Skip to main content
regcurry
regcurry16-Pearl
16-Pearl
Oct 19, 2026 3:23 AM UTC
Solved

Large matrix in solve block i Mathcad 15

Attached is a zip file containing a large matrix in a solve block.

There are two questions in yellow boxes.  Any help is appreciated.

The jpg file shows the geometry of the diffusion equation problem.

terryhendicott
terryhendicottAnswer
21-Topaz II
Oct 19, 2026 4:42 AM UTC

Three Problems

Solve block needs to be a matrix equation the way you have set it up.  I would write them as five separate equations?

Problem two is that the matrix elements have mixed units and this is not permitted in a Mathcad 15 matrix (It is in Prime).  Two rows are unitless. Three rows are 1//m.  See enclosed file

Third five equations in a solve block you would normally search for five unknowns not one?

 

Cheers

Terry

View in thread

4 replies

terryhendicott
terryhendicott21-Topaz IIAnswer
21-Topaz II
Oct 19, 2026 4:42 AM UTC

Three Problems

Solve block needs to be a matrix equation the way you have set it up.  I would write them as five separate equations?

Problem two is that the matrix elements have mixed units and this is not permitted in a Mathcad 15 matrix (It is in Prime).  Two rows are unitless. Three rows are 1//m.  See enclosed file

Third five equations in a solve block you would normally search for five unknowns not one?

 

Cheers

Terry

Translate
    regcurry
    regcurry16-PearlAuthor
    16-Pearl
    Oct 19, 2026 6:15 AM UTC

    Terry,

    Thanks much. You are always a big help.

    The objective is to eliminate the unknown coefficients C, A,B,D,G and use the resulting transcendental equation to fine the values of 𝛼 that satisfy the transcendental equation.

    You helped me 4 months ago with a similar problem for 2 layers and three unknown coefficients instead of 3 layers and five unknown coefficients.

    https://community.ptc.com/mathcad-7/help-with-mathcad-15-solve-block-174320

    Although you had everything unitless, when I used your formalism in my worksheet, I included 1/lenth for the 𝜆s and length for the Rs. That made K1 having  units and K2 no units, I think.🤞  The solve block worked.

    So, if I write it as five separate equations, does that mean every element in each equation can have both units and no units?

    Translate
    regcurry
    regcurry16-PearlAuthor
    16-Pearl
    Oct 19, 2026 4:03 PM UTC

    Mea culpa 

    ‘The 2 layer equations in the solve block are unitless.
     

    Also, the matrix approach to arriving at the correct transcendental equation to solve for 𝛼  in the solve block for the three layer problem is incorrect.  It must be done the way it was done in the 2 layer problem.  It’s a lot more algebra for the three layer problem.  I am using MATHCAD’s symbolic solve→ and simplify→ functions as I did for the 2 layer problem.  When I get it working, I will post it for those with natural curiosity.

    Translate
    regcurry
    regcurry16-PearlAuthor
    16-Pearl
    Oct 25, 2026 8:02 AM UTC

    Here is the promised MATHCAD 15 file for the development of the transcendental equation to solve for 𝛼, the neutron growth rate in an exploding fission weapon.  It extends the model mentioned above in my reply to Terry to include three spherical layers.  It adds a neutron reflector between the fissile core and the heavy tamper.  𝛼 is necessary in the calculation of weapon yield.

    I use this to demonstrate that 90% Highly Enriched Uranium is not necessary for a nuclear weapon of significant yield, as erroneously implied by the media and government “experts,” who should know better. If you are interested in the results of my analysis, check out my YouTube video.
     

    Possible Nuclear Weapons

     

     

    Translate
    Terms & ConditionsAccessibility statement