Technical Report 1999

Executive Summary

This Technical Report describes the magnificent canoe, Krakatoa, Queen's University's 1999 entry to the Canadian National Concrete Canoe Competition. Krakatoa is 5.4 metres in length, 0.80 metres in width, has a centre depth of 0.35 metres and a projected weight of 50 kilograms when completely finished. The thickness of the hull varies with position, but remains within the range of 6 to 9 millimetres. Krakatoa has a unique paint scheme which changes colour depending on the point in time it is being observed. Currently, it appears to be a dull grey, with many black spots. By May 7^th, it could be a completely different colour. We are hoping it will be British Racing Green on that weekend.

The hull shape has very low form drag, as shown by scale model tests. It is an asymmetrical hull, having its widest point aft of the midsection. In cross section, the hull is a gently flared shallow arch, stiffened by the use of angles along the gunwales. There is moderate rocker in the hull to accelerate turns, and a shallow vee to allow straight tracking under power.

Introduction

Queen's University is a mid sized university located at the headwaters of the St. Lawrence River, in Kingston, Ontario. It is home to approximately 15,000 undergraduate students, of which nearly 2,000 are studying Applied Science. Kingston is a recently amalgamated city with a total population of 125,000 people. Some of the major employers in the city are Alcan, Dupont and of course the famous Kingston correctional facilities. For a relatively small city, it is home to a bevy of educational institutions, including the Royal Military College, Queen's and St Lawrence College. Kingston is one of the best sites in North America for sailing, and indeed hosted the Olympic sailing events in 1976.

The Queen's Concrete Canoe Team has been competing in the National Concrete Canoe Competition since 1997. We have never finished last, but alas, never finished first. Our best placing was in 1998, when we placed fourth overall and won the Oral Presentation competition. Krakatoa is a much improved vessel over the previous years' canoes and the team has high hopes of emerging victorious this year. The team has been training hard over the academic year to improve their strength and conditioning in preparation for the arduous canoeing events.

Krakatoa is a volcanic island located close to Indonesia. In the 1880's it exploded, in one of the largest and most violent events ever to occur. The sound it made could be heard as far away as Australia, and the ash it spewed into the atmosphere darkened the skies and cooled the entire planet for the following year. It is our belief that Krakatoa, the canoe, will have the explosive power and energy of Krakatoa the island, and enable Queen's to improve it's ranking in the 1999 Canadian National Concrete Canoe Competition.

Hull Design

Krakatoa is another canoe designed entirely by the students at Queen's University. It is neither a copy nor a modification of an existing design, but a completely new hull shape that has never before seen the light of day. The hull is asymmetrical, has an overall length of 5400 mm, moderate rocker, a maximum width of 800 mm, a depth of 350 mm over most of its length, and has a shallow vee bottom. The composite hull shell has a design thickness of 6 to 8 millimetres.

This hull design is the product of last year's somewhat disastrous results at the National Championships. While Conan, the 1998 design, had outstanding speed and manouverability when piloted by an experienced solo paddler, two people of less experience tended to cause the canoe to fill with water and turn turtle. Not an auspicious method of impressing your fellow competitors, although it did provide several welcome moments of comic relief, and kept the crash boat crew busy.

In light of this experience, it was decided that Krakatoa would have more initial stability and freeboard, while attempting to maintain the performance characteristics of Conan. The largest changes were therefore a wider beam and flatter bottom to increase the freeboard, with a gently flared hull above the soft chine. This increased beam would also tend to increase the initial stability, making it a more forgiving boat for the less-experienced paddlers on the team.

While not a direct 'hull design' decision, the sinking of Conan also led to further changes in the system of preparation for the 1999 competition. We have begun to train earlier in the year to allow all the potential paddlers to acquaint themselves with the handling of canoes.

The asymmetrical hull shape, when viewed from above, is an approximate airfoil. This was done to minimize the form drag while providing decent payload capacity. An asymmetrical hull allows the widest, and therefore highest displacement to depth ratio, section to be placed under the heavier paddler (usually the stern paddler, which will be strictly enforced in this case). This should keep the sheer line of the canoe level fore to aft, meaning that constant freeboard should be maintained.

When viewed from either end, the shallow vee of the hull can be seen. This vee begins about one third of the way aft of the bow, and helps to provide directional stability. By starting the vee further aft, good turning ability does not need to be sacrificed to maintain tracking. The moderate flare of the hull can be seen as well. This deflects the water that might otherwise splash into the boat, and keeps the paddlers dryer and happier.

Although Krakatoa has a flatter bottom than Conan, with a flare rather than a tumblehome cross-section, finite element analysis of the hull indicates that the strength of the hull should not be exceeded by any of the expected loadings. While tumblehome adds some longitudinal stiffness to the hull, it does complicate construction. Krakatoa's mold should be much easier to reuse if necessary, as well as simplify de-molding the canoe.

