Doubts about Maxsurf and the Modeling (Motion Analysis) of a Boathouse.

[IngLaTorre]

Greetings colleagues! I hope you are doing well. I would like to inquire about a doubt I have regarding the hydrodynamic analysis of a "floating house (boathouse)". The house is prismatic in shape with dimensions of Length: 20 feet x Width: 20 feet x Depth: 15 feet. When I conduct the hydrodynamic analysis with Maxsurf Motion, I assume that both pitch and roll should yield "inversely proportional" values for wave directions of 0° and 90°. Let me explain a bit better: I assume, for example, that the pitch for the boathouse with a wave direction of 0° (facing the bow) should be > 0° and the roll should be = 0°... and in case the wave direction is 90° (from the beam), the pitch should be = 0° and the roll should be > 0°... This is considering that the boathouse lacks propulsion (V = 0Kn).

Note: I will provide a video of the simulation to show that under the same conditions, the behavior is not the same.

 

[jehardiman]

Most likely integrating or model stitching error, typical for this sort of thing. Have you ran the model with no input wave and checked the behavior? Did you give it a ramp-up or just hammer it with the wave? Note that MaxSurf is a surface element modeling method. The bow/stern and port/starboard panels must be mirrors of each other, not copies or rotations. If the element centers and normals are not exactly in line, a moment can be generated.

 

[IngLaTorre]

Most likely integrating or model stitching error, typical for this sort of thing. Have you ran the model with no input wave and checked the behavior? Did you give it a ramp-up or just hammer it with the wave? Note that MaxSurf is a surface element modeling method. The bow/stern and port/starboard panels must be mirrors of each other, not copies or rotations. If the element centers and normals are not exactly in line, a moment can be generated.

Hello @jehardiman, first of all, thank you very much for taking the time to respond to me. Yes, the problem occurs even with the lowest moments (0.5ft wave with periods of 2.5s) and they are particularly noticeable with a wave direction of 90° (roll)... I wanted to provide the previous example with 4.5ft waves to make the error more easily observable. As I mentioned before, the pitch and roll angles should be inversely proportional because it is essentially a "box". Regarding the model, I created it in the following way: Maxsurf Motion Advanced: Surfaces - Add shape - Box... Input: 20ft x 20ft x 15ft - Close Ends - Symmetrical - Half Model on the centerline. I don't see any reason for there to be any misalignment in the model.

 

[baeckmo]

So what happens when you use +45 degrees contra -45 degrees then, it is still a symmetrical body and should give equal answers from both directions.....?

 

[IngLaTorre]

So what happens when you use +45 degrees contra -45 degrees then, it is still a symmetrical body and should give equal answers from both directions.....?

Correct, they should yield similar results, but that's not the case. I made a "small workaround" to achieve symmetrical results, and that was by rotating the boathouse 45°, as I'm showing you in the attached image. However, the Pitch and Roll at the corners are still not accurate.

 

[jehardiman]

Correct, they should yield similar results, but that's not the case. I made a "small workaround" to achieve symmetrical results, and that was by rotating the boathouse 45°, as I'm showing you in the attached image. However, the Pitch and Roll at the corners are still not accurate.

I still think there is an asymmetry in the model, either in wave force calculation error or a nodal normal calculation. Try these things.

1) Print out a text of the nodes, every node should have its identical complement, even the bottom. Check the element normals, usually they follow the right hand rule so the node order listed for the element is important.
2) What is the shape and size of your mesh? Try using a smaller mesh, especially near (i.e. a wave amplitude above and below) the waterline so the wave pressures are better defined.
3) Have you done your hydrostatics? What is the model calculated LCB and TCB? What are your assigned LCG and TCG? Is there sufficient GMl and GMt for the wave selected? Have you calculated natural heave period? Do you have a phasing issue?

 

[IngLaTorre]

Could you please provide an email address for me to send a video of my procedure to?

 

[IngLaTorre]

I still think there is an asymmetry in the model, either in wave force calculation error or a nodal normal calculation. Try these things.

1) Print out a text of the nodes, every node should have its identical complement, even the bottom. Check the element normals, usually they follow the right hand rule so the node order listed for the element is important.
2) What is the shape and size of your mesh? Try using a smaller mesh, especially near (i.e. a wave amplitude above and below) the waterline so the wave pressures are better defined.
3) Have you done your hydrostatics? What is the model calculated LCB and TCB? What are your assigned LCG and TCG? Is there sufficient GMl and GMt for the wave selected? Have you calculated natural heave period? Do you have a phasing issue?

