Fabricating velocity probes

[larrycavan]

[larrycavan]

If there's a correction factor worth applying it may be from your shop temperature to the actual average temperature inside the port of the live engine.

Your not looking for a micro measurement but a trend / overall picture of what's happening inside the port.

Poke the thing in the port and have a go...then stop and ask yourself....how am I going to pick up or slow down "xFPS" in this particular area? You'r not going to be asking how can I pick up or loose "x.xxxFPS" in this area?

[Mouse]

As air density decreases, more velocity is required to create a pressure differential across an orifice or a Pitot tube. But the piece under test will flow more air too because of the less dense air. End result: orifice and Pitot tube do not see a change in pressure differential.

Now, you are not seeing the actual increase in flow, unless you factor in the density correction. This is not usefull in flow test comparison, but is usefull if you are plotting fuel curves, or working with HVAC or measuring air speeds in non-ratiometric applications.

[larrycavan]

[mattpatt]

I hope it's ok for me to drag this post back from the dead, but it's definitely interesting.

I actually bought some 1/8" and 1/16" brass tube the other day with the intention of making a pitot tube. I was trying to figure out how to bend it so seeing these little 1/8" benders that some of you have used has cheered me up. I'll have to hunt one down.

As for the holes in the outer tube, I think that I'ver got that pretty well covered. 1/32" has been the number mentioned several time. I don't have a 1/32" drill, but I do have 0.4mm~1+mm in .1 sizes, so should I just jump in there with the 0.8mm, or is there another size? Not sure if 1/32" is "THE" size or not.

Also, concerning the 'ram' hole in the front. Should this also be 1/32" (0.8mm)?

Picked up some good ideas as to how to make the split of the two tubes at the back end, so many thanks for that chaps.

I've only had the bench finished for a couple of weeks, but it's very addictive. The trouble is that just measuring flow is really of little concern if I don't know the velocity. And it would seem from the formula given in this thread that it's not difficult to calculate. Also downloaded the spread sheets on this subject, so don't even need to hunt down my calculator. Fantastic.

I really have to say that this forum is one of the best 'technical' forums I have been a member of. Quite a few others I've tried (not flowbench related) seem to shun you unless you're already an expert, but it would seem that the experts on this particular forum are more than willing to share their vast knowledge and experience with bench virgins like myself.

[slracer]

[color=#000000]Hi all, I too am going to dig this thread up again. Here and in other posts the discussion has been made about where the static holes should go in the probe. In addition, there has been some question of how small the bend can be made. Thinking about these 2 questions, I have created a different design. My thinking goes like this. It appears that one of the limitations for the radius of the bend is the size of the tubing so smaller

[jfholm]

Doug,
After I sat and looked at that and analyzed in my brain I would say "#### good idea" You broke from the paradigm!

It would be easy to make also. Make the bend in the small tube first and then slide the larger tube up from the other end and solder in place and there you have it.

I have made a special little jig for holding the two tubes centered while I soldered them, but this design you could eliminate that step.

btw for brass tubing and those little mandrel benders you can go here It is so far the cheapest place I have found and nice fast shipping.

John

[86rocco1]

I don't think that design is a good idea because there's a fairly large lateral separation between the pressure and static sample ports, in a small port, it's quite possible that there'll be a speed differential between the position of the pressure sample point and the static sample point which would skew the velocity readings somewhat. I think this design would suffer from severe reproducibility problems.

Consider this example: if you were sampling close to the port wall where you'd expect low speed, you could position the static pick up point so that it's closer to the center of the port where the speed is higher (hence lower static pressure's lower) and without moving the tip you could rotate the probe so that the static pick up is also next to the wall, an area of higher static pressure so with the tip of the probe at the same location, you're able to generate two different velocity reading. In this particular example, it's fairly easy to decide with reading is better but I chose this example because it's easy to visualize the problem, if you're sampling in the center of the port, the same sort of thing could happen but it would be much more difficult to make sense of the numbers.

[Tony]

The static port is (or should be) totally insensitive to velocity, that is the whole idea.
The very small flush pin holes just tap into the attached (stationary) boundary layer that clings to the probe body to read static pressure.

As Doug says, our ports have a very steep pressure profile along their length. This truly brilliant idea of Doug's is a very ingenious way of getting around that problem.

This very high pressure gradient in which we are measuring, is a unique problem for us flow bench freaks.

Most common applications of pitot tubes work in an almost constant static pressure environment. So just because nobody else does it this way, in no way diminishes it's advantages for our special application.

Great idea Doug !!

[jfholm]

Proof is in the pudding though Doug. It looks like you will still have to build two. The new style and the old style to prove the theory. :;):

John

[jsa]

Total pressure is supposed to equal static pressure plus dynamic (velocity) pressure.

