At this point, most of the work is complete. If all the steady state values are correct, the vehicle should operate fairly well. The final phase of calibration is improving drivability. This should be thought of as finely polishing a sculpture. Where we were using a chainsaw earlier, we are now using sandpaper. There shouldn’t be any need for drastic changes anywhere. If large changes are needed at this point, chances are that the earlier modeling is not correct or the vehicle has hardware issues.
This Tech Tip is From the Full Book, ENGINE MANAGEMENT: ADVANCED TUNING. For a comprehensive guide on this entire subject you can visit this link:
SHARE THIS ARTICLE: Please feel free to share this article on Facebook, in Forums, or with any Clubs you participate in. You can copy and paste this link to share: http://www.musclecardiy.com/performance/engine-tuning-how-to-improve-drivability/
Integrating Fuel Maps
During the mapping phase, lambda was set to 1.0 for all but wide-open throttle. During WOT testing, an ideal lambda was found that makes the best power without knock. It now becomes time to blend the map for smooth transition between cruise and power. For the majority of the base fuel map, lambda remains around 1.0 for emissions and fuel economy. If the engine is to operate at λ = 0.85 at 90% load at WOT, some sort of blending needs to happen between there and the cruising region. Since light acceleration is usually acceptable at λ = 1, the enrichment should usually start above 50 to 60% load. Try even increments between 1.0 and 0.85 for the cells leading toward stable WOT operation. If engine or exhaust temperatures get too hot when the engine is held in this transition region, more enrichment can be added. If response is still crisp and temperatures are acceptable, slight enleanment in the transition area may help improve fuel economy. This map should look like a smooth function when viewed in 3D. Sharp breaks in commanded lambda usually indicate a cover-up for some other issue such as insufficient acceleration enrichment or improper ignition tuning.
For supercharged engines operating above 100% load, a reasonable estimate of target lambda should be used for the same engine without boost for the 70 to 100% load regions. This means that a supercharged engine operating at λ = 0.80 at WOT (130% load, 150 kPa) should still have a target of λ = 0.87 at 80% load (100 kPa). It is recommended that target lambda remain fairly consistent under boost, but there should still be a smooth ramping of fuel values between cruise and 100%. In some cases, it may be desirable to tune the base fuel map for extra enrichment above a standard boost level. This can be done by ramping in more enrichment relative to load (in the base commanded fuel map) above the calibrated WOT line to provide more cooling and knock resistance during over-boosted conditions.
Integrating Spark Maps
Much the same way the base target fuel map was blended, a similar operation must be done with the base spark map. The areas between cruise and WOT should be smoothly blended. This table should also look like a smooth function if viewed in 3D. Remember that supercharged engines usually tolerate the same spark advance as their naturally aspirated counterparts at approximately 80% load. If the engine is super-charged with a lower static compression ratio than the typical aspirated counterpart, more timing is necessary at lower loads to compensate for the reduced cylinder pressure. The idea is to run the engine close to MBT timing in these transitional regions in order to extract maximum efficiency without knock.
Drivability improvements are done largely by adjusting the transient controls to provide smooth changes between steady state conditions. The first transient to be refined is acceleration enrichment (AE). Adjustment of this parameter can be highly subjective. The object is to supply just enough extra enrichment to allow smooth transition to power without dumping excess fuel and hurting economy and emissions. When calibrating AE, it is helpful to think about how much of the wall film is being evaporated due to the increased airflow and add just enough fuel to keep τ constant. A quick stab at the throttle pedal momentarily shows a lambda leaner than the commanded value for high load. The acceleration enrichment multiplier should be increased until the wideband goes directly to the targeted steady state lambda at the higher load. The larger the camshaft, the more aggressive this function needs to be. Acceleration enrichment is usually adjustable relative to throttle position, so this should be checked with tip-ins starting at idle, part, and medium load. More AE is necessary at closed throttle because of the large amount of area increase when rotating the blade here. Fine-tuning of this function should be done under normal driving conditions with typical tip-in rates. If the vehicle stumbles, then picks up and goes, more AE is likely needed and the wideband should show lean during the stumbling. Adding fuel makes tip-in smoother, but too much can foul plugs, hurt economy, or increase emissions.
