Compound Curvature in Aluminum plate

[fpjeepy05]

.3dm file attached.

Single curvature with one principal curvature zero. Ruling lines are straight.

I meant for the sections.

Any idea how close the finished boat will be to the design you provide?

Your guess is as good as mine.

 

[Mr Efficiency]

28x5ft semi-displacement eco side console fishing boat. Personal use. Mostly calm seas.

Maybe you can find something along the lines of an Atkin sea-bright skiff for ply construction, that you can adapt to alloy, that will satisfy your requirements.

 

[fpjeepy05]

Maybe you can find something along the lines of an Atkin sea-bright skiff for ply construction, that you can adapt to alloy, that will satisfy your requirements.

Maybe, but I think I would rather just design from scratch myself. It's hard to adapt plans when my intended beam is almost half that of other boats it's size.

 

[fpjeepy05]

.3dm file attached.

Single curvature with one principal curvature zero. Ruling lines are straight.

So I'm out of the office right now, so I can't get on Rhino, and your model won't load on the iOS app. But I was able to upload to Grabcad and view there.

I was missing a few things. I thought the ruling lines were defining individual panels. But there is actually all one panel.

So you kind of cheated. You made the edge curves in such a way that the surface has no twist (I'm trusting you that it doesn't). Which is good design. I also see now what you were saying about the curved sections. It's most noticeable near the keel at the transom. It's not a simple arc between the chine and keel.

I was thinking of the opposite approach. Designing a hull with lots of twist, then chopping up the paneling into small enough pieces that they were developable and tangent discontinuities were small enough to have little effect, that the panels were close enough to being developable that they could be twisted into place, or some combination of the two.

Are you drawing ruling lines by hand, then checking twist, then manipulating edge curve control points then rechecking? That seems like a very tedious job that also requires a large understanding of how to produce developable surfaces. There are programs that make it easier...

 

[DCockey]

No "cheating" but some experience.

You made the edge curves in such a way that the surface has no twist (I'm trusting you that it doesn't).

The keel, chine and sheer curves were created with the desired shape of the curves. No twist in the surface was the result of the process used to create the surface, not (repeat not) because of any special properties or manipulation of the edge curves.

I was thinking of the opposite approach. Designing a hull with lots of twist, then chopping up the paneling into small enough pieces that they were developable and tangent discontinuities were small enough to have little effect, that the panels were close enough to being developable that they could be twisted into place, or some combination of the two.

Similar to approaches traditionally used in manual drafting for metal hulls. Also a method for building a hull - piece it together from individual pieces which are small enough to "force" to the desired shape.

Are you drawing ruling lines by hand, then checking twist, then manipulating edge curve control points then rechecking? That seems like a very tedious job that also requires a large understanding of how to produce developable surfaces. There are programs that make it easier...

No.

Keel, chine and sheer curve control points were not manipulated during creation of the surfaces. These curves were created and then the surfaces were created using those curves.

For this surface I used the Rhino V7 WIP command DevLoft. The ruling line determination method I described in the Rhino Forum thread linked above was used to determine the portions of the keel, chine and sheer curves which uniquely determine developable surfaces. In earlier versions of Rhino for Windows I have used the commands DevSrf and Loft with developable option.

To determine the portions of the keel, chine and sheer curves which uniquely determine developable surfaces I used created the no-twist ruling curves from the aft end of the chine curve to the keel and sheer curves and from the forward end of the chine to the sheer. The sheer and keel curves were then split using the ruling lines. The keel as also split where it became straight. (I knew this needed to be done based on experience.) The forward portion of the bottom surface was created using DevLoft with the chine curve, and forward part of the keel curve inputs. The center part of the side surface was created using DevLoft with chine curve and center part of the sheer curve as inputs.

Next the surfaces were temporarily extended, and the stem and transom edge curves were created to be curvature continuous to the extended surfaces. (The only requirements were the curves be curvature continuous with the extended surfaces and end at the ends of the sheer curve. Otherwise these edge curves can be designed as desired which obviously affects the shape of the surface.) The new stem curve and and forward part of the sheer curve were used as inputs to DevLoft to create the forward part of the side surface. The new transom side edge curve and aft part of the sheer curve were used as inputs to DevLoft to create the aft part of the side surface. The new transom bottom edge curve and middle part of the keel curve hwere used as inputs to DevLoft to create the aft part of the bottom surface. (Note that there are alternative ways of creating the portions of the bottom and side surfaces which are not uniquely determined by the keel, chine and sheer curves.)

