Godzilla and hard chines

[Kudzu]

YES, I have read the manual! Have a printed copy on my desk. This may be there but if so, I have not grasped it yet.

Just started playing with Godzilla last night and starting to understand a bit about it. I have a lot to learn obviously. I am designing kayaks and hard chines, some times multiple chines. Looking at the Godzilla manual I didn't understand how to specify it keep a hard chine. Can it?

If so if you just point me in the general direction I will dig a little deeper.

[marshmat]

Hi Kudzu,

To the best of my knowledge, the current version of Godzilla can work by rescaling a table of offsets you provide, or it can create its own hulls using hull shape series 1, 7 or 8. I don't think any of these series support hard chine craft.

You can get reasonably close results for multi-chine craft as follows:
- Perform an optimization with Godzilla for the target displacement, length, speed, etc.
- Copy ship_output_by_speed.mlt into a spreadsheet file (.mlt output tables are encoded in csv format)
- Use off1.mlt or fsoff1.txt to get Godzilla's generated offsets into the hull modeller of your choice- Freeship works well for this purpose.
- Add your chines, make sure the displacement is the same as the original, and export the new offsets to Michlet. Important note- if there is a transom that extends below the waterline, make sure that it shows up in the final line of the offsets table! If the transom shows up in the second-last line and the last line is all zeros, Michlet won't recognize the transom and the results will be off. (Zeros in the last line are OK for a double-ender.)
- Run those offsets in normal Michlet mode, and copy that ship_output_by_speed.mlt to the same spreadsheet as before. Plot the resistance curves on the same graph to verify any discrepancy.

The presence of chines doesn't screw up Michlet, but the mathematical series used by Godzilla generate smooth curves only. As long as you don't screw with the waterplane shape, the curve of areas, etc. the presence of chines does not usually make a particularly significant difference to the overall resistance curve.

[Guest625101138]

Originally Posted by Kudzu

YES, I have read the manual! Have a printed copy on my desk. This may be there but if so, I have not grasped it yet.

Just started playing with Godzilla last night and starting to understand a bit about it. I have a lot to learn obviously. I am designing kayaks and hard chines, some times multiple chines. Looking at the Godzilla manual I didn't understand how to specify it keep a hard chine. Can it?

If so if you just point me in the general direction I will dig a little deeper.

I am assuming you are using the version with 7 shape functions. I believe the second one controls the hull X-section. If you set the upper limit to 0, the same as the lower limit, then you will force a hard chine. Obviously the seed also needs to be zero.

Normally if I am after a developable shape I set the upper limit around 0.2. This gives a slight curve and can be approximated with sloping sides and a single chine or more accurately with a double chine. You have to remember that Michlet is only valid, in a strict sense, where the sides of the hull are perpendicular to the water surface. In practice it really does not matter if they are angled somewhat. Point is Godzilla will force the perpendicular side at the surface but when you render the hull for development you can make the sides angled.

The starting hull will not make any difference to the final result as Godzilla will take it to the optimum and it usually arrives at the same place irrespective of where it starts. Leo advises that there will be two optimums but I have not seen this occur.

Rick W

[marshmat]

I've never encountered the 'double optimum' issue either, Rick.

Doesn't setting f1 = 0 force a rectangular cross section? (I guess that would be hard chine.... so then yes, I suppose you could say that it supports chine hulls, in a sense.) But then it can't adjust the cross sectional shape. Restricting f1 to some small value would probably give something that could more easily be approximated by a chine hull, as Rick says.

[Guest625101138]

Originally Posted by marshmat

I've never encountered the 'double optimum' issue either, Rick.

Doesn't setting f1 = 0 force a rectangular cross section? (I guess that would be hard chine.... so then yes, I suppose you could say that it supports chine hulls, in a sense.) But then it can't adjust the cross sectional shape. Restricting f1 to some small value would probably give something that could more easily be approximated by a chine hull, as Rick says.

Matt
f1 and the second shape factor are the same thing. Leo numbers from f0.

The X-section will change because the draft and beam can be altered. Typically a hull with unconstrained length and low displacement like I play with will have beam to draft of about 1.4. If you want a high speed hull and heavy displacement it will likely give greater draft than beam - very slender but deep hulls.

