CN102927930A - Method for detecting ultra-large-diameter reflector surface errors in splicing mode by adopting collimator - Google Patents

Abstract

The invention relates to a method for detecting large-diameter reflectors, in particular to a method for detecting ultra-large-diameter reflector surface errors in splicing mode by adopting a collimator, and solves the problems that the existing sub-aperture splicing method needs hundreds of and even thousands of plane mirrors when measuring large-diameter plane mirrors, especially ultra-large-diameter plane mirrors in splicing mode, causes error accumulation and is long in test time. The method for detecting ultra-large-diameter reflector surface errors in splicing mode by adopting the collimator is characterized in that standard spherical waves emitted by an interferometer pass through the large-diameter collimator and then are converted to be large-diameter planer waves to be emitted out, the relative position of an emitting pupil and the large-diameter reflector to be detected is accurately adjusted, emitting wave surfaces are gradually emitted into all sub-aperture areas of the large-diameter reflector to be detected and return to the interferometer, and accordingly all sub-aperture phase distribution data are obtained; and a computer extracts all the sub-aperture phase distribution data measured by the interferometer and obtains full aperture surface error distribution of the large-diameter reflector to be detected through triangulation and global optimization splicing algorithm. The method achieves surface detection for ultra-large-diameter reflectors and is low in test cost and high in efficiency.

Description

Adopt the parallel light tube splicing to detect the method for super large caliber reflecting mirror surface shape error

Technical field

The present invention relates to a kind of detection large caliber reflecting mirror, be specifically related to the detection method of super large caliber plane mirror face shape error.

Background technology

Along with the development of space optics technology, large caliber reflecting mirror just more and more is used in the various optical systems.And along with improving constantly of optical system resolution index, require the surface figure accuracy of large caliber reflecting mirror to be better than 1/50 λ rms (λ=632.8nm).

The process of large caliber reflecting mirror mainly comprises milling, grinding and polishing three phases.Catoptron for milling and process of lapping can utilize three-dimensional or contourgraph that its face shape is carried out contact type measurement, can obtain the distributed intelligence of optical surface shape by data analysis, thus for its further processing foundation and guarantee are provided.For the surface shape measurement of catoptron polishing and terminal stage, contact type measurement will be no longer applicable, because its measuring accuracy generally can only reach 1/3 λ rms, this external this stage utilizes contact type measurement to bring certain cut to minute surface.

Because interference detection has the advantages such as high-resolution, high precision, high sensitivity, good reproducibility, so this technology has become the large caliber reflecting mirror polishing and terminal stage face shape is detected method the most commonly used.General using Rui Qikangmangfa realized the interference of heavy-calibre planar catoptron is detected in the past, and the auspicious strange angle of the method produces the detected image compression, so that the astigmatism of the face shape error on tested plane and sphere itself mixes, affects the accuracy of testing result; Large caliber reflecting mirror deforms at the pupil plane of detection system simultaneously, makes the corresponding relation complicated between mirror shape error and the system's emergent pupil optical path difference, and is different with the difference at auspicious strange angle, causes the data processing of testing result to become very complicated.In addition, Rui Qikangmangfa needs bigbore auxiliary spherical surface mirror, the bore of auxiliary spherical surface mirror is generally 1.2-1.3 times of large caliber reflecting mirror minute surface size to be measured, the manufacturing difficult of heavy caliber, high-precision spherical reflector, therefore, the Rui Qikangmangfa of heavy-calibre planar catoptron detects and is difficult to realize.

Utilize small-bore interferometer one by one phase measurement to be carried out in each zone of large caliber reflecting mirror, utilizing sub-aperture stitching to finish detects the unified face shape of large caliber reflecting mirror, but the splicing for the level crossing of heavy caliber especially super large caliber is measured, the number in sub-aperture needs individual even thousands of of hundreds of, this will bring a lot of error accumulations and test duration very long, be subjected to the impact of environment very large, make the method realize existing a lot of difficulties, as shown in Figure 1, take the super large caliber catoptron of Φ 2m bore as example, utilize conventional at present bore to splice 1261 sub-apertures of measurement needs for the Zygo digital interference instrument of Φ 100mm.

