DE102019125304A1 - Method for measuring and qualifying a rack and device therefor - Google Patents

Abstract

This relates to a method for measuring and qualifying a toothed rack, with a toothing on one side thereof, the toothed rack being optically scanned along its course by an optical sensor spaced from the toothed rack Track is measured along the course of the same and, in a second measuring step, the rack is measured mechanically or magnetically in a defined position or clamped against a defined surface lying in at least one track along the course of the same. The further relates to a device for carrying out the method.

Description

The invention relates to a method for measuring and qualifying a toothed rack in order to determine its torsion and curvature, which deviate from the ideal dimension, and a device for this purpose.

Racks are required for different applications: in the construction of robots and machines, weirs and steering gears, wherever a rotary movement is to be converted into a linear movement. A helical toothing of the rack increases the load-bearing capacity of the connection so that higher torques can be used.

• -Fräsen
• -Schruppen
• -Schlichten

There are three processes you need to know when manufacturing gear racks:

• -Milling
• - roughing
• - Finishing

Roughing and finishing can be summarized under the heading of grinding. Opinions differ a little here. Some assign roughing to milling, but you can mill with metal tools to distinguish it. Mineral tools are then used for roughing and finishing.

So first a rack is milled. The material is removed with a metal milling head.

Racks of lower precision are finished after this pass. If you only want to be a little more precise with milling, you can mill at half the depth, for example, and measure how far the set depth deviates from the actual one. You can enter this into the machine as a correction factor and finish milling. Milling causes high temperature input and, through high forces, causes elastic deformation of the machine and workpiece. As a result, milled racks are never very precise. The rack is still straightened (straightened) and is then ready for use. In some cases only milled racks are hardened, but this is rather unusual.

If the result is to be better, i.e. the curvature and torsion minimized, milling is carried out in such a way that the tooth base is already finished (greatest material load) and the flanks and heads are slightly less removed. Then the bar is hardened, the sides sanded and the bar straightened. Then the top is sanded. The rack is given its precise shape with a mineral disk. Racks that are not so precise are ground in one pass. More precise racks can be obtained by first roughing and then finishing. Between these two processes, the rack is stored outside the machine for 24 hours in order to reduce any internal temperature and any other tension. When finishing, the rack is only machined minimally in order to keep the distortion as low as possible.

When grinding, a round tool (mineral grinding wheel) of a certain width x is repeatedly guided over the rack and dressed in certain time segments.

1. 1. Das Werkzeug ist nicht perfekt abgerichtet. Hierbei kann das Zahnstangenprofil insgesamt zu tief oder zu hoch geschliffen werden.
2. 2. Das Werkzeug ist unregelmäßig und ein oder mehrere Zähne sind anders als der Rest (Tiefe, Breite, Versatz)
3. 3. Das Werkzeug wird beim versetzen nicht genau positioniert. Ein Zahnkopf zwischendrin ist also zu breit oder zu schmal.

A rack is usually several times as long as the width of the grinding wheel, which means that it has to be ground in several passes by moving the grinding wheel. Errors can arise here as follows (even if one assumes that the positioning would be perfect):

1. 1. The tool is not perfectly trained. The rack profile can be ground too deep or too high overall.
2. 2. The tool is irregular and one or more teeth are different from the rest (depth, width, offset)
3. 3. The tool is not positioned exactly when it is moved. A tooth tip in between is therefore too wide or too narrow.

1. 1. Die Zahnstangen müssen alle bei einer bestimmten Temperatur gefertigt werden. Dabei gibt es durchaus Abweichungen, die aufgrund sommerlicher oder winterlicher Temperaturen auftreten, oder gar, weil die Maschine erst frisch hochgefahren wurde und noch nicht ganz durchgewärmt ist (Kühlmittel wird auf konstanter Temperatur gehalten). Diese Temperaturunterschiede können nicht erfasst werden und vorab als Korrekturparameter in die Maschine eingegeben werden, da die Fehlerquellen zu vielfältig sind (jede Achse für sich, die Auflagefläche und genauso das Werkstück selbst werden von der unterschiedlichen Temperatur beeinflusst)
2. 2. Das Ausgangsmaterial unterliegt herstellungsbedingt Schwankungen, die nicht erfasst werden können. Selbst innerhalb einer Charge ist jede Zahnstange bei exakt gleicher Behandlung im Endergebnis unterschiedlich. Auch dies kann nicht mit vernünftigem Aufwand und zerstörungsfrei vorab erfasst werden.
3. 3. Das Härten kann nicht hochpräzise wiederholbar mit perfekt gleichem Ergebnis eingestellt werden. Auch hier spielen Temperaturen und Materialeigenschaften eine Rolle.
4. 4. Die Zahnstange wird nach dem Härten gerichtet, um Verzug zu kompensieren. Das Richten ist unmöglich zu perfektionieren. Es wird immer Ungenauigkeiten geben.

