Godzilla vs. Michlet

[LP]

I'm mucking my way through running Michlet in Godzilla mode and need some expertise on the subject.

I've taken the gkayin.mlt file as a template and modified the contents for for 7.6 L X 2.6 B X 0.6 T meter hull. I'm using the 7 parameter function to generate the hull form and would like clarification of the shape function.

To digress though, in the single parameter series, the range of the function is 0.0-1.0. (0.0=square. 0.5=elipse. 1.0=parabola.)

In the 7 parameter function, the variable range is 0-3. A comment in the text states that this allow for hollows in the hull shape without further description. (that I've seen) I would surmise that 0-1 would be the same as before and 1-3 would progress into concavity.

Under "Shape Factors: Min, Max, Seed), the hull is defined. I can get Michlet to run if the first (5) MIN values are set to zero. In an attempt to push values towards a certain shape, I will change the MIN value to non-zero and end up with a "bad seed shape error" in the out.mlt file. I've been playing with the SEED values also and also get the the same error.

Also, after initiallizing Godzilla, my hulls don't appear to change and I'm not getting any hulls that within the fitness criteria. I don't think that I've overly constrained my hull parameters.

Questions:

Does having a smaller seed value in the shape factor area increase or decrease the likelyhood of finding a solution?

Am I misunderstanding the hull shape functions?

Is a non-zero value not going to work for a MIN value?

Is there an output file I can look at to clue me in on why my hulls don't meet the fitness requirements?

[marshmat]

In my experience with Godzilla, you can play with the parameter ranges all you want so long as you keep a few things in mind, among them
- Min must be smaller than Seed must be smaller than Max
- The initial hull as specified in the first part of the input file (the non-godzilla part) should fall within the ranges of the Godzilla parameters; I don't believe this is supposed to matter for the form parameters but it does seem to help
The input format for Godzilla can be pretty finicky and it can be trial-and-error to get it to run.
You say that after initializing the hulls don't appear to change.... does it seem to iterate really quickly while fitness stays at something like 1E7 ? This seems to be a fairly common error and I've manged to cause it using any number of little mistakes.
I'd be happy to look over the file that's giving you trouble or send you some that I know work OK; feel free to PM me on here.

[Leo Lazauskas]

Originally Posted by marshmat

In my experience with Godzilla, you can play with the parameter ranges all you want so long as you keep a few things in mind, among them
- Min must be smaller than Seed must be smaller than Max
- The initial hull as specified in the first part of the input file (the non-godzilla part) should fall within the ranges of the Godzilla parameters; I don't believe this is supposed to matter for the form parameters but it does seem to help
The input format for Godzilla can be pretty finicky and it can be trial-and-error to get it to run.
You say that after initializing the hulls don't appear to change.... does it seem to iterate really quickly while fitness stays at something like 1E7 ? This seems to be a fairly common error and I've manged to cause it using any number of little mistakes.
I'd be happy to look over the file that's giving you trouble or send you some that I know work OK; feel free to PM me on here.

I agree with most of what has been said here, except that it is not an error if the fitness stays at something like 1e7: that fitness is reported because the optimisation routine has not yet found a hull (or vessel) that satisfies *all* of the constraints.

If Marshmat can' help, send me the file and I'll have a look at it.

Regards,
Leo.

[LP]

Matt, Leo,

Thanks for the input. Learning new software is always a pain. This software has been on the back burner for almost a year.

I was looking at the seed parameter as a step function, not as an initial value. Of course, now it makes complete sense to call it seed. I'm also keeping an eye on the hull output to see if a value is falling outside parameter limits. Everything looks promising.

Questions:
In the documentation, there is a statement,

"The shapes of the hulls can be different from each other, however, they are assumed to be symmetric with respect to their own centreplanes. The overall vessel need not be laterally symmetric. "

Not being laterally symmetric, would that be referenceing a multihull vessel?

Any plans for incorporating heeled characteristics?

[Leo Lazauskas]

Originally Posted by LP

Matt, Leo,

Thanks for the input. Learning new software is always a pain. This software has been on the back burner for almost a year.

I was looking at the seed parameter as a step function, not as an initial value. Of course, now it makes complete sense to call it seed. I'm also keeping an eye on the hull output to see if a value is falling outside parameter limits. Everything looks promising.

Questions:
In the documentation, there is a statement,

"The shapes of the hulls can be different from each other, however, they are assumed to be symmetric with respect to their own centreplanes. The overall vessel need not be laterally symmetric. "

Not being laterally symmetric, would that be referenceing a multihull vessel?

Any plans for incorporating heeled characteristics?

1. Yes, individual hulls are symmetric wrt their own centreplanes. For multihulls the hulls can be placed in asymmetric arrangements, eg the example weinblum vessel.

2. Heel isn't much of an issue for my type of work, so I probably won't get around to it for quite a while.

Have fun with Godzilla!
Leo.