How about a test with just a more competition type of valve job. 70-60- proper sized 45-30-15. Nothing else touched. Ohhh maybe a 30 degree cut on the intake valve.
CNC Mill Mark Removal in Intake Port
- Thread starter Earlie A
- Start date
-
gzig5
Well-Known Member
I was going to ask you earlier about this. What would you say your machine is capable of for precision and repeatable accuracy for the CFM measurement? For instance can it determine to a tenth (0.1) of cfm? Then for a test repeated on the same port under same conditions say five times stepping through the lift range, what would you expect to see for variation for at a given lift number? I'm just trying to baseline the changes I see people reporting, vs machine capability. Changing something that on the next run results in a 10 CFM change may or may not be reality. I am/was an engineer and know that when examining test data, the capability of the measurement system needs to be factored into the analysis.
BTW...I'm enjoying your journey through these tests. Very interesting for me and your methodology seems pretty solid. I don't have room for a full flow bench but this Fall I'm hoping to see what I can cobble together with some controller HW and fans I have laying around from old work projects.
Last edited:
No way can this machine (a Superflow 300/600) accurately determine 0.1 cfm. I'll stick to my 1/2/3 cfm story from before. If I were to run back to back tests within a few minutes of each other, I would expect 2-3 cfm of repeatability. There are so many factors that can influence results. Air temp & humidity. Bench temperature. Rigidity of valve opening device. Accuracy of valve opening device. Vacuum motor bearing and brushes. Then there's the entry plate or clay situation. If it was removed, is it back on exactly as before?
For most applications, the 2-3 cfm accuracy (less than 1% of full range) is good enough.
I agree on the 2-3 mostly on account of in my case the valve opener. You have to use a strong enough spring to keep the valves closed and then it’s a feel, visual thing setting your valve opener to zero. I use cut lexan 1/2 entry plates with a little bit of Duct Seal so my plates are pretty consistent. My head adaptors use 4 head studs so they center everything up nicely
And when you sink the valves, the valve stem ‘gets longer’, the valve spring is less compressed, the vacuum can suck the valve open a little bit and you can gain low lift flow. For a while it seems like you just hit the jackpot and found a bunch of flow - until you realize what really happened.
Good stuff man, thanks for sharing this!
dusterbing the peace
FABO pyrite member
Echoing the sentiment, thank you for doing this!!!
Here's a post to clean up a few loose ends from yesterday. Included is a picture of my plywood entry plate and how I use clay to blend any mismatch with the head. Also included is a picture of my manual valve opening jig. I've found it to be quite rigid and repeatable. There are countless ways to do both the entry plates and the openers. They have to be customized for each new family of heads.
With the flow tests I want to address two issues. The first one is repeatability of the flowbench (this is different from accuracy). Tests 593 and 594 were run back to back on a new, untouched, factory CNC port with no changes in between the two tests. The head was not taken off the bench. Bench temp was 76.5 for the first test and 77.6 for the second test. The vacuum motors inside the flow bench heat the bench up slowly when running intake ports and very quickly when running exhaust ports. When running intake ports the heat from the motor inefficiency is exhausted out the back of the bench. On exhaust ports this heat travels through the bench and out the head. Anyway, as can be seen, the repeatability was 1.4 cfm or better. Pretty good for a system measuring turbulent flow.
This test shows the bench is quite repeatable. Most benches are calibrated (for accuracy) by the user on occasion with an orifice plate that comes with the machine. I suppose this yields decent accuracy, but for most of us it doesn't really matter. For most flow tests, I am looking for the change from one test to the next. I want to know if the little bit of grinding or little bit of epoxy I just added helped or hurt. Accuracy would be nice, but it's not worth the trouble. Repeatability is what I want.
The last test 595 is a flow test done after sanding the short side only of the same port used in 593/594 with 180 grit paper. So, the head was removed from the bench, the intake valve removed, and the SSR sanded. The plywood entry plate was not removed. As can be seen, flow picked up a little.
