grand slam - uah · oklahoma state university 1 grand slam ... 1999 team, “the storm,” then set...

Oklahoma State University

2000 Concrete Canoe Team

Grand Slam

Oklahoma State University 1 Grand Slam

2.0 IntroductionOklahoma State University is located in

Stillwater, a northcentral Oklahomacommunity positioned approximately 60miles between Tulsa and Oklahoma City.The university was founded on December 25,1890, as Oklahoma A&M College, just 20months after the Land Run of 1889. Whenthe first students assembled for class, therewere no buildings, no books, and no curricula.Since these humble beginnings, OSU hasgrown to include an 840-acre Stillwatercampus and branch campuses in Tulsa,Oklahoma City, and Okmulgee, with a totalenrollment just exceeding 26,000 students.The 2000 Concrete Canoe team adheres to theOklahoma State tradition of continuing todevelop and grow.

In 1999, OSU swept each of the races andall four academic categories on the way towinning its third straight regional title. The1999 team, “The Storm,” then set its hopes ona top three finish at Nationals. They placed inthe top five in three of the four academiccategories and racked up three top five racefinishes en route to a 3rd place overall finish.

The 2000 OSU Concrete Canoe team,“Grand Slam,” again swept the Mid-

Continent Conference Competition, and hasits eyes focused on winning a National Title.

3.0 Hull Design3.1 Goals

The races developed by the NationalConcrete Canoe Committee require the hullbe designed to satisfy conflicting objectivessuch as straight line tracking and turningmaneuverability. Grand Slam is the thirdcanoe designed by the same design team.

In 1998, the team designed a verymaneuverable canoe to face the challenges ofthe slalom and 180° turn in the sprint.However, this design lacked the tracking andstraight-line speed to be competitive atNationals. Extensive model testing was donein 1999 to determine how tracking andstraight-line speed could be increased withoutlengthening the canoe. A canoe with identicaldimensions--19 feet long, 30 inches wide, and10 inches deep--improved nationally from 8thto 3rd overall in the races. Despite ourimprovements, the OSU faithful watched thelonger canoes catch and sometimes pass thefaster starting and turning 1999 OSU canoe.

The primary goal for Grand Slam wasmore speed. However, eliminating 1999 hull

1.0 Executive SummaryLighter, longer, faster, and stronger, Oklahoma State University

introduces concrete canoe – Grand Slam. It is constructed with 0.31inch thick lightweight concrete, yielding a compressive strength of880psi, and a unit weight of 53pcf. Grand Slam, reinforced with

10 layers of fiberglass scrim cloth, is the first ever inherentlybuoyant canoe to leave Stillwater, Oklahoma.

Grand Slam measures 20.5 feet long, 11 inches deep, 32 inches wide,and tips the scales at a trim 75 pounds. It is painted black and

traditional orange and white finishes highlighting six cross-sectionalribs. The ribs provide lateral stiffness to the elliptical cross-sections

of this strategically designed racing machine.


Model Cross-sectionalShape

LongitudinalShape Rocker

A Flat bottom Full and rounded6 inches, stern

and bow

B Flat bottom Full and rounded none

C elliptical Full and rounded none

D elliptical Sleek and straight3 inches, stern

only

Figure 1 – Attributes of Full Scale Canoes

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.0 1.5 2.0 2.5 3.0 3.5 4.0

Time (seconds)

Vel

oci

ty (

met

ers/

seco

nd

)

Model A (forward)

Model D (forward)

Model D (90% Loaded)

Model D (backward)

Figure 2 – Velocity vs. Time, Flume testing

design flaws and maintaining goodmaneuverability were also high priorities.Other considerations included coed loading,paddling efficiency, stability, and ease ofconstruction.3.2 Shape Optimization

The Hull Design Team utilized the powerof experience and an in-depth literaturereview to isolate several key variables: length,width, cross-section shape, longitudinalshape, and rocker. The key to obtaining asuccessful design is minimizing the TotalDrag Force (TDF).

TDF = C1(Vn/L) + C2(V

m*L)

C1 and C2 are canoe shape factors dependenton width, cross-sectional shape, longitudinalshape, and rocker. C1 has units of mass and isbased on wave drag. Wave drag represents theforce required for a displacement hull toseparate and return water around the hull. C2,based on skin drag, has units of mass per areaand is linearly tied to the wetted area of thehull. By reducing the wetted area, the skindrag can be significantly reduced.

At a constant velocity the TDF is equal tothe paddler input force. Likewise, if thepaddler input force and length are heldconstant, then the canoe with the lowestvalues for C1 and C2 will have the highestvelocities.

