Wide Variations in Air Mass/RPMs -- Intake Air Buffeting or Computer/Testo Lag?

[geargrinder]

R - it’s not just the DME. Phillzo has a blower. So that’s gonna throw intake flow comparo out the window for myriad reasons.

Sadly I still cant help... cuz tho’ TU, blower also... same prob.

[R Shaffner]

Sounds like the wolf in Red Riding Hood. Or me after my wife gets back from this trip she's been on.

(It's a rainy day and I'm waiting for someone who's late. Pls excuse my pensive rambling.)

Actually, a vena contracta is not so charming. For those who don't know, it's the compression of air flow as it passes through an orifice, as shown in the pic below.

Vena Contracta.png
Several times I've simplistically opened up intakes thinking it would help with air flow. (Going back to my youth when I didn't know better, and more recently when I sorta knew better but did it anyway.) Two 540i examples: Removing the snorkel and leaving the opening in the stock air box. Removing the air horn in the exit for the stock air box. Both were done as tests, but seemed to work OK.

As I explained earlier, I was expecting worse air mass numbers when I removed the air horn inside the air box, and I was surprised when they got better. I had the chance to ponder this on some long drives last weekend, and I think the higher air mass numbers (if accurate) might be because of the vena contracta and how close the air box opening is to the MAF.

The sharp corners of the air box exit (without the air horn) restrict the air flow overall, as if the air were entering a smaller tube (about the size of the vena contracta). That also accelerates and compresses the flow in that center area.

Flow-separation-and-vena-contracta-in-sharp-edged-entrance-to-tube-31.png

In short, I think the meter part of the MAF probably sits right in this narrow area of faster, compressed flow. I suspect the overall air mass going to the TB wasn't any better than with BMW's stock air horn.

If I wanted to prove this I guess I could find some way to put more distance between the air box exit and the MAF, but it's not that important to me. Mostly, I didn't want to leave anyone with the wrong impression that removing the air horn from the stock air box makes sense. The acceleration times from 3000 to 6000 rpm in both cases were the same.

Or... maybe the air mass readings weren't accurate in any of these tests, for reasons stated above.

[geargrinder]

Mmmmm SCIENCE! Me likey.

Although I'm pretty sure the MAF tube should have been engineered long enough to take that into account? But maybe not. For sure those grills in the MAFs (you know, the ones that the shade tree guys delete along with the helmholtz and the venturi velocity stacks etc. et. etc. ) are there to straighten turbulence out. Honestly we know all these things are developed on flow benches with extensive fluid dynamics testing by BMW... I'm always amused that people think that somehow BMW engineers and bean counters were deciding "lets put these extra parts in the car for no reason... lets make the intake really restrictive just to eff with people...". And I know 100% you're not saying that, I'm just kvetching as usual.

I've said before in CAI threads with big "but my common sense says..!" claims for why homebrew stuff "must be better", back on one of my old Audi platforms a dude once took what to laymans eyes as a terrible restrictive factory airbox and tested flow repeatedly with various mods and factory parts on/off etc. Actually it was very much like our E39 boxes, a square box shoved in the front corner, with some weird velocity stack jammed into it a way that made it look 'restrictive'. But he pretty much proved that the factory setup was better than any "deletes" or "drill holes" or CAI setup you could cobble together. Didn't stop about 75% of mod bros from doing that crap anyway, but for those paying attention, it pretty much proved those engineers did their homework and probably used a shtload of science that "derp derp open it up!" engineer does not.

[whiteghost1]

Usually with any fluid volume measuring device, there is a minimum required development length on the inlet and outlet side, to provide a stable signal. For the most part this length has to be without contortions, or the measuring device itself is designed to stabilize the flow within, if the space required is limited (ie. the honeycomb that GG mentioned that the shade trees remove) or via a designed profile internally. This is where all the “extra” velocity stack, helmholtz, screen, etc.... work in unison to provide max efficiency within the average median flow profile, while not being a significant detriment to the upper and lower ranges of this profile.
In this case, this is probably where the average driver spends most of his (or her) time in rpms.

[JimLev]

For what it's worth.....
My Dinan CAI and the larger Dinan MAF tube (thank you WG) do not use honeycomb, neither does my LS3 engine.
The LS3 states the MAF must be at least 6" before any tubing bends.

