunneling working condition and prolonging the service life of the hob.

Because the working environment of the tunneling machine is extremely bad, the dynamic characteristics of the hob during rock breaking are extremely complex, and the prior art is extremely difficult to directly collect and test the contact force of the bottom cutter rock of the rolling edge. There is a possibility that the cutter ring state data is obtained by arranging a resistive strain gauge (hereinafter referred to as strain gauge) on the side of the cutter ring (which cannot be laid because the cutter bottom is involved in crushing rock at the moment of extrusion), and thus the cutter bottom contact force distribution characteristics thereof are evaluated and analyzed indirectly from the side. According to test and research experience, when soft rock is crushed and homogenized by rolling under the small cutting depth range of the hob (generally, the cutting depth h is not more than 10-15 mm), the rolling and rock breaking process is relatively stable and mild, and the safety patch area reserved on the hob is larger, so that the blade bottom contact force distribution condition of the hob can be estimated and analyzed indirectly theoretically by a method of arranging strain gauges. However, for TBM, disc cutters are typically operated on hard or extra hard rock and to ensure high driving rates, the cutter depth of the disc cutter is typically set at 10mm and above, which presents significant difficulties and challenges to the cutter bed cutter contact force test work:

on the one hand, if the hob rolls under large cutting depth to crush hard rock, the hob has extremely strong step rock-breaking effect, the rock at the edge side of the hob can continuously generate severe cracking and crushing phenomenon, and in order to avoid damage and collision of sharp rock slag and rock scraps to a laid strain gauge, the patch position of the strain gauge is far higher than the cutting depth, so that the safety area of the side surface of the cutter ring, which can be used for patch, is narrow, and great difficulty is brought to patch work;

On the other hand, unlike the special working condition of small-cutting deep rolling soft rock, when the hob rolls and cuts hard and brittle hard rock, the hob step rock breaking effect is extremely strong, and the rock breaking process cannot be considered as a quasi-steady state process, so that the discreteness of the acquired data is larger, and the later data is difficult to process;

therefore, when hard rock is crushed by rolling under the large cutting depth of the hob, if the contact force distribution characteristics of the bottom tool rock of the hob edge are indirectly monitored and reflected through the arrangement of the strain gauge, the design of a test hardware system, the reasonable setting of a test scheme, the reasonable selection of test parameters, the effective screening and processing of test data and the verification of test results are extremely critical to the test precision and the influence of test success and failure.

At present, direct or indirect test research on distribution characteristics of contact force of a cutting edge bed cutter in the rock breaking process of the hob is rarely reported. Some similar research results are made by simulation tests or theoretical assumptions. For example: based on experience, rosami et al assume that the bottom contact force of the rolling blade is linearly distributed or uniformly distributed, and deduce a hob three-way force calculation formula for guiding hob structural design and shape selection; cho et al simulate the dynamic rock breaking process of the hob by adopting AUTODYN3D, and arrange test points on the tip of the hob to record the change of cutting force; entacher, zhang Ke and the like perform simulation analysis on the hob rock breaking process by adopting ABAQUS to obtain the deformation characteristics of the hob and the rock. Sun and the like simulate cutter ring stress distribution when the hob breaks rock by adopting NX Nastran, so that maximum stress is obtained at the edge of the cutter ring right below the hob. The simulation analysis and theoretical assumption are all urgent to be reliably verified by test means.

In other application fields or industrial application occasions, although reports about contact force testing schemes exist, such as an LED module contact force testing device and a detection method thereof (application number: 201410520251.2), an engine overhead gas distribution cam shaft contact stress testing method (application number: 201410110600.3), a train wheel rail contact force wireless detection device (application number: 201210362394.6), an arch net contact force prediction method based on NARX neural network (application number: 201110436222.4), a contact resistance and contact force synchronous measurement structure and method of MEMS materials (application number: 201410790613. X), a multidimensional contact force and real contact area dynamic synchronous testing system and method (application number: 201611030857.3) and the like, the method cannot be directly or indirectly applied to a special occasion of hob breaking.

Disclosure of Invention

The invention aims to provide a blade bottom contact force distribution characteristic test system and a blade bottom contact force distribution characteristic test method for rolling and crushing hard rock under large cutting depth of a TBM disc cutter, which are used for overcoming the defects of low reliability, poor accuracy, large data dispersion (which is difficult to process), easy damage of a strain gauge and the like when the blade bottom contact force distribution characteristic is obtained by the existing means, guiding the structural design and selection of the cutter, improving the cutter rock breaking efficiency especially under the working condition of hard rock tunneling, and prolonging the service life of a cutter.

The invention relates to a bottom contact force characteristic test system of a large-cutting-depth lower rolling blade, which comprises a hob standard linear cutting experiment table, a data acquisition instrument (hereinafter referred to as a data acquisition instrument), a hob, an eddy current displacement sensor and an industrial personal computer, and is characterized in that:

the hob standard linear cutting experiment table comprises a frame, a movable cross beam, a tool apron, a rock material bin, a horizontal workbench, a vertical oil cylinder, a longitudinal oil cylinder and a horizontal oil cylinder; the hob comprises a hob ring, a hob body, a hob shaft, a bearing and an end cover; the hob is arranged in the hob seat; a three-way force sensor is also arranged between the tool apron and the movable cross beam;

the hob adopts the customization hob, including customization cutter ring, customization cutter body, customization arbor, customization bearing and customization end cover, characterized by: the cross section size of the cutting edge part of the customized cutter ring below the maximum cutting depth is the same as that of the standard cutter ring, and the left side and the right side of the cutting edge part of the customized cutter ring above the maximum cutting depth are processed into side vertical planes for arranging strain gauges; the material, manufacturing process and cutter ring diameter of the customized cutter ring are consistent with those of the standard cutter ring; a given radial arrangement distance h on the side vertical plane of the blade ₁ Arranging the strain gauges at equal circumferential interval angles delta theta, wherein the strain gauges are fully distributed on the side vertical plane of the custom cutter ring and form corresponding stress-strain test points (hereinafter referred to as test points) so as to monitor the stress/strain state of the custom cutter ring in real time;

marking raised points are arranged on the side surface of the customized hob, and the positions of the marking raised points are in one-to-one correspondence with and aligned with the arrangement positions of the strain gauges;

fixedly mounting a probe of the eddy current displacement sensor on one side of the tool apron, and aligning the probe with the mark protruding point rotating to the lowest point;

the data acquisition instrument can acquire output signals of the three-way force sensor, the strain gauge and the eddy current displacement sensor in real time and transmit the output signals to the industrial personal computer so as to process and analyze.

Preferably, a protective cover is fixedly arranged above the strain gauge.

Preferably, strain gauges are symmetrically arranged on the side vertical planes at the left side and the right side of the custom cutter ring respectively, and a full-bridge test circuit is correspondingly formed.