Summary of the invention

The present invention solves to have now owing to the level crossing that adopts the sub-aperture stitching method to heavy caliber especially super large caliber splices to need hundreds of even thousands of individual when measuring, cause long problem of error accumulation and test duration, a kind of method that adopts the parallel light tube splicing to detect super large caliber reflecting mirror surface shape error is provided.

Adopt the parallel light tube splicing to detect the method for super large caliber reflecting mirror surface shape error, the method is realized by following steps:

Step 1, interferometer are converted into the standard flat ripple by the standard ball ground roll of penetrating behind the heavy caliber parallel light tube, place high-precision standard flat catoptron at the emergent pupil place of described plane wave, position by accurate adjustment standard flat catoptron and heavy caliber parallel light tube, the plane wave of heavy caliber parallel light tube outgoing is received by interferometer behind the standard flat catoptron, and described interferometer records the aberration profile data of heavy caliber parallel light tube;

High-precision standard flat catoptron in step 2, the employing large caliber reflecting mirror replacement step one to be measured, be converted into bigbore plane wave behind the standard ball ground roll process heavy caliber parallel light tube of described interferometer outgoing, adjust the emergent pupil of bigbore plane wave and the relative position of large caliber reflecting mirror to be measured, return interferometer after making the plane wave of outgoing incide one by one each sub-aperture area of large caliber reflecting mirror to be measured; The data that adopt interferometer to record deduct the aberration profile data of the heavy caliber parallel light tube that step 1 records; Obtain the PHASE DISTRIBUTION data of each sub-aperture area of large caliber reflecting mirror to be measured;

Step 3, computing machine extract each sub-aperture phase distributed data of the large caliber reflecting mirror to be measured of interferometer measurement acquisition, and the PHASE DISTRIBUTION the data triangulation in each sub-aperture and the stitching algorithm of global optimization are obtained the unified face shape error of large caliber reflecting mirror to be measured.

Beneficial effect of the present invention: the method that employing parallel light tube splicing of the present invention detects super large caliber reflecting mirror surface shape error adopts the heavy caliber parallel light tube, adopt the parallel light tube of Φ 1m bore to splice the super large caliber catoptron that detects Φ 2m bore, only need 9 sub-apertures; Use the high-precision face shape of finishing the super large caliber catoptron of stitching interferometer measurement technology to detect, testing cost is low, efficient is high.

Description of drawings

Fig. 1 is the synoptic diagram that the super large caliber catoptron digital interference instrument of 2m bore in the prior art splices;

Fig. 2 is the splicing synoptic diagram that the method for employing parallel light tube splicing detection super large caliber reflecting mirror surface shape error of the present invention is spliced the super large caliber catoptron of 2m bore;

Fig. 3 is the apparatus structure synoptic diagram that employing parallel light tube splicing of the present invention detects the method for super large caliber reflecting mirror surface shape error;

Fig. 4 is that employing parallel light tube of the present invention splices parallel light tube caliberating device synoptic diagram in the method that detects super large caliber reflecting mirror surface shape error;

Fig. 5 is the process flow diagram that employing parallel light tube splicing of the present invention detects the method for super large caliber reflecting mirror surface shape error.

Among the figure, 1, interferometer, 2, the heavy caliber parallel light tube, 3, large caliber reflecting mirror, the 4, first adjusting mechanism, the 5, second adjusting mechanism, 6, the standard flat catoptron.

Embodiment

Embodiment one, in conjunction with Fig. 2 to Fig. 4 present embodiment is described, adopt the parallel light tube splicing to detect the method for super large caliber reflecting mirror surface shape error, comprise interferometer 1, heavy caliber parallel light tube 2, the first adjusting mechanism 4, the second adjusting mechanism 5, computing machine and large caliber reflecting mirror to be measured 3 in the method; Be converted into bigbore plane wave front outgoing behind the standard ball ground roll process heavy caliber parallel light tube 2 of described interferometer 1 outgoing, relative position by accurate adjustment emergent pupil and large caliber reflecting mirror to be measured 3, so that the outgoing corrugated is incided one by one each sub-aperture area of large caliber reflecting mirror 3 to be measured and returned interferometer 1, obtain thus each sub-aperture phase distributed data; Computing machine extracts each sub-aperture phase distributed data that interferometer 1 is measured, and it is analyzed and processes, and the stitching algorithm by triangulation and global optimization obtains large caliber reflecting mirror 3 unified face shape errors to be measured and distributes.

