CAESES and FS-Flow (DNV-Futureship)

[Mr. Maurizio Rosson 1]

Hi everybody,

 

I post this topic because I guess that people in FRIENDSHIP SYSTEMS know the FutureShip FS-Flow panel code well and I'd like to improve my hullform design procedure and analysis.

 

I work for a small Italian company, active in marine vehicles field and with a leading position in Ro-Ro ships design - but not limited to. About CFD analysis before and after tank tests, I've been using FS-Flow panel code since 2010. Hullforms are designed with LINES (AVEVA-TRIBON) or DELFTShip, the latter allowing different designs quickly - a perfect fairing is not required at this stage. No parametric hullform design - and in the future I'm going to explore CAESES-FFW for that. Back to my question, the usual procedure is:

 

- hullform design with DELFTShip (subdivision surfaces method)

- exporting from DELFTShip (IGES format)

- rebuilding with Rhinoceros
- exporting from Rhinoceros (IGES format)

- importing in FS-Flow

 

The procedure is very tricky because of IGES geometry problems. When exporting an IGES format, DELFTShip tries to 'translate' the subdivision surfaces with less IGES surfaces as possible, but the geometry results in a set of dozens, hundreds of patches (and the quality is not so good, a lot of small holes or overlapping edges). That's a nightmare for FS-Flow: I should create more PolySurfaces and apply the correct orientation to a very, very large number of surfaces! So I rebuild the hullform in Rhinoceros with a network of curves (stations, buttocks and waterlines), also creating more 'groups' for different panel meshes (bulb, middle body, aft...).
I wonder if CAESES-FFW CAD tools could allow me to import and manipulate the IGES model from DELFTShip in order to obtain a simplified set of IGES surfaces. Moreover, I'm interested in the SHIPFLOW format (.shf): I guess it's the same .shf input format for FS-Flow. Is there any tutorial or sample about that? What are your suggestions?

The next step is connecting CAESES-FFW to FS-FLOW. I hope some experienced users can share their know-how.

[Mr. Mattia Brenner 6]

Ciao Maurizio,

 

nice to see you here in our forum and also to hear that you are experimenting with CAESES. Sorry for not responding earlier. Last week was really busy and I kind of forgot about this...

 

I can confirm what you have experienced with the DELFTShip export. I have also imported some IGES or Parasolid files from DELFTShip and they are made up from an extremely high number of small patches and the quality is really bad. Probably, they are fine if you stay within the subdivision surface world, but the problem starts when moving over to the NURBS world. Unfortunately, there is no quick method that comes to my mind to create a clean surface geometry with a small number of patches. The only idea I have is pretty much in line with you procedure in Rhino, namely creating some kind of sections through the geometry and skinning them with new surfaces. But that is also not ideal.

 

Of course, one could argue about the necessity to start in DELFTShip. Why not using Rhino directly? It seems that the reason you are using DELFTShip is the quick generation of variants. Then of course CAESES would be an ideal tool, since this is our special focus. And you can prepare a suitable input for FS-Flow directly...

 

I think (possibly strangely enough) no one here is really familiar with FS-Flow, since it is mainly used internally by Futureship with their own specific workflow (they uses CAESES in batch mode just to create the geometry). The only contact we had with it was about 6 years ago, when we were working on a project together with a common customer. From that time, I have a document prepared by Futureship that describes the connection. I had sent this document to you, Roberto and Giovanna by email on April 4, 2012. Maybe you can still find it, otherwise let me know and I will send it to you again. It might not be completely up to date, but maybe it is a good start. If you have specific questions about the CAESES part, I can help you along. For the FS-Flow part you will have to ask Futureship.

 

If I recall correctly, you can use the Shipflow offset format as input for FS-Flow, too. To prepare this format, you basically just have to create section groups through your surface geometry and put the different groups in an offset group assembly, which can then be exported as SHF file (via the export menu or with the command <assembly>.exportSHF(String filename)). You can see how this input is prepared for Shipflow in these samples:

 

 

I hope this helps, if you have more questions, let me know.

 

Saluti,

Mattia

[Mr. Maurizio Rosson 1]

Hi Mattia,

 

          thanks for your kind reply.
Now I know you can speak italian very well. I could contact you by e-mail or phone but, so far at least, I keep writing here in english - maybe someone else is interested in this and next topics :)

I found your e-mail, I didn't consider your presentation because I was concerned with other questions then. And now I'm very interested in CAESES connection with FS-Flow. I think that FRIENDSHIP Systems has provided a free CAESES-FFW limited version only recently - am I right? It's a good choice for testing and looking for new paths.

