Workaround - Different Units in Find

[Andrewk2]

Trying to use solve block (Given ... Find) to solve a flow network. I'm using units and trying to use mixed units in the Find statement (both pressure and flow). Mathcad complains.

Is there a quick work-around to getting the solve block to work without doing something drastic?

See attached Word document for my Find statement and a "dirty" work-around I tried, but didn't work.

[IRstuff]

The actual worksheet might be a bit more useful than just a picture of a chunk of it.


TTFN,
Eden

[ptc-1368288]

That's not the way to work especially for this kind of engineering where flow sizing is already engineered in the appropriate system of units.
Make a diagram, abstract, give some values in the work sheet. Pump/pipe sizing and valve sizing is an iterative engineering process.
Example: after my valve sizing, a 3" valve in a 12 " pipe indicates wrong pipe sizing [oversize]. Some pipe material are very expensive. I have the Mathcad valve sizing procedure, but as you put it ... unlikely it will agree.

"Save as" Mathcad 11

jmG

[Andrewk2]

It's in Mathcad 6 Standard (no jokes, please). It's got some strange statements because of Mathcad 6's temperature "issues" and "Standard" doesn't have the programming pallette.

The pump flow depends on the back-pressure. Back-pressure depends on flow. There's a regulating valve that controls the flow distribution in the system based on the pressure across the valve. Since the flow and pressure are both dependent variables of each other, it seemed like a solve block was the best way to get a solution.

If you get it to work, please post solutions in PDF.

[AlvaroDíaz]

On 4/5/2010 5:28:39 PM, andrewk8 wrote:
>It's in Mathcad 6 Standard (no
>jokes, please).

Added only a Fahrenheit2 function and used only one time.

Regards. Alvaro.

[Andrewk2]

Thanks for catching that I was already passing unitless-temperature into the function and didn't need to convert it, again. Fixing that problem didn't work, though.

I significantly simplified the problem to try to ONLY balance pump flow with discharge pressure. See the two new functions SimpleNetwork1 and SimpleNetwork2 at the bottom of the worksheet.

With SimpleNetwork1, I get "error in list". With SimpleNetwork2, I get "incompatible units"

[ptc-1368288]

On 4/6/2010 9:23:37 AM, andrewk8 wrote:
>Thanks for catching that I was
>already passing
>unitless-temperature into the
>function and didn't need to
>convert it, again. Fixing
>that problem didn't work,
>though.
>
>I significantly simplified the
>problem to try to ONLY balance
>pump flow with discharge
>pressure. See the two new
>functions SimpleNetwork1 and
>SimpleNetwork2 at the bottom
>of the worksheet.
>
>With SimpleNetwork1, I get
>"error in list". With
>SimpleNetwork2, I get
>"incompatible units"
_____________________________

Conversion to Mathcad 11 gives different error messages "The units in the expression do not match". An Engineering piece of work needs an abstract and a diagram. Not sure what you are doing ... maybe "sizing" a design around a constant pressure discharge via a bypass valve to a pump ? All what you have does not go by valve sizing, neither by pump sizing. Pump sizing goes by pump curve and min/max operating flow. Valve sizing goes by operating min/max flow and select a valve for the rated installed flow rangeability. You have "valve slope", what is that ? You have centistokes and we use cp [centiPoise]. What fluid is it ? ...etc.

Here are two sheets:
1. Colebrook subject to minor revisions
2. Valve sizing procedure.

That's all what is needed for a large plant with 1000's valves.

I understand "spec" for "design = operating = nominal"

Resume: abstract + diagram

jmG

[Andrewk2]

Mathcad supports different units in a solve block Find() statement (see attached), however, it appears that the ability falls apart when you are using a solve block to define a function. My guess is that the solve block is returning an internal array, and Mathcad can't assign the result to your result because of the different units.

My work-around is to strip the units out of the variables in the Find() statement. In doing so, you have to make sure you are using appropriate units inside the solve block.

Attached is my work-around (go to the bottom of the worksheet). It shows the expected result. Because of back-puressure, flow from a "real" pump is less than a pump with 100% efficiency. Consequently, with lower pump output, system pressure is lower than theoretical.

It's "clunky", but it seems to do the trick. If someone has an alternate solution, I'd be curious.

[RichardJ]

On 4/6/2010 1:49:55 PM, andrewk8 wrote:

>It's "clunky", but it seems to
>do the trick. If someone has
>an alternate solution, I'd be
>curious.

Probably not, but it's worth a try. At some point between MC6 and MC11 they changed it so that what you tried does in fact work. Then in 12 they broke it again, so that it no longer works. There is another approach that now does work, but it doesn't work in MC11, and therefore probably no in MC6 either.

Richard

[ValeryOchkov]

Mathcad 2000



Mathcad 11



Mathcad 14


Val
http://twt.mpei.ac.ru/ochkov/v_ochkov.htm

[Al2000]

Mathcad 2001



Saludos,

Al