In conjunction with Fig. 2, the present invention utilizes the parallel light tube of 1m bore to splice detection and only needs 9 sub-apertures.

In conjunction with Fig. 3 and Fig. 4 present embodiment is described, interferometer 1 is placed on the first adjusting mechanism 4, and the rotation by adjusting its three degree of freedom and the translation of three directions make interferometer 1 aim at heavy caliber parallel light tube 2.Large caliber reflecting mirror 3 to be measured is installed on the second adjusting mechanism 5, can make large caliber reflecting mirror 3 arbitrary regions to be measured aim at the emergent pupil of heavy caliber parallel light tube 2 by adjusting it along the translation of X-axis and Y direction and around the rotation of X-axis and Y-axis; Coordinate system definition as shown in Figure 3; Described the first adjusting mechanism 4 adopts present known sextuple accurate adjusting mechanism, pitching that can accurate adjustment interferometer 1, rock and rotate and along directions X, Y-direction with along the translation of Z direction, the second adjusting mechanism 5 adopts known four-dimensional accurate adjusting mechanism, can accurately control large caliber reflecting mirror 3 to be measured along the translation of directions X, Y-direction with around the rotation of X-axis, Y-axis.

The spherical wave front of described interferometer 1 outgoing is through inciding with the standard flat waveshape behind the flat shape light pipe from axle on 3 of the large caliber reflecting mirrors to be measured, the standard flat ripple turns back on the CCD image planes of interferometer 1 along former road after large caliber reflecting mirror 3 reflections to be measured, accurate this regional PHASE DISTRIBUTION of measuring, each regional PHASE DISTRIBUTION on the sequentially determining large caliber reflecting mirror 3 can obtain the unified face shape error information of large caliber reflecting mirror to be measured 3 by the sub-aperture stitching algorithm.

Embodiment two, in conjunction with Fig. 5 present embodiment is described, present embodiment is the embodiment that embodiment one described employing parallel light tube splicing detects the method for super large caliber reflecting mirror surface shape error:

Concrete step is as follows:

One, parallel light tube is demarcated; Before the sub-inside diameter measurement to large caliber reflecting mirror, at first will demarcate heavy caliber parallel light tube 2, the caliberating device synoptic diagram of heavy caliber parallel light tube 2 is as shown in Figure 4; The standard ball ground roll of laser interferometer outgoing is converted into the standard flat wavefront behind heavy caliber parallel light tube 2, place a high-precision standard flat catoptron 6 at its emergent pupil place, standard flat catoptron 6 is placed on the second adjusting mechanism 5, make the standard flat catoptron aim at heavy caliber parallel light tube 2 by accurate adjustment, the emerging wavefront of heavy caliber parallel light tube 2 returns along former road after 6 reflections of standard flat catoptron, finally received by the CCD of interferometer, because the precision of standard flat catoptron 6 is very high, general its face shape PV value is better than 1/10 λ, λ is wavelength, therefore measure the aberration profile that the PHASE DISTRIBUTION that obtains can be used as heavy caliber parallel light tube 2, be made as w _P

Two, plan sub-aperture; According to the outgoing beam size of the bore of large caliber reflecting mirror 3 to be measured and heavy caliber parallel light tube 2, divide sub-aperture area and calculate the size in sub-aperture.In order to find the solution the splicing coefficient, should there be certain overlapping region in each sub-aperture, and in order to guarantee solving precision, the overlapping region area is generally greater than 1/4 of sub-aperture area.

Three, the measurement of sub-aperture phase data; Remove standard flat catoptron 6, at the second adjusting mechanism 5 large caliber reflecting mirror 3 to be measured is installed, by the second adjusting mechanism 5 accurate adjustments large caliber reflecting mirror 3 to be measured, is made its each sub-aperture area aim at heavy caliber parallel light tube 2, thereby can measure each regional PHASE DISTRIBUTION, be made as z _i, i=1～N wherein, N is the number in sub-aperture.