[whiteghost1]

For what it's worth.....
My Dinan CAI and the larger Dinan MAF tube (thank you WG) do not use honeycomb, neither does my LS3 engine.
The LS3 states the MAF must be at least 6" before any tubing bends.

No thank you for your contributions to the community!

It’s ironic you mention the Dinan MAF tube not having a screen, as I was thinking about that point as I was babbling above. Ahhhhh yes......screens.

In the case of your LS3, with it’s required min 6” development length, there may be a constant laminar flow in all flow regimes to not warrant a screen (I don’t know TBH as I’ve never even seen an LS3 intake tract ��), or perhaps is not turbulent enough to begin with, or there is some other designed element of the intact tract to make it a non issue.

Regarding the M62tu Dinan setup, I believe that with the Dinan stage 4 engine mod you speak of = large CAI/MAF/intake tract/TB.....that the engineers at Dinan were going for the sledgehammer approach, and opened up everything wide from fender to intake manifold. And having said that, were focused on obtaining max power in the upper power band, which is a different mandate than BMW had, which is max engine efficiency over a broad range whilst remaining relatively silent. Dinan whistle anyone? In addition, the Dinan stage 4 setup is pretty well smooth bored from air filter-->intake mani, whereas the stock setup has sharper undulations and greater ripples in the intact tract, ie. to enable engine movement in the case of the post MAF intake boot. This may effect the Froude number, which may also be one of the reasons why the stock system has "extra" components to mitigate the turbulence created.

Or it could be in both cases, specifically why they don’t have a screen, could be the difference with or without is inconsequential. I don’t have any emperical evidence to back up the above statements, just throwing out hypotheticals since we don’t have the test data to base an opinion on.

[geargrinder]

Yeah the Dinan tube is long and without edges or constrictions so I could see them deciding it didn’t need a straightener. The aftermarket tube I use definitely has one though (my choice) as I think it prob adds more than it hurts.

[R Shaffner]

About the screen/straightener -- I always thought it was mostly to keep our fingers out, and as a fail-safe for filter/intake problems. Removing it is one of the things I'd kinda like to test, but obviously I can't with the air mass variability I have.

And yes, I agree that many have been simplistic and stupid about CAIs (perhaps more that than anything else). I have too at times, which is why I wanted to be more scientific about it.

We all agree that the BMW engineers know more than we do and have done mostly smart things. But we also agree that they've done some bone-headed things too. I even started a thread about to get others' thoughts:
https://www.bimmerforums.com/forum/s...s-BMW-Has-Done

Progress often isn't linear. We all know many examples of that. (Our chain guides are a great one. They went back to a center sprocket, or sprockets, on the N62.) So the smart BMW engineers will make mistakes and basically acknowledge them with changes in later years.

For me, the 540i snorkel seemed like one of those mistakes that BMW abandoned (modified significantly) in later models. I'll elaborate on that in another post below.

[R Shaffner]

After I'd driven my '97 540i for a couple of years, I'd learned that it didn't like heat. The engine was clearly less powerful in summer stop-and-go traffic with the AC on. (I live in Charlotte.) And every fall when the air turned crisp, it was like it had gotten a much-needed tune-up. Throttle response was better. Acceleration was back. I'd fall in love with the car again every fall.

I remember one hot August afternoon in particular. We were taking another couple to a wedding and had to stop at a store while the girls ran inside. We sat there for 5-10 minutes, idling in 100 degree heat, in the sun, with the AC on full. I was annoyed and when we finally pulled away the car just wouldn't go. I'd give it WOT in 2nd and no one noticed. It was like I was driving normally.

Up until then I'd always figured it was knock-sensor issue -- that the timing and/or fuel were being dialed back to prevent knock. But this was so bad that I finally tracked down where the intake air was coming from. Most here know what the 540 snorkel looks like; here are pics for those who don't:

IMG_2857.jpg IMG_2858.jpg

Of course, the constriction around the bumper support didn't look good, but when I saw the opening was pointed backward, toward the radiator, I had to remove the snorkel and take a look. On the right is what I saw. Half of the opening is blocked (the flaps on the left would open only when air was flowing away from the intake opening. And on the right the aux fan would pull air away from the intake. What the F?

So I removed the snorkel, and I swear it seemed to drive better on hot summer days, especially in stop-and-go traffic. (I know air builds up in front of the radiator at high speed, especially Autobahn speeds. I bet the snorkel works OK in Germany. Not so good for driving in the South.)