About you presentation, the connection procedure seems quite tricky. I understand the workflow, but details are too hard for me at the moment. Anyway, I'm sure that much has changed for FS-Flow and CAESES since then. Our main target is improving the CFD analysis workflow and tools, from hull geometry creation to batch processing analysis (w/o geometry variation). The following topics can be addressed.

 

Hull design and geometry setup

 

DELFTShip is a very quick and simple tool for hull creation, especially for hydrodynamics, but the geometry export is very bad for FS-Flow and even beyond - as I write in the first post. I rarely used .shf offset format, but I could investigate more alternative:

• a code (C++, Python, Microsoft Visual Studio...) for manipulating DELFTShip offset files (more formats)
• a CATIA or Rhinoceros internal script for manipulating DELFTShip 3D curves or surfaces exporting
• CAESES CAD tools

I just started to explore CAESES, but it seems to provide a lot of different tools to obtain offset groups and other more FS-Flow-friendly geometry formats - I have to walk through a lot of CAESES tutorials and samples :)
I suspect that different 'parent hullforms' require different approaches and geometry formats. For instance, NAOS designs often concerns double skeg ships and no boss.

Anyway, that leads to another prospect: hull design in CAESES. I don't know if CAESES allows a very good hull fairing (for shipbuilding), but it's definitely good enough for hydrodinamics. Moreover, CAESES offers parametric CAD functionality. I guess that the free version includes (or excludes at all) only a few variation tools and optimization engines, but that is not a priority at this stage.

 

FS-Flow connection

 

If I'm correct, CAESES-FFW can be connected to FS-Flow in different ways. You developed a procedure in the past (no updates?), other approaches have been investigated in Futureship. I'm in contact with Futureship on these days, but I don't know how much they can share.
I'm interested in FS-FLow connection in order to setup more simulations to run in batch mode (w/o automatic hull variation) and post-process the panel code results (FS-Flow post-processing tools are a little basic).

 

-----------------------------------------------------------------

 

And now a more specific CAESES-FFW question. About the feature GenerateSHFOffsetGroups in CAESES (Features / hull design / connections / shipflow), I can't understand the Tuning section options.

• What does it mean Attach Distance? It seems to affect the beahaviour of an unknow line (see the diagonal line attached picture), connecting one or more section curves, but I can't get rid of it (or setup correctly).
• What does it mean Y-Position of Stem Section? I guess it concerns the semi-breadth of the bow section, when the stem is 'filleted' (not sharp), but I'm not sure about the X-position to define the semi-breadth.

Thanks for your attention, have a good day.

 

Maurizio

 

[Mr. Mattia Brenner 6]

Ciao Maurizio,

 

actually, there are no updates about the FS-Flow connection. What we did 6 years ago was the last and only time we worked on the coupling. The reason is simply that there are hardly any commercial customers for FS-Flow and maybe even less that want to use it together with CAESES. Your request is the first we have received since then. I agree that the coupling that is described in the document is somewhat complicated, though. This is, however, mostly attributed to the FS-Flow part. If I would start from scratch to set up the connection with the current version of CAESES, I would definitely use the Software Connector. You can find instructions about how to use the Software Connector in different tutorials that come with CAESES (in the categories "Getting Started" and "Integrations").

 

About your specific questions:

• Attach Distance is a setting that you can also find in the section group (and this is also where it is internally used in the feature). It is a value that describes the maximal distance between two surfaces that will still create a continuous section. So if a gap between two surfaces in the model is bigger than the attach distance, the sections that are created there are split at the gap.
• The Y-Position of the Stem Section is typically 0. Sometimes, however, the surface model is a bit unclean (maybe due to tolerances in the export from an upstream tool), so that some surfaces or parts of them don't go to y=0 at the stem. Then the stem section will have some gaps. In that case moving the stem section slightly in y-direction (say to 0.0001 or even less) can solve the problem.

In the case shown in your picture there seems to be a problem when the stem section is connected to the bulb part of the first hull section. If you can attach your project, I can try to fix it.

 

Saluti,

Mattia

[Mr. Maurizio Rosson 1]

Hi Mattia,

 

          I see your point. And I guessed that CAESES Software Connector has been developed a lot in last years. I also think that starting from it makes more sense, when FS-Flow set-up and triggering for batch processing is requested.

Anyway I'm focused on the geometry creation/set up on these days. The geometry preparation is the first main question, so I'm walking through the modelling tutorials. I'll study the hull design samples later, then I'm going to make the first tests.

