Some facts about ship surface modeling that won't be told to you.
Working with various customers and on projects of various types of ships, I constantly notice the same problems that arise when modeling the surface of the hull. Most of them relate more to organizational problems. I have a strong feeling that with the reconciliation of computers for ship surface design, we are losing some of the knowledge that many generations of shipbuilders used. This knowledge is just as necessary today, but over-reliance on computer models of the surface plays a trick on us over and over again. So, a few important facts about ship surface modeling.
1. A high-quality model of the ship's surface is a guarantee of high-quality assembly and welding of the ship's hull. Everyone understands this, but few people know how to check the manufacturability of a mathematical surface model. At most shipyards, the quality of surface smoothing is not controlled in any way. Unfortunately, there are almost no specialists left who were engaged in manual lofting and fairing of the ship's hull. The loftsmens knew everything about the ship's surface like no one else. With the transition to computer models, manual lofting knowledge has become a thing of the past, along with the knowledge that modern designers who model ship surfaces still need today.
2. 95% of all ship surface models that I have seen over the past 30 years of work had technological smoothing defects. The quality of computer smoothing of the ship's surface is very often much lower than if smoothing was applied manually on the molded floor. The designer blindly trusts the results of computer smoothing. Working with such surfaces, shipyards have problems when bending shell plates, assembling, and welding hull structures. Finding out the cause of these problems at the stage of assembling the ship's hull is quite problematic. Shipyards that are serious about surface smoothing can save a lot of time and money when assembling and welding the hull. Only welding can save up to 10% of the time, while the build quality of the case is much higher.
3. It is very difficult to identify smoothing errors at the stage of issuing working documentation for the hull since the work is distributed over sections of the ship's hull and does not make it possible to assess the shape of the entire surface as a whole. Correction of the identified errors is often impossible since some of the sections to be changed may have already been sent to the shipyard and cut out in metal. Many companies involved in the production of working documentation do not pay due attention to the quality of the ship's surface. The verification of the ship's surface model is started only when the hull structure modeling programs begin to generate errors in the modeling of parts and hull shell plates.
4. For modeling and smoothing the surface of the hull, programs not intended for this are often used. Often these are programs for modeling classical surfaces used in mechanical engineering. Even automotive CAD is not always a good choice for hull surface modeling. Very often, free or shareware programs are used to model the surface of the hull. Many companies believe that this is enough to build a good hull surface or simply save money without thinking about the result.
5. Modeling and surface smoothing are not taught anywhere. Talking with students at various universities, I found out that no attention is paid to the modeling of the surface of the hull. Often the teacher recommends a particular program and sends students to the software manuals. At best, these are several hours of lectures describing the method invented by the teacher and using the simplest possible program. As a result, students can only make a primitive crumpled hull that is only good for hydrostatic calculations. Subsequently, working in a design company, they continue to use the acquired “skills”.
6. The technology of smoothing the surface of the ship's hull is very dependent on a particular program used. As a rule, without a course of study from a software developer, it is very difficult to master a particular method on your own. Very often, as a result of self-learning of the program, such models appear that consist of randomly located patches of surfaces. This model shows how incorrect partitioning on surfaces leads to catastrophic results. After the designer understands that everything is bad, and the deadlines for the delivery of the surface cannot be transferred, panic begins, plugging holes and new patches on the old ones. Such a model, in addition to obvious smoothing errors, is overloaded with excessive surface areas, an excessive number of surface control points, and gaps between surface areas. There is no need to talk about quality in this case.
7. Modeling the ship's surface is significantly different from building any other three-dimensional models. When modeling, it is important to have not only a good understanding of the mathematical apparatus of the program, but also to know the rules for constructing a classical ship surface. I often observe the neglect of these rules on many new ship’s hulls. The absence, for example, of a dedicated flat side and a flat bottom is clearly visible on the quality of the hull skin. Often such a surface is modeled not by a shipbuilding engineer, but by a designer who knows how to work with the program and uses the simplest modeling techniques to quickly get the result. As a result, the quality of smoothing the surface of the ship's hull suffers.
8. Often, when transferring the next smoothed hull, customers ask me: “What program was this model made in?”. Undoubtedly, surface smoothing programs should have functionality sufficient to obtain a high-quality surface model. At the same time, the designer who works with the program must be able to do this and have the necessary set of knowledge. The acquisition of a particular program will not solve problems if the specialist is not experienced enough. It's like choosing a piano for a beginner pianist.
9. The designer's lack of fundamental knowledge about the surfaces being modeled leads to the fact that he "agrees" with the surface option that the program offers. This is clearly seen in the models sent to me by various customers. In such cases, there is practically no talk of designing the ship's contours - it would be possible to stretch the surface over several frames. In such cases, one has to practically guess what the designer had in mind and what this or that program depicted. On top of this, the presence of gaps between the surface patches and the almost complete absence of exact dimensions where necessary.
Something must be done with all this, otherwise in the next 10-15 years the world will be flooded with ships of a “new unusual design” from new and unusual designers.