In order to keep the errors as small as possible, the first batch of racks is measured in order to obtain settings for the grinding machine that make the next batch of racks more accurate. This is necessary in order to obtain as many racks of higher quality as possible. The racks produced cause the following problems due to the material used:

1. 1. The racks must all be manufactured at a certain temperature. There are definitely deviations that occur due to summer or winter temperatures, or even because the machine has just started up and is not yet fully warmed up (coolant is kept at a constant temperature). These temperature differences cannot be recorded and entered into the machine in advance as correction parameters because the sources of error are too diverse (each axis for itself, the support surface and the workpiece itself are influenced by the different temperature)
2. 2. The raw material is subject to manufacturing-related fluctuations that cannot be recorded. Even within a batch, every rack is exactly the same Treatment in the end result different. This, too, cannot be recorded in advance with reasonable effort and in a non-destructive manner.
3. 3. The hardening cannot be adjusted with high precision and repeatability with perfectly the same result. Here, too, temperatures and material properties play a role.
4. 4. The rack is straightened after hardening to compensate for distortion. Judging is impossible to perfect. There will always be inaccuracies.

1. 1. Die üblichen taktilen Messverfahren sind langsam. Eine Zahnstange zu vermessen dauert erheblich länger (Minimum 30 Minuten) als die Herstellung (3 Minuten). In dieser Zeit kann sich die Maschine wieder abgekühlt haben und die nachfolgenden Zahnstangen werden wieder anders. Die Messzeit muss also im Bereich der Fertigungszeit liegen. Nur dann könnte nicht nur nach einer Charge, sondern vielleicht jede Zahnstange nach der Fertigung gemessen werden und Korrekturparameter sogar innerhalb der Charge eingegeben werden. Ein Messen während der Fertigung oder direkt nach der Fertigung in der Maschine ist aus verschiedenen Gründen nicht möglich:
   • Das Ausspannen des Werkstückes aus der Maschine (lösen des Schraubstockes / der Spannung) kann allein schon eine Verkürzung um 50 µm auf einen Meter bewirken.
   • Kühlwasser spritzt in alle Richtungen und das wäre dem Messkopf nicht zuträglich.
   • Die Stange muss nach der Fertigung eine gewisse Zeit ruhen, um sich zu entspannen und abzukühlen.
2. 2. Eine Zahnstange zu vermessen ist aufwändig. Um die Teilung und den Schrägungswinkel oder die Flankenwinkel zu messen, müssen die in den Firmen vorhandenen Messmaschinen speziell programmiert werden.
3. 3. Die derzeit bekannten taktilen Messmaschinen können maximal auf +/- 5 Mikrometer genau messen. Das reicht für manche Winkel bei kleinen Zähnen nicht aus.
4. 4. Allein die Längung bei Erhöhung der Temperatur um ein Grad beträgt 13 µm. Ein Messen direkt nach der Bearbeitung wird nie genau, da selbst wenn sich die Stange von außen abgekühlt anfühlt sie im Inneren noch erwärmt sein kann.

The most serious factor in increasing the quality of the products remains the measurement of the racks that have already been manufactured and the adjustment of the manufacturing parameters derived from the measurement result. There are the following problems:

1. 1. The usual tactile measurement methods are slow. Measuring a rack takes considerably longer (minimum 30 minutes) than producing it (3 minutes). During this time, the machine may have cooled down again and the subsequent racks will change again. The measurement time must therefore be in the range of the production time. Only then would it be possible to measure not only after a batch, but perhaps every rack after production, and correction parameters even entered within the batch. A measurement during production or directly after production in the machine is not possible for various reasons:
   • Unclamping the workpiece from the machine (loosening the vice / clamping) can reduce the length by 50 µm to one meter.
   • Cooling water splashes in all directions and that would not be beneficial to the measuring head.
   • The rod needs to rest for a while after production to relax and cool down.
2. 2. Measuring a rack is time-consuming. In order to measure the pitch and the helix angle or the flank angle, the measuring machines available in the company must be specially programmed.
3. 3. The currently known tactile measuring machines can measure with a maximum of +/- 5 micrometers. This is not enough for some angles with small teeth.
4. 4. The elongation alone when the temperature is increased by one degree is 13 µm. Measuring directly after machining is never accurate, because even if the bar feels cooled on the outside, it can still be heated inside.

The object of the invention is to provide an improved method and a device for measuring and qualifying a toothed rack, in which the evaluation can be carried out quickly.

This object is achieved by a method for measuring and qualifying a toothed rack according to the features of claim 1 and a device for measuring and qualifying a toothed rack according to the features of claim 8.

• - einem ersten Vermessungsschritt die Zahnstange ungespannt liegend in wenigstens einer Spur entlang des Verlaufs derselben vermessen wird und in
• - einem zweiten Vermessungsschritt die Zahnstange mechanisch oder magnetisch in einer definierten Lage eingespannt bzw. gespannt gegen eine definierte Fläche liegend in wenigstens einer Spur entlang des Verlaufs derselben vermessen wird.

According to the invention, the method provides that the toothed rack is optically scanned along its course by an optical sensor spaced from the toothed rack,
where in

• - In a first measuring step, the rack is measured lying unstressed in at least one track along the course of the same and in
• - In a second measuring step, the rack is mechanically or magnetically clamped in a defined position or clamped against a defined surface in at least one track along the course of the same.

In this way, the detection of the geometric data of the rack is clamped and freely enabled in a particularly effective manner.

It is advantageous to run 2 tracks with the measuring device. Unstressed for the torsion and stressed for the angle and, in the case of helical racks, for the helix angle.

• Es werden mehrere Spuren auf der Zahnstange abgefahren und
• Messpunkte gesammelt. Im ungespannten Zustand zwei Spuren auf der Seite und zwei oben um Verdrehungen, Verbiegungen zu erfassen.

Procedure:

• Several tracks are followed on the rack and
• Measurement points collected. In the untensioned state, two tracks on the side and two on top to detect twisting and bending.

The rack is then clamped magnetically. This removes twists and turns. Other companies that manufacture measuring machines try to subtract the twisting and bending. The effort is high and the result is very bad if the assumed deformations do not apply.

1. 1. Es wird über einen Algorithmus, der die Menge der Punkte in einem schmalen Band parallel zu x-Achse bestimmt, das Maximum über den Zahnköpfen bestimmt.
2. 2. Es wird eine interpolationsgerade durch die Punkte gelegt.
3. 3. Mit dem erhaltenen y-Achsenabschnitt wird ein neues Fenster parallel zur x-Achse berechnet (Zahngrund, m-Faktor, Radius) in dem die Flanken liegen müssen.
4. 4. Ein weiterer Algorithmus, der zusätzlich die x-Achsen-Werte in der Breite PI * Modul /2 filtert können nun die Start- und Endpunkte der Flanken bestimmen.
5. 5. Aus diesen Start und Endpunkten kann die Interpolierte Geradengleichung jeder Flanke berechnet werden.
6. 6. Die Winkelhalbierende jeder dieser Geraden ist die Mitte des Zahnes beziehungsweise des Zahnstangengrundes.
7. 7. Der Abstand der Winkelhalbierenden liefert Werte, die eine Aussage über die Qualität der Messung ermöglichen. Sie müssen in einem bestimmten Fenster zum theoretischen Abstand liegen. Sind sie zu weit Abweichend, ist bei der Messung etwas schief gegangen, oder die Zahnstange ist Ausschuss oder die Parameter wurden falsch eingegeben.
8. 8. Einer dieser Werte (mehr aus der Mitte) wird zum Referenzpunk gemacht. An diesem Referenzpunk hängt man die theoretisch berechnete Ideale Zahnstange auf.
9. 9. Nun lassen sich Berichte ausgeben, auf denen man die theoretische Zahnstange mit den gemessenen Punkten vergleichen kann. Neben einem Schaubild kann man Standardabweichung, Varianz und Unsicherheit der einzelnen Abschnitte darstellen mit farblicher Kodierung bezüglich der Güte der Messung.
10. 10. Nun können die gewünschten Messwerte ausgerechnet werden, die mit den theoretischen verglichen werden. Als da wären:
    • Rollenmaß, Zahnabstand, Teilung, Zahnstangenhöhe, Flankenwinkel, Zahnhöhe. Diese Messwerte können im Programm mit unterschiedlichen Grenzwerten versehen werden, ab denen die Zahnstange eine gewisse Güteklasse erreicht oder Ausschuss ist.

