Gattling-Style Pitot Tube Bench - My new design.. thoughts?

[Ken73]

I'm not the best artiste~ so hopefully you can make sense out of it. The green box up front is the motor box; the black piece is the frame, and the red is the testing box. The blue "turret" looking thing is 4 pitot tubes; 1.5", 2", 2.5", and 3" .. there will be mechanisms in place to seal them tightly but you will be able to pick which one you want to flow with by turning it to the tube you want. The two green tubes in back are for venting. The large green circle in the back area is for selecting exhaust/intake.

The "turret" will have provisions for the manometers to hook up as well, so you won't have to mess with changing lines around.

Keep in mind I haven't worked out 100% of the details, this is just a preliminary design, so if you see a pitfall please feel free to post it. Or if you see an improvement, post it too.

[2seater]

Interesting idea. I do not have any direct suggestions to improve the design, and I can see how it could work. It would seem to be pretty complex and the only real downside, (aside from the basic engineering) would be the motor power requirement will probably be higher than a simpler design. Every change in direction the air flow is required to make adds to the friction loss so the ducting to vary the air flow direction should be as large as possible. Another small area of higher loss will be the smaller diameter pitot tubes will be longer than required for laminar flow. It may not be a big thing, but it does take more power to move air through a longer tube of a particular diameter than a shorter one. I would like to know more as you work it out.

[jsa]

I'm looking at 2 pitot's to cover my expected flow range.

Your 4 way looks interesting, but I wonder how difficult sealing 3 tubes plus 6 pitot ports will prove to be. It seems a lot of the oriface plate guys have trouble sealing one rotating disk.

[2seater]

I would think all tubes would not need to be sealed at the same time, only the operational one. As long as the chamber where the tubes reside is otherwise sealed, the passage for the tube in the active position would be the only one needing a seal at both ends. If there was some way to cam the seal out of the way while rotating and then tension it when in position might be longer lasting, as a sliding seal would tend to wear, of course the more moving parts, the harder it will be to get them sealed as well. Teflon rubbing against formica makes a good low speed, high load bearing. Possibly Teflon o-rings partially recessed around the holes sliding against a formica surface would work? One other possibility would be to install a full flow valve at the exit end of each tube and simply turn them on as needed, tube holder stationary. Could also combine operational tubes, more than one active at a time, for in between readings, with multiple flow measuring devices Interesting engineering to work out.

[Ken73]

I was thinking of using foam - the thin sheets you find at the craft store - for sealing. Your idea of camming the discs so there is no tension when turning it is what I envisioned. The tubes would be attached to a plexiglas disc that would seal against the foam. Rather than cam it, I was thinking of using some sort of cam-bolt type lock on the bottom disc since that's where the leaking would cause erroneous readings. Leakage at the top wouldn't cause erroneous readings, just wouldn't allow as much airflow. Also, you're correct in that only one tube would be in use at a time. The others would be sealed off at the bottom and probably the top as well.

As for the tubes being too long, that could be rectified by having 3" tubes necked down to the tube size so the pitot tube was only as long as needed. If I remember correcly, Bruce told me 6 times the diameter on either side of the pitot - i.e. a 3" diameter tube would need to be 36" long (3"x6=18" - twice as long = 36".) So the 1.5" would only need to be 18" long, the 2" would need to be 24", and the 2.5" would need to be 30" long. The turret would be 3' long for the 3" tube, the rest would be necked down from the 3" tube to their corresponding lengths.

However, the chamber where the tubes reside will be open; no airflow will be present around them though. You have to be able to reach in, and see the tubes so you can turn them to the correct position and line the tubing up.

[2seater]

I think the general rule for laminar flow in a pipe is ten diameters upstream and five diameters downstream, or fifteen diameters overall. This can vary somewhat depending on the lead in and exit, flow straighteners etc.. The largest diameter you use will set the overall dimension, or about 4 ft for a 3" in a conventional setup. The larger diameter lead and exit tubing makes sense and the holes in the plates to hold them could be the same size as a bonus. Another simple expedient would be to use stoppers in the unused tubes as long as you will be accessing them to rotate the assembly, and avoid the sealing difficulties? Not dissing anything, just sort of thinking out loud.

[Ken73]

OK, I bought some 5' sections of PVC.. all I found was 1.5", 2" and 3" .. so this may be a 3-tube setup. Any reasons why I should try to find something in 2.5" ?

15x the diameter would be 45" for the 3", 30" for the 2" and 22.5" for the 1.5" .. I picked up 1.5" > 3" and 2" > 3" adapters as well while I was at it.

Also picked up a 1/4" sheet of lexan to make the turret bottom and top from.

Hoping to pick up another sheet of MDF this week while I'm off.