Based on this initial condition four modelswere constructed and tested to determine whatshapes would provide the lowest values for C1and C2. Each 1/10 scale model had the sameoverall dimensions: 24 inches long, 3 incheswide, and 1 inch deep. More than 75 flumetest runs were performed to look for trends tohelp in the selection of a final design. ModelA was identical to the 1998 canoe; Models B,C, and D represent step-by-step changesaimed at minimizing the shape factors. Abrief description of each canoe’s attributes iscontained in Figure 1.

Drag tests were run in a flume 20 feetlong, 2.5 feet wide, and 1.5 feet deep. Thedrag tests were performed in still water usinga “smart pulley” to measure displacement vstime.

A mass of known quantity was attached tothe scale models with fishing line andsuspended over the “smart pulley.” When themodel was released, the weight of the masswould pull the model through the water and

the “smart pulley” system recorded the time,in seconds, for each .015 meters of modeltravel. This typically amounted to more than200 data points for each test.

Displacement, velocity, and accelerationvs time graphs were generated for each flumetest run. The four models were pulledforward and backward to determine the effectof hull asymmetry. The models were alsotested with different weights (fully loaded and90% loaded) to determine the effect ofpaddler weight. Plots of velocity vs time froma few representative tests are shown in Figure2.

When tested fully loaded and pulledforward the hull with the highest finalvelocity was Model D with a 2%improvement over the hull with the lowestfinal velocity, Model A. This increase invelocity proved that the TDF could bereduced independent of the length. Byreplacing the flat bottom cross sections withelliptical cross sections, the surface-to-volumeratio decreased, reducing the wetted perimeter


and minimizing skin friction. Pulling the hullsbackward clearly illustrated the effect of hullasymmetry on wave drag. By shifting thewidest point of the hull 60% to the stern, theentry line became sharper and smoother, thusreducing the wave drag. When the widestpoint was placed only 40% from the bow(backward) it resulted in a 6% loss in finaltested velocity.

The drag testing also showed that a veryeffective way of increasing velocity is byreducing the total weight of the canoe andpaddlers. It was shown during testing that a10% reduction in total weight resulted in a3% increase in model velocity. While the HullDesign team realized that a 10% reduction inweight would be significant, it clearlyillustrated the importance of a lightweightcanoe and physically fit paddlers.3.3 Length Considerations

The design team again analyzed the TDFequation. However, the optimal lengthchanged with each set of paddlers. This isbecause each set of paddlers has a differentdisplacement requirement, which effects C1and C2, and a different input force, which isdirectly related to TDF.

Therefore, the design team looked at thelength of the canoes that accumulated themost points in National races. The average ofthe top five canoes, excluding our design at19 feet, was 21.25 feet – more than 2 feetlonger than our canoe. This information,coupled with the experience of our paddlingteams at Nationals, prompted the design teamto increase the design length of Grand Slam to20.5 feet. At 20.5 feet long the theoreticalmaximum speed of the canoe will becomparable to the design speed of the canoesthat caught and passed last year’s canoe.3.4 Bow Section Testing

The 1999 OSU Concrete Canoe Teamlearned a valuable lesson during last year’sregional and national races: “Don’t assumesomething will work. Test it.” The 1999design team assumed a sharp deep bowsection with a vertical nose would knifethrough the water. Unfortunately as shown inFigure 3 it acted more like a scoop.

To fix this design flaw, three full scalebow sections were designed and constructed.The sections were 3.5 feet long, 11 inchesdeep, and had an entry line angle of 7°.

The bow sections were tested in a constantvelocity flume. The flume was 6 feet wide, 8feet deep and 80 feet long. Approximately,6.2 million gallons of water were siphonedfrom a local lake to test the bow sections.Video and still photography were used toevaluate the bow sections.

The final bow section had a modestamount of rocker and a sharp slanted nose.Even submerged 9 inches deep, the final bowsection did not take on water.3.5 Prototype Evaluation

As a part of the mold constructiondiscussed in section 6.0, a fiberglass practicecanoe was constructed. This canoe gavepotential paddlers an opportunity to test thehull design.

As projected, the increase in length madethe canoe slightly more difficult to maneuver;however, when a stopwatch was used to gagethe hull’s performance versus the 1999design, most paddlers realized the canoe’spotential.

To date, 3 of the 5 1999 National winningtimes have been beaten in practice. Figure 4displays the final design and a few of its keyattributes.

