Sometimes in the moment of modeling the structure, we lose control of the nodes and the elements created, and more so when the file is manipulated by datagen, until now technical support with the developer's team help has found many cases in which the user presents inconsistencies in the result at the moment of using the collapse module, even though creating a member from a single joint using offsets works in many of the SACS modules, it creates an issue when this is used on a nonlinear collapse analysis.
This is because nonlinear collapse (old version) uses an updated Lagrangian formulation where the nodes are moved in each iteration. Having a member that only has one joint trivially does not work since the two ends of the member have to be moved to different locations.
Next, the principal inconsistencies are presented when the user works with elements that have the same joint at the start and at the end.
1. Inconsistency in total pile head reactions in gravity loads between Pushover and In-service
Users present this inconsistency at the moment to compare the gravity load during a static analysis vs the gravity load given by the collapse analysis.
Next is presented a case in which the user had a member connected to the same joint (member 2105-2105)
Here are some screenshots about this inconsistency in the gravity load result of having this element:
2. Inconsistency in Graphics getting by Postvue
When we use elements that have the same joint at the start and end, another common inconsistency is the bad behavior in the graphics when using Postvue.
Next, a case where the model had a member connected to the same joint (Member 4409-4409). The model is presented in which we have the application of two loads, one gravitational load and the other one regarding the wind, wave, and current effects for a storm condition:
- Load 904 - Gravitational load
- Load 1653 – Wind, wave, and current 90ᵒ degree
Load Sequence
Base Shear Graphics
In the upper graphic, we can see the effect of lateral forces taken while the gravitational steps are coming. This represents a wrong behavior due to that the slope of the shear load has to start at the application of the lateral force in step 1 of the load 1653
Important: if after following all these recommendations you aren't able to get the model up to the convergence, it's probably that the root of the issue will be related to another issue.
3. Impediment to reach the convergence during the ship impact analysis
When we are running a collapse analysis and when the load application instead of increasing, this is reduced, this could be due to the presence of an element with the same node on its edge,