Unable to close or Exit "Distributed Mechanical APDL" Dialog Box

[vnair7317]

Hi,

I have written a script to solve a simple matrix using "ansys.math.core.math". However, when I execute the script, it opens a Mechanical APDL dialog box that does not close automatically after the execution completes. As a result, I have to manually terminate it each time using the Task Manager.
Could you please suggest how I can modify the script or settings so that the dialog box closes automatically once the execution is finished?
Thank you for your help.

Matrix_solver.txt](https://github.com/user-attachments/files/28143773/Matrix_solver.txt)

[RobPasMue]

@germa89 @clatapie could you have a look at this?

[vnair7317]

Any Updates on this @germa89 @clatapie ?

[RobPasMue]

Extracting content from txt in plain sight

import numpy as np
import scipy.sparse as sp
import ansys.math.core.math as pymath

print("--- Starting PyAnsys Math Verification ---")

# 1. Launch the backend engine
try:
    sol_mat = pymath.AnsMath(port = 50053)
    print("Successfully connected to the Ansys Math backend.")
except Exception as e:
    print(f"Initialization failed. Check your Ansys license.\nError: {e}")
    exit()

# 2. Build a 3x3 sparse matrix R using SciPy
#R_scipy = sp.csr_matrix([[1, 2, 3], [4, 5, 6], [7, 8, 8]], dtype=np.float64)
R_scipy = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 8]], dtype=np.float64)
# 3. Build a 3x1 vector Q using NumPy
Q_numpy = np.array([5, 1, 0], dtype=np.float64)
print(f"Created R shape: {R_scipy.shape} | Created Q shape: {Q_numpy.shape}")

# 4. Stream data to Ansys Workspace
R_ansys = sol_mat.matrix(R_scipy, name="R_matrix")
Q_ansys = sol_mat.set_vec(Q_numpy, name="Q_vector") 
X_ansys = sol_mat.vec(size=3, name="X_sol") 

# 5. Direct Factorization and Solving
print("Executing sparse matrix factorization...")
solver = sol_mat.factorize(R_ansys)

print("Solving Rx = Q via Ansys solver backend...")
solver.solve(Q_ansys, X_ansys)

# 6. Extract solution back to Python memory
final_x_vector = X_ansys.asarray()

print("\n--- SOLUTION ---")
print("Target solution vector x:")
print(final_x_vector)

# CORRECTED VALIDATION: The mathematical answer to inv(R) * Q is [-10, 9, -3]
expected_x = np.array([-10.0, 9.0, -3.0])
is_correct = np.allclose(final_x_vector.flatten(), expected_x, atol=1e-5)
print(f"\nIs the solution correct? {is_correct}")

# 7. Corrected Cleanup Routine
# Use the official mapdl reference without an underscore prefix

sol_mat._mapdl.exit()
print("\nAnsys workspace cleared successfully.")

[germa89]

This is duplicated from #4596.

Transfering to PyMAPDL repository.