4.0 Concrete Mixture Design4.1 Target Properties

While the Hull Design team performed itsanalysis to optimize the hull’s hydraulicperformance, the Structural Design team hadto respond to the demands of the fourpaddlers in the Coed Sprint. An ExcelSpreadsheet was programmed to use imported3-D points from AutoCAD to compute theincremental water displacements of GrandSlam. The canoe weight and four paddlerswere placed in the computer model and thespreadsheet computed 42 incrementalbuoyancy, shear, and moment forces–one foreach cross section.

Figure 3 – Design Flaw and Flume Testing


As shown in Figure 5 forces generatedoutside the paddler locations were small andwere not analyzed. However, all other crosssections were checked in shear and bending todetermine necessary concrete compressivestrength and tensile reinforcement strength.During the analysis two conservativeassumptions were made–concrete carries notension and reinforcement carries no

compression.The computed concrete compressive

strength required was 1.9 MPa (276 psi).This was multiplied by a dynamic load factorof 2.0 and a factor of safety of 1.5 to give afinal compressive strength required of 5.7MPa (827 psi). The total tensile forcerequired was 5.3 kN (1.19 kips).

The Excel spreadsheet also computed thetotal volume of concrete needed toconstruct a canoe 8 mm thick. Whenthe desired weight of the canoe, 36 kg,was divided by the volume of concrete,a target concrete density of 865 kg/m3

(54 pcf) was found.The challenge of the unpainted

section in this year’s competition alsoprompted the design team to look atabsorption and re-wetted compressivestrength as design criteria as well.4.2 Material Selection

The design team broke the materialselection into three categories: binders,aggregates and random fibers, andadmixtures. The search for materials anddonors began on the Internet. These

-1000

-800

-600

-400

-200

0

200

400

0 50 100 150 200

Location (inches)

Gen

eral

Sca

le

Moment(lbs*ft)

Shear(lbs)

Bouyancy(lbs)

Figure 5 – Loading on Canoe Hull

Figure 4 – Hull Design Features


searches proved fruitful and informative. Thedesign team initially settled on four bindingmaterials, three aggregate/random fibers, andtwo high range water reducers(superplasticizers).

The binders selected were portlandcement, silica fume, fly ash, and Laticrete

333. The portland cement used in this year’sproject came directly from a cement plantwithout being identified; however, a chemicalcomposition sheet was supplied. The designteam checked ASTM C 150 and historicalaverages of portland cement types anddetermined our cement to be a Type Iportland cement.

Silica fume and fly ash were selected fortheir pozzolonic properties. These pozzolonsreact with the products of hydration toproduce additional binders. They have alsobeen shown to reduce absorption andpermeability in hardened concrete.

Laticrete 333 is a superflexible thin setmortar latex additive that is 31% solids byweight. Latex enhances many of the desiredproperties of concrete canoe mixes includingworkability and durability, and reducesabsorption and unit weight. The design teamwas careful to select a latex that was suitablefor exterior use, since many latexes are watersoluble even after curing.

The aggregates/fibers selected includeEccospheres, Microlite-T, and Stealth

Fibers. The Eccospheres, also referred to asmicrospheres, are silica glass hollow spheresthat are 100% passing a #200 sieve. Themicrospheres were selected as the primaryaggregate used to reduce the concrete unitweight. With a true particle density of only0.227g/cc, they produce tiny, uniform,reinforced, air voids.

Microlite-T and Stealth fibers added tothe concrete mixtures gave them more body,making them more suitable for placement onthe form. Microlite-T is an expandedvolcanic mineral with a specific gravity of0.41. While 90% of this material still passes a#30 sieve, it is much coarser than themicrospheres. Stealth fibers are a poly-propylene material that was quicklyeliminated due to an inability to achieveuniform dispersion.

Reducing the water to cement ratio is aneffective way to increase compressivestrength, and decrease absorption and

shrinkage. This prompted the design team tolook at two superplasticizers: Rheobuild

2000B and Duracem 300M. Both hadadverse reactions in the latex modifiedconcretes; in fact, two mixtures with highdosages never cured. However, in standardmixtures w/c ratios as low as 0.30 were testedwith acceptable workability.4.3 Compression Testing

Forty-five different material combinationswere tested to determine which mixconstituents could be used to best meet thegoals of the concrete mixture design. The wetunit weights of these mixes varied from 795kg/m3 (49.6 pcf) to 1362 kg/m3 (85 pcf). Theamount of cement in each trial mixture designrepresented 75 to 100% of the bindingmaterial, with silica fume, fly ash, and latexreplacing the cement either alone or incombination.