The main target is to build a routine through features or whatever (I'm still not so familiar with CAESES terminology) in order to export a proper geometry for FS-Flow panel meshing (offset or IGES). I guess that a routine is required for each hull 'topology'.

The input geometry source is not so relevant, if the same procedure is applied. The hullform could be created with an external software, but I'm very interested in CAESES CAD tools for hull design. The parametric modelling and geometry variation would allow to generate more different (but similar) hulls to test - and that's would be a big improvement for us. That's why I asked if CAESES really allows a very good fairing. I took a look at the samples: the hulls seem to be quite good for CFD purpouses, but some areas and transition between surfaces sometimes show some discontinuities or strange behaviour. I'm not so expert to understand if that concerns CAESES limits. Hull samples are very good for learning anyway and samples doesn't require a perfect fairing anyway.

About my previous specific questions, I attach the project. The hull geometry comes from CAESES-FFW.
Thanks for your help, regards.

 

Maurizio

ShipflowOffsetTest.fdbc

[Mr. Mattia Brenner 6]

Ciao Maurizio,

 

please see the attached project file, where the offsets are now generated without problems. I basically just slightly adjusted the inputs of the feature.

 

Surely, more attention could be put on precise fairing during the modeling process. However, as you suggested, CAESES is mostly used for the CFD-driven optimization process by the basic design/hydrodynamics people. The optimized hullform is then exported and processed in suitable downstream tools for production. So we mostly model the geometry so that it is good enough for CFD. That is why we also might lack specific fairing tools that you possibly know from other tools, as well as because of our special modeling process. 

 

Saluti, buon Natal e felice anno nuovo,

Mattia

ShipflowOffsetTest.fdbc

[Mr. Maurizio Rosson 1]

Hi Mattia,

 

          thanks for your help. I see your editing in the feature input, just a question of mm!

Quite apart the fairing tools, I keep on walking through CAESES-FFW tutorials and CAD tools (geometric entities, parameter and design variables, engines ...) seem very powerful to me.
I'll post questions and share feedbacks here in the forum on next weeks.

 

Merry Christmas and a Happy New Year!

 

Maurizio

[Mr. Maurizio Rosson 1]

Hi all,

     just a feedback a few months later.

As I wrote, we were interested in including the whole process in CAESES (hull form modelling, geometry preparation and setup, CFD analysis, post-processing). Our current workflow (hull design, CFD analysis and optimization) is well tested, but it involves more softwares and is slow and tricky. The FS-Flow connection and post-processing is not a priority, so we focused on the modelling and geometry preparation aspects.

Unfortunately CAESES didn't include specific tutorials for ship design, just for a very simple yacht. Some useful sample projects are included, but they are too complex for an absolute beginner because it's hard to see the modelling procedure and organization. At the moment I can't stop the work routine to became an actual expert in the CAESES software and I learnt the basics at odd moments. So we had to scale down our ambitions and just looked for a faster geometry preparation. The hull form is modelled with an other software anyway and exported in IGS format.

We started from the Generate SHF Offset Groups feature and created a new not-so-elegant one, suited for FS-Flow pre-processing and setup. In fact, most of our projects concern double-skeg ships and two distinct skeg offset groups are required (one for the area between the skegs and one outside), so some further manual steps are involved.

I've just installed the last CAESES release and found a new tutorial for hull design (fast monohull). Our usual hull forms are more complex but the fast monohull cookbook is useful anyway, now I really can see a typical hull form modelling procedure. I wonder if other ship design tutorials (even just a guidelines colletcion) are available - or any specific documentation about ship design is dedicated only to Friendship Systems training sessions?

[Mr. Claus Abt 14]

Ciao Maurizio,

 

I guess participating in a training or asking us for a parametric model made for your needs seems to be the most economic way.

Regards

Claus

[Matthias Maasch 21]

Dear Maurizio,

 

additional to Claus' comments, there is a new tutorial "Fast Mono hull" which covers the basic modelling approach for a ship hull. It is included in the new 4.0 CAESES version.

 

Best regards

 

Matthias

[Mr. Maurizio Rosson 1]

Hi Matthias,

 

          I already studied the new tutorial, I'm the one who opened the new related thread :)
About Claus reply, the new tutorial was very helpful for a better comprehension and on these days I analyzed the old (and new) samples and things are more clear to me. When I have time, I'll try to set up a couple of templates, according our needs. Anyway, I think that in the future an advanced training session would be really helpful and I'll discuss this with my boss. Is there a possibility for webinars too?

[Mr. Claus Abt 14]

Webinars? sure!

Claus