Solve block fails in function

[aprasun]

Hi all

while solving for a variable in a function calling solve block, I am getting massage that " either CTOL , TOL or guess values shall be modified". I am attaching my mathcad 11 sheet and attached excel file for seeding the variables.

thanks in advance.


Prasun

[ptc-1368288]

"Solve block fails in function"
_______________________________

Because the solve block is undefined !
and there are surely many more things not working.
"Unit Weight Water = 9810kg/[m�*s�]"
I have hard time to see that !
Modofied modulus of elasticity = 230769230769.23PaMPa
Don't understand that one either.
Why an Excel file in a Mathcad work sheet ?

...

In addition to understanding the Given/Find from the qs, you can Google for 1000's examples in this collab, as well as using the unit system before entering a long program which seems a home made collection of compacted modules from some engineering book or unknown sources

jmG

[ptc-1368288]

... also,
you have a 16.7 millions colors for a 574Kb drawing that is not even readable ! Just make a little drawing from Paint, reduced to 8 colors or even 2 colors for a negligible size that can be read.

jmG

[TomGutman]

YOu're flying blind. You are trying to solve a cubic without plotting it. Therefore you have no way of knowing whether there are, in fact, any solutions close to your guess value, or even within the range of interest.

In fact, your cubic is essentially linear and has a solution that is fairly large and negative.

You also have a scale problem. The values for your cubic are huge. Large enough (in the default units) that for i=1 it is not possible to reduce its value to be within CTOL.
__________________
� � � � Tom Gutman

[aprasun]

Thanks Tom

I could modify your file and finally got correct result.
Final version is attached herewith

prasun

[TomGutman]

I would like to note the difference in the size of the file I posted and the one you just posted.

You seem to be working with very different data, with now only one data set actually involving the cubic.

You still have the scale problem. You are just being very lucky that you are getting to an exact zero for the relevant data.

The cubic has clearly changed (I haven't tried to figure out exactly where). It now shows two real roots in the plotted region. Since it is a cubic (I assume that the leading coefficient is not zero, although I haven't checked that) it must have a third real root somewhere, beyond the plot limits. You need to be sure that the root you are getting is the relevant root.
__________________
� � � � Tom Gutman

[aprasun]

only one difference is there. the sign of K3 in given find block was passed on as +ve, but actually it was -Ve. my mistake !



Prasun

[TomGutman]

Ah, the sign bites again. It is extremely easy to make a mistake in a sign when entering or copying (with or without modifications) expressions. There are many examples of such errors in this collaboratory. That is why I look to avoid retyping of expressions or manual manipulations of expressions, and try to have Mathcad do whatever manipulations are necessary.
__________________
� � � � Tom Gutman

[aprasun]

thanks and noted down.

prasun