Radiators and Fans

[Since87]

Originally posted by Les
Would have helped if I had looked at the photos of the radiators.
Yes the Big Momma cannot be 1 Pass, and Rad E cannot be 2 Pass(could be 3,5,7 etc)
. For the record the difference between 1Pass and 2Pass on Big Momma .:-

LOL, was wondering.

Good point on Rad E. Three or five passes may explain the apparently high resistance. (seven or more passes seems odd to me with a 30 tube rad, but could be 4,4,4,4,4,4,6.)

I've been playing around with numbers towards adding a 'barb ID factor' to the spreadsheet. I'm not familiar with the correct equations for this stuff. (Involves someone named Darcy right?)

I've just been fiddling around in Excel trying to match Bill's 2 gpm data from the chart here. The equation:

dP = .00156 * ID^-5

gives the match shown here:



I like my equation. It's simple and hits half the points dead on, unfortunately, I have no idea whether there is any justification for it in fluid dynamics.

Les, would you mind detailing how you applied the correction for Big Momma's barbs? Or, rework the spreadsheet to take a barb ID as user input?

Do you think a 'barb factor' should replace the 'overhead factor' completely or only partially?

[Les]

Originally posted by Since87
LOL, was wondering.

1)I've been playing around with numbers towards adding a 'barb ID factor' to the spreadsheet. I'm not familiar with the correct equations for this stuff. (Involves someone named Darcy right?)

I've just been fiddling around in Excel trying to match Bill's 2 gpm data from the chart here. The equation:

dP = .00156 * ID^-5

gives the match shown here:



I like my equation. It's simple and hits half the points dead on, unfortunately, I have no idea whether there is any justification for it in fluid dynamics.

2) Les, would you mind detailing how you applied the correction for Big Momma's barbs? Or, rework the spreadsheet to take a barb ID as user input?

3) Do you think a 'barb factor' should replace the 'overhead factor' completely or only partially?

1) Think I prefer (From visual inspection of same Billa data)

From distant memory think fourth power has some justification in fluid dynamics( have to check)

2) Again from visual inspection of same Billa data.
Difference at 2gpm between 2x( 0.488"ID) and 2x(0.3" ID) = 1.3 psi = 8.96KPa.
Did rough convert to ~ 0.896m(H20) should be 0.9149m(H2O).

Simply used the 0.015972. Have been using this for the "Cathar Big Arse" v "Big Momma" and has been giving results.
Whether correct is another question - eg using 2 Barb correction (and is Billa data single Barb - think it is but not sure).

3) Ugh and dunno. I am possibly the last person to ask for conciseness/simplification.Think prob wait for more data with which to play(torture).

EDIT1: Deleted inane comment.

EDIT2: Corrected explanation of arrival at. 0.015972 "fudge factor". Used 0.3"(not 0.25" or 0.25- 0.3") as ID of 3/8" Barb. Excuse is it is a long time since did it.

[Les]

Edited previous post.
Used 0.3"(not 0.25" or 0.25- 0.3") as ID of 3/8" Barb. Excuse is it is a long time since did this work.
Am making hard work of this.

[BillA]

sorry for the delay, bit too much on my plate





data being sent to Les and Since87

[Les]

Originally posted by unregistered
sorry for the delay, bit too much on my plat

data being sent to Les and Since87

Delay? . Was not expecting anything.
Data received.Thanks.
Before I dig too deep a hole , am I understanding correctly? :- The differential is the PD differences between radiators fitted with two(2) specimen Barbs(entry and exit) and two(2) reference 12.83mm ID Barbs.

Les

EDIT: Corrected 0.583" ID to 12.83mmID . See Unregistered below.

[Roscal]

BillA > Hi , do you have some news about testing the rad in the two way possible?