When viewed from the side, the rocker of the hull can be seen clearly. This allows for improved turning ability by reducing the polar moment of inertia that resists changes in direction, as well as reducing the wetted surface of the hull, which reduces drag.

Once again, the hull shape of the boat was developed using a three-dimensional computer model to allow interactive discussion of the shape of the hull. A group consensus was formed about what the hull should look like, while the computer model allowed instant comprehension of the effects of changes in dimension. There are many advantages of performing the design in this manner, but the two major ones, from our point of view, are the interface with DASCHUNDT 2.1 and the ability to perform a stress analysis on the hull.

A further benefit of this technique is that any size model can be produced to further examine the hull form. A one tenth scale model was created using basswood, and tested using the facilities at the Queen's Coastal Engineering Research Laboratory. A glass-fibre cantilever force measurement device was developed to allow quantitative measurements to be made on the model during towing tests. The techniques developed are further discussed in the Innovations section. A similar model of Conan was tested alongside the new hull shape, and thus the hulls could be compared.

All indications are that Krakatoa will be a more stable, dry, and enjoyable boat to paddle, especially on a weekend in May when the water temperature is only 3 degrees!

Concrete Mixture Design

The concrete mixture design was based on experience from past years’ mixture designs and some target properties were identified. Specifically, we were looking for a high strength, low density, and moderate workability. An adequate minimum compressive strength of 25 MPa was identified from past experience to produce an adequate bending strength when used in conjunction with sufficient reinforcement. A lower density, and hence lighter, concrete mixture was sought with the same strength. Moderate workability was defined as a mixture that could pass through the wire mesh but had a low enough slump to adhere to the sides of the form.

The density of the concrete was of particular consideration because concrete usually weakens with lower densities. This was addressed by the addition of polypropylene fibres in the concrete mixture. With very little mass, they added tensile strength, hence achieving high strength and low density.

To achieve the desired workability, more super-plasticizer was added to this year’s mixtures. This would enable the mixture to be less viscous without the addition of extra water, which would decrease its strength.

Certain components of each concrete mixture were the same, however different ratios between them were used for different batches. The binding materials used were Type 30 high early cement and silica fumes. Super-plasticizer was used as an admixture and ceramic microspheres made up the aggregate. Table XX shows each trial concrete mixture design proportions. The chosen batch was the 98-3A mixture.

Table 1. Trial Batches

In the interest of economy, standard grout test cubes of 50.8 mm length per edge were used in the concrete mixture test methods, rather than large cylinders. Five trial batches were produced and three cubes were used per trial. Average values of density and strength were measured while the workability was also noted.

Table 2 shows the compressive strengths of the individual mixtures and their unit weights.

Table 2. Trial Batches and Strengths

Reinforcement

A wide range of possible materials was considered for the reinforcement of the canoe this year. While hardware cloth, a trusty standby from years past, does an admirable job, 1998-99 was the season to branch out and broaden the reinforcement horizons at Queen's. The types of reinforcement considered were galvanized hardware cloth of 6.4 mm and 12.7 mm spacing, plastic geogrid, plastic eavestrough shield, and two varieties of expanded metal lathing. These materials were evaluated based on five categories - cost and availability, tensile strength, ductility, weight and composite action. The evaluation procedures are outlined here.

Cost and availability were evaluated very simply. A QUACKER (Queen's University Applied Science Crack Engineering Researcher) was dispatched to survey the stocks of various local hardware and building supply stores, and reported the cost per unit area of all the possible reinforcing material found. In Kingston, we discovered several reliable sources of all the material outlined above. While other materials, such as window screen and larger size meshes of metal and plastic were available, they were not considered due to possible infringement of the rules and aesthetic reasons (mainly the possibility of 'screen-door in your boat' jokes. We have sensitive Newfoundlanders on the team).

The tensile strength and ductility of the raw materials was evaluated using direct tension tests on the material. The material was trimmed to a reduced cross-section and placed in the clutches of a Unite-o-Matic testing machine. Load and deformation data were collected using a computerized data-acquisition system, which allowed analysis of the stress-strain behaviour of the materials

The results of the tensile testing suggested that the plastic materials would have too low a strength to be useful, although their low weight suggests benefits if strength is not a major concern. At this stage, it was noted that the metal lathing had a highly orthotropic behaviour which was investigated further in the composite action tests.

