I met Professor Blair nearly ten years ago while I was working for Hendrick Motorsports. Professor Blair had paid us a visit to demonstrate his 4stHEAD software for Valvetrain Development. He brought along Mel Cahoon to assist in the presentation of the software. He made his presentation to me along with a few members of the engine shop management which took about 45 minutes. Immediately after the presentation, I asked the Professor if he wouldn’t mind answering a few questions I had about Engine Simulation. He was quite agreeable and we convened to a training room where Andy Randolph and Mel watched as I interrogated the Professor for several hours! Actually, it was a great discussion where we all exchanged thoughts, ideas and our experience regarding the best way to characterize a physical engine in a virtual environment. For months preceding his visit to Hendrick, I had been using the Engine Simulation software from Optimum-Power, Virtual Engines. I had taken it upon myself to bring the software home and use it to determine whether or not it could be used to aid in Engine Development. I had also read his book, Design and Simulation of Four Stroke Engines, and I had many questions regarding both the software (based on his work) and his book. Professor Blair shared his experience with me and answered every question in great detail. That was the beginning of a relationship that we enjoyed for many years.

The next time that the Professor came to visit, I was eager to expand on our previous discussion. While I wanted to learn more, I also wanted the Professor to respect me. I was very calculated about the questions that I asked fearing that he would take something I said, in my lack of experience in engine modeling, and rip me to shreds! I did not have decades of experience in engine theory and I felt at a disadvantage. At first, the Professor was very academic, impersonal and direct in his approach. He treated me as though I was a student and he was the Professor. Over time, I expressed my respect for him and his experience and suddenly the walls all came down. He became a different person. Suddenly, he was humble and candid and even made jokes about himself. From that point on, Professor Blair was an absolute pleasure to work with and we both became more comfortable. We became friends. Over the years, we met nearly every time he came to the United States. I also got to know Charlie McCartan, the Professor’s right-hand man – the man responsible for a large portion of the valvetrain work that supported the Professor.

The Professor I came to know was in stark contrast to what I had been told. I don’t think most people ever really knew the real Gordon P. Blair. Certainly, the true Gordon P. Blair was not revealed in his publications. In text, he was very opinionated and steadfast in his beliefs and his stance was unwavering as he spelled out his theories and experience through research publications. In person, however, he was open-minded and respectful of others’ ideas and opinions. He certainly treated me with much respect and I am very grateful for that.

He was quite a character at times. I used to get a kick out of requesting changes to the software. If I would ask for something (anything, actually) he would say “Do you know how much work that would be? Oh my, it would take this and it would take that and all the code and the math and the this and the that…” and then he’d go do it (or Charlie did it). I quickly learned that I had to make requests by expressing how absolutely difficult it would be and that it would be practically impossible to implement these simple changes… and he’d do it (or Charlie again). Either way, he’d always call me up and say “Master Kurn! You must go to the website and download the latest version of the software as I am sure you will find a welcome improvement. Mind you, it was no easy task. Charlie and I have worked tirelessly for hours on end to make these changes” – which probably meant that Charlie worked tirelessly for hours on end!

I consider myself privileged to have known the real Professor Blair. He was a pioneer in fundamental engine research and accomplished a great deal in an area that is, at best, not very well understood. He had a way of solving difficult problems with simplicity while capturing and characterizing complicated phenomena occurring throughout the strokes of an engine cycle – both 2 and 4-stroke engines. He was also a true patriot, a gentleman and a good friend. He was a truly great person who left his mark on the Motorsports world.

He will be sorely missed.

Brian Kurn

Virtual Prototyping is defined as integrating a geometric model and related engineering tools such as analysis, simulation, optimization, and decision making tools, etc., within a computer-generated environment that facilitates multidisciplinary collaborative product development. In other words, its the use of computer simulation tools to test and analyze the performance of real systems. Virtual Prototyping will never, and should never replace physical testing but it can reduce the number of physical prototypes used throughout the product development process. It can also reveal all modes of failure where physical testing can usually only reveal one – the one that failed the prototype!

As we begin to see the light at the end of the tunnel in this economy, returning to profitability is of paramount importance to the decision makers in any company. In any business, a good way to increase profits is to reduce costs. The strength of Virtual Prototyping is the ability to:

Reduce product cost without a reduction in quality
Get products to market faster
Launch products on time
Meet quality targets at design release

.

The Cost of Testing

Think about how much it costs to test a new camshaft design. For some NASCAR teams it can add up to nearly $50,000 to test an engine on the dyno. That cost is only for testing – the test engine will never see the track! Why does it cost so much? In order to test an engine it must be built just like a race engine. It requires new pistons, rings, bearings, head gaskets, pushrods, valve springs, etc., etc… If that wasn’t enough, there’s the cost of the labor for someone to hone the block, CMM (Coordinate Measurement Machine) the pistons, balance the assembly if needed and assemble the engine. Add to that total the cost of dyno operators, racing fuel, oil and let’s not forget the teardown guys after the engine is all used up. I’ve left out many people/departments but you get the idea. What if the cam does not act the way it was designed? What if it doesn’t perform, or worse yet, it runs better but breaks retainers during durability testing? Now you’ve lost even more money! Cam testing for race teams can be very difficult and it is usually expensive.

Granted, usually test engines are not used for one single test. Normally several tests can be conducted on a fresh test engine. In addition to cam testing, other tests can be conducted such as cylinder head and intake manifold tests. Development Engineers for these race teams strive to get the most out of each test while the engine still has life in it and the results are believable, meaning the engine hasn’t lost power and still repeats well. Typically, in order to test one cam, three tests must be conducted:

1. Baseline cam test
2. Candidate cam test
3. Baseline repeat

If the baseline does not repeat then the results from the candidate cam are in question. If the baseline does repeat well then you can believe that the test was valid. Typically, around five cams are tested that are slight variations of the baseline. Out of those five, maybe one or two have enough merit to justify more testing or endurance testing in preparation for racing the design. Sometimes all five fail to deliver an improved power curve. This is a total loss in cam testing. Virtual Prototyping can help make sure this does not happen.

