Another newby! - General flow bench questions...

[5.0joe]

Hello All,

I recently joined the flow bench forum in hopes meeting others with an interest in flow benches. Judging by all of the posts I have read so far, looks like I came to the right place! Anyways, my name is Joe, and I am located in Cincinnati, OH. I am currently a senior in the Mechanical Engineering Technology program at the University of Cincinnati. For my senior design project, I haven chosen to design and build a flow bench, one specifically for testing automotive cylinder heads. When finished, I will more than likely donate the apparatus to the college, and they will be able to use it for laboratory purposes. In order to make my goal obtainable, and since I have to prove out all of the design with formulas, etc., I chose to keep it as simple as possible. The design I am going to follow is for the most part based on the EZ Flow system offered by Performance Trends. I have attached a layout sketch of the system components and their arrangement. Feel free to take a look and critique it or make any suggestions you feel would help out.

Now, I have a few questions and am hoping some of you experts on here can maybe point me in the right direction to get a solid start. First, the facts and values I have semi-determined so far. To prove out my design, I am going to use a Chevy small block cylinder head (availability, cost, etc.) Generally, from my research, a SB stock intake valve will flow around 215 cfm, and the exhaust around 175 cfm. So, in order to accommodate to any potential port and polish work, I want my system to be able to flow a maximum of 300 cfm (28 in H20). However, in order to minimize the amount of vacuums required, I am going to test at 10 in H20, and then convert the calculations up using a conversion factor of 1.670. This means 300 cfm will be approx. 180 cfm instead. This should take the vacuum requirements from 3-4 motors down to probably 2. Does this idea seem practical? Additionally, can anyone recommend some good, reliable motors that will get the job done at a decent price? I am assuming the vacuum motors will need to be able to reverse the flow to test the exhaust port, correct me if I am wrong? Or is there another way of test the exhaust port where I do not need to reverse flow? Sorry I am very new to all of this. For the orifice plate, I have downloaded the orifice calculator spreadsheet on the forum. Depending upon the pressure drop (delta H), I am coming up with orifice hole diameters between 2.8 in-3.0 in. I will be using 4 in. I.D. sch. 40 PVC throughout (see attached sketch for more detail). Any suggestions on the orifice hole diameter and calculating it (if incorrect) would be greatly appreciated. Finally, I wanted to make note, although I listed it in the sketch: I am going to try to avoid the use of a plenum box / expansion chamber by substituting flow straightener (honeycomb) at various intervals in the pipe length. My questions will probably be very vague without seeing the sketch I attached. Thank you all for any feedback or suggestions! It is great to become a member of such a knowledgeable forum, and I look forward to reading and learning so much more.

Thanks,
Joe

[49-1183904562]

[color=#000000]Joe

Welcome to the forum, First off I

[SWR]

One flaw I've found with the PT setup is that you will,every time you go up to the next orifice size,have a "step" in the flow curve. This is due to the pipe distracting (guiding) the flow through the orifice,so your actual discharge coefficient is un-calculable. It will work,and you will see improvements with it,just don't think that your numbers will be directly comparable to say an SF600.. that is,unless the pipe for that 180cfm @ 10" EZ-Flow-system is 8"+ dia,and have atleast 4 feet of straight pipe before the orifice. Then you'll be +/- about 3 cfm between orifice sizes... a 5" pipe with 4 feet of straight pipe before the orifice will be off + or - about 7-9 cfm. Higher depressions (28" ) skews the results even more.

You can then end up in the interesting scenario that you pull 100% on your inclined and have 100 cfm,change just the orifice to the next size and do a pull,and then have 107cfm (or something) as your end result. If you make an accurate 200 cfm orifice,it will also be impossible to use it to flow and calibrate the next smaller size orifice. This goes on until you're down to about an orifice size flowing about 45cfm@ 10" (5" pipe) then it levels out and give consistent results. An 8" pipe will at least be accurate to about 170cfm @ 10". My next bench will have two settling chambers 2 feet wide x 3 feet long x 1.5 feet high...and belive me,there's a reason...

[bruce]

This is kinda off topic but an interesting thing to try if you have an orifice plate to play with. While flowing an orifice plate on top of your bench put your hand or a board about 12" above the orifice plate and slowly move your hand/board down towards the plate while watching your cfm's, you will notice that the hand/board has little to no effect on your readings till about 2-3" above the plate. Now take your hands rest them on the bench top in a cupped shape around the orifice again starting 12" out from the plate and slowly move them towards the orifice hole and see what happens to your flow as you get closer.

This will be the effect a pipe wall has on the Cd of your orifice plate as SWR has stated above. The plate "sees" the flow from the sides not the top so you need the settling chamber to give the plate an area to get clean air from for a consistent Cd.

