1/26/19 Update
I had the good fortune of getting a test mule cylinder head from Fast650.  His generous gift will allow us to learn a lot more about how the budget flow bench works and also to do some preliminary work on the ports prior to removing metal from a good head.  I call the test mule “The Hammer Head”.  As you can see, the name is appropriate.

I wanted to test the Hammer Head and compare the results to my good used head which I will refer to as my “Performance Head”.  If the baseline results are close, then the mule will be a valid tool to show how various modifications work.  Here’s how the mule looked set up on the flow bench for intake port testing.  Note that the exhaust port is sealed off to eliminate any concern over leaking valves.

For exhaust tests, I fabricated a fitting that matches the geometry of the stock header pipe and gasket.

I decided that I would use 10” H2O as my test pressure for all tests.  The rig isn’t robust enough to pull big numbers so 10” seemed like a happy medium.  The plan is to pull the highest pressure I can achieve and then convert to 10” using the DTec conversion tables.  If I can’t pull at least 10”, I will pull what I can achieve and then convert to 10”.  That way, all final results will be expressed at 10” H2O.  

After hours of testing at varying test pressures, I confirmed that I get essentially the same results when I convert to 10”.  For instance, when I test an item at 15” and get 84 CFM, and then test the same item at 6” and get 53 CFM, and I use the conversion tables to convert both results to 10” I end up with 68 CFM @ 10” H2O.  Science, it works every time.

My intent is to check flow at .025” increments to allow matching head modifications to particular camshafts, the idea being that cams that open the valves quickly but have relatively low lift will benefit more from mods that improve flow at the lower lifts.  In addition, the currently available cams lift the valves in the range of .018” to .083” during the overlap period, so having test data in small increments at the low lift positions seems beneficial.

I started with the intake port.  Opening both valves, it was apparent that I would not be able to maintain at least 10” H2O over a complete valve cycle.  I decided to use declining test pressures from low lift to high lift, similar to a floating pressure drop test.  However, I didn’t just allow the test pressure to float, I lowered the pressure using the bleed valves to allow testing at a predetermined pressure for various lift positions.  That way, I would always be testing at the same pressures.  For instance, at low lifts I used 15” H2O test pressure, and will continue to use 15” at low lifts to evaluate mods.  As the lifts increased, I had to use lower test pressures, and will continue to use those exact test pressures at each valve position to evaluate mods.

Opening a single valve, I was able to maintain 15” H2O through a complete valve cycle from .025” through .400” lift.  I decided to test the individual port runners using a fixed test pressure of 15” H2O.

On the intake port, the flow bench performed very well.  It was consistent throughout the testing.  I could walk away from the project for several hours or even overnight and get exactly the same results when I did cross checks later on.

Here's the intake data on the good head.  Pretty close.

Here is an illustration of how the suck through setup is configured.

Here's the graph on the good head.  Pretty close again.