Four, the processing of each sub-aperture phase data; In the process of measuring every sub-aperture plane shape phase place, because the impact of the factors such as impact, the minute surface roughness of tested person environment are inhomogeneous, testing tool precision, so that sub-aperture data certainly exist disappearance, therefore, adopt triangulation to find the solution and fill up sub-aperture phase data.

Before calculating the splicing coefficient, at first with the measurement data unification (mode that reads by mechanical motion or target in the actual measurement is determined the relative position between each sub-aperture) in overall reference frame in each sub-aperture, be defined in the interpolation point (scope of interpolation point is greater than the size of minute surface under the world coordinates) under the world coordinates, the measurement data of antithetical phrase aperture i is carried out the Delaunay triangulation, thereby obtain (the x of predefined data point, y) surface shape value, that is: utilize define a plane at 3, plane equation is ax+by+z+d=0 (coefficient of z is 0 scarcely), utilize equation (1) to solve coefficient a, b, d:

$\left(\begin{array}{ccc}{x}_1& y_1& 1\\ x_2& y_2& 1\\ x_3& y_3& 1\end{array}\right)\left(\begin{array}{c}a\\ b\\ d\end{array}\right)=\left(\begin{array}{c}{-z}_1\\ {-z}_2\\ {-z}_3\end{array}\right)---\left(1\right)$

Obtain the determined plane equation z ' of these three points=-ax-by-d, the coordinate of point (x, y) is brought in this plane equation goes, can obtain for i sub-aperture, the surface shape value z ' of this interpolation point.For not at the point of Delaunay triangle inside, adopt the closest approach interpolation processing, namely the surface shape value z value of this point is got the z value in closest approach.

Mode by Delaunay triangulation and closest approach interpolation obtains the surface shape value of every sub-aperture interpolation points, has namely obtained phase value and the complete phase data distribution z in each sub-aperture of data point in the overlapping region, aperture separately _i

With each sub-aperture phase data F _iDeduct the systematic error of parallel light tube, can obtain the face shape error w in each sub-aperture _i, w _i=z ' _i-W _P

Five, unified complex optimum splicing; Distribute in order accurately to obtain unified face shape error, the present invention will adopt the global optimization stitching algorithm of error homogenizing, and concrete analysis and solution procedure are as follows.

Find the solution the relative alignment error in adjacent two sub-apertures by the data of overlapping region, can realize that sub-aperture splices in twos.Repeatedly utilize the splicing principle in two sub-apertures just can realize the splicing in a plurality of sub-apertures.But tend to like this cause propagation of error and accumulation, thereby reduced the accuracy of detection in whole aperture.Therefore, the problem that in the sub-aperture stitching process, has complex optimum.When splicing regions during greater than two, suppose total M sub-aperture stitching, M is positive integer, can select first wherein the anyon aperture and for the ease of locating and measurement, generally select the sub-aperture of large caliber reflecting mirror central area as the reference standard as benchmark.

Owing to be zero compensation to be carried out in each zone of large caliber reflecting mirror measure, so the relative detuning amount of position is only brought relative translation and inclination between each sub-aperture.If the PHASE DISTRIBUTION in the sub-aperture of benchmark is w ₀, the pass that then other sub-aperture phase distributes and the sub-aperture phase of benchmark distributes is:

w ₀=w ₁+p ₁+a ₁x ₁+b ₁y ₁

=w ₂+p ₂+a ₂x ₂+b ₂y ₂

=w _M-1+p _M-1+a _M-1x _M-1+b _M-1y _M-1 (2)

W wherein ₁, w ₂..., w _M-1The PHASE DISTRIBUTION (all sub-aperture phase values have all been rejected the aberration profile of parallel light tube itself) in other sub-aperture, a _i, b _iAnd p _iIt is respectively other sub-aperture of relative datum, sub-aperture inclination factor, inclination factor in the y-direction and relative translation coefficient in the x-direction.

