Solving equations approximated simultaneously

[egu]

Hi, I started using Mathcad a few days ago, and I´m trying to learn how to made some calculations for a paper. Right now, I´m having a problem to make an approximation with three different equations.

Before, I made this same calculation with excel, but I dont know how to make it now.

In this case, I have to calculate a velocity, and for that, I need a value, calculated by another depending of the velocity.

So, before, on excel, I started with an approximate number for the velocity, then Ifigured the constants, then with that, I obtained the velocity again, and so on until the value of the speed began to tend to a very similar number varying in a few thousandths of m/s.

Now, I´m not sure to do that step with mathcad. So I´m asking here for some help

What I need to is the velocity "uf", and  with that velocity, the constant to calculate are Rep2 and Cd2, the others variables are defined.

I think twith an itteration I could resolve this, but I don´t know how to do it yet

I hope I was clear.

Thanks

Emiliano

[ValeryOchkov]

1. Send please not only pictures but the sheet too!

2. It seems that I have solved this of same task.

See please pictures and sheets 17.11-17.14 here

Study 17. Cogeneration (CHP), trigeneration (CCHP) and quadrogeneration (CCHPI) or how much of mathematic is contain…

[AlanStevens · Accepted solution]

Look up Solve block in Help.  Here's a solution.  I only have Prime 2, so I've shown an image as well (this should also work in Prime 3.1 or whatever version of Prime you have).

Alan

[egu]

Thank you to both for your quick answer!

Were very interesting the examples of Valery. And with Alan program I was able to continue correctly.

May be It was very simple, but I don´t have too much experience with Mathcad.

I take this opportunity to make you another questions.

When Alan defined cd, what does it means NaN? and at the beginning when the values were define, there are any difference if they were another number different of one (except of zero)?

Thanks!

Emiliano

[Werner_E]

When Alan defined cd, what does it means NaN?

NaN means "Not-a-Number" and is the "result" you get, if Rep2 > 200000.

and at the beginning when the values were define, there are any difference if they were another number different of one (except of zero)?

You mean the guess values? Give it a try.

It depends upon the nature of the equations you solve.

Sometimes a system is very stable and you come to the very same solutions no matter what the guess values are.

Sometimes changing the guess values means you get a different (still valid) solution .

And sometimes using a badly chosen guess makes the solve block fail finding no solution at all.

WE

[egu]

Sorry to keep bothering, but I just had a little problem with the program. I put all in the same way than you on my program, but I cant get Cd2. It keep on telling me that the variable is undefined.

Even I tried to copy the same block that Alan attached before and I didn't have lucky.

Any Idea?

[AlanStevens]

The reason for including NaN as a default value (non-value!) for cd is that your original program didn't cover all possibilities.  For example it didn't cover what happens if Rep2 were to be larger than 2*10^5.  You could replace this with whatever default value you thought was appropriate.

I'm afraid I can't tell what the problem is with your version of the solve block as you are using a version of Prime that I can't open.

Alan

[egu]

Alan, it was just a mistake in the program. I wrote in the cd condition only < and > and no "minor and equal"

Since I started to use mathcad I made a lot of that

Thank you for the help to all

Emiliano

[Werner_E]

The Cd2 argument in the Find() has a wrong label ("system"). Set it to "variable".

Next problem: Unit mismatch. Guess value Cd2 probably should not be just 1 but 1 second.

Next problem: No solution found.

have not investigated further.

WE

[AlanStevens]

"Next problem: Unit mismatch. Guess value Cd2 probably should not be just 1 but 1 second."

No, Cd2  should be dimensionless.

Alan

[Werner_E]

So it was a different unit problem which egu seems to have found out himself in the meanwhile.