In v5759 we revised the structural flange analysis methodology. While the previous analysis methodology had been used for many years, the technical basis was confusing and difficult to follow. There is little published information for the design of structural flanges, and so this was a method we had developed internally. The methodology we used for flanges WITHOUT gussets was not completely consistent with the methodology used for flanges WITH gussets. It was not uncommon to design a flange WITHOUT gussets and get a required thickness for the flange, then gussets were added and the required flange thickness was greater that that without gussets. The reason was that the design methodology was not the same for both designs, and so it produced illogical results at times.
There is still not any published information that we know of that specifically addresses the design of structural flanges with or without gussets; however, we have found published information in technical papers and books that do cover aspects of the flange design. In v5758 we re-wrote the structural flange analysis basis, and developed some articles to explain the methodology with links to the supporting technical information. Unfortunately this will cause some discrepancies when compared to structural flanges designed previously; however, we believe this new basis corrects some errors from the past and provides a more technically sound basis for the design of the flange.
The following are articles explaining the design basis for the flanges within MecaStack:
Here are some of the major observations you may notice when comparing structural flanges designed prior to v5759 (ignoring v5756, v5757 and v5758 due to intermediate changes) to flanges designed in v5759 and later:
Flange Thickness Without Gussets:
The old analysis utilized the AISC 9th edition (Green book) criteria for prying on a bolted flange. The method was not simple to apply to our situation, and in the course of implementing mistakes were made in how it was applied. The a' was incorrectly calculated to subtract have the hole diameter, when it should have added it. Many of the calculations utilize a' and so it has a big impact to the calculations. We are now using AISC 15th edition prying and there are some changes, but we also tried to apply it in a manner that is more consistent with the intent of AISC. The 15th edition measures a, b, a', and b' relative to the mid-plane of the stem (in our case this is the cylindrical shell) and in the 9th edition it measured relative to the outside of the stem (OD of our shell). This has an impact in the results as well.
The biggest change was that we also now implemented criteria based upon Yield Line analysis in conjunction with the prying checks. For most flanges this will be beneficial and will result in a flange that is thinner although there could be cases where the thickness is greater.
The prying force that is estimated (qr in new, Q in old) is often less now than it was in the old version. This indirectly affects some of the other calculations.
Flange Thickness With Gussets:
The old analysis considered that when there are gussets that all of the bolt load went to the gussets and that the flange between the gussets behaved as a fixed-fixed beam.
The new analysis uses prying and yield line analysis which is more consistent with the analysis of the flange without gussets. Our research has indicated that when the gussets get too far apart then the flange basically behaves as a flange without gussets, and so we added that check in to the calculations as well and provide the user with guidance on when the gussets are too far apart.
Gussets:
Previously we were only checking the compression case and it was actually using the bolt load (Tension) for that check. We have now broken up the analysis to consider both tension and compression. We have found the results to be slightly more favorable in most analysis for the new calculations.
Bolt Pre-Load:
The Pre-Load checks will be a little different, and this is primarily because the estimated prying force has changed (See Flange Thickness Without Gussets above).
Bolt Tension:
The bolt tension Unity Ratio will often be higher. This is because the old analysis used a more complicated ratio of flange to bolt stiffness to attempt to determine the load on the bolt itself. The method was confusing and we decided that it was better to use the more conservative option of just verifying the tension on the bolt without any consideration of stiffness ratios. In addition, we do not include prying force, because the prying force is considered in the Flange Thickness check which is really a verification of the tension capacity of the connection as a whole.
Legacy Designs:
We regret any problems that these changes may cause you, but we are obligated to fix and/or improve issues when they are discovered. If you have an existing design and do not have the ability to easily transition to this new calculation basis, you can install old versions (only use version 5755 or older; do not use 5756, 5757 or 5758) of the software by following this link.