CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   about boat simulation (http://www.cfd-online.com/Forums/main/16196-about-boat-simulation.html)

hasim January 12, 2009 09:12

about boat simulation
 
hello

i would like to ask some things about boat simulations, firstly if the software is good enought for the user to calculate the hydrodynamic drag and the results if there are good enought, it would be very good if i can have the advice of a boat expert analist to tell me if he made the simulation and then the real expiriment and how much % different have the results.

i also would like to ask if the people of the shipyards are using simulation software for the design of the hull, or using tables???

i would like to have answers from people that already have expirience, we are going to spent a loto of money on this .

best regards hasim

Bikini Girl with Machinegun January 12, 2009 10:43

Re: about boat simulation
 
CFD can't accurately predict the drag on a flat plate at high Reynolds numbers, so don't expect good accuracy for boats.

For a laugh, check what the CFD code uses to predict skin-friction. Some use the ITTC line because their own predictions are so bad.

How would you feel if you paid thousands of dollars for a drag result based on a friction line that was devised 50 years ago as an "interim solution"?

Of course, you'd get pretty pictures for your money, so it might be worth it to you.


hasim January 12, 2009 12:31

Re: about boat simulation
 
yes but how can i calculate the hydrodynamic drag if i know only the shape??? i know the material an the weigh and the center of the weight. so can you tell me please how these people calculate accurate the drag by table?????

thanks a lot

Bikini Girl with Machinegun January 12, 2009 12:54

Re: about boat simulation
 
Some naval architects use empirical formulas or tables if the ship they are building is very similar to previously built ships. For example, the Delft series for yachts.

What method you use depends a lot on the type of boat. Is it a displacement vessel, or a planing hull, or "semi-planing"?

If it is slender, you might be able to use slender body theory; if it is thin, you can use thin-ship theory. If the ship is very bluff, you can try a non-linear method, but it might not always give you an answer. CFD might give you an answer to +/-50%.

If it is a planing hull then you might be able to solve the appropriate equations numerically if the planform of the hull is made up of straight lines. If the planform is made up of curved lines, be very suspicious of Vortex Lattice-like Methods. They shit themselves on lifting surfaces with curved leading edges and trailing edges. Savitsky's empirical formula might give you a reasonable answer. Maybe not.

You sound like you are completely out of your depth on this. Is there a more simple project you could attempt?


Balduin Bankerotti January 12, 2009 14:42

Re: about boat simulation
 
There is at least something to read here.

http://www.cd-adapco.com/applications/marine.html


hasim January 12, 2009 17:14

Re: about boat simulation
 
well

i am mechanical engineer and work on fea more than 15 years, and also in cfd ..... but as you knoe the fea and cfd are more and more needed in more and more fields, that makes as running as tigers to make everything..... i am working with flow 3d , star ccm and cfd ace+ (in more than 3 comoanies at the same time) i am really confuse with al these informations..... i am seeing the car industries working more and more with optimization and fea and cfd (now more than ever) and from the other view i am wacthing the boat designers to work with tables......

of course balduin i study more than 20 times the star ccm with 6dof but i am still watching the people of flow 3d (fdm) and others companies people to advertizing their software but the people of boat design does not work with them, so that is meaning that or the people in the boat design does not know about cfd or the cfd are not good enouph for that....

we all knoe the power of optimization , a friend of mine from a company in greece that is working on cfd ace+ and a code names ioso from russia told me that we can make miracles on every kind of calculations but i still do not know why we do not do them !!!!!! if you want to design a great hull for a yacht then you can make the first design , apply the parameters in the optimization engine and after that just make the simulations steps until find the best solution. so do you all believe that the tables are more good? why? if you put a really boat in the sea you will take everyhour different results .... it depence from more than 2000 reasons why.

i really want to know why we are keeping thi technology out of boat design, and also if anyone from cd adapdo is here could tell us what is the accurancy of star ccm in the boat simulations.... because when you are design a 1 or 50 or 100 millions ship you have no right to make a mistake.... if you can save a unit od hydrodynamic drag with cfd why not use it to save????

thanks a lot for the answers.