In order to overcome the defects of low reliability, poor accuracy, large data dispersion (which is difficult to process) and easy damage of a strain gauge when the bottom contact force distribution characteristic of the disc-shaped rolling blade is obtained by the prior art means and used for rolling and crushing hard rock under large cutting depth, the invention provides a method for testing the bottom contact force characteristic of the rolling blade under large cutting depth, which is matched with the method and is characterized in that:

Step 1: based on a rolling rock breaking simulation test or a pre-cutting test, calculating to obtain average estimated vertical force born by the customized hob when the given cutting depth h and the rock type are obtained;

step 2: calibrating and testing; rolling a horizontally fixed steel plate at the same hob setting uniform angular speed by using the customized hob to perform multiple calibration tests under the action of different set calibration vertical forces, so as to obtain a three-way cutting force signal curve output by the three-way force sensor, a strain signal curve output by the test point and a voltage signal curve output by the eddy current displacement sensor under the action of different vertical line loads;

preferably, the set nominal vertical force in step 2 should not be smaller than the average estimated vertical force obtained in step 1.

Step 3: the method comprises the following steps of:

step 3.1: calculating the measured average calibration vertical force;

step 3.2: selecting the relative error of the actually measured average calibration vertical force relative to the set calibration vertical force and the root mean square error RMSE of the vertical force as evaluation indexes of calibration test stability, and setting corresponding thresholds for the evaluation indexes; when the indexes are smaller than the set threshold, selecting a data sample obtained by the calibration test, otherwise, discarding the data sample due to overlarge vertical force fluctuation error;

Step 4: for the selected calibration data samples, the following data processing is carried out:

step 4.1: according to the working characteristics of the eddy current displacement sensor, calculating the measured average rotational angular speed of the customized hob by utilizing the voltage signal curve output by the obtained eddy current displacement sensor and combining the theoretical contact angle psi of the hob and the interval time and the position of the maximum value of the output signal of the eddy current displacement sensor, and then determining the position angle theta at the test point in the calibration test process;

step 4.2: according to the measured average rotational angular velocity obtained by step 4.1 and the sampling frequency, a change curve of the measured stress/strain value along with the position angle theta in each calibration test can be obtained;

step 4.3: converting the data curve obtained in step 4.2 into a change curve of the actually measured stress/strain value along with the position angle theta under the action of the single-point load on the unit blade width, and further obtaining the change curve of the actually measured average stress/strain value along with the position angle theta under the action of the single-point load by calculating the average value of all selected data samples;

step 5: rolling and rock breaking test; setting a cutting depth h and rock types in the same step 1, setting a uniform angular speed in the same step 2 by a hob, and carrying out a rolling rock breaking test in a single-blade cutting mode; by utilizing the system for testing the bottom contact force characteristics of the large-cutting-depth lower rolling blade, a three-way cutting force signal curve output by the three-way force sensor, a strain signal curve output by the test point and a voltage signal curve output by the eddy current displacement sensor are synchronously recorded in real time;

Step 6: obtaining an actually measured average rotational angular velocity in a rolling rock breaking test, similar to step 4.1; and then similar to step 4.2, according to the measured average rotational angular velocity and sampling frequency, the measured stress/strain signal obtained in step 5 is used for deducing any rolling rock breaking test moment, the position angle theta of each measuring point and the corresponding measured stress/strain value thereof;

step 7: constructing a relational expression of the influence function; and (3) reversely solving the moment of rock breaking by the given rolling by the influence function based on a truncated singular value regularization method by utilizing the position angle theta of each measuring point at any moment of rock breaking by rolling and the corresponding measured stress/strain value obtained in the step (6) and combining the calibration data sample obtained in the step (4)Component x of contact force of knife rock ₁ 、x ₂ 、…、x _n The unknown column vector X is constructed in the following steps:

step 7.1: the theoretical contact area of the edge bed knife rock is discretized into (n-1) equal parts (n=360 degrees/delta theta) according to equal circumferential interval angles, so that the contact force components on the unit knife edge widths received by the end points of the theoretical contact area of the edge bed knife rock at any rolling rock breaking moment (in the order of the position angle theta from large to small) are sequentially x ₁ 、x ₂ 、…、x _n And constitutes a column vector X; the stress/strain value readings sequentially output from each test point (in order of small to large position angle theta) at the moment of rock breaking by the given rolling are s ₁ 、s ₂ 、…、s _n And constitutes a column vector S; according to any rolling rock breaking moment, the stress/strain value s ₁ 、s ₂ 、…、s _n Component x of contact force with unit blade width ₁ 、x ₂ 、…、x _n The mapping relation between the two is defined as an influence function shown in the following formula (1):

CX＝S (1)

wherein, C is a coefficient matrix of an influence function;

step 7.2: fitting the change curve of the measured average stress/strain value obtained by step 4.3 along with the position angle theta to obtain a fitting function relation f of the measured average stress/strain value relative to the position angle theta _σ,ε (θ) constructing the given rolling break moment s as described in step 7.1 ₁ 、s ₂ 、…、s _n Coefficient matrix C of the corresponding influence function corresponding to each measuring point position angle theta one by one and forming element C _ij The general form of (2) is as follows:

in θ _i For a given rolling rock breaking instant element s _i The position angle theta of the corresponding test point; f (f) _σ,ε (θ) can be written as a fitted functional relation f of measured average stress values with respect to the position angle θ _σ (θ), and the measured average strain value is related to the position angleFitting function relation f of θ _ε (θ) two forms;

then the coefficient matrix C is normalized to obtain a normalized coefficient matrix C ₁ The influence function shown in the formula (1) can be further rewritten as formula (3):

in the method, in the process of the invention,is a similar solution to the unknown column vector X.

Step 7.3: obtaining a normalized coefficient matrix C by adopting Singular Value Decomposition (SVD) ₁ The singular value decomposition formula of (2) is shown in the following formula (4):

C ₁ ＝UDV ^T (4)

in the formula, a rank (C) ₁ ) =k, then there is an n-order orthogonal matrix U whose columns are defined byIs set to be u= (U) ₁ ,u ₂ ,…,u _n ) There is an n-order orthogonal matrix V whose rows are defined by +.>Is of the eigenvector composition of v= (V) ₁ ,v ₂ ,…,v _n )，Σ _k ＝diag(σ ₁ ,σ ₂ ,…,σ _k ) And->Eigenvalue lambda ₁ ≥λ ₂ ≥…λ _k > 0 is the matrix->Is not zero, sigma _i (i=1, 2, …, k) is C ₁ Is a singular value of (c).

The least squares solution of the equation shown in the formula (3) is given by the singular value decomposition formula shown in the following formula (5):

wherein:is a least squares solution to the equation shown in equation (3).

Step 7.4: in order to ensure that the equation solving quality has high reliability, comprehensively considering the influence of noise pollution on true solution X and the influence of solution resolution on solving precision, determining the maximum truncated parameter recommended value k of the singular value by using an L curve method ₀ The principle of the method is as follows: for different regularized truncation parameters k ₀ Drawing different normsAnd regularizing the solution normsA graph therebetween; in general, the graph presents an L-shape, and the inflection point of the curve is the maximum recommended value k of the regularized cutoff parameter ₀ 。

Step 7.5: based on truncated singular value regularization (TSVD) method, reducing k correction to k ₁ Least squares solution of unknown column vector X by truncating part of singular valuesCan be approximated by a truncated singular value expression as shown in the following equation (6):

in which k is 1.ltoreq.k ₁ ≤k ₀ 。

Preferably, the truncation parameter k ₁ Not more than 10.

Preferably, the normalized coefficient matrix C ₁ The order n of (2) is not greater than 100.

