|
|
 on the Automotive Assembly line When an automaker wanted to upgrade its inspection process for detecting the absence of connecting rod inserts, it turned to an automated ultrasonic method that has proved accurate, objective, and fast. John Billings, As part of a major overhaul of its piston-stuffing area, a Big Three automobile manufacturer in Romulus, Michigan, wanted to include online, automated inspection for missing connecting rod bearing inserts. Engines missing their bearing inserts require costly rework if left undetected until the hot test; if allowed to reach the consumer, premature, catastrophic engine failure results. The inspection method in use at the time was to manually tap the rod caps with a hammer and listen for the distinctive sound made when inserts are missing. While generally effective, this technique is dependent on the skill and the attention level of the operator and can be affected by the amount of environmental noise present at the time of testing. Conventional Technologies Among the candidate technologies for detecting missing inserts were: While all three methods can potentially identify engines with missing inserts, each has
Listening methods, both automatic and manual, fail because ambient noise components can mimic the characteristic sound of an engine with missing inserts. Spurious noise such as that made by sparks or clanking parts is common in engine manufacturing environments, and has the potential to cause false rejects. Pressurized air systems do an excellent job of catching engines with air leaks, but provide ambiguous results in missing connecting rod bearing insert detection. Failed engines may or may not have missing inserts, and require further inspection to determine the exact cause of the air leak. An additional drawback of a pressurized air system is the difficulty of obtaining a good seal prior to pressurizing the block; a good seal requires complex fixturing and a small amount of oil in the engine. An Ultrasonic Solution
The system ultimately selected for the Romulus installation was provided by Ultrasonic Arrays, Inc., a manufacturer of noncontact thickness, distance, and internal bond measurement equipment. The core of the system is the DMS-1000 distance measuring gauge, accurate to ±0.001 in. or ±0.1% of the distance from the sensor to the target, whichever is greater. The gauge’s ability to maintain its accuracy in changing environments derives from the sensor reference bar system used by the electronic controller. Each time the system takes a measurement, it determines the current environmental conditions by measuring the distance to the reference bar located in front of the face of the sensor. The system then adjusts for speed-of-sound differences caused by changes in the environment—mainly temperature. At the start of the line, each bare engine block is fixtured on a pallet. The block is then transferred on a conveyor to succeeding stations where additional operations and assemblies are performed. On some older lines, called drag lines, the engine is placed on a pallet and transferred continuously past various assembly stations. In this case, the UAI system is mounted on a counterbalance, “clamped” to the block, and follows the engine as the torque-to-turn procedure is performed. UAI’s sensors are permanently mounted perpendicular to the piston surface, out of the way during block transfer, whether the block is presented pan side up, as at the Romulus plant, or down. The benefits are simplified fixturing with no moving parts, resulting in lower installation and maintenance costs, and less downtime due to wear and tear on the fixturing and sensors. The sensor system is installed at the torque-to-turn station, where an electronic signal initializes the DMS-1000, causing the engine to be automatically cycled through three or more revolutions at 50 rpm. While the piston moves through its stroke, the gauge continuously measures the distance to the piston surface without coming into contact with it (see Photo 1). Several UAI installations inspect pistons with valve relief machined into the face; this does not affect system accuracy. The system measures the distance to the face of the piston by sending an ultrasonic pulse from the transducer to the piston surface and timing the return signal (see Figure 1). The microprocessor-based transducer controller converts the time of flight of the return pulse to distance. The highest and lowest measurements to the piston surface are stored in memory by the DMS-1000’s total indicated runout (TIR) software. The low measurement is then subtracted from the high to determine the piston stroke, which is displayed by the DMS-1000. A cylinder with a missing insert will exhibit a shorter stroke than a properly assembled one because the piston does not achieve full travel. The heart of the measurement system, the DMS-1000, is extremely well suited to industrial environments, with excellent immunity to electromagnetic and acoustic noise. The system controller also includes special curve-fitting software that eliminates the possibility that readings caused by noise sources will be considered as valid data. In addition, both maximum and minimum readings are available with the push of a button at the gauge via an RS-232 serial communications port. Unlike systems that listen for a noise signature, the UAI technique uses its own external reference to compensate for environmental changes and is completely self-calibrating. Measurement accuracy depends on the standoff distance to the piston face, speed of rotation of the crank, and number of rotations. A typical installation with a 3.5 in. stroke can achieve accuracies of ±0.003 in., a level significantly beyond that required to detect a missing bearing insert. It is not required that the sensor be an exact distance from the block, since the system uses a differential measurement. This is an important advantage in that fixture movement caused by expansion or contraction does not create any measurable error. ROI Example Payback varies, depending on the method of detection previously used and the number of engines with missing connecting rod bearing inserts produced. For example, in a plant where hot test detects eight engines with missing inserts each week, the following results can be expected. The engines, if caught at hot test, must be torn down and rebuilt, and any parts damaged during the test must be replaced. Assuming each engine involves 20 hr of teardown and rework at a cost of $40/hr, the cost of those eight engines is $6400. The cost for the year is $332,800. Additional costs not included are those for replacing damaged parts and retesting the engine after rework. The UAI system eliminates all these costs, as well as those entailed by conventional methods of inspection. The Ultrasonic Arrays system installed in March 1996 is detecting and identifying cylinders with missing inserts with unprecedented accuracy. Before the installation, defective engines leaving the piston-stuffing area were normally discovered at the hot test station. Detecting the missing inserts prior to hot test at the torque-to-turn station significantly reduces rework expenses when compared to tearing down a fully assembled engine after the hot test. In addition, the system ensures that engines with missing inserts will not be shipped to assembly plants and ultimately reach the consumer. Engine manufacturers can expect to realize the following benefits:
John Billings is President and CEO, Ultrasonic Arrays, Inc., 18538 142nd Ave. NE, Woodinville, WA 98072; 425-481-6611, fax 425-481-4455, info@ultrasonicarrays.com |
|
|