In the clamped state, the measuring points are traversed again. Then the following is done:

1. 1. Using an algorithm that determines the number of points in a narrow band parallel to the x-axis, the maximum over the tooth tips is determined.
2. 2. An interpolation line is drawn through the points.
3. 3. With the y-axis intercept obtained, a new window is calculated parallel to the x-axis (tooth base, m-factor, radius) in which the flanks must lie.
4. 4. Another algorithm that also filters the x-axis values in the width PI * module / 2 can now determine the start and end points of the edges.
5. 5. The interpolated straight line equation for each flank can be calculated from these start and end points.
6. 6. The bisector of each of these straight lines is the center of the tooth or the bottom of the rack.
7. 7. The distance between the bisector provides values that enable a statement to be made about the quality of the measurement. They have to be in a certain window of the theoretical distance. If they deviate too far, something went wrong during the measurement, the rack is scrap or the parameters were entered incorrectly.
8. 8. One of these values (more from the middle) is made the reference point. The theoretically calculated ideal rack is suspended from this reference point.
9. 9. Reports can now be output on which the theoretical rack can be compared with the measured points. In addition to a graph, the standard deviation, variance and uncertainty of the individual sections can be shown with color coding with regard to the quality of the measurement.
10. 10. The desired measured values can now be calculated and compared with the theoretical ones. As there would be:
    • Roll dimension, tooth spacing, pitch, rack height, flank angle, tooth height. These measured values can be assigned different limit values in the program, from which the rack reaches a certain quality class or is rejected.

1. 1. Um die theoretische Zahnstange über die gemessenen Punkte zu legen könnte man anstatt einen Referenzpunkt zu wählen sofort die beiden Zahnstangen übereinander legen und im halbierungsverfahren bestimmen, wann der Abstand aller Punkte zur theoretischen Zahnstange ein Minimum erreicht. Diese Berechnungen sind aufwändiger als die oben beschriebenen zur Bestimmung eines Referenzpunktes, aber das Halbierungsverfahren könnte diese Methode trotzdem schneller machen.
2. 2. Das Halbierungsverfahren könnte auch die oben Beschriebene Bestimmung eines Referenzpunktes beschleunigen.
3. 3. Anstatt Interpolationsgeraden zu verwenden kann man auch jeden aufeinanderfolgenden Punkt mittels einer Geraden verbinden. Dazu sollte stark abweichende Punkte vorher aussortiert werden. Diese Geraden könnten die Punkte zum Vergleich liefern, würden on the fly berechnet und wären somit schneller verfügbar.
4. 4. Ein Schnellmessverfahren könnte nur relevante Punkte auswählen, die zu vergleichen sind: eine Zahnstange wird mit einer Schleifscheibe gewisser Breite gefertigt. Diese Schleifscheibe wird nach Bearbeitung eines Bereiches abgehoben und am anschließenden Bereich wieder aufgesetzt. Es ist also nicht möglich, dass innerhalb der Scheibenbreite großartige Veränderungen passieren. Der Abstand zwischen den Aufsetzstellen ist interessant. Und über die Zahnstange verteilt die Vermessung je eines Zahnes der Scheibe. Hat die Scheibe also 16 Zähne misst man den ersten Zahn auf der ersten Bearbeitungsbreite, den Zweiten auf der zweiten und so weiter. So wird jeder Zahn der Schleifscheibe kontrolliert und man kann trotzdem eine Verkürzung der Zähne der Schleifscheibe bestimmen, wenn man annimmt das alle Zähne gleich hoch abgerichtet sind. Die Genauigkeit kann bei Maschinen +/-1 µm betragen.
5. 5. Stark Abweichend Punkte könnten nach Übereinanderlegen der theoretischen Zahnstange auf die gemessenen Punkte aussortiert werden, wenn sie einen gewissen unmöglichen Abstand haben (Messabstand 40µm: ein angrenzender Punkt kann nie mehr als 40µm nach oben und unten Abweichen (das wären 45°) wenn eine Zahnstange 20° Flankensteilheit hat. Somit kann bei der Punkte-mit-Zwischengeraden-Methode vermieden werden einen Ausreißer zum Vergleich zu erwischen.
6. 6. Um Zeit bei der Messung zu sparen könnten mehrere Messköpfe die Daten parallel erfassen.

