Modified EZ-Flow bench problem

Anything that does not fit into pitot or conventional orifice flowbench design
Post Reply
Malvin
Posts: 451
Joined: Sat Feb 27, 2010 12:50 am
Location: Pennsylvania

Re: Modified EZ-Flow bench problem

Post by Malvin »

SWR

My thoughts are your pipe is not big enough where your orifice plate is If it was me I would build a round cylinder 18" across and 36" long
put my support orifice plate dead center of that cylinder with a 5" machine hole in it place to baffles one below the support plate one above
the support plate each baffle made round 10" or so pressure tap at the top of cylinder and one above the orifice support plate and one below
it . Its just my thoughts but hope it gives you some help or helps trigger ideas for you :)
1960FL
Posts: 1338
Joined: Fri Jan 08, 2010 10:36 pm
Location: Maryland

Re: Modified EZ-Flow bench problem

Post by 1960FL »

SWR,

I am not doubting the accuracy of your software I am questioning it's application based solely on what I have read here in this thread, yes a 27" cubed chamber is plenty big for what you are doing. My point is that if you use some math that allows for calibration then you can see if the problem is related to math or the effect of say DP, Beta, Reynolds number on the CD. I thinking the issue is not linear.

Rick
SWR
Posts: 63
Joined: Sun Feb 28, 2010 10:00 am

Re: Modified EZ-Flow bench problem

Post by SWR »

Malvin wrote:SWR

My thoughts are your pipe is not big enough where your orifice plate is If it was me I would build a round cylinder 18" across and 36" long
put my support orifice plate dead center of that cylinder with a 5" machine hole in it place to baffles one below the support plate one above
the support plate each baffle made round 10" or so pressure tap at the top of cylinder and one above the orifice support plate and one below
it . Its just my thoughts but hope it gives you some help or helps trigger ideas for you :)


Malvin, Thank you. That is basically what I am trying to achieve with my twin-chamber bench-test-thingy. I bought 11-layer laminated ply boards yesterday to make up yet another two-chamber bench with a very low Beta Ratio and all the bells and whistles to eliminate any build quality issues - does it really need a baffle either side? Will that not destroy the Vena Contracta on the post-orifice side? - and also be using the Excel sheet so I can adjust the Cd either way to get things going, like Rick mention in his above post. Then atleast we should see a conclusion to this mess, and I can play with the pipes later when I have time.

But now, gentlemen, it is Christmas Eve here, food and presents and my little family awaits me. I wish all of you and your loved ones a very Merry Christmas. :)

Kind regards,
Bjørn Deildok
SWR Performance
Norway
Brucepts
Site Admin
Posts: 1851
Joined: Fri Jan 08, 2010 3:35 pm
Location: Pennsylvania
Contact:

Re: Modified EZ-Flow bench problem

Post by Brucepts »

SWR wrote:does it really need a baffle either side? Will that not destroy the Vena Contracta on the post-orifice side?
On my plates in a PTS bench you can get within 3" before a baffle effects the numbers. 5-6" is what I reccommend in the plans for baffle distance.
But now, gentlemen, it is Christmas Eve here, food and presents and my little family awaits me. I wish all of you and your loved ones a very Merry Christmas. :)

Kind regards,
Bjørn Deildok
SWR Performance
Norway
Merry Christmas to you also!
Bruce

Who . . . me? I stayed at a Holiday in Express . . .
larrycavan
Posts: 235
Joined: Tue Mar 09, 2010 9:46 pm

Re: Modified EZ-Flow bench problem

Post by larrycavan »

Square root scales are laid out based on the rise. Change the rise and your scale graduations are thrown off.

I don't see the need for special software. Use the spreadsheet Ed posted up here a decade ago. Start with a Cd value of your choice and adjust it from there until you are accurate with readings. If the Cd is very low, then it must be the cylinder is affecting the flow of the orifice significantly. That is providing you know for certain your test pressure readings are accurate.

I don't see any need to make more complicated that that.
Larry C

http://www.cavanaughracing.com
Brucepts
Site Admin
Posts: 1851
Joined: Fri Jan 08, 2010 3:35 pm
Location: Pennsylvania
Contact:

Re: Modified EZ-Flow bench problem

Post by Brucepts »

larrycavan wrote:Square root scales are laid out based on the rise. Change the rise and your scale graduations are thrown off.
Might want to go back and look at the math on this? Once you set the length of the water gauge your percent scale is set based on that length, changing the rise merely changes the angle of the percent scale it does not change the percent scale markings across the length of the scale. If you change the length of the scale then it will change those markings.

Using Ed's spreadsheet you will see this by changing the rise of the scale it does not change the percent marking using the same length. Changing the well diameter will also effect the scales required rise.

