P-little-delta Improvements

v0.0.86

Sign convention update

Removed P-little-delta Effects from Linear Analysis Moment Diagrams

• Changed member moment diagrams to no longer include P-little-delta effects in non-P-Delta analyses. This leads to consistent moment diagrams across members for simple analysis methods. Before, even when P-Delta effects weren't being calculated, P-little-delta effects were included in member moment diagrams, leading to member moment diagrams that didn't quite match up across joints on non-P-Delta models with significant P-Delta effects.

Bug Fix for P-Delta Member Internal Results

• Fixed a bug in P-Delta analysis member moment calculations. Global results were correct, but member internal moments were neglecting the P-little delta effect. The result was correct moments at nodal locations, but incorrect results in between. This issue has been resolved.

Bug Fix for P-Delta Member Internal Results

• Fixed a bug in P-Delta analysis member moment calculations. Global results were correct, but member internal moments were neglecting the P-little delta effect. The result was correct moments at nodal locations, but incorrect results in between. This issue has been resolved.

Performance Improvements

• Sections have been introduced to allow for member stresses to be tracked by the program during nonlinear analysis in the future. This opens the door for other useful features down the line too. Use of Sections is optional, and will only required for pushover analysis when that feature is implemented. You do not need to use this feature yet.
• P-Delta analysis code has been greatly simplified. Performance has also been improved, as redundant iterations are no longer being performed.
• Removed the tol parameter as it is unecessary when using a geometric stiffness matrix instead of an iterative procedure.
• Better documentation for P-Delta analysis.
• Corrections to unit tests that weren't working properly. Added another AISC Benchmark unit test for P-Delta analysis. This should help safeguard the program against some types of bugs being introduced going forward.
• Bug fix for multiple support springs at a single node. When calculating reactions, the program was only considering the effects of one spring at the node, whichever came first in this list: DX, DY, DZ, RX, RY, RZ. This only affected reaction calculations, and has been remedied.
• Bug fix for mesh max/min results. The program was throwing exceptions when the user requested results from the mesh rather than the elements themselves.
• Work has begun on nonlinear pushover analysis. This has proven to be a challenging feature to implement, and still has a long way to go. However, in the process of working on pushover analysis several inneficiencies in the analysis code were identified and fixed. This update includes those changes, and brings the "work-in-progress" pushover code into the main branch of the project. The incomplete pushover branch was brought into to the main branch because critical parts of the analysis code were diverging from the main branch, making it harder to accept pull requests from other users.

Bug Fix for Multiple Support Springs at a Single Node

• Bug fix for multiple support springs at a single node. The program was only considering the effects of one spring at the node, whichever came first in this list: DX, DY, DZ, RX, RY, RZ. This has been remedied.

Tension/Compression Only Support Spring Fix

• Corrections to tension/compression only support springs. v0.0.76 and v0.0.77 were not working as expected. 3rd time's a charm.

P-Delta Bug Fix

• Simplified P-Delta convergence checks
• Fixed P-Delta bug that was introduced in v0.0.76.

Fix for Support Springs & Other Changes

• Simplified P-Delta convergence checks
• Allowed tension/compression-only support springs to reactivate after being deactivated. Erroneous deflections were being reported on very flexible models that experienced a lot of movement with T/C support springs.
• matplotlib is now an optional dependency. You'll only need to install it if you want to view plots for members.
• Documentation for installation has been improved.

PDF Printing Fix

• Bug fix & improvements for PDF printing capabilities. Methods used to print PDF's had fallen behind updates to the rest of the program. It should be fixed now. The way PDF's are printed has changed slightly. Examples have been updated to demonstrate this and the documentation on readthedocs has been updated as well: (https://pynite.readthedocs.io/en/latest/reporting.html).