Over time, I also did a little more research. I learned that the e39 528 has the same snorkel, just on the opposite side:
P4070535 (1).jpg

(continued below)

[R Shaffner]

(Sorry, having pic problems)

Then I learned that the e39 M5 had what looked like the same intake as the 540, times 2. (Twice the intake for 600cc more engine, and a little higher rpm?) So looking at other models suggested that the stock 540i intake might relatively small for the engine.

And then I noticed that the M5 snorkels were different. The 2 snorkel openings do not face the radiator where they are blocked by the flaps and have to fight the fan. And both the left and right oval openings (where the 540 and 528 snorkels fit) are blocked, as shown below.

user12636_pic46818_1487639111.jpg

And finally, I've helped some friends and family fix some newer BMW models, including an e60 550, and I noticed that the newer factory CAIs use a different and much more direct design.

In summary, I don't want to go back to the factory snorkel, and just one M5 snorkel might be half of what I need/want.

I don't really want to wrap my intake system in insulated foil, but some summer days I wish I had. I also wonder if one of the main advantages of Dinan's CAI is lower IAT in real hot conditions. Plus it would be easier to insulate, if one really wanted to go that far.

(Sorry to exhume and once again beat this dead horse. Just wanted to explain the heat angle from my Southern perspective.)

[R Shaffner]

Excuse me for hijacking my own thread. Now back to Testo air mass readings...

I probably should have noticed this before. Below is a graph on RPM to KPH on one of my 2nd gear WOT runs. (I wasn't on the clutch and it's behaving fine.) It's mostly linear, but not perfectly.

KPM and RPM.png


Here's a graph that shown KPH/RPM, by RPM, using the same data.

RPM to KPH.png

From this is seems that there may be a few data gathering and sync issues. (This reminds me of discussions about how an Income Statement is for a period of time, and a Balance Sheet is for a point in time.) So how does the DME record and report each piece of data? And how does Testo get it? Are they for moments, like snapshots? Are they for the same exact moments or ones close together? Or are they for brief periods of time, that may or may not be the same, and that may or may not overlap? Or some combination of the above? Who knows?)

In the example above I was at WOT in 2nd. Everything was changing pretty quickly. It's easy to see how the data might have a jagged appearance. It's less clear why there's a downward trend as RPM picks up, but again, I suspect it's a data definition, gathering or reporting issue. (My clutch doesn't seem to be slipping.) Perhaps it's like a movie of a spoked wheel, where the wheels seem to move forward, then stop, then backward, etc.

And finally, here's a graph of a brief steady-speed segment, where KPH was the same out to 5 decimals:

RPM at 56 KPH.png

So it's clear that the first test I wanted to do with Testo just wasn't a good one. And I don't think I need to be concerned about my MAF readings.

(Please understand that I am NOT knocking Testo. Not at all. It's great and I am very thankful to Mika for developing and supporting it. Every tool has its limitations. I just came across one for this valuable tool.)

[R Shaffner]

Thanks Mika. Yes, I was using File Logger, which works great. And the 7.2 Serial input. I agree, it sounds like an ECU issue -- how it reads and stores data.

For example, if we have two sets of data, where each has a different stair-step function, then when we compare the data we'd see frequent up-and-down motions as the stair steps moved in and out of phase. That would explain what we see in the RPM/KPH charts above.

Again, thanks for the program and your ongoing support!

[geargrinder]

Just looking at your pix... I'd kind of forgotten the factory snorkel blocking the fan port deal (mines been gone for so long for different reasons).

I'm sure the factory design is predicated on the entire space behind the front radiator panel being lightly pressurized at speed. I am sure those rad shroud flaps open up at speed, no problem. And it probably works "under way" fine. However indeed, you would think at a stand-still the fans would be creating a negative pressure zone behind there.

I did remove my 'block off plate' for the Aux fan on the right side but looking at the M5 setup maybe that's a mistake. I probably want to have 2 block off plates there to maximize 'ram' effect cooling at speed, and, to make the Aux fan draw cooler air from the kidneys and under-bumper grill vs from behind the bumper bar. Not gonna run out and mess with it tomorrow, but I am planning an preventative maintenance Aux fan replacement and that would be the time to do it.

[R Shaffner]

Maybe... With the M5 they might have blocked both openings to route cooler air to those snorkel openings. But that wouldn't explain why they have one blocked on the 540 and the other on the 528.

Sometimes this is like wondering why the ancient Egyptians did things the way they did.