Jump to content
Mr. chris baki

Help for Beginners

Recommended Posts

Good morning to everyone,

Hello Mr. Heinrich,

 

Ok, so far so good. I, still, have some technical issues to firure out but i ll get back to those later on.

 

For the next step, i d like to evaluate the smoothness of surfaces 'n curves created. The traditional way. let's say, is to check each section's or waterline's curvature graph, adjust them, update them respectively and run through that cycle several times...

 

1) For now, i still haven't found out the tools provided to check the curvature of curves or surfaces (so i exported an .iges, opened it at another program and assessed the curvature of curves 'n surfaces which in some points isnt good enough or the desired one). How can i isolate each section/waterline and check their curvature since they are offsets generated from surface interpolation? Should i just make new interpolated curves from surface ?

2) Concerning my design, the things that i can change to improve smoothness of the curvature are only the basic curves (Profile, Deck, Up&Low Chines, and maybe some angles at the bow region), am i right? 

3) Do you have any other way of approaching that problem to suggest?

 

Thank you,

Chris

 

Share this post


Link to post
Share on other sites

Hi Chris,

 

Usually you should not have to go through any fairing procedure. Once you have set up your parametric model it should give a fair hull surfaces for any geometry variation. Often times, this means that you will have to create a few designs automatically as a robustness check or manually play around with the design variables to find a parameterization that does not lead to unwanted shapes.

 

For the main part of the hull, your only means of changing the shape is of course to adjust the underlying curves. For the bow region you can play around with the feature I created for you - this was really just meant as a starting point...

 

Of course there are various ways to investigate smoothness in CAESES. You can find them in the display options of the curves and surfaces e.g. display normals, map curvature (gaussian / mean), isophotes, etc...

 

Best regards,

Heinrich

Share this post


Link to post
Share on other sites

Hello Mr. Heinrich,

Thank you for your answer, i understand that curves should be fair enough but my curvatuere porcupine shows some sudden changes of curvature that i dont want anyhow...

 

For the time, my biggest problem is sth else. I just realized that that my 2 chine vessel actually is not a 2 hard chined vessel like it should be. I mean that the knuckles of sections in most cases do not pass exactly from the chines, it has a small curvature near the knuckles and not a sharp corner. Can you help me on that? i read somewhere about the bevel method but this is way too far advanced for me. Is there any other way possible to achieve that sharpness on the edges defined by the Up&Low chines??

 

Thank you very much for your assistance,

Chris

 

*I know i have some problem on the edge of main surface and front surface but i think i gonna resolve that. For now, the sharpess is my number 1 priority...

Share this post


Link to post
Share on other sites

Hello Mr. Heinrich,

Thank you for your answer, i understand that curves should be fair enough but my curvatuere porcupine shows some sudden changes of curvature that i dont want anyhow...

 

For the time, my biggest problem is sth else. I just realized that that my 2 chine vessel actually is not a 2 hard chined vessel like it should be. I mean that the knuckles of sections in most cases do not pass exactly from the chines, it has a small curvature near the knuckles and not a sharp corner. Can you help me on that? i read somewhere about the bevel method but this is way too far advanced for me. Is there any other way possible to achieve that sharpness on the edges defined by the Up&Low chines??

 

Thank you very much for your assistance,

Chris

 

*I know i have some problem on the edge of main surface and front surface but i think i gonna resolve that. For now, the sharpess is my number 1 priority...

boat_FAT.fdbc

Share this post


Link to post
Share on other sites

Hi Chris,

 

when you look at the surfaces in 3D you can see some kind of rounding at the chines as shown in this picture:

post-933-0-50035700-1550821492_thumb.png

However, this rounding comes only from the rendering resolution (the edge gets sharper if you increase the resolution in the display options of the surface). The underlying Nurbs representation exactly meets the specified edge with a distinct sharp corner, as you can see when looking at the brep of the hull:

post-933-0-39964600-1550821501_thumb.png

So there is nothing to be fixed here...

 

Cheers,

Heinrich

Share this post


Link to post
Share on other sites

Hello Mr. Heinrich,

 

I tried to increase the resolution to let's say 500x500 but.. If you allow me, my concern is that when i get the sectons/waterlines as offsets of the surfaces i can see those small differences as shown in the  pictures posted. I notice that the edge of the surface doesnt meet the chines. Is that caused by the resolution ?