Step 8: let kappa be the correction coefficient, let X ₀ For the correction solution of X, then X ₀ Represented by the following formula (7):

let S ₀ To correct the solution X ₀ Substituting the theoretical stress/strain value column vector obtained by calculating the influence function shown in the formula (1), S ₀ Can be represented by the following formula (8):

based on the cyclic trial and error principle, different truncation parameters k are used ₁ Substituting different correction coefficients kappa into the formula (8) to obtain different theoretical stress/strain value column vectors S ₀ The column vector S ₀ Is composed of theoretical stress/strain values corresponding to different position angles theta; comparing the theoretical stress/strain value with the actual measured stress/strain value one by one under each position angle theta, and when the total error is minimum, the corresponding correction coefficient kappa and cutoff parameter k are obtained ₁ Is the optimal parameter; combining (6) and (7) a set of correction solutions X closest to the true value X is calculated from the optimal parameters ₀ 。

Preferably, the step 7 uses radial stress/strain values to construct a relation that affects the function.

Preferably, the step 7 uses radial stress values to construct a relation that affects the function.

The invention has the advantages that: in order to reasonably design a hob ring structure, deeply study a hob abrasion mechanism, reliably predict hob rock breaking load, improve hob service life and cutting efficiency, the invention provides a system and a method for testing bottom contact force characteristics of a hob edge under large cutting depth, and solves a series of special testing problems when the hob is used for rolling and breaking hard rock or extremely hard rock under large cutting depth (namely, when in TBM cutting working condition), including extremely easy damage of a sensor due to rock scraping caused by the strong step breaking characteristic generated by the hob edge bottom, hard layout of the sensor and other hardware limitations; the method also comprises the key technical problems of high acquired data dispersion caused by the characteristic of strong step crushing generated at the bottom of the rolling blade, difficult analysis and processing of later-stage data, serious distortion of test results and the like, and can test and obtain the contact force distribution characteristics of the rolling blade bed knife rock more accurately and reliably.

Drawings

The patent of the invention is further described below with reference to the drawings and examples.

FIG. 1 is a schematic diagram of a hob standard wire cutting experiment table.

Fig. 2 is a schematic diagram of a hob.

Fig. 3 is a schematic diagram of a standard cutter ring for a certain engineering (dotted line on the right side of the figure) and a custom cutter ring blade structure adopted in one implementation (thick solid line on the left side of the figure).

Fig. 4 is a schematic diagram of the layout position of a single strain flower on a cutter ring.

Fig. 5 is an enlarged view of a partial structure of the strained gate at the I position in fig. 4.

Fig. 6 is a schematic layout (front view) of strain gauges (only three sets), eddy current displacement sensors and marking tabs on hob and tool holder.

Fig. 7 is a left side view of fig. 6 with the tool holder omitted (simplified drawing).

Fig. 8 is a schematic diagram showing the layout effect of the actual strain gauge attached to the custom-made cutter ring shown in fig. 3 during the actual test.

FIG. 9 shows a time-varying vertical force signal curve, an eddy current displacement sensor output voltage signal curve, and measured radial stress signal curves at test points 1-3 obtained in a calibration test.

Fig. 10 is a graph showing the change of the measured radial stress value along with the position angle θ at the test point 2 obtained by sieving the calibration data sample obtained in a certain calibration test.

Fig. 11 is a schematic view showing the contact force components per unit blade width received on the theoretical contact area (in order of small to large position angle θ) of the three equal parts of edge bed knife rock at the moment of rock rolling and breaking shown in fig. 7.

FIG. 12 shows the measured average stress σ obtained by the process of FIG. 10 _r A curve of change with position angle θ.

Fig. 13 is an L graph obtained by a certain actual measurement analysis.

Detailed Description

One embodiment is implemented.

An implementation of the invention will now be described in detail with reference to fig. 1 to 13. The invention relates to a bottom contact force characteristic test system of a large-cutting-depth lower rolling blade, which comprises a hob standard linear cutting experiment table (see figure 1), a data acquisition instrument (not shown), a hob (4) (see figure 1), an eddy current displacement sensor (only a probe (10-2) shown in figure 7 is shown) and an industrial personal computer (not shown); as shown in fig. 1, the hob standard linear cutting experiment table comprises a frame (1), a movable cross beam (2), a tool apron (3), a rock material bin (5), a horizontal workbench (6), a vertical oil cylinder (7), a longitudinal oil cylinder (8) and a horizontal oil cylinder (14); the hob shown in fig. 2 is in a general structural form and comprises a hob ring (4-1), a hob body (4-2), a clamping ring (4-3), a hob shaft (4-4), a bearing (4-5), a sealing component (4-6) and an end cover (4-7); in this example, a standard 17 inch Chang Jiemian flat-bladed hob (a standard 17 inch Chang Jiemian flat-bladed hob ring as shown by the right-hand dashed line in fig. 3) which is widely used in engineering was selected as the subject. As shown in fig. 3, the standard cutter ring (hereinafter referred to as a standard cutter ring) of the standard hob (hereinafter referred to as a standard hob) includes: radius R is 216mm, and the transition arc R at the blade part (4-1-3) ₀ 1.8mm, knife edge angle theta ₀ Smaller than 6 DEG, the width of the cutter blade a ₀ 13mm; in the figure, the blade bottom (4-1-1) of the standard knife ring can not be pasted, the pasting area of the two side surfaces (4-1-2) of the blade part (4-1-3) is smaller, and the pasting is inconvenient on the inclined plane.

As shown in fig. 1 and 2, the hob (4) is arranged in the tool apron (3) through the hob shaft (4-4); and a three-way force sensor (9) is also arranged between the tool apron (3) and the movable cross beam (2) and is used for measuring three-way cutting force (vertical force, lateral force and rolling force) born by the hob (4) in the cutting process in real time.

The invention relates to a bottom contact force characteristic test system for a large-cutting-depth lower rolling blade, which is mainly characterized by comprising the following steps:

in order to reliably protect the strain gauge (avoid scraping and damaging the strain gauge by broken rock scraps in large cutting deep rolling and crushing) under the working condition of hard rock, meanwhile, in order to facilitate the surface mounting, the hob (4) adopts a custom hob for nonstandard test (hereinafter referred to as custom hob), and comprises a custom cutter ring, a custom cutter body, a custom cutter shaft, a custom bearing and a custom end cover, and is characterized in that: the method is characterized in that: adopting the cross section outline dimension shown by the thick solid line on the left side in fig. 3, designing and manufacturing a custom cutter ring for non-standard test (hereinafter called custom cutter ring for short), properly adjusting the matching dimension and structural characteristics of the parts such as a cutter body (4-2), a cutter shaft (4-4) and an end cover (4-7) according to the structural characteristics of the custom cutter ring, correspondingly designing and manufacturing other custom parts for test (a clamping ring (4-3) can be omitted if necessary), and assembling the custom cutter by the custom parts; the custom knife ring is at the maximum cutting depth h _max The cross-sectional dimensions of the following blade portions (4-1-3) are exactly the same as the standard cutter ring shown in phantom on the right in fig. 3, while the custom cutter ring is at maximum cutting depth h _max The left side and the right side (4-1-2) of the blade part (4-1-3) are processed into a side vertical plane (4-1-4) for arranging strain gauges; the material, manufacturing process and diameter of the customized cutter ring are consistent with those of the standard cutter ring;

