JP2004101247A - Testing method for tooth profile measuring device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a testing method for a tooth profile measuring device unaffected by a deviation in shape of a testing master. <P>SOLUTION: The testing master 16 is fitted to the tooth profile measuring device as a test object, the radius r<SB>b</SB>of the base circle is set normal (a=0), a tested surface 16a in the same portion of the testing master 16 is automatically measured two or more times, the radius of the base circle is automatically set larger than normal (a=+dr<SB>b</SB>), the tested surface 16a in the same portion of the testing master 16 is measured two or more times, and the radius of the base circle is automatically set smaller than normal (a=-dr<SB>b</SB>), the tested surface 16a in the same portion of the testing master 16 is automatically measured two or more times. The physical contact position of a contact 15 of the tooth profile measuring device contacting the tested surface 16a of the testing master 16 is not moved still in the state of setting the radius of the base circle of the testing master 16 normal (2=0), and the changed value of the radius of the base circle is used only for computing the command data for numerical value controlled drive to determine the translational position of the contact 15 displaced only in the tangential direction of the base circle. <P>COPYRIGHT: (C)2004,JPO

Description

【００８５】
さらに、図７に示されるように、ある測定条件｛ｋ，ａ｝における測定値偏差の、θ点での平均値からのずれｆ^＊＊ _ｋａｉ（θ）を求める。
【００８６】

【００８７】
さらに、図８に示されるように、ある角度位置ｋの測定における測定値偏差の平均値からのずれの全データのばらつきの確率密度関数Ｐ_ｋ（ｆ^＊＊）を求める。
【００８８】

【００８９】
また、図８に示されるように、ある測定条件｛ｋ，ａ｝における測定値偏差の平均値からのずれの平均曲線Ｆ^＊＊ _ｋａ（θ）を求める。
【００９０】

【００９１】
この平均曲線Ｆ^＊＊ _ｋａ（θ）において、ある測定条件｛ｋ，ａ｝における測定値偏差の平均値からのずれの最大値

を求める。
【００９２】

【００９３】
さらに、この平均曲線Ｆ^＊＊ _ｋａ（θ）において、ある測定条件｛ｋ，ａ｝における測定値偏差の平均値からのずれの最小値

を求める。
【００９４】

【００９５】
また、図７に示されるように、ある測定条件｛ｋ，ａ｝における測定値偏差の平均値からのずれの、θ点での標準偏差ｓ_ｋａ（θ）を求める。
【００９６】

【００９７】
同様に、図８に示されるように、ある測定条件｛ｋ，ａ｝における測定値偏差の平均値からのずれの標準偏差Ｓ_ｋａを求める。
【００９８】

【００９９】
また、ある角度位置ｋの測定における、θ点測定値の標準偏差の平均値からのずれの最大値ｓ_ｋ ⁻（θ）を求める。
【０１００】

【０１０１】
さらに、ある角度位置ｋの測定における測定値偏差の平均値からのずれの標準偏差全体の最大値Ｓ_ｋを求める。
【０１０２】

【０１０３】
また、Ｆ^＊＊ _ｋａ（θ）を、

と曲線近似（ｃｕｒｖｅ　ｆｉｔｔｉｎｇ）し、係数ｂ_ｋａ、ｃ_ｋａ、ｄ_ｋａを求める。
【０１０４】
係数ｂ_ｋａは、測定における圧力角検出値の狂いの程度、ｃ_ｋａは、測定における２次成分の狂いの混入の程度、ｄ_ｋａは、測定における３次成分の狂いの混入の程度を表す指標値である。
【０１０５】
次に、検定結果を評価するために、図９乃至図１８に示されるような目に見える形で、具体的にビジュアル表示する。検定結果の評価例を説明する。
【０１０６】
種々の測定セットについて、基礎円半径の設定を正規、＋ｄｒ_ｂ、−ｄｒ_ｂにしたデータ（ａ＝０，＋ｄｒ_ｂ、−ｄｒ_ｂ）の比較結果は次のことを表す。
【０１０７】
（１）Ｆ_ｋａ（θ）曲線の凹凸の類似性は、検定用マスタ１６の形状偏差に対応するものと考えられる。
【０１０８】
（２）Ｆ_ｋａ（θ）曲線の相違は、一定の測定条件｛ｋ，ａ｝での測定値の、θを基準とした歯形測定磯の並進軸のＮＣ制御の不正確さを表し、また、｛ｂ_ｋａ、ｃ_ｋａ、ｄ_ｋａ｝の相違は、これを定量化したものである。
【０１０９】
（３）ｓ_ｋａ（θ）曲線の０（ｚｅｒｏ　ｖａｌｕｅ）からのずれは、歯形測定機の繰返し精度のθの依存性を表す。
【０１１０】
Ｓ_ｋａのａに関するばらつきは、一定の測定条件｛ｋ｝での測定の繰返し精度に及ぼすセンサ押付圧などの影響を表している。すなわち、（ａ＝０，＋ｄｒ_ｂ、−ｄｒ_ｂ）によるｓ_ｋａ（θ）曲線の相違は、わずかの接触圧（あるいは接触子１５の変位）によるセンサの繰返し精度の違いを表す。
【０１１１】
Ｓ_ｋは、検定用マスタ１６がこの角度位置に取付けられて測定する場合の繰返し精度を表している。
【０１１２】
（４）測定セット１と５との差は、繰返し精度と再現性の差の原因の一部を表す。
【０１１３】
測定値の標準偏差を表す図１１における、