Further optimizations:

1. 1. To place the theoretical rack over the measured points, instead of choosing a reference point, you could immediately place the two racks on top of each other and use the halving method to determine when the distance between all points and the theoretical rack reaches a minimum. These calculations are more complex than those described above for determining a reference point, but the halving method could still make this method faster.
2. 2. The halving process could also accelerate the determination of a reference point described above.
3. 3. Instead of using interpolation lines, you can also connect each successive point by means of a straight line. For this purpose, points that differ greatly should be sorted out beforehand. These straight lines could provide the points for comparison, would be calculated on the fly and would therefore be available more quickly.
4. 4. A high-speed measurement method could only select relevant points that are to be compared: a rack is manufactured with a grinding wheel of a certain width. This grinding wheel is lifted off after machining an area and put back on in the subsequent area. So it is not possible for great changes to happen within the slice width. The distance between the touchdown points is interesting. And the measurement of one tooth of the disk is distributed over the rack. So if the disc has 16 teeth, the first tooth is measured on the first machining width, the second on the second and so on. So every tooth of the grinding wheel is checked and you can still determine a shortening of the teeth of the grinding wheel if you assume that all teeth are dressed to the same height. The accuracy of machines can be +/- 1 µm.
5. 5. Strongly deviating points could be sorted out after superimposing the theoretical rack on the measured points if they have a certain impossible distance (measuring distance 40µm: an adjacent point can never deviate more than 40µm up and down (that would be 45 °) The tooth rack has a slope of 20 °, which means that the point-with-straight line method can be used to avoid catching an outlier for comparison.
6. 6. In order to save time during the measurement, several measuring heads could record the data in parallel.

Further advantageous refinements result from the further subclaims or their possible sub-combinations.

• 1 es sind in schematischer Darstellung gängige Zahnstangen in Gerade- und Schrägverzahnung gezeigt,
• 2 eine schematische Darstellung der Parameter eines Zahns einer Zahnstange,
• 3 eine schematische Darstellung der Teilung der Verzahnung einer Zahnstange,
• 4 eine schematische Darstellung der Teilung der Verzahnung einer schrägverzahnten Zahnstange,
• 5 eine schematische Darstellung der Rollenmaßberechnung einer Zahnstange,
• 6 eine schematische Darstellung eines Messplatzes für eine Zahnstange,
• 7 eine schematische Darstellung der Messstrecke an der Seite und oben an der Verzahnung der Zahnstange,
• 8 eine schematische Darstellung der Zusammenwirkung des Messkopfes mit der Zahnstange über eine optische Abtastung,
• 9 Detail der Messanordnung,
• 10 Ergebnis der Analyse zur Abweichung.

The invention is further explained below with reference to the drawings. In detail, the schematic representation shows in:

• 1 Common racks with straight and helical teeth are shown schematically,
• 2 a schematic representation of the parameters of a tooth of a rack,
• 3rd a schematic representation of the pitch of the toothing of a rack,
• 4th a schematic representation of the pitch of the toothing of a helical rack,
• 5 a schematic representation of the roller dimension calculation of a rack,
• 6th a schematic representation of a measuring station for a rack,
• 7th a schematic representation of the measuring section on the side and on top of the toothing of the rack,
• 8th a schematic representation of the interaction of the measuring head with the rack via optical scanning,
• 9 Detail of the measuring arrangement,
• 10 Result of the analysis of the deviation.

The same reference numbers in the figures denote elements that are the same or have the same effect.

In 1 a) a straight toothing is shown in 1 b) a helical toothing.

There are the following parameters: Standard pitch Tn: Distance from one tooth of a rack to the next. Tn = m * pi

$$\text{Zahngrund}=\text{1}\text{,25}*\text{m}$$

$$\text{Zahngrund}=\text{1}\text{,167}*\text{m}$$

Eingriffswinkel 20° ist Standard. Es gibt aber auch Abweichungen.Module m in mm: dimensions for standardizing gears $$\text{R.}=\text{m / 5}$$

$$\text{Tooth base}=\text{1}\text{, 25}*\text{m}$$

$$\text{Tooth base}=\text{1}\text{, 167}*\text{m}$$

Pressure angle of 20 ° is standard. But there are also deviations.