in the present embodiment, as shown in fig. 4-7, the radial placement distance h is given on the side-lobe plane (4-1-4) of the custom cutter ring ₁ The strain gauges are distributed according to equal circumferential interval angles delta theta, and the strain gauges are distributed on the side vertical planes (4-1-4) of the custom cutter ring in a whole circumference manner and form corresponding stress strain test points (hereinafter referred to as test points) so as to monitor the stress/strain state of the custom cutter ring in real time; in the preferred embodiment, the BX120-2CA type strain gauge is selected in consideration of small size (the width of the strain gauge is 1mm, the length and width of the substrate are 7.2 mm) and compact structure; the strain-gauge arrangement of the strain gauge is shown in FIGS. 4 and 5 and comprises a 0 DEG directional (tangential) strain gauge (9-The strain signals output by the 2-1-1), the 45-degree direction strain grid (9-2-1-2) and the 90-degree direction (radial) strain grid (9-2-1-3) are marked as epsilon in sequence _0° 、ε _45° And epsilon _90° ；

Preferably, in this example, BX120-2CA strain gauges are symmetrically arranged on the side-hanging planes (4-1-4) on the left and right sides of the custom cutter ring, and a full-bridge test circuit is correspondingly formed. Taking the three groups of strain relief shown only in fig. 6 as an example, the description is explained as follows: as shown in fig. 6, 3 strain reliefs are symmetrically arranged on the side vertical planes (4-1-4) on the left side and the right side of the custom knife ring (4-1), including the strain relief 1 (10-2-1), the strain relief 2 (10-2-2) and the strain relief 3 (10-2-3) which are shown in fig. 6 and 7 and the strain relief 4 (10-2-4), the strain relief 5 (10-2-5) and the strain relief 6 (10-2-6) which are symmetrical to the strain relief 1 (10-2-1) and the strain relief 3 (10-2-3) which are positioned on the left side surface of the knife ring and the right side surface of the knife ring which are shown in fig. 7; forming a full-bridge circuit by a strain grating and a corresponding temperature compensation sheet in each direction (respectively in 3 radial directions of 0-degree tangential direction, 45-degree direction and 90-degree direction) of a pair of strain flowers at symmetrical positions on the left side surface and the right side surface of the cutter ring, thereby forming 3 test points; wherein each test point can output strain data in 3 radial directions of 0 degree tangential direction, 45 degrees and 90 degrees at the same time, namely epsilon _0° 、ε _45° And epsilon _90° The method comprises the steps of carrying out a first treatment on the surface of the For example, a full-bridge circuit is formed by a 90-degree directional strain grating of a strain gauge 1 (10-2-1) positioned at the left side of the cutter ring as shown in fig. 8, a 90-degree directional strain grating of a strain gauge 4 (10-2-4) positioned at the right side of the cutter ring as shown in fig. 7, and two 90-degree directional strain gratings as temperature compensation sheets (strain gauges of the same type) for measuring and outputting strain data epsilon at the test point in the 90-degree direction _90° The test point is named as test point 1; designating a test point formed by a pair of the strain gauge 2 (10-2-2) and the strain gauge 5 (10-2-5) as a test point 2; and analogizing sequentially, and naming a test point 3; in actual testing, in order to obtain a larger testing resolution, a plurality of series of strain flowers are often required to be pasted on a side vertical plane (4-1-4) of the customized cutter ring (4-1) along the circumferential direction, and as shown in fig. 8, a layout effect schematic diagram after BX120-2CA type strain flowers are actually pasted on the customized cutter ring as shown in fig. 3 in the actual testing is shown; in the figure, BXThe dimensions of the base and custom cutter ring of the 120-2CA strain relief are plotted against the relative scale, when the circumferential spacing angle Δθ=3°, the maximum cutting depth h _max Radial layout distance h =25mm ₁ At =15 mm, up to 120 strain relief can be laid out (if more compact strain relief is chosen, or the radial laying distance h is reduced ₁ More strain relief may also be laid out), it is seen that the requirements for test resolution are substantially met.

Preferably, a protective cover (not shown) is also fixedly arranged above the strain gauge;

similarly, in this example, as shown in FIGS. 6 and 7, a given radial placement distance h from the blade bottom (4-1-1) on the side of the custom hob ₂ Marking raised points (10-1) are distributed at the same circumferential interval angle delta theta, and the positions of the marking raised points are in one-to-one correspondence and aligned with the distribution positions of the strain flowers; more specifically, 3 steel marked protruding points (10-1) with a magnetic base are adopted and firmly adsorbed on the surface of the left end cover (4-7) of the custom hob by utilizing magnetic force, so that the radial layout position of the custom hob is far higher than the height of the strain gauge, and the possibility that a probe (10-2) is contacted with rock is eliminated; the suction positions of the 3 marking protruding points (10-1) are in one-to-one correspondence and alignment with the arrangement positions of the strain gauge flower 1 (10-2-1), the strain gauge flower 2 (10-2-2) and the strain gauge flower 3 (10-2-3);

In the example, ZA-21 series eddy current displacement sensor is adopted, and a probe (10-2) of the eddy current displacement sensor shown in fig. 7 is fixedly arranged on the left side of the tool apron (3), and the probe (10-2) is aligned with a marking protruding point (10-1) rotating to the lowest point; the probe (10-2) of the eddy current displacement sensor maintains a gap of 1-2 mm with the surface of the marking boss (10-1).

In this example, a DH5925 type data acquisition instrument was used. The data acquisition instrument can synchronously acquire and monitor voltage signals output by the three-way force sensor (9), the bridge circuit formed in the test points 1-3 and the eddy current displacement sensor in real time and transmit the voltage signals to the industrial personal computer for later data processing and analysis.

In this example, after finishing the work of customizing the cutter ring patch, marking the convex point to attract, welding the lead wire, sealing and dustproof, as shown in fig. 6 and 7, the signal cable (11) is reversely and neatly wound on the hob with respect to the autorotation direction of the cutter ring (4-1) so as to prevent the signal cable (11) from being excessively pulled in the rolling process.

According to test experience, when the hob continuously roll-break hard rock or superhard rock, especially when working in a large cutting depth range, the dynamic step rock breaking characteristic is obvious, namely the deformation and failure processes of the rock at the edge bottom and the edge side are extremely severe (the phenomena of fracture breaking, extrusion breaking, shearing breaking and rock chip burst exist), and the stress state of the hob cannot be regarded as quasi-static any more (the contact force distribution form and the contact force of the edge bottom tool are greatly different at different rolling rock breaking moments). In order to solve the key technical problems of high acquired data dispersion caused by the characteristic of the intensive step crushing, difficult analysis and processing of later data, serious distortion of test results and the like, the invention provides a large-cutting-depth bottom contact force characteristic test method for a large-cutting-depth bottom rolling blade by means of the large-cutting-depth bottom contact force characteristic test system. In the first embodiment, a maximum allowable cut depth h is given _max =25 mm, given a cut h of 20mm at test; defining a central angle AOC corresponding to a knife rock theoretical contact area arc AC as shown in FIG. 7 as a knife rock theoretical contact angle psi, wherein psi=arccos ((216-20)/216) ≡24.85 degrees according to the geometric relationship; the hard rock sample is made of uniform weathered granite with 175MPa uniaxial compressive strength and has a sampling frequency of 1000Hz. The specific implementation process is introduced as follows:

step 1: based on a rolling rock breaking simulation test or a pre-cutting test, calculating to obtain average estimated vertical force born by the customized hob when the given cutting depth h and the rock type are obtained; in the embodiment, a finite element simulation model of the hard rock sample is established by adopting large commercial transient explicit nonlinear dynamics simulation software ANSYS/LS-DYNA, and the average estimated vertical force is obtained through statistics according to simulation experiment results.