Charles January 13, 2009 18:01

Re: about boat simulation
 
You can get ship drag a great deal more accurately with CFD than Bikini Girl thinks. You do need to know what you are doing, but there is no need to be quite as cynical as that.


Bikini Girl with Machinegun January 13, 2009 18:54

Re: about boat simulation
 
Just make sure that you have experimental results *before* you do your CFD runs. That way you can tweak your grids and other parameters to give good results. Never ever use CFD in a double-blind verification exercise. CFD doesn't work that way. ;-)


alex January 14, 2009 00:17

Re: about boat simulation
 
yep, just like everything else in life, you get experience and tweak stuff to get good results:) even your cherished "experiments"... what's up with a gun though?

Jeff Vader January 14, 2009 08:00

Re: about boat simulation
 
"Just make sure that you have experimental results *before* you do your CFD runs"

I'm not sure that these guys agree: http://www.cape-horn-eng.com/index.html

"At Cape Horn Engineering we are very proud to continue our work for America's Cup and Volvo Ocean Race campaigns. These latest projects rely solely on our CFD and show that tank tests are now unnecessary. Our references, know-how and resources speak for themselves. Find out more about what we can offer by visiting the links at the top of the page."

JV

hasim January 14, 2009 08:23

Re: about boat simulation
 
wait !!!!!!!!!!!!!!!!!!!

do not run so fast......

firstly hes of course we will make some test before building the real 1:1 boat but how are we sure that the expirimental are more accurate than cfd??? we can build a 1: 10 model not more , then the accurancy is 1:10 !!!!! that means 10% at the best results....

second which code are you using at Cape Horn Engineering????

thanks

Jeff Vader January 14, 2009 12:07

Re: about boat simulation
 
"Just make sure that you have experimental results *before* you do your CFD runs"

I'm not sure that these guys agree: http://www.cape-horn-eng.com/index.html

"At Cape Horn Engineering we are very proud to continue our work for America's Cup and Volvo Ocean Race campaigns. These latest projects rely solely on our CFD and show that tank tests are now unnecessary. Our references, know-how and resources speak for themselves. Find out more about what we can offer by visiting the links at the top of the page."

JV

Bikini Girl with Machinegun January 14, 2009 13:12

Re: about boat simulation
 
Hasim, Seriously, search for and read papers by Prof. Frederick Stern. He is an eminent ship hydrodynamicist who is very involved in the verification and validation of CFD in marine engineering.

A recent paper from his group is: Sakamoto, N., Wilson, R.V. and Stern, F., "Reynolds-averaged Navier-Stokes simulations for high-speed Wigley hull in deep and shallow water" J. Ship Research, Vol. 51, No. 3, Sept. 2007, pp. 187-203.

Spend some time reading papers from the ITTC Resistance groups and other special groups. They review the performance of many codes and techniques. Do your own research and don't believe the snake-oil salesmen here who tell you how wonderful their codes are.

Remember, CFD can't accurately predict the skin-friction on a flat plate at ship-scale Reynolds numbers. What accuracy can you expect for a real ship hull?

Consider this too: CFD codes do quite poorly on model-scale mathematically-defined hulls. It is therefore quite unlikely that they will do better with real hulls. If your proposed ship has a large immersed transom, then CFD will not be reliable.

If you want to see how poorly CFD does with transom sterns, look up papers by Kevin Maki, Robert Beck and Lawrence J. Doctors. The last two are very experienced ship hydrodynamicists who don't believe that CFD is a mature, reliable technique for simple applications like resistance prediction and/or the prediction of the surface behind transoms.


sagar.elias.jackey January 14, 2009 13:44

Re: about boat simulation
 
hi

hasim January 14, 2009 18:58

Re: about boat simulation
 
my dear bikini men....

i will follow your opinion... i saw many sales men telling me that this is good and best and other s$@# like this... when i tell that to see the software in one of my problems al of them disappear.... they are telling me that if the solver solves a plate then it can solves an aircraft... but as you know this is not correct , at least on some exersices.