Hello @jehardiman, thank you once again for your kind response. I have answered your questions with numbered images in the same order as the questions... I can see that the model is symmetric, yet I haven't achieved the desired results. Could you please send me an email to inglatorre2013@gmail.com so that I can share a short video with you about the procedure I am applying? Alternatively, if you can assist me here on the forum, I would greatly appreciate it. Best regards.

 

[jehardiman]

Hello @jehardiman, thank you once again for your kind response. I have answered your questions with numbered images in the same order as the questions... I can see that the model is symmetric, yet I haven't achieved the desired results. Could you please send me an email to inglatorre2013@gmail.com so that I can share a short video with you about the procedure I am applying? Alternatively, if you can assist me here on the forum, I would greatly appreciate it. Best regards.

If figures 1 and 3 is the actual data you are using, you have a serious analysis problem....and maybe not just with MaxSurf and the the mesh model. You need to do your weights and hydrostatics. The control points show that the projected wetted element is assumed to be 4ft high, yet the draft is only 0.475m (1 ft - 5/16") so you are not covering the whole element, and may actually be slamming the hull (i.e. the wave amplitude could be greater than the draft). Additionally KG =0 and the BM's are greater than the GM's by 0.372m, which is an obvious error as GM = KB+BM-KG making the values an impossibility.

Read this thread and post for my opinion right now.
In 10 years?? https://www.boatdesign.net/threads/in-10-years.625/page-2#post-923011

 

[IngLaTorre]

If figures 1 and 3 is the actual data you are using, you have a serious analysis problem....and maybe not just with MaxSurf and the the mesh model. You need to do your weights and hydrostatics. The control points show that the projected wetted element is assumed to be 4ft high, yet the draft is only 0.475m (1 ft - 5/16") so you are not covering the whole element, and may actually be slamming the hull (i.e. the wave amplitude could be greater than the draft). Additionally KG =0 and the BM's are greater than the GM's by 0.372m, which is an obvious error as GM = KB+BM-KG making the values an impossibility.

Read this thread and post for my opinion right now.
In 10 years?? https://www.boatdesign.net/threads/in-10-years.625/page-2#post-923011

Dear @jehardiman, I am still reviewing the case, and indeed, the value of KG that the software is giving me is equal to zero (KG = 0). However, I do not understand why it is giving me this unusual value, nor do I understand how to correct it in the software. The VCG of the vessel is 2ft in lightship, and this should automatically change the value of KG (which is not happening).

Regarding the model, I have adapted it to perform the analysis using the "Analysis Type - Panel Method." In this method, I have added several nodes with the geometry of the boathouse and applied a "Marker - Generate TriMesh Surface from..." I have created two components, one with a height of 2.5ft (which is the maximum draft) and another from 2.5ft to the maximum sail surface.

Could there be a problem with the program, or am I doing something wrong in the model? As I mentioned before, I have the Weight, CG, and Draft data, which are the basic parameters required by the Maxsurf Motion software.

 

[baeckmo]

Since you are convinced that the wetted "hull" shape is completely symmetrical, I'm wondering about the weight distribution of the two components you have created. Are the masses also symmetrical, or are they distributed differently in longitudinal and transverse directions? Even if the average CoG of the additional components is located in a symmetrical plane, the inertia can differ in two directions, resulting in different wave response in spite of the symmetrical wetted hull.

 

[IngLaTorre]

If figures 1 and 3 is the actual data you are using, you have a serious analysis problem....and maybe not just with MaxSurf and the the mesh model. You need to do your weights and hydrostatics. The control points show that the projected wetted element is assumed to be 4ft high, yet the draft is only 0.475m (1 ft - 5/16") so you are not covering the whole element, and may actually be slamming the hull (i.e. the wave amplitude could be greater than the draft). Additionally KG =0 and the BM's are greater than the GM's by 0.372m, which is an obvious error as GM = KB+BM-KG making the values an impossibility.

Read this thread and post for my opinion right now.
In 10 years?? https://www.boatdesign.net/threads/in-10-years.625/page-2#post-923011

Dear @baeckmo,

Thank you very much for your kind response. The boat is supposed to be entirely symmetrical in "lightship" condition. Therefore, its LCG and TCG = 0. As I mentioned earlier, the VCG = 2ft. Thanks to the input from my friend @jehardiman, I realized that I was using the wrong method (Strip Theory) and now I am performing the calculations using the "Panel Method"... With this method, I have obtained much more accurate and symmetrical values regarding the wave direction angles of 0° and 90°. However, I have noticed something that caught my attention, which is that the KG is showing as 0... And I am unsure how to correct it considering that the VCG is 2ft.