Three to six 2-inch cubes, and onefiberglass reinforced plate were cast fromeach 1500 gram trial batch. The cubes wereused to test unit weight, absorption, air-drycompressive strength, and re-wettedcompressive strength. The plates wereconstructed mainly to make workabilityobservations; however, many of the plateswere tested in a pinned-pinned beam test.

Loading rates for the compression testswere done in accordance with ASTM C 39,section 7.5. A loading rate between 0.14 and0.34 MPa/s (20 and 50 psi/s) was maintaineduntil yielding. At yielding, no loadingadjustments were made.

As specimens yielded and failed, length ofyield plateau and subsequent residual strengthwere noted. All concrete mixtures containinglatex had significantly longer yield plateausand higher residual strengths than thosemixtures without latex. Microlite-T andpolypropylene fibers also added some residualstrength to non-latex concrete mixtures.4.4 Absorption Testing

Absorption testing was done using amodified ASTM C 642. For this test a samplewas weighed, oven-dried, weighed again, andplaced in a water bath for 48 hours. Weightswere taken after 5, 10, 20, 40, 90, 180, and360 minutes and a final weight was takenafter 48 hours.

The volume of each sample was foundusing a displaced water test, and unit weightswere computed for each of the 8 readings.


Figure 6 shows the increase in unit weight vstime for a latex vs non-latex concrete mixturedesign. Both concrete mixtures had wet unitweights of about 850 kg/m3 (53pcf) at thetime of mixing.

The 7-day air-dried unit weights of thenon-latex and latex mixtures were 41.6 pcfand 48.1 pcf. The total change in unit weightfrom the oven-dry state to the saturated statefor the latex and non-latex mixtures was14.0% and 33.2%, respectively. However, theunit weight of the latex modified concreteincreased only 4.2% over the air dried unitweight, while the non-latex modified concreteincreased an additional 21.2% over its air-dried state.

While the unit weight of both samples

climbed to within 2% of its original wet unitweight, the latex modified concrete tooknearly 48 hours to reach this mark. The non-latex modified reached 98% of original wetunit weight in only 3 hours.

To test the effect of re-wetting oncompressive strength, air driedsamples were immersed for 24 hours.When re-wetted samples were testedin compression, there was a sharpreduction, 25-70%, in strength for theconcrete mixtures with absorptionpercentages greater than 15%.4.5 Final Concrete MixtureSelection

The final concrete mixture designneeded to have a compressivestrength greater then 5.7 MPa (827psi), a wet unit weight less then 865kg/m3 (54 pcf), and an absorptionpercentage less than 15%.

To achieve these target propertiesportland cement provided the compressivestrength; latex decreased the rate ofabsorption, percent absorption, and unitweight. Eccospheres produced microscopic,uniform, reinforced air voids, reducing theunit weight. The addition of Microlite-T

gave the mixture body and further reduced theunit weight. Water, the universal catalyst,puts all reactions in motion. Figure 7 displaystrial mixtures and the proportions for the finalPatch and Grand Slam concrete mixtures.

30

35

40

45

50

55

0 60 120 180 240 300 360

Time (minutes)

Un

it W

eig

ht

(pcf

)

Non-latex Concrete

Latex Concrete

Figure 6 – Unit Weight vs. Time for Absorption

Mix Name 1999 Mix Non-latex Mix Patching Mix Grand Slam MixBinding Materials kg/m3 pcy %BM kg/m3 pcy %BM kg/m3 pcy %BM kg/m3 pcy %BMPortland Cement 427 720 75.2% 355 599 100% 375 632 81.2% 374 632 83.8%Silica Fume 112 190 19.8% 0 0 0.0% 0 0 0.0% 0 0 0.0%Fly Ash 28 47 5.0% 0 0 0.0% 0 0 0.0% 0 0 0.0%Laticrete 333 0 0 0.0% 0 0 0.0% 87 147 18.8% 73 122 16.2%Aggregates kg/m3 lbs./yd3 kg/m3 lbs./yd3 kg/m3 lbs./yd3 kg/m3 lbs./yd3

Eccospheres 82 139 86 145 75 126 66 111Microlite 45 76 20 34 0 0 42 71

Stealth Fibers 0 0 0 0 0 0 0 0AdmixturesSuperplasticizer 11 19 10 17 0 0 0 0Water 345 581 375 633 344 580 292 493Properties7 Day Strength 11.8MPa 1712psi 7.41MPa 1075psi 7.41MPa 1075psi 6.07MPa 880psiWet Unit Weight 1052kg/m3 65.7pcf 847kg/m3 52.9pcf 880kg/m3 55.0pcf 847kg/m3 52.9pcf

Figure 7 – Trail Mixtures and Final Concrete Mixture


5.0 Reinforcement Design5.1 Target Properties

The reinforcement design is perhaps themost difficult aspect of concrete canoe design.The Structural Design team was required toaddress the problems of point loads; hullrigidity; and composite durability, density,and constructability.

