New Member - have question about flow

[superlen]

Tony,

Thanks for the tip on excessive pre-filtering. I have filters in their for my ECU signals (including the one channel I'll use for the DP measurement), but I'll make sure to check the response and back off the filter to make sure I don't end up losing some valid samples. My sensor samples should be here in a few days, but I also have a new pick and place machine showing up at work Thursday so sadly, I may not have time to play with them when the arrive.

Lenny

[Tony]

If you end up having a bouncy pressure reading even after averaging, it is usually due to airflow instability in the bench. It could be blower surge, or something like vortex shedding within the bench.
The best solution is always to cure the problem right at the source, rather than trying to mask it by slowing down the response either in the air lines or the analog section.
You might be surprised how noisy the pressure samples are, but don't worry about it, unless you can see some definite low frequency rythmic surging in the data.

[superlen]

I wondered about resonances showing up at different flow rates. Has anyone seen this? In particular my bench is slightly smaller than the stock plan so resonances or standing waves or some other phenomenon could pop up. I hadn't thought about motor surge, but I could easily see that happening particularly if one was trying to flow right up to the limit of the motors top end. All kinds of related variables could affect this, motor choice, motor controller electronics, box dimensions, orifice size, cfm being flowed, depression needed, ect.

Someone surely must have modeled this in SolidWorks or a similar airflow simulator to play around with what if scenarios. I think simulations often are flawed when trying to get exact numbers on things, but generally can give someone a good feel of how inputs affect outputs. I have no experience in airflow simulation whatsoever, but always thought it was a cool science to understand.

I don't plan on flowing that much air and the air I do flow will be at a tiny depression since the restriction in the AFM is almost nill. Hopefully I don't find some weird dynamic to contend with. My goal is to shoot for the utmost in accuracy. All my fuel delivery tables programmed into the ECU will be based on what I learn from testing these AFMs & I will be calibrating the AFMs back to factory stock using this bench.

Len

[Tony]

I will be very surprised if you have any flow stability problems.
The only reason I mentioned it at all was to make you aware of the difference between the expected high frequency random noise fluctuations, and some quite low frequency cyclic disturbances which would be abnormal.

[Brucepts]

From another post . . . . this

I kept waffling back and forth on assembly as well from "this needs to all be smooth and symmetrical and perfect" to "That can't matter, just gob some glue on, caulk it and put the screws to it".

Thanks again,

Lenny

Trust me on this one the PTS design has been built in various stages of quality and they all end up flowing around the same Cd

[jfholm]

"That can't matter, just gob some glue on, caulk it and put the screws to it".

I think Bruce is referring to mine - I really threw mine together just to see if it worked. I even used melamine so I would not have to paint the inside. Works great! Bruce sent me my first test orifice and he tested it before he sent it and on my bench it only flowed .2 cfm different than he flowed it. That is two tenths of a cfm.

John

[superlen]

Sweet!

I should have built mine even crappier. LOL I did see your bench John. You're right it's not going to win any beauty awards, but It's hard to argue with success.

Len

[jfholm]

Sweet!

I should have built mine even crappier. LOL I did see your bench John. You're right it's not going to win any beauty awards, but It's hard to argue with success.

Len

I gave it to a friend since I am building a new 8 motor bench that I hope will look much better. He was talking about putting some sort of paneling etc on the old bench. Which just goes to prove you can polish a turd.

[superlen]

LOL..

Tell him to staple green shag all over it and complete the 70s retro look.