Step 2: calibrating and testing; rolling a horizontally fixed steel plate (13) by using the customized hob (4) at the same hob setting uniform angular speed as in the step 1 on a hob standard linear cutting experiment table shown in the figure 1 so as to perform multiple calibration tests under the action of different setting calibration vertical forces and obtain three-way cutting force signal curves output by the three-way force sensor, strain signal curves output by the test point and voltage signal curves output by the eddy current displacement sensor under the action of different vertical line loads; the method comprises the following steps:

Step 2.1: on a hob standard linear cutting experiment table shown in fig. 1, a large rock sample block (12) with a flat and intact upper surface is reliably fixed with a rock material bin (5), and then a steel plate (13) is stably placed on the rock sample block (12) and is firmly and reliably fixed on the rock material bin (5) through bolts (not shown); in the example, in order to avoid the bending deformation of the steel plate (13) due to stress in the test process, the calibration test precision is reduced, and a high-performance thick steel plate with larger rigidity is adopted;

step 2.2: for the convenience of measurement and analysis, the custom hob (4) in idle load is manually rotated, so that the strain gauge 1 (10-2-1) is positioned at the lowest position of the hob ring, namely the strain gauge 1 (10-2-1) is positioned at the position corresponding to the probe (10-2) shown in fig. 7; meanwhile, in order to avoid one-side tilting instability of the steel plate (13) due to unbalanced load in the test process, the hob blade is positioned on the middle symmetrical plane of the steel plate (13) and in the middle area 100mm away from the front edge of the steel plate (13); at this time, as an initial state of the calibration test;

step 2.3: the vertical cylinder (7) shown in figure 1 is driven to enable the custom hob (4) to vertically squeeze the steel plate (13) when the set calibration vertical force F is reached _v Then, locking an oil inlet and an oil outlet of the vertical oil cylinder (7); preferably, in this example, the nominal vertical force F is set as described in step 2 _v The average estimated vertical force obtained in the step 1 is not smaller than the average estimated vertical force;

step 2.4: setting the flow of the longitudinal oil cylinder (8) to be 30% of the full range, starting the longitudinal oil cylinder (8), driving the custom hob (4) to roll the steel plate (13) to a certain distance at the same set uniform angular speed, and stopping the test; during the period, synchronously recording three-way force signals, actually measured radial strain signals at all test points and voltage signals output by an eddy current displacement sensor in real time at a sampling frequency of 1000Hz as 1 data sample of one calibration test in the same test group (same set of calibration vertical force); calibratingDuring the test, a set nominal vertical force F is applied _v Can be considered as a vertical line load of constant magnitude. For convenience in describing the data processing procedure in this example, it is assumed that the circumferential spacing angle Δθ in fig. 7 is 120 °, that is, three test points 1 to 3 are set and only; according to the original signals acquired by a certain calibration test, a vertical force change curve (101) along with time, an eddy current displacement sensor output voltage change curve (301) along with time and actual measurement radial stress values change curves (201) to (203) along with time at the test points 1 to 3 are obtained through preliminary conversion; taking test point 1 as an example, the present example gives a conversion calculation process of the measured radial stress based on the strain relief test data: the principal stress sigma at any moment can be obtained by using the formulas (1) and (2) ₁ And sigma (sigma) ₂ Size and angle alpha ₁ The method comprises the steps of carrying out a first treatment on the surface of the Solving the radial stress sigma of the cutting edge at the test point 1 at the corresponding moment according to the Morse circle _r And tangential stress sigma _θ 。

Wherein E is _c Is the elastic modulus of the cutter ring, v _c Is poisson's ratio.

Step 2.5: the customized hob (4) is driven to roll the steel plate (13) in the opposite direction at the same set uniform angular velocity to the same distance, then the test is stopped, other test conditions (such as set calibration vertical force) are the same step by 2.4, and the acquired data are used as the next newly added data sample of the same test group;

step 2.6: in order to eliminate the influence of data fluctuation, repeatedly performing steps from 2.2 to 2.5, namely performing multiple calibration tests under the action of the same set calibration vertical force so as to acquire enough data samples under the same test group;

step 2.7: to eliminate random errors, different set calibrations are setVertical force KF _v (K>1) Repeatedly performing steps from 2.2 to 2.5, namely repeatedly performing calibration tests for a plurality of times under the action of different set calibration vertical forces, so as to collect enough data samples under different test groups;

step 3: the method comprises the following steps of:

step 3.1: for a change curve (101) of the vertical force with time, which is acquired by the three-way force sensor and is shown in fig. 9, the measured average calibration vertical force is obtained, and a straight line (100) of the measured average calibration vertical force with time is drawn;

Step 3.2: selecting the measured average nominal vertical force relative to the set nominal vertical force (e.g., F _v 、KF _v ) The relative error of the (C), the Root Mean Square Error (RMSE) of the vertical force change curve (101) is used as an evaluation index of calibration test stability, and a corresponding threshold value is set for the RMSE; when the indexes are smaller than the set threshold, selecting the calibration data sample, otherwise, discarding the data sample due to overlarge vertical force fluctuation error; in this example, if the fluctuation amplitude of the vertical force variation curve (101) as shown in fig. 9 obtained in the calibration test is small, only the actual measured average calibration vertical force can be selected empirically with respect to the set calibration vertical force (e.g., F _v 、KF _v ) The relative error of (2) is an evaluation index of calibration test stability, and the threshold value is set to be 5%;

step 4.1: because the intensity of the voltage signal output by the eddy current displacement sensor reflects the distance change relation between the probe (10-2) and the mark protruding point (10-1) (for example, when each test point rotates to the lowest point C shown in FIG. 7, the voltage signal output by the eddy current displacement sensor has the largest amplitude and reaches V shown in FIG. 9 _max ) Therefore, according to the working characteristic, by combining the magnitude of the theoretical contact angle psi of the cutter rock and the interval time and the position of the maximum value of the output signal of the eddy current displacement sensor, the measured average rotating angular speed of the hob which is closer to the real test condition can be calculated, and the measured average rotating angular speed of the hob can be determinedThe angular position of each test point relative to the customized hob plumb line OC (see FIG. 7) at any one rolling break moment; for example, in this example, when the test point 2 is rotated from the position shown in FIG. 7 to the plumb line OC position, from the data curve shown in FIG. 9, it is calculated that (0 to t ₂ ) Average rotational angular velocity ω during this time interval ₁ Is delta theta/t ₂ The method comprises the steps of carrying out a first treatment on the surface of the The position angle theta of each point on the side surface of the cutter ring at the moment of rolling and breaking is defined as the included angle between the connecting line of the point and the circle center O in fig. 7 and the plumb line OC, the left side of the plumb line OC is defined as the included obtuse angle, and the right side of the plumb line OC is defined as the included acute angle. As in fig. 9, the position angles θ at the initial time test point 1 and test point 2 are 0 and (360 ° - Δθ), respectively; t is t ₁ At the moment, the position angle θ at test point 1 and test point 2 is t ₁ *Δθ/t ₂ And (360 DEG to delta theta + t) ₁ *Δθ/t ₂ )；

