A menu of Thesis Topics in Naval Architecture

From Mckwiki

Jump to: navigation, search

I often have the opportunity with students to brainstorm ideas for thesis topics. In order to speed up the process, I am now keeping a simple on-line list of ideas.



The foregoing thesis presents a foundation for understanding the problem of innovation in ship design, and the range of tools available for tackling that problem. In the course of developing this foundation I have identified many opportunities for further research. I envision a series of works, starting with the present foundation work, and then followed by (for example) a work specific to TRIZ, or a work specific to the teleology of ship systems, or other specialized study. My hope would be that the present project would establish a framework for publication in the field of Innovation in Naval Engineering, which framework could then be built upon by several subsequent works.

Thorough application of TRIZ to a ship design

The present dissertation is an omnibus treatment of the many available models and algorithms, each of which has some merit even if only in a limited domain. The omnibus work serves as a foundation for the study of innovation in Naval Engineering. Upon this broad foundation it is possible to erect several edifices, and one of them could be devoted to TRIZ. I envision, thus, a second work (either my own or another candidate's) which is specific to the application of TRIZ to naval architecture. This work would focus mainly on applying the TRIZ algorithm and thus 'translating' TRIZ into maritime paradigms. This would be an interesting work and useful to the industry, but as explained above I leave it to a 'phase II' of this line of study, in the belief that the foundation is the necessary starting point.

Find a real ship problem and apply TRIZ rigorously. This would first require the practitioner to be certified in TRIZ, to make sure that the test results are not tainted by the novitiate of the practitioner. Ideally, this would be repeated with several different naval architectural problems, so that we could construct a “map” of where TRIZ was most useful – mapping laterally according to which discipline it was best at, and mapping vertically to see whether it was best at whole-ship problems, or component-level problems.

Historical analyses

Using historical information of US Navy design projects, attempt to reconstruct a mapping between innovation success/failure against a variety of the factors identified in the dissertation, as listed in the sub-sections below.

This research would probably have to be conducted simply by interview with remaining members of the various project teams, and may be rendered useless by various personal factors. I suggest that some sort of industrial psychologist or interview expert may need to be included in the research team.

Tools used

What innovation tools did the historical team use? It is extremely unlikely that any tools were used by name, so the interviewer will need to extract what the creative process was, and then attempt to map that process into the generic tools types (taxonomy) presented above.

The goal would be to learn that tool X seems to be most useful for problems of type Y.

Note that here again both lateral (range of technical discipline) and vertical (scope from component to whole-ship) dimensions should be addressed.

World Sizes

What is the ‘world size’ of naval engineering? Does the world size correlate with innovation success, as has been shown for other creative endeavors? I would love to see the work of Uzzi (2005) repeated in the naval design realm. I find it easy to imagine that the Q for the naval design community is too high for good innovation.

Repeat this using different vessel projects which may be classed as innovations. Examples of innovation success include the USN T-AGOS 19 SWATH, or the recreational J-24. The SWATH-like “SLICE” craft may also be studied, and it is possible to debate whether this should be viewed as a success or a failure.

KAI Scores

Inventory the KAI scores of naval ship design personnel. Separately characterize those personnel on a scale of innovation, by self-identification and peer interview. Do the KAI scores correlate with the scale of innovation? Does the correlation change with ascension of the corporate ladder?

MBTI Types

Same as above but using Meyers-Briggs Types. Again, does MBTI correlate with the scale of innovation? Does the correlation change with ascension of the corporate ladder?

daVincian character in successful innovators

The third of three similar studies, except here we look to the da Vincian characteristics of the people involved.

VV&A of Innovation Success Predictors

Based on the foregoing studies it should be possible to create an “innovation success predictor” that would make some use of the foregoing metrics. This predictor should then be subjected to validation, verification, and accreditation, and it could then be used as aproject management tool

Teleological decomposition of an entire ship

It would be interesting to teleologically decompose an entire ship. What I propose is to develop a comprehensive functional, or ’teleological’, breakdown of a ship, by decomposing an existing US Navy asset. I expect that this process will uncover numerous cases where functions are competing, and where solutions are not synergistic.

Once a comprehensive list of all the ship’s functions has been generated, then we can attempt to combine those functions into logically similar groups, and reconstruct the ship from them. It is possible that this will identify some very novel architectures.

A second follow-on work that I can currently envision would be to attempt a comprehensive detailed teleological decomposition of a ship. By taking a work breakdown system such as the US Navy Standard Work Breakdown System (SWBS, 1985), one could develop a list of all of the functions that are present on a ship, and the functions that they support, and the functions that support them. For example, the man is on board to steer the ship, the HVAC system is on board to support the man, the chilled water plant is there to support the HVAC. A comprehensive map of all these functions – all these teleos – might provide very interesting opportunities to eliminate Muda or pursue multitasking.