Rick W

[marshmat]

f1 and the second shape factor are the same thing. Leo numbers from f0.

I'm aware of that. When I set parameter f1 to zero, I get a rectangle. I just think that letting it choose the shape freely, then trying to approximate that shape with a double chine, is likely to come closer to optimum than if Godzilla isn't allowed to change the sectional shape as well as the beam/draught ratio.

[Guest625101138]

Matt
I always start with only displacement and speed to see what an uncompromised optimum will look like. If I want a developable hull I usually limit the range of f1 as noted above. This will usually give a narrower hull than the unconstrained one. This is for my trimaran set up with non load carrying outriggers so stability is not an issue.

I sometimes limit other shape factors to force rocker for example.

The reason for doing this is to know what each constraint costs in terms of performance. It is always good to have an appreciation of what the lowest drag hull looks like. Then determine how much you are prepared to pay for various compromises.

Rick W

[Kudzu]

I am designing in FreeShip BTW. I have read all this and will need to go back to manual and make sure I follow all of it. It's far from second nature to me yet. But importing it back in and creating chines makes sense.

I have a design we have been talking about that has some promise as a good all around boat for out waters. I would like to squeeze as much performance as possible and practical out of it.

I have michlet 8.07 is that the latest? Of course I guess I could just do a search and find out really quick. Thanks for the input. Will see what I do with this tonight and might have more questions.

Jeff

[Kudzu]

This is the most frustrating software I have ever worked with!!!!!!!

Now it tells me in the out file that my design speed is out of range. I can't even find anything about design speed in there. All the speeds are in range. I have tried this program a couple of times and I end up giving up every time.

[Guest625101138]

It is incredibly powerful software in a tight bundle. I can assure you it is much faster to use it at the start of the boat design process than building successive boats, with forward and backward steps, until you get something half as good as Godzilla will produce.

If you open the in.mlt file and scan down from the top of the page there is a section called CALCULATION PARAMETERS. You can set the speed range here. I usually set it to 1 to 6 or 1 to 11 and use 51 numbers of speed so I get nice even speed intervals. Data from Freeship imports from knots to m/s in Michlet so I always adjust these parameters to a nice range.

The second last parameter in the GODZILLA INPUT DATA is the design speed. This value has to be within the range set in the upper part of the in.mlt file.

If you have these set appropriately then post your in.mlt file and I will fix it.

You may find post #25 on this thread helpful in setting some structure to using Godzilla:
Racing Canoe

Another helpful hint is to add comments with the # in a standard in.mlt file for Godzilla. I have three in.mlt files for Godzilla. One for mono, one for cat and one for tri. I always start with a generalised hull based on the 3 shape function because it goes where it likes anyhow. If you over constrain it will not get started with the optimisation.

Rick W

[Kudzu]

This is strange. I decided to back up to basics. Let Free!ship create a new Michlet file. Saved it as the IN file and now it won't run at all. Even the one that was working last night won't run.

I thought maybe it was something to do with the hard chines vs the rounded one. Obviously not. I feel like I am beating my head against a wall.

Tell if you can see what is wrong. As I said, I know it's good software. I had it running last night. But it is the hardest thing I have ever used too.

[Guest625101138]

That file runs fine in Michlet. It will not run with the Godzilla switch because it has no Godzilla parameters.

I have attached the Michlet drag curve produced by your hull.

Rick W

[Kudzu]

Yes, I know it didn't have the Godzilla code.

Now this is strange. It will not run for me. I even rebooted just in case after I posted it. Now I am confused..... wait..........

Can I go hide under a rock now?? It was my desktop short cut that was the problem. I feel stupid. Wasted an hour on that! Ran perfect this time.

OK, will play some more. Still not sure I under all there is about Godzilla and I am sure I will have questions and Thanks for helping me find I was the problem.

[Guest625101138]

You can get some incredibly difficult conflicts that are hard to sort through once you start playing with constraints. Godzilla knows there is a problem but may not communicate well. Irrespective it is worth the trouble because building a dud boat is a wasteful exercise.

Rick W

[Kudzu]

WooHoo! Got it to finally run.
Thanks again. Now I can start playing.