Step 4.2: according to the measured average rotational angular velocity obtained by step 4.1 and the sampling frequency, a change curve of radial strain along with the position angle theta at each test point in each calibration test can be obtained; as shown in fig. 10, the change curve of the measured radial stress value at the test point 2 along with the position angle θ after the sieving treatment is shown; in the figure, it is assumed that the same set nominal vertical force F _v Under the action, two substantially coincident actually measured radial stress curves, namely a data curve (1021) and a curve (1022), are obtained, and the same set calibration vertical force KF is assumed _v Two substantially coincident measured radial stress curves, namely a data curve (3021) and a curve (3022), are also obtained under action;

step 4.3: converting the data curve obtained in step 4.2 into a change curve of the actually measured stress/strain value along with the position angle theta under the action of the single-point load on the unit blade width, and further obtaining the change curve of the actually measured average stress/strain value along with the position angle theta under the action of the single-point load by calculating the average value of all selected data samples; in this example, the ordinate values of the radial stress curve 1021 and curve 1022 in FIG. 10 may be divided by F _v /a ₀ The ordinate values of both the radial stress curve (3021) and the curve (3022) in fig. 10 are divided by KF _v /a ₀ Can be converted into the form shown in FIG. 11The radial stress variation curve of the test point 2 along with the position angle theta under the single-point load is further obtained to obtain the average radial stress sigma of the test point 2 _r A variation curve with position angle θ;

step 5: rolling and rock breaking test; setting a cutting depth h and rock types in the same step 1, setting a uniform angular speed of a hob in the same step 2, and carrying out a customized hob rolling rock breaking test under a single-blade cutting mode (also called a Unrelieved mode, i.e. adjacent cutting grooves are not affected by each other) on a hob standard linear cutting experiment table shown in fig. 1; by utilizing the system for testing the bottom contact force characteristics of the rolling blade under large cutting depth, disclosed by the invention, a three-way cutting force signal curve output by the three-way force sensor, a strain signal curve output by the test point and a voltage signal curve output by the eddy current displacement sensor are synchronously recorded in real time, and the system comprises the following specific steps of:

Step 5.1: the steel plate (13) shown in figure 1 and arranged on the rock material bin (5) in the step 2.1 is removed;

step 5.2: to facilitate measurement analysis, similar to step 2.2, the strain gauge 1 (10-2-1) is located in the position corresponding to the probe (10-2) shown in FIG. 8; meanwhile, in order to eliminate the size effect of the rock sample (12), the cutting edge part (4-1-3) of the custom hob is positioned on the middle symmetrical plane of the rock sample (12) and in the middle area which is 100mm away from the front edge of the rock sample (12); at this time, the initial state of the rolling rock breaking test is used;

step 5.3: similar to step 2.3, the vertical cylinder (7) shown in fig. 1 is driven so that the oil inlet and outlet of the vertical cylinder (7) are locked when the custom hob (4) invades the rock sample (12) vertically to a given cutting depth h;

step 5.4: similar to the step 2.4, the flow of the longitudinal oil cylinder (8) is set to be 30% of the full range, the longitudinal oil cylinder (8) is started, the custom hob (4) is driven to roll and cut the rock sample (12) to a given distance at the same set uniform angular speed, and then the test is stopped; during the period, synchronously recording three-directional force signals, radial strain signals at each test point and voltage signals output by an eddy current displacement sensor in real time at the same sampling frequency, and taking the signals as 1 data sample of one calibration test in the same test group (same set of calibration vertical force);

Step 5.5: the custom hob (4) moves vertically upwards to leave the rock surface, a horizontal cylinder (14) is started to drive a rock material bin (5) to horizontally move leftwards (or rightwards) by 1 enough knife interval (such as 100 mm), and then the operation of steps 5.2 to 5.5 is repeated until all grooving operations of a whole rock sample (12) are completed;

step 5.6: replacing the 2 nd rock sample made of the same rock material, and repeating the steps from 5.2 to 5.5 until enough data samples are acquired;

step 6: and (3) obtaining the actual measurement average rotational angular velocity in the rolling rock breaking test similar to the step 4.1, and further obtaining the position angle theta of each measuring point and the corresponding actual measurement stress/strain value of any rolling rock breaking test moment according to the actual measurement average rotational angular velocity and the sampling frequency and the actual measurement stress/strain signal obtained in the step 5 similar to the step 4.2.

Step 7: constructing a relational expression of the influence function; and (3) reversely solving the contact force component x of the rock breaking moment of the given rolling by the impact function by using the position angle theta of each measuring point at any moment of rock breaking by rolling and the corresponding measured stress/strain value obtained in the step (6) and combining the calibration data sample obtained in the step (4) based on a truncated singular value regularization method ₁ 、x ₂ 、…、x _n The unknown column vector X is constructed in the following steps:

step 7.1: the theoretical contact area of the edge bed knife rock is discretized into (n-1) equal parts (n=360 degrees/delta theta) according to equal circumferential interval angles, so that the contact force components on the unit knife edge widths received by the end points of the theoretical contact area of the edge bed knife rock at any rolling rock breaking moment (in the order of the position angle theta from large to small) are sequentially x ₁ 、x ₂ 、…、x _n And constitutes a column vector X; the stress/strain value readings sequentially output from each test point (in order of small to large position angle theta) at the moment of rock breaking by the given rolling are s ₁ 、s ₂ 、…、s _n And constitutes a column vector S; at any moment of rock breaking by rolling, all contact force components x on the unit blade width are used for ₁ 、x ₂ 、…、x _n Each test is respectively described aboveThe stress/strain components are correspondingly generated on the cutter ring at the positions of the points (in the order of the small position angle theta, the actual measured stress/strain values of the test points at the moment of rolling rock breaking (in the order of the small position angle theta, the large position angle theta) are actually total stress/strain values obtained by overlapping and summing the stress/strain components at the positions of the test points; according to the measured stress/strain value s ₁ 、s ₂ 、…、s _n Component x of contact force with unit blade width ₁ 、x ₂ 、…、x _n The mapping relation between the two is defined as an influence function shown in the following formula (3):

CX＝S (3)

Wherein, C is a coefficient matrix of an influence function; in this example, taking the instant shown in FIG. 7 as an example, the contact force component on the unit blade width received on the theoretical contact area of each equal portion of edge bed knife is defined as x as shown in FIG. 11 ₁ 、x ₂ 、x ₃ The unknown column vector X is formed, and the position angle theta of the theoretical contact area of each equal-part edge bed knife rock is 360 degrees (or 0), 360 degrees-psi/2 and 360 degrees-psi in sequence; in fig. 7, the position angles θ corresponding to the test points 1, 3 and 2 are 0 (or 360 °), 120 ° and 240 ° in sequence, and the output stress/strain value readings are s ₁ 、s ₂ 、s ₃ And constitutes a column vector S.

