PTS Flowbench Forum

Chad Speier wrote:I say it has something to do with turbulence with the delta p across the square edge plate. Not to mention it's just past the curtain area and your working off the port area which means a difference in pressure.

Turbulence often shows up in fluid manometers in the form of bouncing fluid. With a manometer scale designed for .8xx SG fluid I would think it would show easier than a scale for heavier 1.xx fluids [blue / violet]. Naturally the intensity of the turbulence would be reflected in the amount of bounce.

With my original manometers and without a baffle plate in my MSD bench cabinet. I ended up using distilled fluid and the green wetting fluid for a 1.0 SG and I saw turbulence very plainly on the inclined. It happened with every head I tested above 90% on the scale.

Another issue was the foolish location of the test pressure pickup. They wanted it mounted in the test fixture cylinder.... Worst spot you could ever put one. Move it in or out a little...whatch your test pressure change... I cured all of those issues.

So let's consider the differences in reactions between a fluid manometer vs an electronic setup. This where I expect Bruce to jump in with both feet with all of his electronic calibration experience he's gained over the past few years.

The electronic units use sampling and averaging to arrive at a displayed number. From discussions with Bruce, it's an eye opening experience. One could conclude that to some degree, turbulence effects could be tuned to a lesser degree than they present in reality with an electronic manometer. Don't hold my feet to the fire on that. It's logical speculation on my part.

I will say that as is, I can see turbulence effects taking place when they do. At least I think so because of the way the digital numbers either hold or don't hold steady. Some of that could be because my manual flow control valves are 6" in diameter. Quite touchy sometimes.

The point is when a turbulence problem is suspected to be causing possible read errors with consideration of that fact that we are comparing benches with orifice and cabinet design differences, how much does it affect a particular type of manometer's ability to reflect that situation accurately ?

I'm not picking on DM's or Analog guages. I'm promoting a discussion to lead to a solution....

One more thing. Unless testing is duplicated with the same stand and exact same positioning of the head on that stand, the results can be discarded immediately when we're splitting hairs over 1-2% in 200CFM or greater flow ranges.

I sure wish Tom Vaught would surface with some input There's a walking library of experience who's input would be valuable...

FWIW, the SF600 hasn't been proven to be anything other than "accepted" and other than the assumed quality behind the brand label, I can't think of a reason it should be. Especially when I've seen huge variances between those benches. So for all of those who subscribe to the "tell them what they want to hear to make a buck" line of thinking...put that in your pipe and smoke it

Might be a lot wiser to haul out Darin's criteria list of what makes a good head....CFM is what? 5th...6th on the list.... OOOOOO I just had to throw that in there.....



Larry C

larrycavan wrote:So let's consider the differences in reactions between a fluid manometer vs an electronic setup. This where I expect Bruce to jump in with both feet with all of his electronic calibration experience he's gained over the past few years.

The electronic units use sampling and averaging to arrive at a displayed number. From discussions with Bruce, it's an eye opening experience. One could conclude that to some degree, turbulence effects could be tuned to a lesser degree than they present in reality with an electronic manometer. Don't hold my feet to the fire on that. It's logical speculation on my part.

I will say that as is, I can see turbulence effects taking place when they do. At least I think so because of the way the digital numbers either hold or don't hold steady. Some of that could be because my manual flow control valves are 6" in diameter. Quite touchy sometimes. Larry C

This was a highly discussed topic when we were working on the design of the DM and software among the "team". The digital gives you the opportunity to make this setting how you the end user wants to see it. By changing the sample rate and averaging you can make the display rock steady or move so fast you can't see the numbers change. It's all how you set it up, for you. We did this so the end user would have that option, some use it, some are happy with just having digital, once they get use to it they can customize it to suit their liking.

I personally set the numbers to roll some and do a "mental averaging". If the average is set to high and the samples to low then the numbers change very slowly, now all of this also comes down to how fast your computer is. But, you have options with digital you do not have with analog.