は、繰返し精度の最悪値を表し、再現性の最良値を表すものである。
【０１１４】
（５）回転主軸１１ならびにロータリエンコーダ２６の異なる箇所を使用しての測定セット（回転位相角０，０．５π，π，１．５π，２π，：（ｋ＝１，…，５））の結果の差は、
・ロータリエンコーダ２６の使用場所によるデータの差
・ロータリエンコーダ２６と下部センタ２２を継ぐカップリング２４の影響
・コラムの倒れの影響
・回転主軸１１のセンタ穴の狂いの影響
を表す。
【０１１５】
これらの評価を行うための具体的表示は、図９〜図１８に示されるような表形式またはグラフ形式で表示する。
【０１１６】
なお、ｋ＝１として、ロータリエンコーダ２６やカップリング２４の不整に関する影響を取込まない状態で簡易的に校正することもある。
【０１１７】
以上のように、本検定法は、各種誤差要因の全てに起因する誤差を総合的に含んだ歯形形状偏差の測定データであっても、それがその真の値からどの程度ずれている範囲に収まっているかを、工業的実用性の観点から明らかにするもので、生産歯車の歯形精度検査結果を保証するため、その測定を行った歯形測定機の測定精度を保証する歯形測定機の検定データの１つの見本を提案するものである。
必要な折りには、この検定データは、その生産歯車精度検査結果の添付資料として用いられるべきものである。また、この検定データは、歯形測定機の日頃のメインテナンスを正しく行うための補助に用いることができる。
【０１１８】
【発明の効果】
請求項１記載の発明によれば、基礎円半径を正規に設定した場合も、正規より大きく設定した場合も、正規より小さく設定した場合も、検定用マスタの同一箇所の歯形を測定するので、検定用マスタの形状偏差が検定結果に混入することを防止でき、検定対象の歯形測定機に起因する誤差のみを検定できる。
【０１１９】
請求項２記載の発明によれば、検定用マスタの被検面の位相とロータリエンコーダとの相対的位置関係を一定角度毎ずらしながら検定を繰返し実行することで、ロータリエンコーダの使用場所によるデータの差、ロータリエンコーダの下部センタと回転主軸とを接続するカップリングの影響、回転主軸のセンタ穴の狂いの影響などを知ることができる。
【０１２０】
請求項３記載の発明によれば、歯車測定機の接触子の物理的接触位置は、検定用マスタの基礎円半径を正規に設定したままで動かさず、変更した基礎円半径の値は、基礎円の接線方向のみに変位する接触子のインボリュート曲線に対応する並進変位を決める数値制御駆動のための指令データの計算用のみに使用するので、検定用マスタの形状偏差が検定結果に混入することを防止でき、検定対象の歯形測定機を数値制御駆動する際の誤差のみを検定結果から評価できるとともに、接触子の物理的接触位置を可変調整する必要がないので、この位置を可変調整する従来の場合に比べて、簡単な検定法を提供できる。
【０１２１】
請求項４記載の発明によれば、検定用マスタの被検面の形状偏差に対応する検出信号から、検定用マスタの基礎円半径を正規の値から変更して設定することにより意図的に発生させた圧力角誤差の理論値を差引いた測定値偏差を求めるので、この測定値偏差を処理することによって、検定の対象としている歯形測定機の状態に関する様々な資料を作成することができる。
【０１２２】
請求項５記載の発明によれば、平均曲線の凹凸の類似性は、検定用マスタの形状偏差に対応することがわかり、平均曲線の相違により、歯形測定機の接触子を移動する並進軸方向の制御の不正確さを把握できる。
【０１２３】
請求項６記載の発明によれば、検定マスタの形状偏差、ＮＣ制御の不正確さ、歯形測定機の繰返し精度のθの依存性、測定の繰返し精度に及ぼすセンサ押付圧などの影響、繰返し精度の最悪値、再現性の最良値、ロータリエンコーダの影響、カップリングの影響、コラムの倒れの影響、回転主軸のセンタ穴の狂いの影響などを定量的に分離して検出できる。
【図面の簡単な説明】
【図１】本発明に係る歯形測定機の検定法の一実施の形態を示す説明図である。
【図２】同上検定法の精度を検証するための説明図（その１）である。
【図３】同上検定法の精度を検証するための説明図（その２）である。
【図４】同上検定法の精度を検証するために実測値と理論値とを比較したグラフである。
【図５】同上検定法の具体的測定手順を示すフローチャートである。
【図６】同上検定法で得られた歯車測定検出信号およびこれから理論値を差引いた測定値偏差を示す説明図である。
【図７】同上検定法で得られた測定値偏差の平均値、測定値偏差のθ点での平均値からのずれ、測定値偏差の平均値からのずれのθ点での標準偏差を示す説明図である。
【図８】同上検定法で得られた測定値偏差の平均値からのずれの平均曲線、測定値偏差の平均値からのずれの最大値および最小値、測定値偏差の平均値からのずれの全データのばらつきの確率密度関数を示す説明図である。
【図９】同上検定法で得られた圧力角誤差を示す表形式の表示例である。
【図１０】同上検定法で得られた測定値偏差の平均値からのずれの最大値および最小値を示す表形式の表示例である。
【図１１】同上検定法で得られた測定値偏差の平均値からのずれの標準偏差を示す表形式の表示例である。
【図１２】同上検定法で得られた測定値偏差のθ点での平均値からのずれを示すグラフ形式の表示例である。
【図１３】同上検定法で得られた測定値偏差のθ点での平均値を示すグラフ形式の表示例である。
【図１４】同上検定法で得られた測定値偏差の平均値からのずれのθ点での標準偏差を示すグラフ形式の表示例である。
【図１５】同上検定法で得られた測定値偏差の平均値からのずれの全データのばらつきの確率密度関数を示すグラフ形式の表示例である。
【図１６】同上検定法で得られた圧力角検出値の狂いの直線成分を示す表形式の表示例である。
【図１７】同上検定法で得られた圧力角検出値の狂いの２次成分を示す表形式の表示例である。
【図１８】同上検定法で得られた圧力角検出値の狂いの３次成分を示す表形式の表示例である。
【図１９】（ａ）は、同上検定法の前提となるインボリュート円筒歯車の歯形測定の基本原理を示す説明図、（ｂ）は、その歯形測定例を示すグラフである。
【図２０】インボリュート円筒歯車歯形測定機の装置概略を示す正面図である。
【図２１】同上歯形測定機のシステムを示す概略図である。
【図２２】本発明に係る検定法との比較で従来の検定法を説明するための説明図（その１）である。
【図２３】同上従来の検定法を説明するための説明図（その２）である。
【図２４】同上従来の検定法により得られた実測値と理論値とを比較したグラフである。
【符号の説明】
ｒ_ｂ　　基礎円半径
１１　　回転主軸
１５　　接触子
１６　　検定用マスタ
１６ａ　　被検面
２２　　下部センタ
２６　　ロータリエンコーダ[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a method for testing a tooth profile measuring machine driven by numerical control built in a computer.
[0002]
[Prior art]
Tooth surface accuracy is often taken up in terms of gear design, manufacturing technology, and as a commercial transaction condition.
[0003]
However, the accuracy of the tooth profile measuring device (gear tooth surface shape inspection device) is often insufficient or the accuracy is not sufficiently verified. Insufficient accuracy may cause driving performance problems.
[0004]
In recent years, most of the tooth profile measuring machines used are driven by a numerical control built in a computer (hereinafter, referred to as CNC) (for example, see Patent Document 1).
[0005]
The CNC-driven tooth profile measuring machine in which the measurement results are processed by a computer incorporates a new error source, which is more than a conventional mechanical tooth profile measuring device.
[0006]
FIG. 21 shows this CNC-driven tooth profile measuring machine, which has at least a rotating main shaft 11 and a translational movement axis 12, and a contact 15 for a contactor table 14 of a displacement sensor 13 moved along the translational axis 12. While being movably provided, the contactor 15 is brought into contact with an involute cylindrical gear to be measured fitted to the rotating main shaft 11 or a surface 16a of a test master 16 for inspection, so as to follow the ideal shape of the surface 16a to be tested. Numerical control by synchronizing the rotational angular position θ of the rotary spindle 11 with the translational displacement X or {X, Y} or {X, Y, Z} (hereinafter represented by translational displacement X) of the contact base 14 ( (Hereinafter referred to as NC) to detect the displacement δ of the contact 15 with respect to the contact base 14 and calculate the spatial position coordinates of the test surface 16a calculated from them and the test surface 16a or its This is a tooth profile measuring machine that calculates a shape deviation of the test surface 16a as a difference from a coordinate value of an ideal shape of the cross section of FIG.
[0007]
Such a CNC tooth profile measuring instrument generally has, for example, the following measurement error sources.
[0008]
The accuracy of the rotational angle position θ of the rotating spindle is determined by the unevenness of the joint between the rotary encoder and the lower center, the unevenness of the drive of the spindle, and the vibration of the device, especially the torsional vibration of the rotating spindle. Inaccuracy of θ adoption value based on inaccuracy of encoder scale interpolation, smoothness of spindle drive, error due to difference of actual value from θ command value in case of open control when there is a problem with NC control technology Occurs.
[0009]
The translational displacement X accuracy of the contact table is based on the inconsistency of the shaft drive, the difference between the actual value with respect to the command value caused by the NC control technology, the change over time of the rail condition caused by the wear of the rail due to the use of the same location, the measuring device The error occurs due to the thermal expansion of the tooth profile measuring device due to the change in the temperature distribution of the tooth profile.
[0010]
The θX synchronization accuracy is as follows: shaft drive inaccuracy, synchronization inaccuracy due to NC control technology, difference between rotation angle recognized by rotary encoder and rotation angle of measured surface, eccentricity of measured object, Deflection, declination, improper installation of the workpiece on the rotating spindle, tooth profile measuring machine, defective center hole of the rotating spindle, transmission of non-uniform motion due to dents, column collapse, change in temperature distribution of the tooth profile measuring machine An error occurs due to uneven speed motion of the rotating main shaft, miso-slip motion, orbit, and the like caused by the rotation.
[0011]
