2024 Warp Analysis Bug? Negative shrinkage

[JoshRS]

Summary of Issue

I'm getting a significant negative shrinkage value in my warp results with the 2024 warp solvers. No Shrink compensation was included in any analyses. Model 1 and 2 are similar in design, automotive rocker molding. 3.5mm thick, no b-side features, 2m long. Using a DD mesh, mesh quality is acceptable. 5 gates sequenced middle out.

Material: Lyondell Basell 1168 (CRIMS)

Model 1 Run in 2023: 0.32% Shrink, part shrunk by 0.26mm. (Expected)

Model 1 Run in 2024: -1.26% Shrink, part grew by 1.11mm.

Model 2 Run in 2024: -7.23% Shrink, part grew by 5.54mm.

Material: ADX 2343 (CRIMS)

Model 2 Run in 2024: -4.6% Shrink, part grew by 4.6mm.

Material: ADX 2343 (Uncorrected Residual Stress)

Model 2 Run in 2024: -1.33% Shrink, part grew by 1.09mm.

Cross section view - note, the part expands in every direction in the section view.

Lengthwise - the part shrink normally (~0.6%)

Volumetric shrink appears normal, no negative values, fairly uniform 3-5% over the entire part. 

Cross Section View

Is this a bug in 2024? Any suggestions? 

[mason.myers]

Hello @JoshRS ,

Are these materials in the public database?  I see a D21168 in 2024 but it does not have CRIMS.  I was not able to see either of the ADX materials.  

Mason

[JoshRS]

Hi Mason,

The ADX Material is MFC030 (available in the database) and the 1168 was internally tested, unfortunately I can't share that one as it's confidential however it is similar to the ADX material.

[mason.myers]

Thanks @JoshRS ,

We can only speak for the validity of materials in the public database or materials that we test.  I would suggest creating a support case to get our support team to review the 3 applications.

Mason

[JoshRS]

Thanks @mason.myers,

I'll submit a ticket with Autodesk. I tried another TPO (Hifax 735) from the database and it seems to do this as well, although to a lesser magnitude.

[raalteh]

We believe this likey has to do with the automatic warpage constraints not being set correctly somehow. We made some changes in this area to solve some issues, but the change (in 2023) seems to have causes an issue for this specific model. This results in a badly converged solution.

As a temporary work around for this case you could use the attached script to set constraints manually.

[JoshRS]

Will this work with 2024 as well?