A concrete canoe hull is very similar to asimply supported concrete beam with auniformly distributed load. A typical concretebeam must be designed to carry maximummoment and shear forces along its length. Italso must be able to carry bearing forcesassociated with point loads at the supports.

An acceptable reinforced concrete canoedesign must be able to carry shear andmoment forces along the length of the hull inaddition to redistributing bearing forcesinduced at paddler locations. The analysisdiscussed in section 4.1 generated a requiredtensile force along the gunwale of 5.3 kN(1.19 kips) to carry moment forces. Aftermultiplying by a dynamic load factor of 2.0and a safety factor of 1.5, the design tensileforce required was 15.9 kN (3.57 kips).

To carry the paddler point loads thecomposite section was designed to bend, butnot break. This is similar to holding a bedsheet on all four sides and placing a weight inits center. The bed sheet can carry nobending forces, but does resist the load bydeflecting and placing the entire sheet intension. Allowing the concrete compositesection to deflect with fixed ends, rather thanrequiring the composite section to carryprimarily bending forces, reduces the highcompressive stress in the concrete and createsa higher tensile load in the reinforcement.The design team believed it was much easierto design a composite to take small bendingforces and high tensile forces than develop alightweight concrete with a compressivestrength greater than 10,000 psi.5.2 Material Selection

The design team found numerouspromising reinforcements including steelhardware cloth, carbon fiber meshes, Kevlar

meshes, and fiberglass meshes. Randomcarbon fibers and polypropylene fibers werealso considered in this phase, but wereeliminated due to their high degree of

unpredictability when used as a primarystructural reinforcement.

All steel meshes were eliminated afterreviewing the research done on steel meshcomposites over the past three years at OSU.Research had shown that placing a flexibleconcrete mixture over a rigid steel mesh oftencaused spawling of concrete outside thecomposite and delamination between layers ofsteel. Lightweight concrete mixtures were notstrong enough to safely pass hull stresses tothe steel.

While the manufacture’s data on Kevlar

were impressive, no distributor found waswilling to donate enough material to test. Thedesign team was left with two fiberglassmeshes and one carbon fiber mesh.5.3 Raw Material Testing

The first step in assessing the finalreinforcement was a standard tension test todetermine raw material strength. In astandard grab test, textiles typically fail at theconnection with the loading apparatus. Thedesign team elected to develop a loadingapparatus that would reduce this tendency.

To reduce connection stresses reinforce-ment ends were rolled up on 3/4-inchdiameter rollers and pinned in place. The firstand last 11-inches of each 24-inch test stripwere taped leaving a 2-inch test band.Failures typically occurred in the 2-inch testband. Figure 8 displays each of the materialstested, their specifications, and testedstrengths.

The design team was surprised to find thatthe carbon fiber mesh did not performsignificantly better than the fiberglass mesh.After contacting the carbon fiber meshdistributor, the design team learned thatcarbon fiber meshes are typically only usedafter pre-impregnation. The pre-impregnationbonds tiny fibers together and allows them toact as a group. Hopeful that the concretemixture could help bond the fibers together,the design team used carbon fiber mesh in testplates despite its average performance in theraw material tests.5.4 Pinned-Pinned Plate Testing

By using the pinned-pinned plate testing,plates were subjected to both bending andaxial forces. Depending composite stiffness,the plate beneath the paddlers is subject tovarying degrees of both bending and axialloads.


Figure 9 shows the load vs. deflectiongraph for plates constructed with each type ofreinforcement and the Grand Slam concretemixture. The graph does not show the postfailure loading for Plates A and B.

Plate A, constructed with 4 layers ofuniformly spaced 6 x 12 fiberglass mesh,showed modest deflection vs load. Afterloading the plate with 38.8 pounds, concreteon the compression side slowly crushed untilthe plate deflected 1.8 inches. At this pointthe plate could carry no bending stresses andwas considered failed, but continued to takeload until the test was stopped at 100 poundsof load with a final deflection of 2.0 inches.

