DIY Flow Bench

[350HP930]

I am not too concerned about getting exact CFM numbers or pumping a lot of air through the ports. I just want to make sure that the heads and manifold will flow the same amount of air through each port so that my A/F ratio is consistant across all the cylinders.

I know that I may not be able to create the turbulence that could be an issue at higher revs but between the machining, port matching, removing all casting roughness and good blending work I would be willing to bet that the heads are pretty well matched.

Despite the clean up work I have already done to it I suspect my carrera manifold will be a different story though.

I do have another shop vac so for the sake of science I might put a T in the plumbing and see what I get with double the flow though.

[snowman]

Originally posted by camgrinder
The problem converting to a higher pressure drop is turbulence. Some ports will become turbulent at a higher pressure drop. I have tested heads at 10" and done a back to back test at 28". Sometimes this works fine , other times the port goes nuts and the airflow drops.

Do you have any referances to why there would be turbulance at 28" water vs 10" water? I would suspect the same relative turbulance at both pressures. I haven't seen any myself or have I measured any difference. My suspicians would be a problem with the equipment.

Check out the following links
http://www.superflow.com/support/support-flowbench-faq.htm

http://www.superflow.com/support/support-flowbench-works-how.htm

http://www.superflow.com/support/support-flowbench-which.htm

And finally an explanation of the ammount of error that you might get at much much higher test pressures, Note it should be less than 1%, which on almost any home made bench would be more than wishful thinking.

http://www.superflow.com/support/support-flowbench-what-best-test-pressure.htm

With your engine running at 240" of water and you testing at 10" or 28" whats the difference?

[350HP930]

The faster flows at higher vacuumes are more likely to cause the flow seperations and disturbances that cause turbulence.

[camgrinder]

The problem I have seen only shows up in 2 or 3 heads out of 100. Usually it can be fixed with some attention to the short turn radius.
Its my belief the higher velocity seen at the higher pressure drop
can cause the air detach from the boundry layer and tumble.
The interesting thing is at higher valve lift the flow comes back in line.
I have a 2 freinds with the SF110 flowbench. When they see a slight oscillation on the manometer on thier bench (@ 10"), and I test the same head on my SF600 (@ 28") the oscillations are dramitically multiplied.
FWIW My bench calibrates within .2 % of the setting when it left Superflow in '92. I have also tested an orifice plate on mine and 2 other SF600 benches in the SF Bay area and all 3 benches were within 1%. Flowbenches are like chassis dynos... They are best when compared against themselves and not compared to a bench or a dyno 1000 miles away.

[350HP930]

Well, if I adjust my single test run up to 25" of vacuum I get about 180 CFM of flow through 0.4" of valve lift and my TP tube entry flow stabilizer.

Does anyone have any other 911 flow numbers at 0.4" of lift to compare that to?

[snowman]

911 head flows

Your within 7 cfm. ie 190 vs 197
Try the clay, see pictures.

also see

911T Head Flow Rates

[350HP930]

Well, I had a chance to test the bench this weekend and it was very consistent. Between switching between heads and running the bench for different amounts of time it always produced the same results. As soon as I get back from dinner and visiting my daughter I will post the head flow results on my rebuild thread later tonight.

If my calculations are correct my intake port flows are phenomenal.

[jluetjen]

I think that the value of your flow bench is not to determine if your heads have "phenomenal" flow, but rather if they flowed better or worse after you have made any changes to them. To determine if the heads have "phenomenal" flow you'll need to compare them to a set of stock heads -- which you may or may not have done. But it's not clear from your posting if this is the case.

[350HP930]

I took one before measurement and have calculated a flow improvement of almost 40%.

My current calculated flows of >255 CFM @ 25" vacuum and 0.45" valve lift are also very respectable.

Compared to other numbers I have looked at thats about as good as it gets for head with 49mm intake valves.

The CFM numbers may vary depending on the equipment used and what is used to guide the flow into the port but this whole experiment was all about making sure the head flows were evenly matched.

I am happy to report that all the heads except one are within 1% of each other, and one is about 1.5% off from the rest.

I will be looking to see what is different about my oddball so I can try to correct it.