In a previous Enginology column, we discussed some issues relative to induction and exhaust system design and analysis. In another, the subject of in-cylinder pressure study was also aired. Since then, a few questions have been posed we'd like to address this month, for the purpose of providing some clarification.
On the subject of intake and exhaust system flow, we suggested a relatively simple "rule of thumb" for evaluating (or selecting) passage sizes that could be followed by assuming a value for mean flow velocity. Academically speaking, this approach assumes we are addressing a one-dimensional system and doesn't contemplate such factors as viscous friction, turbulence, or other dynamic disturbances that occur as a function of passage curvature, sudden changes in cross-section areas and related elements. Even though many of the computer-driven simulations that aid the analysis and development of intake and exhaust systems utilize the one-dimensional concept, it's accurate to say there's a practical limitation to the finesse these provide.
Over time, as you read these monthly columns, you'll discover that the treatment we've elected to give topics that have the potential for being infused with academic weight are intentionally delivered in ways to help reduce theory to practice. However, since we've already broached the subject and a couple of more technically-biased questions arose, we'll first dig a bit further into the tuning issue.
For example, flow in inlet and exhaust passages are actually unsteady, interrupted and compressible. Pressure waves, or perturbations, are created that traverse the medium containing them. These move both linearly and non-linearly, based on their respective amplitudes,and add to the complexity in a computational analysis of both intake and exhaust systems.
While we previously suggested intake and exhaust systems could be tuned separately, it's generally believed that an engine's fundamental torque characteristics (essentially the volumetric efficiency curve) is largely governed by its induction system design. And even though similar principles of analysis can be compared to those for an intake system, particularly of a practical nature as previously shared, it's basically during the valve overlap period that an exhaust system might experience initial pressure conditions (atmospheric) as does the induction system. As a result, a comprehensive analysis (or design) of an exhaust system can't be addressed exactly as that for an intake.
However, an assumed mean flow velocity value that's common to both an intake and exhaust system has proven to be a worthwhile approach when simply "ball-parking" a way to examine, compare, or predict the influence of either type of system on an engine's volumetric efficiency or torque curves. Further, believe it or not, it's possible to make a rough determination if an engine has been "over-" or "under-cammed" by comparing actual peak torque rpm to calculated peak torque rpm (using the previously-discussed, one-dimensional approach) and determining if the calculated value (rpm) is above or below the one observed. A condition of "over-camming" is reflected by an observed peak torque rpm that's higher than the one calculated and "under-camming" is indicated by the opposite.
Now, moving over to the inquiries about in-cylinder pressure analysis (engine cycle analysis or ECA), it appears we should have included some thoughts on the effects of combustion cycle-to-cycle variations. In textbook language, this amounts to "cyclic dispersion" over a range of engine speed, time, or both, and relates to variations in peak cylinder pressure as measured from one firing cycle to another, normally sequentially. Why is this important and how can it be addressed?
First, variations in peak cylinder pressure have a direct bearing on net crankshaft torque. Causes can range from variations (cycle to cycle) in air/fuel charge mixture homogeneity, transient changes in localized temperatures within the combustion space, inefficiencies in the removal of combustion residue (non-combustible exhaust gases), engine speed and load. Absent any electronic controls of the type that monitor detonation and make corresponding changes to ignition spark timing (we're dealing with race engines here, not emissions controlled), variations in cycle-to-cycle peak pressure tend to reduce net power.
So, are there any practical issues that can be addressed to minimize this condition? Well, a few might be worth considering. Research has concluded that cycle-to-cycle peak pressure variations are materially influenced by decreases in combustion rate. Fundamentally, we know that "rich" air/fuel mixtures combust more slowly than "lean" mixtures. So as an engine is made to run on the lean side of rich, so to speak, cyclic pressure variations tend to increase. However, if we're concurrently able to improve the atomization of fuel (not only particle size and uniformity but mixture homogeneity) it's possible to reach a level of compromise where combustion of a leaner mixture is made faster and we stand to eliminate some of the cyclic variations.
Following along the same line of thinking, and we've touched on this in prior columns, improvements to mixture homogeneity (atomization and mixing efficiency) not only encourage a faster burn but provide an opportunity for less initial spark timing . . . the results from which, you may recall, are higher IMEP and corresponding net torque.
Once again, putting on our practicality hat, there's another element to consider that can affect both combustion efficiency and cycle-to-cycle pressure conditions: exhaust residue. This inert material, although often containing portions of unburned fuel, does a couple of unwanted things in a racing engine. One is the dilution of fresh air/fuel charges whereby space otherwise available for combustible materials is occupied by unburnable exhaust gas. Technically speaking, this not only reduces the potential for optimizing peak cylinder pressure, but it also reduces flame temperature (useable heat), slows down flame speed, and presents a case for increasing spark timing, none of which match the combustion requirements of a race engine. EGR may be fine for emissions reduction, but that's not currently the goal for your race piece. However, don't discard the idea for some issues we'll discuss later this year.
What can be done about minimizing exhaust contamination of fresh mixtures? An efficiently designed exhaust system in the range of most operated rpm is a place to start. For example, if your engine is most often run in a 4,500-7,500 rpm span, sizing a system to be optimally efficient in this range will help improve the blow-down portion of the exhaust cycle.
Depending upon the point of exhaust valve opening, peak blow-down pressure will occur at different engine speeds and crankshaft angles, but still rely on the sizing of an exhaust system that matches the desired rpm range, particularly the peak torque rpm point. Making certain that a given system is most efficient in this span of engine speed will help minimize combustion residue present at the arrival of each fresh air/fuel mixture charge. Of the conditions this benefits, not slowing down the combustion rate from air/fuel charge contamination trails back to the need for minimizing cycle-to-cycle peak pressure variations. And we've come full circle.
What we'd like to leave you with this month can be boiled down to a couple of issues. The building of a successful race engine is a function of understanding as many of the factors that influence its performance as possible. Two extremes of information populate this requirement: one is purely academic and woven with theoretical and scientific perspectives, and the other is based on pure experience in the building and evaluation of race engine packages. Smokey often told me, you need some of both.
One of the purposes of this column is to attempt a blending of those extremes. To draw benefit from the information Enginology is designed to contain, if I'm able to do this job effectively, requires more than just casual reading. Unless your responses to CT's Editor, Rob Fisher, are to the contrary, we'll begin placing more emphasis on these two informational components: theory and practical applications. By design, this approach will hopefully help you understand the "why" underlying the "what" as you continue the search for more understanding about engine basics and how they can be transposed into winning race engines.