With the flow tests I want to address two issues. The first one is repeatability of the flowbench (this is different from accuracy). Tests 593 and 594 were run back to back on a new, untouched, factory CNC port with no changes in between the two tests. The head was not taken off the bench. Bench temp was 76.5 for the first test and 77.6 for the second test. The vacuum motors inside the flow bench heat the bench up slowly when running intake ports and very quickly when running exhaust ports. When running intake ports the heat from the motor inefficiency is exhausted out the back of the bench. On exhaust ports this heat travels through the bench and out the head. Anyway, as can be seen, the repeatability was 1.4 cfm or better. Pretty good for a system measuring turbulent flow.
This test shows the bench is quite repeatable. Most benches are calibrated (for accuracy) by the user on occasion with an orifice plate that comes with the machine. I suppose this yields decent accuracy, but for most of us it doesn't really matter. For most flow tests, I am looking for the change from one test to the next. I want to know if the little bit of grinding or little bit of epoxy I just added helped or hurt. Accuracy would be nice, but it's not worth the trouble. Repeatability is what I want.
The last test 595 is a flow test done after sanding the short side only of the same port used in 593/594 with 180 grit paper. So, the head was removed from the bench, the intake valve removed, and the SSR sanded. The plywood entry plate was not removed. As can be seen, flow picked up a little.
gzig5
Well-Known Member
Good stuff. Thanks for going through that. I agree with your feelings on the measurement accuracy. That will come out on the time slip, but the repeatability is what will get you there.
I need a calibrated bench as I want to feel confident in my numbers I’m giving customers. Kinda one of the reasons I’ve had three different benches. My bench doesn’t matter if it 90 degrees or -10 degrees.
Dale Davies
Well-Known Member
According to David Vizard, surface finish is not as important as the port shape and efficiency. You need to exercise caution on port volume as a larger volume creates a lazy port flow. Flow rate and the energy areof vital importance to cylinder filling.
Habit of a rough surface causes surface turbulence, which tends to toss fuel clinging to the surface back up into the air flow.
Dale Davies
Well-Known Member
Judging by the numbers, does your foot hurt?
I believe that is the consensus of most of the resident racers and engine builders that responded here. Got to get the right size and shape. Then do not smooth it out much, if at all.
Newbomb Turk
Well-Known Member
That’s the problem with being a reader and not a doer.
Because port surface finish has an impact not only on power but on jet sizing as well.
If you can run 5 sizes smaller jet and make the same (or more) power the car will be quicker.
Darin Morgan already said you can’t make them too rough.
Larry Meaux says the same and he does the exhaust ports and chambers in a rough burr finish. Says he’s making more power the more he does it and the rougher he makes it.
Chad Spierer won’t do a manifold that isn’t burr finished. And some of his intake ports are burr finished.
Once again the claims made by DV and others because he’s not the only one saying it are found to be wanting.
If you don’t test you won’t know it.
Plus, I’ll say with pump gas or any alcohol based fuel today there is power in a burr finish.
You have to test it or you won’t know.
- Joined
- Dec 31, 2005
- Messages
- 8,645
- Reaction score
- 4,675
That depends on who's work you read.
Does the fuel that's impinged on the walls travel as fast as the fuel in the air column?
Is the fuel that's traveling on the walls nice and fine and well atomized?
But my flow bench told me.......
Dale Davies
Well-Known Member
Go to the zoom and climb into the chimpanze compound. They like sitting around picking nits.
Newbomb Turk
Well-Known Member
You must not follow Spierer because he’s changing his thinking some. He is really pushing the new Killer Carbide finishing bur
Newbomb Turk
Well-Known Member
I saw that.
I sure would like to see this quantified with before and after TQ and HP curves and BSFC numbers. Speier, Morgan and the upper echelon of engine builders will spend a crazy amount of time to gain 2 HP.
You think he’s using that finish on the throat and SS as well? That would be a real reset in the ‘it makes sense to me’ category.
Edit: I'm referring to the top of the throat, as in the bottom angles directly beneath the seat angle. Surely the first couple of bottom angles are still smooth.
Last edited:
-