Step 7.2: assume the measured average stress value sigma obtained in step 4.3 _r The curve of change with position angle θ is a curve (4000) as shown in fig. 12, and the curve is fitted to obtain a fitted functional relation of the measured average stress with respect to the position angle θThe function can be further understood as an even function with 360 ° period, the function curve of which can be seen in the curve (5000) in fig. 12; in this case +.A fitting function of the preferred measured average radial stress with respect to the position angle θ>For example, assume that the moment of rock breaking by rolling is as shown in FIG. 7, construct the same step by step as s described in 7.1 ₁ 、s ₂ 、s ₃ The coefficient matrix C of the influence function corresponding to the position angles theta of the measuring points one by one is shown in the following formula (4):

In general, for an actual n-th order coefficient matrix C _n×n Its constituent element c _ij The general form of (2) is represented by the following formula (5):

in θ _i For a given rolling rock breaking instant element s _i The position angle theta of the corresponding test point; f (f) _σ,ε (θ) can be written as a fitted functional relation f of measured average stress values with respect to the position angle θ _σ (θ), and a fitted functional relation f of the measured average strain value with respect to the position angle θ _ε (θ) two forms;

then the coefficient matrix C is normalized to obtain a normalized coefficient matrix C ₁ The influence function shown in the formula (3) can be further rewritten as formula (6):

in the method, in the process of the invention,a similarity solution for the unknown column vector X; in fact, the _on>Proportional to the unknown column vector X.

When C ₁ In the case of an n-th order matrix, the above equation (6) has a solution in theory. However, in consideration of TBM cutting conditions, especially rolling hard or extremely hard rock under large cutting depths of hob, extremely severe step fracture exists in the edge-ground rockFeatures that result in higher randomness and dispersion of the acquired stress/strain data due to noise data being mixed into coefficient matrix C ₁ And in the column vector S, the solved contact force distribution is further caused to present an oscillation form, is severely distorted and cannot be used as an effective solution of an equation. From equation (6), the accuracy of the equation solution is largely determined by And (3) determining. For example, taking a certain real test data as an example, for coefficient matrix C with different orders ₁ Characteristic analysis was performed to obtain a determinant det (C) shown in Table 1 below ₁ ) Condition number cond (C ₁ )：

TABLE 1 different order matrix C ₁ Determinant value and condition number size of (c)

Coefficient matrix C ₁ Order n of (2)	det(C ₁ )	cond(C ₁ )
			15	1.70e-8	431.89
25	4.37e-22	3.10e3
			50	2.82e-61	7.34e3
75	5.34e-90	9.75e3
			100	5.19e-107	2.29e4
125	6.92e-145	1.01e5

As can be seen from Table 1 above, with matrix C ₁ Is increased in order, determinant det (C ₁ ) The sharp decrease approaches 0, while the condition number cond (C ₁ ) Gradually increasing, far above the critical value 100, which is a pathological problem, all result in a coefficient matrix C ₁ The inversion is extremely unstable, so that the data sample characteristics introduced by the remarkable step breaking characteristic during hob rock breaking are combined, and further data processing from step 7.3 to step 7.5 is performed, wherein the specific process is as follows:

step 7.3: obtaining a normalized coefficient matrix C by adopting Singular Value Decomposition (SVD) ₁ The singular value decomposition formula of (2) is shown in the following formula (7):

C ₁ ＝UDV ^T (7)

in the formula, a rank (C) ₁ ) =k, then there is an n-order orthogonal matrix U whose columns are defined byIs set to be u= (U) ₁ ,u ₂ ,…,u _n ) There is an n-order orthogonal matrix V whose rows are defined by +.>Is of the eigenvector composition of v= (V) ₁ ,v ₂ ,…,v _n )，Σ _k ＝diag(σ ₁ ,σ ₂ ,…,σ _k ) And->Eigenvalue lambda ₁ ≥λ ₂ ≥…λ _k > 0 is the matrix->All non-zero eigenvalues, sigma _i (i=1, 2, …, k) is C ₁ Is a singular value of (c).

The least squares solution of the equation shown in the formula (6) is given by the singular value decomposition formula shown in the following formula (8):

wherein:is a least squares solution to the equation shown in equation (6).

Step 7.4: in order to ensure that the equation solving quality has high reliability, comprehensively considering the influence of noise pollution on true solution X and the influence of solution resolution on solving precision, determining the maximum truncated parameter recommended value k of the singular value by using an L curve method ₀ . The principle of the method is as follows: for different regularized truncation parameters k ₀ Drawing different normsAnd regularizing the solution normsA graph therebetween; in general, as shown in FIG. 13, the graph presents an L-shape, with the inflection point of the curve being the maximum recommended value k for the regularized cutoff parameter ₀ For example, the cut-off coefficient k can be found from the inflection point and abscissa data of fig. 13 ₀ 10.

Step by step7.5: based on truncated singular value regularization (TSVD) method, reducing k correction to k ₁ Least squares solution of unknown column vector X by truncating part of singular valuesThe truncated singular value expression can be approximated as shown in the following formula (9):

in which k is 1.ltoreq.k ₁ ≤k ₀ 。

Preferably, the truncation parameter k ₁ Not more than 10.

Preferably, the coefficient matrix C is normalized for the purpose of both the test accuracy and the chip man-hour efficiency ₁ The order n of (2) is not greater than 100.

Step 8: let kappa be the correction coefficient, let X ₀ For the correction solution of X, then X ₀ Represented by the following formula (10):

let S ₀ To correct the solution X ₀ Substituting the theoretical stress/strain value column vector obtained by calculating the influence function shown in the formula (3), S ₀ Can be represented by the following formula (11):

in the example, based on the cyclic trial and error principle, MATLAB scientific calculation software is utilized to realize the purpose of realizing the cyclic automatic implementation of different cutoff parameters k through programming ₁ Substituting different correction coefficients kappa into the (11) to obtain a theoretical stress/strain value column vector S with different moments of rock breaking by rolling ₀ The theoretical column vector S ₀ Is composed of theoretical stress/strain values corresponding to different position angles theta; subsequently, the theoretical stress/strain at each position angle θ is determinedComparing the variable value with the actual stress/strain value one by one, and when the total error is minimum, the corresponding correction coefficient kappa and cutoff parameter k are obtained ₁ Is the optimal parameter; combining (9) and (10) a set of correction solutions X closest to the true value X is calculated from the optimal parameters ₀ 。

The foregoing is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art, who is within the scope of the present invention, should make equivalent substitutions or modifications according to the technical scheme of the present invention and the inventive concept thereof, and should be covered by the scope of the present invention.

