Two highly technical presentations covering exhaust theory and the testing and evaluation of cylinder pressures opened day two of the Advanced Engineering Technology Conference in Indianapolis. Vincent Roman of Burns Stainless spoke on “Development Tools and Testing Methods for the Optimization of Exhaust Performance,” and NASCAR consultant Cecil Stevens covered “How to Use Cylinder Pressure and Valve Motion Measurements to Improve Performance.”

Vincent Roman of Burns Stainless.

Roman started with definitions of critical factors involving exhaust, especially incompressible flow and compressible flow. These are basically distinguishable with changes in density of the exhaust gas, and any formula or header design must take both dynamics into consideration.

“By harnessing the wave energy that’s in the exhaust and intake, we are essentially supercharging the engine,” he added while explaining volumetric efficiency.

The key in header design and tuning is understanding and working with pressure wave dynamics, which are influenced by a number of factors — the least of which is cam timing. Unfortunately, wave tuning in an internal combustion engine does not follow the same classical theories that are found in organ pipes or other similar acoustical instruments.

“What happens in an organ pipe is that we have infinitesimal pressure wave, waves that have almost zero pressure,” explains Roman. “That’s not what’s occurring in the exhaust pipe. It’s probably what’s occurring in the intake. What’s good is that the math is easy and historically it’s what’s been used to design a header.”

More advanced math formulas now consider finite pressure waves that have significant pressure, such as those encountered in an exhaust manifold. Very sophisticated computer modeling software takes these factors into account, but the research and development costs are much higher.

Burns Stainless has developed its X-Design method using a combined approach of classical wave tuning and “adjusting it for what we know works.” The experience has also led to header collector designs that attenuates reflected waves and helps provide a broader power band. The key is analyzing the right information from the customer.

“The most important numbers we need are rpm, because that determines the length of the header. But also, what really is the rpm band of interest?” stresses Roman. “In some cases, such as drag racing, it’s fairly easy. But in other applications like dirt track, a lot of us get blindsided into designing for peak power when the driver is asking for 4,000 to 6,000 rpm power. So, you really have to take care in determining that rpm band.”

Other critical information includes bore, stroke, compression ratio, cam timing, valve sizes and exhaust-port size at the manifold.

“Our program calculates the size of the collector,” adds Roman. “The size of the exhaust port actually dictates whether or not we can run a stepped header. On a modern race engine we’re almost always three steps from the port to the collector.”

During the Q&A period, some interesting notes surfaced, including the tip that header builders should try to get a rotational firing order in the collector. And the future of header design? Think of advanced 3D metal printing.

“Now we’re not limited to round tubing,” raves Roman. “We can actually shape the collector any way we want.”

Determining cylinder pressure

In the morning’s second session, Stevens laid out the preparations and objectives for setting up a dyno test to gather cylinder pressure data. Key requirements include a dyno and operator that can hold a steady rpm, reliable pressure transducers, crankshaft encoder, analysis software, data acquisition and a laptop computer.

Cecil Stevens, Director Combustion Development at Illusions Engine Development

Pressure transducers come in a variety of styles and costs, and then there’s the decision to mount them in the cylinder head or block — which means dedicating that equipment to the dyno — or mounting the transducer to a spark plug. There are also choices in the other required equipment — again, the decisions often come down to cost and research intent. Once the dyno and data gathering essentials are in place, the engine builder will be rewarded with almost 16 million pressure values per minute at 5,500 rpm in an 8-cylinder engine.

Of course, that’s a considerable amount of information that needs to be analyzed with a host of engine design factors and the current tune, down to the valve lash.

Comparing cylinders is a key element in determining where power gains can be made, and the data can be displayed in a variety of ways to help make the comparison relevant to the research. In one example, Stevens showed the results of a premium series engine.

“And we can see that not a single cylinder was happy,” he notes. “And with one chart you can help yourself out a great deal.”

The data can compare cylinder vs. cylinder or cycle vs. cycle, but the key to all of data gathering is organizing it and keeping it available for review.

“Let a high-school gearhead help you,” suggests Stevens.

Much of the analysis returned to a common them of valvetrain dynamics. Stevens suggests to always take measurements of critical components, including valve springs and the cam profile. He especially recommends keeping track of valve spring numbers from out of the box until they’re thrown away. Relevant dyno information should also accompany each test, and all the data compiled on an easy to access Excel spreadsheet.

Stevens focused on a number of valve bounce examples and also dispelled the myth of “valve lofting.

“It’s detrimental to valve springs, rocker arms and lifter-to-lobe interface,” he sums up. “All of the work the exhaust people do. All of the work the cam people do is shortfall if we don’t know that the final outcome of all that work is under control. The valvetrain must reproduce the design.”