Project structure guidelines for innovation in ship design

The foregoing thesis contains enough detail that it can be used to construct a project for maximum innovation potential. This will involve two components: The selection of the project personnel, and the design of the project approach. An example of this is given in the following, but the real intent here is to suggest developing project structure guidelines, of which the following may be predecessor.

Personnel selection

Personnel selection should take place in consideration of the project technical needs, as is well known. Further, in the foregoing we see that the project team should be sufficiently self-contained that all project interfaces will lie within the control of the team. For example, if the project team will need to modify the ship structure as part of the innovation, then the structural engineer should be part of the team.

On the other hand, the corollary to this is that the team should include only those disciplines inherent in the task. And thus if a structural interface is not foreseen for the innovation, then the ship structural engineer should not be included in the innovation task team.

Further, the team members should be surveyed for their MBTI types, so that team cohesion and functionality is maximized. This is basic business school practice and is not a subject addressed in this dissertation.

But in a similar manner, and as addressed herein, the team should also be assembled with due regard for the KAI scores of the participants. Recall that a high KAI-gap can lead to team friction, and also recall that KAI scores are not absolute, but relative. Thus it is possible that, given the other resource constraints, we end up with a project team lead by somebody who is fairly much an adaptor. But if teamed with people with even stronger adaptor tendencies, then even this team leader will tend to be innovative in comparison.

Project Structure

Similar to studying the people engaged in the work, we may suggest studying the structure of the work itself. Without developing the thesis at length, I submit that it would be clearly advantageous to structure the team’s investigations to explicitly parallel the innovation morphology. Thus, rather than conventional feasibility, concept, and detail phases, (or perhaps in addition to them) we would enjoin phases of problem definition, problem generalization, search for solutions, etc.

The product of the proposed study would be a set of Project Structure Guidelines that can be used as a cookbook for creating innovation project teams. These guidelines could be captured in the form of a management practice, or as a lecture series in a management course.

Software: Horizontal integration

Tools exist for several of the invention methodologies discussed herein, but these tools tend to be vertically integrated. Thus for example there is a TRIZ software. What I do not find, however, is a horizontal software such as, for example, a comprehensive tool for the “finding analogies” step that would use all of the methods mentioned: Image analogies, word analogies, patent databases, and word antonyms. Developing such a software would result in a "stereoscopic" tool that would present the inventor with many directions from which to tackle his problem, and would result in a much more robust candidate solution space.

Software: Analogy ranking

Building on the previous software solution, we may then imagine an algebra that would combine the software-found analogs. In this scheme we may imagine that an analog is found in, say, biology and the lexicon and the patent database. The fact that this same analogy is found in three very different types of search might suggest that it is a particularly fruitful one for investigation. Therefore it might be interesting to find a means for counting these "hits" and scoring each analogy with some sort of measure of merit, thus ranking them so that the software automatically presents a 'most-likely to be fruitful' subset. This scheme is similar in effect to the way that Google ranks websearch findings to present a 'most likely to be what you wanted' front page.

Paper: When to avoid fixation.

Many of the fundamental authors have touched upon the need to maintain an open mind as long as possible, to avoid fixating on a design solution. This is, for example, axiomatic in brainstorming wherein the participants are required to suspend judgment until a specific point in the process.

The research question for this paper would be: Is the point up to which we must suspend judgment a common point for all of the methods? I.e., do we suspend judgment up through step {x} of the morphology, regardless of which techniques we are using at each step? Or are there some techniques that are more 'judgment tolerant'?

This could have important application to which techniques one should employ, if one knows, for example, that she is working with a group that is prone to hasty judgment, or prone to avoid judgment.

Engineering by interaction rather than component

As mentioned in the section on the work of Dr. Tyson Browning, (Browing, 2010) I see a relationship between his focus upon interactions in design (which is also found in other models, including TRIZ) and the Cognitive Network Model and its attention to associative distance and interactions between concepts. It would be interesting to apply Browning's philosophy by actively studying, not concepts, but the associations between concepts, explicitly. As a mental placeholder for this future research, imagine studying adjectives instead of nouns. If instead of studying "apples" what if we studied "red"? If instead of studying bearings we studied “things that move?” What would be the engineering benefit of designing a research project based on interactions rather than components?

Which tool for which engineer?

Given that engineers have different cognition characteristics, as measured say by the Meyers-Briggs Type Indicator or the Kirton Adaption/Innovation Inventory, do we find a preference for one innovation toolset over another, that correlates with these personality characteristics?