Hows the saying go . . . "once you go digital, you never go back" or something like that


Do you ever wonder why the automotive magazines do not discuss/address this? Anyone have any contacts?

Call it; Educate the masses of "CFM matters only Sheeple".

Harold's book is a GREAT start on this education! Even shows my plates but, no credits

The good thing is . . . we are here listening and discussing how WE CAN MAKE this better!

I welcome any other flowbench vendor to step up and add their input???

An inclined manometer with a square root scale equal to that of the PTSDM's sensors could be easily made. Hook it up in parallel with the DM. Flow something that will generate turbulence. Compare readings.

Actually, the analog manometer doesn't have to scale 1 to 1 to see the result but it would give inarguable authenticity....

I haven't looked at it in a long time but as I recall, the PTS bench had the orifice lying flat rather than on an incline. To me, that's going to give the orifice a better opportunity at a more even distribution of the flows stream and with that a better pressure distribution balance. Those combined should yield a more accurate measurement. Well, let's just call it a position of lesser influence to flow disruptions

You know it's interesting how you never really see any flow results from a 1200 bench. They're not intended for cylinder heads, I know that but you'd think that someone, somewhere would have access to one and would have some numbers for the intake ports of at least one head that was compared to a 600.

Let's go back to 2006 when I proposed the P.A.P. test..... I see absolutely nothing since that shows any new data that wasn't determinable from that test. Look at the range variances in some of the SF benches compared to our standard, the Ford bench and also to each other. Maybe, just maybe it would be a good idea to incorporate those results into your flow sheets for customers. Then they can see with their own eyes, the reality of this type of testing.

I guess you could technically call the Ford bench a DIY That is to say " it wasn't built in Colorado and it's not blue. Maybe we should just dismiss it altogether..... Or NOT...

So there's no confusion here... There's no PTS bench in those results because it did not exist at the time. Bruce's results were on his old Pitot bench...

Here's a disclaimer straight out of Pipe Max

Due to differences in FlowBenches and Dynos, the required Intake and Exhaust
Flow CFM @28in. may be 20-30 CFM lower and 20-80 HP lower than your Dyno or
your FlowBench . CFM Flow Lag Times will often cause the required Peak CFM
Flow to occur at different Crank Angle Degrees, than what PipeMax predicts.

On my flow bench(SF110 with Flowcom electronics) I use the Performance Trends software to record the data on a PC. I like the software because it "frames" the test and keeps keep me consistent. One other thing that I like is that when I click F1 on the keyboard and tell it to record data it does 30 readings, about 1 a second, averages them and then records the data. This allows it to give a "stability" reading in the data. This can be adjusted to a greater or lesser number but I've always used 30. I think that because I have a little flow bench and can only flow at 10 inches of water that it takes longer for separation to come and go in the port and thus it needs a longer time. On an SF 600 this could probably be reduced to 10 or 15 readings.

What I'm wondering is if by giving it many more samples that it doesn't act as though the digital manometers have a higher sampling rate than they really have as far as the data is concerned? I've never ask Kevin if the averaging formula uses all the data or if it kicks out the highest and lowest "spikes". Could it be that the SF 600 has a turbulence issue that's somehow masked by the heavy fluid in the manometers AND the rather low sampling rate of the electronics, you can't see the "spikes"?

DaveMcLain wrote:On my flow bench(SF110 with Flowcom electronics) I use the Performance Trends software to record the data on a PC. I like the software because it "frames" the test and keeps keep me consistent. One other thing that I like is that when I click F1 on the keyboard and tell it to record data it does 30 readings, about 1 a second, averages them and then records the data. This allows it to give a "stability" reading in the data. This can be adjusted to a greater or lesser number but I've always used 30. I think that because I have a little flow bench and can only flow at 10 inches of water that it takes longer for separation to come and go in the port and thus it needs a longer time. On an SF 600 this could probably be reduced to 10 or 15 readings.