yes the sales men are always press by their bosses to sell , but they do not consider what they are doing....

i try to solve a boat with 5 software that include free surface, 6dof, etc... everysolver gave me different results... the accurate between the max and min per solver results are almost 50% !!!!!

i do not want to have a solver telling me how the sailboat will turn..... the humanity know it from ancient greeks... i need a software telling me that this boat with this shape with this engine can reach 50 knots.... and with this wave it will make this rotations.... i believe that this is very simple.... if they can not solve it accurate then ... i do not want it. my boss will kill me if once the solve tells me that the boat will reach 50 knots and then 30!!!!! and he will have right because he spent more than 50 000 on software...

i saw the same almost with the fea codes , many software gives you accurancy 50%, only the last 5 years we are taking almost 10%....

i do not know if it is good or bad to coupling with cfd .... i believe that these are tools only to see where is the high pressure not to see the number of the pressure... i was looking the coupling of ansys with cfx.... when the ansys has an accurancy of 30% and the cfx accurancy of 20% then the accurancy is 20%*30* that mean.... nothing ... and they sell this..... jesus....

i will start afraid of the constructions that these people advertizing with the spot "" design with (code) it is safe"

hasim January 14, 2009 18:59

Re: about boat simulation
 
my dear bikini men....

i will follow your opinion... i saw many sales men telling me that this is good and best and other s$@# like this... when i tell that to see the software in one of my problems al of them disappear.... they are telling me that if the solver solves a plate then it can solves an aircraft... but as you know this is not correct , at least on some exersices.

yes the sales men are always press by their bosses to sell , but they do not consider what they are doing....

i try to solve a boat with 5 software that include free surface, 6dof, etc... everysolver gave me different results... the accurate between the max and min per solver results are almost 50% !!!!!

i do not want to have a solver telling me how the sailboat will turn..... the humanity know it from ancient greeks... i need a software telling me that this boat with this shape with this engine can reach 50 knots.... and with this wave it will make this rotations.... i believe that this is very simple.... if they can not solve it accurate then ... i do not want it. my boss will kill me if once the solve tells me that the boat will reach 50 knots and then 30!!!!! and he will have right because he spent more than 50 000 on software...

i saw the same almost with the fea codes , many software gives you accurancy 50%, only the last 5 years we are taking almost 10%....

i do not know if it is good or bad to coupling with cfd .... i believe that these are tools only to see where is the high pressure not to see the number of the pressure... i was looking the coupling of ansys with cfx.... when the ansys has an accurancy of 30% and the cfx accurancy of 20% then the accurancy is 20%*30* that mean.... nothing ... and they sell this..... jesus....

i will start afraid of the constructions that these people advertizing with the spot "" design with (code) it is safe"

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


hasim January 14, 2009 19:06

Re: about boat simulation
 
i also have an article in the flow 3d forum becaus ei do not have any ipinion about his software and i would like to have your comments also on this. one person tell me that the accuracy is more than 96 % , they make a lot of expiments and i believe that they might make a code ready for this. i will talk with him since he is more expert on the software to tell me also his opinio

Bikini Girl with Machinegun January 15, 2009 16:11

Re: about boat simulation
 
"These latest projects rely solely on our CFD and show that tank tests are now unnecessary. Our references, know-how and resources speak for themselves."

Yes, they do speak for themselves. And only themselves. They are advertising a product, therefore their self-assessment of their own wonderful, awesome, incredible, predictive powers are not to be believed.


Bikini Girl with Machinegun January 15, 2009 17:29

Re: about boat simulation
 
Only a complete idiot would believe any advertising material from a company trying to sell a product. Why even show us that crap?

I'll believe them when they show their failures along with their claims of awesome accuracy. If companies don't show when CFD fails to give good results then they are just marketing scum and unscientific charlatans.


solarwind January 19, 2009 20:21

Re: about boat simulation
 
Dear bikini girl, can you show us a few articles showing experiments failure? Or is that taboo?



All times are GMT -4. The time now is 19:44.