Design-To Load Envelopes

A typical stress analysis process starts with loads generated by FEA solve. Using HyperX - or a FEM post processing software - a typical inspection is to plot the color gradient of load magnitude. The engineer individually churns through the list of load cases and corresponding load components for each case, updating the color plot, attempting to get an idea of which load-case-load-component combinations are the highest for each part of the structure.

This is cumbersome, and inherently misses the effects of combinations of load, as no single load gradient plot captures this.

HyperLoads offers an alternative approach to understanding Design-to Loads per Zone. Let's look at the two panels below as an example. For each panel, the element loads can be represented as scatter plots in 2D space, as shown. Each point represents the load state of a single element, for a single load case. On the left is a projection of points on the X-Y plane, middle the X-XY, and right Y-XY.

You can use these plots to see, for example, that the the root panel is dominated by compression Nx, while the panel located near the engine pylon experiences more severe multiaxial loading.

A 2D convex hull can be drawn around the extents of these points, creating 2D envelopes in each plane.

HyperLoads generates these 2D envelopes, but can also combine them into one 3D surface envelope, providing the engineer a way to inspect the load magnitude combinations of Nx, Ny, Nxy in one plot, for every element, for every load case.

This 3D aspect is particularly important because, in many cases, the critical load is not able to be seen in any of the 2D envelopes as it is truly a combination of all three load components, like in the case for the panel represented in blue.

Different Design-to envelopes are required for each different load type. These load types include strength at limit load, strength at ultimate load, buckling at limit load, and buckling at ultimate load; and can be selected on the lefthand-side of the viewport. See Strength vs. Buckling loads to understand more about these different load types.

Buckling envelopes are always generated using an element area-weighted-average processing technique. However, the (strength) Design-to envelopes shown for a given Zone are largely dependent on the loads processing technique options. The number of points included, as well as the shape of the envelope change depending on the load processing technique used to generate the envelope.

The loads processing window can be used to select which technique is used to generate Design-to Load envelope(s). A unique envelope is generated for each selection (i.e. 3 techniques selected, there will be 3 unique envelopes). Examples are shown below for the same Zone for each technique. Their respective settings are also summarized.

Filtered Cases

The Envelopes resulting from these options are created with a filtered set of Design Load Cases, based on criticality. Therefore, they contain fewer points compared to the “All Design Cases” options below.

Peak Element (Filtered)

This technique results in one point for each peak element metric. See Peak Element for more information on this loads processing technique and the various metrics it includes.

Toggling the Comprehensive Metrics option will enable 92 additional metrics during the peak loads processing run. So, there will be an additional 92 points in the envelope.

Peak Element (Extended)

This technique - as the name implies - is an extension of the Peak Element (Filtered) method above. It will result in one point, if not more, per peak element metric. It works by:

1. Run Peak Element (Filtered) and take the list of load cases that flagged ANY metric as being critical

2. For each load case on the list, run all enabled Peak Element metrics.

Any element that comes up as critical becomes a point on the envelope.

Toggling the Comprehensive Metrics option will enable 92 additional metrics during the peak loads processing run. So, there will be an additional 92 (if not more) points in the envelope.

Convex Hull

This technique results in one point for each element-load-case-pair that makes up the surface of the convex hull(s).

The Settings list allows you to select which loads you would like to include in the convex hull.

• Membrane - Select to consider points on the Nx, Ny, Nxy convex hull.

• Bending - Select to consider points on the Mx, My, Mxy convex hull.

• Transverse Shear - Select to consider points on the Qx, Qy convex hull.

• Membrane and Bending - Select to consider points on the Nx, Ny, Mx, My convex hull.

All Cases

The Envelopes resulting from these options consider all Design Load Cases – not just a filtered set.

Peak Element (All)

This technique results in at least one point for each design case. See the Peak Element for more information on this loads processing technique and the various metrics it includes.

Toggling the Comprehensive Metrics option will enable 92 additional metrics during the peak loads processing run. So, there will be an additional 92 points in the envelope.

Element-Based (Default)

This technique results in one point for each element, for each design case.

As illustrated in the previous section, the number of points that make up the Design-to Load envelope(s) for a given Zone are dependent on:

Regardless of the number of points, typically the points of interest are those that make up the outer surface of the envelope - or the skin of the potato. Depending on the loads processing technique, there may be several points interior to the envelope. To more readily distinguish interior vs. exterior points on the envelope, select the "Color Green" option on the ribbon. This colors all points that are interior to the envelope green (as in the loads captured by these points are "good"), but leaves exterior, skin points to remain colored by the selected coloring method.

There is also an option to hide all interior points from the plot. Notice this option also removes any Legend items that are no longer being used now that the interior points are hidden. This can be a quick way to understand, for example, whether or not a particular Design Load Case results in any points on the exterior of the envelope.

As with the traditional HyperX Viewport, selecting an item in the Legend (a Design Load Case, in this case) will cause it to be highlighted in the HyperLoad Viewport.

You can also toggle cases on-and-off using the Legend.

But, there is still a need to further quantify the effect of a certain Design Load Case (or Cases) on the overall envelope.

Using the Loads Processing Dialog, users have the ability to select particular load cases of interest to be isolated. Once isolated, case(s) in this list can be colored independently, while still showing the rest of the envelope. See the example in the image below.

Note

Isolating Design Cases works only for load processing techniques that include All Design Cases - trying to isolate a design case while using a Filtered processing technique may result in the isolated case being removed by the filtering process.

A load ratio vector - which numerically quantifies the how close the critical load from a certain case is to the envelope - can be calculated and shown for isolated Design Load Case(s). These are the vectors shown in yellow in the image below. This option is enabled by toggling to "Isolated Design Case" in the dropdown menu on the ribbon.

The load ratio vector for a given design case starts at the controlling load point for that load case and, following the same (Nx, Ny, Nxy) vector orientation as that point, extends to the surface of the envelope.

Said more mathematically, if A is the Design-to Load vector, let B be the vector in the same direction as A, but projected onto the surface of the Design-To envelope. A line is drawn from A to B, and the load ratio is reported as ||A|| / ||B||. This, in effect, is a way of measuring the exterior-ness of a given design case in relation to the envelope surface.

A label containing both the numerical value of the load ratio as well as the controlling element ID can be added using the Probe, which is located on the View tab of the Ribbon. For example, for the NF_Cruise design case shown below, the load ratio is 0.2778. This means that the most critical load in that design case is only ~28% of the actual controlling load case in the same orientation. The higher the load ratio value, the closer the critical load for the given design case is to the surface of the envelope.

In many cases, it is useful to export the Design-to Loads to a .csv file for incorporating into other tools and workflows. This can be done from the Design-To Loads tab of the ribbon, as shown in the figure.

The resulting form presents several options for what loads are exported. Expand the sections below to understand those options.

Include Principal Loads

Use this checkbox, and the subsequent options beneath, to export principal loads for each Design-to Load point in addition to the applied loads.

The membrane principal loads are given by the equations in the image below.

The additional options determine how the moments are calculated...

• Moments at Membrane Principal Angle - This option exports moments that are calculated with respect to the membrane principal axis, given by the equations below.

  

• Principal Moments - This option exports principal moments based as they are traditionally calculated - directly from the applied FEA moments.