Claims

1. A method for testing bottom contact force characteristics of a large-cutting-depth lower rolling blade is characterized by comprising the following steps of:

the method is matched with a bottom contact force characteristic test system of a large-cutting-depth lower rolling blade;

the system for testing the bottom contact force characteristics of the large-cutting-depth lower rolling blade comprises a hob standard linear cutting experiment table, a data acquisition instrument, a hob, an eddy current displacement sensor and an industrial personal computer;

the hob adopts the customization hob, including customization cutter ring, customization cutter body, customization arbor, customization bearing and customization end cover, characterized by: the cross-section size of the cutting edge part of the custom knife ring below the maximum cutting depth is the same as that of the standard knife ring, and the custom knife ring is provided with side vertical planes at the left side and the right side of the cutting edge part above the maximum cutting depth by To lay out strain gauges; the material, manufacturing process and cutter ring diameter of the customized cutter ring are consistent with those of the standard cutter ring; a given radial arrangement distance h on the side vertical plane of the blade ₁ Arranging the strain gauges according to equal circumferential interval angles delta theta, and arranging the strain gauges on the side vertical planes of the custom cutter ring in a whole circumference manner to form corresponding test points so as to monitor the stress/strain state of the custom cutter ring in real time;

the data acquisition instrument can acquire output signals of the three-way force sensor, the strain gauge and the eddy current displacement sensor in real time and transmit the output signals to the industrial personal computer so as to process and analyze;

the method comprises the following steps:

step 3: the method comprises the following steps of:

step 3.1: calculating the measured average calibration vertical force;

step 5: rolling and rock breaking test; setting a cutting depth h and rock types in the same step 1, setting a uniform angular speed in the same step 2 by a hob, and carrying out a rolling rock breaking test in a single-blade cutting mode; synchronously recording a three-way cutting force signal curve output by the three-way force sensor, a strain signal curve output by the test point and a voltage signal curve output by the eddy current displacement sensor in real time by using the large-cutting-depth lower rolling blade bottom contact force characteristic test system;

step 7.1: the theoretical contact area of the edge bed knife rock is discretized into n-1 equal parts according to equal circumferential interval angles, n=360 degrees/delta theta, so that the contact force components on the unit knife edge widths received by the endpoints of the theoretical contact area of the edge bed knife rock in equal parts at any rolling rock breaking moment according to the order of the position angles theta from large to small are sequentially x ₁ 、x ₂ 、…、x _n And constitutes a column vector X; the stress/strain value readings sequentially output by the test points at the moment of rock breaking by rolling are s according to the sequence of the position angles theta from small to large ₁ 、s ₂ 、…、s _n And constitutes a column vector S; according to any rolling rock breaking moment, the stress/strain value s ₁ 、s ₂ 、…、s _n Component x of contact force with unit blade width ₁ 、x ₂ 、…、x _n The mapping relation between the two is defined as an influence function shown in the following formula (1):

CX＝S (1)

wherein, C is a coefficient matrix of an influence function;

step 7.2: fitting the change curve of the measured average stress/strain value obtained by step 4.3 along with the position angle theta to obtain a fitting function relation f of the measured average stress/strain value relative to the position angle theta _σ,ε (θ) constructing the given rolling break moment s as described in step 7.1 ₁ 、s ₂ 、…、s _n Coefficient matrix C of the corresponding influence function corresponding to each measuring point position angle theta one by one and forming element C _ij The form of (2) is as follows:

in the method, in the process of the invention,a similarity solution for the unknown column vector X;

step 7.3: obtaining a normalized coefficient matrix C by adopting a singular value decomposition method ₁ The singular value decomposition formula of (2) is shown in the following formula (4):

C ₁ ＝UDV ^T (4)

in the design ofThere is an n-order orthogonal matrix U whose columns are made up of +.>Is set to be u= (U) ₁ ,u ₂ ,…,u _n ) There is an n-order orthogonal matrix V whose rows are defined by +.>Is of the eigenvector composition of v= (V) ₁ ,v ₂ ,…,v _n )，Σ _k ＝diag(σ ₁ ,σ ₂ ,…,σ _k ) And->Eigenvalue lambda ₁ ≥λ ₂ ≥…λ _k > 0 is the matrix->Is not zero, sigma _i (i=1, 2, …, k) is C ₁ Is a singular value of (2);

Wherein:a least squares solution to the equation shown in equation (3);

step 8: let kappa be the correction coefficient, let X ₀ For the correction solution of X, then X ₀ Represented by the following formula (6):

let S ₀ To correct the solution X ₀ Substituting the theoretical stress/strain value column vector obtained by calculating the influence function shown in the formula (1), S ₀ Can be represented by the following formula (7):

based on cyclic trial and error principleSubstituting different correction coefficients kappa into the formula (7) to obtain different theoretical stress/strain value column vectors S ₀ The column vector S ₀ Is composed of theoretical stress/strain values corresponding to different position angles theta; comparing the theoretical stress/strain value with the actual measured stress/strain value one by one under each position angle theta, and when the total error is minimum, taking the corresponding correction coefficient kappa as an optimal parameter; combining (5) and (6), obtaining a set of correction solutions X closest to the true value X from the optimal parameter calculation ₀ 。

2. The method for testing the bottom contact force characteristics of the large-cutting-depth lower rolling blade according to claim 1, wherein the step 7 further comprises the following sub-steps:

step 7.4: determining the maximum truncated parameter recommended value k of the singular value by using an L-curve method ₀ ；

Step 7.5: based on a truncated singular value regularization method, reducing k correction to a truncated parameter k ₁ Least squares solution for unknown column vector XThe truncated singular value expression can be approximated as shown in the following equation (8):

in which k is 1.ltoreq.k ₁ ≤k ₀ 。

3. The method for testing the bottom contact force characteristics of a large-cutting-depth rolling blade according to claim 2, wherein different cutoff parameters k are based on a cyclic trial-and-error principle ₁ Substituting different correction coefficients kappa into the formula (7) to obtain different theoretical stress/strain value column vectors S ₀ The column vector S ₀ Is composed of theoretical stress/strain values corresponding to different position angles theta; comparing the theoretical stress/strain value with the actual stress/strain value one by one under each position angle theta, and when the overall error occursWhen the difference is minimum, the corresponding correction coefficient k and the truncation parameter k ₁ Is the optimal parameter; combining (5) and (6), obtaining a set of correction solutions X closest to the true value X from the optimal parameter calculation ₀ 。

4. The method according to any one of claims 1, 2 and 3, wherein the set nominal vertical force in step 2 is not smaller than the average estimated vertical force obtained in step 1.

5. The method for testing bottom contact force characteristics of large-cutting-depth rolling blade according to any one of claims 1, 2 and 3, wherein the step 7 uses radial stress/strain values to construct a relation of influence functions.

6. The method for testing bottom contact force characteristics of large-cutting-depth rolling blades according to any one of claims 1, 2 and 3, wherein the step 7 uses radial stress values to construct a relation of influence functions.

7. A method for testing bottom contact force characteristics of large-cutting-depth rolling blade according to any one of claims 2 or 3, wherein said cutoff parameter k ₁ Not more than 10; the normalized coefficient matrix C ₁ The order n of (2) is not greater than 100.

8. The method for testing bottom contact force characteristics of a large-cutting-depth lower rolling blade according to claim 1, wherein a protective cover is fixedly arranged above the strain gauge.

9. The method for testing bottom contact force characteristics of a large-cutting-depth lower rolling blade according to claim 1, wherein strain gauges are symmetrically arranged on side vertical planes on the left side and the right side of the customized cutter ring respectively, and a full-bridge testing circuit is correspondingly formed.