The change in the measured section cross-sectional shape due to the mounting error of the gear to be measured is caused by the difference between the intended measurement position coordinates of the inspection target surface and the actual measurement point coordinates on the tooth profile measuring machine, the eccentricity of the measured object, Deflection, declination, attachment of the workpiece to the rotating spindle, tooth profile measuring machine, failure of the center hole of the rotating spindle, transmission of non-uniform motion due to burrs, column collapse, changes in the temperature distribution of the measuring machine This is caused by the difference between the theoretical and actual reference coordinate values in the resulting shape deviation definition formula.
[0012]
The accuracy (δ read value) of the displacement sensor depends on the dead zone, linearity, bending of the contact and its support due to pressing pressure, and the effect of frictional force. Errors occur due to the effects of the roundness, the difference in sensitivity to lateral displacement, and the vibration of the tooth profile measuring machine.
[0013]
Problems related to measurement execution include the accuracy of axis center setting (origin setting), the accuracy of tooth surface position display origin setting, and the accuracy of placing the contact point of the sensor contact on the basic circle or the theoretical measurement line. It is caused by the dynamic resonance point (the natural frequency consisting of the mass of the contact and the pressing pressure and the uneven frequency of the test surface), the linearity of the AD conversion of the measurement signal, and the response to the measurement.
[0014]
Problems in data processing include the influence of the filter applied to the AD-converted measurement signal (the timing shift of the θ reading, X reading, and δ reading), the theoretical imperfection of the synchronous control NC data (involute cylindrical gear) In other cases), it is caused by the influence of the power supply state and the magnetic field state in the measurement environment.
[0015]
The shape deviation data as the output of the tooth profile measuring instrument includes the errors caused by all of them, but it is necessary to know these individual causes and their influences separately from the measurement results or this test result. Is impossible.
[0016]
Conventionally, in the test using a test master whose tooth profile is very close to involute, there is generally the following problem.
[0017]
(I) When the output is 0, it is impossible to distinguish whether it means high accuracy of the tooth profile measuring machine or that the contact sensor has a dead zone.
[0018]
(Ii) Since the output is close to 0 and the verification master also has some uncertainty regarding its shape accuracy, the quantitative accuracy of the absolute displacement indication value of the output cannot be verified.
[0019]
In the conventional method for measuring tooth profile measuring machines, the master 16 for verification is attached to the tooth measuring device to be verified, and the base circle radius is correctly set to r._bAfter measuring the tooth profile at the same location, set the base circle radius to dr_bLarge r_b+ Dr_bOr the base circle radius is dr_bSmall r_b-Dr_bAnd the physical position where the contact 15 contacts the surface 16a to be measured is the position corresponding to the set base circle radius, that is, the contact The same position (base circle radius r) as used for calculating the command data for the CNC drive that determines the translation displacement X of the table 14_bIf is set correctly, the radius r_bAt the position of the base circle radius r_b+ Dr_bRadius if changed to_b+ Dr_bPosition), and the verification work was performed.
[0020]
In the verification method of the conventional tooth profile measuring machine, the contact point of the contact with the surface to be inspected is set to the base circle radius r_b2The detection value in the case where the detection value changes according to will be described. Here, for the sake of clarity in comparison with the numerical calculation program, the base circle radius r of the verification master_bTo r_blIt is described.
[0021]
In FIG. 22, the hatched involute has a base circle radius r_b1And the involute without hatching is the base circle radius r virtually realized by CNC control._b2Is an involute that serves as a standard for tooth profile measurement and definition of tooth profile accuracy. (In the figure, to avoid misunderstanding in later numerical calculation programs, r_bTo r_blDescribed).
[0022]
dr_b= R_b2-R_b1
[0023]
In actual measurement, first, the tooth surface shape error at a pitch point in the middle of the tooth surface is set to 0, that is, initialization is performed so that both involutes intersect at point P in FIG.
[0024]
The detection value at a point separated by θ in the rotation angle position from this point is the distance between BC in the measurement by the conventional method. The distance between the BCs can be obtained by solving the equations shown in FIGS. 22 and 23 simultaneously.
[0025]
The result is as shown in FIG. 24, and the distance between BC is dr_bIt can be seen that approximation can be made with sufficient accuracy by θ.
[0026]
In this conventional measurement, the base circle radius r, which is a reference for measuring the tooth profile accuracy,_b2= R_b+ Dr_bChange the value of_bIs changed, the corresponding point C of the surface to be measured contacted by the contact changes (see FIG. 22).
[0027]
[Patent Document 1]
Japanese Patent Publication No. 5-51082 (page 4, FIG. 9)
[0028]
[Problems to be solved by the invention]
As described above, when the measurement is performed by changing the base circle radius as in the related art, even if the sampling data is the same θ, it measures different points on the test surface of the verification master. Will be.
[0029]
As a result, no matter how precisely the verification master is manufactured, there is some form deviation, so the shape deviation of the verification master is mixed into the test result, and a slight error occurs in the test result by this processing. When it was detected, it was not possible to separate whether it was caused by the tooth profile measuring machine to be verified or by the shape deviation of the verification master.
[0030]
SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a test method of a tooth profile measuring machine which is not affected by a shape deviation of a test master.
[0031]
[Means for Solving the Problems]
