Anyone who has ever read about a cylinder head knows that the most common way heads are compared to each other is by port flow. Most of you reading this are probably at least slightly familiar with the fact that flow numbers come from putting a cylinder head on a flow bench, and that, of course, higher flow means more power.
While a flow bench can be a powerful tool for developing a precise port design, or by judging the effectiveness of someone else’s port design, it’s not a tool the average enthusiast has much contact with. In this video, Andy Wood of Unity Motorsports Garage pulls back the curtain and offers a peek at what goes into flow testing a set of cylinder heads.
One thing most of us are wary about, is people making their numbers look good, in order to make their products look better on paper. Much like a dyno, a flowbench is only as accurate as the data you enter into it. And sure, if you want to cheat the system, there are ways, but generally, you want the flow bench to give you real, precise data. To get that data, being methodical and diligent in your setup is the key to accurate data.
In the video above (and the subsequent installment below), Wood walks viewers through the steps he takes when performing a flow test. He starts by mentioning that your fixture’s bore diameter needs to match your engine’s bore, in order to get realistic data. If you use an artificially large test fixture, you can get artificially large flow numbers, due to unrealistic unshrouding of the valves.
One thing to note here is that Wood isn’t just running a straight CFM flow test. He’s also taking several measurements to feed into David Vizard’s Induction Optimization Program, which will evaluate port design efficiency. The results from the program will guide Wood on how and where to improve the port’s shape.
The first step is to measure the volume of your intake and exhaust ports. Much like checking the volume of your combustion chamber, this is accomplished with a graduated burette and a sealing plate. Only, you’re filling and measuring your intake and exhaust ports. A note for those who’ve never CC’d an intake port before, unless you have a monster burette, you’ll have to refill your standard 100cc burette at least once.
With the ports CC’d, Wood then measures the port dimensions at the port’s opening. Next, is measuring the length of the runner. Not a simple process, it isn’t overly complicated either, as Wood demonstrates in the video. By measuring the short-side and long-side of the runner, adding the figures, and then dividing the product in half, you get what is referred to as the “port length.” With all the measurements made, it’s time to put the head on the flow bench.
Once bolted in place, a small radius of clay is built up around the intake port being tested. Without that radius in place, the sharp edge of the port would be exposed to the incoming air and would create turbulence, falsely skewing the flow data. Next, you set up the device to open the valve a specific amount, as ports are usually flow tested at different amounts of valve lift to create a more complete picture of how the port performs.
With all the setup completed, the flowbench is fired up and a vacuum of 28 inches of water is applied to the underside of the port. With Superflow’s FlowCom measurement computer, the traditional liquid filled manometers you can see in the background become obsolete, and CFM numbers are generated in real-time.
While most of the time, you’ll hear about a cylinder head’s peak flow numbers. Sometimes, you’ll see a chart with several different data points, usually starting at .100 inch of valve lift, and go up to .600. But as you can see in the video, Woods starts at .025 inch of valve lift, and slowly works his way up, collecting data at each valve-opening increment.
If all you wanted was flow data, you’d be done at this point. However Wood then plugs all the flow data, along with the measurements he took earlier, into the IOP program and starts to dive into theories about where in the lift range flow is most important — which is another article for another day.