If the velocity and static ports are in regions of the ports with different velocities, then the static pressures should be different, skewing readings.

One example of where I see this being an issue is on small ports, where the static port is near the wall and the velocity port is near the port center.

So I'm on the negative side of the fence until you demonstrate otherwise.

[slracer]

John and Tony, Thanks for the comments! I was beginning to worry as over 100 views of this post and no comments! Perhaps because I use a pitot style bench, I may be more used to thinking about pitot design, but the twin tube (inside/outside) has not made sense for me. I originally was thinking about a siameze tube where only the static was bent at the end, but Bruce's "flailing in the wind" comment made me rethink that one. Since the stiffness is esentially a factor of R^4, a 50% larger tube gave 5x the stiffness. On the end at the single tube, the bend and short length (I hope) keep it from moving.

Ed, as Tony says, the static pressure in a plane across the port SHOULD be constant. In the flow element for a pitot bench, the static pickup is in the side wall about 1/2 inch in front of the pitot pickup. In my 1-1/2 inch diameter tube, that puts it WAY away from Bruce's averaging pitot (and a 1-1/2 inch flow element is small). I would expect more variation ALONG the port than across due to wall losses and turning losses. The variation you are describing is the velocity (or dynamic) pressure, not static.

OK John, so now I have to build one! I looked at the Tower Hobbies website. Which bender do you use? It appears the K&S spring tubes are the only ones that go down to 1/16 inch but the DuBro 1/8 says it will go down to 1/16. I wonder about that some as they also make a 5/32? (BTW, is that the ID or OD of the tube?) The K&S also says 4 sizes from 1/16 to 3/16 but the pic shows 5. 5 makes more sense as that would include 3/32, 1/8, and 5/32. 4 leaves one of those out. Both benders also say use on soft brass, hardened brass will not work! Is the brass they sell hard or soft? The description says "easy to bend and form" so I'll assume sift. I'll call them tomorrow to find out, but your input (and anyone else's) on the benders would be appreciated. As for building 2, I think a better answer would be to build 1 new design and send it to Bruce to test against his standard. (Are you lurking out there somewhere Bruce?) After he sees one (if it works), he can improve on the design and then make them for sale! :p

Larry C and Larry M, you are both big users, do you see any trouble spots? I've never done a velocity test in my life so this was just a different way to look at things!

Thanks all, Doug

[bruce]

Yes, I'm "lurking" . . . from a production stand point I know what it takes to build my current design and without divulging to much about my construction techniques it's going to be a bear to build. One question I would ask is; How will you control the tip radius at a constant dimension?

I'll continue to lurk . . .

[jfholm]

Doug,
I bought the Dubro bender in the 1/8" varity and that is o.d. btw. They sent 4 or 5 short pieces of 1/8" tubing with the bender to let you practice. I also bought some 3' lengths of the 1/16" and 1/8" brass tubing they sell. It will bend ok.

A hint, bend slowly and you may want to keep a thumb or finger pressed down firmly in the center of the bend on the mandrel. The first piece I bent I just slammed it home and it kinked flat right in the center. After that when I went slow that little bender made some very tight mandrel bends. It is nice! Some lube may help also.

Yeah send one to Bruce. That is a good idea. Also what the other guys are saying has got me thinking about the static pressure holes. I know the velocities vary quite a bit in my port and would that not make the static reading different also? My understanding is that we are trying to get a pressure differential reading between the pressure port and the static ports and would that not mean we need to be as close as possible to having both ports in the same air stream?

It will be interesting to see. Also I found someplace, but can't find it now, that the static ports should be about 300% of the outer tube diameter from the pressure tip. Since my outer tubing is 1/8" my static ports (holes) are 3/8" from the tip and are about .030" (four equally spaced)

John

[Tony]

One source of very small bore tubing is the refrigeration industry. Many types of refrigerators do not use a proper refrigerant expansion valve, but a length of fairly fine capillary tube.
All domestic air conditioners, refrigerators, and freezers use a length of copper capillary tube, and a fairly wide range of sizes are available.

You may even be able to salvage something from a dead refrigerator.

Google turned up this, just to give an example of the type of stuff I am raving on about:

Copper capillary tubes

Outside Diameter: 1.83 mm - 6.10 mm
Inside Diameter: 0.66 mm - 4.45 mm
Average Walthickness: 0.584 mm - 0.862 mm
Area of Cross Section: 0.343 mm - 15.5 mm

The thin stuff will be pretty flexible, (just like a hollow copper wire), but I would expect it will work harden fairly quickly. It would be fairly easy to work with, but would certainly not look as nice as stainless.

So it may be ideal for a quick do it yourself garage experiment, but it would be a bit lacking for a saleable presentable commercial product.