According to the first aspect of the present invention, the verification master is attached to the tooth profile measuring machine to be verified, the base circle radius is set normally, and the tooth profile at the same location of the verification master is automatically measured a plurality of times. Tooth profile measurement that automatically sets the radius of the circle larger than normal and measures the tooth profile of the same location on the verification master multiple times, and automatically sets the base circle radius smaller than normal and measures the tooth profile of the same location on the verification master multiple times. This is a test method for the machine.The tooth profile of the same location on the test master is measured regardless of whether the base circle radius is set to normal, larger than normal, or smaller than normal. Can be prevented from mixing in the test result, and only the error caused by the tooth profile measuring device to be tested can be tested.
[0032]
According to the invention described in claim 2, the connection between the rotating spindle on which the verification master is mounted and the lower center provided with the rotary encoder for detecting the rotation angle of the verification master is released, and This is a verification method for a tooth profile measuring machine which repeatedly repeats fixing and fixing a relative positional relationship between a phase of a test surface and a rotary encoder by a fixed angle, and performing a verification method according to claim 1. By repeatedly performing the test while shifting the relative positional relationship between the phase of the test surface and the rotary encoder by a certain angle, the data difference depending on the location of the rotary encoder and the connection between the lower center of the rotary encoder and the rotary spindle The influence of the coupling, the influence of the center hole of the rotating main shaft, and the like can be known.
[0033]
According to a third aspect of the present invention, in the method of testing the tooth profile measuring machine according to the first or second aspect, the physical contact position of the contact of the gear measuring machine that comes into contact with the test surface of the master for verification is used for verification. The master's base circle radius is not set and moved, and the changed base circle radius value is used only for calculating command data for numerical control drive that determines the translation displacement corresponding to the involute curve of the contact. Therefore, it is possible to prevent the shape deviation of the test master from being mixed into the test result, and to evaluate only the error when driving the tooth profile measuring device to be verified by numerical control from the test result, Since there is no need to variably adjust the physical contact position of the child, a simpler test method can be provided as compared with the conventional case in which this position is variably adjusted.
[0034]
According to a fourth aspect of the present invention, there is provided a method for testing a tooth profile measuring machine according to any one of the first to third aspects, wherein a basis of the verification master is obtained from a detection signal corresponding to a shape deviation of a test surface of the verification master. This is a test method that obtains a measured value deviation by subtracting the theoretical value of the pressure angle error intentionally generated by changing the circle radius from the normal value and setting it. It is possible to create various data on the condition of the tooth profile measuring machine which is the target of the above.
[0035]
According to a fifth aspect of the present invention, in the test method for a tooth profile measuring machine according to the fourth aspect, a plurality of inspections are performed by subtracting a theoretical value of a pressure angle error intentionally generated from a detection signal corresponding to a shape deviation. This is a test method in which a curve is created, an average curve is determined from the plurality of inspection curves, and from the shape of the average curve, data on the condition of the tooth profile measuring machine to be verified is created. It can be seen that the gender corresponds to the shape deviation of the test master, and the inaccuracy of the control in the translation axis direction for moving the contact of the tooth profile measuring machine can be grasped by the difference in the average curve.
[0036]
According to a sixth aspect of the present invention, in the test method of the tooth profile measuring machine according to the fifth aspect, an amount obtained by subtracting a theoretical value of a pressure angle error intentionally generated from a detection signal corresponding to a shape deviation is provided; This is a test method that statistically processes the difference from the average value to create data on the state of the tooth profile measuring machine that is the subject of the test. Dependence of repeatability on θ, influence of sensor pressing pressure on repeatability of measurement, worst value of repeatability, best value of repeatability, influence of rotary encoder, influence of coupling, influence of column tilt, rotation It can quantitatively separate and detect the influence of the deviation of the center hole of the main shaft.
[0037]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, the present invention will be described in detail with reference to one embodiment shown in FIGS. FIG. 21 used in the description of the prior art is a common drawing used for explaining the present invention.
[0038]
This test method uses the same measurement method as the tooth surface shape inspection of production gears, and the user of the tooth profile measuring instrument uses a reference master whose absolute value of the shape deviation from the involute is known or a verification master with a slightly unknown shape deviation. The purpose of this measurement is to ensure the reliability of the results of the tooth surface shape inspection of production gears that are conducted on a daily basis, and to grasp the health of the tooth profile measuring machine. It is not intended to determine the manufacturing accuracy of the tooth profile measuring machine by individually measuring the accuracy of each component, assembly, alignment, and the like of the tooth profile measuring device as originally performed by the manufacturer.
[0039]
Referring to FIG. 1, an outline of the present verification method will be described. A verification master 16 is attached to a tooth profile measuring machine to be verified, and its base circle radius r_bIs set to normal (a = 0), the test surface 16a at the same position of the test master 16 is automatically measured a plurality of times, and then the base circle radius is automatically set to be larger than normal (a = + dr)._b), Measuring the test surface 16a at the same location of the test master 16 a plurality of times, and then automatically setting the base circle radius smaller than normal (a = −dr)_b), The test surface 16a at the same location of the test master 16 is measured a plurality of times.
[0040]