Utilize least square method, so that the quadratic sum value of all overlapping region phase differential is minimum, can get following formula (3):

$S=\underset{j_1\≠0}{\overset{N_1}{\mathrm{\Σ}}}\underset{i_1\⋐W_0,W_{j_1}}{\overset{n}{\mathrm{\Σ}}}{\{W_0(x_{{1i}_1},y_{{1i}_1})-[W_{j_1}(x_{j_1{i}_1},y_{j_1{i}_1})+p_{j_1}{x}_{j_1{i}_1}+a_{j_1}{x}_{j_1{i}_1}+b_{j_1}{y}_{j_1{i}_1}\left]\right\}}^2$

$+\underset{j_2\∩j_3\≠0}{\overset{N_2}{\mathrm{\Σ}}}\underset{i_2\⋐W_{j_2},W_{j_3}}{\overset{n}{\mathrm{\Σ}}}\left\{\right[W_{j_2}(x_{j_2{i}_2},y_{j_2{i}_2})+p_{j_2}{x}_{j_2{i}_2}+a_{j_2}{x}_{j_2{i}_2}+b_{j_2}{y}_{j_2{i}_2}]$

$-{[W_{j_3}(x_{j_3{i}_2},y_{j_3{i}_2})+p_{j_3}{x}_{j_3{i}_2}+a_{j_3}{x}_{j_3{i}_2}+b_{j_3}{y}_{j_3{i}_2}]\}}^2=\min$

N in the formula ₁The overlapping region number in other sub-aperture and the sub-aperture of benchmark, N ₂Be the overlapping region number between other sub-aperture, n is the sampling number in the overlapping region.

Utilize least square fitting, each coefficient asked respectively local derviation and made that its value is zero can get:

$\left\{\begin{array}{c}\frac{\∂S}{{\∂p}_i}=0\\ \frac{\∂S}{{\∂a}_i}=0\\ \frac{\∂S}{{\∂b}_i}=0\end{array}\right.---\left(4\right)$

1≤i in the formula≤M-1, (3) formula of employing just can obtain the best splicing factor in each sub-aperture of relative datum, sub-aperture, thereby the position phase data in all sub-apertures is proofreaied and correct unified to identical benchmark.

Because the splicing factor p in the sub-aperture of benchmark ₀, a ₀, b ₀Be not included in the equation, so the unknowm coefficient that altogether needs to find the solution is 3 * (M-1) individual.(4) formula is analyzed and found the solution, and the least square equation that can get finally is (5) formula:

$\left[{\left(\underset{k}{\overset{M-1}{\mathrm{\Σ}}}{G}_{\mathrm{ik}}\right)}_i\right]=\left[{\left(Q_{\mathrm{ij}}-{\mathrm{\δ}}_{\mathrm{ij}}\underset{k}{\overset{M-1}{\mathrm{\Σ}}}{Q}_{\mathrm{ik}}\right)}_{\mathrm{ij}}\right]\left[{\left(R_i\right)}_i\right]$

Wherein the expression formula of submatrix G, Q, R is as follows:

$G_{\mathrm{ij}}=\left[\begin{array}{c}\underset{i\∩j}{\mathrm{\Σ}}\mathrm{x\Δw}\\ \underset{i\∩j}{\mathrm{\Σ}}\mathrm{y\Δw}\\ \underset{i\∩j}{\mathrm{\Σ}}\mathrm{\Δw}\end{array}\right]---\left(6\right)$ $Q_{\mathrm{ij}}=\left[\begin{array}{ccc}\underset{i\∩j}{\mathrm{\Σ}}\mathrm{xx}& \underset{i\∩j}{\mathrm{\Σ}}\mathrm{xy}& \underset{i\∩j}{\mathrm{\Σ}}x\\ \underset{i\∩j}{\mathrm{\Σ}}\mathrm{yx}& \underset{i\∩j}{\mathrm{\Σ}}\mathrm{yy}& \underset{i\∩j}{\mathrm{\Σ}}y\\ \underset{i\∩j}{\mathrm{\Σ}}x& \underset{i\∩j}{\mathrm{\Σ}}y& n_{\mathrm{ij}}\end{array}\right]---\left(7\right)$

$Q_{\mathrm{ii}}=\left[\begin{array}{ccc}0& 0& 0\\ 0& 0& 0\\ 0& 0& 0\end{array}\right]---\left(8\right)$ $R_i=\left[\begin{array}{c}{p}_i\\ a_i\\ b_i\end{array}\right]---\left(9\right)$ ${\mathrm{\δ}}_{\mathrm{ij}}=\left\{\begin{array}{cc}1& i=j\\ 0& i\≠j\end{array}\right.---\left(10\right)$

N in the formula _IjSampling number for each overlapping region; I and j are the integers from 1 to M-1; K is the integer from 0 to M-1; Summation in formula (6), (7) is carried out in the overlapping region, if any two sub-aperture w _iAnd w _jThere is not lap, then submatrix G _Ij, Q _IjBe null matrix.

