Floating Depression?

[jfholm]

Just one more quick post. I use Engine Analyzer Pro software from Performance Trends. It is best when entering head flow into it that you be as accurate as possible. As the old saying goes GIGO, garbage in and garbage out

John

[BigBro74]

John- I also want repeatability, without it you have very little I also wish to get pretty accuarate numbers for the same reasons as you with comparison to others work and for the sims. x2 on the GIGO!

ON another note I have made plans to get my hadicrafted orifice plate and one of the heads (the pair i still have here) flowed on another flow bench. I will have to retest the the head on my own bench also, this pair of heads i have little interest in porting (to me-junk, chev 882's) and only used them as something to test on the bench but did not number the ports that i flowed. I only flowed 1 head and averaged the numbers. I then put them back on an engine to keep debris out, and don't remember now which head it was or which port flowed what! To save $ I am going to have them flow one port and use that. The 461 heads i ported I have better notes on as spent alot of time working on them but they are racing friday nights on a guys car that lives about 100mi away. Hopefully this testing can give me/us some insight into the experiment. I will think hard on the PAP's also in the future as another way but same thing to a degree, they are plates not heads, although they are undoubtably made to a tight tolerance and would help correlate to others' work who have flowed them. It may be a week or two by the time i get the stuff shuffled back and forth, the shop with the bench is 36 mi away. Hey...... and now its starting to cost me. pay to play right nothing is really lowbuck!

I have read Dalton now Rick. It is a great read. I especially liked the work on anti-reversion techniques. A lot in there to chew ( ihad a long post on all the good stuff but i got logged off while typing so lemme just say I DO Like it!) on and am glad you reccomended it. That said, I don't think DV would have been expelled for plagairism.They both advocate using a vacuum cleaner to move air through auto parts/ heads and such at different depressions. But, to me there is one big distinction in what they ask the reader to do, even perhaps why. Dalton asks the reader to us a valve to divert some of the flow through a valve in the test rig and test at different flow rates at all valve lifts and such to find when it falls in its a$$. No where does Vizard ask the reader to divert any of the flow anywhere but through the test peice. He even says not to at some point if trying to correlate the numbers to flow at 28" as it will make errors. But his reasoning comes from a depression map of when an engine is pulling different depressions at different valve lifts and crank degrees. Low lift as when overlap period and the headders are pullin big time dep. valve wide open max lift only piston demand, much smaller amount of depression. This did/does make good sense to me as another alternative to static testing.

J

[BigBro74]

OK- so here is what i got (unvarnished). the shop is a reputable engine builder in our area and i know builds top running dirt late model engines (these are unlimited cars 800+ hp). The person that did the testing said the flowbench is a superflow brand, but said when asked if it was a 600 model he said "thousand, it's a larger bench". is there an older model 1000? all i see new are 1020 and 1200. The sheet has written (i am assuming manual bench, no print-out or other fancy stuff) which range it was flowed in and a number and then a cfm number. this is the reading at the manometer and then corrected? I was not in the shop area, they have a customer counter in front (and my 5 and 2 year old kids were with me so ) so i did not get to see their set up. The results may leave me with more questions than answers.

882 intake port.
lift sf flow@28 - My flow
.100 61.9 - 60
.200 118.6 - 110
.300 176.3 - 168
.400 197.6 - 187
.500 200.6 - 195

home made plate flow- this is the part i have real questions about
hole marked - sf flow
5 - 5.2
10 - 9.9
20 - 24.7
40 - 43.3
80 - 93.2
160 - 182.5
all holes open
315 - 374.7 ???? on this the results do not add up. if all holes are tallied=358.8. i understand there may be some venturi effect with all flowing at once. is this in line you think? I expected the plate would flow more than what it is marked as i made it by hand and i dont even have a drill press. The larger two holes especially are somewhat oversized (they are actually not round) by some 5-15 thou. the other holes are closer to the print but not perfect.

all is interesting to say the least. leave me some input on all this.
Jason

[BigBro74]