Plate B contained 8 layers of uniformlyspaced 20 x 10 fiberglass mesh. This platehad the highest deflection vs. load of theplates shown. Similar to Plate A, Plate Bwent through a period of high deflection at aconstant load of 36.6 pounds. However, PlateB’s deflection was due to a failure of 3 of the8 layers of fiberglass mesh. Again, once theplate reached a deflection of about 1.8 inchesit stopped deflecting. The test was ended at aload of 100 pounds and a deflection of 2.2inches.

Plate C, constructed with 1 layer ofcarbon fiber on the compression face and oneon the tension face, was relatively rigid andfailed abruptly. Unlike the previous twoplates, all of the reinforcement broke at a loadof 41.0 pounds.

5.5 Construction ConsiderationsConstructability of a reinforced hull with

predictable thickness was the primary focusof one series of plates. Several plates wereconstructed by placing alternating layers ofconcrete and reinforcement until a desiredhull thickness of 8mm (0.31 inches) wasreached.

The design team found that to achieve areasonably repeatable thickness with theGrand Slam concrete mixture, relatively thinlayers of concrete should be placed over themesh. Using 5 layers of 12 x 6 fiberglass, 10layers of 20 x 10 fiberglass, or 7 layers ofcarbon fiber mesh created a composite ofuniform thickness.5.6 Additional Structural Elements

The stiffness of a section is directlyproportional to the product of the modulus ofelasticity, E, and the moment of inertia, I. Allof the thin reinforced composites tested had alow EI product and therefore low stiffness.

In order for the canoe hull to perform as itwas designed–hydraulically and structurally–it must maintain its shape. To do this, hullstiffness needed to be significantly increased.

There are two ways to increase stiffness:increase E or increase I. Since E was alreadyset by the concrete mixture design, the designteam looked at increasing I. I for arectangular cross section is proportional to thedepth cubed. By using six 2-inch deep ribs

0

5

10

15

20

25

30

35

40

45

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Deflection (inches)

Lo

ad (

po

un

ds)

12 x 6 Fiberglass

20 x 10 Fiberglass

8 x 8 Carbon Fiber

Figure 9 – Load vs. Deflection Graph

12 x 6 Fiberglass

12 x 6 Breaking Strength

sample Direction Tested Data Sheet

Warp (12) 160 lbs. 140 lbs.

Fill (6) 220 lbs. 160 lbs.

Thickness .0096 in .011 in

20 x 10 Fiberglass

20 x 10 Breaking Strength


Warp (20) 85 lbs. 95 lbs.

Fill (10) 95 lbs. 90 lbs.

Thickness .0057 in .0052 in

8 x 8 Carbon Fiber

8X8 Breaking Strength


Warp (8) 240 lbs. N/A

Fill (8) 200 lbs. N/A

Thickness .0126 in N/A

Figure 8 – Samples and Raw Material Data


over the widest 10 feet of the canoe, thestiffness was increased more than 25 times.5.6 Final Reinforcement Design

After reviewing raw material data,pinned-pinned plate testing, and consideringpredictable constructability, the design teamelected to use 10 layers of 20 x 10 fiberglassreinforcement.

One inch of 20 x 10 fiberglassreinforcement can resist a load of 95 pounds.The reinforcement design provides 10 layersof fiberglass along both gunwales. The top 1inch of the gunwales is capable of resisting atensile force of 8.45 kN (1.9 kips), which ismore than the 5.3 kN (1.19 kips) required.

Plate testing showed that the compositewas capable of distributing point loadsbeneath paddlers. Finally, the constructabilityof the composite will simplify constructionand finishing.

6.0 Construction6.1 Mold Construction

Construction of the male mold began withfull-scale templates of hull cross sections.These cross sections were cut from16-gagesteel sheets with a computer-driven plasmacutter at 15.2 cm (6 inches) intervals along the6.2 m (20.5 feet) hull length.

A 6-inch block of polystyrene then was“sandwiched” between two steel templatesand a special hotwire was used to cut thepolystyrene along the templates. Individualpolystyrene cross sections were gluedtogether in seven larger segments. Combiningindividual cross sections into seven largersections allowed for easier removal of themold once the canoe had cured.

These segments were then placed on aspecially designed table that would supportthe segments during casting/curing and allowremoval of the mold. Included in the tableand polystyrene mold was a dual keyway toensure proper alignment of each moldsegment on the table.

The Construction team was tasked withdesigning a mold durable enough for castingnumerous canoes. To provide the requireddurability, the Construction team placed afiberglass shell over the polystyrene mold.