Even more accurate calibration of the AE can be done with high speed logging of wideband lambda and target ratio. While performing a tip-in with a known load change, but at a constant target lambda, AE requirements can be shown. This is best performed at part load where target λ = 1. A change from 15% load up to 50% load should not incur a change in the delivered lambda. By logging the actual lambda during this transition, AE requirements can be determined. If the wideband shows a momentary lean condition during load change, more acceleration enrichment is needed. Likewise, momentary rich conditions indicate excessive AE. This process assumes that all static airflow and fuel mapping has already be done with a high degree of accuracy. In OEM level calibrations, this process is repeated for a large array of speed, load, and temperature points to provide the best possible lambda control for emissions under all possible conditions.
While acceleration enrichment helps to ensure proper fueling during tip-in, it may sometimes be necessary to take additional measures to prevent knock or driveline noise. Burst knock is a phenomenon that can occur during sudden increases in engine load. The rapidly rising cylinder pressures may lead to knock under what would other-wise be stable combustion at steady state operation. Even without knock, the sudden onset of engine torque may lead to driveline noise as the lash is quickly taken up. Many OEMs intentionally reduce the available torque onset to keep noise low or prevent driveline component failure. Both of these functions are typically controlled by momentarily reducing spark advance. If instant throttle response is desired, the spark retardation can be set to zero. Added acceleration enrichment can be used to quench minor knock while retaining full timing and better torque. If no specific function exists in the PCM being used and tip-in retard is desired, the appropriate cells in the base spark map can be reduced. Since the engine is not likely to spend any time cruising above ~50% load at low speed, it is usually safe to reduce timing below MBT here to soften tip-in or reduce burst knock.
Dashpot adjustment can be checked at this point by simply lifting off the gas pedal from an elevated engine speed. The same should also be done by cruising at a steady speed and pushing in the clutch or shifting to neutral. The object is to find the balance between hanging at the elevated speed for too long after lifting versus dropping too quickly past idle speed and stalling. If the engine speed hangs, decrease the IAC position at the same engine speed or increase the decrement rate of the IAC position. If deceleration tends to drop right past idle speed (and the closed throttle position at idle is correct), the commanded IAC position at higher engine speeds should be increased or the decrement rate should be reduced. Initial adjustments to this function should be done in 10 to 20% increments to see enough difference in actual performance.
Some difficult engine combinations may not allow for a quick descent immediately to idle speed. In these cases, the IAC position versus engine speed knee point can be moved to a few hundred RPM above idle. This allows for engine speed to drop to a lower intermediate speed where stalling is less likely and the idle controls can softly move toward the desired target with a softer landing.
Deceleration enleanment (DE) can be adjusted next. Some PCMs have tables specifically for deceleration enleanment. These tables allow for the adjustment of the turn on/off points based on engine speed. These should be set well above idle to avoid stalling. The larger the camshaft over-lap and duration, the bigger the gap between idle and DE threshold should be. Many stock engines that idle around 650 rpm can tolerate DE all the way down to 1,000 rpm or so. A modified engine with high compression and overlap may drive better by only allowing DE above 3,000 rpm.
On PCMs without such tables, the same effect can be accomplished by adjusting the row or column of base fuel map cells for high engine speed and low load (high vacuum). Since the only way the engine could ever operate in these cells would be to have the throttle closed at high speeds, this fills the definition of deceleration. Setting target lambda to an extremely lean value commands lean operation in these conditions. A value of 0% VE also forces fuel shutoff and can be used in simpler speed density systems to achieve the same result. Since normal combustion is not present and any actual loads are very small, spark advance values for these cells can also be relatively high.
DE feel to the driver is tied to dashpot as well. Dashpot should be the primary control of engine torque and speed during deceleration with DE only activating in areas where very fast drops in engine speed are desired. If a gentle drop in engine speed is desired at any point, DE should not be used since it compromises the engine’s ability to produce consistent torque output.
Closing the Loop (or not?)
After airflow has been mapped correctly under most conditions, closed loop operation can be considered. All modern OEM systems use closed loop operation to compensate fuel delivery for weather conditions, part wear, and changes in tolerances. This operation is a very reliable method of keeping the engine in stoichiometric operation as long as the system inputs (primarily from the HEGOs) are reliable.
The timing of the HEGO signal is carefully adjusted in the factory PCM. The PCM makes adjustments to delivered fuel based upon the amount of “flight time” or transport delay between the exhaust valve opening and HEGO sensor measurement. When an engine is operating very close to stoichiometry and only small adjustments are being made, it is important to adjust in the correct direction at the proper time to avoid unstable correction conditions. If modifications to the vehicle have forced moving the HEGOs further away from the cylinder head, it may be necessary to adjust the transport delay functions in the PCM to compensate. Otherwise the fast adjustments being made by the PCM during closed loop operation may become out of phase with the HEGO feedback, forcing the learned values to diverge from ideal. Adjusting the transport delays can put these back in synchronization, improving fuel economy and emissions. Typically, a change from a short manifold to long-tube headers adds about 20 to 30% to the transport delay times due to sensor location.