I always recommend a water gauge be calibrated to plates and not a set rise, knowing the plates cfm and where it should fall on the percent scale based on the internal plate range you set your scale rise to show a plates given percent. Keep in mind that if the pivot point of the scale is not at the 0 mark you will need to move that end also so you have 0 at rest while also being able to meet the percent scale based on your plates cfm rating. This can take some time to calibrate to a range of plates on the same scale. Again, back to my 80% calibration, from my experience this gets you pretty darn close on the first go to having a couple plates match your scale without alot of "back and forth" of the scale rise, least from my experience I have found.
Bruce

Who . . . me? I stayed at a Holiday in Express . . .
SWR
Posts: 63
Joined: Sun Feb 28, 2010 10:00 am

Re: Modified EZ-Flow bench problem

Post by SWR »

Brucepts wrote:
On my plates in a PTS bench you can get within 3" before a baffle effects the numbers. 5-6" is what I reccommend in the plans for baffle distance.
Would you recommend having an as large baffle as possible to make sure there is nothing but lateral flow? I am thinking as the chamber is square, I can fit a circular baffle that almost touches the walls of the box and still have plenty area for flow in the open corners. Don't want to have a mess like the first 2-chamber bench, as optimum as can be made is the deal now... I go for broke this time. :)
Brucepts
Site Admin
Posts: 1851
Joined: Fri Jan 08, 2010 3:35 pm
Location: Pennsylvania
Contact:

Re: Modified EZ-Flow bench problem

Post by Brucepts »

You just want a baffle that keeps the flow from coming straight into your orifice hole, my plans use a 10" square above and below the orifice plate. The one below the plate is for exhaust flow but neither have any effect on flow in either direction.

Some people that have built a PTS bench have come up with their own idea of baffle plate size and shape and so far we have not seen any benches that have a Cd that is out of range with what others are seeing. A few have even changed the plenum sizes to suit their application or shop space and I find the same Cd calibration numbers in the same range. I try and make it a point to ask what someone's Cd calibration is if they go "rogue" on a build. I do not log it more of a mental reference on what is going on with the design.
Bruce

Who . . . me? I stayed at a Holiday in Express . . .
larrycavan
Posts: 235
Joined: Tue Mar 09, 2010 9:46 pm

Re: Modified EZ-Flow bench problem

Post by larrycavan »

Brucepts wrote:
larrycavan wrote:Square root scales are laid out based on the rise. Change the rise and your scale graduations are thrown off.
Might want to go back and look at the math on this? Once you set the length of the water gauge your percent scale is set based on that length, changing the rise merely changes the angle of the percent scale it does not change the percent scale markings across the length of the scale. If you change the length of the scale then it will change those markings.
With all due respect to Ed [because his orifice flow calculation worksheet is priceless]... I disagree with the assessment of laying out a square root scale as you describe.

You lay out a square root scale by rise. Rise determines the length. Make the rise higher, the angled line has to be extended to reach the intersect of the resting plane on the same angle. The scale needs to be done over.

Either the length of the angled line or the angle of that line has to change. Either change invalidates the original scale.

That's the way an engineer showed me how to do it in 1994. I then laid out my scale using Cadkey on my pc. The math, as I understand it is related to the sin of the angle. So, for an inclined angel of 30 degrees you would have [1 / sin 30 = 2] or 1" on the vertical is 2" on the inclined.

Changing the angle to something other than 30 degrees, as you would if you changed the rise, forces the math recalculation as well. If that change was to an angle of 45 degrees then the math would be 1 Inch on the Vertical would be 1 / sin 40 or 1 /.707. Making 1 inch on the vertical equal 1.414 on the inclined.

Anyway, that's my story and I'm sticking to it unless proven wrong...in which case I will stand corrected.
Last edited by larrycavan on Sat Dec 26, 2015 1:36 am, edited 2 times in total.
Larry C

http://www.cavanaughracing.com
Tony
Posts: 1438
Joined: Wed Mar 03, 2010 3:40 pm
Location: Melbourne, Australia

Re: Modified EZ-Flow bench problem

Post by Tony »

This is not correct !

Ten inches rise could be measured on a ten inch scale vertical manometer.
Only disadvantage of doing that is the graduations would be extremely close together and difficult to read.

Or the scale could be a mile long and still have a ten inch rise at the far end.
The angle of slope is irrelevant.
If you have ten inches of pressure it will lift the water ten inches, no matter how long the manometer tube is.

You make your scale any length you want, mark off the square law divisions, and you are set to go.
You then raise one end to give you the required amount of rise.
Also known as the infamous "Warpspeed" on some other Forums.
Post Reply