Method and apparatus for detecting faults in an object

• US 4,876,727 A
• Filed: 01/30/1987
• Issued: 10/24/1989
• Est. Priority Date: 02/10/1986
• Status: Expired due to Fees

• Current Assignee: NUKEM GmbH (Cameco Corporation)
• Original Assignee: NUKEM GmbH (Cameco Corporation)
• Inventors: Fabian, Josef, Maurer, Albrecht
• Primary Examiner(s): Boudreau, Leo H.
• Application Number:
  US07/008,950
• Time in Patent Office: 998 Days
• Field of Search: 382/34, 382/54, 382/30, 382/8, 382/42, 358/106, 73/602, 73/587, 73/596, 73/598, 73/606, 356/237
• US Class Current: 382/149
• CPC Class Codes: G01B 17/06 for measuring contours or curvatures

Abstract

The subject of the invention is a method and a device for detecting faults in or on an object. The object form is measured and then compared with a master form. In this comparison one point and one line of the object form and of the master form with reference to a coordinate system, are brought into coincidence by shifting the form or master form. Partial forms and partial master forms are then displaced relative to one another in partial areas of the coordinate system until a maximum overlap is detected. Then the nonoverlapping points of the partial forms are checked for faults.

Claims

1. A method for nondestructively detecting faults in and an object comprising the steps of:

measuring fault related values at various points of said object;

storing said measured fault related values with respect to a coordinate system in memory locations;

obtaining master values which relate to an object not having faults;

storing said master values with respect to said coordinate system in memory locations;

comparing said measured values and said master values to each other so that corresponding edges or corners are found;

aligning said measured values and said master values by restoring said measured values in said memory so that said corresponding edges and corners have corresponding coordinates to said master values;

defining partial areas of said aligned measured values and corresponding partial areas of said aligned master values so that each of said partial areas and corresponding partial areas have the same coordinate distances from said edge;

aligning said measured values in said partial areas and said master values in said corresponding partial area by restoring said measured values of said corresponding partial areas in memory locations to obtain minimal differences between said measured values and master values; and

determining a fault difference between said twice aligned measured values in said partial areas and said twice aligned master values in said corresponding partial areas to determine if said fault exists.

2. A method according to claim 1 wherein:

said object has a constant thickness; and

said second alignment step includes the step of shifting said master values or said measured values located in said partial area and stored in said memory to obtain maximum overlap of said master values and said measured values in said partial areas.

3. A method according to claim 1 further including the step of displaying said fault difference on a monitor.

4. A method according to claim 1 wherein said second step of aligning includes the steps of:

subtracting corresponding measured values from said master values to obtain first difference values;

shifting said measured values or said master values located in said partial area;

subtracting corresponding measured values and master values after said shifting to obtain second difference values;

comparing said first and second difference values to determine an absolute difference; and

shifting said master values or said measured values located in said partial area and stored in said memory on the basis of said absolute difference to obtain said alignment.

5. A method according to claim 4 wherein:

an integral of said first difference values and second difference values are compared in said difference comparison step;

said substracting and difference comparison steps are repeated until a minimum absolute value is obtained; and

said shifting step uses said minimum absolute value to obtain said alignment.

6. A method according to claim 4 further including the step of using coordinates of said twice aligned values as starting coordinates to determine alignment of measured values and master values in an adjacent partial area.

7. An apparatus for non-destructively detecting faults in an object through comparison with predetermined master values associated with a fault-free object comprising:

means for measuring fault related values at various points of said object;

first memory means for storing said measured values;

second memory means for storing said predetermined master values; and

means for detecting said faults using said measured values and said master values, said detecting means including;

means for associating said measured values and said master values stored in said first memory means and said second memory means with respect to a coordinate system,

means for aligning said measured values and said master values stored in said first and second memory means, respectively,

means for determining partial areas of said measured values and corresponding partial areas of said master values,

means for aligning said measured values and said master values located in said determined partial areas, respectively,

means for detecting faults by comparing said measured values and said master values located in said corresponding partial areas; and

means for indicating when said detecting means detects said fault.

8. An apparatus according to claim 7 wherein said measuring means includes an ultrasonic test head and an ultrasonic test instrument.

9. An apparatus according to claim 8 wherein said ultrasonic test head includes actual position value transmitters.

10. An apparatus according to claim 7 wherein said indicating means is a monitor capable of graphically illustrating said fault.