Some mild performance upgrades may prevent proper closed loop operation at cold temperatures. OEM systems have temperature-based triggers that can be adjusted upward to reduce unwanted adjustments during warmup. In cases where the sensor is moved too far away from the head to remain hot even with its internal heater, the reduced accuracy may prohibit reliable closed loop operation at low speed. Additionally, large camshafts with significant overlap may experience too much charge wash at low speed for the HEGO to determine an accurate exhaust only signal. In some cases, this can be remedied by setting a closed loop operation threshold to an engine speed high enough to maintain good feedback from the HEGOs. If this is not possible, full-time open loop operation may be the only other choice.
If the airflow mapping has been done accurately, there should be very little closed loop correction needed anyway. This is the part of the process that reveals just how precise one was during the airflow mapping. Ideally, the airflow modeling was performed correctly for all speed and load conditions, and closed loop corrections are less than about 3%. Nothing should really change too much at this point. If closed loop operation suddenly causes a shift of 20% in fueling, there may still be more work to be done on the base tables. This is the time to stop and take a closer look at your previous work.
Choosing Cam/Runner Timing
For engines with variable valve timing or intake manifold geometry, measurements should be taken under most possible conditions. This usually means making a power sweep with the cams fixed to minimum and maximum advance or intake port flaps open and closed. By overlaying the resulting torque curves an intersection point can be found. This is usually the best point for switchover between the two conditions.
For infinitely variable systems, this process can become more complex. A good starting point is to perform sweeps for minimum, maximum, and mean positions and compare these to the OEM calibration if available. The OEMs typically hold variable cam timing adjustment to a later point to preserve emissions during low speed test cycles. More torque can often be found by comparing the OEM response curve to a locked sweep to find a better transition.
If the camshaft itself has changed in an infinitely variable system, be prepared for a longer dyno session to find the ideal crossover points. Start with an OEM curve and work from there.
Choosing a Shift Point
By examining the WOT power curves, an ideal transmission shift point can be determined. Many OEM automatic-equipped PCMs also control shift patterns. While the focus of this book is not transmission calibration, engine output plays a significant part in WOT shift patterns. The method of picking this shift point is similar for either automatic or manual transmissions.
If calibrated correctly, most engines exhibit a parabolic curve for both power and torque at WOT. There will be a high-speed point on the curve at which the engine makes peak power. Beyond this point, power drops as speed increases. How fast this drop occurs determines where the shift point should be.
If an engine with a redline of 6,600 rpm makes 300 hp at 6,000 rpm and drops to 260 hp by 6,300 rpm, there is no reason to shift much beyond 6,100 rpm. Shifting before actual redline allows the engine to enter the next higher gear near the meat of the torque curve. However, if another engine makes the same 300 hp at 6,000 rpm but continues to make 295 hp at 6,500 rpm, the shift point should be much later. If this second engine makes significantly less torque than the first, this is even more reason to delay the shift point as much as possible, perhaps close to redline.
Second to detonation, RPM is the largest strain on an engine. If peak power occurs at a lower engine speed, waiting to shift until redline only serves to reduce engine life and slow the vehicle down while making more noise. If engine power is no longer increasing, high RPM is usually not going to help.
Once all of this work has been completed, drive the car. This may sound obvious, but if most of the work was performed on a chassis dynamometer, the road may be different. Even a load bearing dyno is slightly different from the real world due to the added airflow across the car, radiator, and engine compartment. The unique driving style of one person may expose conditions unseen during testing. Try to drive both normally and as awkwardly as possible during the test drive to see as many real-world conditions as you can. Hold the vehicle in high gear and lug the motor with moderate acceleration to check for bucking, misfire, tip-in response, and knock. Perform one WOT pass (if it won’t get you arrested) to check for knock, bucking, or unusual performance. Coast to a stop in neutral to verify dashpot calibration. Stop the car for a minute and check restart with the engine hot. If possible, allow the car to cool completely and check cold starting.
Congratulations. You have successfully calibrated like a professional.
Written by Greg Banish and Posted with Permission of CarTechBooks