Printer Friendly

Force-field sensing for QC.

Force-field sensing for QC

In the late '60s, aerodynamics researchers at the University of California, Berkeley, studying wind-tunnel effects and some later work at MIT discovered the reciprocal relationship between force fields (acoustic and electromagnetic) and objects in those fields. Simply put: An object's shape can be used to determine the field around it, and vice versa. The field "encodes" the object, and if the object changes, even slightly, the field changes.

Here was a way to inspect parts: create a field around the object, measure the field, and correlate changes in that field to changes in the object. The field choices include acoustic (from a fixed-frequency horn), inductive (generating coils), capacitive (charged plates), and microwave (microwave horn). The more sensors measuring the field, the better the object is described. For even more information, multiple field types can be used simultaneously without imteracting or interfering with each other.

As an object passes through multiple fields, it distorts each differently, yielding different information in a single pass. Since only field distortion is being measured, the residual field effect of jigs and fixtures can be ignored.

Multiple fields

The farious fields vary sinusoidally, growing and collapsing at their own fixed frequency, and their matching sensors determine only two values: signal amplitude and phase shift. Acoustic fields range from 20 kHz to 500 kHz, inductive and capacitive from 1 kHz to 10 MHz, and microwave from 1 to 10 GHz. Multiple fields of the same type can choose different fixed frequencies and then use filters to separate the sensor signals.

Because each sensor determines only phase shift and amplitude, determining field distortion caused by the measured object is greatly simplified. Even arrays of hundreds of sensors ar a simple analysis task for a work-station-grade computer. In comparison, machine-vision systems typically analyze half a million values during a single inspection. By choosing appropriate force fields, the number of sensors--and datahandling task--can be kept relatively small, yet fully define the object for inspection purposes.

Random isn't statistical

Where is force-field inspection useful? Many assembly problems are caused by random inconsistencies of batch-produced parts, where common equipment produces many different part numbers, particularly plating, heat-treating, degreasing, deburring, etc. Although wandering dimensional trends can be tracked and corrected for with SPC methods, random manufacturing errors cannot. SPC doesn't help when the bolt before and the bolt after are perfectly good, yet the bolt in between is bad. And if these bad parts are not rejected, they can jam up the automation downstream. It's when the process wanders out of tolerance and produces all bad parts that SPC shines.

Short shots in injection molding, damaged leads on surface-mounted ICs, bar ends on screw-machined parts, and inconsistent heat treatment on stampings are examples of random process errors seldom detected by SPC. Other manufacturing defects that sometimes slip by are broken punches on headers, broken taps on nut formers, and broken bits on turned parts. All of these are candidates for force-field inspection.


Force-field measurements can be used to ensure consistent part quality. For example, consider a connector component made of plastic with embedded metal connector pins. The shape and dimensions of the plastic can be determined from acoustic-field values, and the shape and composition of the metallic pins by inductive-field values.

Changes in values imply changes in a given connector. Like most production parts, no two connectors are exactly the same, but by statistically studying the values for good parts, boundary conditions can be established to sort out reject parts. The acoustic values will judge the plastic portion (objects smaller than the acoustic wavelength have little effect on the acoustic field) and the inductive values will judge the metal-pin portino (inductive fields react only to metals). This system will detect flaws in the plastic, flaws in the pins, and even flaws in the relationships between metal and plastic.

The technology also can be used in the assembly process. Consider an insertion device that picks up parts from a feeder and inserts them into an assembly. Inductive and capacitive sensors can be epoxied into the insertion tip and detect a missed part, a dropped part, a mispositioned part, or an incorrectly assembled part. Depending on the application, force-field checks can be 10 to 100 times/sec.

Generally, an ultrasonic parts-qualifying system can check up to 5 parts/sec wth a resolution of 0.001". Accuracy depends on the distance between sensor and object. Distance distors the field, so accuracy improves as distance diminishes. Inductive and capacitive fields are best for close-range measurement. Their force fields are distorted by objects with standoff distances approximating the sensor dimensions, i.e., a 1"-dia capacitive plate or inductive coil best senses objects an inch or less away. These are exponential fields, their sensitivity falls off exponentially with standoff distance. Acoustic and microwave fields are sinusoidal, falling off more slowly with standoff, and range is not affected by sensor size, but by the wave length of the driven frequency. Acoustic sensors have a practical range of several wave lengths, and microwave sensors an even greater range. Both fields are distorted by objects several inches or more from the field sensors.

Advantages and disadvantages

The biggest advantage of forcefield sensing is that accurate checks can be made without human assistance. The system checks each part against the statistical norm of good parts and rejects any inconsistency. In the insertion application case cited, after monitoring the first few dozen good parts, subsequent assembly operations are checked for consistency. No operator is needed to program the system to detect what is unacceptable. Setup time is a minute or two of operation.

Yet, how certain can you be that the consistency of the assembled objects is actually good enough? The force-field approach makes no absolute measurements. Although it detects when a diameter is a few thousandths off, it does not yield a digital measurement or measure specific material composition. It only determines consistency.

Force-field measurements, however, can be verified by running known rejects through the test station. By determining the values of these known defects or assembly errors, you can confirm that even slight defects will be rejected.
COPYRIGHT 1989 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Buckley, Shawn
Publication:Tooling & Production
Date:Feb 1, 1989
Previous Article:Insert update - the right tool boosts profits.
Next Article:Flexibility in contour cuts - the case for plasma.

Related Articles
Minding your QCs.
The quest for quality: sixth-inning scorecard.
High costs of court action.
Tributes paid to lawyer with music skill.
Asylum kids plan slammed.
STABBED TO DEATH.. FOR LOOKING AT TRAIN THUG; Tom was just minding his own business.
Use some horse sense with QC.
Life for the 'kebab' killer who stabbed his best friend through the heart.
Luxury or necessity? Producers still aren't convinced that automatic inspection is a need, not a want.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters