Rad Testing, conditions, assumptions, etc.

[BillA]

am in the midst of rad testing, and as usual I'm daily reminded of what miserable work this is
much of the difficulty has to do with maintaining steady state conditions, w/o which intelligible data cannot be collected

the question has been raised as to the suitability/appropriateness/necessity of the LMTD characterization of the rad's wet side temp, and upon reflection it would seem appropriate to make the same characterization of the dry side
- for this it would be necessary to know the mean velocity, atmospheric pressure, humidity, inlet and outlet temps, and the pressure drop - but this info requires very sophisticated equipment not at hand
-> the solution would seem to be the continued use of what is standard practice,
using the inlet air and inlet coolant temps to characterize the temp difference, "C"

opinions solicited

the heat dissipated is measured with a defined temp difference between the air and the coolant, this is an arbitrary amount as a very good WCing system with substantial radiator overcapacity could have a temperature differential of 5°C whereas a more typical system would see a 10°C rise in the coolant temp over ambient

this has caused a bit of confusion, my initial testing had a 10° differential, the ThermoChill rads were done with a 5° difference (presumably reflecting 'better' systems), and my present testing is again with a 10° difference (as does JoeC's)
- whatever the differential is, it causes extreme difficulty in testing as both the coolant and air temps respond differently yet must be maintained at a fixed interval
-> I am considering allowing the differential to 'float' within a range of ±1°, calculating the actual C/W, then back calculating the W dissipated with a 10°C differential

this will have the effect of removing the error of the temp differential measurement/control from the dissipation calculation

comments are solicited

the reason behind both of the above changes is the variation in air temps across the inlets of a dual fan rad, and while the phrase "in an environmental chamber" sounds impressive - achieving uniform temps across the dual inlets is not possible
- but knowing what the air temp actually is is not so difficult (4 RTDs, 1 in each quadrant, avg temp = inlet air temp for that fan; same setup for each fan)

one has to speculate as to how others control/quantify the inlet air temp
pH is going to be very unhappy, this is a vexing problem

[ToasterIQ2000]

Dear Mr. Unregistered,
Out here in Lake Woebegone, where us toasters like to vacation, I am set down today to say something about your testing of my in-laws, the Radiators.
I taught writing once upon a time, in a galaxy far far away, to students who were not uniformly prepared to stipulate to the phenomenolgical primacy of The Complete Sentence, or even to it's value as an elaborate and fragile artiface that may at times require effort and dilligence to will into being and persistance. Removing my lobes from my caudal sphincter...
The results that I imagine would be based on large scale sampling across varied real-world conditions, and would take the form of a three dimensional surface for any given radiator: X is wet side flow; Y is airflow; Z is delta t of water. Color of any given Z point indicates delta t between water in and air in for that given Z point. The independent variables are: water flowrate, X; air flowrate, Y; delta t between input temperatures, color. The one dependent variable is delta t of water-in to water-out, as Z. Alternatively, there might be a slider attached to the three dimensional surface to select only one particular delta t between input temperatures at a time, or a pair of sliders to select two points on the delta t between input temperatures independent axis concurrently.

[bigben2k]

Errr. what's the question? :D

I can appreciate the simpler approach, I'm just not sure which 2 changes you're proposing.

[ToasterIQ2000]

While my goofball persona is relatively deflated:
#3:
I didn't have 'a question' per say; though my sense that the grammar, punctuation, page layout, coompositional refinement & etc. of the original post is a step below that which I read in say the Journal of Neuroscience does imply questions.
And I wasn't prorosing two changes that I was aware of...

A big three Dimensional graph.
X is is the flow of water through the radiator.
Y is the flow of air through the radiator.
Z is the measure of water cooling. ( raw degrees or perhaps in watts to sync with X )
Color of Z indicates the difference between intake water temp and air temp.

Importantly, a subset of Z-colors can be chosen for display: say 5, 15 and 25 degrees in blue yellow and red, so as to avoid a huge mud collored cloud.

Set up a system with automatic pump and fan control and data logging. Set it to scroll water flow from min to max and back and forth on one time scale; scroll / adjust airflow independently on another time scale; adjust the water chiller controlling X temperature in on another scale; adjust / scroll / wander air temp in independently as well. Invite some uncontrolled & unmeasured variance: air pressure, humidity, set it near a window where sometimes it is in the sun and sometimes not.

Set this up, start the data logging, and let it run days or weeks. Dump the data into a database and render a few charts. Select out a Z-color of say 10 degrees and render it. Hopefully, it will show a big cloud of points in the 3D space. If the cloud is spikey, holed, or slabbed on X Y or Z, something is amiss: something important uncontrolled / unmeasured, independent variables not wandered properly, or data sensed / logged in error. If it is a nice fluffy pancake on a hill, dense on the inside and thin towards the edges, then it is a neat description of the cooling of a radiator across varied water & air flow & input temps.

Annyway, that is what I thought of on reading Mr. Grumpy's post. Feel free to take it as a non-sequitor, just some tangent that I thought of... I like that he shares what might be more rambling and perhaps less than polished notes / thoughts in progess here. I wouldn't want to be seen as 'complaining' in that regard.