[BillA]

no Les
the ref connectors are a pair of 3/8"NPT x 5/8" barb (in brass) with an ID of 12.83mm/0.505in.

this is something of a gap between the 1/2"CTS (5/8" OD copper pipe) which has an ID of 0.572"
- or thereabouts, depends on the schedule as it is OD based

the only difference in pressure drop between 0.468" and 0.505" was 0.01psi @ 3gpm
would suggest that going to 0.572" ID is not huge, but something rather small

I will dig through my old data for a correlation

Roscal
will try, but am a bit pressed for time (for reasons which will be known shortly)

[Les]

Originally posted by unregistered
no Les
the ref connectors are a pair of 3/8"NPT x 5/8" barb (in brass) with an ID of 12.83mm/0.505in.

Typo (re 0.583"ID Barbs, 12.83mm).
I stay in the kitchen,maybe only for the spinach.

[Les]

Originally posted by unregistered
.......
data being sent to Les and Since87

The usual Dog's Dinner.
Probably only a selfish indulgence and completely useless for "The Simulator"
The best fits with Simulated Data from "SF Pressure Drop5" *:-



Have been using the "Green" simulations in previous graphs.
Maybe OK for small ID Differences( <2mm).

* http://www.pressure-drop.com/

[BillA]

and here the real problem of the measurand being smaller than the rounding error/uncertainity becomes apparent
0.01psi (using a scaling indicator with offset) is too large an increment for this work

an omitted 'note' (sorry Les)
all connector inlets and outlets were chamfered
- am not suggesting that a radius be defined, typically a slope with rounded edges

[BillA]

Roscal
the last test is yours
reversing the inlet/outlets on a Chevette core at 2gpm yielded a (very nominal) 0.005psi (~ 3 or 4cm H2O) difference
perhaps of interest to the analytically inclined, not even slightly to a user
(if you knew which direction was better it would not be worth crossing the hoses)

[Since87]

Some further torturing of Bill's latest data...

I took a subset of the data ( ID's from 6.1mm to 11.18 mm, and flowrates from 3.783 lpm to 11.35 lpm) and graphed dP vs ID for the four flow rates included.

I used log scale for the dP axis because it can clarify (exaggerate?) patterns in the data.



Then I 'normalized' the four curves by dividing out flowrate^2 for each.



The black line I 'painted' in (actually something a bit less sloppy) is what I was inclined to use as a barb factor.

But I really disliked that 8.89mm ID number, so I just plopped 9.89 mm in instead. I really liked the results with 9.89, but I decided to take a look at the Imperial measurement and saw that it was 0.375". (or 9.525mm)

Here's the second graph redone with 9.525mm in place of 8.89mm.



I'm much happier now.

Too late at night for me to be coming up with an equation though.

Les,

How do you generate those polynomials from a set of data points? Is that something Excel can do?

[pHaestus]

click on data in the graph to select. Right click, add trendline, choose the type of line to fit. Then in the options check "show equation" and "show r^2" or something like that

[Roscal]

BillA > Thank you for the test . Which way was better :enter small tank or big tank??

It's a pity that difference between the 2 tanks of the chevette rad was so small , much difference should probably appear in a rad like that (at left) :

[Since87]

Just a quickie.

For each ID, I averaged the dP/Q^2 value for the four flowrates I'd graphed earlier to get a single data set.

[Since87]

I've played around with my experimentation spreadsheet to incorporate Bill's barb data.

The latest simulated vs measured comparison:



I assumed 0.25" ID barbs for the Big Momma.

I've tweaked all the fudge factors a bit.

There is still an overhead factor, but it is only 61% of its former value.

I will update the user friendly spreadsheet tomorrow and make it available for download.

[bigben2k]

If you're interested, here are my results, for Fedco# 2-304:

dP of heatercore @ 0.5 gpm: 1.5" (0.054 psi)
dP of heatercore @ 1.0 gpm: 4.0" (0.14 psi)
dP of heatercore @ 1.5 gpm: 8.5" (0.31 psi)
dP of heatercore @ 3.5 gpm: estimated, 4.48 psi

Not terribly accurate, procedure described here.

[Since87]

Ok, I finally got the new version of the spreadsheet done.

A (color distorted) screenshot:



The main changes are:

Accounting for inlet and outlet ID in determining the pressure drop.

More detailed usage notes, and some background information on the spreadsheet.

The pie chart, which is handy for visualizing the impact of barb ID on the dP of the heatercore.