The graph is here folks All sizes | 3rd gear full load | Flickr - Photo Sharing!
I suggest you download it to view it because Flickr is resizing it in such a way that the fonts are squashed.
The green line is target boost, the bright red line is actual boost. Blue line is RPM, from below 1500 to about 4500.
Target Boost, in general control terms is also known as setpoint. Setpoint is not a measured variable, it is a constructed or derived variable, created by software as a result of applying a function, or functions, to a measured process variable ,PV, or several different measured process variables. Setpoint should be considered to be nothing more than a valid target value and is not necessarily the ideal value for any particular level of performance. The functions applied to the measured PV's should be selected to suit the application.
In the ECU, regardless of changes made to the DPF, no changes have been made to the setpoint, or the PV's or the functions applied to PV's, so control behaviour is unlikely to fail to reach the ideal behaviour, described by the design demands of the system. A lower backpressure should, in theory, enhance the response time of the control system to an instantaneous change (sometimes called an 'impulse')
Until T= 2.1 , target leads actual. (no overboost)
At T = 2.1, the rate of change of Target Boost changes and Actual Boost has to be brought in line to meet the new target
In order to meet the new target at least three things need to happen. The turbo turbine needs to slow or least stop accelerating. The pressure which was being built by the accelerating turbine has to dissipate. And the Boost control mechanism needs to move.
This is why there is a lag between the time when Target Boost changes and Actual Boost comes back to meet Target Boost. Turbo lag applies not only on an accelerating turbine but also a decelerating one, as a function of the mass of the turbine, and not to mention the speed of response of relevant turbo control valves.
At T = 4.1 peak overshoot has occured and the residual pressure has begun to recede (there is a delay from when the turbo slows down to when the pressure it was producing has dissipated due to their being no way to bleed off pressure instantaneously)
At about T = 5 the falling variable called Actual Boost undershoots Target Boost (target boost is rising) and the system responds by starting to increase boost again, to align with the still rising Target Boost value
By T = 6.2 the effect of the response by the control system to the undershoot by Boost Actual takes effect and by T = 8.3 Boost Actual has again converged with boost target.
In general control theory, a system which responds to a step change within one overshoot period can by most standards be said to be 'in process'
Fine tuning of the control loop can be performed, if required, but in this particular instance there doesn't appear to be any need for retuning because the response of Actual Boost to changes by Target Boost appears to be in keeping with the constraints being applied by the mechanical limitations of the system, such as turbine momentum and pressure trapping, and given time, Actual Boost conforms readily to Target Boost
The reason for the spikes in the graph at various points is because the scan rate is fixed in FES free at 1HZ. Using a resolution of about 10mS (sampling at 100Hz) would show a very smooth response, with the same peaks and troughs, but with a nice smooth blending curve between.
I see no reason to remap this ECU because it is responding very well to step changes in demand well enough. The overshoot and undershoot is small, considering the actual response time of the turbine's speed and the rate at which excess pressure can be bled off. Not to mention that at peak boost for a sustained period which is achieved in a test like this, the turbo is reaching it's maximum speed. This test is unlikely to occur in normal driving very often. (but it all depends how you like to drive !)
All in all, nothing unusual. Boost control is coping perfectly well in response to this very high load test. (3rd gear, uphill, from between about 1500rpm to about 4500rpm) The brown line is the throttle pedal position, showing that I floored it.
The blue line is the speed. The rate of change of speed is acceleration and acceleration is only produced by torque. The near constant acceleration should suggest how willing the engine is to deliver torque....the response by the control system as the revs rise beyond about 3800 is to reduce boost, in accordance with the change in demand set by Boost Target. in so doing, the operation of the control system is confirmed, and the response by the controlled variable Actual Boost is fluid and shows a willingness to conform to demands by Target Boost). Following that, the Actual Boost is then falling at the correct rate as the revs continue to rise up to their peak value. As I said before, this is a well controlled system
A less severe test, such as a 50% actuation of the throttle pedal. would more than likeley show that the value of Boost Actual tracks Boost Target yet more closely, when changes to Boost Target take place.
B.Eng (Hons) - Mechatronics (control and automation) , MIET