What I'm wondering is if by giving it many more samples that it doesn't act as though the digital manometers have a higher sampling rate than they really have as far as the data is concerned? I've never ask Kevin if the averaging formula uses all the data or if it kicks out the highest and lowest "spikes". Could it be that the SF 600 has a turbulence issue that's somehow masked by the heavy fluid in the manometers AND the rather low sampling rate of the electronics, you can't see the "spikes"?

Dave,

When you use that F1 function, is that done at one particular lift point, at each lift point, or ??? Could you elaborate a little on that process?

As for the SF600 turbulence issue, it's hard to draw absolute conclusions [for me anyway] because when I look at the PAP results, there's a tendency in the data to support such. Then I look at Larry M's data using both the stock manometer and the FP1

Look at the color coding on the worksheet. That's what tells me the recorded range variances are impossible to correlate from bench to bench to bench to bench. Certainly if only 2 benches are compared, a conclusion is easy to draw. When I look at Chad's posted results they pretty much fall into place with what the SF range variances show in the PAP results.

What is particularly interesting is the variances, percentage wise, fall within SF standards for their bench reading variances. That said, how can anyone crucify any DIY bench that meets those standards?

The procedure that I use when I do a flow test is to pick out a test I did previously on a very similar cylinder head. Then I enter all the information, valve sizes etc into the software, pick what lift points I'm going to use, how many cylinders I'm going to test and then I start.

In the Performance Trends software you press F5 on the keyboard and another little screen pops up over the top of the main screen. I keep the little one stretched out so that the numbers are nice and big. The little screen shows current depression, flow range, temperatures, velocity probe readings, swirl, percentage of flow and actual corrected CFM in real time.

I put the bench into the flow range I intend for the first lift point. I open the valve to that point, allow the manometers to settle for a few seconds and then press the F1 key. The program then records 30 readings, averages them and then steps to the next lift point on the graph. I then look at what range I need, go to that range if it's different, lift the valve, let it stabilize, hit F1. I do this till I'm at the end of the test, say .700 lift. Then I press the ESC key and that takes me back to the main screen where I can graphs or reports to look at and or print out. The software works well and I should do a video and put it on Youtube because I don't think there's one out there and it might be helpful to others.

I have a friend who does some really nice cylinder head work and he's done a lot of development over the years all using a home built bench. He often gets aggravated by those who think that a Superflow bench an especially an SF 600 is somehow superior to what he's using. We've compared our results by passing parts back and forth and our results always agree very closely as do results on another friend's SF600. The home made bench works fine and I would have gone that way if I would have had a resource like this forum and PTS 10-11 years ago when I got my Superflow.

Thanks Dave

In the PAP results, the most useful information came from Darin for the current discussion. He had the wisdom to flow in multiple ranges and report the results.

Here's just Larry M's and Darin's results. Looking at this, if I owned a 600 and was going to compare results with another 600, I'd want to know for certain which ranges were used in testing on the other 600. This pic clearly substantiates the need.

The sad part is the SF600 is the accepted standard by so many people and I'll wager most of the time, the details of the tests are not included when guys swap numbers.

Look at JUST Darin's numbers. Some interesting things take place. I have concluded that flowing air through an orifice inside a chamber of the sizes commonly used in most benches, is probably the most inexact method if absolute, carved in stone, 100% comparable scientific data is the goal. All you're reading is environment conditions for that particular head on that particular bench who's orifice plate is set at a particular range. Nothing more - nothing less. It is however what we have to work with and fits nicely into a budget and has a small foot print in the shop.

I will also go out on a limb and say that IMHO, if all orifices bench cabinets were built to a length of perhaps 5 or 6 feet & if the measuring orifices were placed well away from the the location the head mounts to & with baffles sufficient to really calm the air down so that a very accurate pressure differential measurement across the orifice was possible, the test results would probably differ from those conducted on a traditionally designed orifice bench cabinet.

Doe's any of that actually matter very much? I think not. That is unless you're faced with the situation of CFM driven customers that are armed with inferior information, from which, they "think" they draw a fair conclusion of a head porter's work...

Enough...I'm done with this.... [aren't you glad ]