[106-1194218389]

I actually have the PT EZ Flow system and helped him in the preliminary developement of it. We did at that time find that the bigger the tube then better the results. With his electronics hooked up it does a very good job. Right now I am using 4" PVC and when I tried using the inclined manometer with it it would not read correctly with the inclined manometer. I am in the process of building a 6" tube version and as SWR mentioned an 8" tube would be good, but hang onto your wallet when you start buying PVC that big. 6" gets a bit pricey. The flow straightners are essential. I have found some 1" thick plastic honeycomb with 1/4" cells that works great. I also have 2 feet before and after the orifice. Bottom line is works very well with Performance Trends Black Box. So far have not had good results with inclined manometer.

[5.0joe]

jfholm~

Are you using a settling chamber in your system? It seems pretty conclusive that a settling chamber will probably be necessary after reading the responses, which is somewhat what I expected to hear. This would make the Cd more predictable. Would this, in turn, avoid the need for 8" piping?

[SWR]

[SWR]

[SWR]

[larrycavan]

[quote="1960FL"][color=#000000]Joe

Welcome to the forum, First off I

[115-1172523331]

Good morning Joe, As the builder of one of the most recent pitot type bench completed, let me add my $1.00 worth. (It used to be 2-bits, but inflation caught me.) The final part of my bench is documented here:



and some of the earlier construction here:



Notice that I changed designs between the 2 sets of pics!

The bottom line is that I was also budget limited and finished the bench for about $350. I only play with very small (6 CI) engines so your motor requirements and bench size will different. Rick (1960FL) suggested (above) a Harbor Freight controller for depression control, they work well and are cheap $20 ( $10 on sale). As a suggestion, see if you can add "flexibility" in your design. If you donate it to the university, the professor will appreciate it. {In my bench, I built a variable inclined (actually 2 as the first wasn't working as I wanted) and can also change out the flow element from 1-1/4 ID to 2 ID.} You don't have to build those extra pieces, just show how it would be done. It's possible that it would be worth extra points in evaluation! Best of luck! -- Doug

[5.0joe]

Doug-

I actually read your "finished flowbench" post when I first check out this forum. Great job on giving all the details, by the way! Now that it seems I am going to need the expansion chamber, I was curious as to what guided you in choosing the size of yours? I will be need to flow a bit more than your setup (~300 cfm at 28 in.), so I am assuming I will need a little bigger of an expansion chamber? But is there an equation or specific criterion to follow when building one? Also, I am assuming you probably placed a PVC flange on the inside of the chamber at each of the two pipe outlet locations and bolted them to the box? I know that having a leak proof seal is critical to the accuracy of the flow bench. Thanks for your input!

-Joe

[115-1172523331]

Hi Joe, First, almost all of my design decisions were based on having flexibility in the bench. If I don't use it too much, I have friends in the racing world who would also like to test some. Second, I had some leftover MDF so I made the box as big as I could using that piece. In terms of design criteria, I was an advanced design engineer for aircraft and have a million rules of thumb on airflow. In designing ducts, I have always used a minimum volume of 3 times the displacement for a plenum, i.e., a 5L motor, a 15L airbox. (A 24 inch cube.) I don't remember where it came from and I have seen many smaller numbers used with good results. My settling chamber is 10X because I "might" test larger engines and I had the room and materials. I know Bruce uses a 6" pipe, 18 inches long for his small (but bigger than mine) engines and gets the results he was looking for. You can find PVC "slice gates" (or other names) at most plumbing supply places. I used one I was given here on the forum (from 1960FL, I think). They had a "Repair Kit" for the gates which was the flange used on each side. I got the 2 inch size and added reducers to get to the 1-1/2 inch I am using now. I am building a 1-1/4 for even more sensitivity. They are at each end and have a flat rubber gasket sandwiched to the wall and bottom of the settling chamber. 8 bolts into T-nuts and I have another "range". More flexibility. I have run my inlet blocked to 90+cm depression and show no flow on a 1 inch rise on my inclined manometer so I guess I have no leaks. If you cannot find design criterion, find examples and tell the prof what you did. I used to "review" papers for technical committees and I always thought that initiative was important! (But remember that the a/c I designed were going to fly about 10-15 years after I worked on them. Err on the safe side!)

Another post showing some construction details is here:



My suggestion is to learn from the people here on the Forum, they have taught me a lot! Keep your design simple as it is easier to "debug" and faster to build. Don't be afraid to ask for free stuff and discounts as a student project. If you don't ask, the answer is "No". And, most of all, Have Fun!

Doug

[106-1194218389]

To SWR, I'll test again but it seemed like it would read about 70% and the orifice will actually flow much more than what it says it will when you use the calibration spreadsheet. Nothing wrong with the spreadsheet it is just the way the orifice acts in the tube. But I will actually set it up and redo it.

John

[5.0joe]

Doug-

I am actually in the design engineering dept. at General Electric Aviation, in the CF6, M&I engine group. I am going to work up some revised sketches with a plenum box (after determining the size I wish to go with). I have been racking my brain looking at all kinds of websites with useful information, and there are so many ways I can go with this. But again, keeping it simple and flexible able my main concerns.

Thanks Again,
Joe