With some knowledge of the material properties of the various reinforcements, we began a series of tests to evaluate the performance of the materials in composite action with the chosen concrete mix. Standard forms measuring 400 mm by 120 mm with a depth of 6.4 mm were prepared, and all possible combinations of reinforcement were placed therein. After a week of moist curing, the specimens were tested in single point loading over a span of 350 mm. The data from this testing suggested that the best combination of strength, ductility and unit weight resulted from the use of metal lathing. This led to a second series of tests which evaluated the orthotropic behaviour of the material. These tests investigated the effect of orientation and placement of the layers of material, as well as the effects of thickness on the behaviour of the composite section. This showed that the section had much improved resistance to bending when the tensile layer of reinforcement mimicked the direction of maximum tensile stress. This knowledge allowed the intelligent planning of the reinforcement layout for construction.

The chosen reinforcement material is an expanded metal lath, 4 mm thick, with openings of 5x7 mm, formed of 0.2 mm thick steel sheet. When combined with concrete in the canoe wall, it can resist an equivalent elastic stress of 13.7 MPa before failure.

Construction (or, How I learned to Stop Worrying and Love the Wiener Dog)

The construction of Krakatoa continued the development of the DASCHUNT system, first introduced in 1998. This evolution has been termed DASCHUNT 2.1, due to the significant improvements over the original method. DASCHUNT stands for Differential Axial Sandwich Construction Technique, and the method involves the creation of a three-dimensional computer model of the canoe hull. (INSERT FIGURE HERE) This model is the same used for the finite element analysis of the hull. The computer model is divided into sections at intervals of 51.4 mm. These sections are plotted full scale and two copies of these lines are then traced onto 25.4 mm thick polystyrene sheets. These sheets are then cut out using a saw, notched and placed onto the supporting framework. The notch enables all the sections to line up perfectly, thus forming a stepwise continuous surface that is slightly larger than the planned surface of the hull. This stepped hull form is then sanded smooth using the interfaces between steps as a guide to the correct hull shape. Finish sanding is done to ensure a smooth surface, and as a last step, the formwork is covered in heatshrink plastic.

The superstructure of the formwork is thus plastic wrapped polystyrene, while the base is composed of wood. This wooden base supports to foam, providing a level, rigid datum, and also allows the tensioning of the reinforcement. The base has a key of 2x4 lumber running like a spine, while there are 2x6 cross-bars spaced at 600 mm intervals. These cross-bars provide a spreading effect for the 9.5 mm plywood sheer forms. The sheer forms control the gunwale height of the canoe, and are the anchor points for the screw-type tension adjusters used to tie down the reinforcement. The details of this construction system can be seen the following figure.

The use of a female mold was briefly considered, but ultimately rejected on grounds of cost. The major component of any mold being used by Queen's is expanded polystyrene, which costs on the order of $15.00 per sheet. A female mold would therefore cost approximately three times as much as a male one. A female mold would also introduce some problems with the placement of reinforcement in the mold. DASCHUNT 2.1 avoids these problems and allows very good control over the dimensions and surface quality of the final product.

The reinforcement is initially cut to fit the mold, then tensioned in place by using a system of fishing line and screw (turnbuckle) adjusters. This system uses a pre-fitting method; that is, the reinforcement is then removed from the form before casting begins.

In order to ensure an even smooth surface on the interior of the canoe, a thin layer of concrete is placed on the form, and then the first layer of pre-fitted reinforcing is placed into the concrete. Further concrete is placed to fill the gaps in the rebar. The second layer of mesh is then placed and tensioned into place, and the final surface layer of concrete is placed. Finally, the surface is smoothed into place, and allowed to set. When initial set has taken place, the surface is further smoothed, and then a layer of plastic sheet is squeegeed onto the surface of the concrete. This step produces a very smooth surface. After allowing the concrete to complete its setting overnight, the entire canoe and form are encased in a plastic moisture container, which is then maintained at 100% humidity using ordinary humidifiers. After 3 days, the canoe is removed from the form and replaced into the humidity chamber for a further week, then removed, finish sanded to provide a smooth surface and finally painted.

Project Management

It has been the team's experience that with growing knowledge of the techniques needed to produce a canoe; the organizational style of the group can change. In the first stages, no one knew anything - thus chaos seemed the most appropriate method of management. The second year, with a high turnover rate, one person knew more than everyone else - hence a dictatorship. This year, a larger group of people shared the knowledge gained under the dictatorship and was able to form opinions and make decisions. This situation cried out for democracy!

The team organized itself along the lines of a semi-mythic tropical island tribe called the Krakatoans. Our best research tells us that this tribe was composed of several bureaucratic leaders that made decisions and presented them to the tribe at large. The tribe would then split up into smaller sub-groups to work on the tasks resulting from the bureaucratic decisions and divide the work equitably amongst them, but, of course, they are always free to discuss what they think should be done to achieve the goals of the group.