At this time, the physical contact position of the contact 15 of the gear measuring machine that comes into contact with the test surface 16a of the test master 16 remains in a state where the base circle radius of the test master 16 is set to normal (a = 0). And the changed value of the base circle radius is used only for calculation of command data for numerical control drive for determining the translation position of the contact 15 displaced only in the tangential direction of the base circle.
[0041]
Further, the coupling connecting the rotary spindle 11 on which the verification master 16 is mounted and the lower center 22 on which the rotary encoder 26 for detecting the rotation angle of the verification master 16 is released is released. , The relative positional relationship between the phase of the test surface 16a and the rotary encoder 26 is shifted by a fixed angle and fixed (k = 1 to 5), and the above-described test method is executed for each phase.
[0042]
As shown in FIG. 20, a target tooth profile measuring machine is configured such that a rotating main shaft 11 that supports an involute cylindrical gear or a test master 16 whose tooth profile is measured is formed on a body 21 by a lower center 22 and an upper center 23. The lower center 22 is rotatably supported and connected to the lower center 22 by a coupling 24. The lower center 22 is provided with a drive system 25 for rotating and driving the lower center 22, and a rotary encoder 26 for detecting a rotation angle. I have.
[0043]
The body 21 is provided with a Y-axis direction moving body (column) 28Y that is moved in the Y-axis direction by a Y-axis drive motor 27Y, and the amount of movement is detected by a Y-axis linear encoder 29Y. The body 28Y is provided with an X-axis direction moving body 28X that is moved in the X-axis direction by an X-axis driving motor 27X, and the amount of movement is detected by an X-axis linear encoder 29X. (Hereinafter, the X axis is referred to as a “translation axis”), a Z-axis moving body 28Z that is moved in the Z-axis direction by a Z-axis drive motor 27Z is provided, and the amount of movement is detected by a Z-axis linear encoder 29Z. You.
[0044]
The members from the X-axis drive motor 27X to the Z-axis direction moving body 28Z are translational shaft members that move the contact base 14 of the displacement sensor 13 in the X-axis direction. A contact 15 movable by a contact pressure from the tooth surface is provided.
[0045]
Then, as shown in FIG. 21, while the contact 15 of the displacement sensor 13 is in contact with the test surface 16a, the rotation angle position θ of the rotary spindle 11 and X are adjusted so as to follow the ideal shape of the test surface 16a. The translation displacement X or {X, Y} or {X, Y, Z} (hereinafter represented by translation displacement X) of the contact base 14 that moves in the X-axis direction together with the axial moving body 28X is synchronized. And is driven by numerical control (hereinafter referred to as NC).
[0046]
In FIG. 21, the detection signals detected by the rotary encoder 26 and the X-axis linear encoder 29X are fed back to the NC control unit 33 via the signal conditioner 31 and the counter 32, and are also taken in the personal computer 34.
[0047]
As shown in FIGS. 19A and 19B, the basic principle of measuring the tooth profile of an involute cylindrical gear by this tooth profile measuring machine is to detect the displacement δ of the contact 15 with respect to the contact base 14 of the displacement sensor 13. , The spatial position coordinates of the test surface 16a calculated from the measured values, and the mathematically given coordinate values of the ideal shape of the test surface 16a or its cross section (X = d_hθ / 2), the shape deviation of the test surface 16a is determined.
[0048]
The displacement sensor 13 includes a length displacement such as a differential transformer, an optical interference laser interferometer, a digital displacement sensor (magnescale, inductosin, moiré fringe), a resistive strain gauge, an optical linear encoder, and a photoelectric microscope. A meter and a position detector are also included.
[0049]
Next, an embodiment of a test method for a tooth profile measuring machine for an involute cylindrical gear will be described.
[0050]
The point P (point where both involute curves intersect) where the tooth surface shape error is 0 in FIG. 22 of the conventional test method is the point H on the base circle shown in FIG.₁A on the tangent at₁Initialize to match the points.
[0051]
The detected value at a point θ away from this point at the rotation angle position is A in FIG.₂A₃The same rotational angle position θ of the rotary spindle 11 determines the position of the surface to be measured even when the measurement is performed while changing the base circle radius.
[0052]
That is, if the sampling of the measurement data is performed at the same rotational angle position θ of the rotating main shaft 11, dr_bIs changed, the sampling at the same θ is a measurement value of the same point on the test surface 16 a of the test master 16.
[0053]
As a result, any shape deviation of the verification master 16 is dr at this installation angle position of the verification master 16._bAnd the data obtained by repeated measurements will be added to the test results as an invariant function for θ.
[0054]
Therefore, it is possible to evaluate only the error of the NC control of the tooth profile measuring machine which must be most targeted in the test operation from the test result.
[0055]
In the test result by such a test method, since the error caused by the shape deviation of the test master 16 does not enter at all, the rotation angle of the rotating main shaft 11 of the average value in the result measured by changing the base circle radius is used. If there is a correspondence between fine irregularities with respect to the position θ, this corresponds to the shape deviation of the test master 16 and estimates the accuracy of the sensitivity of the measurement sensor to minute input.
[0056]
Further, in the conventional test method, there is a problem in the robustness of the test result of the target tooth profile measuring machine unless the test master 16 having the absolutely correct accuracy is used. However, in the present method, as described later, By converting the measurement results into data distributed around the average value and performing statistical processing, the reliability of the results can be improved even in the test of a tooth profile measuring machine using the test master 16 having a certain shape deviation. It has the advantage of not causing problems.
[0057]
Next, the point of contact of the contact 15 with the test surface 16a is determined by the set base circle radius r._b2Irrespective of the base circle radius r_b1Verify the detection value when staying at the position. Here, for the sake of clarity in comparison with the numerical calculation program, the base circle radius r of the test master 16 is set._bTo r_blIt is described.