So, after spending last evening using the old grey matter between the ears I come up with a few questions for the shop with the flowbench (question 1: which model= sf 1200 manual bench). Question 2= "What cylinder bore adapter did you flow the head on?" answer= "4.165"" (not what I asked them to use, asked for 4"). question 3= "I noticed on your sheet by the way the numbers are written you proabably flowed the cylinder head first" ans="yes". Question 4="After you did that did you flow the plate on top of that bore adapter?" answer ="yes". I told them when i brought it that the plate is made to go on the bench without a bore adapt, directly on the 5" hole. I guess they didn't take from this that i wanted THEM to do the same?
Now I don't want to say the did a bad job or anything-they were nice, didn't charge much to do a PITA little job, and answered all my questions without hesitation. I just must not have made clear enough EXACTLY what i wished then to do.

This may answer a couple questions i personally had about the results of my test. First is they used a larger bore on the cylinder head, it should have gained a little flow over my results from this which is what happened. Second for me was why the plated would flow so much more than expected? like i said earlier, i expected some higher just not that much. It is my understanding that a bore adapter can do this to an orifice plate?? to what degree i do not know maybe you guys could help me out here? The PAP's on my own bench will be the next logical step when $$ allow i guess.

It was Very interesting to me though to see how the intake port was very similar to my own results just a few % higher overall. My own concern would have been that the low lift flow would be significantly less and the higher lifts (with lower depression on my bench) would have read a larger gain than they did.

Interesting...... Thoughts?..... lemme have it. Bruce ,Rick, are u here?

Jason

[1960FL]

Flow your plate on a bore adapter close to their size and see what you get?


I am not sure what you are trying to achieve here, maybe validate your results? but you have no comparative calibration! You need to have an orifice plate that when you put it on your bench it pulls a consistent depression, say 10" you know the plate (Bruce Made it for you, Maybe) then you have your motor guy flow this plate at the same exact depression at the same location (On the bench not the adapter). Now you will have a standard to compare to no matter who's bench you flow on the math will say so. You have to start with a standard if you want to compare numbers at all! Period. Oh and yes i know you made your plate but unless you made two identical plates (within at least .002) and regression tested against each other to determine the cd it really has little value at this point.

Rick

[BigBro74]

good idea! I will have to make one. (4.165" bore adpt.)

Rick In an earlier post YOU asked if i had run any of these heads on another bench to compare. That simply is why i was doing it- for YOUR own viewing pleasure. I really dont care too much what the thing flows other than rough idea. They wrote .? cfm and that is way too small a number for me. i care that the heads ported using it work well and so far that is the case. saying that I have no known standard is true to a point because i built it myself and was not able to hold perfect tolerances. The point is I did not design the plate. It was designed by another person, who is an egineer, for this specific purpose. I suspect That is the reson it is made in the fashion it is!! With it the bench is always using different sized orifices for different cfm. they are not sharp or square orifices. the two smallest are nozzles. the rest are radiused top/ square bottom. The thing is i just read my change in pressure between the plate and the bench before hand (and after sometimes to check) and recorded it as it should correlate to cfm @ 28"h20 by the plate. The design itself is proven accurate, just my tolerances not.

I know, Rick, that you have probably forgotten more about flow benches than i will ever know(seriously :!:compliment ). I am just trying to lend a little of what i have found on this (not to be a PITA). Did you read D.V.'s whole speal on this? It all makes good sense to me. maybe not everyone agrees as much. I'm not a "crony", never met the guy. like to read his stuff.

[1960FL]

Jason,

I am not bashing you on this you are the first of many to at least follow up on your post what happened to “kennedyrd” so I applaud what you are doing. What I am trying to make clear is that in your thread you have continually tried to extrapolate CFM to something, but for some reason you cannot bring yourself to buy a known standard, in which to compare your findings. If you read this forum and others I will continually preach this fact no mater what bench design you have, FD, Pito, LFE, OP. If one does not have a base standard to compare to then why look at anyone else’s CFM numbers as they are useless. If you continually spoke of % improvement or loss then I can take this design with some value as it is a comparator, but it was you that wanted to look at cfm numbers I think your first post eluded to this.