[2seater]

I think it depends on if you are going to go digital on the measuring instruments or manometers. From my limited experience there seems to be a "sweet spot" or maybe "sweet range" where accuracy of a pitot-manometer combination is the best. I think the digital has better abilty to read the high and low range of any particular flow tube essentially extending the usable range. I haven't gone digital, so I use several different flow tubes: 1", 1.25", 1.5", 2.0", 2.5" and 3.0". The ranges have a fair amount of overlap, but when testing a MAF sensor vs frequency and comparing that to the volume measurement from the pitot, I found a modest range where the readings agreed the best with the mass reading.

[Ken73]

I'll go digital later on, right now I'm not looking for super-duper accuracy, just something to get me in the 2-3CFM range. (However, digital IS in the plans, I just want to get the bench up and running right now.)

Still working on the sealing aspect. ???

I will probably make my own digital setup, as I do a lot of electronic stuff on my own.

[2seater]

How about just borrowing the stopper idea from the orifice guys? You would only need a small access door to add or remove stoppers from either end to prevent flow and would likely be just as quick and more reliable? Almost no moving parts. Simple can be good

[Ken73]

Yea, I'm thinking you're right, but I'm not real big on the stopper thing - how about the knife gates? I did a lot of looking around and seems like a lot of others are using them. I want something I can change while the motors are still on, even.

[2seater]

Nothing wrong with valves, simply a better and more expensive form of stopper. If they are downstream they should have little effect on the stability of the measured flow although adding a little "insurance length" wouldn't hurt. I use an adjustable air bleed and an inline homemade damper valve with a vernier cable to operate it for the flow control. Not elegant or fast, but very reliable. Many more sophisticated ways to skin the same cat and it doesn't matter as long as it works repeatably.

[Ken73]

Okay, so I think I have an idea on how to do this a little better; first, rather than rotate the entire assembly, I'd just rotate a disc at the top, and the three tubes would be stationary. The disc would not only select which pitot tube to use, but the manometer connections as well.

I'm thinking of two pieces of acrylic, one circular with one 3" hole in it, then the lower piece with the three holes in it for the different tubes. The manometer connections would be acheived by drilling holes in the lower plate, and machining corresponding channels in the upper plate to connect them.

To seal them? I'm thinking about spring-loading the top (rotating) plate, and the selector knob on the control panel would simply be pulled up to release the tension. Then you could rotate the knob on the control panel to the desired tube. Release, and the spring goes back into tension, pushing the two lexan plates onto each other solidly, and hopefully sealing the whole mess.

Of course, I have yet to work out how the knob controls everything just yet, but I'm getting there. Still drawing on paper to get the ideas.

I've pretty much decided on a three-tube design though, 1.5", 2", and 3" tubes.

Also, after reading some of the other posts, I did see where someone was using a router control for a single motor for depression control, and single switches for the rest of the motors. I think that along with flapper valves is probably how I'll go about the motor box.

[larrycavan]

Ken,

I didn't read every detail in this thread AND I admire your courage.

I'll offer you a bit of experienced opinion on rotating devices in a flowbench.

First, YES it can be made to work. BUT...and there's always that darn but.... you will have your work cut out for you.

Sealing a rotating device is no easy task. Exhaust sealing is particularly difficult.

If you are convinced that is the way you are going to build it, then maybe the disk is the way to go. I would design it so that the bottom of the disk had a rubber gasket material and the disk had some spring loading to push it down onto the pipes. Maybe a lever[s] to lift the disk, rotate it and then lock it down in place with the pipes remaining stationery.

The opening in the disk would need to be large enough so that no influence from the disk's orifice Cd would affect your readings.

Exhaust pressure on the disk is considerable. You have to clamp cylinder heads to the test stand for exhaust testing and they weigh considerably more than a disk would.

Personally, I hope you proceed with your ideas because it can be done and I'd like to see someone do it.

Larry C

[Ken73]

I didn't realize the pressure was that great.. I'll do a bit of re-thinking and maybe some experimenting. I still have some ideas on how to keep it down though. The nice thing is I don't really have to worry about Cd in mine, since I am using the same size tubing (3") for all three tubes, then necking them down. I think that might be another advantage of this particular pitot design over an orifice plate design.

Springs don't have to be used necessarily either for the clamping pressure - I could also see using some sort of wedge to keep the plate in place. Pull the knob, the wedge(s) retract, releasing pressure. The wedges might be spring loaded so they return to their clamping position.

I can get 6"x6"x1/16" rubber sheets at the hardware store and make a seal with them.

Thanks for the encouraging words; I realize I have my work cut out for me. I've already built one flowbench before (quite inadequate) and I'm working towards one I don't have to do as much to while I'm testing (removing probes, only having one flow range, etc.)