Three layers of tightly woven fiberglasscloth were applied to the mold and coated

with resin. In addition to increasing molddurability, the fiberglass layer also increasedmold rigidity, which facilitated the placementof reinforcement and concrete.6.2 Prototype Construction

The Paddling team’s request for aprototype/practice canoe was met by buildinga fiberglass canoe from the new mold. Thefirst step for constructing the practice canoewas applying a fiberglass release agent to themold. Next, fiberglass mesh was cut to theapproximate shape of the hull and placed onthe mold. Fiberglass resin was applied to themesh with plastic putty knives and repeatedfor five layers of fiberglass.

Once cured, the practice canoe wasremoved from the mold. To provideadditional strength and rigidity to the practicecanoe, 1-inch strap steel was attached alongthe outside of the gunwales and ¼-inch all-thread was used as thwarts. A PVC pipe wassplit and placed over the gunwales to preventpaddlers from getting scratched on the steelgunwales.6.3 Ribs and Edge-Forming System

After the mold was cut into its sevensections, six 2-inch deep ribs were placedbetween the mold sections. The ribs wereconstructed from 1-inch thick plywood.Computer plots were made of each rib andthen glued to the plywood. They were cut outwith a jigsaw, sanded, and covered withpacking tape to prevent concrete frombonding to them.

The edge-forming system consisted of a1-inch by 3/8-inch high density plastic stripattached at the gunwale of the male mold withwood screws. This strip removed any minorlongitudinal waves and created a gunwalewith uniform thickness.6.4 Concrete Canoe Construction

Prior to casting the first layer of scrimcloth, reinforcement was cut to fit the moldand then used as a pattern to cut the additionalnine layers. Casting of the canoe began byfilling the ribs with concrete and reinforce-ment, placing one layer of scrim on the moldand, applying a thin layer of concrete. Thelayer of concrete was placed by hand andworked in a manner so it would penetrate andslightly cover the scrim cloth. The next layerof fiberglass reinforcement was positioned onthe mold and another layer of concrete was


applied. This process was repeated for ninelayers of fiberglass scrim cloth.

The last layer of scrim cloth was placedand a thin sanding layer of concrete wasadded. This produced a uniform finishwithout surface defects and minimized finalsanding. No special curing tents were requiredsince the latex in the mix formed a film on theoutside of the concrete and locked in thewater. The canoe was allowed to cure twoweeks before patching and sanding wereperformed on the exterior of the canoe.

Once the exterior was sanded, the canoewas removed from the mold by removingeach of the rib sections and removing the 7major mold sections.

The canoe’s interior received severaladditional hours of sanding until the desiredfinish was achieved. At this point the entirecanoe was sealed, primed, and painted withautomotive grade materials. The final coat ofpaint was buffed and polished to a high glossshine. In the final step, decals and graphicswere added to depict the school name andtheme.

7.0 Project Management7.1 Organizational Approach

Shortly after last year’s nationalcompetition, a Project Manager was chosenfor the 1999-00 concrete canoe competition.The project management began in Augustwith the development of a basic WorkBreakdown Structure (WBS).

The WBS contains all basic tasksassociated with the concrete canoe design,construction, and documentation. AnOrganizational Breakdown Structure (OBS)was adopted, with the idea that each task inthe WBS must be assigned to a committee orperson.

The OBS was set up with a ProjectManager, Lead Engineer, and six majorcommittees. Each major committee hadnumerous task-oriented sub committees. TheProject Manager scheduled, assigned, andmonitored tasks, while the Lead Engineer wasresponsible for coordinating and reviewingthe total research and design effort.

The Fundraising committee created apromotional brochure to aid in securingdonations and later tracked project costs. The

Hull Design committee tested new hulldesigns to obtain the best hull for bothstraight-line tracking and maneuverability.The Structural Design committee developed aconcrete mix and reinforcement scheme toprovide a strong, durable, final product. TheConstruction committee was responsible forprocurement of materials, constructiontechniques, and display. The Academiccommittee provided the display content, andwrote the design paper and oral presentation.Finally, the Paddle team was responsible forquick starts, fast cruising speed, andoutstanding maneuvering. While eachcommittee had a different leader, many teammembers were active on several committees.7.2 Project Implementation

Once the responsibilities were assigned,the work had to begin. To attract newstudents and inform incoming students, a slidepresentation about the 1999 OSU ConcreteCanoe Team was given at the first ASCEchapter meeting of the fall semester. Thefollowing weekend an orientation meetingwas held for all people interested in being partof the team. The meeting included concretecanoe paddling for new members, a trip to theconcrete canoe lab, a brief description of the2000 OSU Concrete Canoe Team goals, and apicnic provided by the faculty.