I believe that the cognitive style of the person will affect his choice of preferred tool at each step of the process. I can imagine visual thinkers enjoying the visual analogies method, while lexical thinkers enjoy WordTree. This is a fruitful avenue of future research

What kind of engineer takes to each of the various tools or algorithms? Is one tool best suited to the least innovative, and another to the least adaptive? This is interesting to me personally. I find TRIZ a bit tedious, but then I am already a pretty good innovator. In consequence I am reluctant to take the time to slowly learn TRIZ, since I think it would only make me slightly better, not hugely better. In the sections above we discussed the personality characteristics of innovators. At this stage I merely state that it would be interesting to know what type of engineer “takes to” TRIZ the most easily – are these the linear algorithmic thinkers, or are they the wild hare creative ones? This might be a fruitful area for future research.

Right-sizing the information flow in naval design.

Based on von Hippel (1990) we may consider that specialists working on one part of a design problem don't actually need the full capability of the interfacing tasks, but might be able to get away with using relatively simple surrogate models for those tasks. Thus for example does the hull form designer really need to know the characteristics of the propeller, or is it sufficient to know merely the diameter? As one can see by this example, the interface between design tasks would be streamlined in accordance with this philosophy.

Starting with the USN Design Process Model it should be possible to determine information needed by each of the blocks of the process. This information need can then be compared to the information actually produced by the interfacing block. It is my expectation that we may find several cases where information is produced that is actually not used by the interfacing tasks – a sort of “data Muda.”

Sometimes the production of that information is an unavoidable by-product of the task itself, but I wonder to what extent there is data-Muda that can be eliminated, yielding a concomitant streamlining of the design process.

The use of KAI in the design of engineering teams

The Kirton Adaption-Innovation Inventory (KAI) returns a numerical indicator of where a person falls on a spectrum from Adaptor to Innovator (Kirton, 1994). It is conceptually similar (but narrower in focus) to the popular Myers Briggs Type Indicator (Myers, 1995) that is already extensively used in DOD.

As discussed in the body of this dissertation, it should be possible to use KAI scores as one of the parameters in selecting members of an engineering team, and assigning the tasks to those individuals. It would be fruitful to take this simple fact and develop a management-grade short course on how to implement this practice within the naval engineering enterprise.

MOQ in ship design

Robert Pirsig’s Metaphysics of Quality (Pirsig, 1991) has given rise to a substantial body of literature which is focused upon defining Quality, and defining procedures that lead to the production of Quality.

It would be interesting to attempt a journeyman-grade application of those teachings into the field of ship design. The goal would be to define something like “How do you know when you have done a good job, as a ship designer?”

Formalization of this process would be a great boon for practical ship design.

Metacognitive maturity

As I have repeatedly stated "not everybody can do it." It seems to me – and this would be a fruitful avenue for future research – that one has to have progressed pretty far on the knowledge axis in the metacognitive realm, and pretty far along the cognition axis into the creative realm, for invention to be easy. This would be a fruitful topic for statistical study.


Development of a Human Factors Measure of Performance.

I think this might be harvestable from the domain of Intelligent Ship Arrangements for part of its content. (ISA does some arrangement optimization, so it must have an objective function. Can that be used as part of an HFE MOP?)

Elements of Style

Discuss with Trevor Dobbins. What makes a an aesthetically pleasing ship? Varies by mission certainly. Note that it also evolves: 1940s "cool" isn't 21c "cool".

Effect of operator input on seakeeping

Ship seakeeping performance assessments are executed by predicting the ship motion, comparing the motion against some set of limits, and then reporting the percent of all possible speed / heading / direction / wave height combinations in which the limits are not exceeded.

But in the case of high speed craft the actions of the operator are very important. The helmsman carves the wave, the throttleman addresses jumps and crests, and these factors are not refelected in the assocaited ship motion model.

How big are these influences? Can we reflect them in a motion prediction? Can we reflect them in a towing tank test? Does the presence or absence of the operator input make a big difference in the result? (This last one would be a nice straightforward tank testing project, once without operator input, once with operator input, and then compare motion statistics, or peaks, or RAOs, or ... )

Gradients of response surfaces

Holtrop and Fung (and others) have response surface models of ship resistance in the form Resistance = f(Geometry). It might be interesting to create a nested series of reduced-order models of these surfaces, to support VSMs of ship resistance. At minimum we need what, speed and displacement? (McKesson's 2-parameter TF model) Then we add displacement-length ratio? Is the next one CP or B/T ratio? etc. Just knowing the ranking of these parameters would be interesting to the naval architect, as it tells him which ones are the most important to get right. This would actually be a pretty straightforward analysis to run in a tool like SAS.