The end result are no-twist developable surfaces which are curvature continuous and match the input keel, chine and sheer curves.

 

[fpjeepy05]

No "cheating" but some experience.

The keel, chine and sheer curves were created with the desired shape of the curves. No twist in the surface was the result of the process used to create the surface, not (repeat not) because of any special properties or manipulation of the edge curves.


Similar to approaches traditionally used in manual drafting for metal hulls. Also a method for building a hull - piece it together from individual pieces which are small enough to "force" to the desired shape.


No.

Keel, chine and sheer curve control points were not manipulated during creation of the surfaces. These curves were created and then the surfaces were created using those curves.

For this surface I used the Rhino V7 WIP command DevLoft. The ruling line determination method I described in the Rhino Forum thread linked above was used to determine the portions of the keel, chine and sheer curves which uniquely determine developable surfaces. In earlier versions of Rhino for Windows I have used the commands DevSrf and Loft with developable option.

To determine the portions of the keel, chine and sheer curves which uniquely determine developable surfaces I used created the no-twist ruling curves from the aft end of the chine curve to the keel and sheer curves and from the forward end of the chine to the sheer. The sheer and keel curves were then split using the ruling lines. The keel as also split where it became straight. (I knew this needed to be done based on experience.) The forward portion of the bottom surface was created using DevLoft with the chine curve, and forward part of the keel curve inputs. The center part of the side surface was created using DevLoft with chine curve and center part of the sheer curve as inputs.

Next the surfaces were temporarily extended, and the stem and transom edge curves were created to be curvature continuous to the extended surfaces. (The only requirements were the curves be curvature continuous with the extended surfaces and end at the ends of the sheer curve. Otherwise these edge curves can be designed as desired which obviously affects the shape of the surface.) The new stem curve and and forward part of the sheer curve were used as inputs to DevLoft to create the forward part of the side surface. The new transom side edge curve and aft part of the sheer curve were used as inputs to DevLoft to create the aft part of the side surface. The new transom bottom edge curve and middle part of the keel curve hwere used as inputs to DevLoft to create the aft part of the bottom surface. (Note that there are alternative ways of creating the portions of the bottom and side surfaces which are not uniquely determined by the keel, chine and sheer curves.)

The end result are no-twist developable surfaces which are curvature continuous and match the input keel, chine and sheer curves.

I think I understand more clearly now. At what amount of twisting does the ruling line connecting the aftmost point of the chine no longer hit the keel line aft of the stem? I imagine the hull no longer looks nice once the ruling line from the aftmost point of the chine starts to climb up the stem. When does the belly in the sections start to get too much?

Once I'm back on the grid, I'll have to give this a try.

 

[DCockey]

There is zero twist along the ruling lines. The location along the keel of the zero twist ruling line from the aft end of the chine is unique and a function of the shape of direction of a tangent to the chine curve at the end of the chine and the shape of the keel curve. There is a plane which is tangent to both the chine and the keel and the ends of the ruling line between them.

I imagine the hull no longer looks nice once the ruling line from the aftmost point of the chine starts to climb up the stem.

That would require a very strange stem shape.

I don't select the location of the ruling lines. The curves determine the location of the ruling lines.

When does the belly in the sections start to get too much?

That depends on numerous factors including asthetics and ideas about how hydrodynamics work.

 

[fpjeepy05]

There is zero twist along the ruling lines. The location along the keel of the zero twist ruling line from the aft end of the chine is unique and a function of the shape of direction of a tangent to the chine curve at the end of the chine and the shape of the keel curve. There is a plane which is tangent to both the chine and the keel and the ends of the ruling line between them.

That would require a very strange stem shape.

I don't select the location of the ruling lines. The curves determine the location of the ruling lines.

That depends on numerous factors including asthetics and ideas about how hydrodynamics work.

I'm sorry I'm using the wrong terminology. Maybe "warp" is a better term. I guess I will have to learn more by playing with it myself.

Basically, the point I'm driving at is there are shapes that are not developable. I.e a variable deadrise hull with straight sections. A long narrow gentle sweeping variable deadrise hull starting at 20 degrees and flattening out to 10. I think could be "converted" to a developable surface hull, and the result would look and perform very similar. By contrast a short fat hull with aggressive inflections in chine and shear that starts with an 80 degree plumb bow and warps to a 0 degree stern with lots of flare and tumblehome, might not look as nice after being "converted" to a developable surface. Maybe I am wrong. But I guess I was trying to learn in what ranges things place nice and when things start to get wonky?

