Some Nav Arch stuff

Sail4beer

Usual suspect
10,366
3,674
Toms River,NJ
That was very interesting, thanks for posting! Didn’t know how they scaled the hulls down and slowed the video to analyze the date, now I do.
 

Zonker

Super Anarchist
10,133
6,319
Canada
Thanks for posting this. Had no idea of the size of the wave pool there other than "it's big".

When UBC had a towing tank and wave pool the wave pool was more often used for storm scenes in movies (and made more money than towing tests). They would drape the walls with black plastic to make them dark and have hoses to fire rain and extra water at the actors.

bgytr - do you think there is still a case to be made for towing tanks or has CFD greatly reduced its utility? I know in our firm we can do stuff in CFD that we couldn't do in a tank (at lots less cost). We don't go the tank unless a client insists.
 

bgytr

Super Anarchist
5,072
676
Thanks for posting this. Had no idea of the size of the wave pool there other than "it's big".

When UBC had a towing tank and wave pool the wave pool was more often used for storm scenes in movies (and made more money than towing tests). They would drape the walls with black plastic to make them dark and have hoses to fire rain and extra water at the actors.

bgytr - do you think there is still a case to be made for towing tanks or has CFD greatly reduced its utility? I know in our firm we can do stuff in CFD that we couldn't do in a tank (at lots less cost). We don't go the tank unless a client insists.
I'm not an expert on cfd- I had peripheral involvement in some tasks up to the end of my career where cfd was used. We did comparisons of cfd vs. tank results. It has been some years since I have done any comparisons so I can't make a comment on how things have progressed.
My last experience was that cfd was pretty good at relative comparisons, but was sometimes inaccurate as far as absolute numbers depending on initial conditions of input. The case for tank testing is a relative expense and to continue refinement of numerical methods. If you consider the cost of one warship, and further a fleet of warships, methodical tank testing is a drop in the bucket as far as initial investment, with a huge tangible benefit of getting accurate results.
For calm water resistance and powering, cfd might be good enough, but again, I'm not an expert- my concentration was structures and seaway loads. We did a lot of tank testing to measure seaway loads during my career. We made scaled back-spline models to measure primary loads. Sometimes those models had scaled structural panels to measure equivalent static pressure secondary loads on local structures. The overall goal in my group was getting seaway loads for boat and ship structural design.
The cfd for seaway type analysis was very tricky and slow- it actually took longer to run through a time slice than to do the test in real life, and runs often crashed. Of course that is always improving with better computers, improved software, and user experience.
So for big ticket investment, I'd say the tank tests will be worth it for some time to come, especially when out of the realm of experience concepts are put forth, such as DDG1000, etc.
 

Zonker

Super Anarchist
10,133
6,319
Canada
Thanks for your thoughts and sharing more detail. Yeah, in the scope of a 1B warship (or series of them) it makes total sense.

My world is workboats that are generally in the 1-20M range, so a tank test program is often not justified; mostly because absolute lowest resistance isn't a big goal. Fewer tanks out there too, so getting time in the tank is harder. We generally use Vienna because it's small and cheap, and Memorial in NL for bigger programs.

We do a fair bit of seakeeping / motions simulations with CFD.

One area that proved fruitful was dynamic instability of quite fat tugboat hulls - and how to fix it.

We've also simulated self righting lifeboats self righting behaviour. The video of the full scale test of the boat being capsized by a crane and then released was amazingly similar to the CFD - even to the extent of water captured inside the flying bridge structure.

And turning a barge train in S. American pushed by a shallow river tug with 3 Z-drives. We had to show authorities a longer barge train could make a turn.

Then a slow speed waterjet vessel where we modelled the flow inside the waterjet duct (quite draggy we found) because the vessel was only doing 12 knots and the w.j. people didn't have very good thrust data at low speeds like that.

We've spent many $100,000's on computers. The air conditioner for the small server room is something like 15 or 20 kW to give you a feel for the power they are sucking down. Several people just dedicated to using it and we keep adding more. Nice to have it in-house but we use capacity up as fast as we add it.
 

bgytr

Super Anarchist
5,072
676
Thanks for your thoughts and sharing more detail. Yeah, in the scope of a 1B warship (or series of them) it makes total sense.

My world is workboats that are generally in the 1-20M range, so a tank test program is often not justified; mostly because absolute lowest resistance isn't a big goal. Fewer tanks out there too, so getting time in the tank is harder. We generally use Vienna because it's small and cheap, and Memorial in NL for bigger programs.

We do a fair bit of seakeeping / motions simulations with CFD.

One area that proved fruitful was dynamic instability of quite fat tugboat hulls - and how to fix it.

We've also simulated self righting lifeboats self righting behaviour. The video of the full scale test of the boat being capsized by a crane and then released was amazingly similar to the CFD - even to the extent of water captured inside the flying bridge structure.

And turning a barge train in S. American pushed by a shallow river tug with 3 Z-drives. We had to show authorities a longer barge train could make a turn.

Then a slow speed waterjet vessel where we modelled the flow inside the waterjet duct (quite draggy we found) because the vessel was only doing 12 knots and the w.j. people didn't have very good thrust data at low speeds like that.

We've spent many $100,000's on computers. The air conditioner for the small server room is something like 15 or 20 kW to give you a feel for the power they are sucking down. Several people just dedicated to using it and we keep adding more. Nice to have it in-house but we use capacity up as fast as we add it.
Ya for smaller applications, cfd totally makes sense, compared to a 1 to 2 mil tank test program.
And there were some very cool simulations the last few years of my career, I bet they continue to get way better since I have retired. Sounds like you guys have stepped up the computing and get to work on some cool stuff!
 

sam_crocker

Super Anarchist
1,534
87
PNW
I'm not an expert on cfd- I had peripheral involvement in some tasks up to the end of my career where cfd was used. We did comparisons of cfd vs. tank results. It has been some years since I have done any comparisons so I can't make a comment on how things have progressed.
My last experience was that cfd was pretty good at relative comparisons, but was sometimes inaccurate as far as absolute numbers depending on initial conditions of input. The case for tank testing is a relative expense and to continue refinement of numerical methods. If you consider the cost of one warship, and further a fleet of warships, methodical tank testing is a drop in the bucket as far as initial investment, with a huge tangible benefit of getting accurate results.
For calm water resistance and powering, cfd might be good enough, but again, I'm not an expert- my concentration was structures and seaway loads. We did a lot of tank testing to measure seaway loads during my career. We made scaled back-spline models to measure primary loads. Sometimes those models had scaled structural panels to measure equivalent static pressure secondary loads on local structures. The overall goal in my group was getting seaway loads for boat and ship structural design.
The cfd for seaway type analysis was very tricky and slow- it actually took longer to run through a time slice than to do the test in real life, and runs often crashed. Of course that is always improving with better computers, improved software, and user experience.
So for big ticket investment, I'd say the tank tests will be worth it for some time to come, especially when out of the realm of experience concepts are put forth, such as DDG1000, etc.
The Sailing Yacht Research Foundation did a project called the Wide-Light Project which compared a modern hull (Wide and Light :) ) in both a tank and in four CFD programs to compare the results. I don't recall the final results exactly but I think it still matched what you found - accurate for comparison purposes but not exact, especially when things like transom immersion come into play. Just thought I'd throw that out there for you design guys.....
 

Attachments

  • Widelight-from-Seahorse-Feb2016-low.pdf
    1.2 MB · Views: 23

Latest posts




Top