# Proprietary Software

CP&A has developed a suite of applications that all work together to permit engineering designs to be completed efficiently. We have used the proprietary software to design all types of cranes, bulk handling equipment, buildings and other structures. Additionally we have written custom software to write input files and to read the database files from finite element programs. Our software suite leverages SAP2000's API to run code checking for design codes that are not supported by SAP2000. CP&A's crane analysis suite can be considered a custom workbench for SAP2000.

## Templates

Use SAP2000's API to parametrically create commonly used crane configurations

## Pre-processing

Automate unique loading requirements for cranes. Create multiple models such as boom down and boom up.

## sub-modeling

Pass forces and moments from global model to local models to save on computation time.

## Post-processing

Use custom algorithms to implement design checks for crane design codes that are not currently supported by standard commercial analysis packages.

## Crane Software Features

### CP&A Section

CP&A Section is an interactive environment to create and modify frame element sections. Section stores each section by name in a library. Sections may be defined using equations and variables. Sections may be imported into popular FEA packages such as SAP2000.

### Multiple Model Support

Update load cases/combinations and run analysis for multiple models simultaneously. Check all load combinations across multiple models with a single click of a button.

### CP&A Stress and Fatigue

CPA/STRESS & FATIGUE calculates stresses and fatigue damage on sections for different load combinations. CPA/STRESS & FATIGUE combines constant loads, variable loads, lateral loads in two directions, and angled loads (i.e. wind) from the worst angle. A combination table file defines how the analysis loads are combined. Combination tables are used to define the load combinations to be analyzed to calculate the maximum stress and fatigue damage.

### CP&A Thin Buckling

CPA/BUCKLING checks thin panel plate buckling. Buckling may be checked either manually (checking of individual plates) or by using file input (checking multiple members and plates). CPA/BUCKLING reads stresses from CPA/STRESS binary output file.

â€‹

This can help optimize crane designs of all types. Many crane manufacturers can benefit by actually calculating thin panel buckling rather than using the prescriptive methods for stiffener/diaphragm spacing as outlined in CMAA 70 and AIST TR6.

Connections

CP&A stress may pass the governing forces to connection sheets. The connection sheets check multiple load cases and multiple failure modes. These governing forces may also be passed to FEA models of local connections with a fine mesh.

Non-Linear Tie Downs

The most accurate way to analyze tie down demands for STS container cranes is a non-linear analysis. Tie downs are tension only and wheel loads are compression only, this is inherently non-linear. CP&A have developed a program that quickly analyzes tie down demands on the dock.

## Why Beam Elements?

Beam elements offer many advantages over shell and solid elements when it comes to modeling a full crane structure. The largest advantage is code compliant checks related to structural stability and buckling.

â€‹

To the untrained eye, it can appear that beam element models do not handle connection checks. However, forces from the global model can be passed on to local connections checks for more detailed analysis. This is sometimes referred to as breakout or sub-modelling. Creating a convergent mesh for an entire ship to shore crane is not practical or cost effective. Often times a mesh that is not fine enough is covered up by stress averaging between node points. This can produce a nice looking image but the results are not accurate.

â€‹

To summarize, global stability, stress, fatigue and buckling are checked and verified with beam elements. Local connections are performed with either code compliant connection calculations or local (breakout) shell/solid finite element models with convergent meshes.

## Breakout Models

Breakout models help simplify the global model. This fast tracks debugging and allows us to create accurate analysis without sacrificing on mesh size.

For example, a container crane gantry system is modeled separately due to the high number of elements required for an accurate analysis. In the global model, the gantry is represented by an equivalent beam with the same stiffness as the gantry system. Parameter loads are applied to the breakout gantry model and equivalent beam properties (area, moment of inertia, torsion constant, etc.) are calculated for use in the global model.

â€‹

We find this is the most efficient way to accurately model complex geometry. Creating a convergent mesh for an entire container crane is extremely expensive. Several issues can occur if the mesh is not the right shape or size such as shear locking which overestimates the stiffness. Without a convergent mesh the analysis is just a pretty picture. Mesh convergence is extremely important!

## What is Stress Averaging?

Stress Averaging On

Stress Averaging Off

Stress averaging is when the finite finite element analysis (FEA) program averages the stresses calculated at each node point. Each solid element has an associated internal force and/or stress at each joint. If there are large discontinuities sometimes stress averaging is used to cover up these large stress variation. Large stress variations between elements indicates that the model is not properly meshed and needs to be refined to properly capture the variation in stress.

â€‹

The two images above show what stress averaging can do. The image on the left has stress averaging on, and the image to the right does not. Notice that when stress averaging is turned on the element stress reduces from 62 ksi to 41 ksi. The two images are the same load case with the exact same geometry, the only difference is the use of stress averaging. When stress averaging is turned off, it is obvious that the mesh has not converged and needs to be refined. With stress averaging turned off, one element calculates the stress as 62 ksi and the adjacent element calculates 20 ksi at the same intersecting node point. With stress averaging on, these two numbers are averages and the program displays a nice smooth plot with an average stress at 41 ksi.

The mesh should be refined until these two solid objects calculate similar stresses at the joining node point. There is no guarantee a refined mesh will converge to the average stress, this is one of many reasons why mesh convergence is so important.