cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - New to the community? Learn how to post a question and get help from PTC and industry experts! X

One task - how can I use not rkfixed but odesolve?

ValeryOchkov
24-Ruby IV

One task - how can I use not rkfixed but odesolve?

A Task.

A steel core thrown vertically upward at a speed v0 and fell to the ground in 15 seconds. What is the coefficient of air resistance in the equation Ffriction = k S Ro v ^ 2? [The answer k=0.877 ]

I have a solution with the using the rkfixed function (Mathcad Prime with UNITS!- see the attach). In Mathcad 15 we can not use units in this task

1. Is the solution correct?

2. How can I use not rkfixed but odesolve function?

3. I'll take the other comments and suggestions gratefully!

t15.png

1 ACCEPTED SOLUTION

Accepted Solutions

AlanStevens wrote:

Nearest I can manage.

Alan

Thanks for the idea.

I think it is a solution:

t15-d.png

View solution in original post

25 REPLIES 25

Problems with odesolve - see the attach.

Ode solved?

Problem may be redefinition of 'm'

Regards

Andy

A Westerman wrote:

Ode solved?

Problem may be redefinition of 'm'

Regards

Andy

Thanks!

But:

t15blue.png

Problem may be redefinition of 'm'

No, not quite. Look at the Labels of y within Valery's Solve block. They are labelled "function" but should be labelled "variable". Also the "t" in "y(t)" in the argument within odesolve has label "none", but should have label "variable". I find this sort of thing occurs all too easily within Prime, making it still a toy not a real tool.

Meanwhile, the attached shows a way of solving Valery's original problem using odesolve.

Alan

Capture.PNG

Missed that the original had the issue of multiple uses of variable names - definitely a serious weakness in Prime.

But T15 (2) was re-written from 1st principles & lost that in translation.

regards

Andy

A Westerman wrote:

Missed that the original had the issue of multiple uses of variable names - definitely a serious weakness in Prime.

But T15 (2) was re-written from 1st principles & lost that in translation.

regards

Andy

Yes, I often find problems in Prime can be fixed by rewriting from scratch!

Incidentally, Valery's problem can also be fixed by labelling all the y's and all the t's within his solve block as "none"!

Alan

I just saw this thread. In Valery's original file, he used "m" both as a variable and as the unit "m".

He's able to do this with labels. The variable "m" has the label variable, and the unit "m" has the label "m", which you can see by the default blue color.

Valery.png

I'm logging the issue that Alan raised below with some of the labels showing the wrong color.

Mona

Mona Zeftel wrote:

I just saw this thread. In Valery's original file, he used "m" both as a variable and as the unit "m".

He's able to do this with labels. The variable "m" has the label variable, and the unit "m" has the label "m", which you can see by the default blue color.

Valery.png

I'm logging the issue that Alan raised below with some of the labels showing the wrong color.

Mona

Mona,

see please http://communities.ptc.com/message/190187

I think it is not correct (normal) with lables in the odesolve-block in Mathcad Prime and will be fine in Prime 3 or N.

Mona Zeftel wrote:

I just saw this thread. In Valery's original file, he used "m" both as a variable and as the unit "m".

He's able to do this with labels. The variable "m" has the label variable, and the unit "m" has the label "m", which you can see by the default blue color.

Valery.png

I'm logging the issue that Alan raised below with some of the labels showing the wrong color.

Mona

Sorry Mona! But we say in same cases in Russia - you should at first "eat peas" before using the function odelolve with lables in Mathcad Prime

AlanStevens wrote:

Problem may be redefinition of 'm'

I find this sort of thing occurs all too easily within Prime, making it still a toy not a real tool.

Alan

I too!

Mona!

Do you read this?

And second.

I can create the function y(t, k) by the using the fkfiksed function but can not with the odesolve one.

Somebody can?

However we can use the odesolve! With grief in half! Pardon, with the bisection method!

See

t15-odesolve.png

Valery Ochkov wrote:

However we can use the odesolve! With grief in half! Pardon, with the bisection method!

Did you see my file t15b above Valery? That uses the bisection method (automatically, not by hand).

Alan

AlanStevens wrote:

Valery Ochkov wrote:

However we can use the odesolve! With grief in half! Pardon, with the bisection method!

Did you see my file t15b above Valery? That uses the bisection method (automatically, not by hand).

Alan

Thanks, Alan!

But it is very difficult!

I would like to have a function y(t,k) as in the sheet with rkfixed!

As in Mathcad Prime Minister - http://communities.ptc.com/docs/DOC-2999

Valery Ochkov wrote:

I would like to have a function y(t,k) as in the sheet with rkfixed!

I agree. It is frustrating not to have this capability.

Alan

Nearest I can manage.

Alan

AlanStevens wrote:

Nearest I can manage.

Alan

Thanks for the idea.

I think it is a solution:

t15-d.png

So too:

t15-e.png

A last variant (?):

t15odesolve-all.png

A video (an idea) of this task is here

Eureka! I thought of Archimedes!

Eureka.png

So Mathcad calculates faster, but we do not understand how to get this equation:

Eureka-Simplify.png

Valery Ochkov wrote:

So Mathcad calculates faster, but we do not understand how to get this equation:

Eureka-Simplify.png

Haven't you got a y'(t) too many here Valery? i.e. shouldn't y'(t)^2.sign(y'(t)) be just y'(t).sign(y'(t))?

Alan

AlanStevens wrote:

Valery Ochkov wrote:

So Mathcad calculates faster, but we do not understand how to get this equation:

Eureka-Simplify.png

Haven't you got a y'(t) too many here Valery? i.e. shouldn't y'(t)^2.sign(y'(t)) be just y'(t).sign(y'(t))?

Alan

Sorry I do not understand your question!

y'^2.sign(y') is equal y'*|y'| - the vector of the friction force is opposite the vector of velosity!

But what is fester?

Oops! I read sign(y'(t)) as absolute value of y'(t)!

But I don't understand what you are asking. Is what faster than what?

Alan

AlanStevens wrote:

But I don't understand what you are asking. Is what faster than what?

Alan

What is fester? y'^2 * sign(y') or y' * |y'| ?

Top Tags