.

The Cost-Effective Solution

Virtual Prototyping can increase the chances of success and reduce the cost of testing. This procedure is also very efficient and can produce better results in less time than traditional methods. Although Virtual Prototypes can come in many forms, let’s look at high-performance, aftermarket cam design using Virtual Prototyping.

Explore the Design Space

Using Engine Simulation, the duration and lobe phasing can be determined in order to achieve the desired power goals. Using the existing cam as a baseline, lift and duration multipliers are used within the Engine model to simulate the increase/decrease of duration and lift of the cam. The intake and exhaust lobe phasing is also varied and the results of all the variables are recorded and analyzed in order to determine the cam specs that will deliver the desired power curve. Literally thousands of lift/duration/lobe phasing combinations can be tested in a few hours. Subsequent engine models will be created and used to design cams for different applications of the same engine. Such applications typically include:

Forced induction (Supercharged, Turbocharged)
Naturally aspirated
Stock engine – installing new cam in unmodified engine
Modified engine – installing new cam in a modified engine
Nitrous Oxide

.

Design, Run and Analyze the Virtual Prototype

Once the cam specs have been defined, a Valvetrain Simulation model is created and lobes are designed and tested until they meet the design criteria of the Engine Simulation model and do not exhibit any undesirable dynamic behavior. This iterative process continues until all goals of stability, durability and performance are obtained. The final prototype lift curves are then rerun in Engine Simulation to verify that they indeed perform as they were designed.

Manufacture and Test

After the design has been verified, the lobe data is sent to the cam grinder and a dyno test is scheduled. The likelihood of a positive outcome is ensured, especially in a street car application where performance is easy to predict with a high level of accuracy. Undesirable dynamic behavior has been minimized by designing the components to operate as a well designed system instead of a collection of components. Valvetrain behavior is typically dominated by valve spring frequencies and its harmonics as a function of speed. If any component of the valvetrain system operates at a frequency near that of the valve spring then the valve spring can become excited when the two frequencies align. When that occurs bad things happen. Using simulation tools enables us to design around the natural frequency of the spring and select components that do not excite the spring. Additionally, we can choose a valve spring whose negative frequency characteristics lie outside the operating speed-range of interest.

Start Production

When properly done, Virtual Prototyping will lead to more productive testing. The big cost savings are in the fact that testing of candidate cams is reduced or eliminated except for design validation and durability testing. Components can be designed to use less material without sacrificing their structural integrity. In many cases, using the traditional method, cams are not even tested on a valvetrain test rig or dynamometer – they are sent to the track or the street based solely on the experience from previous designs. Although this is common practice among the big cam grinding companies, it is not a very efficient method when compared to Virtual Prototyping. Failure in any form of racing or street application can be very expensive! It is well proven that Valvetrain Simulation can accurately predict dynamic valve and spring behavior with very high accuracy in race applications. In comparison, high-performance street applications are far less volatile which also make them even more predictable. Using Virtual Prototypes to design the valvetrain components can allow manufacturers to reduce risk, shorten the design process (time and cost) and increase performance without sacrificing durability. Win, win, win, win.

A typical Engine Simulation project consists of the creation of a baseline model, correlating that baseline to measured data and then making small changes to the model and on the dyno to ensure the variable modifications are being modeled properly. Simple modifications such as header primary lengths and cam timing are good examples of changes that should be easy to predict for a well-built baseline. Once a solid baseline has been achieved, then the variable  exploration and optimization process begins. Goals are specified and a strategy is defined based on a range of variable values that are within manufacturability. Many times the variables believed to be important turn out to have only a small effect on the outcome.

Contact us to discuss your project.

The June 2009 issue of PRI magazine features an article on cutting-edge CAD/CAM Software that companies are using to aid them in their businesses. Four Stroke Design, LLC was one of the companies featured in the article.

Excerpts of the article include: “Brian Kurn of Four Stroke Design, in Charlotte, North Carolina, specializes in ground-up race engine design. “The only thing that changes is the size of the models, which is directly related to two things; the time it takes to solve and the accuracy of the result,” said Kurn. “As computer technology advances and prices get lower for faster hardware, the accuracy of simulation improves and the time to solve is reduced. This makes the process more cost effective.”

Perhaps the biggest benefit is the ability to eliminate trial-and-error proto-typing. “The days of physical prototyping are numbered,” said Four Stroke Design’s Kurn, who utilizes simulation technology including CAD (SolidWorks), Engine Simulation (Optimum-Power’s Automated Design), Valvetrain Simulation (Blair’s 4stHEAD), Dieter Zuck’s CDS, RecurDyn from FunctionSim, and CFD (Fluent/Ansys and Star-CCM+ from CD-Adapco). “Before Valvetrain Simulation, for example, everything had to be manufactured and physically tested in order to find out whether or not your design was effective. Using Valvetrain Simulation, several designs can be evaluated in a short period of time.”

Simulations can also shortcut the testing procedure, significantly reducing or even eliminating testing costs. “NASCAR engines use pushrod-operated valvetrain systems,” Kurn continued. “The dynamic valve lift curves are radically different when compared to the static valve lift curve of a given pushrod valvetrain system. Being able to predict dynamic valve and spring behavior is very valuable in a market as highly competitive as NASCAR Sprint Cup racing. With Valvetrain Simulation, you can explore the possibilities with a high degree of confidence and identify possible design flaws.”