[0058]
In FIG. 2, the hatched involute has a base circle radius r_blAnd the involute without hatching is the base circle radius r virtually realized by CNC control._b2, Ie, an involute serving as a criterion for defining tooth profile accuracy. Where dr_b= R_b2-R_blIt is.
[0059]
In actual measurement, first, a tooth surface shape error at a pitch point in the middle of the tooth surface is set to 0, that is, A in FIG.₁Initialize so that both involutes intersect at a point.
[0060]
The detected value at a point separated by θ from the point at the rotation angle position is A in the measurement of this assay.₂A₃Is the distance between them.
[0061]
This A₂A₃The distance between them can be obtained by solving the equations shown in FIGS. 2 and 3 simultaneously. The result is as shown in FIG.₂A₃The distance between them is dr_bIt can be seen that approximation can be made with sufficient accuracy by θ.
[0062]
That is, in the related art shown in FIGS. 22 to 24, the distance between BCs is dr._bIt can be understood that approximation can be made with sufficient accuracy by θ, and as shown in FIGS. 2 to 4, in the test method according to the present invention, A₂A₃The distance between them is dr_bIt can be seen that approximation can be made with sufficient accuracy by θ. As a result, it is possible to grasp the accuracy that was not known in the related art.
[0063]
Next, an example of a specific measurement procedure will be described with reference to the flowchart shown in FIG. Note that circled numbers represent step numbers.
[0064]
Normally,
(1) The verification master 16 is measured in accordance with the routine procedure for inspecting the tooth surface shape of the production gear. First, the verification master 16 is set at the reference position (0 °) (step 1).
[0065]
(2) Base circle radius r_bIs set normally, and the tooth profile at the same location is measured continuously, for example, eight times on the left tooth surface and eight times on the right tooth surface (step 2).
[0066]
(3) Base circle radius r_bIs set larger than normal by, for example, 0.05 mm, and the tooth profile at the same location is measured continuously, for example, eight times on the left flank and eight times on the right flank (step 3).
[0067]
(4) Base circle radius r_bIs set to, for example, 0.05 mm smaller than normal, and the tooth profile at the same location is continuously measured, for example, eight times on the left tooth surface and eight times on the right tooth surface (step 4).
[0068]
Above, (2), (3), and (4) are referred to as a measurement set.
[0069]
Further, when the influence of the accuracy of the rotary encoder 26 is concerned, the following measurements (5) to (8) are performed.
[0070]
(5) When the measurement set 1 is completed, the position of the test surface of the verification master 16 is rotated by 90 ° relative to the reference center with respect to the lower center 22 connected to the rotary encoder 26 of the tooth profile measuring machine (step 5, 6), and the same measurement as in (2), (3), and (4) above is performed (measurement set 2).
[0071]
(6) When the measurement set 2 is completed, the position of the test surface of the test master 16 is further rotated by 90 ° with respect to the lower center 22 connected to the rotary encoder 26 of the tooth profile measuring machine (steps 7 and 8), and The same measurement as (2), (3), and (4) above is performed at a position 180 ° from the position (measurement set 3).
[0072]
(7) When the measurement set 3 is completed, the position of the test surface of the verification master 16 is further rotated by 90 ° with respect to the lower center 22 connected to the rotary encoder 26 of the tooth profile measuring machine (steps 9 and 10), and The same measurement as (2), (3) and (4) is performed at a position 270 ° from the position (measurement set 4).
[0073]
(8) When the measurement set 4 is completed, the position of the test surface of the verification master 16 is further rotated by 90 ° with respect to the lower center 22 connected to the rotary encoder 26 of the tooth profile measuring machine, and returned to the initial position (step 11). , 12) and the same measurements as in (2), (3), and (4) above (measurement set 5).
[0074]
(9) The test is always performed on both left and right tooth surfaces continuously.
[0075]
Next, a specific data processing method will be described with reference to FIGS.
[0076]
Now, the rotational angle position of the rotary spindle 11 at the time of tooth profile measurement is θ, and the deviation of the set base circle radius from the normal value is dr._bAs shown in FIG. 6, the detection signal corresponding to the shape deviation value of the rotational angle position θ of the rotary spindle 11 is represented by f_kai(Θ).
[0077]
As shown in FIG. 1, the subscript k represents data from measurement sets 1 to 5 (k = 1,..., 5), and the rotary encoder 26 is shown with respect to the verification master 16. In this case (first measurement state), k = 1, the coupling 24 connecting the rotary spindle 11 of the verification master 16 and the lower center 22 of the rotary encoder 26 is disconnected, and the verification master 16 is K = 2 when the rotary encoder 26 is relatively rotated by π / 2 and then re-coupled.... The rotary encoder 26 is rotated once relative to the verification master 16 and is coupled in the same state as k = 1. In this case, k = 5.
[0078]
The subscript a indicates that the setting of the base circle radius is normal, + dr_b, -Dr_b(A = 0, + dr)_b, -Dr_b), Base circle radius r_b, (R_b+ Dr_b), (R_b-Dr_b), The contact 15 of the displacement sensor 13 is set to x = r_bWhen driving with θ, a = 0 and x = (r_b+ Dr_bA = + dr when driving with θ)_bAnd x = (r_b-Dr_bA) = − dr when driving with θ_bAnd However, the position of the contact 15 with respect to the tooth surface is not changed, and the generating master 16 and the contact 15 are caused to perform a generating motion by using these three types of base circle radii.
[0079]
Further, the subscript i indicates that the data is data of repeated measurement under the same conditions (for example, i = 1,..., 8).
[0080]
As described with reference to FIG. 2 to FIG._bSince the tooth surface shape deviation sufficiently close to θ should be measured, the length of the tooth profile inspection range on the action line is h, and the pressure angle error of the inspection curve in that range is f._HαThen f_Hα/ L_αIs dr_b/ R_bShould match.
[0081]
The theoretical value of the generated pressure angle error is subtracted from the measurement signal of the sensor, and as shown in FIG. 6, the measurement value deviation f under a certain measurement condition {k, a}^* _kai(Θ) is determined.
[0082]
f^* _kai(Θ) = f_kai(Θ) -dr_b・ Θ
[0083]
With respect to this value, as shown in FIG. 7, the average value F at the θ point of the deviation of the measurement value under a certain measurement condition {k, a}_ka(Θ) is determined.
[0084]

