For engineers who have to compare many possible design variations, I'm curious how you actually handle it in industry.

For example, if you have hundreds or thousands of possible geometry/parameter combinations:

• How do you decide which designs are worth simulating?
• Do you actually run simulations for all of them, or use some kind of screening/optimization method?
• What tools do you use for this?
• What part of the process takes the most time?
• Is computational cost the main limitation, or is setting up/running/post-processing the simulations the bigger bottleneck?
• Do you use surrogate models, reduced-order models, optimization software, custom Python scripts, etc.?
• What happens when the design space becomes too large to explore practically?

I'm particularly interested in how this is handled in industry and what the biggest bottleneck is in practice.

For engineers who have to compare many possible design variations, I'm curious how you actually handle it in industry.

For example, if you have hundreds or thousands of possible geometry/parameter combinations:

• How do you decide which designs are worth simulating?
• Do you actually run simulations for all of them, or use some kind of screening/optimization method?
• What tools do you use for this?
• What part of the process takes the most time?
• Is computational cost the main limitation, or is setting up/running/post-processing the simulations the bigger bottleneck?
• Do you use surrogate models, reduced-order models, optimization software, custom Python scripts, etc.?
• What happens when the design space becomes too large to explore practically?

I'm particularly interested in how this is handled in industry and what the biggest bottleneck is in practice.

I'm interested in how FEA/CFD teams deal with the gap between simulation results and real-world test results.

For example, if a simulation predicts one result but the physical test gives something noticeably different:

• How do you determine what caused the discrepancy?
• Do you have a systematic process for correlating the simulation to the test?
• What data/tools do you normally use?
• Is this usually handled manually by the engineer?
• How much time can this investigation take?
• Are there particular parts of this process that are consistently painful?

I'm especially interested in how this works in industry rather than academic/research settings.

I'm trying to understand whether this is actually a significant engineering problem or just something that looks more painful from the outside.

[supprimé]

I'm researching the workflow engineers use to manage CFD/FEA simulations, particularly when you have dozens or hundreds of runs with different meshes, solver settings, boundary conditions, models, etc.

I'm considering building a tool that would:

• track each simulation run and its settings
• keep runs organized/versioned
• compare results between runs
• automatically check things like convergence and mesh independence
• generate verification/validation reports

Before building anything, I want to understand how people actually handle this today.

For those of you who regularly use Fluent, ANSYS, OpenFOAM, Abaqus, etc.:

1. How do you currently organize your simulation runs?
2. Do you use folders, Excel, scripts, Git, a PLM system, or something else?
3. How do you compare a large number of runs?
4. How much of your verification/convergence/mesh-independence analysis is manual?
5. Do you have to manually create reports documenting your simulation methodology and results?
6. What's the most annoying part of this workflow?
7. If a tool automated some of this, what would it need to do for you to actually use it?
8. Would this be something you'd pay for? If so, what pricing model would make sense?

I'm particularly interested in hearing from people who have dealt with 20–100+ simulation runs rather than just a few cases.

I'm not trying to sell anything here — I'm trying to find out whether this is actually a problem worth solving before I spend months building it.

I'm researching the workflow engineers use to manage CFD/FEA simulations, particularly when you have dozens or hundreds of runs with different meshes, solver settings, boundary conditions, models, etc.

I'm considering building a tool that would:

• track each simulation run and its settings
• keep runs organized/versioned
• compare results between runs
• automatically check things like convergence and mesh independence
• generate verification/validation reports

Before building anything, I want to understand how people actually handle this today.

For those of you who regularly use Fluent, ANSYS, OpenFOAM, Abaqus, etc.:

1. How do you currently organize your simulation runs?
2. Do you use folders, Excel, scripts, Git, a PLM system, or something else?
3. How do you compare a large number of runs?
4. How much of your verification/convergence/mesh-independence analysis is manual?
5. Do you have to manually create reports documenting your simulation methodology and results?
6. What's the most annoying part of this workflow?
7. If a tool automated some of this, what would it need to do for you to actually use it?
8. Would this be something you'd pay for? If so, what pricing model would make sense?

I'm particularly interested in hearing from people who have dealt with 20–100+ simulation runs rather than just a few cases.

I'm not trying to sell anything here — I'm trying to find out whether this is actually a problem worth solving before I spend months building it.