 

Thank you very much for your support.

 

Kind regards,

Chris

post-1088-0-20158900-1550836176_thumb.png

post-1088-0-39670000-1550836183_thumb.png

post-1088-0-87947800-1550836191_thumb.png

post-1088-0-39384100-1550836204_thumb.png

Share this post


Link to post
Share on other sites

Good afternoon mr. Heinrich, thank you for your tips. I was really depressed when i could not see the chines properly aligned with the knuckles of the surface. Now, please i think i need some more hints.

 

When i try to create my skeg-> Projection Skeg top (as in fast monohull) i notice that the end projection point appears at z=-0.003 (x=11.4) approx. .When i zoom in a bit more, i see that the main surface in some points along x-pos (CPC_1 curve) takes negative z-values?? Can this be happening? Is that related with the failure at the created skeg? Moreover, can that be related with the other warning i get about an internal pole in main surface or in "prostego" ??

 

I am grateful for your help,

Chris

boat_FAT_evolution.fdbc

Share this post


Link to post
Share on other sites

hi all !

 

Im trying to model a hull from existing IGS file and want to create many variants of it and check the best hullform for the constant displacement.

 

After importing the IGS file how do we proceed further 

 

Thank you!

Share this post


Link to post
Share on other sites

Hi,

I can tell a few things based on my up-to-now, realtively small, experience.

It ,really, depends on which curves you inserted, which parameters you want to adjust, what kind of transformations you want to apply, and in what extent you want to parametrize your project.

If your imported curves have relatively few points and you can control them, you can just hold a copy of them as reference and then just start to insert parameters on the points you want to adjust. On my case though,where the imported nurbs curves had lots of points, difficult to adjust, the way that i went thought is the following: Recreated the basic curves and maybe a few auxilliary ones(bsplines,fsplines/lines etc.) and then started to create the surface in order to achiece to close it nice and create a starting hull as similar as possible to the imported one. Then parameters...

Are you going to recreate the hull in caeses for cfd calcs where meshing surface is critical or just the curves (waterlines,section,cpc etc.)? I am telling this cause i still have some problems in closing my surface/hull perfeclty.. I think you can just create your offsets even if the hull is ot not closed fully for your hydrostatics. Also,obviously, the more curves you have, its easier to control your vessel but changes in shape will be less probably.

 

You can check, as well, the files i exchanged through here in order to get some ideas or try things.

 

Have fun,

Chris

Share this post


Link to post
Share on other sites

Hi Chris,

 

yes, that can in fact be the case. Just add a rail to the meta surface to make it exactly match the cpc.

Regarding the pole warning in PROSTEGO/s1, that refers to the curve c6. I changed the input curve 'imagecurve1curve' (it now gets the actual surface edge as input) to fix this...

You now have another pole warning tracing back to your imported deck contour --> you should probably create a nice reparametrization of this...

 

Now on to priyankas question. Regarding the modelling, you can choose to follow the way Chris has chosen and build up a fully parametric model based on the geometry you have imported. The alternative would be a partial parametric hull where you can choose from a wide variety of free form deformation and shift techniques inside CAESES. Of course, also a combination of both approaches is possible.

 

The tricky issue - keeping the displacement constant while changing the geometry - can easily be achieved through a nested optimization. You can read how that works and download an example from this post:

 

https://www.caeses.com/forum/index.php?/topic/734-constant-volume-kept-during-dakota-optimization/

 

If you choose a Lackenby shift you can directly adjust the XCB (longitudinal center of buoyancy) and displacement of your vessel - take a look at this video:

 

https://www.caeses.com/blog/2016/video-adjusting-displacement-and-lcb-in-5-minutes/

 

Hope this helps.

 

Best regards to both of you,

Heinrich

 

 

 

boat_FAT_evolution.fdbc

Share this post


Link to post
Share on other sites

Thank you for your answer,

 

1)I inserted a rail curve (an image curve of the CPC_1) but the problem remains. To be honest, when i did that it seemed that problem got solved but after 1 day and changing some other things irrelevant it seems that i am in the same situation. So, can u help me fix that and basically explain me why that happens? Since the main meta surface gets CPC_1 as a curve (CPC_1 runs for sure at z= 0) how can that happen? Also when i try to create an image curve of CPC_1 again if i zoom i see the image curve at different pos compared to CPC_1, also if i change the Parametrization it moves a bit up or down...

