CentralCoaster

Ok, before you roll your eyes and think I'm an idiot.. listen up.

I'm trying to model the vacuum pressure in the cylinder on the intake stroke. My preliminary plan is to use a compression fitting to hook a beefy vacuum line up to one spark plug hole and measure the dynamic vacuum pressure over a range of engine speeds, ideally up to or near redline. I would of course, unplug the injector on that cylinder. My professor first suggested using the brake booster line, but the vacuum there doesn't fluctuate as much, and is affected by all the cylinders--I only want to model one. I'll be collecting the data on the computer, if it can record it fast enough. This is all part of a senior project I'm working on BTW. ;)

I'm open to any other suggestions! I might discover I'm wasting my time trying to measure vaccuum at these engine speeds though. The only other way I could think of (without drilling holes in my manifold!) is to bend a pitot tube just right and feed it through the plenum and partway down into a runner tube. Then I wouldn't have to run it with a plug out.

I don't see any reason it would hurt anything, except for the engine being underpowered.. But the computer may try and compensate, or throw me some hoops to jump through. It'll probably run it too rich first of all, and this might cause knocking... hmm, I'm speculating here by myself.. I could use some feedback!

Worst case scenario, I don't run the test. I just don't want to give up on this phase of the project unless it's unavoidable.

CorvetteZ51Racer

I wouldn't do it because you're running a now out-of-balance firing order and putting a lot of stress on the bottom end. I do the types of measurements you're trying to do for the purposes of engine research. Unless you have a BIG budget, you're not going to be able to measure the pressures quickly enough to make any kind of useful model. It's normal to take these measurements at every 1/2 degree of crank rotation, which at 4500 rpm is 54k times per second. Also, if you were trying to model it in relation to crankshaft position, which requires an encoder...which is a good $1000 easy for a cheap one. And if you're going to run the engine for any length of time, you'll need more than one, because they really won't live for too long (we break ours every 15 or so minutes taking measurements at 9000 rpm. The DAQ board that you would need which would do sample and hold at 54 kHz per channel is also going to run a couple thousand $$. We do in-cylinder pressure sensing data acquisition on dyno cells, measuring the pressure inside each cylinder at every 1/2 degree of crank rotation at 9000+ rpm. The equipment we have to just get STARTED on such a project is in the range of $30-40k.

That all being said, if you're not too concerned about the time correlation of the data, and you don't need it sampled so frequently for the purposes of your model (much lower of a sampling frequency will actually give you aliased data), then I would try to figure out a way to do it with all 8 cylinders running.

CentralCoaster

Well crap, that's what I tried to tell the professor... the data aquisition wouldn't be fast enough. I get the feeling he wasn't listening to a word I said anyways. :crazy: Thanks for the tip, I'll have to settle with manifold vacuum pressure, and just vary the rpms for that. That should be relatively constant, (or at least pulsing so fast that the transducer can't react!) so I'll just measure that across the rpm range.

Even with aliasing, I could probably still zero in on a maximum cylinder vacuum, but like you said, I couldn't get any useful data without enough resolution. .. do you know off hand what a typical SBC cylinder vacuum pressure would be, or what I could expect? (or is it classified :D )

I guess it's possible that the one of the DAQs here at the university is fast enough, they aren't afraid of spending tens of thousands of dollars of the taxpayer's money on lab equipment. Half a degree is overkill for what I'm looking for, where something like 5kHz might work well enough.

Also, couldn't you model the crank position by just setting up an inductor to plug wire #1? Aww crap, then you'd have to know the spark advance... hmm but that's electronic, and that can't be too hard to measure if you're calibrated, as it doesn't change too quickly.

I think I'm getting in over my head here. I had the ignorant idea I was gonna design and build an intake manifold the old fashion way. That didn't fly well with my project advisor though.

Thanks for your help! Glad someone knows what's going on around here...
:sleep:

CorvetteZ51Racer

If you intent is to design an intake manifold, and you do not need to do the DAQ as part of the project, design it for an average manifold vacuum (and average runner vacuum with IV open of 20.54" H20 (which is equivalent to 300 fps airflow velocity). If you try to sample at 5kHz, you'll miss EVERYTHING! Let me put it this way. During IVO, intake runner pressure changes from 0" up to ~41-42" and back down to 0" (or even a positive pressure if your intake valve closing time isn't just right)

If you decide to go with manifold vacuum (as in plenum vaccum), you'll need to measure it as far away from a runner opening as you can, otherwise you'll get interference in your readings (assuming you're going for an overall average vacuum with this scheme).

If you are wondering about average cylinder pressures, I can tell you for non-race engines that they run at around 10-12 bar or so. I can't tell you for the race engines I have hands on experience with.

Oh, and FWIW, the equipment that we have for in-cylinder pressure measurements were obtained through grants and sponsored research, though it didn't hurt that Smokey Yunick donated a hand-built motoring dyno to us.

Rynda

Sounds interesting and would be a lot of fun to do. You mentioned getting a pitot tube into a runner, can a pitot tube be small enough to fit through a notched intake manifold gasket? Even if you drilled a small hole through a runner, epoxy steel{or aluminum?} would patch it up quite readily.
If you approach this like an engineer and not a physists, I'm sure you will find a method that works and is practicle to implement.

CorvetteZ51Racer

You don't want to measure the pressure in the intake runner using a pitot tube through a gasket. They do make pitot tubes that small, but any time you're putting something between the media and the sensor, you're introducing potential error. The frequency of the pressure changes is high enough that you'll start getting resonance in the pitot tube giving you VERY bad results.

If you want to measure the manifold pressure, my suggestion would be to disconnect the vacuum fitting where the EGR plumbs into the plenum, and get a pressure sensor that you can screw into that threaded hole. You won't need the EGR while the engine is running far above idle (and won't notice its absence when it's near idle anyway), and using that line will not open up a vacuum leak in the overall system (if memory serves on the vacuum hose routing.

Remember that what you will actually be measuring in the intake manifold is inches of vacuum, not positive (above atmospheric) pressure, unless you have forced induction.