 

[DCockey]

Two different approaches to creating a developable hull shape:

a) Create an arbitrary hull shape, and then try to "convert" it to a developable shape. Depending on the initial hull shape the developable version may be very similar to the original, or major compromises maybe required (changing a round bilge shape to multi-chine for example).

b) Start the process keeping the limitations of developable shapes in mind. For example flare is generally impossible with a developable shape. This may require considerable experimentation aka "trial and error" to learn what can and cannot be done with developable shapes. For instance in the hull shape I created for this thread only the aft portion of the keel is straight. That was built in from the beginning of the design. Also the stem is raked. For a non-developable design I might keep the keel straight for most of it's length, and have an almost vertical stem.

 

[fpjeepy05]

Two different approaches to creating a developable hull shape:

a) Create an arbitrary hull shape, and then try to "convert" it to a developable shape. Depending on the initial hull shape the developable version may be very similar to the original, or major compromises maybe required (changing a round bilge shape to multi-chine for example).

b) Start the process keeping the limitations of developable shapes in mind. For example flare is generally impossible with a developable shape. This may require considerable experimentation aka "trial and error" to learn what can and cannot be done with developable shapes. For instance in the hull shape I created for this thread only the aft portion of the keel is straight. That was built in from the beginning of the design. Also the stem is raked. For a non-developable design I might keep the keel straight for most of it's length, and have an almost vertical stem.

Agreed. I understand "converting" is not a good practice.

Also I think it is not good practice to not offer designs because they aren't easy enough for the construction method. Your plumb bow example. If a client wants a plumb bow on an aluminum boat, I should try to find a way to do it. It doesn't seem smart to reply to the client "We build out of aluminum. We do raked stems." Or if they want a boat with a 70 degree deadrise forward and 8 degrees aft. I can't tell them "We like surfaces to be developable. 20 degrees forward and 12 degrees aft is plenty good enough." They want those things because they have been proven to be beneficial on hulls of other construction types.

 

[DCockey]

Agreed. I understand "converting" is not a good practice.

Also I think it is not good practice to not offer designs because they aren't easy enough for the construction method. Your plumb bow example. If a client wants a plumb bow on an aluminum boat, I should try to find a way to do it. It doesn't seem smart to reply to the client "We build out of aluminum. We do raked stems." Or if they want a boat with a 70 degree deadrise forward and 8 degrees aft. I can't tell them "We like surfaces to be developable. 20 degrees forward and 12 degrees aft is plenty good enough." They want those things because they have been proven to be beneficial on hulls of other construction types.

Who are you refering to - designers or builders? It's easier and quick to design hull shapes which are not developable than are developable. A good designer will be knowledgeable about the construction methods they design for, both limitations and how to design for efficient construction. If a customer wants something which will significantly impact cost, performance or choice of builders a knowledgable designer might present several alternatives and the tradeoffs between them. For instance if the custome requests a straight, vertical stem and a warped bottom with straight sections the designer should let them know about added costs and any performance tradeoffs might be compared to a design with a curved, raking stem and some curvature in the bottom sections.

Similar for a builder. It usually does not work out well for anybody if a builder takes on a project very far outside of their compotency. A builder frequently can also provide good feedback on how to reduce the cost of potentially improve the quality of a design - for example by saying introducing come curvature in the bottom sections can signficantly reduce the amount of welding and provide a surface with less distortion.

 

[Mr Efficiency]

What size and type of power for this boat ?

 

[DCockey]

What size and type of power for this boat ?

I'm curious about how the size and type of power of a particular boat is related to a general discussion about designing hull shapes so that the shape is developable.

 

[Mr Efficiency]

He has mentioned the dimensions he is interested in, and talks of a "semi-displacement" hull, I want to know what he intends to power it with.

 

[fpjeepy05]

He has mentioned the dimensions he is interested in, and talks of a "semi-displacement" hull, I want to know what he intends to power it with.

I apologize I guess I was referencing it in this thread Mr. E.

I would love to power it with a little straight shaft diesel, but its not likely in the cards. I am thinking about importing a diesel outboard. Otherwise, it will just be a 50hp gas.