[0085]
Further, as shown in FIG. 7, the deviation f of the measured value deviation under a certain measurement condition {k, a} from the average value at the θ point is shown.^** _kai(Θ) is determined.
[0086]

[0087]
Further, as shown in FIG. 8, the probability density function P of the dispersion of all data of deviation from the average value of the measured value deviation in the measurement of a certain angular position k_k(F^**).
[0088]

[0089]
Also, as shown in FIG. 8, the average curve F of the deviation from the average value of the measured value deviation under a certain measurement condition {k, a}.^** _ka(Θ) is determined.
[0090]

[0091]
This average curve F^** _kaIn (θ), the maximum value of the deviation of the measured value deviation from the average value under a certain measuring condition {k, a}

Ask for.
[0092]

[0093]
Further, the average curve F^** _kaIn (θ), the minimum value of the deviation from the average value of the measurement value deviation under a certain measurement condition {k, a}

Ask for.
[0094]

[0095]
Further, as shown in FIG. 7, the standard deviation s at the θ point of the deviation of the deviation of the measured value from the average value under a certain measurement condition {k, a}._ka(Θ) is determined.
[0096]

[0097]
Similarly, as shown in FIG. 8, the standard deviation S of the deviation of the measured value deviation from the average value under a certain measurement condition {k, a}_kaAsk for.
[0098]

[0099]
Further, the maximum value s of the deviation from the average value of the standard deviation of the θ point measurement value in the measurement of a certain angular position k_k ⁻(Θ) is determined.
[0100]

[0101]
Further, the maximum value S of the entire standard deviation of the deviation of the measured value deviation from the average value in the measurement at a certain angular position k_kAsk for.
[0102]

[0103]
Also, F^** _ka(Θ),

Curve approximation (curve @ fitting) and the coefficient b_ka, C_ka, D_kaAsk for.
[0104]
Coefficient b_kaIs the degree of deviation of the detected pressure angle in the measurement, c_kaIs the degree of contamination of the secondary components in the measurement, d_kaIs an index value indicating the degree of mixing of the tertiary component in the measurement.
[0105]
Next, in order to evaluate the test result, it is visually displayed in a visible form as shown in FIGS. 9 to 18. An evaluation example of the test result will be described.
[0106]
For various measurement sets, set the base circle radius to normal, + dr_b, -Dr_bData (a = 0, + dr_b, -Dr_b) Indicates the following.
[0107]
(1) F_kaThe similarity of the unevenness of the (θ) curve is considered to correspond to the shape deviation of the test master 16.
[0108]
(2) F_kaThe difference in the (θ) curve indicates the inaccuracy of the NC control of the translation axis of the tooth profile measuring rock on the basis of θ of the measurement value under the constant measurement conditions {k, a}, and_ka, C_ka, D_kaThe difference in｝ is a quantification of this.
[0109]
(3) s_kaThe deviation of the (θ) curve from 0 (zero value) indicates the dependence of the repetition accuracy of the tooth profile measuring machine on θ.
[0110]
S_kaThe variation with respect to a represents the influence of the sensor pressing pressure and the like on the repeatability of the measurement under the constant measurement condition {k}. That is, (a = 0, + dr_b, -Dr_b) By s_kaThe difference in the (θ) curve indicates a difference in the repeatability of the sensor due to a slight contact pressure (or a displacement of the contact 15).
[0111]
S_kRepresents the repetition accuracy when measurement is performed with the verification master 16 attached to this angular position.
[0112]
(4) The difference between the measurement sets 1 and 5 represents part of the cause of the difference in repeatability and reproducibility.
[0113]
In FIG. 11 showing the standard deviation of the measured values,

Represents the worst value of the repeatability and the best value of the reproducibility.
[0114]
(5) Measurement set (rotation phase angles 0, 0.5π, π, 1.5π, 2π,: (k = 1,..., 5)) using different positions of the rotary spindle 11 and the rotary encoder 26 The difference between the results is
・ Data difference depending on where the rotary encoder 26 is used
・ Effect of coupling 24 connecting rotary encoder 26 and lower center 22
・ Effect of column collapse
・ Influence of center hole deviation of rotating spindle 11
Represents
[0115]
Specific displays for performing these evaluations are displayed in a table format or a graph format as shown in FIGS.
[0116]
Note that, with k = 1, the calibration may be performed simply without taking into account the influence of the irregularity of the rotary encoder 26 or the coupling 24.
[0117]
As described above, according to this test method, even if the measured data of the tooth profile deviation including the errors caused by all of the various error factors comprehensively, it falls within a range that deviates from its true value by a certain degree. It is to clarify whether it fits in from the viewpoint of industrial practicality, and in order to guarantee the result of the tooth profile accuracy inspection of the production gear, the test data of the tooth profile measuring machine that guarantees the measurement accuracy of the tooth profile measuring machine that performed the measurement Is proposed.
When necessary, the test data should be used as an attachment to the production gear accuracy test results. Further, the test data can be used for assisting correct daily maintenance of the tooth profile measuring machine.
[0118]
【The invention's effect】
According to the first aspect of the present invention, the tooth profile of the same location of the test master is measured regardless of whether the base circle radius is set to normal, larger than normal, or smaller than normal. It is possible to prevent the shape deviation of the test master from being mixed into the test result, and to test only the error caused by the tooth profile measuring device to be tested.
[0119]
According to the second aspect of the present invention, the test is repeatedly performed while shifting the relative positional relationship between the phase of the test surface of the test master and the rotary encoder by a predetermined angle, so that the data of the rotary encoder according to the use location can be obtained. It is possible to know the difference, the influence of the coupling connecting the lower center of the rotary encoder and the rotary spindle, and the influence of the deviation of the center hole of the rotary spindle.
[0120]
According to the third aspect of the present invention, the physical contact position of the contact of the gear measuring machine does not move while the base circle radius of the verification master is normally set, and the value of the changed base circle radius is Since it is used only for calculating command data for numerical control drive that determines the translational displacement corresponding to the involute curve of the contact displaced only in the tangential direction of the circle, the shape deviation of the test master may be mixed in the test result. In addition to being able to evaluate only the error when driving the tooth profile measuring device to be verified by numerical control from the verification result, there is no need to variably adjust the physical contact position of the contactors. A simple assay can be provided as compared with the case of
[0121]
According to the fourth aspect of the invention, the detection signal corresponding to the shape deviation of the test surface of the test master is intentionally generated by changing the base circle radius of the test master from a normal value. Since the measured value deviation is obtained by subtracting the theoretical value of the pressure angle error, various data relating to the state of the tooth profile measuring machine to be verified can be prepared by processing the measured value deviation.
[0122]
According to the fifth aspect of the present invention, it can be understood that the similarity of the unevenness of the average curve corresponds to the shape deviation of the test master. Control inaccuracy can be understood.
[0123]
According to the invention of claim 6, the shape deviation of the verification master, the inaccuracy of the NC control, the dependence of the repetition accuracy of the tooth profile measuring machine on θ, the influence of the sensor pressing pressure on the repetition accuracy of the measurement, the repetition accuracy Value, the best value of reproducibility, the effect of the rotary encoder, the effect of the coupling, the effect of the column falling, the effect of the center hole of the rotating spindle, and the like can be quantitatively separated and detected.
[Brief description of the drawings]
FIG. 1 is an explanatory view showing an embodiment of a test method for a tooth profile measuring machine according to the present invention.
FIG. 2 is an explanatory diagram (part 1) for verifying the accuracy of the test method.
FIG. 3 is an explanatory diagram (part 2) for verifying the accuracy of the above-mentioned test method.
FIG. 4 is a graph comparing actual measurement values and theoretical values in order to verify the accuracy of the above-mentioned test method.
FIG. 5 is a flowchart showing a specific measurement procedure of the same assay method.
FIG. 6 is an explanatory diagram showing a gear measurement detection signal obtained by the above-mentioned test method and a measurement value deviation obtained by subtracting a theoretical value from the gear measurement detection signal.
FIG. 7 shows the average value of the measured value deviation, the deviation of the measured value deviation from the average value at the θ point, and the standard deviation at the θ point of the deviation of the measured value deviation from the average value obtained by the above-mentioned test method. FIG.
FIG. 8 shows an average curve of deviation from the average of measured value deviations obtained by the above-mentioned test method, a maximum value and a minimum value of deviation of the measured value deviation from the average value, and a deviation of the measured value deviation from the average value. FIG. 9 is an explanatory diagram showing a probability density function of variation of all data.
FIG. 9 is a tabular display example showing a pressure angle error obtained by the above-mentioned test method.
FIG. 10 is a display example in a table format showing a maximum value and a minimum value of a deviation from an average value of measured value deviations obtained by the same test method.
FIG. 11 is a display example in a table format showing a standard deviation of a deviation of a measured value deviation from an average value obtained by the above-described test method.
FIG. 12 is a display example in a graph format showing a deviation of a measured value deviation obtained from the above-described test method from an average value at a θ point.
FIG. 13 is a display example in the form of a graph showing an average value of measured value deviations obtained by the above-mentioned test method at point θ.
FIG. 14 is a display example in a graph format showing a standard deviation at a θ point of a deviation from a mean value of a measured value deviation obtained by the above-described test method.
FIG. 15 is a display example in a graph format showing a probability density function of variation of all data of deviation from the average value of measured value deviation obtained by the above-described test method.
FIG. 16 is a display example of a tabular form showing a deviation linear component of the detected pressure angle value obtained by the above-described test method.
FIG. 17 is a display example of a tabular form showing a secondary component of a deviation of the detected pressure angle obtained by the same test method.
FIG. 18 is a display example in a tabular form showing a tertiary component of a deviation of a detected pressure angle obtained by the above-described test method.
FIG. 19 (a) is an explanatory diagram showing the basic principle of tooth profile measurement of an involute cylindrical gear as a premise of the above-mentioned test method, and FIG. 19 (b) is a graph showing an example of tooth profile measurement.
FIG. 20 is a front view schematically showing an apparatus of an involute cylindrical gear tooth profile measuring device.
FIG. 21 is a schematic view showing a system of the tooth profile measuring machine.
FIG. 22 is an explanatory diagram (part 1) for explaining a conventional test method in comparison with the test method according to the present invention.
FIG. 23 is an explanatory view (No. 2) for describing the conventional assay method according to the above.
FIG. 24 is a graph comparing measured values and theoretical values obtained by the conventional test method.
[Explanation of symbols]
r_bBase circle radius
11 rotating spindle
15 contact
16 verification master
16a Test surface
22 lower center
26 rotary encoder

Claims (6)

Attach the master for verification to the tooth profile measuring machine to be verified,
Normally set the base circle radius and automatically measure the tooth profile of the same location of the verification master multiple times,
Automatically set the base circle radius larger than normal and measure the tooth profile at the same location of the verification master multiple times,
A test method for a tooth profile measuring machine, wherein a tooth profile at the same position of a test master is measured a plurality of times by automatically setting a base circle radius smaller than a normal value. The connection between the rotary spindle on which the verification master is mounted and the lower center provided with the rotary encoder for detecting the rotation angle of the verification master is released, and the phase of the test surface of the verification master and the rotary encoder are disconnected. Fixing the relative positional relationship by shifting it by a certain angle,
A test method for a tooth profile measuring machine, characterized by repeating the steps of executing the test method according to claim 1. The physical contact position of the contact of the gear measuring machine that comes into contact with the test surface of the verification master does not move while the base circle radius of the verification master is set to normal, and the value of the changed base circle radius is 3. The method according to claim 1, wherein the method is used only for calculating command data for a numerical control drive for determining a translational displacement corresponding to an involute curve of a child. From the detection signal corresponding to the shape deviation of the test surface of the test master, the theoretical value of the pressure angle error intentionally generated by changing and setting the base circle radius of the test master from the normal value is subtracted. 4. The method according to claim 1, wherein the measured value deviation is obtained. From the detection signal corresponding to the shape deviation, subtract the theoretical value of the pressure angle error intentionally generated to create a plurality of inspection curves,
An average curve is obtained from these multiple test curves,
5. The test method for a tooth profile measuring machine according to claim 4, wherein data on the condition of the tooth profile measuring device to be verified is created from the shape of the average curve. By statistically processing the difference between the theoretical value of the pressure angle error intentionally generated from the detection signal corresponding to the shape deviation and the average value, the state of the tooth profile measuring machine to be tested 6. A method for testing a tooth profile measuring machine according to claim 5, wherein a material relating to the tooth profile is prepared.

Images