This is a very long post with lots of pictures.  Please hold your comments and questions until I have made the final entry, which will be obvious.

In preparation for future performance upgrades, I decided that I was going to need a better tool to measure improvements in air flow.   I want to improve the cylinder head flow.  I could always just hack away at the ports, valves & guides, but it seems more prudent to try and see if I can monitor the changes to verify that they are in-fact improving flow.  It’s very hard to put the material back on once you grind it off.

Armen was gracious enough to send me a sectioned LS650 cylinder head to play around with.  It had been subjected to one of those disconnected cam chains that we keep reading about on the forum.  As such, it was toast, so Armen seized the opportunity and cut the thing in half, right smack down the center of the ports.  It’s a thing of beauty.  You can now really get your eyeballs around the problem areas.

Fast 650 held my hand through the basics of port flow and served as a mentor.  He is a fabulous resource and provided a wealth of hard-earned knowledge on the ins & outs of cylinder head mods.  He fixed me up with a great article from an outfit named “DTec” that provides all the details for fabrication of an inexpensive flow bench.

Up to this point, I was trying to measure my flow improvements by simply attaching a vacuum cleaner with a u-tube water manometer in-line.  It worked OK and I could pretty much see big changes in pressure which indicate a change in flow, but I had no way to keep the test pressure constant, and the resolution on the manometer was poor.  Also, changes in atmospheric conditions and variations in line voltage and frequency seemed to make repeatability difficult.

After reading the DTec article on the DIY flow bench three or four times, I felt that I could fabricate a very simple tool that would allow measurement of flow with good resolution so that very small changes (good or bad) would be easy to detect.  The tool would also allow me to test each change at exactly the same pressure, so there would be no doubt about other things affecting the results (things like atmospheric pressure, humidity, temperature, electrical voltage & frequency, vacuum cleaner performance, etc.).  If you always adjust to the same test pressure, the playing field will be level, and the results should be consistent and accurate.

I used the info in the DTec article to fabricate a smaller and cheaper version of their flow bench.  I wanted something that would be very simple, compact for easy storage, provide good resolution to make interpretation fool proof, cheap (of course), and not necessarily accurate.  I’m not worried about accuracy in terms of cubic feet per minute (CFM).  All I want is a tool that shows me if I made the flow better, or if I made the flow worse.  I want the tool to give consistent, repeatable results.  The budget flow bench meets those requirements.

The sketch is pretty crude but I think it will do.  The heart of this system is the orifice with an inclined manometer that measure the difference in pressure across the upstream and downstream sides of the orifice.  I used the DTec article to provide the various sizes for the orifice and also the details for the scale on the inclined manometer (very important, won’t work without the scale).
 
The orifice and inclined manometer are a team, and they really don’t care what the test pressure is.  They only sense flow.  The inclined manometer scale is graduated in a manner that provides an accurate reading of the percentage of flow associated with a given differential.  This relationship between percentage of flow and differential pressure remains the same, regardless of orifice size or test pressure.

I will provide the details about the rig and its construction later in this post.  For now, let’s get to the meat of the tests.
 
I wanted to test valves and guides to see what low hanging fruit might be available.  These parts are easily removed and easily replaced.  If I perform a mod on a valve or a guide, and I don’t like the results, I can simply throw the part away and install a new one.  If I grind on the head and I don’t like the results, it gets expensive real fast.  The guides and valves provided good test specimens to play with.  They also allowed me to practice and learn more about how to set-up and operate the rig.

The stock exhaust valves are typical.  They are 1.1” diameter and have 7mm stems.  I had two old junk Honda valves laying around and they just happened to fit Armen’s head.   So, I left one valve with stock geometry and I did a 30 degree back-cut on the other.

Here is a picture of the valves.

The exhaust gas flows directly at that guide.  We might be able to improve things by simply tapering or rounding the guide.

Her is another view looking through the port outlet. See that nasty blunt guide sticking out of the guide boss?  We should be able to improve that.

I tested ten combinations, five with the stock valve, and five with the back-cut valve.  The inclined manometer reads out in percent.  So, if you are using a 100 CFM orifice, and the manometer reads 81, then the flow across the orifice is 81 CFM (.81 x 100 = 81).  Keep in mind that the CFM numbers are not accurate.  This contraption is not calibrated.  However, I believe that the change in CFM from one combination to the next is pretty accurate.  Anyway you look at it, the contraption certainly shows that improvements are lurking in the guides and valves.  

Here are the results of the first five tests using a valve with stock geometry.

Once again, the flush guide performed best.  If you calculate the percent of flow to actual CFM, you find that the flush guide with back-cut valve flows about 2 CFM better than stock at low lift and about 1.3 CFM better at max lift.  I’m only testing one exhaust valve so if both valves are considered the improvement works out to about 3 to 4 CFM.  Not earth shattering but it’s a cumulative process.  The final package is the result of many small improvements.  As they say, every little bit counts.

I ran some special tests just to fool around.  I initially borrowed Armen’s head to see if I could fabricate a removable insert to bridge the gap between the 1.3” opening in the exhaust port and the 1.27” inner pipe in the exhaust header.  The intent was to eliminate the immediate & drastic increase/decrease in the cross section between the port and inner header pipe.  So, I did one test with a plastic insert that tried to simulate the insert I had in mind.  It didn’t work too good but it’s hard to do a proper test with only half-a-port.  I also did a test on the half-tapered guide where I reversed its position 180 degrees.  That flowed better than when it was installed as intended.  Finally, I filled in the threaded hole for the head stud.  I used modeling clay to fill the hole and smoothed it flush with the surrounding port material.  That flowed the best at high lift.

Here are the results of the special tests.

The Budget Flow Bench is intended to be set up with the cylinder head resting on top.  Then you suck through the head to test the intake ports, and you reverse the flow and blow through for the exhaust ports.  For the tests I did, I had to suck through the exhaust port (in the normal direction of flow) because I only had half-a-head.  I’m pretty sure the results may be a bit different once I test a complete head as intended.
 
To be able to test in both directions, I will have to add a bleed valve chamber so that when blowing through a test specimen the bleed valves will be located on the vacuum cleaner exhaust rather than the cannister (which is on the suction side of the impeller).  The manometer lines will also have to be reversed.

This sketch shows it configured for blow-thru.  

On 1/15/19 I changed the sketch.  I had the u-tube hooked up to the wrong side of the orifice.  It should be connected between the test specimen and the orifice because the u-tube is supposed to measure the differential across the test specimen.

Use a hole saw to cut a 2-3/8” hole in a section of the 4” pipe to accept the PVC bushing.  Cut the hole with the end cap in place to make sure that the edge of the end cap doesn’t overlap the hole for the bushing.  The vacuum cleaner hose will be attached to the bushing.