Exacting Standards

Measuring perpendicularity and parallelism (E1 and E2) of compression springs using vision technology

By Larry Sheiman
SAS Inc., a DTechXion Ltd. company

With a background in the aircraft industry, I am acutely aware of the concept of extended product liability. This liability is generally associated with the reliability of primary flight-critical components that would, if not manufactured to exacting tolerances, cause the loss of the aircraft or, at best, a serious accident. Having now been associated with the spring industry for 20 years, it is apparent that the motor industry is driving extended product liability to the same exacting standards as aviation. What this means to the embattled springmaker is strict adherence to various tolerances and, in more recent times, the measurement and control of perpendicularity and parallelism (E1 and E2) in compression springs (among a plethora of other variables).

Manual Measurement

More often than not, E1 and E2 measurement is performed on straight squared-and-ground compression springs. Conical springs can also be inspected for E2, but concentricity replaces E1 requirements for such springs. Every spring company has its own (often proprietary) method of determining E1 and E2. In general, these methods are all able to cope with the measurements when the compression spring is greater than 1" (25.4mm) in free length and the spring’s outside diameter is greater than 0.4" (10mm). Springs conforming to these approximate dimensions can be held using the forefinger and thumb by most quality inspectors. When springs are smaller than this, grasping and measuring them can be a problem.

Another challenge for inspectors is that frequently (certainly when it comes to DIN and now EN standards) the tolerances for E1 and E2 are expressed as a percentage of L0 (free length) and OD (outside diameter). Grade 1 tolerances (the highest quality standards for E1 and E2) call for tolerances of 3% of L0 and 1.5% of OD. Now, not only is it nearly impossible to measure the springs, but also the tolerances are approaching the limit of the inspector to visually verify in-tolerance conditions (E1 within 0.03"/0.76mm and E2 within 0.006"/0.15mm).

The most common method used for determining E1 is the use of a right-angle block and a feeler gauge expressing the limit of E1. The spring is placed within the right angle and rotated until the gap between the vertical section of the right angle and the edge of the spring (top or bottom) is a maximum. If it is possible to insert the feeler gauge in the described gap, then the spring falls outside the E1 tolerance requirement. The spring is then flipped and the process repeated. This process can take 30 seconds to a minute, certainly, when the gap is borderline with respect to the tolerance. Furthermore, repeatability of inspection between inspectors is affected by the decision as to the rotational position of the maximum gap and the process of inserting the feeler gauge, possibly moving the spring.

E2 is generally measured by placing the spring on a flat surface, and then rotating it while inspecting the change in height of the top surface with respect to the rotated position using a regular height gauge. One of the influential factors determining reliability of this method is that the contact force on the spring (from touching the spring using the height-gauge crosshead) varies between inspectors, resulting in a direct variation in the E2 determination.

Optical Comparator Inspection

An alternative to the right-angle block and height gauge is the use of an optical comparator. In most cases, however, these comparators do not include a turntable to automatically rotate the spring so that the maximum E1 and E2 conditions can be observed before a measurement is taken defining their disposition. It is time consuming to find this maximum condition using the comparator. Furthermore, comparators (though accurate devices) can suffer from repeatability issues when different operators extract the E1 and E2 measurements using two independent axes (X and Z). The repeatability of the comparator is based on the subjective decision of operators to assess the position in 2D space of the edges of the springs after consideration of the maximum E1 and E2 spring rotation procedure. Inspection times for a single spring can certainly be minutes. When batch inspection (or 100% inspection) is required, this can impact the price and profitability of producing springs to E1 and E2 requirements.

Digital-Optical Inspection

A solution to the described measurement uncertainties and time constraints is the use of digital-optical inspection methods (vision technology). There is one absolute requirement when using such a method, however; namely, the incorporation of a telecentric lens within the inspection equipment. A telecentric lens minimizes the perspective distortion created by “normal” lenses, since its ray angle is kept to a minimum (telecentricity should be less than 0.3º). The CTV1600 optical inspection system (from SAS Inc.) incorporates such a lens. This lens allows it to measure compression springs with L0 up to 8" (200mm) and OD up to 1.7" (45mm) using its standard supplied telecentric machine-vision digital-optical system. Repeatability of inspection for both E1 and E2 is typically within 0.001" (0.025mm) with accuracy of 0.0004" (0.01mm), independent of the operator.

A system such as the CTV1600 (Figure 1, page 57) is typically programmed per part number. The spring to be inspected is selected by the operator from a pre-programmed list of parts. The spring is placed on a servo-controlled turntable, and the turntable rotates automatically while the spring’s entire side-form is digitally captured using an industrial CMOS camera (the standard camera supplied with the system is a 2 megapixel variant). A proprietary patented algorithm is used to measure E1 and E2 during the rotation process. Typical inspection time for both measurements is 20 seconds or less, if the patented auto-flip algorithm is used to mathematically flip the spring for E1 measurements. The results for E1 and E2 are stored for SPC and reporting purposes. OD and L0 can also be selected for inspection and storage.

The advantages of such a system are its repeatability and speed, when compared with conventional methods. Furthermore, test-result automatic storage for SPC and general quality reporting further reduce the workload on quality-control personnel. Most importantly, it is easy to “convince” the customer that E1 and E2 conform to quality requirements, since the customer can perform the inspection using the equipment in the same way and with the same method as the quality inspectors.

Larry Sheiman is managing partner in SAS Inc., a DtechXion Ltd company. He holds a master’s degree in aeronautical engineering and was introduced to the spring industry in the mid ’80s while developing a spring-design program for springmakers. In 1988, he developed a fully servo-controlled compression/tension/torsion spring testing system. Subsequently, he has been involved in developing non-destructive testing solutions and has been instrumental in the supply of robotic industrial ultrasonic inspection systems to the aircraft industry.

The SAS Inc line of equipment includes: compression, tension and torsion systems; fatigue-testing machines; pitch trace and OD map optical-inspection machines; compression spring E1, E2 and general measurement optical-inspection machines; and generalized XYZ automated optical-inspection systems for 2D geometric measurement of parts. SAS is represented in North America by Forming Systems Inc. Readers may contact Forming Systems by e-mail at info@formingsystemsinc.com, phone at (269) 679-3557 or fax at (269) 679-3567.