Method for automatic compensation of probe offset in a coordinate measuring machine

• US 4,866,643 A
• Filed: 10/09/1987
• Issued: 09/12/1989
• Est. Priority Date: 10/09/1987
• Status: Expired due to Term

• Current Assignee: Hexagon Metrology AB (Hexagon AB)
• Original Assignee: Brown & Sharpe Manufacturing Company (BNS Holding Incorporated)
• Inventors: Dutler, Werner
• Primary Examiner(s): Lall, Parshotam S.
• Assistant Examiner(s): MELNICK, STEVEN A
• Application Number:
  US07/107,253
• Time in Patent Office: 704 Days
• Field of Search: 364/571.02, 364/560, 364/559, 364/474.36, 364/474.35, 364/474.37, 364/474.03, 364/167.01, 364/474.18, 33/503, 33/504, 33/505
• US Class Current: 702/95
• CPC Class Codes: G01B 21/045 Correction of measurements G01B9/02055 takes precedence
• CPC Class Codes: G05B 19/401 characterised by control arrangements for measuring, e.g. calibration and initialisation, measuring workpiece for machining purposes G05B19/19 takes precedence
• CPC Class Codes: G05B 2219/37043 Touch probe, store position of touch point on surface
• CPC Class Codes: G05B 2219/37193 Multicoordinate measuring system, machine, cmm
• CPC Class Codes: G05B 2219/50336 Tool, probe offset for curves, surfaces, contouring

Abstract

A method for automatic probe offset compensation in a machine for measuring coordinates of a workpiece surface by contacting points on the surface with a probe having a prescribed diameter and sensing and recording the coordinates of the probe. The method can be applied to planar surfaces and circularly curved surfaces. The method includes determining a direction point in space just prior to contact with the workpiece surface, determining a directional vector between the direction point and the first contact point, and determining the angle of the directional vector relative to a normal to the surface. When the angle is within a predetermined range of acceptance, the direction point is considered valid. In the case of a planar surface, the measured surface is shifted by the probe offset in a direction opposite the normal vector so that the measured surface and the workpiece surface coincide. In the case of a curved surface, the distance from the direction point to the centerpoint of the measured surface is compared with the radius of the measured surface, and the probe offset is added to or subtracted from the radius so that the measured surface and the workpiece surface coincide. In a further refinement, the direction point is considered invalid if it is within the distribution of measured contact points.

Claims

1. In a machine for measuring coordinates of a generally planar surface on a workplace by contacting points with a probe having a prescribed diameter and continuously sampling and recording the coordinates of the probe, a method for automatically compensating the planar surface being measured for probe offset comprising the steps of:

(a) moving the probe toward the planar surface and measuring the coordinates of a direction point D just before contact between the probe and the planar surface;

(b) moving the probe along the planar surface and measuring the coordinates of at least three contact points C₁ -C_N with the probe and the planar surface in contact;

(c) fitting by electronic data processing means a plane S to the contact points C₁ -C_N, the plane S being defined by a normal vector V_n passing through the first contact point C₁ and a coordinate of the intersection between plane S and vector V_n ;

(d) calculating by electronic data processing means a directional vector V_d between the first contact point C₁ and direction point D;

(e) calculating by electronic data processing means an angle A between the normal vector V_n and the directional vector V_d, and when the angle A is greater than 90°

, reversing the direction of the normal vector V_n ;

(f) when the angle A is in the range between the predetermined range of acceptance and 90°

, rejecting the direction point D as invalid and repeating the compensation procedure; and

(g) when the angle A is within a predetermined range of acceptance, moving the plane S by electronic data processing means by the dimension of the probe offset in a direction opposite the direction of the normal vector V_n so that the plane S coincides with the planar surface, thereby compensating for probe offset.

2. A method for probe offset compensation as defined in claim 1 wherein the step of measuring the coordinates of direction point D includes the steps of measuring the coordinates of two sample points just before contact between the probe and the planar surface and averaging the coordinates of the two sample points.

3. A method for probe offset compensation as defined in claim 1 wherein the step of fitting a plane S to the contact points C₁ -C_N includes the step of utilizing a least squares fit algorithm to calculate a definition of the plane S.

4. A method for probe offset compensation as defined in claim 1 wherein the predetermined range of acceptance is about ±

88 degrees relative to the normal vector V_n.

5. A method for probe offset compensation as defined in claim 1 further including the steps of

defining a plane S_max parallel to plane S and passing through the contact point determined by electronic data processing means to have maximum deviation from plane S in the direction of the normal vector V_n,

defining a plane S_min parallel to plane S and passing through the contact point determined by electronic data processing means to have maximum deviation from plane S in the direction opposite to the normal vector V_n, and

rejecting the direction point D as invalid it is located between the planes S_max and S_min and repeating the compensation procedure.

6. A method for probe offset compensation as defined in claim 1 wherein the step of repeating the compensation procedure further includes the steps of

measuring the coordinates of a new direction point D'"'"' approximately perpendicular to the last contact point C_N by moving the probe off the surface in a direction approximately perpendicular to the contact point C_N and taking a coordinate measurement,

defining a new normal vector V_n '"'"' passing through the last contact point C_N,

calculating a new directional vector V_d '"'"' between the last contact point C_N and the new direction point D'"'"', and

repeating steps (e)-(g) with vectors V_n and V_d replaced by vectors V_n '"'"' and V_d '"'"', respectively.

7. A method for probe offset compensation as defined in claim 5 wherein the step of repeating the compensation procedure further includes the steps of

measuring the coordinates of a new direction point D'"'"' approximately perpendicular to the last contact point C_N by moving the probe off the surface in a direction approximately perpendicular to the contact point C_N and taking a coordinate measurement,

defining a new normal vector V_n '"'"' passing through last contact point C_N,

calculating a new directional vector V_d '"'"' between the last contact point C_N and the new direction point D'"'"', and

repeating steps (e)-(g) with the vectors V_n and V_d replaced by vectors V_n '"'"' and V_d '"'"', respectively.

8. In a machine for measuring coordinates of a curved surface on a workpiece by contacting points on the curved surface with a probe having a prescribed diameter and continuously sampling and recording the coordinates of the probe, a method for automatically compensating the curved surface being measured for probe offset comprising the steps of:

(a) moving the probe toward the curved surface measuring the coordinates of a direction point D just before contact between the probe and the curved surface;

(b) moving the probe along the curved surface and measuring the coordinates of at least three contact points C₁ -C_N with the probe and the curved surface in contact;

(c) fitting by electronic data processing means a curve S to the contact points C₁ -C_N, the curve S being defined by a center O and a radius R;

(d) calculating by electronic data processing means a tangential vector V_t tangent to the curve S at the first contact point C₁ ;

(e) calculating by electronic data processing means a directional vector V_d between the first contact point C₁ and direction point D;

(f) calculating by electronic data processing means an angle A between the tangential vector V_t and the directional vector V_d ;

(g) when the angle A is outside the predetermined range of acceptance, rejecting the direction point D as invalid and repeating the compensation procedure;

(h) when the angle A is within a predetermined range of acceptance and the distance between direction point D and the center O of curve S is greater than the radius R of curve S, subtracting the probe offset from the radius of the curve S by electronic data processing means so that the curve S coincides with the curved surface, thereby compensating for probe offset; and

(i) when the angle A is within the predetermined range of acceptance and the distance between the direction point D and the center O of curve S is less than the radius R of curve S. adding the probe offset to the radius R of the curve S by electronic data processing means so that the curve S coincides with the curved surface, thereby compensating for probe offset.

9. A method for probe offset compensation as defined in claim 8 wherein the step of measuring the coordinates of direction point D includes the steps of measuring the coordinates of two sample points just before contact between the probe and the curved surface and averaging the coordinates of the two sample points.

10. A method for probe offset compensation as defined in claim 8 wherein the step of fitting a curve S to the contact points C₁ -C_N includes the step of utilizing a least squares fit algorithm to calculate a definition of the curve S.

11. A method for probe offset compensation as defined in claim 8 wherein the predetermined range of acceptance is between about 2 and 178 degrees relative to the tangential vector V_t.

12. A method for probe offset compensation as defined in claim 8 further including the steps of

defining a curve S_max parallel to curve S and passing through the contact point determined by electronic data processing means to have maximum deviation from curve S in one radial direction,

defining a curve S_min parallel to curve S and passing through the contact point determined by electronic data processing means to have maximum deviation from curve S in the opposite radial direction;

and

rejecting the direction point D as invalid if it is located between the curves S_max and S_min and repeating the compensation procedure.

13. A method for probe offset compensation as defined in claim 8 wherein the step of repeating the compensation procedure further includes the steps of

measuring the coordinates of a new direction point D'"'"' approximately perpendicular to the last contact point C_N by moving the probe off the surface in a direction approximately perpendicular to the contact point C_N and taking a coordinate measurement,

calculating a new tangential vector V_t '"'"' passing through the last contact point C_N and tangent to the curve S,

calculating a new directional vector V_d '"'"' between the last contact point C_N and the new direction point D'"'"', and

repeating steps (f)-(i) with the vectors V_t and V_d replaced by vectors V_t '"'"' and V_d '"'"', respectively.

14. A method for probe offset compensation as defined in claim 12 wherein the step of repeating the compensation procedure further includes the steps of

measuring the coordinates of a new direction point D'"'"' approximately perpendicular to the last contact point C_N by moving the probe off the surface in a direction approximately perpendicular to the contact point C_N and taking a coordinate measurement,

calculating a new tangential vector V_t '"'"' passing through the last contact point C_N and tangent to the curve S,

calculating a new directional vector V_d '"'"' between the last contact point C_N and the new direction point D'"'"', and

repeating steps (f)-(i) with the vectors V_t and V_d replaced by vectors V_t '"'"' and V_d '"'"', respectively.