Over the next two weeks, individuals wereassigned to teams based on their interest,knowledge, and experience. Team memberswith the most expertise were appointedcommittee leaders and were asked to educatenew members. Teaching was the mostimportant responsibility of committee leaders.By passing on the knowledge of pastsuccesses and failures, team leadersguaranteed success for future canoe projects.7.3 Project Schedule

A detailed critical path diagram wasdeveloped in August to show theinterrelationships between various tasks eachcommittee was assigned to complete. Thisdiagram was simplified into a bar chart withspecific completion dates for the major tasks.

8.0 Cost AssessmentAll materials used were cataloged in a

project notebook kept at the concrete canoelab. A simple time sheet that containedspaces for name, hours spent, and specificactivities performed was also kept in the


project notebook. These records wereperiodically compiled in a spreadsheet andthen removed from the project notebook.

Using the labor and material rates outlinedin the 2000 Rules and Regulations, thecompiled records were converted to labor andmaterial costs detailed in the appendix. Laborcosts for research/development and concretecanoe construction came to $41,979.73.Material costs for research/development andconcrete canoe construction added to$3,504.02.

The grand total for development andconstruction of the Grand Slam was$45,384.75.

9.0 Innovative FeaturesThe 2000 OSU Concrete Canoe project

features innovations in research approach,flume testing, reinforcement design,construction techniques, and competitionpreparation.

One Thing at a Time – Throughout thecourse of the Hull, Concrete Mix, andReinforcement Designs, the design teamsused a thorough and methodical process oftesting. By changing only one variable at atime, design teams could pinpoint the effectof each variable without the influence ofanother, ensuring that trends were identified.

“Smart Pulley” Drag Testing – Onedrawback with constant velocity flume testingis that all information about drag forcesduring acceleration is lost. Using the “smartpulley” system allowed us to look at totaldisplacement, instantaneous velocity, andinstantaneous acceleration.

Perfect Practice Makes Perfect – For thepast several years the first time the paddlersexperienced how the competition canoehandled was during competition at regionals.This year’s team has been practicing in aperfectly shaped canoe since February. Thisgave the students an opportunity to becomecomfortable with the features of the newdesign over the course of 3 months, ratherthen 3 minutes.

Accurate Construction – The success ofa design often depends on its implementation.To ensure that the mold was built to exactspecifications, a dual keyway was cut intoevery polystyrene cross section during itsproduction. The reinforcement scheme was

chosen with construction in mind and aprototype was built to verify and refine allconstruction steps.

Fiberglass Scrim Reinforcement – Adurable, ductile, strong, and accurate concretecomposite was created. This year’scomposite is designed not to crack with theaddition of latex modified concrete. The unitweight decreased by 10% and allowed OSUto construct its first ever inherently buoyantcanoe. Additionally, the lay-up procedureresulted in a uniform thickness and reducedsanding/patching time by 200%.

Strength through Shape, Not WeightAdding material and weight is an easy way toincrease strength and stiffness. However,adding a little material and weight in the rightplace is far superior. Two-inch deep cast-in-place ribs were implemented to providestiffness through shape. The stiffness of thecentral 10 feet of the canoe was increasedmore than 25 times by using six ribs in thatsection.

10.0 SummaryThe primary goal of the 2000 Oklahoma

State University Concrete Canoe team was todesign and construct the best canoe to everleave Stillwater, Oklahoma. By building thelongest, lightest, and fastest canoe in teamhistory, the team has met this goal.

In 1997 OSU earned a trip to Nationalswith a canoe that weighted 145 pounds. Thecanoe made cracking noises each time it washandled, and posted a 1:24 in the men’ssprint-5 seconds slower than the bestwomen’s time.

Grand Slam weighs in at only 75 poundsand has posted times faster than 3 of the 51999 National winning times. Diligentresearch created advancements in concretemix, reinforcement, and hull designs. Asuccessful top down management structureensured that the project was completed ontime and under budget. Additional hoursspent designing and refining constructiontechniques, resulted in a perfect mold forGrand Slam.

The 2000 OSU Concrete Canoe Team hasloaded the bases and will be looking for aGrand Slam when it steps to the plate inGolden, Colorado.

grand slam - uah · oklahoma state university 1 grand slam ... 1999 team, “the storm,” then set...

Documents