I have my own opinions and sometimes they are strong, I believe in standards and one is that I do not believe you can take flow numbers at 10” and say they are the same as those tested at 28” or 40” or 100” yes this math may work for sharp edge orifice plates but not for a port with turbulence. This is the core of my disbelief in this style of testing, you are changing to many variables at one time and I personally believe air speed and CD each have a profound effect on flow and induced restriction. Thus in this design as you think you are improving your head flow but you are by the nature of the design slowing down the air speed and thus reducing the effect of upstream turbulence. This is one of the main factors for testing at a fixed depression as it sets the standard for the base air speed across the port.

In this style of bench you have no standard to compare to you are only comparing to past data, the other styles of bench you have the internal standard (Pito, LFE, Orifice) and you are comparing your changing CD/CSA (As you port) against a know standard the flow is across this standard and not across the tested part.

I am in no way devaluing the learning process of what you are doing, in my early days I used to do real wet flow, as a kid I used the garden hose to push water through the port and try to figure out what I was seeing. I still do this some times I just can’t help myself.

Jason you are very close to having a viable DIY flow bench it does not have to be a PTS design look at the merc dog it uses hole saws to create the internal orifice but once calibrated it works and can repeat.


Rick

I have another post i will put up later but it explains some the history here also.

[Michael Moore]

I only flowed 1 head and averaged the numbers. I then put them back on an engine to keep debris out, and don't remember now which head it was or which port flowed what!

A friend suggested that I keep a "flow check" head on hand. It really only needs to have one useable port with guide so you can saw it out of a bigger head or find a cheap junk head somewhere. It doesn't matter if it is stock, ported or what, you just want it to flow the same every time you put it on the bench.

Even with the use of a calibration orifice he likes to also do a quick run on the checking head just to make sure that all the numbers are coming out looking like they should. He says that the check head has alerted him to new leaks or similar problems before he'd spent a bunch of time with a "real" head.

It seems like a reasonable thing to do.

cheers,
Michael

[BigBro74]

Rick - I find that i agreee with you on several points! first, i dont believe for a second that flowing a port at 10" is the same as flowing at 40" or 100" etc. I think it is this very thing that first piqed my interest in any (and especially floating dep.) flow bench design.
Engines NEVER pull a fixed depression on a port. for this reason alone i believe that testing at a floating depression with a high starting depression is more appropriate. It MUCH more closely simulates what happens in a running engine.
Finding a way to measure the velocity in a known tube or orifice would be a really great way to actually measure the cfm
but in this application the cfm is (like you point out) just a number that is useless until converted to a fixed depression to have any way to compare to something else. I wish i had a great idea here as to how to correlate all this but i do not. it is very much an apples to mixed fruit comparison..... I agree.
The only way i know to relate to everyone what is happening is to use a number of cfm corrected to a known depression. I will and want to get a set of bruces plates to help this relationship i am just real close cash now and will future. I did find it interesting that the port flow had a track that was very similar to mine with a standard dep. verses a floating one regardless of the numbers assigned to it. I think vigerous discussion is a way to come up with answers and ideas! Best.

[Michael Moore]

You can convert CFM at one depression to CFM at another depression, you just use an appropriate multiplier.

It sounds like the concern is more that pulling a real 100" at .300" lift may give different air behavior in the port than 10" at the same lift. That is probably the case, as velocities may change and a SSR that is fine at low speeds has separation at high speeds. But in the ideal straight tube test port different depressions probably shouldn't be giving any difference when they are all corrected to some standard depression.

I'm still trying to figure out how to get a valve at the end of that perfectly straight port.

A friend hooked my Flow Quik to his orifice bench and could read the same port setup with both in quick succession. He saw no problems with the FQ numbers vis a vis his bench.

How many people are going to be shipping a head back and forth to someone else and therefore need to be able to compare numbers off of two different benches? I think that for most of us repeatability within our own bench's numbers is going to be of far more importance than traceability to NIST.

Maybe if we're a Pro Stock team with big money at stake it would become more important to have a 100" bench. On the other hand, having any bench that lets us quantify flow is a big step up over nothing at all.

I don't need a better bench, I need to understand what the bench is telling me and what I need to do to make the port happier.

cheers,
Michael