Reference profile: The reference profile corresponds to the theoretical rack profile on which the gear wheel rolls without play. In practice, it is the shape of the tool that is used to produce the gear in the hobbing process. The reference profile of gears in use today is standardized in DIN 867.

See also 3rd (Straight teeth G) and 4th (Helical toothing S)

Calculation formula: $$\begin{array}{l}{\text{T}}_{\text{st}}=\underset{\_}{\text{Division normal}\left({\text{T}}_{\text{n}}\right)}\\ \beta \end{array}$$

example

5 Eine beispielhafte Rollenmaßberechnung mit Eingriffswinkel 20°.

α =

Eingriffswinkel

D_M =

Messrollendurchmesser

M_P =

Prüfmaß

m =

Modul

M_K =

Rollenmaß über Profilmitte

M_Z =

Rollenmaß über Zahnkopf

a =

Maß von Unterseite bis Profilmitte

Module 3 helix angle 19.5281 ° $$\begin{array}{ll}\text{Division normal}\left({\text{T}}_{\text{n}}\right)\frac{\mathrm{9,424778}}{0.94247766}\hfill & \hfill \\ \begin{array}{l}\text{Cosine}\left(\mathrm{\beta }\right)\\ 10.0\text{mm}\end{array}\hfill & =\text{Forehead division}\left({\text{T}}_{\text{st}}\right)\hfill \end{array}$$

5 An exemplary roll dimension calculation with a pressure angle of 20 °.

α =

Pressure angle

D_M =

Measuring roller diameter

M_P =

Test dimension

m =

module

M_K =

Roll dimension over profile center

M_Z =

Roll dimension over tooth head

a =

Measure from the bottom to the middle of the profile

$$\text{M}\_\text{P }=\text{ M}\_\text{K }+\text{ a}$$

$$\text{M}\_\text{Z }=\text{ M}\_\text{K }+\text{ m}$$

Vereinfachung bei Eingriffswinkel α = 20° $$\begin{array}{cccc}\text{M\_K }=& \underset{\_}{\text{D\_M}}& +\underset{\_}{}\text{D\_M}& -\mathrm{2,1578638}\\ x\text{ m}& & & \\ & 2& \mathrm{0,6840403}& \end{array}$$

$$\text{M\_K }=\mathrm{1,9619}\times \text{ D\_M }-\mathrm{2,1578638}\times \text{ m}$$

Formula: $$\begin{array}{ccccc}\text{M.}\_\text{K}& =& \text{D.}\_\text{M.}& +\text{D.}\_\text{M.}& -\mathrm{\pi }\times \text{m}\\ & & \overline{2}& \overline{2\times \text{sin}\mathrm{\alpha }}& \mathrm{4th}\overline{\times \text{tan}}\mathrm{\alpha }\end{array}$$

$$\text{M.}\_\text{P}=\text{M.}\_\text{K}+\text{a}$$

$$\text{M.}\_\text{Z}=\text{M.}\_\text{K}+\text{m}$$

Simplification for pressure angle α = 20 ° $$\begin{array}{cccc}\text{M\_K}=& \underset{\_}{\text{DM}}& +\underset{\_}{}\text{DM}& -2.1578638\\ x\text{m}& & & \\ & 2& 0.6840403& \end{array}$$

$$\text{M\_K}=1.9619\times \text{DM}-2.1578638\times \text{m}$$

6th shows a rack measuring machine that can measure the individual pitch, total pitch, pitch line (pitch circle), roll dimension, flank angle (pressure angle), helix angle, torsion and straightness on racks. You can put together a measurement protocol that contains the required data.

