Here's what happens too often in engine plants all over the world. After many assembly steps, a gas or diesel engine comes off the line. An operator places it in a hot-test stand to make sure all parts are together correctly. However, the test reveals that oil pressure is too low, though it doesn't say why. So the engine goes to the repair stand, and all the time and money spent building it is now wasted.
Here's what can happen using a new method called cold-testing. Specific defects, such as valve leaks, under or over clearanced bearings, missing or damaged piston rings, and other problems are exposed with remarkable precision.
Manufacturers have known about cold-testing for years. But typical ac drives cannot control the speeds and load variations it requires. So cold-test systems were little more than sketches on paper until the development of alldigital, ac-vector drives.
From hot to cold
When it comes to detecting problems, hot testing will tell you where to look, but that's about all. It identifies the symptoms, such as low oil pressure, but not the causes, such as a defective oil pump. And if there is a flaw, the high loads and speeds required in hot testing can cause further damage.
Hot test also requires a lot of time and floor space, as well as large amounts of exhaust, coolant, fuel, and fire protection. So to eliminate or reduce these problems, many engine manufacturers would prefer cold
On the other hand, cold-testing can pinpoint specific defects, reduce the risk of engine damage during testing, provide real-time data on the manufacturing process, and save money because it's done before an engine is completely assembled and dressed.
Cold-test also eliminates fuel combustion impulses, which, combined with low test speeds, provides extreme signal sensitivity. This is how the method gets to the root of the problem, such as a loose spark wire in cylinder No. 5, or an exhaust valve seat interface problem in cylinder No. 2.
In addition, lower speeds and loads reduce the potential for collateral damage to a defective engine. The testing takes less time too because there's no need for engine warm-up. Lastly, there's less chance of exhaust, fuel, and coolant leaks contaminating the workplace.
The only thing missing has been the right drive. Cold testing requires an external electric drive that can rotate an engine at low speeds, typically 50 to 200 rpm. While the engine turns, analog to digital conversion hardware, sampling at rates from 200 kHz to 1 MHz, measures engine responses. The data are synchronized to crankshaft position with resolutions of 1 to 88 points per degree. A data set usually represents one engine cycle or two crankshaft revolutions.
A typical test measures the torque required to rotate the engine crankshaft while driving the piston. This torqueto- turn test will verify whether the engine was properly assembled.
Driving the solution
Maintaining crankshaft velocity within close tolerances is critical to cold testing accuracy. Before ac vector drives, accuracy was low because the early drive systems couldn't maintain engine speed against a load that varied drastically during engine cycles. For example, velocity might fluctuate ±40% at each cylinder compression.
Test system manufacturers were able to reduce these fluctuations to ±20% by adding a flywheel to the drive line. However, this increased drive system complexity and cost, and velocity variations were still too high for good diagnostic accuracy.
By using a digital ac vector drive, with some modifications for test stand and dynamometer functions, developers have found a way to reduce these variations. The drive's closed-loop control updates motor position, velocity, and torque based on feedback from an incremental encoder sampling at 4 kHz. Encoder feedback maintains velocity fluctuations to within ±3%, compared to ±20% with flywheels.
Other special features of this modified drive include a torque estimator and a digital current regulator. By monitoring motor current and voltage, the estimator can detect fluctuations faster than the position encoder. It also compensates for temperature changes and other variations. And, because it eliminates the need for an expensive torque transducer, it reduces the size and cost of the entire drive line.
The digital current regulator operating at 10 kHz increases the response time of the motor drive. The net result is that the drive reacts to changing loads within about 10 msec, reducing velocity variations to about ±3%.
Engine mapping is a critical step in cold testing. It's the process of determining expected engine pressure, load, and flow signatures while rotating the engine at a specific speed. With the information they gather, engineers can optimize final test speeds and measurement configurations – such as open versus closed intake, single or multiple cylinders per sensor, and forward or reverse rotation – for maximum diagnostic sensitivity. Engine mapping also helps determine variability among a batch of engines.
Here's a closer look at what coldtesting can reveal.
When evaluating valve-train and cylinder integrity, the test compares exhaust pressure on the closed valves throughout the cycle to detect leaks. A pressure drop when the valve first opens can help determine the exhaust valve opening angle.
Peak port pressure is a function of cylinder compression integrity. A pressure drop indicates intake valve opening position. Mean port pressure is an indicator of valve leakage.
An exhaust rocker arm is "loose" if the exhaust opening position is 3 to 40 degrees late and there's a high peak exhaust pressure. An exhaust valve face or seat leak shows up as an increase in exhaust port pressure during the cylinder compression cycle accompanied by high opening pressure.
Also worth mentioning is the ability to measure camshaft advance. An advance of one tooth can cause the exhaust valve to open early, by approximately 22 degrees. The vector drive's ability to maintain constant velocity helps prevent speed variations that could mask these defects, particularly during the compression portion of the cycle.
By monitoring oil gallery flow and pressure, oil pump flow and pressure, and oil level control, cold-testing identifies such major defects as improperly sized or missing bearings and plugs. Oil pumps are not very efficient at low speeds. And at high speeds, the relief valve can mask defects. So, cold-testing usually uses an external supply system that provides oil at constant pressure and temperature.
As the crankshaft rotates slowly, oil flow is monitored to produce high fidelity signatures. For example, a lower than normal pump pressure that fluctuates in sequence with the engine cycle identifies a missing connecting-rod lower shell. Just 0.010 in. of additional bearing clearance in the connecting rod, will result in high oil gallery flow before top dead center.
Cold-testing identifies ignition defects using primary, ignition-coil induction, and direct injection methods. The primary method monitors the primary core voltage waveform. It will detect missing or shorted spark plug wires and give their location.
The ignition coil-induction method uses low impedance, non-contact sensors. This approach does not require electrical connections to the ignition coil yet it yields the same high fidelity signal as the primary method.
Direct injection fires the high-voltage coil system without engine rotation. Plug gap sensitivity measurements can be increased to ±0.005 in. in most applications.
There are limitations, however. For example, cold-testing provides no means for directly measuring engine horsepower or torque output, although power output can be ensured by tight acceptance limits. Also, the only heat produced is due to pumping losses and friction in the engine, so seals and gaskets do not reach hot running temperatures. Peak cylinder pressures are also low because there's no combustion in the cylinders. Fortunately, neither are important for testing assembly.
Dave Monteith is vice president and John Gagneur is senior product development engineer at Elcon Systems, Farmington, Mich. Greg DeBaker is automotive manager, Unico Inc.