 

One more question concerning the skeg, Which images sould i create? I mean i create aft_sub1 and aft_sub 2 but thy present differences than the original metasurface. Specially, aft_sub2 seems to do a little zig-zag close to the stern. Also, if i try to trim metasurface (create image of it and put the projection point p2 as in fast monohull exemple) it seems to change even more...Why that happens? Which is the correct way of advancing on that matter?

 

2) concerning FrontSurfaceTry i see that there is a small hole at the connection between main_metasurface and FrontsurfaceTry at z=2m  approx. Why is that happening ? rail curves etc. are all set...

 

Thank you very much for your help,

Chris

 

boat_FAT_evolution9.fdbc

Share this post


Link to post
Share on other sites

Hi Chris,

 

1: I checked your project (with the rail 'ProfileImage') and I can't see any problems at all.

Without using a rail it is normal to have a slight oscillation in the surface. This can be reduced by increasing the number of interpolated sections ('Interpolated points in surface direction' input for the meta surface). If you add a rail you will have no oscillations at all on the surface edge.

Zooming in is not always the best way to judge things due to the limited render resolution. (Increasing the resolution in the display options of curves and surfaces does NOT change the geometry but only the rendering...)

 

About the skeg: the parametrization of the surfaces was not ideal. For ruled surfaces it is always helpful, if the rulings are not warped too much... I replaced the input curves by imagecurves for which I set the parametrization to 'unit speed' (you can do that in the object editor in the 'general' options...)

 

From here I suggest (instead of creating subsurfaces etc... ) you continue with Brep modelling as this makes it very easy to combine skeg and hull, as well as close the hull and possibly create a computational domain for you CFD at a later stage...

 

2. See my answer to question 1. There is no hole - you can also visually check the surface by when looking at the Brep 'hull'

 

Hope you are making good progress. Cheers,

Heinrich

 

boat_FAT_evolution9.fdbc

Share this post


Link to post
Share on other sites

Hello thank you for your answer but new queries keep comin,

 

1) Can u explain a bit or give some hints(maybe a link etc.) why u prefer to use the image curves and why the result is better?

2) I changed a little bit the for part(CL_FOR) of Chine Low (decreased breadth 5-6cm) and the upadated Brep presents a hole at the for end of main metasurface? If i just create sections from surfaces i have no problem though...

3) Brep is really fine but if i wanted to create the subsurfaces etc. what would be the right way of thinkin? I mean should i have splitted the main surface in smaller parts?

 

Thank you,

Chris

Share this post


Link to post
Share on other sites

Dear Priyanka,

 

I can not see any attachments, but since this thread is more about fully parametric modelling while you are interested in a partial parametric approach - why don't you just open another thread to keep things a bit more organized. Feel free to sent me a PM when you have done so and I will take a closer look.

 

@Chris:

 

1. It is not about the image curves - they are just used to allow trimming.

2. Try increasing the number of sections of the main meta surface. Keep in mind, that you will have to ensure a horizontal tangent at the end of this chine curve other wise you will loose continuity towards the bow region.

3. You will most likely need something like the 'subsurface from surface curve' (in surfaces / sub surface / more / ...)

 

post-933-0-74483600-1552384342_thumb.gif

 

Cheers,

Heinrich

Share this post


Link to post
Share on other sites

Hello mr. Heinrich,

 

1) I asked about the image curves because when the ruled surface was created from the multisegmentsmoothcurves the result wasnot that good. On the contrary with the use of image curves 9and unit speed trck) the resulted surface is better, so i guess there is some difference...

2) Anyhow, I have still the same problem for which i asked help for the skeg connection. I mean when i use Brep i get the same problem concering the sections produced over the skeg region (coupling of hull and skeg is not correct, the section line describes both hull and skeg which is wrong... - See attached image problem1)

3) As far as it concerns the front region of main metasurface, i increased the number of interpolatedpointsinsurfacesirections from 15->75 but the result is the same (attached image problem2) and of course if i increase the corresponding number in frontmetasurface, the bow sections change/almost competely dearranged..