1. 1. Die Zahnstange wird gereinigt.
2. 2. Die Seite der Zahnstange wird im entspannten Zustand an der seitlichen Anlagefläche (Rechtwinkligkeit zur Wälzebene) in zwei Ebenen gemessen, um die Geradheit und Torsion zu erfassen.
3. 3. Die Zahnstange wird auf den Zahnköpfen wie unter 2. gemessen.
4. 4. Die Zahnstange wird magnetisch gespannt.
5. 5. Der konfokale Wegmesssensor vermisst die Zähne. Es werden verschiedene Parameter errechnet und eine Exportdatei erzeugt.
6. 6. Die exportierten Teildaten werden in einem zu entwickelnden Programm zusammengefasst, ausgewertet und zu einem Messprotokoll aufgearbeitet, das auf einem Drucker ausgegeben werden kann.

A measurement consists of the following sections:

1. 1. The rack is cleaned.
2. 2. The side of the rack is measured in the relaxed state on the lateral contact surface (perpendicularity to the rolling plane) in two planes in order to determine the straightness and torsion.
3. 3. The rack is measured on the tooth tips as under 2..
4. 4. The rack is clamped magnetically.
5. 5. The confocal displacement sensor measures the teeth. Various parameters are calculated and an export file is generated.
6. 6. The exported partial data are summarized in a program to be developed, evaluated and processed into a measurement protocol that can be output on a printer.

7th shows: for the measurements on the side and above, the head was tilted. The distortion is factored out.

In 8th the measuring head is shown measured with a rack.

The entire measurement process including the output of the measurement protocol for a rack takes no longer than one to two minutes. Depending on the machine size and design, several racks can also be measured; the racks are either positioned one behind the other or next to one another in the case of several measuring positions. The data could also be passed on digitally to production planning and control systems (PPS). Also optionally, the measuring machine can measure racks fully automatically using an automatic loading and unloading device, output their measurement protocols, communicate with the production planning and control systems (PPS) and sort out the racks that are not dimensionally accurate. This means that the measuring machine can be perfectly integrated into the rack production process.

9 shows the measuring head in a closer view.

Rough measurement

The first measurement is made in two rows with points on the rack heads and in the relaxed state. If the deviations are too great, the rod must be straightened before it is clamped and measured with the Keyence measuring head. By means of these points, a torsion and a straightness can be shown, which flow into the protocol as shown below.

10 shows an exemplary measurement result of a fine measurement.

The rack is clamped diagonally with a special magnet. In the process, both a bending in the x or y direction and a torsion are drawn straight.

For the measurement data, the diagrams for the protocol are generated in Excel, for example. A diagram should show an adjustable upper and lower limit value by means of two lines. The measuring line is then located in between. If it crosses the limit line, it is scrap or a rack of poor quality, depending on the setting. If the rod remains within the lines, the printed report serves as a certificate of quality.

Claims (10)

Method for measuring and qualifying a toothed rack, with a toothing on one side thereof, characterized in that the toothed rack is optically scanned along its course by an optical sensor spaced from the toothed rack, with the toothed rack lying unstressed in at least one in a first measuring step Track is measured along the course of the same and in a second measuring step the rack is mechanically or magnetically clamped in a defined position or clamped against a defined surface lying in at least one track along the course of the same. Procedure according to Claim 1 , characterized in that the optical sensor is designed to be movable along the rack by means of a measuring mechanism. Procedure according to Claim 1 or 2 , characterized in that the optical sensor is formed by an interference measuring head. Method according to one of the Claims 1 to 3rd , characterized in that at least one track along the toothing side is measured in a measurement run. Procedure according to Claim 4 , characterized in that at least one track along a side surface adjoining the toothing side is measured in a measurement run. Method according to one of the preceding claims, characterized in that at least in one of the measuring steps a first measuring run is carried out by means of the optical sensor of 90 ° against the surface to be measured, and in a second measuring run by means of the optical sensor at an angle smaller or greater than 90 ° takes place against the surface to be measured. Method according to one of the preceding claims, characterized in that the first measuring step is offset against the second measuring step, in particular in a differential method, whereby the torsion and curvature of the rack, which deviate from the ideal dimension, is calculated. Device for practicing the method according to one of the Claims 1 to 7th , characterized in that a flat measuring table is provided with a contact edge, and that an optical sensor is designed to be movable along the measuring table by means of a measuring mechanism. Device for carrying out the method according to one of the 1 to Claims 8 , characterized in that the optical sensor is formed by an interference measuring head. Device for carrying out the method according to one of the 1 to Claims 8 , characterized in that an electric magnet is provided for defined clamping and holding under a preselectable force against the flat measuring table and the contact edge, which is able to apply a magnetic field to a rack to be measured.

Images