Utilize equation (5)-(10) to try to achieve submatrix R _i, just can obtain the best splicing factor in each sub-aperture of relative datum, sub-aperture, thereby the position phase data in all sub-apertures be proofreaied and correct unified to identical reference surface.More than be a kind of sub-aperture stitching mode of global optimization, the splicing cumulative errors can be opened in whole splicing regions " diffusion ", played the effect of error homogenizing, thereby can reduce widely to splice cumulative errors, realize well the splicing in a plurality of sub-apertures.

Six, error and precision analysis; The method of utilizing the parallel light tube splicing to detect the super large caliber catoptron is subjected to the impact of several factors: such as impacts such as interferometer precision, parallel light tube stated accuracy, sub-aperture stitching arithmetic accuracy, institutional adjustment precision and test environments.We will carry out precision analysis and differentiate with it obtaining accurately face shape error distribution of large caliber reflecting mirror to testing result.

The present invention carries out precision analysis to testing result and differentiates obtaining accurately with it that aspherical reflector surface shape error to be measured distributes, and accuracy of detection can reach λ/more than 50, λ=632.8nm.

Claims (3)

1. adopt the parallel light tube splicing to detect the method for super large caliber reflecting mirror surface shape error, it is characterized in that the method is realized by following steps:

The standard ball ground roll of step 1, interferometer (1) outgoing is converted into the standard flat ripple behind heavy caliber parallel light tube (2), place high-precision standard flat catoptron (6) at the emergent pupil place of described plane wave, by the position of accurate adjustment standard flat catoptron (6) with heavy caliber parallel light tube (2), the plane wave of heavy caliber parallel light tube (2) outgoing is received by interferometer (1) behind standard flat catoptron (6), and described interferometer (1) records the aberration profile data of heavy caliber parallel light tube (2);

Step 2, adopt high-precision standard flat catoptron (6) in large caliber reflecting mirror to be measured (3) replacement step one, be converted into bigbore plane wave behind the standard ball ground roll process heavy caliber parallel light tube (2) of described interferometer (1) outgoing, adjust the emergent pupil of bigbore plane wave and the relative position of large caliber reflecting mirror to be measured (3), return interferometer (1) after making the plane wave of outgoing incide one by one each sub-aperture area of large caliber reflecting mirror to be measured (3); The data that adopt interferometer (1) to record deduct the aberration profile data of the heavy caliber parallel light tube (2) that step 1 records; Obtain the PHASE DISTRIBUTION data of each sub-aperture area of large caliber reflecting mirror to be measured (3);

Step 3, computing machine extract each sub-aperture phase distributed data that interferometer (1) is measured the large caliber reflecting mirror to be measured (3) that obtains, and the PHASE DISTRIBUTION the data triangulation in each sub-aperture and the stitching algorithm of global optimization are obtained the unified face shape error of large caliber reflecting mirror to be measured (3).

2. employing parallel light tube splicing according to claim 1 detects the method for super large caliber reflecting mirror surface shape error, it is characterized in that, described interferometer (1) is fixed on the first adjusting mechanism (4), by adjusting the first adjusting mechanism (4) interferometer (1) is aimed at heavy caliber parallel light tube (2).

3. employing parallel light tube splicing according to claim 1 detects the method for super large caliber reflecting mirror surface shape error, it is characterized in that, described large caliber reflecting mirror to be measured (3) is fixed on the second adjusting mechanism (5), by adjusting the second adjusting mechanism (5) along the translation of X-axis and Y direction and the emergent pupil that makes the arbitrary region of large caliber reflecting mirror to be measured (3) aim at heavy caliber parallel light tube (2) around the rotation of X-axis and Y-axis.

Images