CFD software options for Hull Resistance curves

[ram68ocean]

I am looking to obtain accurate hull resistance curves (kN vs. boat speed in knots) for a semi displacement hull of 12m (39ft) and speeds up to 12 knots.

What CFD software options are recommended without spending too much?

Orca3D seems to be best I have found so far. Any others?

 

[TANSL]

(kN vs. boat speed in knots)

Could you be more specific as to what you are really looking for because, at first glance, I do not see any relationship between the KNs values and the speed of the boat (although it is in knots)?
If what you need is to obtain the power-speed curve of a semi-displacement boat, you do not need any CFD software or anything like that. There are spreadsheets, even in this forum, with the Holtrop-Mennen method that would be perfect for your plans.

 

[DCockey]

I am looking to obtain accurate hull resistance curves (kN vs. boat speed in knots) for a semi displacement hull of 12m (39ft) and speeds up to 12 knots.

Could you be more specific as to what you are really looking for because, at first glance, I do not see any relationship between the KNs values and the speed of the boat (although it is in knots)?

I assume ram68ocean would like the resistance in kilonewtons (kN) as a function of the speed in knots.

What CFD software options are recommended without spending too much?

Orca3D seems to be best I have found so far. Any others?

How much is too much? How much time are you willing and able to devote to learning how to use CFD software?

The Orca3D Parametric Speed/Power Analysis module uses regression analysis, not CFD. The accuracy of the predictions depends on how closely the vessel fits within the range of intended uses.

Orca3D Marine CFD uses Simerics-MP (Multi-Purpose) CFD software, and more than doubles the cost of Orca.

 

[ram68ocean]

Could you be more specific as to what you are really looking for because, at first glance, I do not see any relationship between the KNs values and the speed of the boat (although it is in knots)?
If what you need is to obtain the power-speed curve of a semi-displacement boat, you do not need any CFD software or anything like that. There are spreadsheets, even in this forum, with the Holtrop-Mennen method that would be perfect for your plans.

kN = kilo Newtons (a force)
I am after the hull resistance (in kN) in calm water at different speeds up to 12 knots (1kn, 2kn, 3kn,...,12kn), in order to estimate the installed power requirements (basically selecting an engine or genset).
Power = Force * speed
Power_required = Force * speed * efficiency (I would typically start with a 50% efficiency)

Holtrop-Mennen method I don't perceive it as accurate as doing CFD. And it seems to be a method for larger ships. Here (link) shows it limits of applicability.
Have you used this method for smaller boats successfully?

I will go and read the other CFD similar threads. When I did earlier a forum search on CFD nothing was coming up... strange.

 

[Alik]

Holtrop-Mennen method I don't perceive it as accurate as doing CFD. And it seems to be a method for larger ships. Here (link) shows it limits of applicability.
Have you used this method for smaller boats successfully?

CFD can give any results, it is all about settings. I saw many CFD calcs giving 20-40% lower resistance.

Holtrop-Mennen - read carefully description of the method, better search for original publication(s). The authors clearly specified the applicability of the method. In my opinion, not perfect for small craft, however due to inclusion in Orca I saw some mistakenly use it for catamaran hulls

 

[TANSL]

@ram68ocean, my sincere apologies for such a stupid mistake. I am used to talking about towing horsepower in terms of HP (EHP, BHP, SHP), not kilo newtons, hence my mistake.
The Holtrop-Mennen method is applicable to boats between 9 and 80 m in length, therefore it could be valid for your boat.
I have used the method for small displacement and semi-displacemente boats (less than 24 m in length) and the results are satisfactory although it is not easy to validate them by sea trials. Experience can help to create corrective factors for the results obtained with the method (just a suggestion, not a statement).

 

[wet feet]

Its a long way above my capability but Open Foam is free and won't be beaten on price.

 

[Ad Hoc]

I am looking to obtain accurate hull resistance curves (kN vs. boat speed in knots) for a semi displacement hull of 12m (39ft) and speeds up to 12 knots....

How accurate is acceptable to you shall depend upon the route you take.
Nothing is an absolute...so you need to factor in your expectations and the actual data you have to hand.

Firstly on that length and speed, your Fn = 0.57....so that eliminates most if not all the "big ship" regression methods.
For many reason, too numerous to mention, as Alik also notes. But the obvious one being on that Fn the hull shape will be as you note "semi-displacement"...which large ocean going vessels are not!

I would suggest ignoring CFD, and the ideal basis for this is the Bailey NPL series; that data set will give you what you seek and as accurate as you need too.
You can find it HERE.

 

[ram68ocean]

How much is too much? How much time are you willing and able to devote to learning how to use CFD software?
The Orca3D Parametric Speed/Power Analysis module uses regression analysis, not CFD. The accuracy of the predictions depends on how closely the vessel fits within the range of intended uses.
Orca3D Marine CFD uses Simerics-MP (Multi-Purpose) CFD software, and more than doubles the cost of Orca.

Below $5000 ideally.
I was meaning indeed Orca3D Marine CFD https://orca3d.com/pages/orca-marine-cfd

How accurate is acceptable to you shall depend upon the route you take.
Nothing is an absolute...so you need to factor in your expectations and the actual data you have to hand.