This organizational structure provided the framework for the canoe team this year. The Chief Krakatoan was in charge of defining overall goals and negotiating with outside parties for supplies. The Medicine Man, being old and having much experience, led the research team searching for ways to improve the canoe. The Hut Builder provided planning and organization for constructing all the infrastructure of the project. The Bongo Master sent out messages to inform the entire tribe of the events that were taking place.

This system of project management was much more effective than either dictatorship or chaos, and helped to provide a better split of the labour among the group. There exists a foundation after this year for an even more advanced management culture to develop. Who knows, it could even become a republic!

Cost Assessment

Careful tracking of the time and money spent by the canoe team allows accurate assessment of the total cost of the project. As in years past, the overwhelming expense, as calculated by the standards imposed in the rules, is the labour. As this cost is borne entirely by the volunteer efforts of the team, the high apparent cost of the canoe is not so high after all.

The labour rates as specified in Section III are used to develop the costs for the individual hours worked. These costs are then summed and collated with respect to the task performed and its category as defined by the rules. The Raw Labour Rate multiplier as developed by Queen's is 2.875. This multiplier accounts for the overhead, direct and indirect costs associated with the project. Material costs are multiplied by a markup factor of 1.1.

As additional information, the total time spent working on this canoe project is 732 person hours.

A more comprehensive breakdown is presented in the Detailed Cost Analysis appendix.

Innovations

Krakatoa is the product of evolutions in the design and build process. To outline these evolutionary innovations, it is best to begin at the beginning.

After moderate success of scale modeling in 1998, a better arrangement was needed to produce quantitative results. To this end, a force measurement rig was designed and built by the team. This system allowed the sensitivity of the rig to be adjusted by varying the length of a glass-fibre cantilever arm. In this manner, increments of drag force as small as 0.25 N could be measured. This sensitivity allowed better use of the results and more confidence in them. A schematic diagram of this measurement device is presented here.

While previous reinforcement was adequate for the purpose, a better method of evaluating reinforcement performance was sought. While tensile testing is useful for comparing materials; the ultimate pudding proof as it were is in the performance of the reinforcing in composite with the concrete mix. For this reason, a very large number of composite sections were cast to determine the best overall combination of reinforcing materials. A three-point bending test was used to evaluate these composite sections, which showed a variability of over 10 times in bending resistance. We believe that this composite bending test is the only accurate way of assessing reinforcement performance.

The method used to design and build the hull has been more fully integrated in the 1998-1999 canoe campaign, and as such has been renamed DASCHUNDT 2.1. This stands for Differential Axial Sandwich Construction and Design Technique, Generation 2.1. For a further description of this system, please see the sections on Hull Design and Construction, as only the revised aspects will be presented here.

A fitted sheer plate was bent around the formwork of the canoe. This plate defined the gunwale position of the boat and allowed the reinforcing material to be tensioned around the formwork using turnbuckles developed by the canoe team. These turnbuckles and the use of heavy duty monofilament line allowed the steel mesh to be tensioned into position very well - eliminating unsightly bulges that had been present in earlier designs. This refinement of the construction technique allowed the sequential placing of the steel after the layers of concrete had been applied. This resulted in a very smooth, low void finish on the interior of the canoe, as compared with previous years. A further consequence of the sheer plate was that it allowed continuous finishing of the hull right up to the gunwale line and past it. This produced a very regular surface up to and including the gunwale line.

After casting of the bare hull, some portions of the concrete were removed and further structural elements were added. This significantly enhanced the rigidity of the hull and added much needed stability. This two stage construction process allowed the use of overhanging structural elements that could not be incorporated otherwise due to form work interference.

Curing of the hull was completed in a constant humidity environment. This was maintained through the use of vaporizers inside a polyethylene tent. For the first three days after casting, a framework of wood supported the tent. After these three days, the bare hull was removed from the form and placed, inverted on two sections of the formwork. This left an air chamber beneath the hull. By placing the vaporizer beneath the hull and running it constantly, a 100% relative humidity condition was maintained for the next 9 days. Several layers of burlap cloth were placed atop the hull to ensure that the moisture, when condensed out of the atmosphere onto the polyethylene sheet, would drip back onto the hull and maintain the outer surface at the same soaking wet humidity. Further refinements to this technique are planned which will allow recycling of the water inside the tent and not require refilling.

The finishing of the canoe will take place using the facilities of a local auto body repair business. This will allow the very rapid and precise painting of the canoe to a uniform thickness, and reduce the surface roughness of the hull considerably, when compared to previous years.

Finally, there has been an innovation in the preparation of the team members. In the past, we have tended to neglect the paddling aspect of the competition and concentrate on producing the canoe. This year our focus was considerably changed. Team members participated in strength and fitness training throughout the year in the hopes of improving their endurance and speed on the race course. On water training began almost as soon as the ice had melted on the Cataraqui River, and it is planned to continue this training until the time of the competition.