Synthesis model

Build a ship synthesis model (for the mission of student's choice.) The framework can be used by following student's. The model itself may be useful in some teaching situations. And the graduate takes the model with him and builds upon it in his employment.

Add NPS POFacets RCS signature prediction capability into the UNO Synthesis model

Propeller Cups

I know of no systematic series or mathematical models to allow sizing of the cupped trailing edge on high speed boat propellers.

Numerical Surface Props

Develop a numerical model for the design of surface-piercing propellers. Validate by comparison with experimental data published by Rose & Kruppa

Non-Parallel Steps

In simple terms, what is the effect of having planing hull steps be not parallel? For example, sea plane floats (which admittedly have very special requirements) have different keel lines forward and aft of the step.

Non-Equal Deadrise Steps

What if the deadrise aft of the step is different from the deadrise forward?

Hybrid V/Inverse-V

What if a stepped hull has a V-hull forward, and an inverse-V / Sea Sled shape aft of the step? Or the opposite (Sea Sled forward, V aft)? Note that this results in a tri-hull shape rather like an unlimited hydroplane.

Shape of trailing edge

Most steps have trailing edges that are straight athwartships. Is this optimal?

Revise and document the Vorus Cat/Sea/Air program

Design from the bottom up

As naval architects we tend to draw the hull of the boat and then analyze its planing performance. What if we were to sketch instead the desired shape of the planing surface (drawings shapes that look like airplanes) and then developed the associated hull form from that? Better still, we should draw the desired pressure distribution....

Dynamic models of JetSkis / Personal Watercraft

Add a WIG capability (using Matveev's equations) to Cat/Sea/Air

Dihedral in Tunnel Boats

The aerodynamic platform in most tunnel-hulls is flat bottomed (athwartships.) What would be the effect of having dihedral (or anhedral) in this surface?

Study of the Transonic(tm) hull and its cousins

A systematic series of cheese-wedge hulls, in both towing tank experiment and CFD.

Performance of the Möbius Hull Form


Remotely Manned Merchant Vessels

NEW (June 2014): Rolls Royce seem to have taken this idea and run with it! See: this article.

The principle of Muda elimination gives rise to the idea of a remotely-operated merchant ship. We are invited to imagine a line-haul container ship which is manned via a satellite link from the flag state.

For purposes of visualization, we assume the vessel is piloted to the sea buoy via a pilot close by on the pilot boat. Once the pilot disengages control, the control transfers to the remote site.

In the event of a commlink failure at sea, the vessel is programmed to go into a station-keeping mode, perhaps steaming slowly in a circle, while chirping a distress code.

The Muda being eliminated is “world tours for sailors.” The advantages of eliminating the onboard humans include the following:

  • Space
  • Weight
  • Systems eliminated:
    • Berthing
    • Sanitary
    • Food service and stowage
    • Non-cargo HVAC
    • Lighting
    • Lifesaving
  • Systems redesigned:
    • Different motion limits, since no humans
    • Different structural loads, due to different motions
    • Different survival / risk strategy, since no loss-of-life danger

A very cursory preliminary investigation of this idea for a corvette suggests that a reduction of 25% in ship displacement is possible, with an attendant 20%+ reduction in fuel consumption. I have not run the numbers on a container ship, but I believe that the weight (and thus fuel) savings will be smaller, because the deadweight fraction of these ships is much higher. Nevertheless, the savings will be real, and may be married to a set of substantial savings in crew costs as well.

Developing this concept should be within the skills of a team of undergraduates.


Concept design of an ice-hulled ship. What are the structural properties of ice? How about ice mixed with a fiber? Given those properties, what kind of weight fractions result?

One the face of it, such a vessel looks very green from a shipbreaking point of view, but is it really? How much energy (carbon emissions) would it take to make and maintain the ice? Contrast that with the loss of value of recyclable steel, and where does it net out?

Note that "Project Habakkuk" was a WW-II project to use ice structure to build an aircraft carrier, to lilypad short-range aircraft across the Atlantic. So what would today's mission be? Arctic drilling maybe?

Concept Design of a Pedagogical Ship

It would be lovely to make naval architectural science accessible to the seat of the pants. I have drawn upon my small boat experience all of my career, in order to 'make sense' of the results from the mathematics. So how can I compress a few years of sea time into a classroom experience? How about a vessel that is specifically designed to exemplify lessons.

I imagine:

  • A flexible structure that deforms visibly in response to both static loading and wave action
  • A raisable KG so that we can experience the slow roll as we approach instability
  • A shiftable LCP so that we can demonstrate directional instability
  • Direct metering of thrust and power so that we can see a speed / drag / power relationship in real time.
  • ... more
Personal tools