 

My main problem is the skeg region because i can get the right sections towards chineLow end if i just insert the a groupSurface as source at sections instead of Brep... Still, though i dont get why if i utilize Surfaces as input the sections are ok (at x 32m approx.) and if i use Brep the problem remains...

 

Thank you,

Cheers,

Chris

post-1088-0-99399900-1552389980_thumb.png

post-1088-0-81060800-1552389998_thumb.png

boat10.fdbc

Share this post


Link to post
Share on other sites

Dear Chris,

 

1. In that case, it is the parametrization of the curve that is important. For ruled surfaces it is always a good idea to have at least a somewhat similar parameter distribution on both rails.

2. That is because your Brep is not trimmed either.

3. It works for me - which CAESES version do you use?

 

I put the Brep together so you can understand better how this is supposed to work in CAESES and I also created a set of sections as I understand this is what you want to export.

 

post-933-0-70470100-1552401616_thumb.png post-933-0-96347000-1552401622_thumb.png

 

P.S There are filters on the bottom of the 3D view to turn on/of the sections / offsets / surfaces etc... so you don't have to turn off the section visualization manually

 

Best regards,

Heinrich

 

 

boat10.fdbc

Share this post


Link to post
Share on other sites

Good Afternoon mr. Heinrich,

Thank you a lot for your quick response.

 

1) ... That's the role of unit speed? To make it easier for the ruled surface between 2 curves with different length?

 

2) I see, yes now it is ok. But how did u refer at the Hull Brep to cut and trim Skeg Brep? If i open at Hull Brep menu i see nowhere Skeg Brep as input....

 

   Sth else to mention: If i tried to implement the other way (the one mentioned at fastmonohull exemple) in order to cut n trim skeg the biggest problem i have is that the edges of the subsurface seem really different than the original ones. I increased the resolution properties to 500x500 but the same problem.. (see attached images prob3&4.png)

 

3) i use version 4.4.1. For me the problem remains...(see attached .probRemain.png).

 

Moreover, i want to ask for one more thing concerning my version. 50% of the times i try to open a second Caeses project i cannot. How can i resolve this? Because i would like sometimes to check 2 projects simultaneously and now i can say that possibility depends on chance... ?!?

 

Thank you very much,

Chris

post-1088-0-93910700-1552410724_thumb.png

post-1088-0-41887100-1552410738_thumb.png

post-1088-0-39564000-1552410744_thumb.png

boat_evolution10.fdbc

Share this post


Link to post
Share on other sites

Hi Chris,

 

1. The 'role' of unit speed is nothing more than parameterizing the curve such that the internal parameter t[0,1] runs with equal 'speed' per curve length. Since ruled surfaces create straight rulings between equi-parametric curve positions the parameterization allows you to massively influence the shape of the resulting suface - see example of two circles and the ruled surface in between for different parameterizations.

post-933-0-72114600-1552462683_thumb.png post-933-0-24413700-1552462689_thumb.png

 

2. Check the object editor of the Brep. The 3 steps I used are:

 

1. add sources (bare hull)

post-933-0-09223000-1552463013_thumb.png

2. project and trim (skeg contour)

post-933-0-77473500-1552463025_thumb.png

3. boolean sum (skeg)

post-933-0-12590000-1552463036_thumb.png

 

3. In the file you attached, the 'number of interpolated points in surface direction' for the meta surface main was set back to 15. I checked with 4.4.1 and it works for me.

 

 

Cheers,

Heinrich

Share this post


Link to post
Share on other sites

Hello mr.Heinrich,

 

Thank you for your detailed explanations.

 

You are right, i had it set back at 15 because nomatter what i tried yesterday did not wok out, cant tell why..

 

Anyhow, i tried again now and the result is fine. If i set the number of interpolted points in surface direction of MainSurface above 35 and leave that of FrontSurface at 15 everything seems to run smoothly.

 

Thank you very much for your up-to-now assistance,

Chris

Share this post


Link to post
Share on other sites

Hi,

 

I am new to CAESES. I wanted to create a surface of this type as shown in the attached figure (for simplicity is have attached simple geometry). Can you please tell me the option that I need to use to create such a surface. I have created the curves as shown in image 2

 

Thanks,

Anup

post-1106-0-39507800-1555381687_thumb.jpg

post-1106-0-73572200-1555381696_thumb.jpg

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...