Firstly on that length and speed, your Fn = 0.57....so that eliminates most if not all the "big ship" regression methods.
For many reason, too numerous to mention, as Alik also notes. But the obvious one being on that Fn the hull shape will be as you note "semi-displacement"...which large ocean going vessels are not!

I would suggest ignoring CFD, and the ideal basis for this is the Bailey NPL series; that data set will give you what you seek and as accurate as you need too.
You can find it HERE.

Thank you Ad Hoc for the Bailey NPL series, I wasn't aware of it. I found a direct download link to the pdf here: https://www.boatdesign.net/attachments/npl-pdf.93909/
I will need like a week to digest all that info though. You won't like my next question but do you know if there is a spreadsheet already available for the NPL method ?

So, my friend TANSL is happy to use Holtrop here but in your opinion Holtrop doesn't apply for this boat and NPL should give better results?
That is why I want to use CFD, and if possible more than one software to cross check.

I am not planning to become a CFD expert and I don't want to go into the mighty details of CFD. That's why at least from the distance the Orca3d CFD package looks the most user friendly option for people wanting the simplest workflow (input = hullshape and load conditions, output = hull resistance or power vs. speed).

Reference:
I have tried this spreadsheet https://www.boatdesign.net/threads/semi-displacement-resistance.57725/ but don't think it works well for this case.
Holtrop spreadsheets:
https://www.boatdesign.net/threads/holtrop-mennen.57771/
https://www.boatdesign.net/threads/holtrop-method.57755/

 

[Ad Hoc]

....You won't like my next question but do you know if there is a spreadsheet already available for the NPL method ?...

You will miss out on all the real learning and understanding of what you're doing, if you just go direct to an excel spreadsheet.
But no.. I dont have one. Just use the graphs... simple.

...Holtrop doesn't apply for this boat and NPL should give better results?
That is why I want to use CFD, and if possible more than one software to cross check....

Correct on NPL v HM.
As for CFD..how will you validate it??...you need a data set to validate it, otherwise it is just pretty colours.
ERGO... use NPL..

 

[TANSL]

So, my friend TANSL is happy to use Holtrop here but in your opinion Holtrop doesn't apply for this boat and NPL should give better results?

I don't know if I would use Holtrop "here" because I don't have enough data on your ship.
I have used the Holtrop-Mennen method in many projects and I am perfectly aware of its limits. As I see that you do not know them, I am attaching a table with them. As you can see, except that it not only works for big ships but also for not so big ones, I agree with what Ad Hoc says about limits. You said that it was not a method applicable to smaller ships and that is not correct. That is the only thing I had said so far, which is applicable for boats between 9 and 80 m. Of course, as you now know, there are other limits, but I think you did not refer to them because you did not know them. As for the accuracy of this method, or any other, to affirm something in that sense is a very delicate matter because the conditions in which measurements can be made in sea trials are very different from "laboratory" conditions.
And, yes, it pleases me that you treat me as "my friend". I also hold you in high regard.

 

[DCockey]

Holtrop-Mennen - read carefully description of the method, better search for original publication(s). The authors clearly specified the applicability of the method. In my opinion, not perfect for small craft, however due to inclusion in Orca I saw some mistakenly use it for catamaran hulls

Posted on the Rhino forum by Larry Leibman of Orca3D Welcome to Orca3D https://discourse.mcneel.com/t/welcome-to-orca3d/56539/5

The last item is a little more philosophical in nature. As you know the Holtrop method was not designed for multihull resistance prediction. We do have users sometimes use it for a catamaran like you are doing, but you must remember that there were no catamarans in the database of experimental results used to develop the Holtrop regression. Therefore I would view any results with a somewhat “suspicious eye” since things like hull wave interference effects and other physical effects are not captured. Having said that, if you are going to try to use the Holtrop for a catamaran the idea is that you are treating one hull of the catamaran as a vessel by itself, and then double the results to account for the other hull. Therefore, you need to take what appears to be the starboard hull in your model, and split it down the centerline and then only use the starboard half of that hull. Again, remember you will need to double the resulting resistance prediction to account for the other hull.

 

[Alik]

Posted on the Rhino forum by Larry Leibman of Orca3D Welcome to Orca3D https://discourse.mcneel.com/t/welcome-to-orca3d/56539/5

Good comment for dummy software users However, 'double' is not enough.

 

[DCockey]

Good comment for dummy software users However, 'double' is not enough.

He did say "As you know the Holtrop method was not designed for multihull resistance prediction ...... you must remember that there were no catamarans in the database of experimental results used to develop the Holtrop regression. Therefore I would view any results with a somewhat “suspicious eye” since things like hull wave interference effects and other physical effects are not captured."

 

[Alik]

He did say "As you know the Holtrop method was not designed for multihull resistance prediction ...... you must remember that there were no catamarans in the database of experimental results used to develop the Holtrop regression. Therefore I would view any results with a somewhat “suspicious eye” since things like hull wave interference effects and other physical effects are not captured."

David, that's clear, but why does not he recommend appropriate methods for catamarans? The student can read the original paper, program it and use... Not the way for software users, I understand. As a result, I see design package from ... where cat is calculated using Holtrop.